diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,86221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 12313,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.121497604158207e-05,
+      "grad_norm": 6.562383818863084,
+      "learning_rate": 1.3513513513513516e-08,
+      "loss": 1.0675,
+      "step": 1
+    },
+    {
+      "epoch": 0.00016242995208316414,
+      "grad_norm": 8.277066733441483,
+      "learning_rate": 2.702702702702703e-08,
+      "loss": 1.3477,
+      "step": 2
+    },
+    {
+      "epoch": 0.0002436449281247462,
+      "grad_norm": 5.566945299122935,
+      "learning_rate": 4.0540540540540545e-08,
+      "loss": 0.982,
+      "step": 3
+    },
+    {
+      "epoch": 0.0003248599041663283,
+      "grad_norm": 20.123631677375503,
+      "learning_rate": 5.405405405405406e-08,
+      "loss": 1.1009,
+      "step": 4
+    },
+    {
+      "epoch": 0.00040607488020791033,
+      "grad_norm": 7.516929368064332,
+      "learning_rate": 6.756756756756757e-08,
+      "loss": 1.1767,
+      "step": 5
+    },
+    {
+      "epoch": 0.0004872898562494924,
+      "grad_norm": 6.534930634637095,
+      "learning_rate": 8.108108108108109e-08,
+      "loss": 0.9279,
+      "step": 6
+    },
+    {
+      "epoch": 0.0005685048322910744,
+      "grad_norm": 8.547137659048731,
+      "learning_rate": 9.459459459459461e-08,
+      "loss": 1.1229,
+      "step": 7
+    },
+    {
+      "epoch": 0.0006497198083326566,
+      "grad_norm": 8.40063081083683,
+      "learning_rate": 1.0810810810810812e-07,
+      "loss": 0.9784,
+      "step": 8
+    },
+    {
+      "epoch": 0.0007309347843742387,
+      "grad_norm": 7.113188011014463,
+      "learning_rate": 1.2162162162162163e-07,
+      "loss": 1.007,
+      "step": 9
+    },
+    {
+      "epoch": 0.0008121497604158207,
+      "grad_norm": 9.511299992179831,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 1.0264,
+      "step": 10
+    },
+    {
+      "epoch": 0.0008933647364574028,
+      "grad_norm": 6.234366333859408,
+      "learning_rate": 1.4864864864864866e-07,
+      "loss": 1.1152,
+      "step": 11
+    },
+    {
+      "epoch": 0.0009745797124989848,
+      "grad_norm": 8.300258843057527,
+      "learning_rate": 1.6216216216216218e-07,
+      "loss": 0.9787,
+      "step": 12
+    },
+    {
+      "epoch": 0.0010557946885405669,
+      "grad_norm": 6.394938331200344,
+      "learning_rate": 1.756756756756757e-07,
+      "loss": 1.0764,
+      "step": 13
+    },
+    {
+      "epoch": 0.0011370096645821489,
+      "grad_norm": 5.924100005331387,
+      "learning_rate": 1.8918918918918921e-07,
+      "loss": 1.0187,
+      "step": 14
+    },
+    {
+      "epoch": 0.001218224640623731,
+      "grad_norm": 7.076697509559219,
+      "learning_rate": 2.0270270270270273e-07,
+      "loss": 1.3016,
+      "step": 15
+    },
+    {
+      "epoch": 0.001299439616665313,
+      "grad_norm": 7.949795337583316,
+      "learning_rate": 2.1621621621621625e-07,
+      "loss": 1.1151,
+      "step": 16
+    },
+    {
+      "epoch": 0.001380654592706895,
+      "grad_norm": 7.324802587536972,
+      "learning_rate": 2.2972972972972977e-07,
+      "loss": 1.0339,
+      "step": 17
+    },
+    {
+      "epoch": 0.0014618695687484773,
+      "grad_norm": 7.644757636442122,
+      "learning_rate": 2.4324324324324326e-07,
+      "loss": 1.0653,
+      "step": 18
+    },
+    {
+      "epoch": 0.0015430845447900593,
+      "grad_norm": 19.41567123436543,
+      "learning_rate": 2.567567567567568e-07,
+      "loss": 1.2616,
+      "step": 19
+    },
+    {
+      "epoch": 0.0016242995208316413,
+      "grad_norm": 7.096406348798535,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 1.0637,
+      "step": 20
+    },
+    {
+      "epoch": 0.0017055144968732233,
+      "grad_norm": 9.633919383798274,
+      "learning_rate": 2.837837837837838e-07,
+      "loss": 1.185,
+      "step": 21
+    },
+    {
+      "epoch": 0.0017867294729148055,
+      "grad_norm": 10.180555284457757,
+      "learning_rate": 2.972972972972973e-07,
+      "loss": 1.0474,
+      "step": 22
+    },
+    {
+      "epoch": 0.0018679444489563875,
+      "grad_norm": 8.634935943189125,
+      "learning_rate": 3.1081081081081084e-07,
+      "loss": 0.9087,
+      "step": 23
+    },
+    {
+      "epoch": 0.0019491594249979695,
+      "grad_norm": 7.712256121372059,
+      "learning_rate": 3.2432432432432436e-07,
+      "loss": 1.2098,
+      "step": 24
+    },
+    {
+      "epoch": 0.0020303744010395515,
+      "grad_norm": 7.212382477186713,
+      "learning_rate": 3.378378378378379e-07,
+      "loss": 0.9682,
+      "step": 25
+    },
+    {
+      "epoch": 0.0021115893770811338,
+      "grad_norm": 8.965382654928913,
+      "learning_rate": 3.513513513513514e-07,
+      "loss": 0.9742,
+      "step": 26
+    },
+    {
+      "epoch": 0.002192804353122716,
+      "grad_norm": 9.333155078594814,
+      "learning_rate": 3.648648648648649e-07,
+      "loss": 0.9463,
+      "step": 27
+    },
+    {
+      "epoch": 0.0022740193291642978,
+      "grad_norm": 5.770329415577018,
+      "learning_rate": 3.7837837837837843e-07,
+      "loss": 0.9216,
+      "step": 28
+    },
+    {
+      "epoch": 0.00235523430520588,
+      "grad_norm": 7.19909286290371,
+      "learning_rate": 3.9189189189189195e-07,
+      "loss": 0.9299,
+      "step": 29
+    },
+    {
+      "epoch": 0.002436449281247462,
+      "grad_norm": 8.198959079831749,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 1.2135,
+      "step": 30
+    },
+    {
+      "epoch": 0.002517664257289044,
+      "grad_norm": 7.469329596056761,
+      "learning_rate": 4.18918918918919e-07,
+      "loss": 1.162,
+      "step": 31
+    },
+    {
+      "epoch": 0.002598879233330626,
+      "grad_norm": 6.064802318250161,
+      "learning_rate": 4.324324324324325e-07,
+      "loss": 1.0726,
+      "step": 32
+    },
+    {
+      "epoch": 0.0026800942093722084,
+      "grad_norm": 7.561733885808792,
+      "learning_rate": 4.45945945945946e-07,
+      "loss": 0.9132,
+      "step": 33
+    },
+    {
+      "epoch": 0.00276130918541379,
+      "grad_norm": 5.510848606396675,
+      "learning_rate": 4.5945945945945953e-07,
+      "loss": 0.8835,
+      "step": 34
+    },
+    {
+      "epoch": 0.0028425241614553724,
+      "grad_norm": 5.05147854422008,
+      "learning_rate": 4.7297297297297305e-07,
+      "loss": 1.025,
+      "step": 35
+    },
+    {
+      "epoch": 0.0029237391374969546,
+      "grad_norm": 7.449384043723888,
+      "learning_rate": 4.864864864864865e-07,
+      "loss": 1.0634,
+      "step": 36
+    },
+    {
+      "epoch": 0.0030049541135385364,
+      "grad_norm": 5.554970539302449,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.9037,
+      "step": 37
+    },
+    {
+      "epoch": 0.0030861690895801186,
+      "grad_norm": 5.648676923146146,
+      "learning_rate": 5.135135135135135e-07,
+      "loss": 1.0197,
+      "step": 38
+    },
+    {
+      "epoch": 0.0031673840656217004,
+      "grad_norm": 5.581517917957071,
+      "learning_rate": 5.270270270270271e-07,
+      "loss": 0.9723,
+      "step": 39
+    },
+    {
+      "epoch": 0.0032485990416632827,
+      "grad_norm": 6.811513573858421,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 1.0647,
+      "step": 40
+    },
+    {
+      "epoch": 0.003329814017704865,
+      "grad_norm": 13.02023304402827,
+      "learning_rate": 5.540540540540542e-07,
+      "loss": 0.9535,
+      "step": 41
+    },
+    {
+      "epoch": 0.0034110289937464467,
+      "grad_norm": 5.903154525353516,
+      "learning_rate": 5.675675675675676e-07,
+      "loss": 0.9207,
+      "step": 42
+    },
+    {
+      "epoch": 0.003492243969788029,
+      "grad_norm": 5.713506629308762,
+      "learning_rate": 5.810810810810812e-07,
+      "loss": 0.901,
+      "step": 43
+    },
+    {
+      "epoch": 0.003573458945829611,
+      "grad_norm": 7.568618399596037,
+      "learning_rate": 5.945945945945947e-07,
+      "loss": 0.8387,
+      "step": 44
+    },
+    {
+      "epoch": 0.003654673921871193,
+      "grad_norm": 4.684278865041181,
+      "learning_rate": 6.081081081081082e-07,
+      "loss": 0.9369,
+      "step": 45
+    },
+    {
+      "epoch": 0.003735888897912775,
+      "grad_norm": 4.719765340896398,
+      "learning_rate": 6.216216216216217e-07,
+      "loss": 0.9138,
+      "step": 46
+    },
+    {
+      "epoch": 0.0038171038739543573,
+      "grad_norm": 6.215441613276706,
+      "learning_rate": 6.351351351351353e-07,
+      "loss": 0.8665,
+      "step": 47
+    },
+    {
+      "epoch": 0.003898318849995939,
+      "grad_norm": 5.80882937072084,
+      "learning_rate": 6.486486486486487e-07,
+      "loss": 0.8277,
+      "step": 48
+    },
+    {
+      "epoch": 0.003979533826037522,
+      "grad_norm": 4.4336706590480786,
+      "learning_rate": 6.621621621621623e-07,
+      "loss": 0.9194,
+      "step": 49
+    },
+    {
+      "epoch": 0.004060748802079103,
+      "grad_norm": 6.287882016956921,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 0.997,
+      "step": 50
+    },
+    {
+      "epoch": 0.004141963778120685,
+      "grad_norm": 5.639526953982635,
+      "learning_rate": 6.891891891891893e-07,
+      "loss": 0.9637,
+      "step": 51
+    },
+    {
+      "epoch": 0.0042231787541622675,
+      "grad_norm": 5.51310604984265,
+      "learning_rate": 7.027027027027028e-07,
+      "loss": 0.8232,
+      "step": 52
+    },
+    {
+      "epoch": 0.00430439373020385,
+      "grad_norm": 4.517639703226352,
+      "learning_rate": 7.162162162162164e-07,
+      "loss": 0.8983,
+      "step": 53
+    },
+    {
+      "epoch": 0.004385608706245432,
+      "grad_norm": 6.666343453939387,
+      "learning_rate": 7.297297297297298e-07,
+      "loss": 0.7985,
+      "step": 54
+    },
+    {
+      "epoch": 0.004466823682287013,
+      "grad_norm": 5.422191276786231,
+      "learning_rate": 7.432432432432434e-07,
+      "loss": 0.9058,
+      "step": 55
+    },
+    {
+      "epoch": 0.0045480386583285955,
+      "grad_norm": 4.624429643253304,
+      "learning_rate": 7.567567567567569e-07,
+      "loss": 0.7943,
+      "step": 56
+    },
+    {
+      "epoch": 0.004629253634370178,
+      "grad_norm": 4.391972463636353,
+      "learning_rate": 7.702702702702704e-07,
+      "loss": 0.8046,
+      "step": 57
+    },
+    {
+      "epoch": 0.00471046861041176,
+      "grad_norm": 7.0317366536589745,
+      "learning_rate": 7.837837837837839e-07,
+      "loss": 0.96,
+      "step": 58
+    },
+    {
+      "epoch": 0.004791683586453342,
+      "grad_norm": 6.253750275868285,
+      "learning_rate": 7.972972972972974e-07,
+      "loss": 0.8267,
+      "step": 59
+    },
+    {
+      "epoch": 0.004872898562494924,
+      "grad_norm": 4.2004573670066545,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.7541,
+      "step": 60
+    },
+    {
+      "epoch": 0.004954113538536506,
+      "grad_norm": 4.843922828461304,
+      "learning_rate": 8.243243243243244e-07,
+      "loss": 0.8652,
+      "step": 61
+    },
+    {
+      "epoch": 0.005035328514578088,
+      "grad_norm": 6.663532165513732,
+      "learning_rate": 8.37837837837838e-07,
+      "loss": 0.9765,
+      "step": 62
+    },
+    {
+      "epoch": 0.00511654349061967,
+      "grad_norm": 5.518475179846077,
+      "learning_rate": 8.513513513513514e-07,
+      "loss": 0.8343,
+      "step": 63
+    },
+    {
+      "epoch": 0.005197758466661252,
+      "grad_norm": 4.16743993524973,
+      "learning_rate": 8.64864864864865e-07,
+      "loss": 0.8835,
+      "step": 64
+    },
+    {
+      "epoch": 0.005278973442702835,
+      "grad_norm": 4.724546088137992,
+      "learning_rate": 8.783783783783785e-07,
+      "loss": 0.8031,
+      "step": 65
+    },
+    {
+      "epoch": 0.005360188418744417,
+      "grad_norm": 5.965079554752788,
+      "learning_rate": 8.91891891891892e-07,
+      "loss": 0.8464,
+      "step": 66
+    },
+    {
+      "epoch": 0.005441403394785998,
+      "grad_norm": 7.388225521826432,
+      "learning_rate": 9.054054054054055e-07,
+      "loss": 0.9254,
+      "step": 67
+    },
+    {
+      "epoch": 0.00552261837082758,
+      "grad_norm": 4.486467374218727,
+      "learning_rate": 9.189189189189191e-07,
+      "loss": 0.7852,
+      "step": 68
+    },
+    {
+      "epoch": 0.005603833346869163,
+      "grad_norm": 4.818486349074325,
+      "learning_rate": 9.324324324324325e-07,
+      "loss": 0.6409,
+      "step": 69
+    },
+    {
+      "epoch": 0.005685048322910745,
+      "grad_norm": 12.591926984529453,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 0.6555,
+      "step": 70
+    },
+    {
+      "epoch": 0.005766263298952327,
+      "grad_norm": 3.582506346281597,
+      "learning_rate": 9.594594594594596e-07,
+      "loss": 0.7594,
+      "step": 71
+    },
+    {
+      "epoch": 0.005847478274993909,
+      "grad_norm": 6.316272242206502,
+      "learning_rate": 9.72972972972973e-07,
+      "loss": 0.8258,
+      "step": 72
+    },
+    {
+      "epoch": 0.005928693251035491,
+      "grad_norm": 7.768203703731759,
+      "learning_rate": 9.864864864864867e-07,
+      "loss": 0.7343,
+      "step": 73
+    },
+    {
+      "epoch": 0.006009908227077073,
+      "grad_norm": 4.411865830101082,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.8984,
+      "step": 74
+    },
+    {
+      "epoch": 0.006091123203118655,
+      "grad_norm": 6.221797107498061,
+      "learning_rate": 1.0135135135135136e-06,
+      "loss": 0.8646,
+      "step": 75
+    },
+    {
+      "epoch": 0.006172338179160237,
+      "grad_norm": 4.668292899364404,
+      "learning_rate": 1.027027027027027e-06,
+      "loss": 0.7082,
+      "step": 76
+    },
+    {
+      "epoch": 0.0062535531552018195,
+      "grad_norm": 5.307505469950394,
+      "learning_rate": 1.0405405405405408e-06,
+      "loss": 0.7584,
+      "step": 77
+    },
+    {
+      "epoch": 0.006334768131243401,
+      "grad_norm": 5.411978113961716,
+      "learning_rate": 1.0540540540540542e-06,
+      "loss": 0.8715,
+      "step": 78
+    },
+    {
+      "epoch": 0.006415983107284983,
+      "grad_norm": 11.725200388824204,
+      "learning_rate": 1.0675675675675677e-06,
+      "loss": 0.8187,
+      "step": 79
+    },
+    {
+      "epoch": 0.006497198083326565,
+      "grad_norm": 7.066619804350216,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.7657,
+      "step": 80
+    },
+    {
+      "epoch": 0.0065784130593681475,
+      "grad_norm": 5.206447831213838,
+      "learning_rate": 1.0945945945945948e-06,
+      "loss": 1.0916,
+      "step": 81
+    },
+    {
+      "epoch": 0.00665962803540973,
+      "grad_norm": 6.270913752130497,
+      "learning_rate": 1.1081081081081083e-06,
+      "loss": 0.695,
+      "step": 82
+    },
+    {
+      "epoch": 0.006740843011451312,
+      "grad_norm": 5.602218933207931,
+      "learning_rate": 1.1216216216216218e-06,
+      "loss": 0.7361,
+      "step": 83
+    },
+    {
+      "epoch": 0.006822057987492893,
+      "grad_norm": 5.592916777746583,
+      "learning_rate": 1.1351351351351352e-06,
+      "loss": 0.6823,
+      "step": 84
+    },
+    {
+      "epoch": 0.0069032729635344755,
+      "grad_norm": 6.7335919368870565,
+      "learning_rate": 1.148648648648649e-06,
+      "loss": 0.7927,
+      "step": 85
+    },
+    {
+      "epoch": 0.006984487939576058,
+      "grad_norm": 7.941076590684681,
+      "learning_rate": 1.1621621621621624e-06,
+      "loss": 0.8417,
+      "step": 86
+    },
+    {
+      "epoch": 0.00706570291561764,
+      "grad_norm": 5.969244160464252,
+      "learning_rate": 1.1756756756756758e-06,
+      "loss": 0.689,
+      "step": 87
+    },
+    {
+      "epoch": 0.007146917891659222,
+      "grad_norm": 5.118618058340234,
+      "learning_rate": 1.1891891891891893e-06,
+      "loss": 0.9608,
+      "step": 88
+    },
+    {
+      "epoch": 0.007228132867700804,
+      "grad_norm": 8.55651778885645,
+      "learning_rate": 1.2027027027027028e-06,
+      "loss": 0.6376,
+      "step": 89
+    },
+    {
+      "epoch": 0.007309347843742386,
+      "grad_norm": 8.9312254966113,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 0.8781,
+      "step": 90
+    },
+    {
+      "epoch": 0.007390562819783968,
+      "grad_norm": 5.435421842215974,
+      "learning_rate": 1.22972972972973e-06,
+      "loss": 0.7296,
+      "step": 91
+    },
+    {
+      "epoch": 0.00747177779582555,
+      "grad_norm": 7.00575424705863,
+      "learning_rate": 1.2432432432432434e-06,
+      "loss": 0.7004,
+      "step": 92
+    },
+    {
+      "epoch": 0.007552992771867132,
+      "grad_norm": 5.126051571596671,
+      "learning_rate": 1.2567567567567568e-06,
+      "loss": 0.6933,
+      "step": 93
+    },
+    {
+      "epoch": 0.007634207747908715,
+      "grad_norm": 7.546722570854549,
+      "learning_rate": 1.2702702702702705e-06,
+      "loss": 0.7443,
+      "step": 94
+    },
+    {
+      "epoch": 0.007715422723950297,
+      "grad_norm": 5.82298529351105,
+      "learning_rate": 1.2837837837837838e-06,
+      "loss": 0.6609,
+      "step": 95
+    },
+    {
+      "epoch": 0.007796637699991878,
+      "grad_norm": 5.395134658996866,
+      "learning_rate": 1.2972972972972974e-06,
+      "loss": 0.6966,
+      "step": 96
+    },
+    {
+      "epoch": 0.007877852676033461,
+      "grad_norm": 5.715845533979447,
+      "learning_rate": 1.310810810810811e-06,
+      "loss": 0.8075,
+      "step": 97
+    },
+    {
+      "epoch": 0.007959067652075043,
+      "grad_norm": 4.346663727077859,
+      "learning_rate": 1.3243243243243246e-06,
+      "loss": 0.8382,
+      "step": 98
+    },
+    {
+      "epoch": 0.008040282628116624,
+      "grad_norm": 4.510360891483355,
+      "learning_rate": 1.3378378378378378e-06,
+      "loss": 0.8409,
+      "step": 99
+    },
+    {
+      "epoch": 0.008121497604158206,
+      "grad_norm": 6.252214797052545,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.8968,
+      "step": 100
+    },
+    {
+      "epoch": 0.008202712580199788,
+      "grad_norm": 4.133668140760274,
+      "learning_rate": 1.364864864864865e-06,
+      "loss": 0.7697,
+      "step": 101
+    },
+    {
+      "epoch": 0.00828392755624137,
+      "grad_norm": 5.13995005645122,
+      "learning_rate": 1.3783783783783786e-06,
+      "loss": 0.8302,
+      "step": 102
+    },
+    {
+      "epoch": 0.008365142532282953,
+      "grad_norm": 5.645133972443234,
+      "learning_rate": 1.391891891891892e-06,
+      "loss": 0.8553,
+      "step": 103
+    },
+    {
+      "epoch": 0.008446357508324535,
+      "grad_norm": 5.768137707230809,
+      "learning_rate": 1.4054054054054056e-06,
+      "loss": 0.7269,
+      "step": 104
+    },
+    {
+      "epoch": 0.008527572484366117,
+      "grad_norm": 3.9384737579894358,
+      "learning_rate": 1.418918918918919e-06,
+      "loss": 0.8465,
+      "step": 105
+    },
+    {
+      "epoch": 0.0086087874604077,
+      "grad_norm": 6.608951862752225,
+      "learning_rate": 1.4324324324324327e-06,
+      "loss": 0.7358,
+      "step": 106
+    },
+    {
+      "epoch": 0.008690002436449282,
+      "grad_norm": 4.8856541626722425,
+      "learning_rate": 1.445945945945946e-06,
+      "loss": 0.9122,
+      "step": 107
+    },
+    {
+      "epoch": 0.008771217412490864,
+      "grad_norm": 3.744812822198274,
+      "learning_rate": 1.4594594594594596e-06,
+      "loss": 0.7847,
+      "step": 108
+    },
+    {
+      "epoch": 0.008852432388532446,
+      "grad_norm": 4.555012637892476,
+      "learning_rate": 1.4729729729729731e-06,
+      "loss": 0.6125,
+      "step": 109
+    },
+    {
+      "epoch": 0.008933647364574027,
+      "grad_norm": 5.417265029398178,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 0.6426,
+      "step": 110
+    },
+    {
+      "epoch": 0.009014862340615609,
+      "grad_norm": 6.978404404546592,
+      "learning_rate": 1.5e-06,
+      "loss": 0.8351,
+      "step": 111
+    },
+    {
+      "epoch": 0.009096077316657191,
+      "grad_norm": 7.486793674946231,
+      "learning_rate": 1.5135135135135137e-06,
+      "loss": 0.9104,
+      "step": 112
+    },
+    {
+      "epoch": 0.009177292292698773,
+      "grad_norm": 6.427379250554024,
+      "learning_rate": 1.5270270270270272e-06,
+      "loss": 0.7182,
+      "step": 113
+    },
+    {
+      "epoch": 0.009258507268740356,
+      "grad_norm": 7.581127803220287,
+      "learning_rate": 1.5405405405405409e-06,
+      "loss": 0.7966,
+      "step": 114
+    },
+    {
+      "epoch": 0.009339722244781938,
+      "grad_norm": 4.51084883429586,
+      "learning_rate": 1.5540540540540541e-06,
+      "loss": 0.7349,
+      "step": 115
+    },
+    {
+      "epoch": 0.00942093722082352,
+      "grad_norm": 5.521276273858657,
+      "learning_rate": 1.5675675675675678e-06,
+      "loss": 0.6476,
+      "step": 116
+    },
+    {
+      "epoch": 0.009502152196865102,
+      "grad_norm": 10.232542435537884,
+      "learning_rate": 1.5810810810810812e-06,
+      "loss": 0.7003,
+      "step": 117
+    },
+    {
+      "epoch": 0.009583367172906684,
+      "grad_norm": 5.670487468601197,
+      "learning_rate": 1.5945945945945947e-06,
+      "loss": 0.921,
+      "step": 118
+    },
+    {
+      "epoch": 0.009664582148948267,
+      "grad_norm": 5.136554782606984,
+      "learning_rate": 1.6081081081081082e-06,
+      "loss": 0.6932,
+      "step": 119
+    },
+    {
+      "epoch": 0.009745797124989849,
+      "grad_norm": 3.81643952542769,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.7676,
+      "step": 120
+    },
+    {
+      "epoch": 0.009827012101031431,
+      "grad_norm": 7.129104413729071,
+      "learning_rate": 1.6351351351351353e-06,
+      "loss": 0.7903,
+      "step": 121
+    },
+    {
+      "epoch": 0.009908227077073012,
+      "grad_norm": 5.89283670831236,
+      "learning_rate": 1.6486486486486488e-06,
+      "loss": 0.6741,
+      "step": 122
+    },
+    {
+      "epoch": 0.009989442053114594,
+      "grad_norm": 4.728889968775353,
+      "learning_rate": 1.6621621621621622e-06,
+      "loss": 0.7493,
+      "step": 123
+    },
+    {
+      "epoch": 0.010070657029156176,
+      "grad_norm": 5.400043416105325,
+      "learning_rate": 1.675675675675676e-06,
+      "loss": 0.6254,
+      "step": 124
+    },
+    {
+      "epoch": 0.010151872005197758,
+      "grad_norm": 5.7442992971209605,
+      "learning_rate": 1.6891891891891894e-06,
+      "loss": 0.6588,
+      "step": 125
+    },
+    {
+      "epoch": 0.01023308698123934,
+      "grad_norm": 5.225132595645112,
+      "learning_rate": 1.7027027027027028e-06,
+      "loss": 0.721,
+      "step": 126
+    },
+    {
+      "epoch": 0.010314301957280923,
+      "grad_norm": 4.917827752874381,
+      "learning_rate": 1.7162162162162163e-06,
+      "loss": 0.6719,
+      "step": 127
+    },
+    {
+      "epoch": 0.010395516933322505,
+      "grad_norm": 5.96105392795946,
+      "learning_rate": 1.72972972972973e-06,
+      "loss": 0.8192,
+      "step": 128
+    },
+    {
+      "epoch": 0.010476731909364087,
+      "grad_norm": 5.348491477184876,
+      "learning_rate": 1.7432432432432432e-06,
+      "loss": 0.741,
+      "step": 129
+    },
+    {
+      "epoch": 0.01055794688540567,
+      "grad_norm": 4.046650669596968,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 0.6943,
+      "step": 130
+    },
+    {
+      "epoch": 0.010639161861447251,
+      "grad_norm": 5.045700573097575,
+      "learning_rate": 1.7702702702702704e-06,
+      "loss": 0.5737,
+      "step": 131
+    },
+    {
+      "epoch": 0.010720376837488834,
+      "grad_norm": 4.332973323075942,
+      "learning_rate": 1.783783783783784e-06,
+      "loss": 0.7721,
+      "step": 132
+    },
+    {
+      "epoch": 0.010801591813530414,
+      "grad_norm": 12.053968494279287,
+      "learning_rate": 1.7972972972972973e-06,
+      "loss": 0.7081,
+      "step": 133
+    },
+    {
+      "epoch": 0.010882806789571996,
+      "grad_norm": 4.314977931658641,
+      "learning_rate": 1.810810810810811e-06,
+      "loss": 0.7476,
+      "step": 134
+    },
+    {
+      "epoch": 0.010964021765613579,
+      "grad_norm": 4.372992825514325,
+      "learning_rate": 1.8243243243243245e-06,
+      "loss": 0.5623,
+      "step": 135
+    },
+    {
+      "epoch": 0.01104523674165516,
+      "grad_norm": 5.06501074840964,
+      "learning_rate": 1.8378378378378381e-06,
+      "loss": 0.8409,
+      "step": 136
+    },
+    {
+      "epoch": 0.011126451717696743,
+      "grad_norm": 5.486341378807148,
+      "learning_rate": 1.8513513513513514e-06,
+      "loss": 0.7097,
+      "step": 137
+    },
+    {
+      "epoch": 0.011207666693738325,
+      "grad_norm": 5.885624763272731,
+      "learning_rate": 1.864864864864865e-06,
+      "loss": 0.8605,
+      "step": 138
+    },
+    {
+      "epoch": 0.011288881669779907,
+      "grad_norm": 5.810017768183461,
+      "learning_rate": 1.8783783783783785e-06,
+      "loss": 0.7589,
+      "step": 139
+    },
+    {
+      "epoch": 0.01137009664582149,
+      "grad_norm": 5.241258116276019,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.6821,
+      "step": 140
+    },
+    {
+      "epoch": 0.011451311621863072,
+      "grad_norm": 5.141543904486299,
+      "learning_rate": 1.9054054054054054e-06,
+      "loss": 0.6702,
+      "step": 141
+    },
+    {
+      "epoch": 0.011532526597904654,
+      "grad_norm": 6.614627634564522,
+      "learning_rate": 1.918918918918919e-06,
+      "loss": 0.7541,
+      "step": 142
+    },
+    {
+      "epoch": 0.011613741573946236,
+      "grad_norm": 4.270716146835968,
+      "learning_rate": 1.9324324324324326e-06,
+      "loss": 0.6118,
+      "step": 143
+    },
+    {
+      "epoch": 0.011694956549987819,
+      "grad_norm": 4.833439173611212,
+      "learning_rate": 1.945945945945946e-06,
+      "loss": 0.6631,
+      "step": 144
+    },
+    {
+      "epoch": 0.011776171526029399,
+      "grad_norm": 6.670666111898159,
+      "learning_rate": 1.9594594594594595e-06,
+      "loss": 0.5792,
+      "step": 145
+    },
+    {
+      "epoch": 0.011857386502070981,
+      "grad_norm": 5.320699935552583,
+      "learning_rate": 1.9729729729729734e-06,
+      "loss": 0.6649,
+      "step": 146
+    },
+    {
+      "epoch": 0.011938601478112564,
+      "grad_norm": 10.048210540269071,
+      "learning_rate": 1.9864864864864864e-06,
+      "loss": 0.6055,
+      "step": 147
+    },
+    {
+      "epoch": 0.012019816454154146,
+      "grad_norm": 4.8358064531705525,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7224,
+      "step": 148
+    },
+    {
+      "epoch": 0.012101031430195728,
+      "grad_norm": 8.585169822222198,
+      "learning_rate": 2.013513513513514e-06,
+      "loss": 0.7546,
+      "step": 149
+    },
+    {
+      "epoch": 0.01218224640623731,
+      "grad_norm": 4.847625342598578,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 0.6867,
+      "step": 150
+    },
+    {
+      "epoch": 0.012263461382278892,
+      "grad_norm": 3.7722782135223594,
+      "learning_rate": 2.0405405405405407e-06,
+      "loss": 0.7262,
+      "step": 151
+    },
+    {
+      "epoch": 0.012344676358320475,
+      "grad_norm": 5.149292525910975,
+      "learning_rate": 2.054054054054054e-06,
+      "loss": 0.8792,
+      "step": 152
+    },
+    {
+      "epoch": 0.012425891334362057,
+      "grad_norm": 6.26503383092344,
+      "learning_rate": 2.0675675675675677e-06,
+      "loss": 0.7837,
+      "step": 153
+    },
+    {
+      "epoch": 0.012507106310403639,
+      "grad_norm": 4.188912763837703,
+      "learning_rate": 2.0810810810810815e-06,
+      "loss": 0.5692,
+      "step": 154
+    },
+    {
+      "epoch": 0.012588321286445221,
+      "grad_norm": 5.04456436284514,
+      "learning_rate": 2.0945945945945946e-06,
+      "loss": 0.652,
+      "step": 155
+    },
+    {
+      "epoch": 0.012669536262486802,
+      "grad_norm": 7.405968257695848,
+      "learning_rate": 2.1081081081081085e-06,
+      "loss": 0.6266,
+      "step": 156
+    },
+    {
+      "epoch": 0.012750751238528384,
+      "grad_norm": 5.310169429627866,
+      "learning_rate": 2.121621621621622e-06,
+      "loss": 0.6542,
+      "step": 157
+    },
+    {
+      "epoch": 0.012831966214569966,
+      "grad_norm": 5.046936455927635,
+      "learning_rate": 2.1351351351351354e-06,
+      "loss": 0.8108,
+      "step": 158
+    },
+    {
+      "epoch": 0.012913181190611548,
+      "grad_norm": 4.036458244440137,
+      "learning_rate": 2.148648648648649e-06,
+      "loss": 0.6339,
+      "step": 159
+    },
+    {
+      "epoch": 0.01299439616665313,
+      "grad_norm": 5.254253720134002,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.7129,
+      "step": 160
+    },
+    {
+      "epoch": 0.013075611142694713,
+      "grad_norm": 4.842001153737952,
+      "learning_rate": 2.175675675675676e-06,
+      "loss": 0.7669,
+      "step": 161
+    },
+    {
+      "epoch": 0.013156826118736295,
+      "grad_norm": 3.45883013736874,
+      "learning_rate": 2.1891891891891897e-06,
+      "loss": 0.7403,
+      "step": 162
+    },
+    {
+      "epoch": 0.013238041094777877,
+      "grad_norm": 6.615557242600742,
+      "learning_rate": 2.2027027027027027e-06,
+      "loss": 0.6072,
+      "step": 163
+    },
+    {
+      "epoch": 0.01331925607081946,
+      "grad_norm": 5.295390317262873,
+      "learning_rate": 2.2162162162162166e-06,
+      "loss": 0.4518,
+      "step": 164
+    },
+    {
+      "epoch": 0.013400471046861042,
+      "grad_norm": 4.553708607425031,
+      "learning_rate": 2.22972972972973e-06,
+      "loss": 0.7405,
+      "step": 165
+    },
+    {
+      "epoch": 0.013481686022902624,
+      "grad_norm": 5.730983978090373,
+      "learning_rate": 2.2432432432432435e-06,
+      "loss": 0.7216,
+      "step": 166
+    },
+    {
+      "epoch": 0.013562900998944206,
+      "grad_norm": 6.28068035640207,
+      "learning_rate": 2.256756756756757e-06,
+      "loss": 0.543,
+      "step": 167
+    },
+    {
+      "epoch": 0.013644115974985787,
+      "grad_norm": 5.534240505988875,
+      "learning_rate": 2.2702702702702705e-06,
+      "loss": 0.7273,
+      "step": 168
+    },
+    {
+      "epoch": 0.013725330951027369,
+      "grad_norm": 10.022692075346896,
+      "learning_rate": 2.283783783783784e-06,
+      "loss": 0.6259,
+      "step": 169
+    },
+    {
+      "epoch": 0.013806545927068951,
+      "grad_norm": 4.133335838910277,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 0.5538,
+      "step": 170
+    },
+    {
+      "epoch": 0.013887760903110533,
+      "grad_norm": 6.849412426150848,
+      "learning_rate": 2.310810810810811e-06,
+      "loss": 0.5594,
+      "step": 171
+    },
+    {
+      "epoch": 0.013968975879152115,
+      "grad_norm": 7.2240081386637955,
+      "learning_rate": 2.3243243243243247e-06,
+      "loss": 0.6406,
+      "step": 172
+    },
+    {
+      "epoch": 0.014050190855193698,
+      "grad_norm": 4.948498140450114,
+      "learning_rate": 2.337837837837838e-06,
+      "loss": 0.697,
+      "step": 173
+    },
+    {
+      "epoch": 0.01413140583123528,
+      "grad_norm": 3.8421671175611465,
+      "learning_rate": 2.3513513513513517e-06,
+      "loss": 0.709,
+      "step": 174
+    },
+    {
+      "epoch": 0.014212620807276862,
+      "grad_norm": 4.9263286743150925,
+      "learning_rate": 2.364864864864865e-06,
+      "loss": 0.5944,
+      "step": 175
+    },
+    {
+      "epoch": 0.014293835783318444,
+      "grad_norm": 4.181352300851268,
+      "learning_rate": 2.3783783783783786e-06,
+      "loss": 0.7951,
+      "step": 176
+    },
+    {
+      "epoch": 0.014375050759360027,
+      "grad_norm": 5.217958869841234,
+      "learning_rate": 2.391891891891892e-06,
+      "loss": 0.8847,
+      "step": 177
+    },
+    {
+      "epoch": 0.014456265735401609,
+      "grad_norm": 5.53287305794427,
+      "learning_rate": 2.4054054054054055e-06,
+      "loss": 0.6506,
+      "step": 178
+    },
+    {
+      "epoch": 0.01453748071144319,
+      "grad_norm": 4.977958862414195,
+      "learning_rate": 2.418918918918919e-06,
+      "loss": 0.649,
+      "step": 179
+    },
+    {
+      "epoch": 0.014618695687484771,
+      "grad_norm": 7.079003805106819,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.5173,
+      "step": 180
+    },
+    {
+      "epoch": 0.014699910663526354,
+      "grad_norm": 6.4006738662761675,
+      "learning_rate": 2.4459459459459463e-06,
+      "loss": 0.7425,
+      "step": 181
+    },
+    {
+      "epoch": 0.014781125639567936,
+      "grad_norm": 7.223867008336038,
+      "learning_rate": 2.45945945945946e-06,
+      "loss": 0.7116,
+      "step": 182
+    },
+    {
+      "epoch": 0.014862340615609518,
+      "grad_norm": 4.74150217223079,
+      "learning_rate": 2.4729729729729733e-06,
+      "loss": 0.8202,
+      "step": 183
+    },
+    {
+      "epoch": 0.0149435555916511,
+      "grad_norm": 5.838162332180424,
+      "learning_rate": 2.4864864864864867e-06,
+      "loss": 0.7202,
+      "step": 184
+    },
+    {
+      "epoch": 0.015024770567692683,
+      "grad_norm": 7.3164412014837845,
+      "learning_rate": 2.5e-06,
+      "loss": 0.6318,
+      "step": 185
+    },
+    {
+      "epoch": 0.015105985543734265,
+      "grad_norm": 4.0466470539819674,
+      "learning_rate": 2.5135135135135137e-06,
+      "loss": 0.6161,
+      "step": 186
+    },
+    {
+      "epoch": 0.015187200519775847,
+      "grad_norm": 4.20696930651959,
+      "learning_rate": 2.527027027027027e-06,
+      "loss": 0.8844,
+      "step": 187
+    },
+    {
+      "epoch": 0.01526841549581743,
+      "grad_norm": 4.619062654793844,
+      "learning_rate": 2.540540540540541e-06,
+      "loss": 0.7893,
+      "step": 188
+    },
+    {
+      "epoch": 0.015349630471859011,
+      "grad_norm": 4.028095158003242,
+      "learning_rate": 2.554054054054054e-06,
+      "loss": 0.9053,
+      "step": 189
+    },
+    {
+      "epoch": 0.015430845447900594,
+      "grad_norm": 5.474147278368468,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 0.6785,
+      "step": 190
+    },
+    {
+      "epoch": 0.015512060423942174,
+      "grad_norm": 7.068401125839219,
+      "learning_rate": 2.581081081081081e-06,
+      "loss": 0.6353,
+      "step": 191
+    },
+    {
+      "epoch": 0.015593275399983756,
+      "grad_norm": 6.211801926350057,
+      "learning_rate": 2.594594594594595e-06,
+      "loss": 0.6171,
+      "step": 192
+    },
+    {
+      "epoch": 0.01567449037602534,
+      "grad_norm": 6.8766856866833805,
+      "learning_rate": 2.6081081081081083e-06,
+      "loss": 0.6291,
+      "step": 193
+    },
+    {
+      "epoch": 0.015755705352066923,
+      "grad_norm": 4.450746282218491,
+      "learning_rate": 2.621621621621622e-06,
+      "loss": 0.7322,
+      "step": 194
+    },
+    {
+      "epoch": 0.015836920328108503,
+      "grad_norm": 6.737626992175194,
+      "learning_rate": 2.6351351351351353e-06,
+      "loss": 0.8431,
+      "step": 195
+    },
+    {
+      "epoch": 0.015918135304150087,
+      "grad_norm": 5.202217528227061,
+      "learning_rate": 2.648648648648649e-06,
+      "loss": 0.6027,
+      "step": 196
+    },
+    {
+      "epoch": 0.015999350280191667,
+      "grad_norm": 6.136140183171829,
+      "learning_rate": 2.662162162162162e-06,
+      "loss": 0.7227,
+      "step": 197
+    },
+    {
+      "epoch": 0.016080565256233248,
+      "grad_norm": 6.20592135137412,
+      "learning_rate": 2.6756756756756757e-06,
+      "loss": 0.8206,
+      "step": 198
+    },
+    {
+      "epoch": 0.016161780232274832,
+      "grad_norm": 5.542234344965935,
+      "learning_rate": 2.689189189189189e-06,
+      "loss": 0.5755,
+      "step": 199
+    },
+    {
+      "epoch": 0.016242995208316412,
+      "grad_norm": 6.481129594142135,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.9124,
+      "step": 200
+    },
+    {
+      "epoch": 0.016324210184357996,
+      "grad_norm": 8.819932985194717,
+      "learning_rate": 2.7162162162162165e-06,
+      "loss": 0.7077,
+      "step": 201
+    },
+    {
+      "epoch": 0.016405425160399577,
+      "grad_norm": 3.7133046294352496,
+      "learning_rate": 2.72972972972973e-06,
+      "loss": 0.7187,
+      "step": 202
+    },
+    {
+      "epoch": 0.01648664013644116,
+      "grad_norm": 4.768904441511917,
+      "learning_rate": 2.7432432432432434e-06,
+      "loss": 0.7079,
+      "step": 203
+    },
+    {
+      "epoch": 0.01656785511248274,
+      "grad_norm": 5.4109232849671445,
+      "learning_rate": 2.7567567567567573e-06,
+      "loss": 0.8741,
+      "step": 204
+    },
+    {
+      "epoch": 0.016649070088524325,
+      "grad_norm": 11.784275280928684,
+      "learning_rate": 2.7702702702702703e-06,
+      "loss": 0.6088,
+      "step": 205
+    },
+    {
+      "epoch": 0.016730285064565906,
+      "grad_norm": 6.596444076197084,
+      "learning_rate": 2.783783783783784e-06,
+      "loss": 0.7085,
+      "step": 206
+    },
+    {
+      "epoch": 0.01681150004060749,
+      "grad_norm": 4.7347497389094055,
+      "learning_rate": 2.7972972972972973e-06,
+      "loss": 0.7641,
+      "step": 207
+    },
+    {
+      "epoch": 0.01689271501664907,
+      "grad_norm": 4.895873979702855,
+      "learning_rate": 2.810810810810811e-06,
+      "loss": 0.613,
+      "step": 208
+    },
+    {
+      "epoch": 0.01697392999269065,
+      "grad_norm": 4.112688584540271,
+      "learning_rate": 2.8243243243243246e-06,
+      "loss": 0.6551,
+      "step": 209
+    },
+    {
+      "epoch": 0.017055144968732235,
+      "grad_norm": 4.280798819098727,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 0.8091,
+      "step": 210
+    },
+    {
+      "epoch": 0.017136359944773815,
+      "grad_norm": 4.953151146240816,
+      "learning_rate": 2.851351351351351e-06,
+      "loss": 0.6225,
+      "step": 211
+    },
+    {
+      "epoch": 0.0172175749208154,
+      "grad_norm": 4.858934778537565,
+      "learning_rate": 2.8648648648648654e-06,
+      "loss": 0.7894,
+      "step": 212
+    },
+    {
+      "epoch": 0.01729878989685698,
+      "grad_norm": 3.776464089377743,
+      "learning_rate": 2.8783783783783785e-06,
+      "loss": 0.6714,
+      "step": 213
+    },
+    {
+      "epoch": 0.017380004872898563,
+      "grad_norm": 5.927211431887128,
+      "learning_rate": 2.891891891891892e-06,
+      "loss": 0.6115,
+      "step": 214
+    },
+    {
+      "epoch": 0.017461219848940144,
+      "grad_norm": 4.290882863865101,
+      "learning_rate": 2.9054054054054054e-06,
+      "loss": 0.8019,
+      "step": 215
+    },
+    {
+      "epoch": 0.017542434824981728,
+      "grad_norm": 5.014996919723652,
+      "learning_rate": 2.9189189189189193e-06,
+      "loss": 0.9165,
+      "step": 216
+    },
+    {
+      "epoch": 0.01762364980102331,
+      "grad_norm": 4.678677528211619,
+      "learning_rate": 2.9324324324324328e-06,
+      "loss": 0.7034,
+      "step": 217
+    },
+    {
+      "epoch": 0.017704864777064892,
+      "grad_norm": 5.856721537107752,
+      "learning_rate": 2.9459459459459462e-06,
+      "loss": 0.6187,
+      "step": 218
+    },
+    {
+      "epoch": 0.017786079753106473,
+      "grad_norm": 5.675292275975791,
+      "learning_rate": 2.9594594594594593e-06,
+      "loss": 0.6282,
+      "step": 219
+    },
+    {
+      "epoch": 0.017867294729148053,
+      "grad_norm": 4.026148472844569,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.6984,
+      "step": 220
+    },
+    {
+      "epoch": 0.017948509705189637,
+      "grad_norm": 3.7553668002568115,
+      "learning_rate": 2.9864864864864866e-06,
+      "loss": 0.7059,
+      "step": 221
+    },
+    {
+      "epoch": 0.018029724681231218,
+      "grad_norm": 5.010656407185249,
+      "learning_rate": 3e-06,
+      "loss": 0.6676,
+      "step": 222
+    },
+    {
+      "epoch": 0.0181109396572728,
+      "grad_norm": 5.174403860939526,
+      "learning_rate": 3.0135135135135135e-06,
+      "loss": 0.6528,
+      "step": 223
+    },
+    {
+      "epoch": 0.018192154633314382,
+      "grad_norm": 9.300176785765524,
+      "learning_rate": 3.0270270270270274e-06,
+      "loss": 0.6981,
+      "step": 224
+    },
+    {
+      "epoch": 0.018273369609355966,
+      "grad_norm": 6.1866583084404185,
+      "learning_rate": 3.040540540540541e-06,
+      "loss": 0.704,
+      "step": 225
+    },
+    {
+      "epoch": 0.018354584585397547,
+      "grad_norm": 7.4629748217872605,
+      "learning_rate": 3.0540540540540544e-06,
+      "loss": 0.593,
+      "step": 226
+    },
+    {
+      "epoch": 0.01843579956143913,
+      "grad_norm": 4.530659236075315,
+      "learning_rate": 3.0675675675675674e-06,
+      "loss": 0.5932,
+      "step": 227
+    },
+    {
+      "epoch": 0.01851701453748071,
+      "grad_norm": 5.748440319813538,
+      "learning_rate": 3.0810810810810817e-06,
+      "loss": 0.7238,
+      "step": 228
+    },
+    {
+      "epoch": 0.018598229513522295,
+      "grad_norm": 40.72028295804037,
+      "learning_rate": 3.0945945945945947e-06,
+      "loss": 0.7017,
+      "step": 229
+    },
+    {
+      "epoch": 0.018679444489563875,
+      "grad_norm": 8.886281188290711,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 0.5761,
+      "step": 230
+    },
+    {
+      "epoch": 0.018760659465605456,
+      "grad_norm": 4.343869317099014,
+      "learning_rate": 3.1216216216216217e-06,
+      "loss": 0.8567,
+      "step": 231
+    },
+    {
+      "epoch": 0.01884187444164704,
+      "grad_norm": 4.744970167283222,
+      "learning_rate": 3.1351351351351356e-06,
+      "loss": 0.7763,
+      "step": 232
+    },
+    {
+      "epoch": 0.01892308941768862,
+      "grad_norm": 7.495714776851687,
+      "learning_rate": 3.148648648648649e-06,
+      "loss": 0.8186,
+      "step": 233
+    },
+    {
+      "epoch": 0.019004304393730204,
+      "grad_norm": 4.102840820567404,
+      "learning_rate": 3.1621621621621625e-06,
+      "loss": 0.7489,
+      "step": 234
+    },
+    {
+      "epoch": 0.019085519369771785,
+      "grad_norm": 3.8421345026888027,
+      "learning_rate": 3.1756756756756755e-06,
+      "loss": 0.6933,
+      "step": 235
+    },
+    {
+      "epoch": 0.01916673434581337,
+      "grad_norm": 7.387916364375026,
+      "learning_rate": 3.1891891891891894e-06,
+      "loss": 0.6052,
+      "step": 236
+    },
+    {
+      "epoch": 0.01924794932185495,
+      "grad_norm": 3.849533522847183,
+      "learning_rate": 3.202702702702703e-06,
+      "loss": 0.7804,
+      "step": 237
+    },
+    {
+      "epoch": 0.019329164297896533,
+      "grad_norm": 5.396298934498324,
+      "learning_rate": 3.2162162162162164e-06,
+      "loss": 0.6641,
+      "step": 238
+    },
+    {
+      "epoch": 0.019410379273938114,
+      "grad_norm": 4.940398668509269,
+      "learning_rate": 3.22972972972973e-06,
+      "loss": 0.5532,
+      "step": 239
+    },
+    {
+      "epoch": 0.019491594249979698,
+      "grad_norm": 6.577994445331441,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.7118,
+      "step": 240
+    },
+    {
+      "epoch": 0.019572809226021278,
+      "grad_norm": 5.519259337961804,
+      "learning_rate": 3.256756756756757e-06,
+      "loss": 0.7421,
+      "step": 241
+    },
+    {
+      "epoch": 0.019654024202062862,
+      "grad_norm": 6.0576611783888215,
+      "learning_rate": 3.2702702702702706e-06,
+      "loss": 0.6083,
+      "step": 242
+    },
+    {
+      "epoch": 0.019735239178104443,
+      "grad_norm": 8.457628617936836,
+      "learning_rate": 3.2837837837837837e-06,
+      "loss": 0.678,
+      "step": 243
+    },
+    {
+      "epoch": 0.019816454154146023,
+      "grad_norm": 7.364746035461257,
+      "learning_rate": 3.2972972972972976e-06,
+      "loss": 0.5896,
+      "step": 244
+    },
+    {
+      "epoch": 0.019897669130187607,
+      "grad_norm": 7.750984536568317,
+      "learning_rate": 3.310810810810811e-06,
+      "loss": 0.6219,
+      "step": 245
+    },
+    {
+      "epoch": 0.019978884106229187,
+      "grad_norm": 4.981685980750705,
+      "learning_rate": 3.3243243243243245e-06,
+      "loss": 0.6942,
+      "step": 246
+    },
+    {
+      "epoch": 0.02006009908227077,
+      "grad_norm": 6.278603739216485,
+      "learning_rate": 3.337837837837838e-06,
+      "loss": 0.7533,
+      "step": 247
+    },
+    {
+      "epoch": 0.020141314058312352,
+      "grad_norm": 4.801507815370232,
+      "learning_rate": 3.351351351351352e-06,
+      "loss": 0.6608,
+      "step": 248
+    },
+    {
+      "epoch": 0.020222529034353936,
+      "grad_norm": 7.021222735394233,
+      "learning_rate": 3.3648648648648653e-06,
+      "loss": 0.8164,
+      "step": 249
+    },
+    {
+      "epoch": 0.020303744010395516,
+      "grad_norm": 5.832060700035103,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 0.6886,
+      "step": 250
+    },
+    {
+      "epoch": 0.0203849589864371,
+      "grad_norm": 4.387381573209668,
+      "learning_rate": 3.391891891891892e-06,
+      "loss": 0.6018,
+      "step": 251
+    },
+    {
+      "epoch": 0.02046617396247868,
+      "grad_norm": 4.588515154439844,
+      "learning_rate": 3.4054054054054057e-06,
+      "loss": 0.8009,
+      "step": 252
+    },
+    {
+      "epoch": 0.020547388938520265,
+      "grad_norm": 5.9906722613866865,
+      "learning_rate": 3.418918918918919e-06,
+      "loss": 0.6694,
+      "step": 253
+    },
+    {
+      "epoch": 0.020628603914561845,
+      "grad_norm": 4.955892303287632,
+      "learning_rate": 3.4324324324324326e-06,
+      "loss": 0.8309,
+      "step": 254
+    },
+    {
+      "epoch": 0.020709818890603426,
+      "grad_norm": 4.824146659749785,
+      "learning_rate": 3.445945945945946e-06,
+      "loss": 0.8692,
+      "step": 255
+    },
+    {
+      "epoch": 0.02079103386664501,
+      "grad_norm": 5.5270982189809015,
+      "learning_rate": 3.45945945945946e-06,
+      "loss": 0.5992,
+      "step": 256
+    },
+    {
+      "epoch": 0.02087224884268659,
+      "grad_norm": 6.204903985337891,
+      "learning_rate": 3.4729729729729734e-06,
+      "loss": 0.777,
+      "step": 257
+    },
+    {
+      "epoch": 0.020953463818728174,
+      "grad_norm": 3.695647833466083,
+      "learning_rate": 3.4864864864864865e-06,
+      "loss": 0.6435,
+      "step": 258
+    },
+    {
+      "epoch": 0.021034678794769755,
+      "grad_norm": 5.291080353938334,
+      "learning_rate": 3.5e-06,
+      "loss": 0.6796,
+      "step": 259
+    },
+    {
+      "epoch": 0.02111589377081134,
+      "grad_norm": 4.5185266685147285,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.6448,
+      "step": 260
+    },
+    {
+      "epoch": 0.02119710874685292,
+      "grad_norm": 4.851027601896295,
+      "learning_rate": 3.5270270270270273e-06,
+      "loss": 0.682,
+      "step": 261
+    },
+    {
+      "epoch": 0.021278323722894503,
+      "grad_norm": 5.6100378122852925,
+      "learning_rate": 3.5405405405405408e-06,
+      "loss": 0.7084,
+      "step": 262
+    },
+    {
+      "epoch": 0.021359538698936083,
+      "grad_norm": 5.135851719796815,
+      "learning_rate": 3.5540540540540542e-06,
+      "loss": 0.8671,
+      "step": 263
+    },
+    {
+      "epoch": 0.021440753674977667,
+      "grad_norm": 5.735868784715722,
+      "learning_rate": 3.567567567567568e-06,
+      "loss": 0.5465,
+      "step": 264
+    },
+    {
+      "epoch": 0.021521968651019248,
+      "grad_norm": 4.041664795202519,
+      "learning_rate": 3.5810810810810816e-06,
+      "loss": 0.6117,
+      "step": 265
+    },
+    {
+      "epoch": 0.02160318362706083,
+      "grad_norm": 4.742239360128328,
+      "learning_rate": 3.5945945945945946e-06,
+      "loss": 0.6543,
+      "step": 266
+    },
+    {
+      "epoch": 0.021684398603102412,
+      "grad_norm": 5.6376075863303114,
+      "learning_rate": 3.608108108108108e-06,
+      "loss": 0.7132,
+      "step": 267
+    },
+    {
+      "epoch": 0.021765613579143993,
+      "grad_norm": 4.108413552937677,
+      "learning_rate": 3.621621621621622e-06,
+      "loss": 0.9163,
+      "step": 268
+    },
+    {
+      "epoch": 0.021846828555185577,
+      "grad_norm": 7.767807764665652,
+      "learning_rate": 3.6351351351351354e-06,
+      "loss": 0.6385,
+      "step": 269
+    },
+    {
+      "epoch": 0.021928043531227157,
+      "grad_norm": 5.864609873584057,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 0.6538,
+      "step": 270
+    },
+    {
+      "epoch": 0.02200925850726874,
+      "grad_norm": 4.820511968405664,
+      "learning_rate": 3.6621621621621624e-06,
+      "loss": 0.6993,
+      "step": 271
+    },
+    {
+      "epoch": 0.02209047348331032,
+      "grad_norm": 4.967071259608292,
+      "learning_rate": 3.6756756756756763e-06,
+      "loss": 0.8915,
+      "step": 272
+    },
+    {
+      "epoch": 0.022171688459351906,
+      "grad_norm": 3.7176252188194723,
+      "learning_rate": 3.6891891891891897e-06,
+      "loss": 0.906,
+      "step": 273
+    },
+    {
+      "epoch": 0.022252903435393486,
+      "grad_norm": 7.432800095587209,
+      "learning_rate": 3.7027027027027028e-06,
+      "loss": 0.5487,
+      "step": 274
+    },
+    {
+      "epoch": 0.02233411841143507,
+      "grad_norm": 4.725582603425111,
+      "learning_rate": 3.7162162162162162e-06,
+      "loss": 0.6497,
+      "step": 275
+    },
+    {
+      "epoch": 0.02241533338747665,
+      "grad_norm": 7.338793955644148,
+      "learning_rate": 3.72972972972973e-06,
+      "loss": 0.6416,
+      "step": 276
+    },
+    {
+      "epoch": 0.02249654836351823,
+      "grad_norm": 5.306407989783936,
+      "learning_rate": 3.7432432432432436e-06,
+      "loss": 0.6278,
+      "step": 277
+    },
+    {
+      "epoch": 0.022577763339559815,
+      "grad_norm": 5.1495205257748164,
+      "learning_rate": 3.756756756756757e-06,
+      "loss": 0.7679,
+      "step": 278
+    },
+    {
+      "epoch": 0.022658978315601395,
+      "grad_norm": 5.708997533226292,
+      "learning_rate": 3.7702702702702705e-06,
+      "loss": 0.7532,
+      "step": 279
+    },
+    {
+      "epoch": 0.02274019329164298,
+      "grad_norm": 8.83003188884847,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.7133,
+      "step": 280
+    },
+    {
+      "epoch": 0.02282140826768456,
+      "grad_norm": 5.539341071050379,
+      "learning_rate": 3.797297297297298e-06,
+      "loss": 0.7097,
+      "step": 281
+    },
+    {
+      "epoch": 0.022902623243726144,
+      "grad_norm": 8.240407300947401,
+      "learning_rate": 3.810810810810811e-06,
+      "loss": 0.6048,
+      "step": 282
+    },
+    {
+      "epoch": 0.022983838219767724,
+      "grad_norm": 4.587467280833204,
+      "learning_rate": 3.824324324324324e-06,
+      "loss": 0.8175,
+      "step": 283
+    },
+    {
+      "epoch": 0.02306505319580931,
+      "grad_norm": 7.179515697903426,
+      "learning_rate": 3.837837837837838e-06,
+      "loss": 0.8436,
+      "step": 284
+    },
+    {
+      "epoch": 0.02314626817185089,
+      "grad_norm": 5.309912678695326,
+      "learning_rate": 3.851351351351352e-06,
+      "loss": 0.625,
+      "step": 285
+    },
+    {
+      "epoch": 0.023227483147892473,
+      "grad_norm": 4.941339777332484,
+      "learning_rate": 3.864864864864865e-06,
+      "loss": 0.6989,
+      "step": 286
+    },
+    {
+      "epoch": 0.023308698123934053,
+      "grad_norm": 5.2052407828367215,
+      "learning_rate": 3.878378378378378e-06,
+      "loss": 0.5503,
+      "step": 287
+    },
+    {
+      "epoch": 0.023389913099975637,
+      "grad_norm": 4.238862573234204,
+      "learning_rate": 3.891891891891892e-06,
+      "loss": 0.8404,
+      "step": 288
+    },
+    {
+      "epoch": 0.023471128076017218,
+      "grad_norm": 9.415691207617115,
+      "learning_rate": 3.905405405405406e-06,
+      "loss": 0.6679,
+      "step": 289
+    },
+    {
+      "epoch": 0.023552343052058798,
+      "grad_norm": 4.698876605885647,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 0.5115,
+      "step": 290
+    },
+    {
+      "epoch": 0.023633558028100382,
+      "grad_norm": 8.590289244453471,
+      "learning_rate": 3.932432432432433e-06,
+      "loss": 0.8333,
+      "step": 291
+    },
+    {
+      "epoch": 0.023714773004141963,
+      "grad_norm": 3.9911010631765764,
+      "learning_rate": 3.945945945945947e-06,
+      "loss": 0.5686,
+      "step": 292
+    },
+    {
+      "epoch": 0.023795987980183547,
+      "grad_norm": 4.926542248398489,
+      "learning_rate": 3.95945945945946e-06,
+      "loss": 0.6576,
+      "step": 293
+    },
+    {
+      "epoch": 0.023877202956225127,
+      "grad_norm": 7.808087195520982,
+      "learning_rate": 3.972972972972973e-06,
+      "loss": 0.7971,
+      "step": 294
+    },
+    {
+      "epoch": 0.02395841793226671,
+      "grad_norm": 4.140299491048756,
+      "learning_rate": 3.986486486486487e-06,
+      "loss": 0.649,
+      "step": 295
+    },
+    {
+      "epoch": 0.02403963290830829,
+      "grad_norm": 6.598136262319362,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6101,
+      "step": 296
+    },
+    {
+      "epoch": 0.024120847884349875,
+      "grad_norm": 6.657448338847534,
+      "learning_rate": 4.013513513513514e-06,
+      "loss": 0.5807,
+      "step": 297
+    },
+    {
+      "epoch": 0.024202062860391456,
+      "grad_norm": 6.049362925798347,
+      "learning_rate": 4.027027027027028e-06,
+      "loss": 0.614,
+      "step": 298
+    },
+    {
+      "epoch": 0.02428327783643304,
+      "grad_norm": 7.1798282609319415,
+      "learning_rate": 4.040540540540541e-06,
+      "loss": 0.6665,
+      "step": 299
+    },
+    {
+      "epoch": 0.02436449281247462,
+      "grad_norm": 15.049209513289172,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.6214,
+      "step": 300
+    },
+    {
+      "epoch": 0.0244457077885162,
+      "grad_norm": 3.8053242555293796,
+      "learning_rate": 4.067567567567568e-06,
+      "loss": 0.9301,
+      "step": 301
+    },
+    {
+      "epoch": 0.024526922764557785,
+      "grad_norm": 5.676808515237133,
+      "learning_rate": 4.0810810810810815e-06,
+      "loss": 0.7352,
+      "step": 302
+    },
+    {
+      "epoch": 0.024608137740599365,
+      "grad_norm": 3.930706133352666,
+      "learning_rate": 4.0945945945945945e-06,
+      "loss": 0.6721,
+      "step": 303
+    },
+    {
+      "epoch": 0.02468935271664095,
+      "grad_norm": 4.260405802198655,
+      "learning_rate": 4.108108108108108e-06,
+      "loss": 0.6061,
+      "step": 304
+    },
+    {
+      "epoch": 0.02477056769268253,
+      "grad_norm": 3.8039307407950544,
+      "learning_rate": 4.121621621621622e-06,
+      "loss": 0.7034,
+      "step": 305
+    },
+    {
+      "epoch": 0.024851782668724114,
+      "grad_norm": 4.487160345392378,
+      "learning_rate": 4.135135135135135e-06,
+      "loss": 0.703,
+      "step": 306
+    },
+    {
+      "epoch": 0.024932997644765694,
+      "grad_norm": 7.679594903469531,
+      "learning_rate": 4.148648648648649e-06,
+      "loss": 0.6759,
+      "step": 307
+    },
+    {
+      "epoch": 0.025014212620807278,
+      "grad_norm": 3.8651431834889203,
+      "learning_rate": 4.162162162162163e-06,
+      "loss": 0.6642,
+      "step": 308
+    },
+    {
+      "epoch": 0.02509542759684886,
+      "grad_norm": 7.531093080992881,
+      "learning_rate": 4.175675675675676e-06,
+      "loss": 0.6803,
+      "step": 309
+    },
+    {
+      "epoch": 0.025176642572890442,
+      "grad_norm": 5.135386924538244,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 0.9062,
+      "step": 310
+    },
+    {
+      "epoch": 0.025257857548932023,
+      "grad_norm": 6.817605774173451,
+      "learning_rate": 4.202702702702703e-06,
+      "loss": 0.5363,
+      "step": 311
+    },
+    {
+      "epoch": 0.025339072524973603,
+      "grad_norm": 4.999220532436134,
+      "learning_rate": 4.216216216216217e-06,
+      "loss": 0.7114,
+      "step": 312
+    },
+    {
+      "epoch": 0.025420287501015187,
+      "grad_norm": 5.1328146811650495,
+      "learning_rate": 4.22972972972973e-06,
+      "loss": 0.6691,
+      "step": 313
+    },
+    {
+      "epoch": 0.025501502477056768,
+      "grad_norm": 5.421243989150761,
+      "learning_rate": 4.243243243243244e-06,
+      "loss": 0.5742,
+      "step": 314
+    },
+    {
+      "epoch": 0.025582717453098352,
+      "grad_norm": 5.473638240229686,
+      "learning_rate": 4.256756756756757e-06,
+      "loss": 0.6078,
+      "step": 315
+    },
+    {
+      "epoch": 0.025663932429139932,
+      "grad_norm": 4.005836520564742,
+      "learning_rate": 4.270270270270271e-06,
+      "loss": 0.6261,
+      "step": 316
+    },
+    {
+      "epoch": 0.025745147405181516,
+      "grad_norm": 4.474936034323712,
+      "learning_rate": 4.283783783783784e-06,
+      "loss": 0.7862,
+      "step": 317
+    },
+    {
+      "epoch": 0.025826362381223097,
+      "grad_norm": 7.022757009649874,
+      "learning_rate": 4.297297297297298e-06,
+      "loss": 0.7268,
+      "step": 318
+    },
+    {
+      "epoch": 0.02590757735726468,
+      "grad_norm": 4.756298282781381,
+      "learning_rate": 4.310810810810811e-06,
+      "loss": 0.9672,
+      "step": 319
+    },
+    {
+      "epoch": 0.02598879233330626,
+      "grad_norm": 3.7972155322987797,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.7338,
+      "step": 320
+    },
+    {
+      "epoch": 0.026070007309347845,
+      "grad_norm": 6.4304814258504,
+      "learning_rate": 4.3378378378378385e-06,
+      "loss": 0.618,
+      "step": 321
+    },
+    {
+      "epoch": 0.026151222285389426,
+      "grad_norm": 4.985880984448781,
+      "learning_rate": 4.351351351351352e-06,
+      "loss": 0.752,
+      "step": 322
+    },
+    {
+      "epoch": 0.02623243726143101,
+      "grad_norm": 6.965800077960176,
+      "learning_rate": 4.364864864864865e-06,
+      "loss": 0.6961,
+      "step": 323
+    },
+    {
+      "epoch": 0.02631365223747259,
+      "grad_norm": 5.990584472824659,
+      "learning_rate": 4.378378378378379e-06,
+      "loss": 0.68,
+      "step": 324
+    },
+    {
+      "epoch": 0.02639486721351417,
+      "grad_norm": 3.9564290883110647,
+      "learning_rate": 4.391891891891892e-06,
+      "loss": 0.6824,
+      "step": 325
+    },
+    {
+      "epoch": 0.026476082189555755,
+      "grad_norm": 4.712991398333253,
+      "learning_rate": 4.4054054054054054e-06,
+      "loss": 0.7004,
+      "step": 326
+    },
+    {
+      "epoch": 0.026557297165597335,
+      "grad_norm": 4.011893314732997,
+      "learning_rate": 4.418918918918919e-06,
+      "loss": 0.5713,
+      "step": 327
+    },
+    {
+      "epoch": 0.02663851214163892,
+      "grad_norm": 10.789632918330955,
+      "learning_rate": 4.432432432432433e-06,
+      "loss": 0.5751,
+      "step": 328
+    },
+    {
+      "epoch": 0.0267197271176805,
+      "grad_norm": 5.480368608439147,
+      "learning_rate": 4.445945945945946e-06,
+      "loss": 0.621,
+      "step": 329
+    },
+    {
+      "epoch": 0.026800942093722083,
+      "grad_norm": 5.7028472331846,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.6287,
+      "step": 330
+    },
+    {
+      "epoch": 0.026882157069763664,
+      "grad_norm": 4.278443647971252,
+      "learning_rate": 4.472972972972973e-06,
+      "loss": 0.7512,
+      "step": 331
+    },
+    {
+      "epoch": 0.026963372045805248,
+      "grad_norm": 5.540973909463085,
+      "learning_rate": 4.486486486486487e-06,
+      "loss": 0.6958,
+      "step": 332
+    },
+    {
+      "epoch": 0.02704458702184683,
+      "grad_norm": 8.48363442658351,
+      "learning_rate": 4.5e-06,
+      "loss": 0.6622,
+      "step": 333
+    },
+    {
+      "epoch": 0.027125801997888412,
+      "grad_norm": 5.265277541309016,
+      "learning_rate": 4.513513513513514e-06,
+      "loss": 0.7521,
+      "step": 334
+    },
+    {
+      "epoch": 0.027207016973929993,
+      "grad_norm": 6.782051797558551,
+      "learning_rate": 4.527027027027027e-06,
+      "loss": 0.5695,
+      "step": 335
+    },
+    {
+      "epoch": 0.027288231949971573,
+      "grad_norm": 7.707736747999393,
+      "learning_rate": 4.540540540540541e-06,
+      "loss": 0.7251,
+      "step": 336
+    },
+    {
+      "epoch": 0.027369446926013157,
+      "grad_norm": 9.996464962694214,
+      "learning_rate": 4.554054054054055e-06,
+      "loss": 0.7112,
+      "step": 337
+    },
+    {
+      "epoch": 0.027450661902054738,
+      "grad_norm": 5.660779752011064,
+      "learning_rate": 4.567567567567568e-06,
+      "loss": 0.7041,
+      "step": 338
+    },
+    {
+      "epoch": 0.02753187687809632,
+      "grad_norm": 6.58420296764706,
+      "learning_rate": 4.581081081081081e-06,
+      "loss": 0.7357,
+      "step": 339
+    },
+    {
+      "epoch": 0.027613091854137902,
+      "grad_norm": 6.154533828404867,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.8672,
+      "step": 340
+    },
+    {
+      "epoch": 0.027694306830179486,
+      "grad_norm": 3.6916825043347874,
+      "learning_rate": 4.608108108108109e-06,
+      "loss": 0.7116,
+      "step": 341
+    },
+    {
+      "epoch": 0.027775521806221067,
+      "grad_norm": 5.536606059040381,
+      "learning_rate": 4.621621621621622e-06,
+      "loss": 0.5388,
+      "step": 342
+    },
+    {
+      "epoch": 0.02785673678226265,
+      "grad_norm": 5.0131032158949305,
+      "learning_rate": 4.635135135135136e-06,
+      "loss": 0.63,
+      "step": 343
+    },
+    {
+      "epoch": 0.02793795175830423,
+      "grad_norm": 8.645264812499297,
+      "learning_rate": 4.6486486486486495e-06,
+      "loss": 0.6898,
+      "step": 344
+    },
+    {
+      "epoch": 0.028019166734345815,
+      "grad_norm": 8.845007826985789,
+      "learning_rate": 4.6621621621621625e-06,
+      "loss": 0.7366,
+      "step": 345
+    },
+    {
+      "epoch": 0.028100381710387395,
+      "grad_norm": 6.800329318246896,
+      "learning_rate": 4.675675675675676e-06,
+      "loss": 0.6626,
+      "step": 346
+    },
+    {
+      "epoch": 0.028181596686428976,
+      "grad_norm": 3.843752947734245,
+      "learning_rate": 4.6891891891891895e-06,
+      "loss": 0.9352,
+      "step": 347
+    },
+    {
+      "epoch": 0.02826281166247056,
+      "grad_norm": 4.717520289005672,
+      "learning_rate": 4.702702702702703e-06,
+      "loss": 0.558,
+      "step": 348
+    },
+    {
+      "epoch": 0.02834402663851214,
+      "grad_norm": 5.18843657142752,
+      "learning_rate": 4.716216216216216e-06,
+      "loss": 0.6733,
+      "step": 349
+    },
+    {
+      "epoch": 0.028425241614553724,
+      "grad_norm": 6.203969455338302,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 0.7618,
+      "step": 350
+    },
+    {
+      "epoch": 0.028506456590595305,
+      "grad_norm": 5.787025528310779,
+      "learning_rate": 4.743243243243243e-06,
+      "loss": 0.6203,
+      "step": 351
+    },
+    {
+      "epoch": 0.02858767156663689,
+      "grad_norm": 5.74015922758883,
+      "learning_rate": 4.756756756756757e-06,
+      "loss": 0.7899,
+      "step": 352
+    },
+    {
+      "epoch": 0.02866888654267847,
+      "grad_norm": 3.468789863957611,
+      "learning_rate": 4.770270270270271e-06,
+      "loss": 0.5911,
+      "step": 353
+    },
+    {
+      "epoch": 0.028750101518720053,
+      "grad_norm": 4.222539599204078,
+      "learning_rate": 4.783783783783784e-06,
+      "loss": 0.6077,
+      "step": 354
+    },
+    {
+      "epoch": 0.028831316494761634,
+      "grad_norm": 5.987825387101823,
+      "learning_rate": 4.797297297297297e-06,
+      "loss": 0.631,
+      "step": 355
+    },
+    {
+      "epoch": 0.028912531470803218,
+      "grad_norm": 3.494510978483797,
+      "learning_rate": 4.810810810810811e-06,
+      "loss": 0.6175,
+      "step": 356
+    },
+    {
+      "epoch": 0.028993746446844798,
+      "grad_norm": 5.341804933211992,
+      "learning_rate": 4.824324324324325e-06,
+      "loss": 0.8241,
+      "step": 357
+    },
+    {
+      "epoch": 0.02907496142288638,
+      "grad_norm": 5.332752685486192,
+      "learning_rate": 4.837837837837838e-06,
+      "loss": 0.6647,
+      "step": 358
+    },
+    {
+      "epoch": 0.029156176398927963,
+      "grad_norm": 4.174093461478239,
+      "learning_rate": 4.851351351351352e-06,
+      "loss": 0.681,
+      "step": 359
+    },
+    {
+      "epoch": 0.029237391374969543,
+      "grad_norm": 4.796111353710757,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.7803,
+      "step": 360
+    },
+    {
+      "epoch": 0.029318606351011127,
+      "grad_norm": 4.430094054311749,
+      "learning_rate": 4.878378378378379e-06,
+      "loss": 0.5628,
+      "step": 361
+    },
+    {
+      "epoch": 0.029399821327052707,
+      "grad_norm": 4.736436195711088,
+      "learning_rate": 4.891891891891893e-06,
+      "loss": 0.6637,
+      "step": 362
+    },
+    {
+      "epoch": 0.02948103630309429,
+      "grad_norm": 5.52908335708342,
+      "learning_rate": 4.905405405405406e-06,
+      "loss": 0.6167,
+      "step": 363
+    },
+    {
+      "epoch": 0.029562251279135872,
+      "grad_norm": 5.203669203589008,
+      "learning_rate": 4.91891891891892e-06,
+      "loss": 0.7164,
+      "step": 364
+    },
+    {
+      "epoch": 0.029643466255177456,
+      "grad_norm": 5.632857406640382,
+      "learning_rate": 4.932432432432433e-06,
+      "loss": 0.6265,
+      "step": 365
+    },
+    {
+      "epoch": 0.029724681231219036,
+      "grad_norm": 4.525898082221645,
+      "learning_rate": 4.9459459459459466e-06,
+      "loss": 0.6532,
+      "step": 366
+    },
+    {
+      "epoch": 0.02980589620726062,
+      "grad_norm": 3.993279600481055,
+      "learning_rate": 4.95945945945946e-06,
+      "loss": 0.6465,
+      "step": 367
+    },
+    {
+      "epoch": 0.0298871111833022,
+      "grad_norm": 4.976392222325483,
+      "learning_rate": 4.9729729729729735e-06,
+      "loss": 0.6768,
+      "step": 368
+    },
+    {
+      "epoch": 0.029968326159343785,
+      "grad_norm": 7.027194066182947,
+      "learning_rate": 4.986486486486487e-06,
+      "loss": 0.7201,
+      "step": 369
+    },
+    {
+      "epoch": 0.030049541135385365,
+      "grad_norm": 5.6159327874829605,
+      "learning_rate": 5e-06,
+      "loss": 0.6772,
+      "step": 370
+    },
+    {
+      "epoch": 0.030130756111426946,
+      "grad_norm": 4.876745048221957,
+      "learning_rate": 4.999999913506616e-06,
+      "loss": 0.6551,
+      "step": 371
+    },
+    {
+      "epoch": 0.03021197108746853,
+      "grad_norm": 11.157591685761751,
+      "learning_rate": 4.999999654026468e-06,
+      "loss": 0.5551,
+      "step": 372
+    },
+    {
+      "epoch": 0.03029318606351011,
+      "grad_norm": 4.729581532775085,
+      "learning_rate": 4.999999221559576e-06,
+      "loss": 0.7666,
+      "step": 373
+    },
+    {
+      "epoch": 0.030374401039551694,
+      "grad_norm": 5.683222450900519,
+      "learning_rate": 4.9999986161059685e-06,
+      "loss": 0.7107,
+      "step": 374
+    },
+    {
+      "epoch": 0.030455616015593275,
+      "grad_norm": 20.13916455542101,
+      "learning_rate": 4.9999978376656875e-06,
+      "loss": 0.5651,
+      "step": 375
+    },
+    {
+      "epoch": 0.03053683099163486,
+      "grad_norm": 4.663636959572572,
+      "learning_rate": 4.999996886238788e-06,
+      "loss": 0.6166,
+      "step": 376
+    },
+    {
+      "epoch": 0.03061804596767644,
+      "grad_norm": 3.547301470431729,
+      "learning_rate": 4.999995761825335e-06,
+      "loss": 0.6379,
+      "step": 377
+    },
+    {
+      "epoch": 0.030699260943718023,
+      "grad_norm": 4.026761047432411,
+      "learning_rate": 4.999994464425406e-06,
+      "loss": 0.6795,
+      "step": 378
+    },
+    {
+      "epoch": 0.030780475919759603,
+      "grad_norm": 4.177399049954423,
+      "learning_rate": 4.99999299403909e-06,
+      "loss": 0.6356,
+      "step": 379
+    },
+    {
+      "epoch": 0.030861690895801187,
+      "grad_norm": 8.689971440287565,
+      "learning_rate": 4.999991350666491e-06,
+      "loss": 0.5655,
+      "step": 380
+    },
+    {
+      "epoch": 0.030942905871842768,
+      "grad_norm": 4.286085314862286,
+      "learning_rate": 4.999989534307722e-06,
+      "loss": 0.7988,
+      "step": 381
+    },
+    {
+      "epoch": 0.03102412084788435,
+      "grad_norm": 4.468190941115001,
+      "learning_rate": 4.999987544962908e-06,
+      "loss": 0.6375,
+      "step": 382
+    },
+    {
+      "epoch": 0.031105335823925932,
+      "grad_norm": 11.088261247896137,
+      "learning_rate": 4.999985382632186e-06,
+      "loss": 0.6834,
+      "step": 383
+    },
+    {
+      "epoch": 0.031186550799967513,
+      "grad_norm": 4.746884755745409,
+      "learning_rate": 4.9999830473157065e-06,
+      "loss": 0.6825,
+      "step": 384
+    },
+    {
+      "epoch": 0.03126776577600909,
+      "grad_norm": 5.042936323367531,
+      "learning_rate": 4.9999805390136315e-06,
+      "loss": 0.6341,
+      "step": 385
+    },
+    {
+      "epoch": 0.03134898075205068,
+      "grad_norm": 4.6209559647373375,
+      "learning_rate": 4.999977857726135e-06,
+      "loss": 0.5923,
+      "step": 386
+    },
+    {
+      "epoch": 0.03143019572809226,
+      "grad_norm": 5.563156193085517,
+      "learning_rate": 4.999975003453401e-06,
+      "loss": 0.6442,
+      "step": 387
+    },
+    {
+      "epoch": 0.031511410704133845,
+      "grad_norm": 5.076759188588105,
+      "learning_rate": 4.999971976195628e-06,
+      "loss": 0.6856,
+      "step": 388
+    },
+    {
+      "epoch": 0.03159262568017542,
+      "grad_norm": 3.752599670476179,
+      "learning_rate": 4.999968775953025e-06,
+      "loss": 0.7078,
+      "step": 389
+    },
+    {
+      "epoch": 0.031673840656217006,
+      "grad_norm": 4.540276130175966,
+      "learning_rate": 4.999965402725812e-06,
+      "loss": 0.6747,
+      "step": 390
+    },
+    {
+      "epoch": 0.03175505563225859,
+      "grad_norm": 6.85554781706532,
+      "learning_rate": 4.999961856514226e-06,
+      "loss": 0.58,
+      "step": 391
+    },
+    {
+      "epoch": 0.031836270608300174,
+      "grad_norm": 4.470561799528824,
+      "learning_rate": 4.99995813731851e-06,
+      "loss": 0.544,
+      "step": 392
+    },
+    {
+      "epoch": 0.03191748558434175,
+      "grad_norm": 3.381515929810499,
+      "learning_rate": 4.999954245138921e-06,
+      "loss": 0.6551,
+      "step": 393
+    },
+    {
+      "epoch": 0.031998700560383335,
+      "grad_norm": 5.642691704207667,
+      "learning_rate": 4.99995017997573e-06,
+      "loss": 0.7476,
+      "step": 394
+    },
+    {
+      "epoch": 0.03207991553642492,
+      "grad_norm": 4.423437336138073,
+      "learning_rate": 4.999945941829217e-06,
+      "loss": 0.6501,
+      "step": 395
+    },
+    {
+      "epoch": 0.032161130512466496,
+      "grad_norm": 5.470944485898885,
+      "learning_rate": 4.999941530699675e-06,
+      "loss": 0.5093,
+      "step": 396
+    },
+    {
+      "epoch": 0.03224234548850808,
+      "grad_norm": 6.234514653013104,
+      "learning_rate": 4.999936946587412e-06,
+      "loss": 0.5785,
+      "step": 397
+    },
+    {
+      "epoch": 0.032323560464549664,
+      "grad_norm": 4.217022001267068,
+      "learning_rate": 4.999932189492741e-06,
+      "loss": 0.6715,
+      "step": 398
+    },
+    {
+      "epoch": 0.03240477544059125,
+      "grad_norm": 3.997512319922242,
+      "learning_rate": 4.999927259415994e-06,
+      "loss": 0.6798,
+      "step": 399
+    },
+    {
+      "epoch": 0.032485990416632825,
+      "grad_norm": 5.231675871788692,
+      "learning_rate": 4.99992215635751e-06,
+      "loss": 0.581,
+      "step": 400
+    },
+    {
+      "epoch": 0.03256720539267441,
+      "grad_norm": 6.544773650290789,
+      "learning_rate": 4.999916880317645e-06,
+      "loss": 0.608,
+      "step": 401
+    },
+    {
+      "epoch": 0.03264842036871599,
+      "grad_norm": 5.966992739092824,
+      "learning_rate": 4.999911431296762e-06,
+      "loss": 0.6306,
+      "step": 402
+    },
+    {
+      "epoch": 0.03272963534475758,
+      "grad_norm": 9.143165980782102,
+      "learning_rate": 4.999905809295239e-06,
+      "loss": 0.7834,
+      "step": 403
+    },
+    {
+      "epoch": 0.032810850320799154,
+      "grad_norm": 5.104500215594029,
+      "learning_rate": 4.999900014313464e-06,
+      "loss": 0.5444,
+      "step": 404
+    },
+    {
+      "epoch": 0.03289206529684074,
+      "grad_norm": 3.982825484617581,
+      "learning_rate": 4.999894046351839e-06,
+      "loss": 0.6663,
+      "step": 405
+    },
+    {
+      "epoch": 0.03297328027288232,
+      "grad_norm": 9.429533561374097,
+      "learning_rate": 4.999887905410775e-06,
+      "loss": 0.6111,
+      "step": 406
+    },
+    {
+      "epoch": 0.0330544952489239,
+      "grad_norm": 6.206743996051552,
+      "learning_rate": 4.9998815914907e-06,
+      "loss": 0.6354,
+      "step": 407
+    },
+    {
+      "epoch": 0.03313571022496548,
+      "grad_norm": 8.266987795691529,
+      "learning_rate": 4.9998751045920494e-06,
+      "loss": 0.5287,
+      "step": 408
+    },
+    {
+      "epoch": 0.033216925201007066,
+      "grad_norm": 5.87817379773943,
+      "learning_rate": 4.999868444715271e-06,
+      "loss": 0.7346,
+      "step": 409
+    },
+    {
+      "epoch": 0.03329814017704865,
+      "grad_norm": 5.524749698493624,
+      "learning_rate": 4.999861611860827e-06,
+      "loss": 0.702,
+      "step": 410
+    },
+    {
+      "epoch": 0.03337935515309023,
+      "grad_norm": 4.172513273585029,
+      "learning_rate": 4.99985460602919e-06,
+      "loss": 0.6599,
+      "step": 411
+    },
+    {
+      "epoch": 0.03346057012913181,
+      "grad_norm": 4.750265866846706,
+      "learning_rate": 4.9998474272208445e-06,
+      "loss": 0.5904,
+      "step": 412
+    },
+    {
+      "epoch": 0.033541785105173395,
+      "grad_norm": 5.899694142476412,
+      "learning_rate": 4.999840075436286e-06,
+      "loss": 0.5972,
+      "step": 413
+    },
+    {
+      "epoch": 0.03362300008121498,
+      "grad_norm": 8.320548493491238,
+      "learning_rate": 4.999832550676026e-06,
+      "loss": 0.5492,
+      "step": 414
+    },
+    {
+      "epoch": 0.033704215057256556,
+      "grad_norm": 4.509397796756542,
+      "learning_rate": 4.999824852940583e-06,
+      "loss": 0.7786,
+      "step": 415
+    },
+    {
+      "epoch": 0.03378543003329814,
+      "grad_norm": 5.616284004220855,
+      "learning_rate": 4.999816982230491e-06,
+      "loss": 0.7635,
+      "step": 416
+    },
+    {
+      "epoch": 0.033866645009339724,
+      "grad_norm": 5.599669292826977,
+      "learning_rate": 4.999808938546294e-06,
+      "loss": 0.759,
+      "step": 417
+    },
+    {
+      "epoch": 0.0339478599853813,
+      "grad_norm": 6.6642180967818945,
+      "learning_rate": 4.999800721888548e-06,
+      "loss": 0.6626,
+      "step": 418
+    },
+    {
+      "epoch": 0.034029074961422885,
+      "grad_norm": 5.882885604144428,
+      "learning_rate": 4.999792332257822e-06,
+      "loss": 0.8418,
+      "step": 419
+    },
+    {
+      "epoch": 0.03411028993746447,
+      "grad_norm": 4.363979980545484,
+      "learning_rate": 4.999783769654697e-06,
+      "loss": 0.6134,
+      "step": 420
+    },
+    {
+      "epoch": 0.03419150491350605,
+      "grad_norm": 4.776238235905251,
+      "learning_rate": 4.999775034079765e-06,
+      "loss": 0.7331,
+      "step": 421
+    },
+    {
+      "epoch": 0.03427271988954763,
+      "grad_norm": 4.945197802004096,
+      "learning_rate": 4.99976612553363e-06,
+      "loss": 0.692,
+      "step": 422
+    },
+    {
+      "epoch": 0.034353934865589214,
+      "grad_norm": 3.1538233404162628,
+      "learning_rate": 4.999757044016909e-06,
+      "loss": 0.7029,
+      "step": 423
+    },
+    {
+      "epoch": 0.0344351498416308,
+      "grad_norm": 5.660874509416241,
+      "learning_rate": 4.99974778953023e-06,
+      "loss": 0.6652,
+      "step": 424
+    },
+    {
+      "epoch": 0.03451636481767238,
+      "grad_norm": 3.5726277574942857,
+      "learning_rate": 4.9997383620742354e-06,
+      "loss": 0.7189,
+      "step": 425
+    },
+    {
+      "epoch": 0.03459757979371396,
+      "grad_norm": 5.076928873914476,
+      "learning_rate": 4.9997287616495745e-06,
+      "loss": 0.6376,
+      "step": 426
+    },
+    {
+      "epoch": 0.03467879476975554,
+      "grad_norm": 7.805331425608438,
+      "learning_rate": 4.999718988256913e-06,
+      "loss": 0.6691,
+      "step": 427
+    },
+    {
+      "epoch": 0.03476000974579713,
+      "grad_norm": 6.422161317682913,
+      "learning_rate": 4.999709041896927e-06,
+      "loss": 0.6528,
+      "step": 428
+    },
+    {
+      "epoch": 0.034841224721838704,
+      "grad_norm": 4.561480725928831,
+      "learning_rate": 4.9996989225703055e-06,
+      "loss": 0.5714,
+      "step": 429
+    },
+    {
+      "epoch": 0.03492243969788029,
+      "grad_norm": 5.799407616311238,
+      "learning_rate": 4.9996886302777466e-06,
+      "loss": 0.6321,
+      "step": 430
+    },
+    {
+      "epoch": 0.03500365467392187,
+      "grad_norm": 6.86134449223638,
+      "learning_rate": 4.9996781650199655e-06,
+      "loss": 0.6045,
+      "step": 431
+    },
+    {
+      "epoch": 0.035084869649963456,
+      "grad_norm": 4.7834450971379034,
+      "learning_rate": 4.999667526797685e-06,
+      "loss": 0.8719,
+      "step": 432
+    },
+    {
+      "epoch": 0.03516608462600503,
+      "grad_norm": 8.624024638101803,
+      "learning_rate": 4.9996567156116395e-06,
+      "loss": 0.7095,
+      "step": 433
+    },
+    {
+      "epoch": 0.03524729960204662,
+      "grad_norm": 4.366180557531383,
+      "learning_rate": 4.9996457314625794e-06,
+      "loss": 0.5537,
+      "step": 434
+    },
+    {
+      "epoch": 0.0353285145780882,
+      "grad_norm": 5.128282050398408,
+      "learning_rate": 4.9996345743512635e-06,
+      "loss": 0.8241,
+      "step": 435
+    },
+    {
+      "epoch": 0.035409729554129785,
+      "grad_norm": 5.611359253780588,
+      "learning_rate": 4.999623244278464e-06,
+      "loss": 0.6241,
+      "step": 436
+    },
+    {
+      "epoch": 0.03549094453017136,
+      "grad_norm": 14.430464729028463,
+      "learning_rate": 4.999611741244965e-06,
+      "loss": 0.5835,
+      "step": 437
+    },
+    {
+      "epoch": 0.035572159506212946,
+      "grad_norm": 5.4664011456664285,
+      "learning_rate": 4.999600065251563e-06,
+      "loss": 0.6353,
+      "step": 438
+    },
+    {
+      "epoch": 0.03565337448225453,
+      "grad_norm": 3.9622611695813172,
+      "learning_rate": 4.999588216299065e-06,
+      "loss": 0.5976,
+      "step": 439
+    },
+    {
+      "epoch": 0.03573458945829611,
+      "grad_norm": 5.613516770561363,
+      "learning_rate": 4.999576194388292e-06,
+      "loss": 0.7192,
+      "step": 440
+    },
+    {
+      "epoch": 0.03581580443433769,
+      "grad_norm": 4.503094326677596,
+      "learning_rate": 4.999563999520075e-06,
+      "loss": 0.587,
+      "step": 441
+    },
+    {
+      "epoch": 0.035897019410379274,
+      "grad_norm": 4.636439079968228,
+      "learning_rate": 4.999551631695257e-06,
+      "loss": 0.4959,
+      "step": 442
+    },
+    {
+      "epoch": 0.03597823438642086,
+      "grad_norm": 4.842745847770766,
+      "learning_rate": 4.999539090914696e-06,
+      "loss": 0.8282,
+      "step": 443
+    },
+    {
+      "epoch": 0.036059449362462435,
+      "grad_norm": 19.890875208483,
+      "learning_rate": 4.999526377179259e-06,
+      "loss": 0.6707,
+      "step": 444
+    },
+    {
+      "epoch": 0.03614066433850402,
+      "grad_norm": 4.903212810732332,
+      "learning_rate": 4.999513490489824e-06,
+      "loss": 0.5994,
+      "step": 445
+    },
+    {
+      "epoch": 0.0362218793145456,
+      "grad_norm": 4.394920629425536,
+      "learning_rate": 4.999500430847284e-06,
+      "loss": 0.667,
+      "step": 446
+    },
+    {
+      "epoch": 0.03630309429058719,
+      "grad_norm": 3.62470443817531,
+      "learning_rate": 4.9994871982525425e-06,
+      "loss": 0.703,
+      "step": 447
+    },
+    {
+      "epoch": 0.036384309266628764,
+      "grad_norm": 6.417116392361057,
+      "learning_rate": 4.999473792706516e-06,
+      "loss": 0.6643,
+      "step": 448
+    },
+    {
+      "epoch": 0.03646552424267035,
+      "grad_norm": 4.999543742426304,
+      "learning_rate": 4.999460214210131e-06,
+      "loss": 0.7914,
+      "step": 449
+    },
+    {
+      "epoch": 0.03654673921871193,
+      "grad_norm": 5.788280407646206,
+      "learning_rate": 4.999446462764327e-06,
+      "loss": 0.6302,
+      "step": 450
+    },
+    {
+      "epoch": 0.03662795419475351,
+      "grad_norm": 4.2978695329143495,
+      "learning_rate": 4.999432538370057e-06,
+      "loss": 0.6356,
+      "step": 451
+    },
+    {
+      "epoch": 0.03670916917079509,
+      "grad_norm": 4.224970563305071,
+      "learning_rate": 4.999418441028283e-06,
+      "loss": 0.7923,
+      "step": 452
+    },
+    {
+      "epoch": 0.03679038414683668,
+      "grad_norm": 6.438067501026122,
+      "learning_rate": 4.9994041707399794e-06,
+      "loss": 0.4936,
+      "step": 453
+    },
+    {
+      "epoch": 0.03687159912287826,
+      "grad_norm": 4.579554653548933,
+      "learning_rate": 4.999389727506137e-06,
+      "loss": 0.58,
+      "step": 454
+    },
+    {
+      "epoch": 0.03695281409891984,
+      "grad_norm": 4.4684771869350035,
+      "learning_rate": 4.999375111327753e-06,
+      "loss": 0.5475,
+      "step": 455
+    },
+    {
+      "epoch": 0.03703402907496142,
+      "grad_norm": 7.907710639614374,
+      "learning_rate": 4.999360322205838e-06,
+      "loss": 0.6459,
+      "step": 456
+    },
+    {
+      "epoch": 0.037115244051003006,
+      "grad_norm": 5.848886945428208,
+      "learning_rate": 4.999345360141417e-06,
+      "loss": 0.6723,
+      "step": 457
+    },
+    {
+      "epoch": 0.03719645902704459,
+      "grad_norm": 3.055736509730621,
+      "learning_rate": 4.999330225135525e-06,
+      "loss": 0.7079,
+      "step": 458
+    },
+    {
+      "epoch": 0.03727767400308617,
+      "grad_norm": 4.84395863310794,
+      "learning_rate": 4.999314917189209e-06,
+      "loss": 0.8212,
+      "step": 459
+    },
+    {
+      "epoch": 0.03735888897912775,
+      "grad_norm": 4.4914093742426,
+      "learning_rate": 4.999299436303527e-06,
+      "loss": 0.7551,
+      "step": 460
+    },
+    {
+      "epoch": 0.037440103955169335,
+      "grad_norm": 5.747535926175868,
+      "learning_rate": 4.999283782479552e-06,
+      "loss": 0.755,
+      "step": 461
+    },
+    {
+      "epoch": 0.03752131893121091,
+      "grad_norm": 4.124807094235786,
+      "learning_rate": 4.999267955718367e-06,
+      "loss": 0.6151,
+      "step": 462
+    },
+    {
+      "epoch": 0.037602533907252496,
+      "grad_norm": 4.951699574302496,
+      "learning_rate": 4.999251956021066e-06,
+      "loss": 0.6636,
+      "step": 463
+    },
+    {
+      "epoch": 0.03768374888329408,
+      "grad_norm": 3.634905291149469,
+      "learning_rate": 4.999235783388757e-06,
+      "loss": 0.7115,
+      "step": 464
+    },
+    {
+      "epoch": 0.037764963859335664,
+      "grad_norm": 7.068387813096013,
+      "learning_rate": 4.999219437822559e-06,
+      "loss": 0.689,
+      "step": 465
+    },
+    {
+      "epoch": 0.03784617883537724,
+      "grad_norm": 4.404856833972122,
+      "learning_rate": 4.999202919323603e-06,
+      "loss": 0.6214,
+      "step": 466
+    },
+    {
+      "epoch": 0.037927393811418825,
+      "grad_norm": 4.578684856908467,
+      "learning_rate": 4.9991862278930315e-06,
+      "loss": 0.7647,
+      "step": 467
+    },
+    {
+      "epoch": 0.03800860878746041,
+      "grad_norm": 4.898513292964633,
+      "learning_rate": 4.9991693635320005e-06,
+      "loss": 0.6884,
+      "step": 468
+    },
+    {
+      "epoch": 0.03808982376350199,
+      "grad_norm": 4.575368944744301,
+      "learning_rate": 4.999152326241675e-06,
+      "loss": 0.6647,
+      "step": 469
+    },
+    {
+      "epoch": 0.03817103873954357,
+      "grad_norm": 4.238070688134001,
+      "learning_rate": 4.999135116023236e-06,
+      "loss": 0.6667,
+      "step": 470
+    },
+    {
+      "epoch": 0.038252253715585154,
+      "grad_norm": 6.035566032840847,
+      "learning_rate": 4.999117732877873e-06,
+      "loss": 0.6754,
+      "step": 471
+    },
+    {
+      "epoch": 0.03833346869162674,
+      "grad_norm": 3.607010706319493,
+      "learning_rate": 4.9991001768067895e-06,
+      "loss": 0.8115,
+      "step": 472
+    },
+    {
+      "epoch": 0.03841468366766832,
+      "grad_norm": 6.191780527301594,
+      "learning_rate": 4.9990824478112e-06,
+      "loss": 0.6837,
+      "step": 473
+    },
+    {
+      "epoch": 0.0384958986437099,
+      "grad_norm": 5.355617747355118,
+      "learning_rate": 4.999064545892331e-06,
+      "loss": 0.7097,
+      "step": 474
+    },
+    {
+      "epoch": 0.03857711361975148,
+      "grad_norm": 4.233673909331852,
+      "learning_rate": 4.999046471051422e-06,
+      "loss": 0.5915,
+      "step": 475
+    },
+    {
+      "epoch": 0.038658328595793066,
+      "grad_norm": 4.780173065160517,
+      "learning_rate": 4.999028223289724e-06,
+      "loss": 0.9316,
+      "step": 476
+    },
+    {
+      "epoch": 0.03873954357183464,
+      "grad_norm": 4.063084409048435,
+      "learning_rate": 4.999009802608497e-06,
+      "loss": 0.6671,
+      "step": 477
+    },
+    {
+      "epoch": 0.03882075854787623,
+      "grad_norm": 4.851439856688829,
+      "learning_rate": 4.998991209009019e-06,
+      "loss": 0.8034,
+      "step": 478
+    },
+    {
+      "epoch": 0.03890197352391781,
+      "grad_norm": 7.105889513775342,
+      "learning_rate": 4.998972442492575e-06,
+      "loss": 0.532,
+      "step": 479
+    },
+    {
+      "epoch": 0.038983188499959395,
+      "grad_norm": 3.8451222439770683,
+      "learning_rate": 4.9989535030604615e-06,
+      "loss": 0.5993,
+      "step": 480
+    },
+    {
+      "epoch": 0.03906440347600097,
+      "grad_norm": 5.621997119878575,
+      "learning_rate": 4.998934390713994e-06,
+      "loss": 0.5273,
+      "step": 481
+    },
+    {
+      "epoch": 0.039145618452042556,
+      "grad_norm": 4.81307738692358,
+      "learning_rate": 4.9989151054544905e-06,
+      "loss": 0.6318,
+      "step": 482
+    },
+    {
+      "epoch": 0.03922683342808414,
+      "grad_norm": 3.849217538759433,
+      "learning_rate": 4.998895647283287e-06,
+      "loss": 0.6886,
+      "step": 483
+    },
+    {
+      "epoch": 0.039308048404125724,
+      "grad_norm": 3.909888024088152,
+      "learning_rate": 4.99887601620173e-06,
+      "loss": 0.7443,
+      "step": 484
+    },
+    {
+      "epoch": 0.0393892633801673,
+      "grad_norm": 5.4327964670836835,
+      "learning_rate": 4.9988562122111785e-06,
+      "loss": 0.6783,
+      "step": 485
+    },
+    {
+      "epoch": 0.039470478356208885,
+      "grad_norm": 4.60325431470408,
+      "learning_rate": 4.998836235313001e-06,
+      "loss": 0.6589,
+      "step": 486
+    },
+    {
+      "epoch": 0.03955169333225047,
+      "grad_norm": 4.5476339338388945,
+      "learning_rate": 4.998816085508582e-06,
+      "loss": 0.5747,
+      "step": 487
+    },
+    {
+      "epoch": 0.039632908308292046,
+      "grad_norm": 4.970415173334207,
+      "learning_rate": 4.9987957627993145e-06,
+      "loss": 0.6026,
+      "step": 488
+    },
+    {
+      "epoch": 0.03971412328433363,
+      "grad_norm": 4.315065332489597,
+      "learning_rate": 4.998775267186605e-06,
+      "loss": 0.7375,
+      "step": 489
+    },
+    {
+      "epoch": 0.039795338260375214,
+      "grad_norm": 5.5992123809226,
+      "learning_rate": 4.998754598671871e-06,
+      "loss": 0.6427,
+      "step": 490
+    },
+    {
+      "epoch": 0.0398765532364168,
+      "grad_norm": 3.7014622620168405,
+      "learning_rate": 4.998733757256544e-06,
+      "loss": 0.7897,
+      "step": 491
+    },
+    {
+      "epoch": 0.039957768212458375,
+      "grad_norm": 6.7379942751786235,
+      "learning_rate": 4.998712742942065e-06,
+      "loss": 0.6393,
+      "step": 492
+    },
+    {
+      "epoch": 0.04003898318849996,
+      "grad_norm": 4.857719068955753,
+      "learning_rate": 4.998691555729888e-06,
+      "loss": 0.6606,
+      "step": 493
+    },
+    {
+      "epoch": 0.04012019816454154,
+      "grad_norm": 4.0539838554657885,
+      "learning_rate": 4.9986701956214804e-06,
+      "loss": 0.7552,
+      "step": 494
+    },
+    {
+      "epoch": 0.04020141314058313,
+      "grad_norm": 4.957439145405436,
+      "learning_rate": 4.998648662618318e-06,
+      "loss": 0.666,
+      "step": 495
+    },
+    {
+      "epoch": 0.040282628116624704,
+      "grad_norm": 4.953757290265213,
+      "learning_rate": 4.998626956721894e-06,
+      "loss": 0.6508,
+      "step": 496
+    },
+    {
+      "epoch": 0.04036384309266629,
+      "grad_norm": 3.84200026766362,
+      "learning_rate": 4.998605077933706e-06,
+      "loss": 0.6307,
+      "step": 497
+    },
+    {
+      "epoch": 0.04044505806870787,
+      "grad_norm": 6.038380469894045,
+      "learning_rate": 4.998583026255272e-06,
+      "loss": 0.794,
+      "step": 498
+    },
+    {
+      "epoch": 0.04052627304474945,
+      "grad_norm": 5.862309248926957,
+      "learning_rate": 4.998560801688116e-06,
+      "loss": 0.5471,
+      "step": 499
+    },
+    {
+      "epoch": 0.04060748802079103,
+      "grad_norm": 5.376571375980098,
+      "learning_rate": 4.998538404233776e-06,
+      "loss": 0.6685,
+      "step": 500
+    },
+    {
+      "epoch": 0.04068870299683262,
+      "grad_norm": 3.972813169207523,
+      "learning_rate": 4.998515833893801e-06,
+      "loss": 0.5943,
+      "step": 501
+    },
+    {
+      "epoch": 0.0407699179728742,
+      "grad_norm": 5.457428585886831,
+      "learning_rate": 4.998493090669754e-06,
+      "loss": 0.6821,
+      "step": 502
+    },
+    {
+      "epoch": 0.04085113294891578,
+      "grad_norm": 5.526128969258129,
+      "learning_rate": 4.998470174563208e-06,
+      "loss": 0.6268,
+      "step": 503
+    },
+    {
+      "epoch": 0.04093234792495736,
+      "grad_norm": 5.00358557141072,
+      "learning_rate": 4.9984470855757485e-06,
+      "loss": 0.5511,
+      "step": 504
+    },
+    {
+      "epoch": 0.041013562900998946,
+      "grad_norm": 9.228964746118006,
+      "learning_rate": 4.998423823708974e-06,
+      "loss": 0.7192,
+      "step": 505
+    },
+    {
+      "epoch": 0.04109477787704053,
+      "grad_norm": 7.043937824113614,
+      "learning_rate": 4.998400388964494e-06,
+      "loss": 0.6898,
+      "step": 506
+    },
+    {
+      "epoch": 0.041175992853082106,
+      "grad_norm": 7.469052434654404,
+      "learning_rate": 4.998376781343929e-06,
+      "loss": 0.6673,
+      "step": 507
+    },
+    {
+      "epoch": 0.04125720782912369,
+      "grad_norm": 5.819947119265823,
+      "learning_rate": 4.998353000848913e-06,
+      "loss": 0.8352,
+      "step": 508
+    },
+    {
+      "epoch": 0.041338422805165274,
+      "grad_norm": 4.632267764249478,
+      "learning_rate": 4.998329047481093e-06,
+      "loss": 0.6627,
+      "step": 509
+    },
+    {
+      "epoch": 0.04141963778120685,
+      "grad_norm": 5.04622538690542,
+      "learning_rate": 4.998304921242124e-06,
+      "loss": 0.6928,
+      "step": 510
+    },
+    {
+      "epoch": 0.041500852757248435,
+      "grad_norm": 4.929054550494181,
+      "learning_rate": 4.998280622133677e-06,
+      "loss": 0.6864,
+      "step": 511
+    },
+    {
+      "epoch": 0.04158206773329002,
+      "grad_norm": 5.630142199208348,
+      "learning_rate": 4.998256150157433e-06,
+      "loss": 0.5752,
+      "step": 512
+    },
+    {
+      "epoch": 0.0416632827093316,
+      "grad_norm": 4.327018939734424,
+      "learning_rate": 4.998231505315085e-06,
+      "loss": 0.6679,
+      "step": 513
+    },
+    {
+      "epoch": 0.04174449768537318,
+      "grad_norm": 5.207414191020043,
+      "learning_rate": 4.998206687608339e-06,
+      "loss": 0.5235,
+      "step": 514
+    },
+    {
+      "epoch": 0.041825712661414764,
+      "grad_norm": 5.078181936565571,
+      "learning_rate": 4.998181697038912e-06,
+      "loss": 0.5066,
+      "step": 515
+    },
+    {
+      "epoch": 0.04190692763745635,
+      "grad_norm": 4.596975319564036,
+      "learning_rate": 4.998156533608531e-06,
+      "loss": 0.5846,
+      "step": 516
+    },
+    {
+      "epoch": 0.04198814261349793,
+      "grad_norm": 6.0532848525291545,
+      "learning_rate": 4.998131197318942e-06,
+      "loss": 0.5737,
+      "step": 517
+    },
+    {
+      "epoch": 0.04206935758953951,
+      "grad_norm": 4.35820106724424,
+      "learning_rate": 4.998105688171893e-06,
+      "loss": 0.6762,
+      "step": 518
+    },
+    {
+      "epoch": 0.04215057256558109,
+      "grad_norm": 6.664678597912458,
+      "learning_rate": 4.998080006169153e-06,
+      "loss": 0.5852,
+      "step": 519
+    },
+    {
+      "epoch": 0.04223178754162268,
+      "grad_norm": 8.897090441557413,
+      "learning_rate": 4.9980541513124966e-06,
+      "loss": 0.7535,
+      "step": 520
+    },
+    {
+      "epoch": 0.042313002517664254,
+      "grad_norm": 5.576078105545307,
+      "learning_rate": 4.998028123603714e-06,
+      "loss": 0.5836,
+      "step": 521
+    },
+    {
+      "epoch": 0.04239421749370584,
+      "grad_norm": 5.498839425247008,
+      "learning_rate": 4.998001923044605e-06,
+      "loss": 0.638,
+      "step": 522
+    },
+    {
+      "epoch": 0.04247543246974742,
+      "grad_norm": 7.495880701431175,
+      "learning_rate": 4.997975549636985e-06,
+      "loss": 0.6051,
+      "step": 523
+    },
+    {
+      "epoch": 0.042556647445789006,
+      "grad_norm": 8.634014617651163,
+      "learning_rate": 4.997949003382676e-06,
+      "loss": 0.5373,
+      "step": 524
+    },
+    {
+      "epoch": 0.04263786242183058,
+      "grad_norm": 3.6876707339560153,
+      "learning_rate": 4.997922284283517e-06,
+      "loss": 0.8439,
+      "step": 525
+    },
+    {
+      "epoch": 0.04271907739787217,
+      "grad_norm": 4.386768122021177,
+      "learning_rate": 4.997895392341356e-06,
+      "loss": 0.9273,
+      "step": 526
+    },
+    {
+      "epoch": 0.04280029237391375,
+      "grad_norm": 4.887505724579102,
+      "learning_rate": 4.997868327558053e-06,
+      "loss": 1.0235,
+      "step": 527
+    },
+    {
+      "epoch": 0.042881507349955335,
+      "grad_norm": 4.002537001522704,
+      "learning_rate": 4.997841089935482e-06,
+      "loss": 0.7851,
+      "step": 528
+    },
+    {
+      "epoch": 0.04296272232599691,
+      "grad_norm": 7.16699779080668,
+      "learning_rate": 4.997813679475528e-06,
+      "loss": 0.7571,
+      "step": 529
+    },
+    {
+      "epoch": 0.043043937302038496,
+      "grad_norm": 4.775024972507428,
+      "learning_rate": 4.997786096180086e-06,
+      "loss": 0.721,
+      "step": 530
+    },
+    {
+      "epoch": 0.04312515227808008,
+      "grad_norm": 5.099681936395401,
+      "learning_rate": 4.997758340051066e-06,
+      "loss": 0.5692,
+      "step": 531
+    },
+    {
+      "epoch": 0.04320636725412166,
+      "grad_norm": 4.818659941047612,
+      "learning_rate": 4.997730411090387e-06,
+      "loss": 0.643,
+      "step": 532
+    },
+    {
+      "epoch": 0.04328758223016324,
+      "grad_norm": 6.604728440380809,
+      "learning_rate": 4.997702309299983e-06,
+      "loss": 0.7149,
+      "step": 533
+    },
+    {
+      "epoch": 0.043368797206204825,
+      "grad_norm": 5.6047753258513415,
+      "learning_rate": 4.997674034681799e-06,
+      "loss": 0.6094,
+      "step": 534
+    },
+    {
+      "epoch": 0.04345001218224641,
+      "grad_norm": 5.772050070771113,
+      "learning_rate": 4.99764558723779e-06,
+      "loss": 0.5527,
+      "step": 535
+    },
+    {
+      "epoch": 0.043531227158287986,
+      "grad_norm": 6.695215883564865,
+      "learning_rate": 4.997616966969925e-06,
+      "loss": 0.8073,
+      "step": 536
+    },
+    {
+      "epoch": 0.04361244213432957,
+      "grad_norm": 5.0342584361934035,
+      "learning_rate": 4.997588173880184e-06,
+      "loss": 0.7011,
+      "step": 537
+    },
+    {
+      "epoch": 0.043693657110371154,
+      "grad_norm": 4.746592585436567,
+      "learning_rate": 4.99755920797056e-06,
+      "loss": 0.6319,
+      "step": 538
+    },
+    {
+      "epoch": 0.04377487208641274,
+      "grad_norm": 5.2897074099894095,
+      "learning_rate": 4.997530069243057e-06,
+      "loss": 0.5789,
+      "step": 539
+    },
+    {
+      "epoch": 0.043856087062454314,
+      "grad_norm": 3.8007318490045416,
+      "learning_rate": 4.997500757699691e-06,
+      "loss": 0.6953,
+      "step": 540
+    },
+    {
+      "epoch": 0.0439373020384959,
+      "grad_norm": 5.144861861634209,
+      "learning_rate": 4.9974712733424905e-06,
+      "loss": 0.621,
+      "step": 541
+    },
+    {
+      "epoch": 0.04401851701453748,
+      "grad_norm": 5.905282410348868,
+      "learning_rate": 4.997441616173495e-06,
+      "loss": 0.6511,
+      "step": 542
+    },
+    {
+      "epoch": 0.04409973199057906,
+      "grad_norm": 5.309964350881816,
+      "learning_rate": 4.997411786194758e-06,
+      "loss": 0.5526,
+      "step": 543
+    },
+    {
+      "epoch": 0.04418094696662064,
+      "grad_norm": 11.149541991220845,
+      "learning_rate": 4.997381783408343e-06,
+      "loss": 0.6067,
+      "step": 544
+    },
+    {
+      "epoch": 0.04426216194266223,
+      "grad_norm": 5.410844505382304,
+      "learning_rate": 4.9973516078163256e-06,
+      "loss": 0.6418,
+      "step": 545
+    },
+    {
+      "epoch": 0.04434337691870381,
+      "grad_norm": 5.430764606424544,
+      "learning_rate": 4.997321259420793e-06,
+      "loss": 0.6921,
+      "step": 546
+    },
+    {
+      "epoch": 0.04442459189474539,
+      "grad_norm": 11.412874034754271,
+      "learning_rate": 4.997290738223847e-06,
+      "loss": 0.5558,
+      "step": 547
+    },
+    {
+      "epoch": 0.04450580687078697,
+      "grad_norm": 4.133154872685264,
+      "learning_rate": 4.9972600442275985e-06,
+      "loss": 0.6049,
+      "step": 548
+    },
+    {
+      "epoch": 0.044587021846828556,
+      "grad_norm": 5.592148797903198,
+      "learning_rate": 4.997229177434171e-06,
+      "loss": 0.5138,
+      "step": 549
+    },
+    {
+      "epoch": 0.04466823682287014,
+      "grad_norm": 8.487263757967648,
+      "learning_rate": 4.997198137845702e-06,
+      "loss": 0.7193,
+      "step": 550
+    },
+    {
+      "epoch": 0.04474945179891172,
+      "grad_norm": 7.830506522230207,
+      "learning_rate": 4.997166925464337e-06,
+      "loss": 0.7594,
+      "step": 551
+    },
+    {
+      "epoch": 0.0448306667749533,
+      "grad_norm": 8.206186247100188,
+      "learning_rate": 4.997135540292237e-06,
+      "loss": 0.5762,
+      "step": 552
+    },
+    {
+      "epoch": 0.044911881750994885,
+      "grad_norm": 4.437021611163288,
+      "learning_rate": 4.997103982331574e-06,
+      "loss": 0.5914,
+      "step": 553
+    },
+    {
+      "epoch": 0.04499309672703646,
+      "grad_norm": 5.38722516332683,
+      "learning_rate": 4.997072251584531e-06,
+      "loss": 0.5335,
+      "step": 554
+    },
+    {
+      "epoch": 0.045074311703078046,
+      "grad_norm": 7.791956117401669,
+      "learning_rate": 4.997040348053304e-06,
+      "loss": 0.6441,
+      "step": 555
+    },
+    {
+      "epoch": 0.04515552667911963,
+      "grad_norm": 4.66481520418914,
+      "learning_rate": 4.9970082717401e-06,
+      "loss": 0.8322,
+      "step": 556
+    },
+    {
+      "epoch": 0.045236741655161214,
+      "grad_norm": 5.615716066994191,
+      "learning_rate": 4.9969760226471385e-06,
+      "loss": 0.6918,
+      "step": 557
+    },
+    {
+      "epoch": 0.04531795663120279,
+      "grad_norm": 5.424151701577734,
+      "learning_rate": 4.9969436007766514e-06,
+      "loss": 0.5777,
+      "step": 558
+    },
+    {
+      "epoch": 0.045399171607244375,
+      "grad_norm": 4.7203985695060195,
+      "learning_rate": 4.9969110061308826e-06,
+      "loss": 0.7466,
+      "step": 559
+    },
+    {
+      "epoch": 0.04548038658328596,
+      "grad_norm": 4.730334252790068,
+      "learning_rate": 4.996878238712087e-06,
+      "loss": 0.679,
+      "step": 560
+    },
+    {
+      "epoch": 0.04556160155932754,
+      "grad_norm": 3.979712937320014,
+      "learning_rate": 4.996845298522531e-06,
+      "loss": 0.5825,
+      "step": 561
+    },
+    {
+      "epoch": 0.04564281653536912,
+      "grad_norm": 6.730522946218954,
+      "learning_rate": 4.996812185564496e-06,
+      "loss": 0.6467,
+      "step": 562
+    },
+    {
+      "epoch": 0.045724031511410704,
+      "grad_norm": 3.33620867884503,
+      "learning_rate": 4.99677889984027e-06,
+      "loss": 0.7073,
+      "step": 563
+    },
+    {
+      "epoch": 0.04580524648745229,
+      "grad_norm": 5.926130876855334,
+      "learning_rate": 4.996745441352159e-06,
+      "loss": 0.667,
+      "step": 564
+    },
+    {
+      "epoch": 0.04588646146349387,
+      "grad_norm": 5.572153026812714,
+      "learning_rate": 4.996711810102478e-06,
+      "loss": 0.8168,
+      "step": 565
+    },
+    {
+      "epoch": 0.04596767643953545,
+      "grad_norm": 8.174580316906296,
+      "learning_rate": 4.996678006093553e-06,
+      "loss": 0.6412,
+      "step": 566
+    },
+    {
+      "epoch": 0.04604889141557703,
+      "grad_norm": 4.09571754078007,
+      "learning_rate": 4.996644029327723e-06,
+      "loss": 0.6742,
+      "step": 567
+    },
+    {
+      "epoch": 0.04613010639161862,
+      "grad_norm": 4.109112417478182,
+      "learning_rate": 4.996609879807341e-06,
+      "loss": 0.8325,
+      "step": 568
+    },
+    {
+      "epoch": 0.046211321367660194,
+      "grad_norm": 4.752061462760559,
+      "learning_rate": 4.9965755575347665e-06,
+      "loss": 0.6972,
+      "step": 569
+    },
+    {
+      "epoch": 0.04629253634370178,
+      "grad_norm": 3.9521387750727524,
+      "learning_rate": 4.996541062512377e-06,
+      "loss": 0.5081,
+      "step": 570
+    },
+    {
+      "epoch": 0.04637375131974336,
+      "grad_norm": 7.487630136380704,
+      "learning_rate": 4.996506394742559e-06,
+      "loss": 0.5242,
+      "step": 571
+    },
+    {
+      "epoch": 0.046454966295784945,
+      "grad_norm": 5.41182957872003,
+      "learning_rate": 4.996471554227711e-06,
+      "loss": 0.6066,
+      "step": 572
+    },
+    {
+      "epoch": 0.04653618127182652,
+      "grad_norm": 4.448050180267281,
+      "learning_rate": 4.996436540970243e-06,
+      "loss": 0.647,
+      "step": 573
+    },
+    {
+      "epoch": 0.046617396247868106,
+      "grad_norm": 7.285104243616152,
+      "learning_rate": 4.99640135497258e-06,
+      "loss": 0.5929,
+      "step": 574
+    },
+    {
+      "epoch": 0.04669861122390969,
+      "grad_norm": 3.4657873419662413,
+      "learning_rate": 4.996365996237155e-06,
+      "loss": 0.616,
+      "step": 575
+    },
+    {
+      "epoch": 0.046779826199951274,
+      "grad_norm": 6.130204069422916,
+      "learning_rate": 4.996330464766414e-06,
+      "loss": 0.7037,
+      "step": 576
+    },
+    {
+      "epoch": 0.04686104117599285,
+      "grad_norm": 4.045238400737803,
+      "learning_rate": 4.996294760562817e-06,
+      "loss": 0.735,
+      "step": 577
+    },
+    {
+      "epoch": 0.046942256152034435,
+      "grad_norm": 4.541687240117253,
+      "learning_rate": 4.996258883628834e-06,
+      "loss": 0.5942,
+      "step": 578
+    },
+    {
+      "epoch": 0.04702347112807602,
+      "grad_norm": 3.8724612963620326,
+      "learning_rate": 4.996222833966947e-06,
+      "loss": 0.5935,
+      "step": 579
+    },
+    {
+      "epoch": 0.047104686104117596,
+      "grad_norm": 6.548773937549966,
+      "learning_rate": 4.996186611579652e-06,
+      "loss": 0.5016,
+      "step": 580
+    },
+    {
+      "epoch": 0.04718590108015918,
+      "grad_norm": 4.80061928529426,
+      "learning_rate": 4.996150216469454e-06,
+      "loss": 0.7185,
+      "step": 581
+    },
+    {
+      "epoch": 0.047267116056200764,
+      "grad_norm": 3.5697724511020974,
+      "learning_rate": 4.996113648638872e-06,
+      "loss": 0.6544,
+      "step": 582
+    },
+    {
+      "epoch": 0.04734833103224235,
+      "grad_norm": 8.657160084841168,
+      "learning_rate": 4.996076908090435e-06,
+      "loss": 0.712,
+      "step": 583
+    },
+    {
+      "epoch": 0.047429546008283925,
+      "grad_norm": 5.811863937951256,
+      "learning_rate": 4.9960399948266865e-06,
+      "loss": 0.6283,
+      "step": 584
+    },
+    {
+      "epoch": 0.04751076098432551,
+      "grad_norm": 4.144290801020178,
+      "learning_rate": 4.9960029088501814e-06,
+      "loss": 0.5942,
+      "step": 585
+    },
+    {
+      "epoch": 0.04759197596036709,
+      "grad_norm": 11.532211557025384,
+      "learning_rate": 4.995965650163485e-06,
+      "loss": 0.7404,
+      "step": 586
+    },
+    {
+      "epoch": 0.04767319093640868,
+      "grad_norm": 4.498137515882452,
+      "learning_rate": 4.995928218769174e-06,
+      "loss": 0.5951,
+      "step": 587
+    },
+    {
+      "epoch": 0.047754405912450254,
+      "grad_norm": 5.1824043128703465,
+      "learning_rate": 4.99589061466984e-06,
+      "loss": 0.6123,
+      "step": 588
+    },
+    {
+      "epoch": 0.04783562088849184,
+      "grad_norm": 5.787634937924542,
+      "learning_rate": 4.995852837868086e-06,
+      "loss": 0.5376,
+      "step": 589
+    },
+    {
+      "epoch": 0.04791683586453342,
+      "grad_norm": 7.146634422767184,
+      "learning_rate": 4.995814888366523e-06,
+      "loss": 0.599,
+      "step": 590
+    },
+    {
+      "epoch": 0.047998050840575,
+      "grad_norm": 4.672683860091777,
+      "learning_rate": 4.995776766167781e-06,
+      "loss": 0.7119,
+      "step": 591
+    },
+    {
+      "epoch": 0.04807926581661658,
+      "grad_norm": 4.546708179061631,
+      "learning_rate": 4.9957384712744935e-06,
+      "loss": 0.5744,
+      "step": 592
+    },
+    {
+      "epoch": 0.04816048079265817,
+      "grad_norm": 5.406893771451819,
+      "learning_rate": 4.9957000036893124e-06,
+      "loss": 0.6749,
+      "step": 593
+    },
+    {
+      "epoch": 0.04824169576869975,
+      "grad_norm": 4.6310196800989685,
+      "learning_rate": 4.9956613634149e-06,
+      "loss": 0.9589,
+      "step": 594
+    },
+    {
+      "epoch": 0.04832291074474133,
+      "grad_norm": 5.64526302845874,
+      "learning_rate": 4.995622550453929e-06,
+      "loss": 0.7278,
+      "step": 595
+    },
+    {
+      "epoch": 0.04840412572078291,
+      "grad_norm": 7.582504845192103,
+      "learning_rate": 4.995583564809086e-06,
+      "loss": 0.56,
+      "step": 596
+    },
+    {
+      "epoch": 0.048485340696824496,
+      "grad_norm": 4.631867662632177,
+      "learning_rate": 4.995544406483067e-06,
+      "loss": 0.7885,
+      "step": 597
+    },
+    {
+      "epoch": 0.04856655567286608,
+      "grad_norm": 4.646870659330431,
+      "learning_rate": 4.9955050754785835e-06,
+      "loss": 0.7185,
+      "step": 598
+    },
+    {
+      "epoch": 0.04864777064890766,
+      "grad_norm": 4.724376677010658,
+      "learning_rate": 4.995465571798356e-06,
+      "loss": 0.6945,
+      "step": 599
+    },
+    {
+      "epoch": 0.04872898562494924,
+      "grad_norm": 3.7877986447817684,
+      "learning_rate": 4.995425895445118e-06,
+      "loss": 0.7329,
+      "step": 600
+    },
+    {
+      "epoch": 0.048810200600990825,
+      "grad_norm": 4.6352382568877335,
+      "learning_rate": 4.995386046421614e-06,
+      "loss": 0.6972,
+      "step": 601
+    },
+    {
+      "epoch": 0.0488914155770324,
+      "grad_norm": 4.7603716353361305,
+      "learning_rate": 4.9953460247306035e-06,
+      "loss": 0.502,
+      "step": 602
+    },
+    {
+      "epoch": 0.048972630553073986,
+      "grad_norm": 4.752679614641838,
+      "learning_rate": 4.995305830374854e-06,
+      "loss": 0.9242,
+      "step": 603
+    },
+    {
+      "epoch": 0.04905384552911557,
+      "grad_norm": 3.4417838568029855,
+      "learning_rate": 4.995265463357147e-06,
+      "loss": 0.7566,
+      "step": 604
+    },
+    {
+      "epoch": 0.04913506050515715,
+      "grad_norm": 6.888094078901035,
+      "learning_rate": 4.995224923680277e-06,
+      "loss": 0.6201,
+      "step": 605
+    },
+    {
+      "epoch": 0.04921627548119873,
+      "grad_norm": 4.161212593195377,
+      "learning_rate": 4.995184211347046e-06,
+      "loss": 0.7392,
+      "step": 606
+    },
+    {
+      "epoch": 0.049297490457240314,
+      "grad_norm": 7.5719084378043515,
+      "learning_rate": 4.995143326360274e-06,
+      "loss": 0.6514,
+      "step": 607
+    },
+    {
+      "epoch": 0.0493787054332819,
+      "grad_norm": 10.224202349098311,
+      "learning_rate": 4.99510226872279e-06,
+      "loss": 0.5716,
+      "step": 608
+    },
+    {
+      "epoch": 0.04945992040932348,
+      "grad_norm": 5.767252392060593,
+      "learning_rate": 4.995061038437434e-06,
+      "loss": 0.6206,
+      "step": 609
+    },
+    {
+      "epoch": 0.04954113538536506,
+      "grad_norm": 5.065547686279448,
+      "learning_rate": 4.995019635507059e-06,
+      "loss": 0.5774,
+      "step": 610
+    },
+    {
+      "epoch": 0.04962235036140664,
+      "grad_norm": 5.65097481771209,
+      "learning_rate": 4.9949780599345295e-06,
+      "loss": 0.5498,
+      "step": 611
+    },
+    {
+      "epoch": 0.04970356533744823,
+      "grad_norm": 6.523300211900963,
+      "learning_rate": 4.994936311722723e-06,
+      "loss": 0.6711,
+      "step": 612
+    },
+    {
+      "epoch": 0.049784780313489804,
+      "grad_norm": 6.2556246954187715,
+      "learning_rate": 4.994894390874527e-06,
+      "loss": 0.7809,
+      "step": 613
+    },
+    {
+      "epoch": 0.04986599528953139,
+      "grad_norm": 5.630465680189662,
+      "learning_rate": 4.994852297392845e-06,
+      "loss": 0.5909,
+      "step": 614
+    },
+    {
+      "epoch": 0.04994721026557297,
+      "grad_norm": 5.019843596472891,
+      "learning_rate": 4.994810031280587e-06,
+      "loss": 0.8296,
+      "step": 615
+    },
+    {
+      "epoch": 0.050028425241614556,
+      "grad_norm": 6.614977325258197,
+      "learning_rate": 4.994767592540678e-06,
+      "loss": 0.5424,
+      "step": 616
+    },
+    {
+      "epoch": 0.05010964021765613,
+      "grad_norm": 6.173835775193037,
+      "learning_rate": 4.9947249811760555e-06,
+      "loss": 0.6781,
+      "step": 617
+    },
+    {
+      "epoch": 0.05019085519369772,
+      "grad_norm": 7.520930623536426,
+      "learning_rate": 4.994682197189667e-06,
+      "loss": 0.6022,
+      "step": 618
+    },
+    {
+      "epoch": 0.0502720701697393,
+      "grad_norm": 8.187916646181824,
+      "learning_rate": 4.994639240584474e-06,
+      "loss": 0.6612,
+      "step": 619
+    },
+    {
+      "epoch": 0.050353285145780885,
+      "grad_norm": 7.453069836355382,
+      "learning_rate": 4.994596111363448e-06,
+      "loss": 0.7443,
+      "step": 620
+    },
+    {
+      "epoch": 0.05043450012182246,
+      "grad_norm": 7.747733385771382,
+      "learning_rate": 4.994552809529573e-06,
+      "loss": 0.5906,
+      "step": 621
+    },
+    {
+      "epoch": 0.050515715097864046,
+      "grad_norm": 9.01741469170529,
+      "learning_rate": 4.994509335085847e-06,
+      "loss": 0.591,
+      "step": 622
+    },
+    {
+      "epoch": 0.05059693007390563,
+      "grad_norm": 6.323489129605991,
+      "learning_rate": 4.994465688035276e-06,
+      "loss": 0.9078,
+      "step": 623
+    },
+    {
+      "epoch": 0.05067814504994721,
+      "grad_norm": 13.299317733180853,
+      "learning_rate": 4.994421868380881e-06,
+      "loss": 0.525,
+      "step": 624
+    },
+    {
+      "epoch": 0.05075936002598879,
+      "grad_norm": 4.296125414620111,
+      "learning_rate": 4.994377876125695e-06,
+      "loss": 0.5857,
+      "step": 625
+    },
+    {
+      "epoch": 0.050840575002030375,
+      "grad_norm": 3.877492783088978,
+      "learning_rate": 4.994333711272761e-06,
+      "loss": 0.6115,
+      "step": 626
+    },
+    {
+      "epoch": 0.05092178997807196,
+      "grad_norm": 7.011247326243034,
+      "learning_rate": 4.9942893738251355e-06,
+      "loss": 0.5045,
+      "step": 627
+    },
+    {
+      "epoch": 0.051003004954113536,
+      "grad_norm": 7.046298816782971,
+      "learning_rate": 4.994244863785887e-06,
+      "loss": 0.6668,
+      "step": 628
+    },
+    {
+      "epoch": 0.05108421993015512,
+      "grad_norm": 13.568359594461246,
+      "learning_rate": 4.994200181158093e-06,
+      "loss": 0.5775,
+      "step": 629
+    },
+    {
+      "epoch": 0.051165434906196704,
+      "grad_norm": 7.5605209242829545,
+      "learning_rate": 4.9941553259448475e-06,
+      "loss": 0.5511,
+      "step": 630
+    },
+    {
+      "epoch": 0.05124664988223829,
+      "grad_norm": 7.085727266285509,
+      "learning_rate": 4.994110298149253e-06,
+      "loss": 0.6272,
+      "step": 631
+    },
+    {
+      "epoch": 0.051327864858279865,
+      "grad_norm": 4.890560092384863,
+      "learning_rate": 4.994065097774426e-06,
+      "loss": 0.6843,
+      "step": 632
+    },
+    {
+      "epoch": 0.05140907983432145,
+      "grad_norm": 5.665397086701724,
+      "learning_rate": 4.994019724823495e-06,
+      "loss": 0.9322,
+      "step": 633
+    },
+    {
+      "epoch": 0.05149029481036303,
+      "grad_norm": 12.520866191934164,
+      "learning_rate": 4.993974179299597e-06,
+      "loss": 0.9683,
+      "step": 634
+    },
+    {
+      "epoch": 0.05157150978640461,
+      "grad_norm": 6.353771093218972,
+      "learning_rate": 4.993928461205885e-06,
+      "loss": 0.7176,
+      "step": 635
+    },
+    {
+      "epoch": 0.051652724762446194,
+      "grad_norm": 6.398842006345174,
+      "learning_rate": 4.993882570545523e-06,
+      "loss": 0.6833,
+      "step": 636
+    },
+    {
+      "epoch": 0.05173393973848778,
+      "grad_norm": 4.720771899259599,
+      "learning_rate": 4.993836507321686e-06,
+      "loss": 0.554,
+      "step": 637
+    },
+    {
+      "epoch": 0.05181515471452936,
+      "grad_norm": 6.5529107721965945,
+      "learning_rate": 4.9937902715375605e-06,
+      "loss": 0.8592,
+      "step": 638
+    },
+    {
+      "epoch": 0.05189636969057094,
+      "grad_norm": 6.335582258551991,
+      "learning_rate": 4.993743863196348e-06,
+      "loss": 0.6642,
+      "step": 639
+    },
+    {
+      "epoch": 0.05197758466661252,
+      "grad_norm": 5.488427009047834,
+      "learning_rate": 4.993697282301256e-06,
+      "loss": 0.6354,
+      "step": 640
+    },
+    {
+      "epoch": 0.052058799642654106,
+      "grad_norm": 4.435230831471148,
+      "learning_rate": 4.99365052885551e-06,
+      "loss": 0.6361,
+      "step": 641
+    },
+    {
+      "epoch": 0.05214001461869569,
+      "grad_norm": 3.973892603431311,
+      "learning_rate": 4.9936036028623465e-06,
+      "loss": 0.6415,
+      "step": 642
+    },
+    {
+      "epoch": 0.05222122959473727,
+      "grad_norm": 11.239148633823167,
+      "learning_rate": 4.99355650432501e-06,
+      "loss": 0.6895,
+      "step": 643
+    },
+    {
+      "epoch": 0.05230244457077885,
+      "grad_norm": 4.274914346041189,
+      "learning_rate": 4.993509233246761e-06,
+      "loss": 0.5007,
+      "step": 644
+    },
+    {
+      "epoch": 0.052383659546820435,
+      "grad_norm": 6.237193126781422,
+      "learning_rate": 4.9934617896308675e-06,
+      "loss": 0.6882,
+      "step": 645
+    },
+    {
+      "epoch": 0.05246487452286202,
+      "grad_norm": 3.98788426140263,
+      "learning_rate": 4.993414173480617e-06,
+      "loss": 0.6089,
+      "step": 646
+    },
+    {
+      "epoch": 0.052546089498903596,
+      "grad_norm": 14.335138449996226,
+      "learning_rate": 4.9933663847993005e-06,
+      "loss": 0.7351,
+      "step": 647
+    },
+    {
+      "epoch": 0.05262730447494518,
+      "grad_norm": 7.149335425005816,
+      "learning_rate": 4.9933184235902275e-06,
+      "loss": 0.7582,
+      "step": 648
+    },
+    {
+      "epoch": 0.052708519450986764,
+      "grad_norm": 4.593224187186037,
+      "learning_rate": 4.993270289856714e-06,
+      "loss": 0.5349,
+      "step": 649
+    },
+    {
+      "epoch": 0.05278973442702834,
+      "grad_norm": 11.086443268562299,
+      "learning_rate": 4.993221983602093e-06,
+      "loss": 0.5782,
+      "step": 650
+    },
+    {
+      "epoch": 0.052870949403069925,
+      "grad_norm": 4.566408185176231,
+      "learning_rate": 4.993173504829705e-06,
+      "loss": 0.5413,
+      "step": 651
+    },
+    {
+      "epoch": 0.05295216437911151,
+      "grad_norm": 4.3786114129219165,
+      "learning_rate": 4.993124853542906e-06,
+      "loss": 0.7764,
+      "step": 652
+    },
+    {
+      "epoch": 0.05303337935515309,
+      "grad_norm": 4.4667797149749955,
+      "learning_rate": 4.993076029745061e-06,
+      "loss": 0.5315,
+      "step": 653
+    },
+    {
+      "epoch": 0.05311459433119467,
+      "grad_norm": 4.402424708224538,
+      "learning_rate": 4.99302703343955e-06,
+      "loss": 0.6889,
+      "step": 654
+    },
+    {
+      "epoch": 0.053195809307236254,
+      "grad_norm": 5.124489236968013,
+      "learning_rate": 4.992977864629762e-06,
+      "loss": 0.7257,
+      "step": 655
+    },
+    {
+      "epoch": 0.05327702428327784,
+      "grad_norm": 5.821195473769332,
+      "learning_rate": 4.9929285233191005e-06,
+      "loss": 0.6547,
+      "step": 656
+    },
+    {
+      "epoch": 0.05335823925931942,
+      "grad_norm": 6.888571998941501,
+      "learning_rate": 4.992879009510978e-06,
+      "loss": 0.5126,
+      "step": 657
+    },
+    {
+      "epoch": 0.053439454235361,
+      "grad_norm": 5.619280693409271,
+      "learning_rate": 4.992829323208822e-06,
+      "loss": 0.6526,
+      "step": 658
+    },
+    {
+      "epoch": 0.05352066921140258,
+      "grad_norm": 6.967418216632179,
+      "learning_rate": 4.992779464416069e-06,
+      "loss": 0.596,
+      "step": 659
+    },
+    {
+      "epoch": 0.05360188418744417,
+      "grad_norm": 4.9434745741217645,
+      "learning_rate": 4.992729433136171e-06,
+      "loss": 0.6319,
+      "step": 660
+    },
+    {
+      "epoch": 0.053683099163485744,
+      "grad_norm": 8.648403799491454,
+      "learning_rate": 4.992679229372588e-06,
+      "loss": 0.5502,
+      "step": 661
+    },
+    {
+      "epoch": 0.05376431413952733,
+      "grad_norm": 11.185178358108589,
+      "learning_rate": 4.9926288531287946e-06,
+      "loss": 0.6938,
+      "step": 662
+    },
+    {
+      "epoch": 0.05384552911556891,
+      "grad_norm": 5.6934860327185115,
+      "learning_rate": 4.992578304408278e-06,
+      "loss": 0.6992,
+      "step": 663
+    },
+    {
+      "epoch": 0.053926744091610496,
+      "grad_norm": 5.313881451433935,
+      "learning_rate": 4.992527583214533e-06,
+      "loss": 0.738,
+      "step": 664
+    },
+    {
+      "epoch": 0.05400795906765207,
+      "grad_norm": 6.054202350907215,
+      "learning_rate": 4.992476689551071e-06,
+      "loss": 0.6326,
+      "step": 665
+    },
+    {
+      "epoch": 0.05408917404369366,
+      "grad_norm": 7.4133417376434725,
+      "learning_rate": 4.992425623421414e-06,
+      "loss": 0.7014,
+      "step": 666
+    },
+    {
+      "epoch": 0.05417038901973524,
+      "grad_norm": 4.221490444985979,
+      "learning_rate": 4.992374384829094e-06,
+      "loss": 0.7564,
+      "step": 667
+    },
+    {
+      "epoch": 0.054251603995776825,
+      "grad_norm": 10.257715188715267,
+      "learning_rate": 4.992322973777658e-06,
+      "loss": 0.6686,
+      "step": 668
+    },
+    {
+      "epoch": 0.0543328189718184,
+      "grad_norm": 4.701020797968543,
+      "learning_rate": 4.992271390270662e-06,
+      "loss": 0.5668,
+      "step": 669
+    },
+    {
+      "epoch": 0.054414033947859985,
+      "grad_norm": 5.172562987556883,
+      "learning_rate": 4.992219634311677e-06,
+      "loss": 0.563,
+      "step": 670
+    },
+    {
+      "epoch": 0.05449524892390157,
+      "grad_norm": 8.921514201488444,
+      "learning_rate": 4.992167705904282e-06,
+      "loss": 0.694,
+      "step": 671
+    },
+    {
+      "epoch": 0.054576463899943146,
+      "grad_norm": 3.807857859349466,
+      "learning_rate": 4.992115605052072e-06,
+      "loss": 0.5912,
+      "step": 672
+    },
+    {
+      "epoch": 0.05465767887598473,
+      "grad_norm": 3.2898787814816357,
+      "learning_rate": 4.992063331758651e-06,
+      "loss": 0.4573,
+      "step": 673
+    },
+    {
+      "epoch": 0.054738893852026314,
+      "grad_norm": 3.4263761182268526,
+      "learning_rate": 4.9920108860276375e-06,
+      "loss": 0.6525,
+      "step": 674
+    },
+    {
+      "epoch": 0.0548201088280679,
+      "grad_norm": 4.749436044976401,
+      "learning_rate": 4.991958267862659e-06,
+      "loss": 0.51,
+      "step": 675
+    },
+    {
+      "epoch": 0.054901323804109475,
+      "grad_norm": 5.875542464144077,
+      "learning_rate": 4.991905477267356e-06,
+      "loss": 0.6024,
+      "step": 676
+    },
+    {
+      "epoch": 0.05498253878015106,
+      "grad_norm": 5.5252942081115695,
+      "learning_rate": 4.991852514245384e-06,
+      "loss": 0.6975,
+      "step": 677
+    },
+    {
+      "epoch": 0.05506375375619264,
+      "grad_norm": 3.5503049002474336,
+      "learning_rate": 4.991799378800404e-06,
+      "loss": 0.7191,
+      "step": 678
+    },
+    {
+      "epoch": 0.05514496873223423,
+      "grad_norm": 3.9287521299019947,
+      "learning_rate": 4.9917460709360955e-06,
+      "loss": 0.6743,
+      "step": 679
+    },
+    {
+      "epoch": 0.055226183708275804,
+      "grad_norm": 7.458997142187155,
+      "learning_rate": 4.991692590656146e-06,
+      "loss": 0.6277,
+      "step": 680
+    },
+    {
+      "epoch": 0.05530739868431739,
+      "grad_norm": 5.196449898312582,
+      "learning_rate": 4.991638937964257e-06,
+      "loss": 0.5941,
+      "step": 681
+    },
+    {
+      "epoch": 0.05538861366035897,
+      "grad_norm": 7.808673495814518,
+      "learning_rate": 4.9915851128641405e-06,
+      "loss": 0.4626,
+      "step": 682
+    },
+    {
+      "epoch": 0.05546982863640055,
+      "grad_norm": 10.233712958106928,
+      "learning_rate": 4.991531115359519e-06,
+      "loss": 0.6285,
+      "step": 683
+    },
+    {
+      "epoch": 0.05555104361244213,
+      "grad_norm": 4.467794770096275,
+      "learning_rate": 4.991476945454133e-06,
+      "loss": 0.5607,
+      "step": 684
+    },
+    {
+      "epoch": 0.05563225858848372,
+      "grad_norm": 5.521835573749263,
+      "learning_rate": 4.991422603151727e-06,
+      "loss": 0.5919,
+      "step": 685
+    },
+    {
+      "epoch": 0.0557134735645253,
+      "grad_norm": 5.041631495422818,
+      "learning_rate": 4.991368088456062e-06,
+      "loss": 0.5745,
+      "step": 686
+    },
+    {
+      "epoch": 0.05579468854056688,
+      "grad_norm": 4.822878210770882,
+      "learning_rate": 4.99131340137091e-06,
+      "loss": 0.7395,
+      "step": 687
+    },
+    {
+      "epoch": 0.05587590351660846,
+      "grad_norm": 8.016561518905647,
+      "learning_rate": 4.991258541900058e-06,
+      "loss": 0.587,
+      "step": 688
+    },
+    {
+      "epoch": 0.055957118492650046,
+      "grad_norm": 8.66516164876571,
+      "learning_rate": 4.991203510047299e-06,
+      "loss": 0.7209,
+      "step": 689
+    },
+    {
+      "epoch": 0.05603833346869163,
+      "grad_norm": 4.158292161529495,
+      "learning_rate": 4.991148305816441e-06,
+      "loss": 0.5705,
+      "step": 690
+    },
+    {
+      "epoch": 0.05611954844473321,
+      "grad_norm": 8.004317430680866,
+      "learning_rate": 4.991092929211305e-06,
+      "loss": 0.6297,
+      "step": 691
+    },
+    {
+      "epoch": 0.05620076342077479,
+      "grad_norm": 7.148231924748558,
+      "learning_rate": 4.9910373802357214e-06,
+      "loss": 0.5579,
+      "step": 692
+    },
+    {
+      "epoch": 0.056281978396816375,
+      "grad_norm": 4.294720627427241,
+      "learning_rate": 4.990981658893535e-06,
+      "loss": 0.5622,
+      "step": 693
+    },
+    {
+      "epoch": 0.05636319337285795,
+      "grad_norm": 7.230336818000405,
+      "learning_rate": 4.990925765188602e-06,
+      "loss": 0.8874,
+      "step": 694
+    },
+    {
+      "epoch": 0.056444408348899536,
+      "grad_norm": 3.9400202583073662,
+      "learning_rate": 4.9908696991247885e-06,
+      "loss": 0.6678,
+      "step": 695
+    },
+    {
+      "epoch": 0.05652562332494112,
+      "grad_norm": 4.144892106492086,
+      "learning_rate": 4.990813460705975e-06,
+      "loss": 0.6598,
+      "step": 696
+    },
+    {
+      "epoch": 0.056606838300982704,
+      "grad_norm": 5.029300780914533,
+      "learning_rate": 4.990757049936051e-06,
+      "loss": 0.5704,
+      "step": 697
+    },
+    {
+      "epoch": 0.05668805327702428,
+      "grad_norm": 5.440381358402857,
+      "learning_rate": 4.990700466818923e-06,
+      "loss": 0.4935,
+      "step": 698
+    },
+    {
+      "epoch": 0.056769268253065865,
+      "grad_norm": 64.88422852309522,
+      "learning_rate": 4.990643711358504e-06,
+      "loss": 0.6541,
+      "step": 699
+    },
+    {
+      "epoch": 0.05685048322910745,
+      "grad_norm": 9.761223335464411,
+      "learning_rate": 4.990586783558722e-06,
+      "loss": 0.6356,
+      "step": 700
+    },
+    {
+      "epoch": 0.05693169820514903,
+      "grad_norm": 8.471075449596462,
+      "learning_rate": 4.990529683423515e-06,
+      "loss": 0.6032,
+      "step": 701
+    },
+    {
+      "epoch": 0.05701291318119061,
+      "grad_norm": 8.610029295506125,
+      "learning_rate": 4.990472410956835e-06,
+      "loss": 0.6641,
+      "step": 702
+    },
+    {
+      "epoch": 0.057094128157232193,
+      "grad_norm": 8.074345676422041,
+      "learning_rate": 4.9904149661626456e-06,
+      "loss": 0.7349,
+      "step": 703
+    },
+    {
+      "epoch": 0.05717534313327378,
+      "grad_norm": 6.475287058051118,
+      "learning_rate": 4.99035734904492e-06,
+      "loss": 0.601,
+      "step": 704
+    },
+    {
+      "epoch": 0.057256558109315354,
+      "grad_norm": 5.1928926651854335,
+      "learning_rate": 4.990299559607646e-06,
+      "loss": 0.6079,
+      "step": 705
+    },
+    {
+      "epoch": 0.05733777308535694,
+      "grad_norm": 5.520091117650318,
+      "learning_rate": 4.990241597854822e-06,
+      "loss": 0.7949,
+      "step": 706
+    },
+    {
+      "epoch": 0.05741898806139852,
+      "grad_norm": 6.225321905485309,
+      "learning_rate": 4.99018346379046e-06,
+      "loss": 0.5618,
+      "step": 707
+    },
+    {
+      "epoch": 0.057500203037440106,
+      "grad_norm": 4.444351207370442,
+      "learning_rate": 4.99012515741858e-06,
+      "loss": 0.5987,
+      "step": 708
+    },
+    {
+      "epoch": 0.05758141801348168,
+      "grad_norm": 11.383212330047579,
+      "learning_rate": 4.990066678743219e-06,
+      "loss": 0.5409,
+      "step": 709
+    },
+    {
+      "epoch": 0.05766263298952327,
+      "grad_norm": 5.0174724436449765,
+      "learning_rate": 4.9900080277684224e-06,
+      "loss": 0.8077,
+      "step": 710
+    },
+    {
+      "epoch": 0.05774384796556485,
+      "grad_norm": 6.82758331108689,
+      "learning_rate": 4.989949204498248e-06,
+      "loss": 0.7131,
+      "step": 711
+    },
+    {
+      "epoch": 0.057825062941606435,
+      "grad_norm": 4.599805153792023,
+      "learning_rate": 4.989890208936767e-06,
+      "loss": 0.7143,
+      "step": 712
+    },
+    {
+      "epoch": 0.05790627791764801,
+      "grad_norm": 5.615806754693292,
+      "learning_rate": 4.98983104108806e-06,
+      "loss": 0.4989,
+      "step": 713
+    },
+    {
+      "epoch": 0.057987492893689596,
+      "grad_norm": 5.944167262991299,
+      "learning_rate": 4.989771700956223e-06,
+      "loss": 0.6002,
+      "step": 714
+    },
+    {
+      "epoch": 0.05806870786973118,
+      "grad_norm": 5.363075574799061,
+      "learning_rate": 4.989712188545362e-06,
+      "loss": 0.5249,
+      "step": 715
+    },
+    {
+      "epoch": 0.05814992284577276,
+      "grad_norm": 5.71188897749546,
+      "learning_rate": 4.989652503859592e-06,
+      "loss": 0.6282,
+      "step": 716
+    },
+    {
+      "epoch": 0.05823113782181434,
+      "grad_norm": 6.4060919441849835,
+      "learning_rate": 4.989592646903047e-06,
+      "loss": 0.8009,
+      "step": 717
+    },
+    {
+      "epoch": 0.058312352797855925,
+      "grad_norm": 5.806526365359627,
+      "learning_rate": 4.989532617679866e-06,
+      "loss": 0.5443,
+      "step": 718
+    },
+    {
+      "epoch": 0.05839356777389751,
+      "grad_norm": 5.459430108939258,
+      "learning_rate": 4.989472416194204e-06,
+      "loss": 0.5308,
+      "step": 719
+    },
+    {
+      "epoch": 0.058474782749939086,
+      "grad_norm": 5.688152424438068,
+      "learning_rate": 4.9894120424502254e-06,
+      "loss": 0.6284,
+      "step": 720
+    },
+    {
+      "epoch": 0.05855599772598067,
+      "grad_norm": 7.784647717876923,
+      "learning_rate": 4.989351496452109e-06,
+      "loss": 0.5055,
+      "step": 721
+    },
+    {
+      "epoch": 0.058637212702022254,
+      "grad_norm": 4.6787370535165245,
+      "learning_rate": 4.9892907782040435e-06,
+      "loss": 0.6513,
+      "step": 722
+    },
+    {
+      "epoch": 0.05871842767806384,
+      "grad_norm": 6.799832486685082,
+      "learning_rate": 4.9892298877102305e-06,
+      "loss": 0.7293,
+      "step": 723
+    },
+    {
+      "epoch": 0.058799642654105415,
+      "grad_norm": 5.66494166268401,
+      "learning_rate": 4.989168824974884e-06,
+      "loss": 0.5597,
+      "step": 724
+    },
+    {
+      "epoch": 0.058880857630147,
+      "grad_norm": 6.3608187122909206,
+      "learning_rate": 4.989107590002228e-06,
+      "loss": 0.6171,
+      "step": 725
+    },
+    {
+      "epoch": 0.05896207260618858,
+      "grad_norm": 6.134461895108829,
+      "learning_rate": 4.989046182796501e-06,
+      "loss": 0.5414,
+      "step": 726
+    },
+    {
+      "epoch": 0.05904328758223017,
+      "grad_norm": 7.100189240185222,
+      "learning_rate": 4.988984603361949e-06,
+      "loss": 0.6406,
+      "step": 727
+    },
+    {
+      "epoch": 0.059124502558271744,
+      "grad_norm": 5.1992788570507535,
+      "learning_rate": 4.988922851702837e-06,
+      "loss": 0.6249,
+      "step": 728
+    },
+    {
+      "epoch": 0.05920571753431333,
+      "grad_norm": 4.3545253550275635,
+      "learning_rate": 4.988860927823436e-06,
+      "loss": 0.8036,
+      "step": 729
+    },
+    {
+      "epoch": 0.05928693251035491,
+      "grad_norm": 5.3206883020912,
+      "learning_rate": 4.988798831728031e-06,
+      "loss": 0.5943,
+      "step": 730
+    },
+    {
+      "epoch": 0.05936814748639649,
+      "grad_norm": 6.089917627663601,
+      "learning_rate": 4.9887365634209186e-06,
+      "loss": 0.7094,
+      "step": 731
+    },
+    {
+      "epoch": 0.05944936246243807,
+      "grad_norm": 3.956275748674879,
+      "learning_rate": 4.9886741229064075e-06,
+      "loss": 0.5626,
+      "step": 732
+    },
+    {
+      "epoch": 0.05953057743847966,
+      "grad_norm": 4.511627354156277,
+      "learning_rate": 4.988611510188818e-06,
+      "loss": 0.5764,
+      "step": 733
+    },
+    {
+      "epoch": 0.05961179241452124,
+      "grad_norm": 6.4116489379834025,
+      "learning_rate": 4.988548725272482e-06,
+      "loss": 0.5982,
+      "step": 734
+    },
+    {
+      "epoch": 0.05969300739056282,
+      "grad_norm": 6.311885375734862,
+      "learning_rate": 4.988485768161746e-06,
+      "loss": 0.5339,
+      "step": 735
+    },
+    {
+      "epoch": 0.0597742223666044,
+      "grad_norm": 3.9308120481243902,
+      "learning_rate": 4.988422638860964e-06,
+      "loss": 0.584,
+      "step": 736
+    },
+    {
+      "epoch": 0.059855437342645985,
+      "grad_norm": 4.880711425301184,
+      "learning_rate": 4.988359337374505e-06,
+      "loss": 0.5078,
+      "step": 737
+    },
+    {
+      "epoch": 0.05993665231868757,
+      "grad_norm": 5.281870836321762,
+      "learning_rate": 4.988295863706751e-06,
+      "loss": 0.5754,
+      "step": 738
+    },
+    {
+      "epoch": 0.060017867294729146,
+      "grad_norm": 5.0230810112537485,
+      "learning_rate": 4.988232217862091e-06,
+      "loss": 0.6391,
+      "step": 739
+    },
+    {
+      "epoch": 0.06009908227077073,
+      "grad_norm": 6.955537743254657,
+      "learning_rate": 4.988168399844931e-06,
+      "loss": 0.5283,
+      "step": 740
+    },
+    {
+      "epoch": 0.060180297246812314,
+      "grad_norm": 4.147704339162062,
+      "learning_rate": 4.988104409659685e-06,
+      "loss": 0.653,
+      "step": 741
+    },
+    {
+      "epoch": 0.06026151222285389,
+      "grad_norm": 7.537477012145328,
+      "learning_rate": 4.988040247310783e-06,
+      "loss": 0.6525,
+      "step": 742
+    },
+    {
+      "epoch": 0.060342727198895475,
+      "grad_norm": 4.863974888473347,
+      "learning_rate": 4.987975912802663e-06,
+      "loss": 0.5931,
+      "step": 743
+    },
+    {
+      "epoch": 0.06042394217493706,
+      "grad_norm": 5.366327422788624,
+      "learning_rate": 4.9879114061397784e-06,
+      "loss": 0.4838,
+      "step": 744
+    },
+    {
+      "epoch": 0.06050515715097864,
+      "grad_norm": 4.656417177398882,
+      "learning_rate": 4.987846727326591e-06,
+      "loss": 0.681,
+      "step": 745
+    },
+    {
+      "epoch": 0.06058637212702022,
+      "grad_norm": 20.788942947113796,
+      "learning_rate": 4.987781876367576e-06,
+      "loss": 0.6331,
+      "step": 746
+    },
+    {
+      "epoch": 0.060667587103061804,
+      "grad_norm": 7.459258545982763,
+      "learning_rate": 4.987716853267222e-06,
+      "loss": 0.5166,
+      "step": 747
+    },
+    {
+      "epoch": 0.06074880207910339,
+      "grad_norm": 7.51821304798715,
+      "learning_rate": 4.9876516580300285e-06,
+      "loss": 0.4729,
+      "step": 748
+    },
+    {
+      "epoch": 0.06083001705514497,
+      "grad_norm": 6.15090790030722,
+      "learning_rate": 4.987586290660506e-06,
+      "loss": 0.5615,
+      "step": 749
+    },
+    {
+      "epoch": 0.06091123203118655,
+      "grad_norm": 5.226446559415324,
+      "learning_rate": 4.987520751163176e-06,
+      "loss": 0.4638,
+      "step": 750
+    },
+    {
+      "epoch": 0.06099244700722813,
+      "grad_norm": 6.752685506869791,
+      "learning_rate": 4.9874550395425764e-06,
+      "loss": 0.5201,
+      "step": 751
+    },
+    {
+      "epoch": 0.06107366198326972,
+      "grad_norm": 4.989408238882795,
+      "learning_rate": 4.987389155803252e-06,
+      "loss": 0.7494,
+      "step": 752
+    },
+    {
+      "epoch": 0.061154876959311294,
+      "grad_norm": 6.750783797733722,
+      "learning_rate": 4.987323099949763e-06,
+      "loss": 0.65,
+      "step": 753
+    },
+    {
+      "epoch": 0.06123609193535288,
+      "grad_norm": 7.609585837462416,
+      "learning_rate": 4.9872568719866795e-06,
+      "loss": 0.6736,
+      "step": 754
+    },
+    {
+      "epoch": 0.06131730691139446,
+      "grad_norm": 4.452864412048146,
+      "learning_rate": 4.987190471918584e-06,
+      "loss": 0.5907,
+      "step": 755
+    },
+    {
+      "epoch": 0.061398521887436046,
+      "grad_norm": 4.734835124703949,
+      "learning_rate": 4.98712389975007e-06,
+      "loss": 0.653,
+      "step": 756
+    },
+    {
+      "epoch": 0.06147973686347762,
+      "grad_norm": 5.928107826920242,
+      "learning_rate": 4.987057155485746e-06,
+      "loss": 0.7772,
+      "step": 757
+    },
+    {
+      "epoch": 0.06156095183951921,
+      "grad_norm": 9.884828624607355,
+      "learning_rate": 4.98699023913023e-06,
+      "loss": 0.5529,
+      "step": 758
+    },
+    {
+      "epoch": 0.06164216681556079,
+      "grad_norm": 4.2764549097941,
+      "learning_rate": 4.986923150688151e-06,
+      "loss": 0.6011,
+      "step": 759
+    },
+    {
+      "epoch": 0.061723381791602375,
+      "grad_norm": 7.233144313683277,
+      "learning_rate": 4.986855890164152e-06,
+      "loss": 0.6427,
+      "step": 760
+    },
+    {
+      "epoch": 0.06180459676764395,
+      "grad_norm": 5.536503055820394,
+      "learning_rate": 4.986788457562887e-06,
+      "loss": 0.6915,
+      "step": 761
+    },
+    {
+      "epoch": 0.061885811743685536,
+      "grad_norm": 5.309270582449383,
+      "learning_rate": 4.986720852889021e-06,
+      "loss": 0.5991,
+      "step": 762
+    },
+    {
+      "epoch": 0.06196702671972712,
+      "grad_norm": 4.284356637879479,
+      "learning_rate": 4.9866530761472335e-06,
+      "loss": 0.8062,
+      "step": 763
+    },
+    {
+      "epoch": 0.0620482416957687,
+      "grad_norm": 5.643793776418088,
+      "learning_rate": 4.986585127342214e-06,
+      "loss": 0.7215,
+      "step": 764
+    },
+    {
+      "epoch": 0.06212945667181028,
+      "grad_norm": 4.295093117802471,
+      "learning_rate": 4.986517006478663e-06,
+      "loss": 0.4926,
+      "step": 765
+    },
+    {
+      "epoch": 0.062210671647851865,
+      "grad_norm": 5.881778143353826,
+      "learning_rate": 4.986448713561295e-06,
+      "loss": 0.5777,
+      "step": 766
+    },
+    {
+      "epoch": 0.06229188662389345,
+      "grad_norm": 3.5376491877561533,
+      "learning_rate": 4.986380248594835e-06,
+      "loss": 0.8141,
+      "step": 767
+    },
+    {
+      "epoch": 0.062373101599935026,
+      "grad_norm": 15.197506269463709,
+      "learning_rate": 4.9863116115840215e-06,
+      "loss": 0.6347,
+      "step": 768
+    },
+    {
+      "epoch": 0.06245431657597661,
+      "grad_norm": 5.5635913852411765,
+      "learning_rate": 4.986242802533603e-06,
+      "loss": 0.7381,
+      "step": 769
+    },
+    {
+      "epoch": 0.06253553155201819,
+      "grad_norm": 4.921640572698924,
+      "learning_rate": 4.986173821448341e-06,
+      "loss": 0.5331,
+      "step": 770
+    },
+    {
+      "epoch": 0.06261674652805978,
+      "grad_norm": 4.243356448186626,
+      "learning_rate": 4.9861046683330085e-06,
+      "loss": 0.7414,
+      "step": 771
+    },
+    {
+      "epoch": 0.06269796150410135,
+      "grad_norm": 4.7676286369533685,
+      "learning_rate": 4.986035343192389e-06,
+      "loss": 0.7152,
+      "step": 772
+    },
+    {
+      "epoch": 0.06277917648014295,
+      "grad_norm": 5.052963380949943,
+      "learning_rate": 4.985965846031283e-06,
+      "loss": 0.5741,
+      "step": 773
+    },
+    {
+      "epoch": 0.06286039145618452,
+      "grad_norm": 5.830703109939469,
+      "learning_rate": 4.985896176854496e-06,
+      "loss": 0.5442,
+      "step": 774
+    },
+    {
+      "epoch": 0.0629416064322261,
+      "grad_norm": 28.565367004080375,
+      "learning_rate": 4.9858263356668505e-06,
+      "loss": 0.6674,
+      "step": 775
+    },
+    {
+      "epoch": 0.06302282140826769,
+      "grad_norm": 5.2243621774870475,
+      "learning_rate": 4.985756322473178e-06,
+      "loss": 0.5452,
+      "step": 776
+    },
+    {
+      "epoch": 0.06310403638430927,
+      "grad_norm": 5.380977330836419,
+      "learning_rate": 4.9856861372783236e-06,
+      "loss": 0.5991,
+      "step": 777
+    },
+    {
+      "epoch": 0.06318525136035084,
+      "grad_norm": 6.151320235729201,
+      "learning_rate": 4.9856157800871455e-06,
+      "loss": 0.661,
+      "step": 778
+    },
+    {
+      "epoch": 0.06326646633639244,
+      "grad_norm": 4.520649099692102,
+      "learning_rate": 4.985545250904509e-06,
+      "loss": 0.5931,
+      "step": 779
+    },
+    {
+      "epoch": 0.06334768131243401,
+      "grad_norm": 3.8505363861712336,
+      "learning_rate": 4.985474549735296e-06,
+      "loss": 0.6474,
+      "step": 780
+    },
+    {
+      "epoch": 0.06342889628847559,
+      "grad_norm": 4.383396583650249,
+      "learning_rate": 4.985403676584397e-06,
+      "loss": 0.7205,
+      "step": 781
+    },
+    {
+      "epoch": 0.06351011126451718,
+      "grad_norm": 5.432225271309565,
+      "learning_rate": 4.985332631456719e-06,
+      "loss": 0.6348,
+      "step": 782
+    },
+    {
+      "epoch": 0.06359132624055876,
+      "grad_norm": 8.317084663016479,
+      "learning_rate": 4.9852614143571755e-06,
+      "loss": 0.6672,
+      "step": 783
+    },
+    {
+      "epoch": 0.06367254121660035,
+      "grad_norm": 6.050699233785058,
+      "learning_rate": 4.985190025290696e-06,
+      "loss": 0.5816,
+      "step": 784
+    },
+    {
+      "epoch": 0.06375375619264192,
+      "grad_norm": 4.054546774991104,
+      "learning_rate": 4.985118464262219e-06,
+      "loss": 0.5229,
+      "step": 785
+    },
+    {
+      "epoch": 0.0638349711686835,
+      "grad_norm": 6.331916697325334,
+      "learning_rate": 4.985046731276697e-06,
+      "loss": 0.6044,
+      "step": 786
+    },
+    {
+      "epoch": 0.06391618614472509,
+      "grad_norm": 5.448136315607667,
+      "learning_rate": 4.984974826339093e-06,
+      "loss": 0.6758,
+      "step": 787
+    },
+    {
+      "epoch": 0.06399740112076667,
+      "grad_norm": 3.0701662409934065,
+      "learning_rate": 4.984902749454382e-06,
+      "loss": 0.5999,
+      "step": 788
+    },
+    {
+      "epoch": 0.06407861609680825,
+      "grad_norm": 4.6802757341120556,
+      "learning_rate": 4.9848305006275525e-06,
+      "loss": 0.6773,
+      "step": 789
+    },
+    {
+      "epoch": 0.06415983107284984,
+      "grad_norm": 5.59540917186297,
+      "learning_rate": 4.984758079863603e-06,
+      "loss": 0.5526,
+      "step": 790
+    },
+    {
+      "epoch": 0.06424104604889141,
+      "grad_norm": 4.319537345372907,
+      "learning_rate": 4.984685487167544e-06,
+      "loss": 0.4746,
+      "step": 791
+    },
+    {
+      "epoch": 0.06432226102493299,
+      "grad_norm": 4.812410544795939,
+      "learning_rate": 4.9846127225444e-06,
+      "loss": 0.5985,
+      "step": 792
+    },
+    {
+      "epoch": 0.06440347600097458,
+      "grad_norm": 3.8747262789638,
+      "learning_rate": 4.984539785999205e-06,
+      "loss": 0.8711,
+      "step": 793
+    },
+    {
+      "epoch": 0.06448469097701616,
+      "grad_norm": 9.168257138239873,
+      "learning_rate": 4.984466677537007e-06,
+      "loss": 0.6323,
+      "step": 794
+    },
+    {
+      "epoch": 0.06456590595305775,
+      "grad_norm": 9.621282105988284,
+      "learning_rate": 4.984393397162862e-06,
+      "loss": 0.6117,
+      "step": 795
+    },
+    {
+      "epoch": 0.06464712092909933,
+      "grad_norm": 4.3783026160070895,
+      "learning_rate": 4.984319944881844e-06,
+      "loss": 0.6991,
+      "step": 796
+    },
+    {
+      "epoch": 0.0647283359051409,
+      "grad_norm": 4.35590663167429,
+      "learning_rate": 4.984246320699033e-06,
+      "loss": 0.7034,
+      "step": 797
+    },
+    {
+      "epoch": 0.0648095508811825,
+      "grad_norm": 4.746827833135994,
+      "learning_rate": 4.984172524619525e-06,
+      "loss": 0.6776,
+      "step": 798
+    },
+    {
+      "epoch": 0.06489076585722407,
+      "grad_norm": 5.905725252167514,
+      "learning_rate": 4.984098556648425e-06,
+      "loss": 0.5483,
+      "step": 799
+    },
+    {
+      "epoch": 0.06497198083326565,
+      "grad_norm": 4.596246589585779,
+      "learning_rate": 4.984024416790852e-06,
+      "loss": 0.688,
+      "step": 800
+    },
+    {
+      "epoch": 0.06505319580930724,
+      "grad_norm": 4.9534579923413595,
+      "learning_rate": 4.983950105051936e-06,
+      "loss": 0.6562,
+      "step": 801
+    },
+    {
+      "epoch": 0.06513441078534882,
+      "grad_norm": 5.123985851109989,
+      "learning_rate": 4.9838756214368185e-06,
+      "loss": 0.6707,
+      "step": 802
+    },
+    {
+      "epoch": 0.0652156257613904,
+      "grad_norm": 5.402799215185337,
+      "learning_rate": 4.9838009659506535e-06,
+      "loss": 0.6378,
+      "step": 803
+    },
+    {
+      "epoch": 0.06529684073743199,
+      "grad_norm": 4.492757760289716,
+      "learning_rate": 4.983726138598608e-06,
+      "loss": 0.5916,
+      "step": 804
+    },
+    {
+      "epoch": 0.06537805571347356,
+      "grad_norm": 3.95999131962753,
+      "learning_rate": 4.9836511393858575e-06,
+      "loss": 0.5023,
+      "step": 805
+    },
+    {
+      "epoch": 0.06545927068951515,
+      "grad_norm": 3.920207373866997,
+      "learning_rate": 4.983575968317593e-06,
+      "loss": 0.7991,
+      "step": 806
+    },
+    {
+      "epoch": 0.06554048566555673,
+      "grad_norm": 7.315668961691789,
+      "learning_rate": 4.983500625399017e-06,
+      "loss": 0.5301,
+      "step": 807
+    },
+    {
+      "epoch": 0.06562170064159831,
+      "grad_norm": 3.912121127147346,
+      "learning_rate": 4.98342511063534e-06,
+      "loss": 0.7372,
+      "step": 808
+    },
+    {
+      "epoch": 0.0657029156176399,
+      "grad_norm": 7.552130637632774,
+      "learning_rate": 4.983349424031789e-06,
+      "loss": 0.5946,
+      "step": 809
+    },
+    {
+      "epoch": 0.06578413059368148,
+      "grad_norm": 6.533691238009157,
+      "learning_rate": 4.983273565593601e-06,
+      "loss": 0.642,
+      "step": 810
+    },
+    {
+      "epoch": 0.06586534556972305,
+      "grad_norm": 9.916532299872616,
+      "learning_rate": 4.983197535326024e-06,
+      "loss": 0.6117,
+      "step": 811
+    },
+    {
+      "epoch": 0.06594656054576464,
+      "grad_norm": 5.29822144361094,
+      "learning_rate": 4.983121333234321e-06,
+      "loss": 0.576,
+      "step": 812
+    },
+    {
+      "epoch": 0.06602777552180622,
+      "grad_norm": 7.011546578891193,
+      "learning_rate": 4.983044959323763e-06,
+      "loss": 0.6305,
+      "step": 813
+    },
+    {
+      "epoch": 0.0661089904978478,
+      "grad_norm": 7.6729419826090695,
+      "learning_rate": 4.982968413599635e-06,
+      "loss": 0.5282,
+      "step": 814
+    },
+    {
+      "epoch": 0.06619020547388939,
+      "grad_norm": 3.568801069591388,
+      "learning_rate": 4.982891696067234e-06,
+      "loss": 0.5766,
+      "step": 815
+    },
+    {
+      "epoch": 0.06627142044993097,
+      "grad_norm": 3.822836645624077,
+      "learning_rate": 4.9828148067318675e-06,
+      "loss": 0.647,
+      "step": 816
+    },
+    {
+      "epoch": 0.06635263542597256,
+      "grad_norm": 4.578694026237904,
+      "learning_rate": 4.982737745598857e-06,
+      "loss": 0.8134,
+      "step": 817
+    },
+    {
+      "epoch": 0.06643385040201413,
+      "grad_norm": 6.7533794019944065,
+      "learning_rate": 4.982660512673534e-06,
+      "loss": 0.6404,
+      "step": 818
+    },
+    {
+      "epoch": 0.06651506537805571,
+      "grad_norm": 5.791281720391061,
+      "learning_rate": 4.982583107961243e-06,
+      "loss": 0.5909,
+      "step": 819
+    },
+    {
+      "epoch": 0.0665962803540973,
+      "grad_norm": 4.161592546084501,
+      "learning_rate": 4.982505531467339e-06,
+      "loss": 0.5977,
+      "step": 820
+    },
+    {
+      "epoch": 0.06667749533013888,
+      "grad_norm": 6.394531347329611,
+      "learning_rate": 4.982427783197191e-06,
+      "loss": 0.5928,
+      "step": 821
+    },
+    {
+      "epoch": 0.06675871030618045,
+      "grad_norm": 8.488158051535715,
+      "learning_rate": 4.982349863156179e-06,
+      "loss": 0.7508,
+      "step": 822
+    },
+    {
+      "epoch": 0.06683992528222205,
+      "grad_norm": 5.428324952116931,
+      "learning_rate": 4.982271771349694e-06,
+      "loss": 0.5392,
+      "step": 823
+    },
+    {
+      "epoch": 0.06692114025826362,
+      "grad_norm": 4.025855738456331,
+      "learning_rate": 4.98219350778314e-06,
+      "loss": 0.4991,
+      "step": 824
+    },
+    {
+      "epoch": 0.0670023552343052,
+      "grad_norm": 5.231237358948681,
+      "learning_rate": 4.982115072461932e-06,
+      "loss": 0.522,
+      "step": 825
+    },
+    {
+      "epoch": 0.06708357021034679,
+      "grad_norm": 3.4702942997245616,
+      "learning_rate": 4.9820364653914964e-06,
+      "loss": 0.6092,
+      "step": 826
+    },
+    {
+      "epoch": 0.06716478518638837,
+      "grad_norm": 8.455649789106133,
+      "learning_rate": 4.981957686577275e-06,
+      "loss": 0.8485,
+      "step": 827
+    },
+    {
+      "epoch": 0.06724600016242996,
+      "grad_norm": 6.0384535186038955,
+      "learning_rate": 4.981878736024716e-06,
+      "loss": 0.5669,
+      "step": 828
+    },
+    {
+      "epoch": 0.06732721513847154,
+      "grad_norm": 3.4751635293718732,
+      "learning_rate": 4.981799613739284e-06,
+      "loss": 0.5404,
+      "step": 829
+    },
+    {
+      "epoch": 0.06740843011451311,
+      "grad_norm": 4.516175343460089,
+      "learning_rate": 4.981720319726453e-06,
+      "loss": 0.512,
+      "step": 830
+    },
+    {
+      "epoch": 0.0674896450905547,
+      "grad_norm": 6.121636308285373,
+      "learning_rate": 4.981640853991712e-06,
+      "loss": 0.7,
+      "step": 831
+    },
+    {
+      "epoch": 0.06757086006659628,
+      "grad_norm": 4.916730473198547,
+      "learning_rate": 4.981561216540556e-06,
+      "loss": 0.6203,
+      "step": 832
+    },
+    {
+      "epoch": 0.06765207504263786,
+      "grad_norm": 6.140736960651221,
+      "learning_rate": 4.981481407378498e-06,
+      "loss": 0.6689,
+      "step": 833
+    },
+    {
+      "epoch": 0.06773329001867945,
+      "grad_norm": 3.651255962877865,
+      "learning_rate": 4.981401426511059e-06,
+      "loss": 0.6711,
+      "step": 834
+    },
+    {
+      "epoch": 0.06781450499472103,
+      "grad_norm": 4.433646055135197,
+      "learning_rate": 4.981321273943775e-06,
+      "loss": 0.5962,
+      "step": 835
+    },
+    {
+      "epoch": 0.0678957199707626,
+      "grad_norm": 4.214671257503501,
+      "learning_rate": 4.98124094968219e-06,
+      "loss": 0.6696,
+      "step": 836
+    },
+    {
+      "epoch": 0.0679769349468042,
+      "grad_norm": 8.938425585700115,
+      "learning_rate": 4.981160453731864e-06,
+      "loss": 0.5597,
+      "step": 837
+    },
+    {
+      "epoch": 0.06805814992284577,
+      "grad_norm": 5.5327509204918215,
+      "learning_rate": 4.981079786098365e-06,
+      "loss": 0.6733,
+      "step": 838
+    },
+    {
+      "epoch": 0.06813936489888736,
+      "grad_norm": 4.203367930591512,
+      "learning_rate": 4.980998946787276e-06,
+      "loss": 0.6717,
+      "step": 839
+    },
+    {
+      "epoch": 0.06822057987492894,
+      "grad_norm": 4.743933277670276,
+      "learning_rate": 4.98091793580419e-06,
+      "loss": 0.6467,
+      "step": 840
+    },
+    {
+      "epoch": 0.06830179485097052,
+      "grad_norm": 5.519327116976564,
+      "learning_rate": 4.9808367531547144e-06,
+      "loss": 0.5875,
+      "step": 841
+    },
+    {
+      "epoch": 0.0683830098270121,
+      "grad_norm": 5.421051052782078,
+      "learning_rate": 4.980755398844464e-06,
+      "loss": 0.5742,
+      "step": 842
+    },
+    {
+      "epoch": 0.06846422480305368,
+      "grad_norm": 5.205786851548457,
+      "learning_rate": 4.980673872879069e-06,
+      "loss": 0.734,
+      "step": 843
+    },
+    {
+      "epoch": 0.06854543977909526,
+      "grad_norm": 7.795985083428892,
+      "learning_rate": 4.980592175264172e-06,
+      "loss": 0.6009,
+      "step": 844
+    },
+    {
+      "epoch": 0.06862665475513685,
+      "grad_norm": 4.005038090148275,
+      "learning_rate": 4.9805103060054235e-06,
+      "loss": 0.5548,
+      "step": 845
+    },
+    {
+      "epoch": 0.06870786973117843,
+      "grad_norm": 5.440366650852374,
+      "learning_rate": 4.980428265108491e-06,
+      "loss": 0.5983,
+      "step": 846
+    },
+    {
+      "epoch": 0.06878908470722,
+      "grad_norm": 6.265163613283854,
+      "learning_rate": 4.980346052579049e-06,
+      "loss": 0.5767,
+      "step": 847
+    },
+    {
+      "epoch": 0.0688702996832616,
+      "grad_norm": 4.18497253484256,
+      "learning_rate": 4.9802636684227875e-06,
+      "loss": 0.6295,
+      "step": 848
+    },
+    {
+      "epoch": 0.06895151465930317,
+      "grad_norm": 36.607837712422636,
+      "learning_rate": 4.980181112645407e-06,
+      "loss": 0.5392,
+      "step": 849
+    },
+    {
+      "epoch": 0.06903272963534476,
+      "grad_norm": 4.250875496889967,
+      "learning_rate": 4.9800983852526195e-06,
+      "loss": 0.7019,
+      "step": 850
+    },
+    {
+      "epoch": 0.06911394461138634,
+      "grad_norm": 4.20052946210975,
+      "learning_rate": 4.980015486250149e-06,
+      "loss": 0.6024,
+      "step": 851
+    },
+    {
+      "epoch": 0.06919515958742792,
+      "grad_norm": 8.36262034310868,
+      "learning_rate": 4.979932415643733e-06,
+      "loss": 0.505,
+      "step": 852
+    },
+    {
+      "epoch": 0.06927637456346951,
+      "grad_norm": 3.9569054991402206,
+      "learning_rate": 4.9798491734391185e-06,
+      "loss": 0.6134,
+      "step": 853
+    },
+    {
+      "epoch": 0.06935758953951109,
+      "grad_norm": 8.784347823002246,
+      "learning_rate": 4.9797657596420655e-06,
+      "loss": 0.6008,
+      "step": 854
+    },
+    {
+      "epoch": 0.06943880451555266,
+      "grad_norm": 6.296297313209,
+      "learning_rate": 4.979682174258346e-06,
+      "loss": 0.5597,
+      "step": 855
+    },
+    {
+      "epoch": 0.06952001949159425,
+      "grad_norm": 5.72675894986214,
+      "learning_rate": 4.979598417293743e-06,
+      "loss": 0.7964,
+      "step": 856
+    },
+    {
+      "epoch": 0.06960123446763583,
+      "grad_norm": 9.153121757631224,
+      "learning_rate": 4.979514488754053e-06,
+      "loss": 0.6276,
+      "step": 857
+    },
+    {
+      "epoch": 0.06968244944367741,
+      "grad_norm": 3.4432015000564156,
+      "learning_rate": 4.979430388645083e-06,
+      "loss": 0.6616,
+      "step": 858
+    },
+    {
+      "epoch": 0.069763664419719,
+      "grad_norm": 4.626309196229603,
+      "learning_rate": 4.979346116972653e-06,
+      "loss": 0.6686,
+      "step": 859
+    },
+    {
+      "epoch": 0.06984487939576058,
+      "grad_norm": 4.3625603630162555,
+      "learning_rate": 4.979261673742592e-06,
+      "loss": 0.7034,
+      "step": 860
+    },
+    {
+      "epoch": 0.06992609437180217,
+      "grad_norm": 3.987951551641592,
+      "learning_rate": 4.9791770589607455e-06,
+      "loss": 0.7321,
+      "step": 861
+    },
+    {
+      "epoch": 0.07000730934784374,
+      "grad_norm": 4.806090340673884,
+      "learning_rate": 4.979092272632968e-06,
+      "loss": 0.6409,
+      "step": 862
+    },
+    {
+      "epoch": 0.07008852432388532,
+      "grad_norm": 8.252678583793005,
+      "learning_rate": 4.979007314765124e-06,
+      "loss": 0.6183,
+      "step": 863
+    },
+    {
+      "epoch": 0.07016973929992691,
+      "grad_norm": 7.048613756795895,
+      "learning_rate": 4.978922185363095e-06,
+      "loss": 0.5649,
+      "step": 864
+    },
+    {
+      "epoch": 0.07025095427596849,
+      "grad_norm": 7.107709811977146,
+      "learning_rate": 4.97883688443277e-06,
+      "loss": 0.7172,
+      "step": 865
+    },
+    {
+      "epoch": 0.07033216925201007,
+      "grad_norm": 5.254387891978319,
+      "learning_rate": 4.9787514119800515e-06,
+      "loss": 0.756,
+      "step": 866
+    },
+    {
+      "epoch": 0.07041338422805166,
+      "grad_norm": 4.158496862178936,
+      "learning_rate": 4.9786657680108545e-06,
+      "loss": 0.4973,
+      "step": 867
+    },
+    {
+      "epoch": 0.07049459920409323,
+      "grad_norm": 4.086601599664647,
+      "learning_rate": 4.978579952531104e-06,
+      "loss": 0.7624,
+      "step": 868
+    },
+    {
+      "epoch": 0.07057581418013481,
+      "grad_norm": 4.066778591214372,
+      "learning_rate": 4.978493965546738e-06,
+      "loss": 0.4797,
+      "step": 869
+    },
+    {
+      "epoch": 0.0706570291561764,
+      "grad_norm": 4.458730486420523,
+      "learning_rate": 4.9784078070637076e-06,
+      "loss": 0.7739,
+      "step": 870
+    },
+    {
+      "epoch": 0.07073824413221798,
+      "grad_norm": 4.2724723822092425,
+      "learning_rate": 4.978321477087974e-06,
+      "loss": 0.5737,
+      "step": 871
+    },
+    {
+      "epoch": 0.07081945910825957,
+      "grad_norm": 4.54951709242148,
+      "learning_rate": 4.97823497562551e-06,
+      "loss": 0.5447,
+      "step": 872
+    },
+    {
+      "epoch": 0.07090067408430115,
+      "grad_norm": 4.839083574408698,
+      "learning_rate": 4.978148302682301e-06,
+      "loss": 0.7224,
+      "step": 873
+    },
+    {
+      "epoch": 0.07098188906034272,
+      "grad_norm": 5.90958403636814,
+      "learning_rate": 4.978061458264346e-06,
+      "loss": 0.6231,
+      "step": 874
+    },
+    {
+      "epoch": 0.07106310403638431,
+      "grad_norm": 4.950954034585027,
+      "learning_rate": 4.977974442377652e-06,
+      "loss": 0.5298,
+      "step": 875
+    },
+    {
+      "epoch": 0.07114431901242589,
+      "grad_norm": 4.705158462536222,
+      "learning_rate": 4.977887255028241e-06,
+      "loss": 0.5893,
+      "step": 876
+    },
+    {
+      "epoch": 0.07122553398846747,
+      "grad_norm": 6.237199358592749,
+      "learning_rate": 4.977799896222148e-06,
+      "loss": 0.6342,
+      "step": 877
+    },
+    {
+      "epoch": 0.07130674896450906,
+      "grad_norm": 5.842398936618045,
+      "learning_rate": 4.977712365965414e-06,
+      "loss": 0.6228,
+      "step": 878
+    },
+    {
+      "epoch": 0.07138796394055064,
+      "grad_norm": 4.674400039131552,
+      "learning_rate": 4.9776246642640965e-06,
+      "loss": 0.639,
+      "step": 879
+    },
+    {
+      "epoch": 0.07146917891659221,
+      "grad_norm": 6.64025918584984,
+      "learning_rate": 4.977536791124267e-06,
+      "loss": 0.6026,
+      "step": 880
+    },
+    {
+      "epoch": 0.0715503938926338,
+      "grad_norm": 6.036395710793269,
+      "learning_rate": 4.9774487465520025e-06,
+      "loss": 0.8198,
+      "step": 881
+    },
+    {
+      "epoch": 0.07163160886867538,
+      "grad_norm": 4.761294703683568,
+      "learning_rate": 4.977360530553397e-06,
+      "loss": 0.5251,
+      "step": 882
+    },
+    {
+      "epoch": 0.07171282384471697,
+      "grad_norm": 4.922109235469366,
+      "learning_rate": 4.977272143134554e-06,
+      "loss": 0.6486,
+      "step": 883
+    },
+    {
+      "epoch": 0.07179403882075855,
+      "grad_norm": 4.672737077545084,
+      "learning_rate": 4.97718358430159e-06,
+      "loss": 0.6457,
+      "step": 884
+    },
+    {
+      "epoch": 0.07187525379680013,
+      "grad_norm": 8.289465047137561,
+      "learning_rate": 4.977094854060631e-06,
+      "loss": 0.6234,
+      "step": 885
+    },
+    {
+      "epoch": 0.07195646877284172,
+      "grad_norm": 5.587306517848276,
+      "learning_rate": 4.977005952417818e-06,
+      "loss": 0.6662,
+      "step": 886
+    },
+    {
+      "epoch": 0.0720376837488833,
+      "grad_norm": 3.711622831241775,
+      "learning_rate": 4.9769168793793036e-06,
+      "loss": 0.7314,
+      "step": 887
+    },
+    {
+      "epoch": 0.07211889872492487,
+      "grad_norm": 7.5781498502969145,
+      "learning_rate": 4.976827634951249e-06,
+      "loss": 0.6133,
+      "step": 888
+    },
+    {
+      "epoch": 0.07220011370096646,
+      "grad_norm": 4.823514400373474,
+      "learning_rate": 4.976738219139831e-06,
+      "loss": 0.6408,
+      "step": 889
+    },
+    {
+      "epoch": 0.07228132867700804,
+      "grad_norm": 8.761667850604423,
+      "learning_rate": 4.976648631951236e-06,
+      "loss": 0.6452,
+      "step": 890
+    },
+    {
+      "epoch": 0.07236254365304962,
+      "grad_norm": 3.571671073730977,
+      "learning_rate": 4.976558873391663e-06,
+      "loss": 0.6598,
+      "step": 891
+    },
+    {
+      "epoch": 0.0724437586290912,
+      "grad_norm": 5.322524151802511,
+      "learning_rate": 4.976468943467323e-06,
+      "loss": 0.5386,
+      "step": 892
+    },
+    {
+      "epoch": 0.07252497360513278,
+      "grad_norm": 5.66163748773043,
+      "learning_rate": 4.976378842184439e-06,
+      "loss": 0.5516,
+      "step": 893
+    },
+    {
+      "epoch": 0.07260618858117437,
+      "grad_norm": 5.353033651730894,
+      "learning_rate": 4.9762885695492454e-06,
+      "loss": 0.6012,
+      "step": 894
+    },
+    {
+      "epoch": 0.07268740355721595,
+      "grad_norm": 6.855863397519137,
+      "learning_rate": 4.976198125567988e-06,
+      "loss": 0.6303,
+      "step": 895
+    },
+    {
+      "epoch": 0.07276861853325753,
+      "grad_norm": 4.84169944440175,
+      "learning_rate": 4.976107510246925e-06,
+      "loss": 0.528,
+      "step": 896
+    },
+    {
+      "epoch": 0.07284983350929912,
+      "grad_norm": 5.092992144693499,
+      "learning_rate": 4.976016723592328e-06,
+      "loss": 0.5648,
+      "step": 897
+    },
+    {
+      "epoch": 0.0729310484853407,
+      "grad_norm": 4.858637678276697,
+      "learning_rate": 4.975925765610476e-06,
+      "loss": 0.5921,
+      "step": 898
+    },
+    {
+      "epoch": 0.07301226346138227,
+      "grad_norm": 4.995201775650152,
+      "learning_rate": 4.975834636307667e-06,
+      "loss": 0.6145,
+      "step": 899
+    },
+    {
+      "epoch": 0.07309347843742386,
+      "grad_norm": 6.2150093032360365,
+      "learning_rate": 4.975743335690203e-06,
+      "loss": 0.5049,
+      "step": 900
+    },
+    {
+      "epoch": 0.07317469341346544,
+      "grad_norm": 5.737916586640195,
+      "learning_rate": 4.975651863764403e-06,
+      "loss": 0.5949,
+      "step": 901
+    },
+    {
+      "epoch": 0.07325590838950702,
+      "grad_norm": 5.21474627504135,
+      "learning_rate": 4.975560220536596e-06,
+      "loss": 0.8498,
+      "step": 902
+    },
+    {
+      "epoch": 0.07333712336554861,
+      "grad_norm": 3.9495934594747877,
+      "learning_rate": 4.975468406013124e-06,
+      "loss": 0.6854,
+      "step": 903
+    },
+    {
+      "epoch": 0.07341833834159019,
+      "grad_norm": 7.580396023436531,
+      "learning_rate": 4.97537642020034e-06,
+      "loss": 0.585,
+      "step": 904
+    },
+    {
+      "epoch": 0.07349955331763178,
+      "grad_norm": 6.081348176726435,
+      "learning_rate": 4.9752842631046075e-06,
+      "loss": 0.5681,
+      "step": 905
+    },
+    {
+      "epoch": 0.07358076829367335,
+      "grad_norm": 7.6219497191818695,
+      "learning_rate": 4.975191934732304e-06,
+      "loss": 0.5283,
+      "step": 906
+    },
+    {
+      "epoch": 0.07366198326971493,
+      "grad_norm": 7.644719180949304,
+      "learning_rate": 4.975099435089819e-06,
+      "loss": 0.544,
+      "step": 907
+    },
+    {
+      "epoch": 0.07374319824575652,
+      "grad_norm": 4.967279914364186,
+      "learning_rate": 4.975006764183552e-06,
+      "loss": 0.6976,
+      "step": 908
+    },
+    {
+      "epoch": 0.0738244132217981,
+      "grad_norm": 4.727348872362556,
+      "learning_rate": 4.974913922019916e-06,
+      "loss": 0.6466,
+      "step": 909
+    },
+    {
+      "epoch": 0.07390562819783968,
+      "grad_norm": 6.6416921364354895,
+      "learning_rate": 4.974820908605336e-06,
+      "loss": 0.5407,
+      "step": 910
+    },
+    {
+      "epoch": 0.07398684317388127,
+      "grad_norm": 4.934581673705169,
+      "learning_rate": 4.974727723946245e-06,
+      "loss": 0.6653,
+      "step": 911
+    },
+    {
+      "epoch": 0.07406805814992284,
+      "grad_norm": 3.2488987014947286,
+      "learning_rate": 4.974634368049094e-06,
+      "loss": 0.5007,
+      "step": 912
+    },
+    {
+      "epoch": 0.07414927312596442,
+      "grad_norm": 6.227255018738214,
+      "learning_rate": 4.974540840920341e-06,
+      "loss": 0.5501,
+      "step": 913
+    },
+    {
+      "epoch": 0.07423048810200601,
+      "grad_norm": 4.977844807059578,
+      "learning_rate": 4.974447142566458e-06,
+      "loss": 0.7246,
+      "step": 914
+    },
+    {
+      "epoch": 0.07431170307804759,
+      "grad_norm": 4.9582732140375105,
+      "learning_rate": 4.974353272993929e-06,
+      "loss": 0.6714,
+      "step": 915
+    },
+    {
+      "epoch": 0.07439291805408918,
+      "grad_norm": 5.123191354056708,
+      "learning_rate": 4.974259232209249e-06,
+      "loss": 0.7354,
+      "step": 916
+    },
+    {
+      "epoch": 0.07447413303013076,
+      "grad_norm": 5.206554108430837,
+      "learning_rate": 4.9741650202189245e-06,
+      "loss": 0.6421,
+      "step": 917
+    },
+    {
+      "epoch": 0.07455534800617233,
+      "grad_norm": 8.459450047833085,
+      "learning_rate": 4.9740706370294755e-06,
+      "loss": 0.8359,
+      "step": 918
+    },
+    {
+      "epoch": 0.07463656298221392,
+      "grad_norm": 4.78007740463755,
+      "learning_rate": 4.973976082647432e-06,
+      "loss": 0.6941,
+      "step": 919
+    },
+    {
+      "epoch": 0.0747177779582555,
+      "grad_norm": 4.06407753833687,
+      "learning_rate": 4.9738813570793365e-06,
+      "loss": 0.6078,
+      "step": 920
+    },
+    {
+      "epoch": 0.07479899293429708,
+      "grad_norm": 6.4634082754205915,
+      "learning_rate": 4.973786460331744e-06,
+      "loss": 0.6072,
+      "step": 921
+    },
+    {
+      "epoch": 0.07488020791033867,
+      "grad_norm": 4.794171216936538,
+      "learning_rate": 4.973691392411221e-06,
+      "loss": 0.6489,
+      "step": 922
+    },
+    {
+      "epoch": 0.07496142288638025,
+      "grad_norm": 4.048880670381879,
+      "learning_rate": 4.973596153324346e-06,
+      "loss": 0.6415,
+      "step": 923
+    },
+    {
+      "epoch": 0.07504263786242182,
+      "grad_norm": 4.260077508939967,
+      "learning_rate": 4.973500743077707e-06,
+      "loss": 0.5398,
+      "step": 924
+    },
+    {
+      "epoch": 0.07512385283846341,
+      "grad_norm": 4.630184924560269,
+      "learning_rate": 4.9734051616779085e-06,
+      "loss": 0.5155,
+      "step": 925
+    },
+    {
+      "epoch": 0.07520506781450499,
+      "grad_norm": 6.107378791436437,
+      "learning_rate": 4.973309409131564e-06,
+      "loss": 0.5784,
+      "step": 926
+    },
+    {
+      "epoch": 0.07528628279054658,
+      "grad_norm": 4.991440001564157,
+      "learning_rate": 4.973213485445298e-06,
+      "loss": 0.6463,
+      "step": 927
+    },
+    {
+      "epoch": 0.07536749776658816,
+      "grad_norm": 3.9596877989124017,
+      "learning_rate": 4.973117390625746e-06,
+      "loss": 0.5694,
+      "step": 928
+    },
+    {
+      "epoch": 0.07544871274262974,
+      "grad_norm": 5.9210563584390385,
+      "learning_rate": 4.9730211246795614e-06,
+      "loss": 0.5044,
+      "step": 929
+    },
+    {
+      "epoch": 0.07552992771867133,
+      "grad_norm": 6.400935356399938,
+      "learning_rate": 4.9729246876134015e-06,
+      "loss": 0.5711,
+      "step": 930
+    },
+    {
+      "epoch": 0.0756111426947129,
+      "grad_norm": 6.074570162257794,
+      "learning_rate": 4.9728280794339426e-06,
+      "loss": 0.7171,
+      "step": 931
+    },
+    {
+      "epoch": 0.07569235767075448,
+      "grad_norm": 7.525931570888455,
+      "learning_rate": 4.972731300147867e-06,
+      "loss": 0.5338,
+      "step": 932
+    },
+    {
+      "epoch": 0.07577357264679607,
+      "grad_norm": 5.22558586071865,
+      "learning_rate": 4.972634349761873e-06,
+      "loss": 0.5591,
+      "step": 933
+    },
+    {
+      "epoch": 0.07585478762283765,
+      "grad_norm": 10.047768244682082,
+      "learning_rate": 4.972537228282668e-06,
+      "loss": 0.6477,
+      "step": 934
+    },
+    {
+      "epoch": 0.07593600259887923,
+      "grad_norm": 6.876714921725991,
+      "learning_rate": 4.972439935716972e-06,
+      "loss": 0.5482,
+      "step": 935
+    },
+    {
+      "epoch": 0.07601721757492082,
+      "grad_norm": 6.550177057547305,
+      "learning_rate": 4.972342472071518e-06,
+      "loss": 0.5923,
+      "step": 936
+    },
+    {
+      "epoch": 0.0760984325509624,
+      "grad_norm": 5.709520994341491,
+      "learning_rate": 4.97224483735305e-06,
+      "loss": 0.6293,
+      "step": 937
+    },
+    {
+      "epoch": 0.07617964752700399,
+      "grad_norm": 4.727570596744493,
+      "learning_rate": 4.972147031568322e-06,
+      "loss": 0.5793,
+      "step": 938
+    },
+    {
+      "epoch": 0.07626086250304556,
+      "grad_norm": 5.251994860118902,
+      "learning_rate": 4.972049054724104e-06,
+      "loss": 0.7869,
+      "step": 939
+    },
+    {
+      "epoch": 0.07634207747908714,
+      "grad_norm": 7.152014425413055,
+      "learning_rate": 4.9719509068271755e-06,
+      "loss": 0.5164,
+      "step": 940
+    },
+    {
+      "epoch": 0.07642329245512873,
+      "grad_norm": 9.90794133427569,
+      "learning_rate": 4.971852587884325e-06,
+      "loss": 0.5625,
+      "step": 941
+    },
+    {
+      "epoch": 0.07650450743117031,
+      "grad_norm": 5.980631957938455,
+      "learning_rate": 4.97175409790236e-06,
+      "loss": 0.6737,
+      "step": 942
+    },
+    {
+      "epoch": 0.07658572240721188,
+      "grad_norm": 11.229665313768436,
+      "learning_rate": 4.97165543688809e-06,
+      "loss": 0.6511,
+      "step": 943
+    },
+    {
+      "epoch": 0.07666693738325348,
+      "grad_norm": 15.507076859457626,
+      "learning_rate": 4.971556604848346e-06,
+      "loss": 0.595,
+      "step": 944
+    },
+    {
+      "epoch": 0.07674815235929505,
+      "grad_norm": 4.875894631175207,
+      "learning_rate": 4.971457601789966e-06,
+      "loss": 0.5992,
+      "step": 945
+    },
+    {
+      "epoch": 0.07682936733533664,
+      "grad_norm": 4.406952012061629,
+      "learning_rate": 4.9713584277198e-06,
+      "loss": 0.5121,
+      "step": 946
+    },
+    {
+      "epoch": 0.07691058231137822,
+      "grad_norm": 5.211822267384949,
+      "learning_rate": 4.97125908264471e-06,
+      "loss": 0.521,
+      "step": 947
+    },
+    {
+      "epoch": 0.0769917972874198,
+      "grad_norm": 6.646319909846443,
+      "learning_rate": 4.97115956657157e-06,
+      "loss": 0.5977,
+      "step": 948
+    },
+    {
+      "epoch": 0.07707301226346139,
+      "grad_norm": 5.951258269254235,
+      "learning_rate": 4.971059879507268e-06,
+      "loss": 0.6193,
+      "step": 949
+    },
+    {
+      "epoch": 0.07715422723950296,
+      "grad_norm": 15.849437156314497,
+      "learning_rate": 4.970960021458699e-06,
+      "loss": 0.7991,
+      "step": 950
+    },
+    {
+      "epoch": 0.07723544221554454,
+      "grad_norm": 4.470056142523615,
+      "learning_rate": 4.9708599924327735e-06,
+      "loss": 0.5992,
+      "step": 951
+    },
+    {
+      "epoch": 0.07731665719158613,
+      "grad_norm": 7.357239226247135,
+      "learning_rate": 4.970759792436414e-06,
+      "loss": 0.574,
+      "step": 952
+    },
+    {
+      "epoch": 0.07739787216762771,
+      "grad_norm": 6.193524739615741,
+      "learning_rate": 4.970659421476553e-06,
+      "loss": 0.6871,
+      "step": 953
+    },
+    {
+      "epoch": 0.07747908714366929,
+      "grad_norm": 6.192316599355833,
+      "learning_rate": 4.970558879560137e-06,
+      "loss": 0.6069,
+      "step": 954
+    },
+    {
+      "epoch": 0.07756030211971088,
+      "grad_norm": 7.025324708867777,
+      "learning_rate": 4.97045816669412e-06,
+      "loss": 0.45,
+      "step": 955
+    },
+    {
+      "epoch": 0.07764151709575245,
+      "grad_norm": 5.808792647688984,
+      "learning_rate": 4.970357282885473e-06,
+      "loss": 0.7007,
+      "step": 956
+    },
+    {
+      "epoch": 0.07772273207179405,
+      "grad_norm": 5.800444379893061,
+      "learning_rate": 4.970256228141177e-06,
+      "loss": 0.5379,
+      "step": 957
+    },
+    {
+      "epoch": 0.07780394704783562,
+      "grad_norm": 4.764091635977128,
+      "learning_rate": 4.970155002468223e-06,
+      "loss": 0.6805,
+      "step": 958
+    },
+    {
+      "epoch": 0.0778851620238772,
+      "grad_norm": 5.223528426735964,
+      "learning_rate": 4.970053605873616e-06,
+      "loss": 0.6757,
+      "step": 959
+    },
+    {
+      "epoch": 0.07796637699991879,
+      "grad_norm": 7.117293048534837,
+      "learning_rate": 4.969952038364372e-06,
+      "loss": 0.5716,
+      "step": 960
+    },
+    {
+      "epoch": 0.07804759197596037,
+      "grad_norm": 8.849272822881796,
+      "learning_rate": 4.96985029994752e-06,
+      "loss": 0.7158,
+      "step": 961
+    },
+    {
+      "epoch": 0.07812880695200194,
+      "grad_norm": 4.323247038552405,
+      "learning_rate": 4.969748390630097e-06,
+      "loss": 0.7014,
+      "step": 962
+    },
+    {
+      "epoch": 0.07821002192804354,
+      "grad_norm": 3.15078482423997,
+      "learning_rate": 4.969646310419157e-06,
+      "loss": 0.6552,
+      "step": 963
+    },
+    {
+      "epoch": 0.07829123690408511,
+      "grad_norm": 5.142051674034742,
+      "learning_rate": 4.9695440593217635e-06,
+      "loss": 0.5731,
+      "step": 964
+    },
+    {
+      "epoch": 0.07837245188012669,
+      "grad_norm": 4.530754006814891,
+      "learning_rate": 4.96944163734499e-06,
+      "loss": 0.4648,
+      "step": 965
+    },
+    {
+      "epoch": 0.07845366685616828,
+      "grad_norm": 5.999396192869684,
+      "learning_rate": 4.969339044495925e-06,
+      "loss": 0.6981,
+      "step": 966
+    },
+    {
+      "epoch": 0.07853488183220986,
+      "grad_norm": 4.570316628434694,
+      "learning_rate": 4.969236280781667e-06,
+      "loss": 0.5834,
+      "step": 967
+    },
+    {
+      "epoch": 0.07861609680825145,
+      "grad_norm": 29.36059594145154,
+      "learning_rate": 4.9691333462093264e-06,
+      "loss": 0.6442,
+      "step": 968
+    },
+    {
+      "epoch": 0.07869731178429303,
+      "grad_norm": 4.95470560540118,
+      "learning_rate": 4.969030240786026e-06,
+      "loss": 0.4758,
+      "step": 969
+    },
+    {
+      "epoch": 0.0787785267603346,
+      "grad_norm": 3.893402872277116,
+      "learning_rate": 4.9689269645189e-06,
+      "loss": 0.6544,
+      "step": 970
+    },
+    {
+      "epoch": 0.0788597417363762,
+      "grad_norm": 5.3148301362547885,
+      "learning_rate": 4.968823517415095e-06,
+      "loss": 0.611,
+      "step": 971
+    },
+    {
+      "epoch": 0.07894095671241777,
+      "grad_norm": 5.03652657763463,
+      "learning_rate": 4.9687198994817685e-06,
+      "loss": 0.6703,
+      "step": 972
+    },
+    {
+      "epoch": 0.07902217168845935,
+      "grad_norm": 4.100966531743051,
+      "learning_rate": 4.9686161107260906e-06,
+      "loss": 0.5816,
+      "step": 973
+    },
+    {
+      "epoch": 0.07910338666450094,
+      "grad_norm": 5.116827473240173,
+      "learning_rate": 4.968512151155242e-06,
+      "loss": 0.5917,
+      "step": 974
+    },
+    {
+      "epoch": 0.07918460164054252,
+      "grad_norm": 3.2335435030359596,
+      "learning_rate": 4.968408020776419e-06,
+      "loss": 0.5538,
+      "step": 975
+    },
+    {
+      "epoch": 0.07926581661658409,
+      "grad_norm": 7.5591035161901035,
+      "learning_rate": 4.968303719596823e-06,
+      "loss": 0.461,
+      "step": 976
+    },
+    {
+      "epoch": 0.07934703159262568,
+      "grad_norm": 10.075832303176476,
+      "learning_rate": 4.9681992476236725e-06,
+      "loss": 0.6354,
+      "step": 977
+    },
+    {
+      "epoch": 0.07942824656866726,
+      "grad_norm": 5.561625086524589,
+      "learning_rate": 4.968094604864198e-06,
+      "loss": 0.5383,
+      "step": 978
+    },
+    {
+      "epoch": 0.07950946154470885,
+      "grad_norm": 5.6568874477353095,
+      "learning_rate": 4.967989791325639e-06,
+      "loss": 0.6593,
+      "step": 979
+    },
+    {
+      "epoch": 0.07959067652075043,
+      "grad_norm": 6.416186383122935,
+      "learning_rate": 4.967884807015247e-06,
+      "loss": 0.833,
+      "step": 980
+    },
+    {
+      "epoch": 0.079671891496792,
+      "grad_norm": 5.7530564697103355,
+      "learning_rate": 4.967779651940289e-06,
+      "loss": 0.7025,
+      "step": 981
+    },
+    {
+      "epoch": 0.0797531064728336,
+      "grad_norm": 6.850642228085424,
+      "learning_rate": 4.967674326108039e-06,
+      "loss": 0.5582,
+      "step": 982
+    },
+    {
+      "epoch": 0.07983432144887517,
+      "grad_norm": 4.208095970275836,
+      "learning_rate": 4.9675688295257855e-06,
+      "loss": 0.514,
+      "step": 983
+    },
+    {
+      "epoch": 0.07991553642491675,
+      "grad_norm": 5.1148065579680075,
+      "learning_rate": 4.967463162200828e-06,
+      "loss": 0.6708,
+      "step": 984
+    },
+    {
+      "epoch": 0.07999675140095834,
+      "grad_norm": 7.806584044546556,
+      "learning_rate": 4.967357324140479e-06,
+      "loss": 0.639,
+      "step": 985
+    },
+    {
+      "epoch": 0.08007796637699992,
+      "grad_norm": 7.822904916898733,
+      "learning_rate": 4.967251315352062e-06,
+      "loss": 0.7296,
+      "step": 986
+    },
+    {
+      "epoch": 0.0801591813530415,
+      "grad_norm": 4.251509260453048,
+      "learning_rate": 4.9671451358429115e-06,
+      "loss": 0.6169,
+      "step": 987
+    },
+    {
+      "epoch": 0.08024039632908309,
+      "grad_norm": 3.9712450920352014,
+      "learning_rate": 4.967038785620374e-06,
+      "loss": 0.6307,
+      "step": 988
+    },
+    {
+      "epoch": 0.08032161130512466,
+      "grad_norm": 5.013282143950785,
+      "learning_rate": 4.96693226469181e-06,
+      "loss": 0.6896,
+      "step": 989
+    },
+    {
+      "epoch": 0.08040282628116625,
+      "grad_norm": 6.309239309855388,
+      "learning_rate": 4.966825573064589e-06,
+      "loss": 0.5816,
+      "step": 990
+    },
+    {
+      "epoch": 0.08048404125720783,
+      "grad_norm": 17.18761480554951,
+      "learning_rate": 4.9667187107460934e-06,
+      "loss": 0.7024,
+      "step": 991
+    },
+    {
+      "epoch": 0.08056525623324941,
+      "grad_norm": 8.511668880504516,
+      "learning_rate": 4.966611677743719e-06,
+      "loss": 0.6545,
+      "step": 992
+    },
+    {
+      "epoch": 0.080646471209291,
+      "grad_norm": 7.07740712784662,
+      "learning_rate": 4.96650447406487e-06,
+      "loss": 0.6241,
+      "step": 993
+    },
+    {
+      "epoch": 0.08072768618533258,
+      "grad_norm": 6.4092146264144505,
+      "learning_rate": 4.966397099716965e-06,
+      "loss": 0.6096,
+      "step": 994
+    },
+    {
+      "epoch": 0.08080890116137415,
+      "grad_norm": 3.371066265022369,
+      "learning_rate": 4.9662895547074345e-06,
+      "loss": 0.6807,
+      "step": 995
+    },
+    {
+      "epoch": 0.08089011613741574,
+      "grad_norm": 11.15283861894202,
+      "learning_rate": 4.96618183904372e-06,
+      "loss": 0.69,
+      "step": 996
+    },
+    {
+      "epoch": 0.08097133111345732,
+      "grad_norm": 5.584728402833689,
+      "learning_rate": 4.966073952733273e-06,
+      "loss": 0.6545,
+      "step": 997
+    },
+    {
+      "epoch": 0.0810525460894989,
+      "grad_norm": 4.928404998187488,
+      "learning_rate": 4.965965895783561e-06,
+      "loss": 0.7363,
+      "step": 998
+    },
+    {
+      "epoch": 0.08113376106554049,
+      "grad_norm": 5.18162526836767,
+      "learning_rate": 4.96585766820206e-06,
+      "loss": 0.7405,
+      "step": 999
+    },
+    {
+      "epoch": 0.08121497604158207,
+      "grad_norm": 5.108535543984558,
+      "learning_rate": 4.965749269996258e-06,
+      "loss": 0.4666,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08129619101762366,
+      "grad_norm": 4.081215207318422,
+      "learning_rate": 4.965640701173657e-06,
+      "loss": 0.6248,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08137740599366523,
+      "grad_norm": 4.819602648972772,
+      "learning_rate": 4.9655319617417674e-06,
+      "loss": 0.6029,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08145862096970681,
+      "grad_norm": 3.672836794950573,
+      "learning_rate": 4.965423051708116e-06,
+      "loss": 0.7782,
+      "step": 1003
+    },
+    {
+      "epoch": 0.0815398359457484,
+      "grad_norm": 4.726346322374398,
+      "learning_rate": 4.965313971080237e-06,
+      "loss": 0.5324,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08162105092178998,
+      "grad_norm": 5.155470667426197,
+      "learning_rate": 4.96520471986568e-06,
+      "loss": 0.666,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08170226589783156,
+      "grad_norm": 4.923143643344893,
+      "learning_rate": 4.965095298072001e-06,
+      "loss": 0.5542,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08178348087387315,
+      "grad_norm": 4.1341959962667305,
+      "learning_rate": 4.964985705706775e-06,
+      "loss": 0.5911,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08186469584991472,
+      "grad_norm": 5.790946529810634,
+      "learning_rate": 4.964875942777584e-06,
+      "loss": 0.7514,
+      "step": 1008
+    },
+    {
+      "epoch": 0.0819459108259563,
+      "grad_norm": 5.188744184884793,
+      "learning_rate": 4.964766009292022e-06,
+      "loss": 0.704,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08202712580199789,
+      "grad_norm": 9.318471720754397,
+      "learning_rate": 4.9646559052576985e-06,
+      "loss": 0.6622,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08210834077803947,
+      "grad_norm": 4.902314685975934,
+      "learning_rate": 4.9645456306822285e-06,
+      "loss": 0.5447,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08218955575408106,
+      "grad_norm": 5.828651116573771,
+      "learning_rate": 4.964435185573245e-06,
+      "loss": 0.5288,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08227077073012264,
+      "grad_norm": 2.864291283393185,
+      "learning_rate": 4.96432456993839e-06,
+      "loss": 0.6403,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08235198570616421,
+      "grad_norm": 5.284219787159016,
+      "learning_rate": 4.964213783785317e-06,
+      "loss": 0.6291,
+      "step": 1014
+    },
+    {
+      "epoch": 0.0824332006822058,
+      "grad_norm": 4.43415065344841,
+      "learning_rate": 4.9641028271216905e-06,
+      "loss": 0.8372,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08251441565824738,
+      "grad_norm": 4.37746143116903,
+      "learning_rate": 4.9639916999551905e-06,
+      "loss": 0.5823,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08259563063428896,
+      "grad_norm": 15.987254695136924,
+      "learning_rate": 4.963880402293506e-06,
+      "loss": 0.6104,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08267684561033055,
+      "grad_norm": 5.7244481019012925,
+      "learning_rate": 4.963768934144336e-06,
+      "loss": 0.7552,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08275806058637213,
+      "grad_norm": 4.217271904051636,
+      "learning_rate": 4.963657295515396e-06,
+      "loss": 0.6159,
+      "step": 1019
+    },
+    {
+      "epoch": 0.0828392755624137,
+      "grad_norm": 7.382613542004634,
+      "learning_rate": 4.963545486414411e-06,
+      "loss": 0.6277,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0829204905384553,
+      "grad_norm": 5.353844268940195,
+      "learning_rate": 4.963433506849115e-06,
+      "loss": 0.738,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08300170551449687,
+      "grad_norm": 10.18166117587427,
+      "learning_rate": 4.963321356827258e-06,
+      "loss": 0.6173,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08308292049053846,
+      "grad_norm": 7.319414111585189,
+      "learning_rate": 4.9632090363565995e-06,
+      "loss": 0.6377,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08316413546658004,
+      "grad_norm": 6.252416872046548,
+      "learning_rate": 4.963096545444913e-06,
+      "loss": 0.5095,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08324535044262162,
+      "grad_norm": 4.976076953773711,
+      "learning_rate": 4.962983884099981e-06,
+      "loss": 0.5225,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0833265654186632,
+      "grad_norm": 6.865653287550914,
+      "learning_rate": 4.9628710523296e-06,
+      "loss": 0.5425,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08340778039470478,
+      "grad_norm": 5.6514607039180005,
+      "learning_rate": 4.962758050141576e-06,
+      "loss": 0.5492,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08348899537074636,
+      "grad_norm": 8.200634823957119,
+      "learning_rate": 4.962644877543729e-06,
+      "loss": 0.5683,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08357021034678795,
+      "grad_norm": 4.089877382417978,
+      "learning_rate": 4.96253153454389e-06,
+      "loss": 0.7186,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08365142532282953,
+      "grad_norm": 4.918100288171265,
+      "learning_rate": 4.9624180211499004e-06,
+      "loss": 0.4817,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0837326402988711,
+      "grad_norm": 5.979359137308161,
+      "learning_rate": 4.962304337369618e-06,
+      "loss": 0.575,
+      "step": 1031
+    },
+    {
+      "epoch": 0.0838138552749127,
+      "grad_norm": 5.9147018999524645,
+      "learning_rate": 4.962190483210906e-06,
+      "loss": 0.5979,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08389507025095427,
+      "grad_norm": 3.8882956927363614,
+      "learning_rate": 4.962076458681642e-06,
+      "loss": 0.5231,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08397628522699586,
+      "grad_norm": 8.61999834460005,
+      "learning_rate": 4.96196226378972e-06,
+      "loss": 0.6006,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08405750020303744,
+      "grad_norm": 4.813530840618122,
+      "learning_rate": 4.961847898543038e-06,
+      "loss": 0.7872,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08413871517907902,
+      "grad_norm": 6.218523263233032,
+      "learning_rate": 4.96173336294951e-06,
+      "loss": 0.6882,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08421993015512061,
+      "grad_norm": 5.495336585299689,
+      "learning_rate": 4.961618657017063e-06,
+      "loss": 0.828,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08430114513116219,
+      "grad_norm": 6.714525494166789,
+      "learning_rate": 4.961503780753633e-06,
+      "loss": 0.8362,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08438236010720376,
+      "grad_norm": 3.212772209979488,
+      "learning_rate": 4.9613887341671675e-06,
+      "loss": 0.4224,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08446357508324535,
+      "grad_norm": 7.183836025951974,
+      "learning_rate": 4.961273517265629e-06,
+      "loss": 0.5646,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08454479005928693,
+      "grad_norm": 6.286950241807605,
+      "learning_rate": 4.961158130056989e-06,
+      "loss": 0.5711,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08462600503532851,
+      "grad_norm": 5.85967031646826,
+      "learning_rate": 4.961042572549232e-06,
+      "loss": 0.4819,
+      "step": 1042
+    },
+    {
+      "epoch": 0.0847072200113701,
+      "grad_norm": 4.788980234170179,
+      "learning_rate": 4.960926844750353e-06,
+      "loss": 0.5843,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08478843498741168,
+      "grad_norm": 6.955485093136964,
+      "learning_rate": 4.960810946668362e-06,
+      "loss": 0.5618,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08486964996345327,
+      "grad_norm": 5.484890003960757,
+      "learning_rate": 4.960694878311276e-06,
+      "loss": 0.6486,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08495086493949484,
+      "grad_norm": 5.542258816933063,
+      "learning_rate": 4.960578639687129e-06,
+      "loss": 0.4978,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08503207991553642,
+      "grad_norm": 5.931712433067937,
+      "learning_rate": 4.960462230803961e-06,
+      "loss": 0.6843,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08511329489157801,
+      "grad_norm": 5.759837139562491,
+      "learning_rate": 4.960345651669829e-06,
+      "loss": 0.5678,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08519450986761959,
+      "grad_norm": 8.019301092937635,
+      "learning_rate": 4.960228902292799e-06,
+      "loss": 0.5629,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08527572484366117,
+      "grad_norm": 4.806150258046436,
+      "learning_rate": 4.96011198268095e-06,
+      "loss": 0.4966,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08535693981970276,
+      "grad_norm": 6.278721005688989,
+      "learning_rate": 4.959994892842371e-06,
+      "loss": 0.6009,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08543815479574433,
+      "grad_norm": 4.006330485950786,
+      "learning_rate": 4.959877632785166e-06,
+      "loss": 0.538,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08551936977178591,
+      "grad_norm": 6.0015668910114,
+      "learning_rate": 4.959760202517446e-06,
+      "loss": 0.5693,
+      "step": 1053
+    },
+    {
+      "epoch": 0.0856005847478275,
+      "grad_norm": 17.347793241402837,
+      "learning_rate": 4.959642602047339e-06,
+      "loss": 0.5231,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08568179972386908,
+      "grad_norm": 4.788609767567765,
+      "learning_rate": 4.959524831382981e-06,
+      "loss": 0.6282,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08576301469991067,
+      "grad_norm": 10.272501989404129,
+      "learning_rate": 4.9594068905325225e-06,
+      "loss": 0.7099,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08584422967595225,
+      "grad_norm": 5.360284497181169,
+      "learning_rate": 4.959288779504122e-06,
+      "loss": 0.5964,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08592544465199382,
+      "grad_norm": 4.376816971311538,
+      "learning_rate": 4.959170498305955e-06,
+      "loss": 0.6034,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08600665962803541,
+      "grad_norm": 7.410246905461928,
+      "learning_rate": 4.959052046946203e-06,
+      "loss": 0.5496,
+      "step": 1059
+    },
+    {
+      "epoch": 0.08608787460407699,
+      "grad_norm": 3.455502795284755,
+      "learning_rate": 4.958933425433065e-06,
+      "loss": 0.6964,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08616908958011857,
+      "grad_norm": 5.754779871120552,
+      "learning_rate": 4.958814633774747e-06,
+      "loss": 0.5759,
+      "step": 1061
+    },
+    {
+      "epoch": 0.08625030455616016,
+      "grad_norm": 5.304809027548407,
+      "learning_rate": 4.95869567197947e-06,
+      "loss": 0.807,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08633151953220174,
+      "grad_norm": 6.315153689798026,
+      "learning_rate": 4.958576540055464e-06,
+      "loss": 0.6127,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08641273450824331,
+      "grad_norm": 5.159058407796672,
+      "learning_rate": 4.958457238010974e-06,
+      "loss": 0.5677,
+      "step": 1064
+    },
+    {
+      "epoch": 0.0864939494842849,
+      "grad_norm": 6.440705718742274,
+      "learning_rate": 4.958337765854254e-06,
+      "loss": 0.6875,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08657516446032648,
+      "grad_norm": 5.084236342123366,
+      "learning_rate": 4.958218123593572e-06,
+      "loss": 0.7366,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08665637943636807,
+      "grad_norm": 3.7240825747904447,
+      "learning_rate": 4.958098311237205e-06,
+      "loss": 0.674,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08673759441240965,
+      "grad_norm": 5.058020669894491,
+      "learning_rate": 4.9579783287934445e-06,
+      "loss": 0.6569,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08681880938845123,
+      "grad_norm": 4.0274556261885515,
+      "learning_rate": 4.957858176270591e-06,
+      "loss": 0.6855,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08690002436449282,
+      "grad_norm": 7.179073356157699,
+      "learning_rate": 4.957737853676961e-06,
+      "loss": 0.58,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0869812393405344,
+      "grad_norm": 4.03973215146239,
+      "learning_rate": 4.957617361020879e-06,
+      "loss": 0.515,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08706245431657597,
+      "grad_norm": 7.191602127334415,
+      "learning_rate": 4.9574966983106824e-06,
+      "loss": 0.7152,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08714366929261756,
+      "grad_norm": 4.274773215099363,
+      "learning_rate": 4.95737586555472e-06,
+      "loss": 0.6789,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08722488426865914,
+      "grad_norm": 4.114285344199829,
+      "learning_rate": 4.957254862761354e-06,
+      "loss": 0.7913,
+      "step": 1074
+    },
+    {
+      "epoch": 0.08730609924470072,
+      "grad_norm": 6.861324021554996,
+      "learning_rate": 4.957133689938955e-06,
+      "loss": 0.7393,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08738731422074231,
+      "grad_norm": 6.978671101537034,
+      "learning_rate": 4.95701234709591e-06,
+      "loss": 0.6303,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08746852919678388,
+      "grad_norm": 9.578370134137916,
+      "learning_rate": 4.956890834240613e-06,
+      "loss": 0.6739,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08754974417282547,
+      "grad_norm": 4.953780236460816,
+      "learning_rate": 4.956769151381474e-06,
+      "loss": 0.6609,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08763095914886705,
+      "grad_norm": 6.102476761076953,
+      "learning_rate": 4.9566472985269125e-06,
+      "loss": 0.5512,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08771217412490863,
+      "grad_norm": 4.373415052928376,
+      "learning_rate": 4.956525275685358e-06,
+      "loss": 0.5459,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08779338910095022,
+      "grad_norm": 5.532793766061284,
+      "learning_rate": 4.9564030828652565e-06,
+      "loss": 0.6867,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0878746040769918,
+      "grad_norm": 3.8285573439072254,
+      "learning_rate": 4.956280720075062e-06,
+      "loss": 0.7538,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08795581905303337,
+      "grad_norm": 6.492766812771604,
+      "learning_rate": 4.9561581873232415e-06,
+      "loss": 0.5405,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08803703402907496,
+      "grad_norm": 12.542445075286674,
+      "learning_rate": 4.956035484618272e-06,
+      "loss": 0.5867,
+      "step": 1084
+    },
+    {
+      "epoch": 0.08811824900511654,
+      "grad_norm": 5.8788039197140005,
+      "learning_rate": 4.955912611968648e-06,
+      "loss": 0.5364,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08819946398115812,
+      "grad_norm": 6.820020186044584,
+      "learning_rate": 4.955789569382866e-06,
+      "loss": 0.5554,
+      "step": 1086
+    },
+    {
+      "epoch": 0.08828067895719971,
+      "grad_norm": 5.44097898708547,
+      "learning_rate": 4.955666356869445e-06,
+      "loss": 0.6424,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08836189393324129,
+      "grad_norm": 6.451838410522676,
+      "learning_rate": 4.955542974436908e-06,
+      "loss": 0.6376,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08844310890928288,
+      "grad_norm": 16.847875335305353,
+      "learning_rate": 4.955419422093792e-06,
+      "loss": 0.5683,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08852432388532445,
+      "grad_norm": 5.250679262222032,
+      "learning_rate": 4.955295699848649e-06,
+      "loss": 0.5726,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08860553886136603,
+      "grad_norm": 3.708739891574932,
+      "learning_rate": 4.955171807710037e-06,
+      "loss": 0.664,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08868675383740762,
+      "grad_norm": 6.291250688681625,
+      "learning_rate": 4.955047745686529e-06,
+      "loss": 0.5727,
+      "step": 1092
+    },
+    {
+      "epoch": 0.0887679688134492,
+      "grad_norm": 9.32506667375153,
+      "learning_rate": 4.954923513786711e-06,
+      "loss": 0.6444,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08884918378949078,
+      "grad_norm": 5.9130602901334735,
+      "learning_rate": 4.954799112019178e-06,
+      "loss": 0.6713,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08893039876553237,
+      "grad_norm": 7.27169259231534,
+      "learning_rate": 4.9546745403925385e-06,
+      "loss": 0.598,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08901161374157394,
+      "grad_norm": 4.39443693321864,
+      "learning_rate": 4.954549798915412e-06,
+      "loss": 0.5987,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08909282871761552,
+      "grad_norm": 5.108191362159325,
+      "learning_rate": 4.95442488759643e-06,
+      "loss": 0.5794,
+      "step": 1097
+    },
+    {
+      "epoch": 0.08917404369365711,
+      "grad_norm": 4.394532094744748,
+      "learning_rate": 4.954299806444236e-06,
+      "loss": 0.6292,
+      "step": 1098
+    },
+    {
+      "epoch": 0.08925525866969869,
+      "grad_norm": 5.785973974044764,
+      "learning_rate": 4.954174555467484e-06,
+      "loss": 0.7976,
+      "step": 1099
+    },
+    {
+      "epoch": 0.08933647364574028,
+      "grad_norm": 8.854881338118322,
+      "learning_rate": 4.954049134674842e-06,
+      "loss": 0.4992,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08941768862178186,
+      "grad_norm": 10.211232540471157,
+      "learning_rate": 4.953923544074987e-06,
+      "loss": 0.612,
+      "step": 1101
+    },
+    {
+      "epoch": 0.08949890359782343,
+      "grad_norm": 3.8402408710819724,
+      "learning_rate": 4.953797783676611e-06,
+      "loss": 0.7663,
+      "step": 1102
+    },
+    {
+      "epoch": 0.08958011857386503,
+      "grad_norm": 4.979965541200029,
+      "learning_rate": 4.9536718534884136e-06,
+      "loss": 0.5334,
+      "step": 1103
+    },
+    {
+      "epoch": 0.0896613335499066,
+      "grad_norm": 5.601550825200515,
+      "learning_rate": 4.9535457535191104e-06,
+      "loss": 0.7872,
+      "step": 1104
+    },
+    {
+      "epoch": 0.08974254852594818,
+      "grad_norm": 7.030742327798493,
+      "learning_rate": 4.953419483777427e-06,
+      "loss": 0.5551,
+      "step": 1105
+    },
+    {
+      "epoch": 0.08982376350198977,
+      "grad_norm": 11.753328029891325,
+      "learning_rate": 4.953293044272099e-06,
+      "loss": 0.5917,
+      "step": 1106
+    },
+    {
+      "epoch": 0.08990497847803135,
+      "grad_norm": 4.23164973424406,
+      "learning_rate": 4.953166435011876e-06,
+      "loss": 0.7211,
+      "step": 1107
+    },
+    {
+      "epoch": 0.08998619345407292,
+      "grad_norm": 4.427467426240099,
+      "learning_rate": 4.953039656005519e-06,
+      "loss": 0.725,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09006740843011452,
+      "grad_norm": 6.656857308336458,
+      "learning_rate": 4.9529127072618e-06,
+      "loss": 0.7282,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09014862340615609,
+      "grad_norm": 4.606494946768081,
+      "learning_rate": 4.952785588789504e-06,
+      "loss": 0.6008,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09022983838219768,
+      "grad_norm": 4.861377291580884,
+      "learning_rate": 4.9526583005974275e-06,
+      "loss": 0.7185,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09031105335823926,
+      "grad_norm": 5.1172810841957785,
+      "learning_rate": 4.952530842694375e-06,
+      "loss": 0.451,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09039226833428084,
+      "grad_norm": 5.226532353483265,
+      "learning_rate": 4.95240321508917e-06,
+      "loss": 0.658,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09047348331032243,
+      "grad_norm": 5.970819410829736,
+      "learning_rate": 4.952275417790641e-06,
+      "loss": 0.6415,
+      "step": 1114
+    },
+    {
+      "epoch": 0.090554698286364,
+      "grad_norm": 5.235598768790702,
+      "learning_rate": 4.95214745080763e-06,
+      "loss": 0.5784,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09063591326240558,
+      "grad_norm": 4.456877671742799,
+      "learning_rate": 4.952019314148995e-06,
+      "loss": 0.6458,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09071712823844717,
+      "grad_norm": 5.531350494822164,
+      "learning_rate": 4.951891007823601e-06,
+      "loss": 0.7039,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09079834321448875,
+      "grad_norm": 3.056795429652102,
+      "learning_rate": 4.951762531840325e-06,
+      "loss": 0.6661,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09087955819053034,
+      "grad_norm": 5.721632966434159,
+      "learning_rate": 4.951633886208057e-06,
+      "loss": 0.591,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09096077316657192,
+      "grad_norm": 8.349255186085697,
+      "learning_rate": 4.951505070935699e-06,
+      "loss": 0.6245,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0910419881426135,
+      "grad_norm": 4.572535564011483,
+      "learning_rate": 4.951376086032166e-06,
+      "loss": 0.7326,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09112320311865509,
+      "grad_norm": 4.793988248982188,
+      "learning_rate": 4.95124693150638e-06,
+      "loss": 0.6501,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09120441809469666,
+      "grad_norm": 4.303548964129903,
+      "learning_rate": 4.951117607367281e-06,
+      "loss": 0.507,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09128563307073824,
+      "grad_norm": 3.9997657093256613,
+      "learning_rate": 4.9509881136238144e-06,
+      "loss": 0.647,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09136684804677983,
+      "grad_norm": 6.458649539085403,
+      "learning_rate": 4.950858450284943e-06,
+      "loss": 0.7264,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09144806302282141,
+      "grad_norm": 8.005259138006465,
+      "learning_rate": 4.950728617359637e-06,
+      "loss": 0.7207,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09152927799886298,
+      "grad_norm": 10.787406113255107,
+      "learning_rate": 4.950598614856882e-06,
+      "loss": 0.7295,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09161049297490458,
+      "grad_norm": 7.384428417264019,
+      "learning_rate": 4.950468442785672e-06,
+      "loss": 0.7061,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09169170795094615,
+      "grad_norm": 6.688743726208685,
+      "learning_rate": 4.9503381011550145e-06,
+      "loss": 0.5785,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09177292292698774,
+      "grad_norm": 7.409666261351277,
+      "learning_rate": 4.950207589973929e-06,
+      "loss": 0.6196,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09185413790302932,
+      "grad_norm": 4.151052125921888,
+      "learning_rate": 4.950076909251445e-06,
+      "loss": 0.6793,
+      "step": 1131
+    },
+    {
+      "epoch": 0.0919353528790709,
+      "grad_norm": 6.945429669541111,
+      "learning_rate": 4.949946058996606e-06,
+      "loss": 0.7033,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09201656785511249,
+      "grad_norm": 5.827624293232605,
+      "learning_rate": 4.949815039218467e-06,
+      "loss": 0.673,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09209778283115407,
+      "grad_norm": 7.170155218321833,
+      "learning_rate": 4.949683849926092e-06,
+      "loss": 0.5155,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09217899780719564,
+      "grad_norm": 4.143518961770841,
+      "learning_rate": 4.949552491128559e-06,
+      "loss": 0.4783,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09226021278323723,
+      "grad_norm": 4.759778607114406,
+      "learning_rate": 4.9494209628349585e-06,
+      "loss": 0.7102,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09234142775927881,
+      "grad_norm": 4.218386626424805,
+      "learning_rate": 4.94928926505439e-06,
+      "loss": 0.7196,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09242264273532039,
+      "grad_norm": 3.168695853330223,
+      "learning_rate": 4.949157397795967e-06,
+      "loss": 0.5415,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09250385771136198,
+      "grad_norm": 3.826938431817392,
+      "learning_rate": 4.949025361068814e-06,
+      "loss": 0.7162,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09258507268740356,
+      "grad_norm": 6.290926748077408,
+      "learning_rate": 4.9488931548820685e-06,
+      "loss": 0.5282,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09266628766344515,
+      "grad_norm": 7.032750557333736,
+      "learning_rate": 4.9487607792448765e-06,
+      "loss": 0.5308,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09274750263948672,
+      "grad_norm": 5.947078730213946,
+      "learning_rate": 4.948628234166398e-06,
+      "loss": 0.5358,
+      "step": 1142
+    },
+    {
+      "epoch": 0.0928287176155283,
+      "grad_norm": 5.843171180813921,
+      "learning_rate": 4.948495519655805e-06,
+      "loss": 0.5523,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09290993259156989,
+      "grad_norm": 4.227961968923302,
+      "learning_rate": 4.948362635722281e-06,
+      "loss": 0.6363,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09299114756761147,
+      "grad_norm": 7.599291580837134,
+      "learning_rate": 4.948229582375021e-06,
+      "loss": 0.5243,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09307236254365304,
+      "grad_norm": 6.482980832600019,
+      "learning_rate": 4.948096359623229e-06,
+      "loss": 0.5535,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09315357751969464,
+      "grad_norm": 8.221446398590253,
+      "learning_rate": 4.9479629674761265e-06,
+      "loss": 0.603,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09323479249573621,
+      "grad_norm": 3.448295224589768,
+      "learning_rate": 4.947829405942942e-06,
+      "loss": 0.667,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09331600747177779,
+      "grad_norm": 3.719043167127091,
+      "learning_rate": 4.947695675032919e-06,
+      "loss": 0.5841,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09339722244781938,
+      "grad_norm": 4.156557050698605,
+      "learning_rate": 4.947561774755307e-06,
+      "loss": 0.6107,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09347843742386096,
+      "grad_norm": 4.192630167978793,
+      "learning_rate": 4.947427705119375e-06,
+      "loss": 0.4772,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09355965239990255,
+      "grad_norm": 5.508607665282113,
+      "learning_rate": 4.947293466134399e-06,
+      "loss": 0.6472,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09364086737594413,
+      "grad_norm": 6.1116074527140185,
+      "learning_rate": 4.947159057809668e-06,
+      "loss": 0.5252,
+      "step": 1153
+    },
+    {
+      "epoch": 0.0937220823519857,
+      "grad_norm": 4.348443564825216,
+      "learning_rate": 4.9470244801544794e-06,
+      "loss": 0.6287,
+      "step": 1154
+    },
+    {
+      "epoch": 0.0938032973280273,
+      "grad_norm": 4.398778822812722,
+      "learning_rate": 4.94688973317815e-06,
+      "loss": 0.6383,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09388451230406887,
+      "grad_norm": 8.481255729214475,
+      "learning_rate": 4.946754816889999e-06,
+      "loss": 0.6607,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09396572728011045,
+      "grad_norm": 8.572467423080703,
+      "learning_rate": 4.946619731299365e-06,
+      "loss": 0.5745,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09404694225615204,
+      "grad_norm": 7.588248988044493,
+      "learning_rate": 4.946484476415593e-06,
+      "loss": 0.5236,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09412815723219362,
+      "grad_norm": 6.803366010970415,
+      "learning_rate": 4.946349052248044e-06,
+      "loss": 0.4852,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09420937220823519,
+      "grad_norm": 3.614463840238927,
+      "learning_rate": 4.946213458806088e-06,
+      "loss": 0.4897,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09429058718427678,
+      "grad_norm": 4.806863459042568,
+      "learning_rate": 4.946077696099107e-06,
+      "loss": 0.6462,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09437180216031836,
+      "grad_norm": 5.367629362476872,
+      "learning_rate": 4.945941764136494e-06,
+      "loss": 0.5871,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09445301713635995,
+      "grad_norm": 3.4996137215165826,
+      "learning_rate": 4.945805662927657e-06,
+      "loss": 0.5799,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09453423211240153,
+      "grad_norm": 5.909531771722533,
+      "learning_rate": 4.9456693924820124e-06,
+      "loss": 0.672,
+      "step": 1164
+    },
+    {
+      "epoch": 0.0946154470884431,
+      "grad_norm": 11.436424144343768,
+      "learning_rate": 4.945532952808989e-06,
+      "loss": 0.5327,
+      "step": 1165
+    },
+    {
+      "epoch": 0.0946966620644847,
+      "grad_norm": 3.885963408729381,
+      "learning_rate": 4.945396343918027e-06,
+      "loss": 0.6624,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09477787704052627,
+      "grad_norm": 6.229973134745262,
+      "learning_rate": 4.945259565818582e-06,
+      "loss": 0.5427,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09485909201656785,
+      "grad_norm": 4.17803636708481,
+      "learning_rate": 4.9451226185201155e-06,
+      "loss": 0.5436,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09494030699260944,
+      "grad_norm": 3.7870816927723228,
+      "learning_rate": 4.9449855020321045e-06,
+      "loss": 0.6648,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09502152196865102,
+      "grad_norm": 5.063697581248885,
+      "learning_rate": 4.944848216364036e-06,
+      "loss": 0.6312,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0951027369446926,
+      "grad_norm": 7.690605712577253,
+      "learning_rate": 4.944710761525411e-06,
+      "loss": 0.62,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09518395192073419,
+      "grad_norm": 5.020085939206364,
+      "learning_rate": 4.94457313752574e-06,
+      "loss": 0.5703,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09526516689677576,
+      "grad_norm": 4.5490891920736525,
+      "learning_rate": 4.944435344374544e-06,
+      "loss": 0.7024,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09534638187281735,
+      "grad_norm": 5.913707485877068,
+      "learning_rate": 4.944297382081361e-06,
+      "loss": 0.5042,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09542759684885893,
+      "grad_norm": 3.759445496054164,
+      "learning_rate": 4.944159250655734e-06,
+      "loss": 0.7242,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09550881182490051,
+      "grad_norm": 4.721355611086034,
+      "learning_rate": 4.944020950107224e-06,
+      "loss": 0.6527,
+      "step": 1176
+    },
+    {
+      "epoch": 0.0955900268009421,
+      "grad_norm": 4.718530932065967,
+      "learning_rate": 4.943882480445398e-06,
+      "loss": 0.6651,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09567124177698368,
+      "grad_norm": 3.7323028514817764,
+      "learning_rate": 4.943743841679839e-06,
+      "loss": 0.4774,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09575245675302525,
+      "grad_norm": 7.764398624185714,
+      "learning_rate": 4.943605033820138e-06,
+      "loss": 0.5687,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09583367172906684,
+      "grad_norm": 5.7460830112551715,
+      "learning_rate": 4.943466056875903e-06,
+      "loss": 0.6266,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09591488670510842,
+      "grad_norm": 6.12456544392407,
+      "learning_rate": 4.943326910856749e-06,
+      "loss": 0.9285,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09599610168115,
+      "grad_norm": 5.9339500045832505,
+      "learning_rate": 4.943187595772302e-06,
+      "loss": 0.572,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09607731665719159,
+      "grad_norm": 5.327929545992472,
+      "learning_rate": 4.943048111632205e-06,
+      "loss": 0.7426,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09615853163323317,
+      "grad_norm": 4.422671758582798,
+      "learning_rate": 4.942908458446107e-06,
+      "loss": 0.5256,
+      "step": 1184
+    },
+    {
+      "epoch": 0.09623974660927476,
+      "grad_norm": 4.210710659594965,
+      "learning_rate": 4.942768636223674e-06,
+      "loss": 0.6544,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09632096158531633,
+      "grad_norm": 5.86594483722807,
+      "learning_rate": 4.94262864497458e-06,
+      "loss": 0.6285,
+      "step": 1186
+    },
+    {
+      "epoch": 0.09640217656135791,
+      "grad_norm": 11.95231315351679,
+      "learning_rate": 4.94248848470851e-06,
+      "loss": 0.5907,
+      "step": 1187
+    },
+    {
+      "epoch": 0.0964833915373995,
+      "grad_norm": 4.610476439616043,
+      "learning_rate": 4.9423481554351636e-06,
+      "loss": 0.5891,
+      "step": 1188
+    },
+    {
+      "epoch": 0.09656460651344108,
+      "grad_norm": 15.02936414694867,
+      "learning_rate": 4.9422076571642516e-06,
+      "loss": 0.6007,
+      "step": 1189
+    },
+    {
+      "epoch": 0.09664582148948266,
+      "grad_norm": 7.408105632703543,
+      "learning_rate": 4.942066989905494e-06,
+      "loss": 0.5263,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09672703646552425,
+      "grad_norm": 6.6983744546854105,
+      "learning_rate": 4.941926153668626e-06,
+      "loss": 0.6209,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09680825144156582,
+      "grad_norm": 12.37964028499769,
+      "learning_rate": 4.941785148463391e-06,
+      "loss": 0.5479,
+      "step": 1192
+    },
+    {
+      "epoch": 0.0968894664176074,
+      "grad_norm": 7.949353882372964,
+      "learning_rate": 4.941643974299547e-06,
+      "loss": 0.6356,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09697068139364899,
+      "grad_norm": 12.259554197141354,
+      "learning_rate": 4.941502631186863e-06,
+      "loss": 0.5876,
+      "step": 1194
+    },
+    {
+      "epoch": 0.09705189636969057,
+      "grad_norm": 4.5151134917162254,
+      "learning_rate": 4.941361119135118e-06,
+      "loss": 0.6081,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09713311134573216,
+      "grad_norm": 5.027871122019048,
+      "learning_rate": 4.941219438154103e-06,
+      "loss": 0.7848,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09721432632177374,
+      "grad_norm": 6.044100876833535,
+      "learning_rate": 4.941077588253624e-06,
+      "loss": 0.6873,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09729554129781531,
+      "grad_norm": 5.889644758237491,
+      "learning_rate": 4.940935569443496e-06,
+      "loss": 0.5557,
+      "step": 1198
+    },
+    {
+      "epoch": 0.0973767562738569,
+      "grad_norm": 15.297541275971485,
+      "learning_rate": 4.940793381733544e-06,
+      "loss": 0.547,
+      "step": 1199
+    },
+    {
+      "epoch": 0.09745797124989848,
+      "grad_norm": 12.207281968202853,
+      "learning_rate": 4.940651025133607e-06,
+      "loss": 0.7011,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09753918622594006,
+      "grad_norm": 5.195830064818431,
+      "learning_rate": 4.9405084996535376e-06,
+      "loss": 0.7094,
+      "step": 1201
+    },
+    {
+      "epoch": 0.09762040120198165,
+      "grad_norm": 4.519169304548104,
+      "learning_rate": 4.940365805303195e-06,
+      "loss": 0.5373,
+      "step": 1202
+    },
+    {
+      "epoch": 0.09770161617802323,
+      "grad_norm": 4.18578492619036,
+      "learning_rate": 4.940222942092455e-06,
+      "loss": 0.5373,
+      "step": 1203
+    },
+    {
+      "epoch": 0.0977828311540648,
+      "grad_norm": 5.871433014218688,
+      "learning_rate": 4.940079910031201e-06,
+      "loss": 0.6525,
+      "step": 1204
+    },
+    {
+      "epoch": 0.0978640461301064,
+      "grad_norm": 7.770798201538124,
+      "learning_rate": 4.939936709129333e-06,
+      "loss": 0.6541,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09794526110614797,
+      "grad_norm": 10.00163900865208,
+      "learning_rate": 4.939793339396756e-06,
+      "loss": 0.6166,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09802647608218956,
+      "grad_norm": 6.793365007116543,
+      "learning_rate": 4.939649800843394e-06,
+      "loss": 0.5834,
+      "step": 1207
+    },
+    {
+      "epoch": 0.09810769105823114,
+      "grad_norm": 3.8998039709967927,
+      "learning_rate": 4.939506093479176e-06,
+      "loss": 0.654,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09818890603427272,
+      "grad_norm": 3.5977727046620473,
+      "learning_rate": 4.939362217314048e-06,
+      "loss": 0.5947,
+      "step": 1209
+    },
+    {
+      "epoch": 0.0982701210103143,
+      "grad_norm": 3.430672786433987,
+      "learning_rate": 4.939218172357965e-06,
+      "loss": 0.4971,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09835133598635588,
+      "grad_norm": 4.703232976940314,
+      "learning_rate": 4.9390739586208926e-06,
+      "loss": 0.6256,
+      "step": 1211
+    },
+    {
+      "epoch": 0.09843255096239746,
+      "grad_norm": 8.889711877291688,
+      "learning_rate": 4.938929576112812e-06,
+      "loss": 0.6425,
+      "step": 1212
+    },
+    {
+      "epoch": 0.09851376593843905,
+      "grad_norm": 5.215270092203213,
+      "learning_rate": 4.938785024843712e-06,
+      "loss": 0.5402,
+      "step": 1213
+    },
+    {
+      "epoch": 0.09859498091448063,
+      "grad_norm": 4.480490370978996,
+      "learning_rate": 4.938640304823596e-06,
+      "loss": 0.4592,
+      "step": 1214
+    },
+    {
+      "epoch": 0.0986761958905222,
+      "grad_norm": 4.5775356603291995,
+      "learning_rate": 4.938495416062477e-06,
+      "loss": 0.5733,
+      "step": 1215
+    },
+    {
+      "epoch": 0.0987574108665638,
+      "grad_norm": 5.374607118889353,
+      "learning_rate": 4.93835035857038e-06,
+      "loss": 0.4699,
+      "step": 1216
+    },
+    {
+      "epoch": 0.09883862584260537,
+      "grad_norm": 8.062779666258928,
+      "learning_rate": 4.938205132357344e-06,
+      "loss": 0.6582,
+      "step": 1217
+    },
+    {
+      "epoch": 0.09891984081864696,
+      "grad_norm": 5.276936843727084,
+      "learning_rate": 4.938059737433416e-06,
+      "loss": 0.4957,
+      "step": 1218
+    },
+    {
+      "epoch": 0.09900105579468854,
+      "grad_norm": 5.321609718799565,
+      "learning_rate": 4.9379141738086575e-06,
+      "loss": 0.5664,
+      "step": 1219
+    },
+    {
+      "epoch": 0.09908227077073012,
+      "grad_norm": 4.479754054431104,
+      "learning_rate": 4.9377684414931415e-06,
+      "loss": 0.7467,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09916348574677171,
+      "grad_norm": 4.74529856971756,
+      "learning_rate": 4.937622540496951e-06,
+      "loss": 0.5793,
+      "step": 1221
+    },
+    {
+      "epoch": 0.09924470072281329,
+      "grad_norm": 3.911535284660022,
+      "learning_rate": 4.937476470830181e-06,
+      "loss": 0.6115,
+      "step": 1222
+    },
+    {
+      "epoch": 0.09932591569885486,
+      "grad_norm": 4.166632824497643,
+      "learning_rate": 4.937330232502939e-06,
+      "loss": 0.5927,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09940713067489645,
+      "grad_norm": 5.587700034937294,
+      "learning_rate": 4.937183825525346e-06,
+      "loss": 0.9838,
+      "step": 1224
+    },
+    {
+      "epoch": 0.09948834565093803,
+      "grad_norm": 4.01441412316951,
+      "learning_rate": 4.937037249907529e-06,
+      "loss": 0.6088,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09956956062697961,
+      "grad_norm": 5.422702256253332,
+      "learning_rate": 4.9368905056596336e-06,
+      "loss": 0.8051,
+      "step": 1226
+    },
+    {
+      "epoch": 0.0996507756030212,
+      "grad_norm": 4.730421753529755,
+      "learning_rate": 4.936743592791812e-06,
+      "loss": 0.8022,
+      "step": 1227
+    },
+    {
+      "epoch": 0.09973199057906278,
+      "grad_norm": 6.417481219598529,
+      "learning_rate": 4.936596511314229e-06,
+      "loss": 0.7359,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09981320555510437,
+      "grad_norm": 5.936442260399057,
+      "learning_rate": 4.936449261237064e-06,
+      "loss": 0.4835,
+      "step": 1229
+    },
+    {
+      "epoch": 0.09989442053114594,
+      "grad_norm": 5.161367234189236,
+      "learning_rate": 4.936301842570505e-06,
+      "loss": 0.6098,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09997563550718752,
+      "grad_norm": 6.171812579141409,
+      "learning_rate": 4.936154255324751e-06,
+      "loss": 0.5208,
+      "step": 1231
+    },
+    {
+      "epoch": 0.10005685048322911,
+      "grad_norm": 6.898043493723934,
+      "learning_rate": 4.936006499510016e-06,
+      "loss": 0.6272,
+      "step": 1232
+    },
+    {
+      "epoch": 0.10013806545927069,
+      "grad_norm": 5.279799808361505,
+      "learning_rate": 4.935858575136525e-06,
+      "loss": 0.6761,
+      "step": 1233
+    },
+    {
+      "epoch": 0.10021928043531227,
+      "grad_norm": 6.1780863372983,
+      "learning_rate": 4.935710482214512e-06,
+      "loss": 0.5666,
+      "step": 1234
+    },
+    {
+      "epoch": 0.10030049541135386,
+      "grad_norm": 4.127005405020764,
+      "learning_rate": 4.935562220754224e-06,
+      "loss": 0.7762,
+      "step": 1235
+    },
+    {
+      "epoch": 0.10038171038739543,
+      "grad_norm": 6.939945712344684,
+      "learning_rate": 4.935413790765919e-06,
+      "loss": 0.5601,
+      "step": 1236
+    },
+    {
+      "epoch": 0.10046292536343701,
+      "grad_norm": 5.4373785364212965,
+      "learning_rate": 4.935265192259871e-06,
+      "loss": 0.5489,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1005441403394786,
+      "grad_norm": 3.19887214936069,
+      "learning_rate": 4.935116425246359e-06,
+      "loss": 0.6456,
+      "step": 1238
+    },
+    {
+      "epoch": 0.10062535531552018,
+      "grad_norm": 7.543707007832977,
+      "learning_rate": 4.934967489735679e-06,
+      "loss": 0.5061,
+      "step": 1239
+    },
+    {
+      "epoch": 0.10070657029156177,
+      "grad_norm": 4.008848618860738,
+      "learning_rate": 4.934818385738135e-06,
+      "loss": 0.6719,
+      "step": 1240
+    },
+    {
+      "epoch": 0.10078778526760335,
+      "grad_norm": 4.006099731057698,
+      "learning_rate": 4.934669113264044e-06,
+      "loss": 0.6852,
+      "step": 1241
+    },
+    {
+      "epoch": 0.10086900024364492,
+      "grad_norm": 8.793769739342025,
+      "learning_rate": 4.934519672323737e-06,
+      "loss": 0.5916,
+      "step": 1242
+    },
+    {
+      "epoch": 0.10095021521968651,
+      "grad_norm": 5.1218716406597835,
+      "learning_rate": 4.9343700629275525e-06,
+      "loss": 0.4989,
+      "step": 1243
+    },
+    {
+      "epoch": 0.10103143019572809,
+      "grad_norm": 4.3513298577507245,
+      "learning_rate": 4.934220285085843e-06,
+      "loss": 0.5374,
+      "step": 1244
+    },
+    {
+      "epoch": 0.10111264517176967,
+      "grad_norm": 5.858277317975228,
+      "learning_rate": 4.934070338808974e-06,
+      "loss": 0.4365,
+      "step": 1245
+    },
+    {
+      "epoch": 0.10119386014781126,
+      "grad_norm": 4.971141537823225,
+      "learning_rate": 4.933920224107319e-06,
+      "loss": 0.6175,
+      "step": 1246
+    },
+    {
+      "epoch": 0.10127507512385284,
+      "grad_norm": 7.572976492490089,
+      "learning_rate": 4.933769940991266e-06,
+      "loss": 0.6484,
+      "step": 1247
+    },
+    {
+      "epoch": 0.10135629009989441,
+      "grad_norm": 7.466443178328625,
+      "learning_rate": 4.933619489471213e-06,
+      "loss": 0.6078,
+      "step": 1248
+    },
+    {
+      "epoch": 0.101437505075936,
+      "grad_norm": 3.837978031503154,
+      "learning_rate": 4.933468869557572e-06,
+      "loss": 0.672,
+      "step": 1249
+    },
+    {
+      "epoch": 0.10151872005197758,
+      "grad_norm": 4.587774401604302,
+      "learning_rate": 4.933318081260763e-06,
+      "loss": 0.6828,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10159993502801917,
+      "grad_norm": 5.728533677177494,
+      "learning_rate": 4.933167124591222e-06,
+      "loss": 0.5304,
+      "step": 1251
+    },
+    {
+      "epoch": 0.10168115000406075,
+      "grad_norm": 7.693599571430989,
+      "learning_rate": 4.9330159995593926e-06,
+      "loss": 0.6125,
+      "step": 1252
+    },
+    {
+      "epoch": 0.10176236498010233,
+      "grad_norm": 5.284936139752091,
+      "learning_rate": 4.9328647061757326e-06,
+      "loss": 0.6306,
+      "step": 1253
+    },
+    {
+      "epoch": 0.10184357995614392,
+      "grad_norm": 4.502661220996693,
+      "learning_rate": 4.932713244450712e-06,
+      "loss": 0.6503,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1019247949321855,
+      "grad_norm": 30.435010504189158,
+      "learning_rate": 4.932561614394809e-06,
+      "loss": 0.4843,
+      "step": 1255
+    },
+    {
+      "epoch": 0.10200600990822707,
+      "grad_norm": 3.2680414669887226,
+      "learning_rate": 4.932409816018516e-06,
+      "loss": 0.5335,
+      "step": 1256
+    },
+    {
+      "epoch": 0.10208722488426866,
+      "grad_norm": 5.550536415470609,
+      "learning_rate": 4.932257849332337e-06,
+      "loss": 0.6274,
+      "step": 1257
+    },
+    {
+      "epoch": 0.10216843986031024,
+      "grad_norm": 6.490325484791516,
+      "learning_rate": 4.932105714346788e-06,
+      "loss": 0.7055,
+      "step": 1258
+    },
+    {
+      "epoch": 0.10224965483635182,
+      "grad_norm": 4.635391717749872,
+      "learning_rate": 4.931953411072395e-06,
+      "loss": 0.7677,
+      "step": 1259
+    },
+    {
+      "epoch": 0.10233086981239341,
+      "grad_norm": 5.213698736457178,
+      "learning_rate": 4.931800939519697e-06,
+      "loss": 0.6283,
+      "step": 1260
+    },
+    {
+      "epoch": 0.10241208478843498,
+      "grad_norm": 12.81670585201401,
+      "learning_rate": 4.931648299699245e-06,
+      "loss": 0.5411,
+      "step": 1261
+    },
+    {
+      "epoch": 0.10249329976447658,
+      "grad_norm": 4.7979729375829425,
+      "learning_rate": 4.931495491621598e-06,
+      "loss": 0.6,
+      "step": 1262
+    },
+    {
+      "epoch": 0.10257451474051815,
+      "grad_norm": 5.738885766711591,
+      "learning_rate": 4.931342515297333e-06,
+      "loss": 0.6029,
+      "step": 1263
+    },
+    {
+      "epoch": 0.10265572971655973,
+      "grad_norm": 4.5708865991196586,
+      "learning_rate": 4.931189370737033e-06,
+      "loss": 0.6877,
+      "step": 1264
+    },
+    {
+      "epoch": 0.10273694469260132,
+      "grad_norm": 4.292203823138975,
+      "learning_rate": 4.931036057951295e-06,
+      "loss": 0.6054,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1028181596686429,
+      "grad_norm": 5.094248884490348,
+      "learning_rate": 4.930882576950728e-06,
+      "loss": 1.016,
+      "step": 1266
+    },
+    {
+      "epoch": 0.10289937464468447,
+      "grad_norm": 7.74744023190108,
+      "learning_rate": 4.930728927745954e-06,
+      "loss": 0.6266,
+      "step": 1267
+    },
+    {
+      "epoch": 0.10298058962072607,
+      "grad_norm": 7.178060412399686,
+      "learning_rate": 4.930575110347601e-06,
+      "loss": 0.5436,
+      "step": 1268
+    },
+    {
+      "epoch": 0.10306180459676764,
+      "grad_norm": 4.8955814375005,
+      "learning_rate": 4.9304211247663135e-06,
+      "loss": 0.6069,
+      "step": 1269
+    },
+    {
+      "epoch": 0.10314301957280922,
+      "grad_norm": 6.4873010558747675,
+      "learning_rate": 4.930266971012748e-06,
+      "loss": 0.5739,
+      "step": 1270
+    },
+    {
+      "epoch": 0.10322423454885081,
+      "grad_norm": 5.62659346934576,
+      "learning_rate": 4.930112649097569e-06,
+      "loss": 0.7317,
+      "step": 1271
+    },
+    {
+      "epoch": 0.10330544952489239,
+      "grad_norm": 5.596039021432254,
+      "learning_rate": 4.929958159031457e-06,
+      "loss": 0.4922,
+      "step": 1272
+    },
+    {
+      "epoch": 0.10338666450093398,
+      "grad_norm": 5.488922386676459,
+      "learning_rate": 4.9298035008251e-06,
+      "loss": 0.5276,
+      "step": 1273
+    },
+    {
+      "epoch": 0.10346787947697555,
+      "grad_norm": 5.79593069742807,
+      "learning_rate": 4.929648674489201e-06,
+      "loss": 0.5529,
+      "step": 1274
+    },
+    {
+      "epoch": 0.10354909445301713,
+      "grad_norm": 9.122388238453604,
+      "learning_rate": 4.929493680034472e-06,
+      "loss": 0.6647,
+      "step": 1275
+    },
+    {
+      "epoch": 0.10363030942905872,
+      "grad_norm": 4.654226716442984,
+      "learning_rate": 4.929338517471638e-06,
+      "loss": 0.6721,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1037115244051003,
+      "grad_norm": 4.921506870025976,
+      "learning_rate": 4.929183186811436e-06,
+      "loss": 0.4812,
+      "step": 1277
+    },
+    {
+      "epoch": 0.10379273938114188,
+      "grad_norm": 8.206987541104693,
+      "learning_rate": 4.9290276880646144e-06,
+      "loss": 0.6183,
+      "step": 1278
+    },
+    {
+      "epoch": 0.10387395435718347,
+      "grad_norm": 6.664251212109854,
+      "learning_rate": 4.928872021241932e-06,
+      "loss": 0.5676,
+      "step": 1279
+    },
+    {
+      "epoch": 0.10395516933322504,
+      "grad_norm": 5.059968838894893,
+      "learning_rate": 4.92871618635416e-06,
+      "loss": 0.5516,
+      "step": 1280
+    },
+    {
+      "epoch": 0.10403638430926662,
+      "grad_norm": 4.462582725976913,
+      "learning_rate": 4.928560183412081e-06,
+      "loss": 0.637,
+      "step": 1281
+    },
+    {
+      "epoch": 0.10411759928530821,
+      "grad_norm": 13.013107165252704,
+      "learning_rate": 4.928404012426491e-06,
+      "loss": 0.585,
+      "step": 1282
+    },
+    {
+      "epoch": 0.10419881426134979,
+      "grad_norm": 4.877077629277337,
+      "learning_rate": 4.9282476734081955e-06,
+      "loss": 0.4416,
+      "step": 1283
+    },
+    {
+      "epoch": 0.10428002923739138,
+      "grad_norm": 5.339858862218509,
+      "learning_rate": 4.928091166368013e-06,
+      "loss": 0.534,
+      "step": 1284
+    },
+    {
+      "epoch": 0.10436124421343296,
+      "grad_norm": 4.638726269664911,
+      "learning_rate": 4.927934491316771e-06,
+      "loss": 0.5402,
+      "step": 1285
+    },
+    {
+      "epoch": 0.10444245918947453,
+      "grad_norm": 5.279866806349571,
+      "learning_rate": 4.927777648265313e-06,
+      "loss": 0.8809,
+      "step": 1286
+    },
+    {
+      "epoch": 0.10452367416551613,
+      "grad_norm": 10.172012498738642,
+      "learning_rate": 4.927620637224489e-06,
+      "loss": 0.6804,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1046048891415577,
+      "grad_norm": 4.0125921901692605,
+      "learning_rate": 4.927463458205167e-06,
+      "loss": 0.484,
+      "step": 1288
+    },
+    {
+      "epoch": 0.10468610411759928,
+      "grad_norm": 5.038605450511745,
+      "learning_rate": 4.9273061112182195e-06,
+      "loss": 0.4823,
+      "step": 1289
+    },
+    {
+      "epoch": 0.10476731909364087,
+      "grad_norm": 11.514242344873601,
+      "learning_rate": 4.9271485962745356e-06,
+      "loss": 0.5736,
+      "step": 1290
+    },
+    {
+      "epoch": 0.10484853406968245,
+      "grad_norm": 5.2507131812652705,
+      "learning_rate": 4.9269909133850146e-06,
+      "loss": 0.4464,
+      "step": 1291
+    },
+    {
+      "epoch": 0.10492974904572404,
+      "grad_norm": 3.587842597108573,
+      "learning_rate": 4.926833062560566e-06,
+      "loss": 0.5488,
+      "step": 1292
+    },
+    {
+      "epoch": 0.10501096402176562,
+      "grad_norm": 4.9325898595523405,
+      "learning_rate": 4.926675043812115e-06,
+      "loss": 0.7937,
+      "step": 1293
+    },
+    {
+      "epoch": 0.10509217899780719,
+      "grad_norm": 4.089069618738186,
+      "learning_rate": 4.926516857150593e-06,
+      "loss": 0.6927,
+      "step": 1294
+    },
+    {
+      "epoch": 0.10517339397384878,
+      "grad_norm": 4.157543455041517,
+      "learning_rate": 4.926358502586948e-06,
+      "loss": 0.6499,
+      "step": 1295
+    },
+    {
+      "epoch": 0.10525460894989036,
+      "grad_norm": 4.591475297442821,
+      "learning_rate": 4.9261999801321345e-06,
+      "loss": 0.6702,
+      "step": 1296
+    },
+    {
+      "epoch": 0.10533582392593194,
+      "grad_norm": 5.076099157718857,
+      "learning_rate": 4.9260412897971225e-06,
+      "loss": 0.5649,
+      "step": 1297
+    },
+    {
+      "epoch": 0.10541703890197353,
+      "grad_norm": 5.92574396887693,
+      "learning_rate": 4.9258824315928935e-06,
+      "loss": 0.6207,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1054982538780151,
+      "grad_norm": 4.295889777812483,
+      "learning_rate": 4.925723405530439e-06,
+      "loss": 0.6171,
+      "step": 1299
+    },
+    {
+      "epoch": 0.10557946885405668,
+      "grad_norm": 6.0185008543132055,
+      "learning_rate": 4.925564211620764e-06,
+      "loss": 0.6708,
+      "step": 1300
+    },
+    {
+      "epoch": 0.10566068383009827,
+      "grad_norm": 4.288628260503543,
+      "learning_rate": 4.9254048498748804e-06,
+      "loss": 0.5312,
+      "step": 1301
+    },
+    {
+      "epoch": 0.10574189880613985,
+      "grad_norm": 5.912622580277487,
+      "learning_rate": 4.925245320303819e-06,
+      "loss": 0.6256,
+      "step": 1302
+    },
+    {
+      "epoch": 0.10582311378218144,
+      "grad_norm": 4.036381733012363,
+      "learning_rate": 4.925085622918618e-06,
+      "loss": 0.6512,
+      "step": 1303
+    },
+    {
+      "epoch": 0.10590432875822302,
+      "grad_norm": 11.770792105747262,
+      "learning_rate": 4.924925757730324e-06,
+      "loss": 0.6243,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1059855437342646,
+      "grad_norm": 7.6314249401022085,
+      "learning_rate": 4.924765724750002e-06,
+      "loss": 0.5521,
+      "step": 1305
+    },
+    {
+      "epoch": 0.10606675871030619,
+      "grad_norm": 4.238420473450789,
+      "learning_rate": 4.9246055239887255e-06,
+      "loss": 0.626,
+      "step": 1306
+    },
+    {
+      "epoch": 0.10614797368634776,
+      "grad_norm": 5.023366885034705,
+      "learning_rate": 4.924445155457578e-06,
+      "loss": 0.7469,
+      "step": 1307
+    },
+    {
+      "epoch": 0.10622918866238934,
+      "grad_norm": 5.927044022062121,
+      "learning_rate": 4.924284619167657e-06,
+      "loss": 0.6673,
+      "step": 1308
+    },
+    {
+      "epoch": 0.10631040363843093,
+      "grad_norm": 27.667866009821907,
+      "learning_rate": 4.924123915130072e-06,
+      "loss": 0.4591,
+      "step": 1309
+    },
+    {
+      "epoch": 0.10639161861447251,
+      "grad_norm": 6.067033293424752,
+      "learning_rate": 4.92396304335594e-06,
+      "loss": 0.7129,
+      "step": 1310
+    },
+    {
+      "epoch": 0.10647283359051408,
+      "grad_norm": 4.48662582291358,
+      "learning_rate": 4.923802003856395e-06,
+      "loss": 0.7251,
+      "step": 1311
+    },
+    {
+      "epoch": 0.10655404856655568,
+      "grad_norm": 8.910396684118473,
+      "learning_rate": 4.923640796642578e-06,
+      "loss": 0.5621,
+      "step": 1312
+    },
+    {
+      "epoch": 0.10663526354259725,
+      "grad_norm": 10.709244615053429,
+      "learning_rate": 4.923479421725646e-06,
+      "loss": 0.7027,
+      "step": 1313
+    },
+    {
+      "epoch": 0.10671647851863884,
+      "grad_norm": 6.558186461825374,
+      "learning_rate": 4.923317879116764e-06,
+      "loss": 0.6991,
+      "step": 1314
+    },
+    {
+      "epoch": 0.10679769349468042,
+      "grad_norm": 4.402808201885828,
+      "learning_rate": 4.923156168827109e-06,
+      "loss": 0.5193,
+      "step": 1315
+    },
+    {
+      "epoch": 0.106878908470722,
+      "grad_norm": 3.9531074248473677,
+      "learning_rate": 4.922994290867872e-06,
+      "loss": 0.5984,
+      "step": 1316
+    },
+    {
+      "epoch": 0.10696012344676359,
+      "grad_norm": 5.634957726444024,
+      "learning_rate": 4.922832245250254e-06,
+      "loss": 0.5759,
+      "step": 1317
+    },
+    {
+      "epoch": 0.10704133842280517,
+      "grad_norm": 4.687012040425469,
+      "learning_rate": 4.922670031985467e-06,
+      "loss": 0.5291,
+      "step": 1318
+    },
+    {
+      "epoch": 0.10712255339884674,
+      "grad_norm": 7.331234418642,
+      "learning_rate": 4.922507651084736e-06,
+      "loss": 0.631,
+      "step": 1319
+    },
+    {
+      "epoch": 0.10720376837488833,
+      "grad_norm": 6.179571341056749,
+      "learning_rate": 4.9223451025592965e-06,
+      "loss": 0.5047,
+      "step": 1320
+    },
+    {
+      "epoch": 0.10728498335092991,
+      "grad_norm": 4.797755420619416,
+      "learning_rate": 4.9221823864203955e-06,
+      "loss": 0.5616,
+      "step": 1321
+    },
+    {
+      "epoch": 0.10736619832697149,
+      "grad_norm": 4.7796380411767885,
+      "learning_rate": 4.922019502679292e-06,
+      "loss": 0.7038,
+      "step": 1322
+    },
+    {
+      "epoch": 0.10744741330301308,
+      "grad_norm": 4.880297830435732,
+      "learning_rate": 4.921856451347258e-06,
+      "loss": 0.8187,
+      "step": 1323
+    },
+    {
+      "epoch": 0.10752862827905466,
+      "grad_norm": 4.4975518069828295,
+      "learning_rate": 4.9216932324355755e-06,
+      "loss": 0.6288,
+      "step": 1324
+    },
+    {
+      "epoch": 0.10760984325509625,
+      "grad_norm": 5.693686439938185,
+      "learning_rate": 4.921529845955537e-06,
+      "loss": 0.5948,
+      "step": 1325
+    },
+    {
+      "epoch": 0.10769105823113782,
+      "grad_norm": 9.014157893199854,
+      "learning_rate": 4.9213662919184495e-06,
+      "loss": 0.7244,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1077722732071794,
+      "grad_norm": 17.656772040857238,
+      "learning_rate": 4.921202570335629e-06,
+      "loss": 0.5731,
+      "step": 1327
+    },
+    {
+      "epoch": 0.10785348818322099,
+      "grad_norm": 7.338056929014862,
+      "learning_rate": 4.921038681218405e-06,
+      "loss": 0.5082,
+      "step": 1328
+    },
+    {
+      "epoch": 0.10793470315926257,
+      "grad_norm": 4.851712422566327,
+      "learning_rate": 4.920874624578118e-06,
+      "loss": 0.7288,
+      "step": 1329
+    },
+    {
+      "epoch": 0.10801591813530415,
+      "grad_norm": 5.876176027514485,
+      "learning_rate": 4.920710400426118e-06,
+      "loss": 0.4265,
+      "step": 1330
+    },
+    {
+      "epoch": 0.10809713311134574,
+      "grad_norm": 5.880882787741879,
+      "learning_rate": 4.920546008773771e-06,
+      "loss": 0.4902,
+      "step": 1331
+    },
+    {
+      "epoch": 0.10817834808738731,
+      "grad_norm": 5.58193279698704,
+      "learning_rate": 4.920381449632451e-06,
+      "loss": 0.7027,
+      "step": 1332
+    },
+    {
+      "epoch": 0.10825956306342889,
+      "grad_norm": 5.173776451487993,
+      "learning_rate": 4.920216723013544e-06,
+      "loss": 0.646,
+      "step": 1333
+    },
+    {
+      "epoch": 0.10834077803947048,
+      "grad_norm": 4.388587188343882,
+      "learning_rate": 4.920051828928448e-06,
+      "loss": 0.4977,
+      "step": 1334
+    },
+    {
+      "epoch": 0.10842199301551206,
+      "grad_norm": 4.776839443402457,
+      "learning_rate": 4.919886767388573e-06,
+      "loss": 0.5767,
+      "step": 1335
+    },
+    {
+      "epoch": 0.10850320799155365,
+      "grad_norm": 3.4824258344375725,
+      "learning_rate": 4.919721538405341e-06,
+      "loss": 0.533,
+      "step": 1336
+    },
+    {
+      "epoch": 0.10858442296759523,
+      "grad_norm": 4.742639309949176,
+      "learning_rate": 4.919556141990186e-06,
+      "loss": 0.6688,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1086656379436368,
+      "grad_norm": 4.653988617546264,
+      "learning_rate": 4.919390578154551e-06,
+      "loss": 0.631,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1087468529196784,
+      "grad_norm": 5.238136207553621,
+      "learning_rate": 4.919224846909891e-06,
+      "loss": 0.5437,
+      "step": 1339
+    },
+    {
+      "epoch": 0.10882806789571997,
+      "grad_norm": 6.815351329648775,
+      "learning_rate": 4.919058948267677e-06,
+      "loss": 0.8166,
+      "step": 1340
+    },
+    {
+      "epoch": 0.10890928287176155,
+      "grad_norm": 5.2640299180243355,
+      "learning_rate": 4.918892882239384e-06,
+      "loss": 0.6044,
+      "step": 1341
+    },
+    {
+      "epoch": 0.10899049784780314,
+      "grad_norm": 4.358517148619114,
+      "learning_rate": 4.918726648836507e-06,
+      "loss": 0.7873,
+      "step": 1342
+    },
+    {
+      "epoch": 0.10907171282384472,
+      "grad_norm": 6.052945574950095,
+      "learning_rate": 4.918560248070547e-06,
+      "loss": 0.5615,
+      "step": 1343
+    },
+    {
+      "epoch": 0.10915292779988629,
+      "grad_norm": 4.566253758686044,
+      "learning_rate": 4.918393679953018e-06,
+      "loss": 0.5893,
+      "step": 1344
+    },
+    {
+      "epoch": 0.10923414277592788,
+      "grad_norm": 3.497997250474178,
+      "learning_rate": 4.918226944495445e-06,
+      "loss": 0.6582,
+      "step": 1345
+    },
+    {
+      "epoch": 0.10931535775196946,
+      "grad_norm": 6.920939485791569,
+      "learning_rate": 4.918060041709366e-06,
+      "loss": 0.563,
+      "step": 1346
+    },
+    {
+      "epoch": 0.10939657272801105,
+      "grad_norm": 13.15024709863077,
+      "learning_rate": 4.917892971606329e-06,
+      "loss": 0.5954,
+      "step": 1347
+    },
+    {
+      "epoch": 0.10947778770405263,
+      "grad_norm": 4.208801375695993,
+      "learning_rate": 4.917725734197896e-06,
+      "loss": 0.548,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1095590026800942,
+      "grad_norm": 4.648874348064858,
+      "learning_rate": 4.917558329495636e-06,
+      "loss": 0.6602,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1096402176561358,
+      "grad_norm": 4.308899285098474,
+      "learning_rate": 4.917390757511136e-06,
+      "loss": 0.5401,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10972143263217737,
+      "grad_norm": 5.129516631257819,
+      "learning_rate": 4.917223018255989e-06,
+      "loss": 0.5193,
+      "step": 1351
+    },
+    {
+      "epoch": 0.10980264760821895,
+      "grad_norm": 6.107816197257616,
+      "learning_rate": 4.917055111741802e-06,
+      "loss": 0.5366,
+      "step": 1352
+    },
+    {
+      "epoch": 0.10988386258426054,
+      "grad_norm": 4.707618481744731,
+      "learning_rate": 4.916887037980193e-06,
+      "loss": 0.5354,
+      "step": 1353
+    },
+    {
+      "epoch": 0.10996507756030212,
+      "grad_norm": 4.1837375718082255,
+      "learning_rate": 4.916718796982793e-06,
+      "loss": 0.7129,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1100462925363437,
+      "grad_norm": 3.4672179946418975,
+      "learning_rate": 4.916550388761242e-06,
+      "loss": 0.5567,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11012750751238529,
+      "grad_norm": 4.01249160395805,
+      "learning_rate": 4.916381813327194e-06,
+      "loss": 0.5612,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11020872248842686,
+      "grad_norm": 4.478566912854505,
+      "learning_rate": 4.916213070692312e-06,
+      "loss": 0.5274,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11028993746446845,
+      "grad_norm": 7.752567209797171,
+      "learning_rate": 4.916044160868273e-06,
+      "loss": 0.5645,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11037115244051003,
+      "grad_norm": 4.915949482614922,
+      "learning_rate": 4.915875083866766e-06,
+      "loss": 0.5816,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11045236741655161,
+      "grad_norm": 4.872300208909787,
+      "learning_rate": 4.915705839699488e-06,
+      "loss": 0.5693,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1105335823925932,
+      "grad_norm": 6.8669686866426645,
+      "learning_rate": 4.915536428378152e-06,
+      "loss": 0.6373,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11061479736863478,
+      "grad_norm": 9.217173762118762,
+      "learning_rate": 4.915366849914479e-06,
+      "loss": 0.5895,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11069601234467635,
+      "grad_norm": 5.851967895961872,
+      "learning_rate": 4.915197104320203e-06,
+      "loss": 0.6099,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11077722732071794,
+      "grad_norm": 6.051580089476815,
+      "learning_rate": 4.915027191607069e-06,
+      "loss": 0.6278,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11085844229675952,
+      "grad_norm": 4.182706621669257,
+      "learning_rate": 4.914857111786835e-06,
+      "loss": 0.7995,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1109396572728011,
+      "grad_norm": 6.4377665961998,
+      "learning_rate": 4.9146868648712694e-06,
+      "loss": 0.5338,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11102087224884269,
+      "grad_norm": 4.1795950899223255,
+      "learning_rate": 4.914516450872152e-06,
+      "loss": 0.4989,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11110208722488427,
+      "grad_norm": 5.590546828289337,
+      "learning_rate": 4.914345869801276e-06,
+      "loss": 0.5973,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11118330220092586,
+      "grad_norm": 6.538514315603108,
+      "learning_rate": 4.914175121670443e-06,
+      "loss": 0.5177,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11126451717696743,
+      "grad_norm": 3.691433117466898,
+      "learning_rate": 4.914004206491467e-06,
+      "loss": 0.5844,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11134573215300901,
+      "grad_norm": 4.522420652396943,
+      "learning_rate": 4.913833124276177e-06,
+      "loss": 0.5717,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1114269471290506,
+      "grad_norm": 5.582588983452335,
+      "learning_rate": 4.9136618750364105e-06,
+      "loss": 0.5929,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11150816210509218,
+      "grad_norm": 3.5822825901443145,
+      "learning_rate": 4.913490458784016e-06,
+      "loss": 0.6584,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11158937708113376,
+      "grad_norm": 6.955323322713761,
+      "learning_rate": 4.913318875530855e-06,
+      "loss": 0.7448,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11167059205717535,
+      "grad_norm": 7.999732286808689,
+      "learning_rate": 4.9131471252887995e-06,
+      "loss": 0.5943,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11175180703321692,
+      "grad_norm": 5.699553884482589,
+      "learning_rate": 4.912975208069735e-06,
+      "loss": 0.5888,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1118330220092585,
+      "grad_norm": 9.937720162090995,
+      "learning_rate": 4.912803123885555e-06,
+      "loss": 0.579,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11191423698530009,
+      "grad_norm": 6.426480257301145,
+      "learning_rate": 4.912630872748171e-06,
+      "loss": 0.4428,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11199545196134167,
+      "grad_norm": 4.6331109760310625,
+      "learning_rate": 4.912458454669498e-06,
+      "loss": 0.5904,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11207666693738326,
+      "grad_norm": 5.28940520266137,
+      "learning_rate": 4.912285869661467e-06,
+      "loss": 0.5371,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11215788191342484,
+      "grad_norm": 3.8233787152045418,
+      "learning_rate": 4.912113117736022e-06,
+      "loss": 0.7556,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11223909688946641,
+      "grad_norm": 5.065689586582994,
+      "learning_rate": 4.911940198905114e-06,
+      "loss": 0.6213,
+      "step": 1382
+    },
+    {
+      "epoch": 0.112320311865508,
+      "grad_norm": 6.1114483413633245,
+      "learning_rate": 4.91176711318071e-06,
+      "loss": 0.6723,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11240152684154958,
+      "grad_norm": 7.783537136958568,
+      "learning_rate": 4.911593860574785e-06,
+      "loss": 0.5786,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11248274181759116,
+      "grad_norm": 8.312210011930208,
+      "learning_rate": 4.911420441099329e-06,
+      "loss": 0.6476,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11256395679363275,
+      "grad_norm": 4.230328963619588,
+      "learning_rate": 4.911246854766341e-06,
+      "loss": 0.686,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11264517176967433,
+      "grad_norm": 12.499607763263779,
+      "learning_rate": 4.911073101587831e-06,
+      "loss": 0.4509,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1127263867457159,
+      "grad_norm": 4.97136722337538,
+      "learning_rate": 4.9108991815758225e-06,
+      "loss": 0.8017,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1128076017217575,
+      "grad_norm": 4.68367623760023,
+      "learning_rate": 4.9107250947423516e-06,
+      "loss": 0.574,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11288881669779907,
+      "grad_norm": 4.5482183003098,
+      "learning_rate": 4.910550841099462e-06,
+      "loss": 0.6234,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11297003167384066,
+      "grad_norm": 3.734134582087491,
+      "learning_rate": 4.910376420659211e-06,
+      "loss": 0.7674,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11305124664988224,
+      "grad_norm": 9.222540748041622,
+      "learning_rate": 4.91020183343367e-06,
+      "loss": 0.6176,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11313246162592382,
+      "grad_norm": 7.2087300631114095,
+      "learning_rate": 4.910027079434917e-06,
+      "loss": 0.7806,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11321367660196541,
+      "grad_norm": 4.209780015944576,
+      "learning_rate": 4.909852158675045e-06,
+      "loss": 0.6478,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11329489157800698,
+      "grad_norm": 4.672108189183994,
+      "learning_rate": 4.9096770711661575e-06,
+      "loss": 0.7015,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11337610655404856,
+      "grad_norm": 5.391522206407413,
+      "learning_rate": 4.90950181692037e-06,
+      "loss": 0.6014,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11345732153009015,
+      "grad_norm": 6.512018993656574,
+      "learning_rate": 4.909326395949809e-06,
+      "loss": 0.6083,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11353853650613173,
+      "grad_norm": 5.621883708458305,
+      "learning_rate": 4.909150808266613e-06,
+      "loss": 0.6495,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1136197514821733,
+      "grad_norm": 7.16365741465048,
+      "learning_rate": 4.908975053882931e-06,
+      "loss": 0.5368,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1137009664582149,
+      "grad_norm": 5.860792942986273,
+      "learning_rate": 4.908799132810924e-06,
+      "loss": 0.7422,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11378218143425647,
+      "grad_norm": 3.673990330967191,
+      "learning_rate": 4.9086230450627655e-06,
+      "loss": 0.61,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11386339641029807,
+      "grad_norm": 3.927055438430478,
+      "learning_rate": 4.908446790650641e-06,
+      "loss": 0.7079,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11394461138633964,
+      "grad_norm": 7.144929200227564,
+      "learning_rate": 4.908270369586744e-06,
+      "loss": 0.5993,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11402582636238122,
+      "grad_norm": 4.458920511681134,
+      "learning_rate": 4.908093781883283e-06,
+      "loss": 0.7028,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11410704133842281,
+      "grad_norm": 4.699619928678632,
+      "learning_rate": 4.9079170275524765e-06,
+      "loss": 0.5911,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11418825631446439,
+      "grad_norm": 5.941174948945086,
+      "learning_rate": 4.907740106606557e-06,
+      "loss": 0.5615,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11426947129050596,
+      "grad_norm": 6.471908401673319,
+      "learning_rate": 4.9075630190577634e-06,
+      "loss": 0.5378,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11435068626654755,
+      "grad_norm": 3.9643546125770825,
+      "learning_rate": 4.907385764918351e-06,
+      "loss": 0.6547,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11443190124258913,
+      "grad_norm": 4.878410036784967,
+      "learning_rate": 4.907208344200585e-06,
+      "loss": 0.6994,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11451311621863071,
+      "grad_norm": 5.156488131990337,
+      "learning_rate": 4.907030756916741e-06,
+      "loss": 0.5712,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1145943311946723,
+      "grad_norm": 4.790335390127153,
+      "learning_rate": 4.906853003079108e-06,
+      "loss": 0.6316,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11467554617071388,
+      "grad_norm": 4.153017717675372,
+      "learning_rate": 4.9066750826999855e-06,
+      "loss": 0.6906,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11475676114675547,
+      "grad_norm": 3.87389513168029,
+      "learning_rate": 4.906496995791684e-06,
+      "loss": 0.512,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11483797612279704,
+      "grad_norm": 4.564186801844,
+      "learning_rate": 4.906318742366527e-06,
+      "loss": 0.5193,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11491919109883862,
+      "grad_norm": 5.555381562592889,
+      "learning_rate": 4.906140322436849e-06,
+      "loss": 0.7675,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11500040607488021,
+      "grad_norm": 8.412815182964685,
+      "learning_rate": 4.9059617360149936e-06,
+      "loss": 0.4897,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11508162105092179,
+      "grad_norm": 5.9497894212896325,
+      "learning_rate": 4.905782983113321e-06,
+      "loss": 0.8408,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11516283602696337,
+      "grad_norm": 8.78483062159324,
+      "learning_rate": 4.905604063744197e-06,
+      "loss": 0.6732,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11524405100300496,
+      "grad_norm": 9.970264715659578,
+      "learning_rate": 4.905424977920004e-06,
+      "loss": 0.532,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11532526597904653,
+      "grad_norm": 5.671192067715544,
+      "learning_rate": 4.9052457256531325e-06,
+      "loss": 0.5852,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11540648095508811,
+      "grad_norm": 7.150271447974999,
+      "learning_rate": 4.905066306955986e-06,
+      "loss": 0.6486,
+      "step": 1421
+    },
+    {
+      "epoch": 0.1154876959311297,
+      "grad_norm": 6.958143840228272,
+      "learning_rate": 4.904886721840981e-06,
+      "loss": 0.6012,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11556891090717128,
+      "grad_norm": 4.760774004700414,
+      "learning_rate": 4.904706970320542e-06,
+      "loss": 0.6706,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11565012588321287,
+      "grad_norm": 4.299376204301504,
+      "learning_rate": 4.904527052407107e-06,
+      "loss": 0.633,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11573134085925445,
+      "grad_norm": 4.095629515769702,
+      "learning_rate": 4.904346968113126e-06,
+      "loss": 0.8379,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11581255583529602,
+      "grad_norm": 6.181276965379973,
+      "learning_rate": 4.904166717451059e-06,
+      "loss": 0.7148,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11589377081133762,
+      "grad_norm": 3.8573979500957676,
+      "learning_rate": 4.90398630043338e-06,
+      "loss": 0.6601,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11597498578737919,
+      "grad_norm": 5.572617436631759,
+      "learning_rate": 4.903805717072572e-06,
+      "loss": 0.8698,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11605620076342077,
+      "grad_norm": 6.9707268541736775,
+      "learning_rate": 4.90362496738113e-06,
+      "loss": 0.5288,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11613741573946236,
+      "grad_norm": 4.5564529770959945,
+      "learning_rate": 4.9034440513715605e-06,
+      "loss": 0.4828,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11621863071550394,
+      "grad_norm": 4.109199531053192,
+      "learning_rate": 4.9032629690563835e-06,
+      "loss": 0.7356,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11629984569154551,
+      "grad_norm": 4.437378918259062,
+      "learning_rate": 4.903081720448128e-06,
+      "loss": 0.5222,
+      "step": 1432
+    },
+    {
+      "epoch": 0.1163810606675871,
+      "grad_norm": 6.399358574915106,
+      "learning_rate": 4.902900305559336e-06,
+      "loss": 0.5954,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11646227564362868,
+      "grad_norm": 4.343118949951745,
+      "learning_rate": 4.9027187244025594e-06,
+      "loss": 0.5096,
+      "step": 1434
+    },
+    {
+      "epoch": 0.11654349061967027,
+      "grad_norm": 4.3409440129123515,
+      "learning_rate": 4.902536976990364e-06,
+      "loss": 0.5396,
+      "step": 1435
+    },
+    {
+      "epoch": 0.11662470559571185,
+      "grad_norm": 6.602474988405446,
+      "learning_rate": 4.902355063335324e-06,
+      "loss": 0.5902,
+      "step": 1436
+    },
+    {
+      "epoch": 0.11670592057175343,
+      "grad_norm": 5.5514520331211425,
+      "learning_rate": 4.902172983450029e-06,
+      "loss": 0.8389,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11678713554779502,
+      "grad_norm": 6.703176030649621,
+      "learning_rate": 4.901990737347076e-06,
+      "loss": 0.8458,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1168683505238366,
+      "grad_norm": 4.454570008772595,
+      "learning_rate": 4.901808325039077e-06,
+      "loss": 0.6968,
+      "step": 1439
+    },
+    {
+      "epoch": 0.11694956549987817,
+      "grad_norm": 4.842565250482216,
+      "learning_rate": 4.901625746538653e-06,
+      "loss": 0.5741,
+      "step": 1440
+    },
+    {
+      "epoch": 0.11703078047591976,
+      "grad_norm": 8.75110385587914,
+      "learning_rate": 4.901443001858438e-06,
+      "loss": 0.5164,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11711199545196134,
+      "grad_norm": 4.468121329238175,
+      "learning_rate": 4.901260091011076e-06,
+      "loss": 0.6103,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11719321042800292,
+      "grad_norm": 7.4646000814738045,
+      "learning_rate": 4.901077014009225e-06,
+      "loss": 0.4323,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11727442540404451,
+      "grad_norm": 3.3258631715598055,
+      "learning_rate": 4.900893770865552e-06,
+      "loss": 0.5174,
+      "step": 1444
+    },
+    {
+      "epoch": 0.11735564038008608,
+      "grad_norm": 8.792100552657674,
+      "learning_rate": 4.900710361592737e-06,
+      "loss": 0.6525,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11743685535612768,
+      "grad_norm": 4.961731173202635,
+      "learning_rate": 4.9005267862034695e-06,
+      "loss": 0.6754,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11751807033216925,
+      "grad_norm": 3.814651002868712,
+      "learning_rate": 4.900343044710453e-06,
+      "loss": 0.6968,
+      "step": 1447
+    },
+    {
+      "epoch": 0.11759928530821083,
+      "grad_norm": 4.911489896369547,
+      "learning_rate": 4.900159137126402e-06,
+      "loss": 0.5924,
+      "step": 1448
+    },
+    {
+      "epoch": 0.11768050028425242,
+      "grad_norm": 7.2633083972944945,
+      "learning_rate": 4.899975063464042e-06,
+      "loss": 0.5496,
+      "step": 1449
+    },
+    {
+      "epoch": 0.117761715260294,
+      "grad_norm": 5.005595908510799,
+      "learning_rate": 4.899790823736108e-06,
+      "loss": 0.6644,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11784293023633557,
+      "grad_norm": 4.702266439098655,
+      "learning_rate": 4.89960641795535e-06,
+      "loss": 0.8283,
+      "step": 1451
+    },
+    {
+      "epoch": 0.11792414521237717,
+      "grad_norm": 6.422288029967876,
+      "learning_rate": 4.899421846134529e-06,
+      "loss": 0.5909,
+      "step": 1452
+    },
+    {
+      "epoch": 0.11800536018841874,
+      "grad_norm": 6.425496839266026,
+      "learning_rate": 4.899237108286414e-06,
+      "loss": 0.499,
+      "step": 1453
+    },
+    {
+      "epoch": 0.11808657516446033,
+      "grad_norm": 3.4591295878059247,
+      "learning_rate": 4.8990522044237884e-06,
+      "loss": 0.5702,
+      "step": 1454
+    },
+    {
+      "epoch": 0.11816779014050191,
+      "grad_norm": 4.321892903241153,
+      "learning_rate": 4.898867134559448e-06,
+      "loss": 0.4269,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11824900511654349,
+      "grad_norm": 6.352156069028349,
+      "learning_rate": 4.898681898706197e-06,
+      "loss": 0.5002,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11833022009258508,
+      "grad_norm": 4.739299344449501,
+      "learning_rate": 4.898496496876854e-06,
+      "loss": 0.7199,
+      "step": 1457
+    },
+    {
+      "epoch": 0.11841143506862666,
+      "grad_norm": 5.475577877089898,
+      "learning_rate": 4.898310929084247e-06,
+      "loss": 0.6848,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11849265004466823,
+      "grad_norm": 4.312752315143542,
+      "learning_rate": 4.898125195341217e-06,
+      "loss": 0.6698,
+      "step": 1459
+    },
+    {
+      "epoch": 0.11857386502070982,
+      "grad_norm": 6.101152761282113,
+      "learning_rate": 4.897939295660615e-06,
+      "loss": 0.5399,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1186550799967514,
+      "grad_norm": 5.730964159782436,
+      "learning_rate": 4.897753230055304e-06,
+      "loss": 0.5391,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11873629497279298,
+      "grad_norm": 4.602835724556902,
+      "learning_rate": 4.89756699853816e-06,
+      "loss": 0.6377,
+      "step": 1462
+    },
+    {
+      "epoch": 0.11881750994883457,
+      "grad_norm": 5.577323364531521,
+      "learning_rate": 4.8973806011220695e-06,
+      "loss": 0.6413,
+      "step": 1463
+    },
+    {
+      "epoch": 0.11889872492487615,
+      "grad_norm": 6.109634102169282,
+      "learning_rate": 4.897194037819928e-06,
+      "loss": 0.5454,
+      "step": 1464
+    },
+    {
+      "epoch": 0.11897993990091774,
+      "grad_norm": 4.473657871957378,
+      "learning_rate": 4.897007308644647e-06,
+      "loss": 0.5516,
+      "step": 1465
+    },
+    {
+      "epoch": 0.11906115487695931,
+      "grad_norm": 4.6381132522690995,
+      "learning_rate": 4.896820413609146e-06,
+      "loss": 0.6565,
+      "step": 1466
+    },
+    {
+      "epoch": 0.11914236985300089,
+      "grad_norm": 5.386818421013147,
+      "learning_rate": 4.896633352726357e-06,
+      "loss": 0.7063,
+      "step": 1467
+    },
+    {
+      "epoch": 0.11922358482904248,
+      "grad_norm": 3.1581048679010646,
+      "learning_rate": 4.896446126009224e-06,
+      "loss": 0.7565,
+      "step": 1468
+    },
+    {
+      "epoch": 0.11930479980508406,
+      "grad_norm": 6.401004722074878,
+      "learning_rate": 4.896258733470702e-06,
+      "loss": 0.5026,
+      "step": 1469
+    },
+    {
+      "epoch": 0.11938601478112564,
+      "grad_norm": 2.7519458408541797,
+      "learning_rate": 4.896071175123758e-06,
+      "loss": 0.6255,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11946722975716723,
+      "grad_norm": 4.404610542747091,
+      "learning_rate": 4.8958834509813706e-06,
+      "loss": 0.5504,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1195484447332088,
+      "grad_norm": 6.077687424149937,
+      "learning_rate": 4.8956955610565275e-06,
+      "loss": 0.6899,
+      "step": 1472
+    },
+    {
+      "epoch": 0.11962965970925038,
+      "grad_norm": 5.523332768156323,
+      "learning_rate": 4.895507505362231e-06,
+      "loss": 0.5743,
+      "step": 1473
+    },
+    {
+      "epoch": 0.11971087468529197,
+      "grad_norm": 5.840235758929547,
+      "learning_rate": 4.895319283911492e-06,
+      "loss": 0.4483,
+      "step": 1474
+    },
+    {
+      "epoch": 0.11979208966133355,
+      "grad_norm": 3.9853018352679177,
+      "learning_rate": 4.895130896717336e-06,
+      "loss": 0.6511,
+      "step": 1475
+    },
+    {
+      "epoch": 0.11987330463737514,
+      "grad_norm": 4.087084936629081,
+      "learning_rate": 4.894942343792799e-06,
+      "loss": 0.6048,
+      "step": 1476
+    },
+    {
+      "epoch": 0.11995451961341672,
+      "grad_norm": 4.9948060996575325,
+      "learning_rate": 4.894753625150927e-06,
+      "loss": 0.5193,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12003573458945829,
+      "grad_norm": 5.148303769586792,
+      "learning_rate": 4.894564740804777e-06,
+      "loss": 0.4947,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12011694956549988,
+      "grad_norm": 4.926289780739354,
+      "learning_rate": 4.89437569076742e-06,
+      "loss": 0.641,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12019816454154146,
+      "grad_norm": 4.420640360809511,
+      "learning_rate": 4.894186475051938e-06,
+      "loss": 0.6427,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12027937951758304,
+      "grad_norm": 4.552192514027042,
+      "learning_rate": 4.893997093671422e-06,
+      "loss": 0.6684,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12036059449362463,
+      "grad_norm": 3.6702893426104306,
+      "learning_rate": 4.893807546638979e-06,
+      "loss": 0.6778,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1204418094696662,
+      "grad_norm": 7.860150046179448,
+      "learning_rate": 4.893617833967721e-06,
+      "loss": 0.6479,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12052302444570778,
+      "grad_norm": 3.6225218661319083,
+      "learning_rate": 4.893427955670778e-06,
+      "loss": 0.5258,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12060423942174937,
+      "grad_norm": 7.286179190372548,
+      "learning_rate": 4.893237911761287e-06,
+      "loss": 0.5499,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12068545439779095,
+      "grad_norm": 5.426351678323337,
+      "learning_rate": 4.893047702252399e-06,
+      "loss": 0.4595,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12076666937383254,
+      "grad_norm": 4.641027529381159,
+      "learning_rate": 4.892857327157275e-06,
+      "loss": 0.7437,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12084788434987412,
+      "grad_norm": 5.910005928653577,
+      "learning_rate": 4.892666786489087e-06,
+      "loss": 0.5187,
+      "step": 1488
+    },
+    {
+      "epoch": 0.1209290993259157,
+      "grad_norm": 4.433359476670774,
+      "learning_rate": 4.8924760802610215e-06,
+      "loss": 0.6245,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12101031430195729,
+      "grad_norm": 16.649957190107745,
+      "learning_rate": 4.8922852084862734e-06,
+      "loss": 0.6885,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12109152927799886,
+      "grad_norm": 5.101531362441216,
+      "learning_rate": 4.892094171178049e-06,
+      "loss": 0.4662,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12117274425404044,
+      "grad_norm": 5.1657125457609725,
+      "learning_rate": 4.891902968349568e-06,
+      "loss": 0.5948,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12125395923008203,
+      "grad_norm": 4.342894827579331,
+      "learning_rate": 4.8917116000140614e-06,
+      "loss": 0.5409,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12133517420612361,
+      "grad_norm": 3.8330951289725124,
+      "learning_rate": 4.8915200661847695e-06,
+      "loss": 0.52,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12141638918216519,
+      "grad_norm": 5.179925909022087,
+      "learning_rate": 4.891328366874946e-06,
+      "loss": 0.5504,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12149760415820678,
+      "grad_norm": 5.330823987896905,
+      "learning_rate": 4.891136502097855e-06,
+      "loss": 0.6273,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12157881913424835,
+      "grad_norm": 5.585945845508732,
+      "learning_rate": 4.890944471866774e-06,
+      "loss": 0.6236,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12166003411028994,
+      "grad_norm": 3.83199043102917,
+      "learning_rate": 4.890752276194989e-06,
+      "loss": 0.7202,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12174124908633152,
+      "grad_norm": 3.6411956112594677,
+      "learning_rate": 4.890559915095798e-06,
+      "loss": 0.5213,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1218224640623731,
+      "grad_norm": 4.786252322922742,
+      "learning_rate": 4.890367388582514e-06,
+      "loss": 0.9004,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12190367903841469,
+      "grad_norm": 4.979342968709159,
+      "learning_rate": 4.890174696668458e-06,
+      "loss": 0.6452,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12198489401445627,
+      "grad_norm": 4.11868570254399,
+      "learning_rate": 4.889981839366962e-06,
+      "loss": 0.4785,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12206610899049784,
+      "grad_norm": 4.908815740621781,
+      "learning_rate": 4.889788816691372e-06,
+      "loss": 0.6467,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12214732396653943,
+      "grad_norm": 6.3622155832910625,
+      "learning_rate": 4.889595628655044e-06,
+      "loss": 0.61,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12222853894258101,
+      "grad_norm": 3.6609655168066064,
+      "learning_rate": 4.8894022752713445e-06,
+      "loss": 0.7134,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12230975391862259,
+      "grad_norm": 4.846573857319596,
+      "learning_rate": 4.8892087565536535e-06,
+      "loss": 0.6646,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12239096889466418,
+      "grad_norm": 5.788899246359694,
+      "learning_rate": 4.889015072515361e-06,
+      "loss": 0.7547,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12247218387070576,
+      "grad_norm": 4.583902093938661,
+      "learning_rate": 4.888821223169869e-06,
+      "loss": 0.5206,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12255339884674735,
+      "grad_norm": 3.3647325384813915,
+      "learning_rate": 4.888627208530592e-06,
+      "loss": 0.5928,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12263461382278892,
+      "grad_norm": 3.8539584897978654,
+      "learning_rate": 4.8884330286109535e-06,
+      "loss": 0.5397,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1227158287988305,
+      "grad_norm": 15.266199610075871,
+      "learning_rate": 4.88823868342439e-06,
+      "loss": 0.6754,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12279704377487209,
+      "grad_norm": 5.095675930247071,
+      "learning_rate": 4.888044172984349e-06,
+      "loss": 0.7183,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12287825875091367,
+      "grad_norm": 5.589937550753531,
+      "learning_rate": 4.887849497304289e-06,
+      "loss": 0.7005,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12295947372695525,
+      "grad_norm": 6.394170257036341,
+      "learning_rate": 4.8876546563976825e-06,
+      "loss": 0.6316,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12304068870299684,
+      "grad_norm": 5.521671875850051,
+      "learning_rate": 4.88745965027801e-06,
+      "loss": 0.6913,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12312190367903841,
+      "grad_norm": 4.969659762897969,
+      "learning_rate": 4.887264478958765e-06,
+      "loss": 0.5295,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12320311865507999,
+      "grad_norm": 6.2253045417503845,
+      "learning_rate": 4.887069142453453e-06,
+      "loss": 0.6538,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12328433363112158,
+      "grad_norm": 9.570082889106901,
+      "learning_rate": 4.886873640775588e-06,
+      "loss": 0.5829,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12336554860716316,
+      "grad_norm": 5.609163036466005,
+      "learning_rate": 4.886677973938701e-06,
+      "loss": 0.5647,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12344676358320475,
+      "grad_norm": 7.818406834616618,
+      "learning_rate": 4.886482141956329e-06,
+      "loss": 0.6357,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12352797855924633,
+      "grad_norm": 3.1695857831672067,
+      "learning_rate": 4.8862861448420234e-06,
+      "loss": 0.5878,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1236091935352879,
+      "grad_norm": 4.384307163664372,
+      "learning_rate": 4.886089982609345e-06,
+      "loss": 0.5771,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1236904085113295,
+      "grad_norm": 4.469922202011447,
+      "learning_rate": 4.885893655271869e-06,
+      "loss": 0.5124,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12377162348737107,
+      "grad_norm": 4.594952235845705,
+      "learning_rate": 4.885697162843179e-06,
+      "loss": 0.6882,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12385283846341265,
+      "grad_norm": 11.952868311490398,
+      "learning_rate": 4.8855005053368715e-06,
+      "loss": 0.5141,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12393405343945424,
+      "grad_norm": 5.107326813624944,
+      "learning_rate": 4.885303682766554e-06,
+      "loss": 0.6077,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12401526841549582,
+      "grad_norm": 3.90801909322924,
+      "learning_rate": 4.885106695145846e-06,
+      "loss": 0.739,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1240964833915374,
+      "grad_norm": 4.458856274877279,
+      "learning_rate": 4.884909542488377e-06,
+      "loss": 0.4853,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12417769836757898,
+      "grad_norm": 7.488057909444421,
+      "learning_rate": 4.88471222480779e-06,
+      "loss": 0.5267,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12425891334362056,
+      "grad_norm": 10.709236752099061,
+      "learning_rate": 4.8845147421177375e-06,
+      "loss": 0.5767,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12434012831966215,
+      "grad_norm": 4.227280902938617,
+      "learning_rate": 4.8843170944318855e-06,
+      "loss": 0.6582,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12442134329570373,
+      "grad_norm": 5.304472983158379,
+      "learning_rate": 4.88411928176391e-06,
+      "loss": 0.4994,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1245025582717453,
+      "grad_norm": 6.083933660892841,
+      "learning_rate": 4.8839213041274955e-06,
+      "loss": 0.454,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1245837732477869,
+      "grad_norm": 7.150062978858339,
+      "learning_rate": 4.8837231615363455e-06,
+      "loss": 0.7561,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12466498822382847,
+      "grad_norm": 6.602096765289932,
+      "learning_rate": 4.883524854004168e-06,
+      "loss": 0.6597,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12474620319987005,
+      "grad_norm": 4.853497195307292,
+      "learning_rate": 4.883326381544686e-06,
+      "loss": 0.5989,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12482741817591164,
+      "grad_norm": 5.57662629799743,
+      "learning_rate": 4.88312774417163e-06,
+      "loss": 0.555,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12490863315195322,
+      "grad_norm": 4.64783655989287,
+      "learning_rate": 4.882928941898748e-06,
+      "loss": 0.5814,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1249898481279948,
+      "grad_norm": 8.657184018704008,
+      "learning_rate": 4.882729974739794e-06,
+      "loss": 0.5567,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12507106310403637,
+      "grad_norm": 9.124561847833686,
+      "learning_rate": 4.882530842708537e-06,
+      "loss": 0.5428,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12515227808007798,
+      "grad_norm": 5.93162727377288,
+      "learning_rate": 4.882331545818755e-06,
+      "loss": 0.5641,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12523349305611955,
+      "grad_norm": 4.417596995333459,
+      "learning_rate": 4.882132084084238e-06,
+      "loss": 0.6554,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12531470803216113,
+      "grad_norm": 5.238384789515402,
+      "learning_rate": 4.8819324575187875e-06,
+      "loss": 0.8369,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1253959230082027,
+      "grad_norm": 5.490158000040072,
+      "learning_rate": 4.881732666136217e-06,
+      "loss": 0.7737,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12547713798424429,
+      "grad_norm": 18.610112814184124,
+      "learning_rate": 4.881532709950352e-06,
+      "loss": 0.5733,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1255583529602859,
+      "grad_norm": 4.3346186230095585,
+      "learning_rate": 4.8813325889750275e-06,
+      "loss": 0.518,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12563956793632747,
+      "grad_norm": 5.394377171648292,
+      "learning_rate": 4.881132303224091e-06,
+      "loss": 0.4973,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12572078291236904,
+      "grad_norm": 6.281282652469185,
+      "learning_rate": 4.880931852711401e-06,
+      "loss": 0.53,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12580199788841062,
+      "grad_norm": 5.063604963463662,
+      "learning_rate": 4.880731237450828e-06,
+      "loss": 0.543,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1258832128644522,
+      "grad_norm": 5.589066320656067,
+      "learning_rate": 4.880530457456252e-06,
+      "loss": 0.5307,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12596442784049378,
+      "grad_norm": 5.313990606526498,
+      "learning_rate": 4.880329512741568e-06,
+      "loss": 0.6096,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12604564281653538,
+      "grad_norm": 5.093580296509499,
+      "learning_rate": 4.88012840332068e-06,
+      "loss": 0.6667,
+      "step": 1552
+    },
+    {
+      "epoch": 0.12612685779257696,
+      "grad_norm": 3.074951611991778,
+      "learning_rate": 4.879927129207502e-06,
+      "loss": 0.669,
+      "step": 1553
+    },
+    {
+      "epoch": 0.12620807276861853,
+      "grad_norm": 9.209748336487731,
+      "learning_rate": 4.8797256904159625e-06,
+      "loss": 0.7399,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1262892877446601,
+      "grad_norm": 5.6697335990100965,
+      "learning_rate": 4.87952408696e-06,
+      "loss": 0.6407,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1263705027207017,
+      "grad_norm": 8.545213357048508,
+      "learning_rate": 4.879322318853564e-06,
+      "loss": 0.5582,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1264517176967433,
+      "grad_norm": 7.569125385036642,
+      "learning_rate": 4.879120386110616e-06,
+      "loss": 0.6649,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12653293267278487,
+      "grad_norm": 16.126868678635603,
+      "learning_rate": 4.878918288745128e-06,
+      "loss": 0.6249,
+      "step": 1558
+    },
+    {
+      "epoch": 0.12661414764882645,
+      "grad_norm": 5.161708427963795,
+      "learning_rate": 4.878716026771086e-06,
+      "loss": 0.5061,
+      "step": 1559
+    },
+    {
+      "epoch": 0.12669536262486802,
+      "grad_norm": 4.94640934937331,
+      "learning_rate": 4.878513600202483e-06,
+      "loss": 0.6149,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1267765776009096,
+      "grad_norm": 5.628454265223888,
+      "learning_rate": 4.878311009053328e-06,
+      "loss": 0.5789,
+      "step": 1561
+    },
+    {
+      "epoch": 0.12685779257695118,
+      "grad_norm": 5.099342067120917,
+      "learning_rate": 4.878108253337638e-06,
+      "loss": 0.6344,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12693900755299278,
+      "grad_norm": 5.258000831985147,
+      "learning_rate": 4.877905333069442e-06,
+      "loss": 0.5775,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12702022252903436,
+      "grad_norm": 6.236383483926228,
+      "learning_rate": 4.877702248262782e-06,
+      "loss": 0.5334,
+      "step": 1564
+    },
+    {
+      "epoch": 0.12710143750507594,
+      "grad_norm": 4.048534369160581,
+      "learning_rate": 4.87749899893171e-06,
+      "loss": 0.7184,
+      "step": 1565
+    },
+    {
+      "epoch": 0.12718265248111751,
+      "grad_norm": 4.398427223579315,
+      "learning_rate": 4.8772955850902914e-06,
+      "loss": 0.6671,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1272638674571591,
+      "grad_norm": 5.637044630053773,
+      "learning_rate": 4.877092006752599e-06,
+      "loss": 0.621,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1273450824332007,
+      "grad_norm": 6.7178341766236125,
+      "learning_rate": 4.876888263932721e-06,
+      "loss": 0.5828,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12742629740924227,
+      "grad_norm": 5.624169626907889,
+      "learning_rate": 4.876684356644754e-06,
+      "loss": 0.6779,
+      "step": 1569
+    },
+    {
+      "epoch": 0.12750751238528385,
+      "grad_norm": 4.476842415269882,
+      "learning_rate": 4.876480284902807e-06,
+      "loss": 0.4934,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12758872736132543,
+      "grad_norm": 3.7336585247616885,
+      "learning_rate": 4.8762760487210035e-06,
+      "loss": 0.5301,
+      "step": 1571
+    },
+    {
+      "epoch": 0.127669942337367,
+      "grad_norm": 8.61154812995492,
+      "learning_rate": 4.876071648113473e-06,
+      "loss": 0.6757,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12775115731340858,
+      "grad_norm": 6.969888194514512,
+      "learning_rate": 4.875867083094359e-06,
+      "loss": 0.7971,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12783237228945019,
+      "grad_norm": 7.7890484350762765,
+      "learning_rate": 4.875662353677818e-06,
+      "loss": 0.4918,
+      "step": 1574
+    },
+    {
+      "epoch": 0.12791358726549176,
+      "grad_norm": 4.545675942015323,
+      "learning_rate": 4.875457459878014e-06,
+      "loss": 0.6069,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12799480224153334,
+      "grad_norm": 7.615402665241847,
+      "learning_rate": 4.875252401709126e-06,
+      "loss": 0.5996,
+      "step": 1576
+    },
+    {
+      "epoch": 0.12807601721757492,
+      "grad_norm": 5.198690570704765,
+      "learning_rate": 4.8750471791853436e-06,
+      "loss": 0.6422,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1281572321936165,
+      "grad_norm": 5.875388513113994,
+      "learning_rate": 4.874841792320865e-06,
+      "loss": 0.4701,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1282384471696581,
+      "grad_norm": 6.491950200516407,
+      "learning_rate": 4.874636241129904e-06,
+      "loss": 0.5815,
+      "step": 1579
+    },
+    {
+      "epoch": 0.12831966214569968,
+      "grad_norm": 4.293195229926921,
+      "learning_rate": 4.874430525626682e-06,
+      "loss": 0.6365,
+      "step": 1580
+    },
+    {
+      "epoch": 0.12840087712174125,
+      "grad_norm": 8.402150286140357,
+      "learning_rate": 4.874224645825435e-06,
+      "loss": 0.4979,
+      "step": 1581
+    },
+    {
+      "epoch": 0.12848209209778283,
+      "grad_norm": 6.3159806141726,
+      "learning_rate": 4.874018601740407e-06,
+      "loss": 0.504,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1285633070738244,
+      "grad_norm": 5.865695583703645,
+      "learning_rate": 4.873812393385856e-06,
+      "loss": 0.5678,
+      "step": 1583
+    },
+    {
+      "epoch": 0.12864452204986598,
+      "grad_norm": 4.259448849973863,
+      "learning_rate": 4.873606020776051e-06,
+      "loss": 0.637,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1287257370259076,
+      "grad_norm": 4.795031800681299,
+      "learning_rate": 4.873399483925272e-06,
+      "loss": 0.6518,
+      "step": 1585
+    },
+    {
+      "epoch": 0.12880695200194917,
+      "grad_norm": 6.982468505996596,
+      "learning_rate": 4.8731927828478085e-06,
+      "loss": 0.5015,
+      "step": 1586
+    },
+    {
+      "epoch": 0.12888816697799074,
+      "grad_norm": 4.194815922907224,
+      "learning_rate": 4.872985917557965e-06,
+      "loss": 0.6631,
+      "step": 1587
+    },
+    {
+      "epoch": 0.12896938195403232,
+      "grad_norm": 6.979913270961702,
+      "learning_rate": 4.872778888070055e-06,
+      "loss": 0.477,
+      "step": 1588
+    },
+    {
+      "epoch": 0.1290505969300739,
+      "grad_norm": 6.539545690632459,
+      "learning_rate": 4.872571694398403e-06,
+      "loss": 0.5583,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1291318119061155,
+      "grad_norm": 7.207244340353426,
+      "learning_rate": 4.872364336557348e-06,
+      "loss": 0.6374,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12921302688215708,
+      "grad_norm": 5.271822346774599,
+      "learning_rate": 4.8721568145612355e-06,
+      "loss": 0.5826,
+      "step": 1591
+    },
+    {
+      "epoch": 0.12929424185819866,
+      "grad_norm": 6.699431722524916,
+      "learning_rate": 4.8719491284244256e-06,
+      "loss": 0.5468,
+      "step": 1592
+    },
+    {
+      "epoch": 0.12937545683424023,
+      "grad_norm": 4.303099582925413,
+      "learning_rate": 4.871741278161291e-06,
+      "loss": 0.5465,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1294566718102818,
+      "grad_norm": 3.318090123286297,
+      "learning_rate": 4.87153326378621e-06,
+      "loss": 0.5562,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1295378867863234,
+      "grad_norm": 3.8336332872072507,
+      "learning_rate": 4.87132508531358e-06,
+      "loss": 0.6022,
+      "step": 1595
+    },
+    {
+      "epoch": 0.129619101762365,
+      "grad_norm": 6.539483107346492,
+      "learning_rate": 4.871116742757803e-06,
+      "loss": 0.4905,
+      "step": 1596
+    },
+    {
+      "epoch": 0.12970031673840657,
+      "grad_norm": 3.2101449234700072,
+      "learning_rate": 4.870908236133297e-06,
+      "loss": 0.6639,
+      "step": 1597
+    },
+    {
+      "epoch": 0.12978153171444815,
+      "grad_norm": 11.166157350084525,
+      "learning_rate": 4.870699565454489e-06,
+      "loss": 0.5098,
+      "step": 1598
+    },
+    {
+      "epoch": 0.12986274669048972,
+      "grad_norm": 4.137968227782281,
+      "learning_rate": 4.870490730735818e-06,
+      "loss": 0.4853,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1299439616665313,
+      "grad_norm": 3.555952898094085,
+      "learning_rate": 4.870281731991733e-06,
+      "loss": 0.5433,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1300251766425729,
+      "grad_norm": 6.292559801998863,
+      "learning_rate": 4.870072569236697e-06,
+      "loss": 0.5833,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13010639161861448,
+      "grad_norm": 4.245679851338393,
+      "learning_rate": 4.869863242485183e-06,
+      "loss": 0.5839,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13018760659465606,
+      "grad_norm": 10.695591016769804,
+      "learning_rate": 4.8696537517516754e-06,
+      "loss": 0.5284,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13026882157069763,
+      "grad_norm": 5.501068173317865,
+      "learning_rate": 4.869444097050668e-06,
+      "loss": 0.5927,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1303500365467392,
+      "grad_norm": 5.717664746661781,
+      "learning_rate": 4.8692342783966706e-06,
+      "loss": 0.5258,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1304312515227808,
+      "grad_norm": 5.029968109519692,
+      "learning_rate": 4.869024295804199e-06,
+      "loss": 0.6064,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1305124664988224,
+      "grad_norm": 3.7931205356074575,
+      "learning_rate": 4.868814149287785e-06,
+      "loss": 0.5764,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13059368147486397,
+      "grad_norm": 4.944616527342133,
+      "learning_rate": 4.868603838861969e-06,
+      "loss": 0.5991,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13067489645090555,
+      "grad_norm": 5.178859852231613,
+      "learning_rate": 4.868393364541301e-06,
+      "loss": 0.6465,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13075611142694712,
+      "grad_norm": 4.567090745940825,
+      "learning_rate": 4.868182726340349e-06,
+      "loss": 0.6649,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1308373264029887,
+      "grad_norm": 5.802293496504625,
+      "learning_rate": 4.867971924273685e-06,
+      "loss": 0.5043,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1309185413790303,
+      "grad_norm": 4.914755653217928,
+      "learning_rate": 4.8677609583558956e-06,
+      "loss": 0.5295,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13099975635507188,
+      "grad_norm": 7.208070575519608,
+      "learning_rate": 4.867549828601579e-06,
+      "loss": 0.6164,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13108097133111346,
+      "grad_norm": 7.77772857949289,
+      "learning_rate": 4.8673385350253454e-06,
+      "loss": 0.5208,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13116218630715504,
+      "grad_norm": 8.264639243565528,
+      "learning_rate": 4.867127077641813e-06,
+      "loss": 0.6235,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13124340128319661,
+      "grad_norm": 5.062219960669989,
+      "learning_rate": 4.866915456465615e-06,
+      "loss": 0.57,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1313246162592382,
+      "grad_norm": 16.478111147319943,
+      "learning_rate": 4.866703671511395e-06,
+      "loss": 0.5729,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1314058312352798,
+      "grad_norm": 4.4225512989681475,
+      "learning_rate": 4.8664917227938056e-06,
+      "loss": 0.5722,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13148704621132137,
+      "grad_norm": 5.041669688436434,
+      "learning_rate": 4.866279610327514e-06,
+      "loss": 0.7651,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13156826118736295,
+      "grad_norm": 5.8825838835020505,
+      "learning_rate": 4.8660673341271966e-06,
+      "loss": 0.6381,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13164947616340453,
+      "grad_norm": 4.034188407645285,
+      "learning_rate": 4.865854894207541e-06,
+      "loss": 0.6949,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1317306911394461,
+      "grad_norm": 3.6403419129784864,
+      "learning_rate": 4.865642290583249e-06,
+      "loss": 0.5172,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1318119061154877,
+      "grad_norm": 4.836824286973765,
+      "learning_rate": 4.86542952326903e-06,
+      "loss": 0.6615,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1318931210915293,
+      "grad_norm": 4.388993168301412,
+      "learning_rate": 4.865216592279607e-06,
+      "loss": 0.6131,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13197433606757086,
+      "grad_norm": 6.591216319051455,
+      "learning_rate": 4.865003497629713e-06,
+      "loss": 0.6583,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13205555104361244,
+      "grad_norm": 4.389020157372549,
+      "learning_rate": 4.8647902393340955e-06,
+      "loss": 0.6026,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13213676601965402,
+      "grad_norm": 4.250808042920454,
+      "learning_rate": 4.864576817407507e-06,
+      "loss": 0.6323,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1322179809956956,
+      "grad_norm": 6.881832060317491,
+      "learning_rate": 4.864363231864717e-06,
+      "loss": 0.6207,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1322991959717372,
+      "grad_norm": 10.412011238660916,
+      "learning_rate": 4.864149482720505e-06,
+      "loss": 0.5453,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13238041094777878,
+      "grad_norm": 5.774868373588398,
+      "learning_rate": 4.863935569989662e-06,
+      "loss": 0.5582,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13246162592382035,
+      "grad_norm": 5.245799631843284,
+      "learning_rate": 4.863721493686987e-06,
+      "loss": 0.6431,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13254284089986193,
+      "grad_norm": 4.177337851592638,
+      "learning_rate": 4.8635072538272954e-06,
+      "loss": 0.5681,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1326240558759035,
+      "grad_norm": 3.517896942856102,
+      "learning_rate": 4.863292850425409e-06,
+      "loss": 0.6479,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1327052708519451,
+      "grad_norm": 5.350168873979354,
+      "learning_rate": 4.863078283496167e-06,
+      "loss": 0.6847,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1327864858279867,
+      "grad_norm": 6.986629354106734,
+      "learning_rate": 4.862863553054413e-06,
+      "loss": 0.6294,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13286770080402827,
+      "grad_norm": 4.997326232828624,
+      "learning_rate": 4.862648659115007e-06,
+      "loss": 0.7014,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13294891578006984,
+      "grad_norm": 4.2346207052457485,
+      "learning_rate": 4.8624336016928175e-06,
+      "loss": 0.5626,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13303013075611142,
+      "grad_norm": 4.331783623394118,
+      "learning_rate": 4.8622183808027255e-06,
+      "loss": 0.6618,
+      "step": 1638
+    },
+    {
+      "epoch": 0.133111345732153,
+      "grad_norm": 5.008484418976238,
+      "learning_rate": 4.8620029964596234e-06,
+      "loss": 0.6353,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1331925607081946,
+      "grad_norm": 6.50139784970913,
+      "learning_rate": 4.861787448678416e-06,
+      "loss": 0.486,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13327377568423618,
+      "grad_norm": 6.369833370069241,
+      "learning_rate": 4.861571737474015e-06,
+      "loss": 0.5904,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13335499066027776,
+      "grad_norm": 3.8537585558162317,
+      "learning_rate": 4.8613558628613494e-06,
+      "loss": 0.8424,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13343620563631933,
+      "grad_norm": 6.904987154180493,
+      "learning_rate": 4.8611398248553554e-06,
+      "loss": 0.5671,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1335174206123609,
+      "grad_norm": 3.1433510533347415,
+      "learning_rate": 4.860923623470981e-06,
+      "loss": 0.4704,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13359863558840251,
+      "grad_norm": 5.418253266261615,
+      "learning_rate": 4.860707258723187e-06,
+      "loss": 0.5313,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1336798505644441,
+      "grad_norm": 5.2359535500265615,
+      "learning_rate": 4.860490730626945e-06,
+      "loss": 0.5742,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13376106554048567,
+      "grad_norm": 11.436418505055565,
+      "learning_rate": 4.860274039197237e-06,
+      "loss": 0.5654,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13384228051652725,
+      "grad_norm": 6.262589762855426,
+      "learning_rate": 4.860057184449057e-06,
+      "loss": 0.4724,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13392349549256882,
+      "grad_norm": 5.518995358402323,
+      "learning_rate": 4.85984016639741e-06,
+      "loss": 0.5537,
+      "step": 1649
+    },
+    {
+      "epoch": 0.1340047104686104,
+      "grad_norm": 5.088469039450566,
+      "learning_rate": 4.859622985057313e-06,
+      "loss": 0.5638,
+      "step": 1650
+    },
+    {
+      "epoch": 0.134085925444652,
+      "grad_norm": 4.614369100716233,
+      "learning_rate": 4.859405640443793e-06,
+      "loss": 0.6239,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13416714042069358,
+      "grad_norm": 6.877855199561925,
+      "learning_rate": 4.85918813257189e-06,
+      "loss": 0.5406,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13424835539673516,
+      "grad_norm": 6.336440018776646,
+      "learning_rate": 4.858970461456655e-06,
+      "loss": 0.5815,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13432957037277674,
+      "grad_norm": 6.858031875975209,
+      "learning_rate": 4.858752627113148e-06,
+      "loss": 0.6946,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1344107853488183,
+      "grad_norm": 3.403378239632881,
+      "learning_rate": 4.8585346295564425e-06,
+      "loss": 0.7919,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13449200032485992,
+      "grad_norm": 6.487088297683663,
+      "learning_rate": 4.858316468801624e-06,
+      "loss": 0.4289,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1345732153009015,
+      "grad_norm": 9.756045801760866,
+      "learning_rate": 4.858098144863786e-06,
+      "loss": 0.5673,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13465443027694307,
+      "grad_norm": 4.769994844405226,
+      "learning_rate": 4.857879657758037e-06,
+      "loss": 0.6929,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13473564525298465,
+      "grad_norm": 4.567963041913676,
+      "learning_rate": 4.857661007499493e-06,
+      "loss": 0.6402,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13481686022902623,
+      "grad_norm": 7.539959586550248,
+      "learning_rate": 4.857442194103287e-06,
+      "loss": 0.6195,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1348980752050678,
+      "grad_norm": 3.320972169580362,
+      "learning_rate": 4.8572232175845574e-06,
+      "loss": 0.5644,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1349792901811094,
+      "grad_norm": 6.466693615210596,
+      "learning_rate": 4.857004077958456e-06,
+      "loss": 0.6244,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13506050515715098,
+      "grad_norm": 5.503111817650875,
+      "learning_rate": 4.8567847752401476e-06,
+      "loss": 0.5756,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13514172013319256,
+      "grad_norm": 5.858104142956886,
+      "learning_rate": 4.8565653094448054e-06,
+      "loss": 0.6822,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13522293510923414,
+      "grad_norm": 4.782981817213062,
+      "learning_rate": 4.856345680587616e-06,
+      "loss": 0.7815,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13530415008527572,
+      "grad_norm": 15.4570240471019,
+      "learning_rate": 4.856125888683775e-06,
+      "loss": 0.4739,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13538536506131732,
+      "grad_norm": 4.282142809688187,
+      "learning_rate": 4.855905933748492e-06,
+      "loss": 0.5324,
+      "step": 1667
+    },
+    {
+      "epoch": 0.1354665800373589,
+      "grad_norm": 5.095680068440331,
+      "learning_rate": 4.855685815796989e-06,
+      "loss": 0.5745,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13554779501340047,
+      "grad_norm": 4.141977804675802,
+      "learning_rate": 4.855465534844494e-06,
+      "loss": 0.7417,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13562900998944205,
+      "grad_norm": 7.6969307870140895,
+      "learning_rate": 4.8552450909062494e-06,
+      "loss": 0.5206,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13571022496548363,
+      "grad_norm": 6.450650416903025,
+      "learning_rate": 4.855024483997509e-06,
+      "loss": 0.655,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1357914399415252,
+      "grad_norm": 4.810253355009561,
+      "learning_rate": 4.85480371413354e-06,
+      "loss": 0.5323,
+      "step": 1672
+    },
+    {
+      "epoch": 0.1358726549175668,
+      "grad_norm": 3.3253261110937116,
+      "learning_rate": 4.8545827813296154e-06,
+      "loss": 0.5753,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1359538698936084,
+      "grad_norm": 3.8466882509432496,
+      "learning_rate": 4.8543616856010235e-06,
+      "loss": 0.7926,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13603508486964996,
+      "grad_norm": 4.8978637912258,
+      "learning_rate": 4.854140426963064e-06,
+      "loss": 0.5541,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13611629984569154,
+      "grad_norm": 5.229352786538515,
+      "learning_rate": 4.853919005431046e-06,
+      "loss": 0.4727,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13619751482173312,
+      "grad_norm": 2.9378106286602095,
+      "learning_rate": 4.85369742102029e-06,
+      "loss": 0.7667,
+      "step": 1677
+    },
+    {
+      "epoch": 0.13627872979777472,
+      "grad_norm": 9.698335327128627,
+      "learning_rate": 4.8534756737461305e-06,
+      "loss": 0.7053,
+      "step": 1678
+    },
+    {
+      "epoch": 0.1363599447738163,
+      "grad_norm": 4.258566186707566,
+      "learning_rate": 4.853253763623909e-06,
+      "loss": 0.7551,
+      "step": 1679
+    },
+    {
+      "epoch": 0.13644115974985788,
+      "grad_norm": 7.742141939352821,
+      "learning_rate": 4.853031690668982e-06,
+      "loss": 0.6012,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13652237472589945,
+      "grad_norm": 7.324297389610232,
+      "learning_rate": 4.852809454896715e-06,
+      "loss": 0.8128,
+      "step": 1681
+    },
+    {
+      "epoch": 0.13660358970194103,
+      "grad_norm": 6.79191875565859,
+      "learning_rate": 4.852587056322485e-06,
+      "loss": 0.4649,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1366848046779826,
+      "grad_norm": 4.346976001633204,
+      "learning_rate": 4.852364494961684e-06,
+      "loss": 0.5741,
+      "step": 1683
+    },
+    {
+      "epoch": 0.1367660196540242,
+      "grad_norm": 5.599252491473373,
+      "learning_rate": 4.852141770829707e-06,
+      "loss": 0.5591,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1368472346300658,
+      "grad_norm": 7.657958990969802,
+      "learning_rate": 4.851918883941969e-06,
+      "loss": 0.5636,
+      "step": 1685
+    },
+    {
+      "epoch": 0.13692844960610737,
+      "grad_norm": 3.4450921267982126,
+      "learning_rate": 4.851695834313892e-06,
+      "loss": 0.7366,
+      "step": 1686
+    },
+    {
+      "epoch": 0.13700966458214894,
+      "grad_norm": 5.273690756756951,
+      "learning_rate": 4.851472621960909e-06,
+      "loss": 0.5808,
+      "step": 1687
+    },
+    {
+      "epoch": 0.13709087955819052,
+      "grad_norm": 4.952280541512243,
+      "learning_rate": 4.851249246898465e-06,
+      "loss": 0.6648,
+      "step": 1688
+    },
+    {
+      "epoch": 0.13717209453423213,
+      "grad_norm": 7.707004969920144,
+      "learning_rate": 4.851025709142018e-06,
+      "loss": 0.6268,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1372533095102737,
+      "grad_norm": 4.430949277747921,
+      "learning_rate": 4.850802008707034e-06,
+      "loss": 0.4935,
+      "step": 1690
+    },
+    {
+      "epoch": 0.13733452448631528,
+      "grad_norm": 4.0862917668071175,
+      "learning_rate": 4.8505781456089926e-06,
+      "loss": 0.7303,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13741573946235686,
+      "grad_norm": 8.15399495830509,
+      "learning_rate": 4.850354119863384e-06,
+      "loss": 0.573,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13749695443839843,
+      "grad_norm": 4.4854550799183,
+      "learning_rate": 4.850129931485709e-06,
+      "loss": 0.6696,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13757816941444,
+      "grad_norm": 6.062165358428213,
+      "learning_rate": 4.849905580491481e-06,
+      "loss": 0.506,
+      "step": 1694
+    },
+    {
+      "epoch": 0.13765938439048162,
+      "grad_norm": 4.30707595006544,
+      "learning_rate": 4.849681066896224e-06,
+      "loss": 0.5298,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1377405993665232,
+      "grad_norm": 4.645280994473025,
+      "learning_rate": 4.849456390715471e-06,
+      "loss": 0.5178,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13782181434256477,
+      "grad_norm": 4.408267231508574,
+      "learning_rate": 4.849231551964771e-06,
+      "loss": 0.5725,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13790302931860635,
+      "grad_norm": 9.45812967426135,
+      "learning_rate": 4.849006550659681e-06,
+      "loss": 0.6328,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13798424429464792,
+      "grad_norm": 4.491014201207106,
+      "learning_rate": 4.84878138681577e-06,
+      "loss": 0.5886,
+      "step": 1699
+    },
+    {
+      "epoch": 0.13806545927068953,
+      "grad_norm": 6.727338627008777,
+      "learning_rate": 4.848556060448617e-06,
+      "loss": 0.5413,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1381466742467311,
+      "grad_norm": 6.965832541226805,
+      "learning_rate": 4.848330571573815e-06,
+      "loss": 0.582,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13822788922277268,
+      "grad_norm": 5.263090461409852,
+      "learning_rate": 4.848104920206964e-06,
+      "loss": 0.5413,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13830910419881426,
+      "grad_norm": 4.916463781199789,
+      "learning_rate": 4.847879106363681e-06,
+      "loss": 0.5164,
+      "step": 1703
+    },
+    {
+      "epoch": 0.13839031917485584,
+      "grad_norm": 4.020484866288476,
+      "learning_rate": 4.847653130059591e-06,
+      "loss": 0.5987,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1384715341508974,
+      "grad_norm": 7.925299595269813,
+      "learning_rate": 4.847426991310327e-06,
+      "loss": 0.5019,
+      "step": 1705
+    },
+    {
+      "epoch": 0.13855274912693902,
+      "grad_norm": 4.765085041505615,
+      "learning_rate": 4.84720069013154e-06,
+      "loss": 0.6024,
+      "step": 1706
+    },
+    {
+      "epoch": 0.1386339641029806,
+      "grad_norm": 3.5389423477868704,
+      "learning_rate": 4.846974226538887e-06,
+      "loss": 0.5936,
+      "step": 1707
+    },
+    {
+      "epoch": 0.13871517907902217,
+      "grad_norm": 4.437272587336693,
+      "learning_rate": 4.846747600548039e-06,
+      "loss": 0.6592,
+      "step": 1708
+    },
+    {
+      "epoch": 0.13879639405506375,
+      "grad_norm": 6.923436463835136,
+      "learning_rate": 4.8465208121746775e-06,
+      "loss": 0.7827,
+      "step": 1709
+    },
+    {
+      "epoch": 0.13887760903110533,
+      "grad_norm": 4.075078776605313,
+      "learning_rate": 4.846293861434494e-06,
+      "loss": 0.6561,
+      "step": 1710
+    },
+    {
+      "epoch": 0.13895882400714693,
+      "grad_norm": 4.965543277677793,
+      "learning_rate": 4.846066748343193e-06,
+      "loss": 0.6394,
+      "step": 1711
+    },
+    {
+      "epoch": 0.1390400389831885,
+      "grad_norm": 4.952847589325617,
+      "learning_rate": 4.84583947291649e-06,
+      "loss": 0.5429,
+      "step": 1712
+    },
+    {
+      "epoch": 0.13912125395923008,
+      "grad_norm": 6.231762044145288,
+      "learning_rate": 4.84561203517011e-06,
+      "loss": 0.6042,
+      "step": 1713
+    },
+    {
+      "epoch": 0.13920246893527166,
+      "grad_norm": 7.90195822251666,
+      "learning_rate": 4.8453844351197906e-06,
+      "loss": 0.6262,
+      "step": 1714
+    },
+    {
+      "epoch": 0.13928368391131324,
+      "grad_norm": 5.163740585708867,
+      "learning_rate": 4.845156672781283e-06,
+      "loss": 0.4433,
+      "step": 1715
+    },
+    {
+      "epoch": 0.13936489888735482,
+      "grad_norm": 6.772106592052472,
+      "learning_rate": 4.844928748170343e-06,
+      "loss": 0.5625,
+      "step": 1716
+    },
+    {
+      "epoch": 0.13944611386339642,
+      "grad_norm": 7.967404236571118,
+      "learning_rate": 4.844700661302745e-06,
+      "loss": 0.5634,
+      "step": 1717
+    },
+    {
+      "epoch": 0.139527328839438,
+      "grad_norm": 4.952354575898061,
+      "learning_rate": 4.844472412194271e-06,
+      "loss": 0.4938,
+      "step": 1718
+    },
+    {
+      "epoch": 0.13960854381547957,
+      "grad_norm": 4.882182783614095,
+      "learning_rate": 4.844244000860713e-06,
+      "loss": 0.514,
+      "step": 1719
+    },
+    {
+      "epoch": 0.13968975879152115,
+      "grad_norm": 4.517018593698755,
+      "learning_rate": 4.844015427317878e-06,
+      "loss": 0.5567,
+      "step": 1720
+    },
+    {
+      "epoch": 0.13977097376756273,
+      "grad_norm": 5.353241935249897,
+      "learning_rate": 4.84378669158158e-06,
+      "loss": 0.6774,
+      "step": 1721
+    },
+    {
+      "epoch": 0.13985218874360433,
+      "grad_norm": 4.71686990816346,
+      "learning_rate": 4.843557793667647e-06,
+      "loss": 0.5591,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1399334037196459,
+      "grad_norm": 4.136924741202888,
+      "learning_rate": 4.843328733591918e-06,
+      "loss": 0.711,
+      "step": 1723
+    },
+    {
+      "epoch": 0.1400146186956875,
+      "grad_norm": 5.450901778292648,
+      "learning_rate": 4.843099511370243e-06,
+      "loss": 0.6455,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14009583367172906,
+      "grad_norm": 5.05521301047581,
+      "learning_rate": 4.842870127018482e-06,
+      "loss": 0.5929,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14017704864777064,
+      "grad_norm": 5.278566981624996,
+      "learning_rate": 4.842640580552508e-06,
+      "loss": 0.5624,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14025826362381222,
+      "grad_norm": 3.187653564030086,
+      "learning_rate": 4.842410871988204e-06,
+      "loss": 0.4295,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14033947859985382,
+      "grad_norm": 4.96925386948935,
+      "learning_rate": 4.842181001341465e-06,
+      "loss": 0.6622,
+      "step": 1728
+    },
+    {
+      "epoch": 0.1404206935758954,
+      "grad_norm": 4.620002867331911,
+      "learning_rate": 4.8419509686281965e-06,
+      "loss": 0.6541,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14050190855193698,
+      "grad_norm": 4.351427980680414,
+      "learning_rate": 4.841720773864315e-06,
+      "loss": 0.5794,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14058312352797855,
+      "grad_norm": 6.055475530739616,
+      "learning_rate": 4.84149041706575e-06,
+      "loss": 0.6507,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14066433850402013,
+      "grad_norm": 4.39137607633036,
+      "learning_rate": 4.8412598982484396e-06,
+      "loss": 0.5077,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14074555348006174,
+      "grad_norm": 5.347025072258548,
+      "learning_rate": 4.8410292174283356e-06,
+      "loss": 0.5427,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1408267684561033,
+      "grad_norm": 5.982794959612033,
+      "learning_rate": 4.840798374621399e-06,
+      "loss": 0.6222,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1409079834321449,
+      "grad_norm": 4.786923948878069,
+      "learning_rate": 4.8405673698436046e-06,
+      "loss": 0.5887,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14098919840818647,
+      "grad_norm": 4.34136443925787,
+      "learning_rate": 4.840336203110934e-06,
+      "loss": 0.8337,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14107041338422804,
+      "grad_norm": 6.40416263003894,
+      "learning_rate": 4.840104874439385e-06,
+      "loss": 0.4646,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14115162836026962,
+      "grad_norm": 4.993914992295412,
+      "learning_rate": 4.839873383844964e-06,
+      "loss": 0.6091,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14123284333631123,
+      "grad_norm": 4.072501067349767,
+      "learning_rate": 4.839641731343688e-06,
+      "loss": 0.5803,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1413140583123528,
+      "grad_norm": 4.916864007591204,
+      "learning_rate": 4.839409916951586e-06,
+      "loss": 0.6042,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14139527328839438,
+      "grad_norm": 7.127573823763706,
+      "learning_rate": 4.839177940684699e-06,
+      "loss": 0.5263,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14147648826443596,
+      "grad_norm": 4.634413084773449,
+      "learning_rate": 4.838945802559079e-06,
+      "loss": 0.6171,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14155770324047753,
+      "grad_norm": 4.578421538253424,
+      "learning_rate": 4.8387135025907885e-06,
+      "loss": 0.5441,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14163891821651914,
+      "grad_norm": 4.586524758850179,
+      "learning_rate": 4.8384810407959e-06,
+      "loss": 0.5399,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14172013319256072,
+      "grad_norm": 5.918670545030279,
+      "learning_rate": 4.8382484171905006e-06,
+      "loss": 0.4686,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1418013481686023,
+      "grad_norm": 5.704207192778562,
+      "learning_rate": 4.8380156317906855e-06,
+      "loss": 0.6141,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14188256314464387,
+      "grad_norm": 4.921937978250558,
+      "learning_rate": 4.837782684612562e-06,
+      "loss": 0.5936,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14196377812068545,
+      "grad_norm": 8.366604146884352,
+      "learning_rate": 4.83754957567225e-06,
+      "loss": 0.5511,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14204499309672702,
+      "grad_norm": 3.91804902399355,
+      "learning_rate": 4.837316304985879e-06,
+      "loss": 0.6865,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14212620807276863,
+      "grad_norm": 7.888393422464691,
+      "learning_rate": 4.8370828725695885e-06,
+      "loss": 0.5439,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1422074230488102,
+      "grad_norm": 4.847280666549644,
+      "learning_rate": 4.836849278439532e-06,
+      "loss": 0.6774,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14228863802485178,
+      "grad_norm": 5.563719824565103,
+      "learning_rate": 4.836615522611874e-06,
+      "loss": 0.5419,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14236985300089336,
+      "grad_norm": 9.244694159995396,
+      "learning_rate": 4.8363816051027875e-06,
+      "loss": 0.5761,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14245106797693494,
+      "grad_norm": 4.37956947048255,
+      "learning_rate": 4.8361475259284604e-06,
+      "loss": 0.6103,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14253228295297654,
+      "grad_norm": 11.159158566456547,
+      "learning_rate": 4.8359132851050875e-06,
+      "loss": 0.4938,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14261349792901812,
+      "grad_norm": 5.121940842643738,
+      "learning_rate": 4.835678882648878e-06,
+      "loss": 0.7047,
+      "step": 1756
+    },
+    {
+      "epoch": 0.1426947129050597,
+      "grad_norm": 4.634689837382904,
+      "learning_rate": 4.8354443185760505e-06,
+      "loss": 0.5441,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14277592788110127,
+      "grad_norm": 4.696643113982221,
+      "learning_rate": 4.835209592902837e-06,
+      "loss": 0.6011,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 33.651614779388595,
+      "learning_rate": 4.834974705645478e-06,
+      "loss": 0.5516,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14293835783318443,
+      "grad_norm": 7.739691900140675,
+      "learning_rate": 4.834739656820228e-06,
+      "loss": 0.6453,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14301957280922603,
+      "grad_norm": 6.791535657044752,
+      "learning_rate": 4.83450444644335e-06,
+      "loss": 0.7097,
+      "step": 1761
+    },
+    {
+      "epoch": 0.1431007877852676,
+      "grad_norm": 5.503204544875674,
+      "learning_rate": 4.834269074531119e-06,
+      "loss": 0.5335,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14318200276130919,
+      "grad_norm": 4.4158123615058935,
+      "learning_rate": 4.834033541099822e-06,
+      "loss": 0.5697,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14326321773735076,
+      "grad_norm": 8.45132186133372,
+      "learning_rate": 4.833797846165758e-06,
+      "loss": 0.575,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14334443271339234,
+      "grad_norm": 4.117636505890885,
+      "learning_rate": 4.833561989745232e-06,
+      "loss": 0.7003,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14342564768943394,
+      "grad_norm": 5.425663068958977,
+      "learning_rate": 4.833325971854568e-06,
+      "loss": 0.5898,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14350686266547552,
+      "grad_norm": 4.870653806817092,
+      "learning_rate": 4.8330897925100966e-06,
+      "loss": 0.6641,
+      "step": 1767
+    },
+    {
+      "epoch": 0.1435880776415171,
+      "grad_norm": 4.047961871711326,
+      "learning_rate": 4.8328534517281575e-06,
+      "loss": 0.6453,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14366929261755867,
+      "grad_norm": 4.458252548214351,
+      "learning_rate": 4.832616949525107e-06,
+      "loss": 0.4695,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14375050759360025,
+      "grad_norm": 7.708237827828306,
+      "learning_rate": 4.832380285917309e-06,
+      "loss": 0.5208,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14383172256964183,
+      "grad_norm": 6.499621307199643,
+      "learning_rate": 4.8321434609211386e-06,
+      "loss": 0.6735,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14391293754568343,
+      "grad_norm": 5.872483962693705,
+      "learning_rate": 4.831906474552983e-06,
+      "loss": 0.467,
+      "step": 1772
+    },
+    {
+      "epoch": 0.143994152521725,
+      "grad_norm": 3.669849200249159,
+      "learning_rate": 4.831669326829242e-06,
+      "loss": 0.6378,
+      "step": 1773
+    },
+    {
+      "epoch": 0.1440753674977666,
+      "grad_norm": 4.3559205214849674,
+      "learning_rate": 4.831432017766323e-06,
+      "loss": 0.652,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14415658247380816,
+      "grad_norm": 4.444082580638693,
+      "learning_rate": 4.831194547380647e-06,
+      "loss": 0.6826,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14423779744984974,
+      "grad_norm": 7.791669775176109,
+      "learning_rate": 4.830956915688647e-06,
+      "loss": 0.5328,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14431901242589135,
+      "grad_norm": 3.828853400773109,
+      "learning_rate": 4.830719122706764e-06,
+      "loss": 0.7301,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14440022740193292,
+      "grad_norm": 4.69526829505865,
+      "learning_rate": 4.830481168451453e-06,
+      "loss": 0.6871,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1444814423779745,
+      "grad_norm": 3.7376839261484927,
+      "learning_rate": 4.830243052939179e-06,
+      "loss": 0.6172,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14456265735401608,
+      "grad_norm": 5.5007109017935365,
+      "learning_rate": 4.830004776186419e-06,
+      "loss": 0.6784,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14464387233005765,
+      "grad_norm": 9.156018680302578,
+      "learning_rate": 4.82976633820966e-06,
+      "loss": 0.5217,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14472508730609923,
+      "grad_norm": 3.661157732313108,
+      "learning_rate": 4.829527739025399e-06,
+      "loss": 0.7003,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14480630228214084,
+      "grad_norm": 6.275804371905447,
+      "learning_rate": 4.829288978650149e-06,
+      "loss": 0.5237,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1448875172581824,
+      "grad_norm": 6.549033525463337,
+      "learning_rate": 4.829050057100428e-06,
+      "loss": 0.5854,
+      "step": 1784
+    },
+    {
+      "epoch": 0.144968732234224,
+      "grad_norm": 3.236921347753896,
+      "learning_rate": 4.82881097439277e-06,
+      "loss": 0.6303,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14504994721026557,
+      "grad_norm": 5.188648638663512,
+      "learning_rate": 4.828571730543718e-06,
+      "loss": 0.5043,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14513116218630714,
+      "grad_norm": 4.430201943407788,
+      "learning_rate": 4.828332325569825e-06,
+      "loss": 0.6505,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14521237716234875,
+      "grad_norm": 4.8572855854705645,
+      "learning_rate": 4.828092759487658e-06,
+      "loss": 0.6374,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14529359213839033,
+      "grad_norm": 18.826306527694214,
+      "learning_rate": 4.827853032313793e-06,
+      "loss": 0.4907,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1453748071144319,
+      "grad_norm": 7.143933819169128,
+      "learning_rate": 4.827613144064819e-06,
+      "loss": 0.5695,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14545602209047348,
+      "grad_norm": 5.6480765506427755,
+      "learning_rate": 4.827373094757334e-06,
+      "loss": 0.6081,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14553723706651506,
+      "grad_norm": 3.6245079382385987,
+      "learning_rate": 4.827132884407948e-06,
+      "loss": 0.6509,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14561845204255663,
+      "grad_norm": 4.493541124449762,
+      "learning_rate": 4.826892513033283e-06,
+      "loss": 0.6714,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14569966701859824,
+      "grad_norm": 4.222014821722098,
+      "learning_rate": 4.8266519806499705e-06,
+      "loss": 0.6869,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14578088199463982,
+      "grad_norm": 5.40579336748145,
+      "learning_rate": 4.826411287274655e-06,
+      "loss": 0.5754,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1458620969706814,
+      "grad_norm": 7.803642190908184,
+      "learning_rate": 4.82617043292399e-06,
+      "loss": 0.5751,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14594331194672297,
+      "grad_norm": 5.10891502640099,
+      "learning_rate": 4.825929417614643e-06,
+      "loss": 0.5584,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14602452692276455,
+      "grad_norm": 5.342228163843412,
+      "learning_rate": 4.825688241363289e-06,
+      "loss": 0.6378,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14610574189880615,
+      "grad_norm": 3.4939671214065506,
+      "learning_rate": 4.825446904186619e-06,
+      "loss": 0.7725,
+      "step": 1799
+    },
+    {
+      "epoch": 0.14618695687484773,
+      "grad_norm": 7.740693551879523,
+      "learning_rate": 4.825205406101328e-06,
+      "loss": 0.5815,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1462681718508893,
+      "grad_norm": 3.8127593882048676,
+      "learning_rate": 4.824963747124132e-06,
+      "loss": 0.6857,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14634938682693088,
+      "grad_norm": 7.341843917850864,
+      "learning_rate": 4.824721927271747e-06,
+      "loss": 0.5974,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14643060180297246,
+      "grad_norm": 5.649042550435367,
+      "learning_rate": 4.8244799465609095e-06,
+      "loss": 0.6106,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14651181677901404,
+      "grad_norm": 6.644722885021702,
+      "learning_rate": 4.82423780500836e-06,
+      "loss": 0.5848,
+      "step": 1804
+    },
+    {
+      "epoch": 0.14659303175505564,
+      "grad_norm": 3.4917280474695853,
+      "learning_rate": 4.823995502630857e-06,
+      "loss": 0.5819,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14667424673109722,
+      "grad_norm": 4.938957664863549,
+      "learning_rate": 4.823753039445164e-06,
+      "loss": 0.6797,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1467554617071388,
+      "grad_norm": 3.8798922539217955,
+      "learning_rate": 4.823510415468059e-06,
+      "loss": 0.6352,
+      "step": 1807
+    },
+    {
+      "epoch": 0.14683667668318037,
+      "grad_norm": 4.629237568082465,
+      "learning_rate": 4.82326763071633e-06,
+      "loss": 0.6874,
+      "step": 1808
+    },
+    {
+      "epoch": 0.14691789165922195,
+      "grad_norm": 4.720441802830114,
+      "learning_rate": 4.8230246852067784e-06,
+      "loss": 0.587,
+      "step": 1809
+    },
+    {
+      "epoch": 0.14699910663526355,
+      "grad_norm": 5.910430878822207,
+      "learning_rate": 4.822781578956212e-06,
+      "loss": 0.4856,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14708032161130513,
+      "grad_norm": 4.138152132805585,
+      "learning_rate": 4.8225383119814526e-06,
+      "loss": 0.6528,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1471615365873467,
+      "grad_norm": 4.151700339228477,
+      "learning_rate": 4.822294884299335e-06,
+      "loss": 0.552,
+      "step": 1812
+    },
+    {
+      "epoch": 0.14724275156338829,
+      "grad_norm": 4.1614483287452915,
+      "learning_rate": 4.822051295926701e-06,
+      "loss": 0.7129,
+      "step": 1813
+    },
+    {
+      "epoch": 0.14732396653942986,
+      "grad_norm": 6.707224091798943,
+      "learning_rate": 4.821807546880407e-06,
+      "loss": 0.6328,
+      "step": 1814
+    },
+    {
+      "epoch": 0.14740518151547144,
+      "grad_norm": 6.368579491491943,
+      "learning_rate": 4.8215636371773186e-06,
+      "loss": 0.7908,
+      "step": 1815
+    },
+    {
+      "epoch": 0.14748639649151304,
+      "grad_norm": 4.199767598115914,
+      "learning_rate": 4.821319566834314e-06,
+      "loss": 0.5206,
+      "step": 1816
+    },
+    {
+      "epoch": 0.14756761146755462,
+      "grad_norm": 4.709792455873804,
+      "learning_rate": 4.82107533586828e-06,
+      "loss": 0.6195,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1476488264435962,
+      "grad_norm": 4.319113558733222,
+      "learning_rate": 4.820830944296117e-06,
+      "loss": 0.6602,
+      "step": 1818
+    },
+    {
+      "epoch": 0.14773004141963778,
+      "grad_norm": 5.544772853522585,
+      "learning_rate": 4.820586392134735e-06,
+      "loss": 0.5952,
+      "step": 1819
+    },
+    {
+      "epoch": 0.14781125639567935,
+      "grad_norm": 4.015521528170778,
+      "learning_rate": 4.820341679401057e-06,
+      "loss": 0.5967,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14789247137172096,
+      "grad_norm": 3.9474000516118215,
+      "learning_rate": 4.820096806112015e-06,
+      "loss": 0.6292,
+      "step": 1821
+    },
+    {
+      "epoch": 0.14797368634776253,
+      "grad_norm": 4.8859410343320615,
+      "learning_rate": 4.8198517722845524e-06,
+      "loss": 0.4472,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1480549013238041,
+      "grad_norm": 4.277013788450521,
+      "learning_rate": 4.819606577935626e-06,
+      "loss": 0.6099,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1481361162998457,
+      "grad_norm": 4.8797029682772095,
+      "learning_rate": 4.8193612230822e-06,
+      "loss": 0.598,
+      "step": 1824
+    },
+    {
+      "epoch": 0.14821733127588727,
+      "grad_norm": 7.742169376350744,
+      "learning_rate": 4.819115707741252e-06,
+      "loss": 0.6474,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14829854625192884,
+      "grad_norm": 5.12268577416012,
+      "learning_rate": 4.818870031929771e-06,
+      "loss": 0.4876,
+      "step": 1826
+    },
+    {
+      "epoch": 0.14837976122797045,
+      "grad_norm": 3.7709678599136605,
+      "learning_rate": 4.818624195664756e-06,
+      "loss": 0.727,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14846097620401202,
+      "grad_norm": 2.5702762627716833,
+      "learning_rate": 4.818378198963218e-06,
+      "loss": 0.7224,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1485421911800536,
+      "grad_norm": 3.097076580981341,
+      "learning_rate": 4.81813204184218e-06,
+      "loss": 0.6025,
+      "step": 1829
+    },
+    {
+      "epoch": 0.14862340615609518,
+      "grad_norm": 7.135529520333667,
+      "learning_rate": 4.817885724318671e-06,
+      "loss": 0.6685,
+      "step": 1830
+    },
+    {
+      "epoch": 0.14870462113213676,
+      "grad_norm": 3.4538569552830003,
+      "learning_rate": 4.817639246409738e-06,
+      "loss": 0.6588,
+      "step": 1831
+    },
+    {
+      "epoch": 0.14878583610817836,
+      "grad_norm": 4.3881359378371965,
+      "learning_rate": 4.817392608132435e-06,
+      "loss": 0.4665,
+      "step": 1832
+    },
+    {
+      "epoch": 0.14886705108421994,
+      "grad_norm": 3.7661484270198304,
+      "learning_rate": 4.817145809503828e-06,
+      "loss": 0.5136,
+      "step": 1833
+    },
+    {
+      "epoch": 0.1489482660602615,
+      "grad_norm": 5.494631448466639,
+      "learning_rate": 4.816898850540995e-06,
+      "loss": 0.5539,
+      "step": 1834
+    },
+    {
+      "epoch": 0.1490294810363031,
+      "grad_norm": 10.552148803818461,
+      "learning_rate": 4.816651731261023e-06,
+      "loss": 0.6492,
+      "step": 1835
+    },
+    {
+      "epoch": 0.14911069601234467,
+      "grad_norm": 4.963897986383459,
+      "learning_rate": 4.816404451681012e-06,
+      "loss": 0.6898,
+      "step": 1836
+    },
+    {
+      "epoch": 0.14919191098838624,
+      "grad_norm": 4.656896578348469,
+      "learning_rate": 4.816157011818073e-06,
+      "loss": 0.6373,
+      "step": 1837
+    },
+    {
+      "epoch": 0.14927312596442785,
+      "grad_norm": 6.1635228629484935,
+      "learning_rate": 4.815909411689326e-06,
+      "loss": 0.61,
+      "step": 1838
+    },
+    {
+      "epoch": 0.14935434094046943,
+      "grad_norm": 5.077711445021842,
+      "learning_rate": 4.815661651311905e-06,
+      "loss": 0.5255,
+      "step": 1839
+    },
+    {
+      "epoch": 0.149435555916511,
+      "grad_norm": 3.5697847643358926,
+      "learning_rate": 4.815413730702953e-06,
+      "loss": 0.5737,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14951677089255258,
+      "grad_norm": 4.3625630887558025,
+      "learning_rate": 4.8151656498796245e-06,
+      "loss": 0.5621,
+      "step": 1841
+    },
+    {
+      "epoch": 0.14959798586859416,
+      "grad_norm": 3.632848050826748,
+      "learning_rate": 4.814917408859087e-06,
+      "loss": 0.6579,
+      "step": 1842
+    },
+    {
+      "epoch": 0.14967920084463576,
+      "grad_norm": 5.537325255170911,
+      "learning_rate": 4.8146690076585145e-06,
+      "loss": 0.6326,
+      "step": 1843
+    },
+    {
+      "epoch": 0.14976041582067734,
+      "grad_norm": 4.502954580803916,
+      "learning_rate": 4.8144204462950985e-06,
+      "loss": 0.5227,
+      "step": 1844
+    },
+    {
+      "epoch": 0.14984163079671892,
+      "grad_norm": 9.989837223985806,
+      "learning_rate": 4.8141717247860355e-06,
+      "loss": 0.5365,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1499228457727605,
+      "grad_norm": 9.268156392807445,
+      "learning_rate": 4.813922843148537e-06,
+      "loss": 0.7256,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15000406074880207,
+      "grad_norm": 4.723473199759222,
+      "learning_rate": 4.813673801399825e-06,
+      "loss": 0.6814,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15008527572484365,
+      "grad_norm": 5.682840300388035,
+      "learning_rate": 4.81342459955713e-06,
+      "loss": 0.6083,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15016649070088525,
+      "grad_norm": 3.7968600757264417,
+      "learning_rate": 4.813175237637697e-06,
+      "loss": 0.5635,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15024770567692683,
+      "grad_norm": 4.69369222204755,
+      "learning_rate": 4.812925715658779e-06,
+      "loss": 0.5237,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1503289206529684,
+      "grad_norm": 4.060309446961183,
+      "learning_rate": 4.812676033637643e-06,
+      "loss": 0.6771,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15041013562900998,
+      "grad_norm": 4.068231340082502,
+      "learning_rate": 4.812426191591565e-06,
+      "loss": 0.4976,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15049135060505156,
+      "grad_norm": 2.940489927194899,
+      "learning_rate": 4.812176189537833e-06,
+      "loss": 0.6372,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15057256558109317,
+      "grad_norm": 18.577063761435014,
+      "learning_rate": 4.811926027493745e-06,
+      "loss": 0.5688,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15065378055713474,
+      "grad_norm": 9.15886334831888,
+      "learning_rate": 4.811675705476613e-06,
+      "loss": 0.5274,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15073499553317632,
+      "grad_norm": 8.116139489468573,
+      "learning_rate": 4.811425223503755e-06,
+      "loss": 0.6006,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1508162105092179,
+      "grad_norm": 5.340837408995612,
+      "learning_rate": 4.811174581592506e-06,
+      "loss": 0.4627,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15089742548525947,
+      "grad_norm": 7.861074956324514,
+      "learning_rate": 4.810923779760207e-06,
+      "loss": 0.6517,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15097864046130105,
+      "grad_norm": 5.046613794107396,
+      "learning_rate": 4.810672818024212e-06,
+      "loss": 0.8007,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15105985543734265,
+      "grad_norm": 3.78641177578102,
+      "learning_rate": 4.810421696401889e-06,
+      "loss": 0.4462,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15114107041338423,
+      "grad_norm": 5.953501964167005,
+      "learning_rate": 4.810170414910611e-06,
+      "loss": 0.6483,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1512222853894258,
+      "grad_norm": 4.508454868931279,
+      "learning_rate": 4.809918973567767e-06,
+      "loss": 0.4402,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15130350036546739,
+      "grad_norm": 6.1091026815566645,
+      "learning_rate": 4.809667372390755e-06,
+      "loss": 0.5401,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15138471534150896,
+      "grad_norm": 4.269870818997486,
+      "learning_rate": 4.809415611396984e-06,
+      "loss": 0.6984,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15146593031755057,
+      "grad_norm": 8.488877241499456,
+      "learning_rate": 4.809163690603877e-06,
+      "loss": 0.6111,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15154714529359214,
+      "grad_norm": 5.571115307591249,
+      "learning_rate": 4.808911610028861e-06,
+      "loss": 0.5755,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15162836026963372,
+      "grad_norm": 5.592522309071258,
+      "learning_rate": 4.808659369689384e-06,
+      "loss": 0.7285,
+      "step": 1867
+    },
+    {
+      "epoch": 0.1517095752456753,
+      "grad_norm": 3.3631906733071535,
+      "learning_rate": 4.808406969602895e-06,
+      "loss": 0.6066,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15179079022171688,
+      "grad_norm": 5.558011990131933,
+      "learning_rate": 4.8081544097868615e-06,
+      "loss": 0.599,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15187200519775845,
+      "grad_norm": 4.196854563685466,
+      "learning_rate": 4.8079016902587586e-06,
+      "loss": 0.6429,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15195322017380006,
+      "grad_norm": 5.663489456766142,
+      "learning_rate": 4.807648811036073e-06,
+      "loss": 0.4956,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15203443514984163,
+      "grad_norm": 4.348733009474466,
+      "learning_rate": 4.807395772136303e-06,
+      "loss": 0.5331,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1521156501258832,
+      "grad_norm": 5.237186173256271,
+      "learning_rate": 4.807142573576958e-06,
+      "loss": 0.7043,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1521968651019248,
+      "grad_norm": 5.71550907696815,
+      "learning_rate": 4.806889215375556e-06,
+      "loss": 0.491,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15227808007796637,
+      "grad_norm": 5.112913674384685,
+      "learning_rate": 4.80663569754963e-06,
+      "loss": 0.5455,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15235929505400797,
+      "grad_norm": 4.422551589450753,
+      "learning_rate": 4.806382020116721e-06,
+      "loss": 0.6936,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15244051003004955,
+      "grad_norm": 4.418250640226396,
+      "learning_rate": 4.806128183094383e-06,
+      "loss": 0.6371,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15252172500609112,
+      "grad_norm": 4.611528104096768,
+      "learning_rate": 4.805874186500179e-06,
+      "loss": 0.5783,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1526029399821327,
+      "grad_norm": 4.3459558003142345,
+      "learning_rate": 4.805620030351686e-06,
+      "loss": 0.6227,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15268415495817428,
+      "grad_norm": 6.170103993151758,
+      "learning_rate": 4.805365714666489e-06,
+      "loss": 0.5419,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15276536993421586,
+      "grad_norm": 2.958664035946074,
+      "learning_rate": 4.805111239462185e-06,
+      "loss": 0.7513,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15284658491025746,
+      "grad_norm": 13.416473280689857,
+      "learning_rate": 4.8048566047563835e-06,
+      "loss": 0.4584,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15292779988629904,
+      "grad_norm": 4.99824146528109,
+      "learning_rate": 4.8046018105667024e-06,
+      "loss": 0.6614,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15300901486234061,
+      "grad_norm": 9.102398270856987,
+      "learning_rate": 4.8043468569107735e-06,
+      "loss": 0.7377,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1530902298383822,
+      "grad_norm": 3.746424579208254,
+      "learning_rate": 4.804091743806237e-06,
+      "loss": 0.4525,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15317144481442377,
+      "grad_norm": 6.368130414333825,
+      "learning_rate": 4.803836471270748e-06,
+      "loss": 0.5629,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15325265979046537,
+      "grad_norm": 6.446103551439812,
+      "learning_rate": 4.803581039321966e-06,
+      "loss": 0.5416,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15333387476650695,
+      "grad_norm": 5.229434076566928,
+      "learning_rate": 4.803325447977568e-06,
+      "loss": 0.5842,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15341508974254853,
+      "grad_norm": 5.954690104051001,
+      "learning_rate": 4.80306969725524e-06,
+      "loss": 0.5898,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1534963047185901,
+      "grad_norm": 6.150359963901614,
+      "learning_rate": 4.802813787172678e-06,
+      "loss": 0.5839,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15357751969463168,
+      "grad_norm": 5.125992935197739,
+      "learning_rate": 4.802557717747588e-06,
+      "loss": 0.5668,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15365873467067329,
+      "grad_norm": 4.364632434646274,
+      "learning_rate": 4.802301488997691e-06,
+      "loss": 0.4172,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15373994964671486,
+      "grad_norm": 4.882188374905198,
+      "learning_rate": 4.802045100940715e-06,
+      "loss": 0.6242,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15382116462275644,
+      "grad_norm": 3.3401070060949403,
+      "learning_rate": 4.801788553594403e-06,
+      "loss": 0.5872,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15390237959879802,
+      "grad_norm": 4.868516648657934,
+      "learning_rate": 4.801531846976504e-06,
+      "loss": 0.5688,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1539835945748396,
+      "grad_norm": 5.105870944801884,
+      "learning_rate": 4.801274981104781e-06,
+      "loss": 0.6434,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15406480955088117,
+      "grad_norm": 3.8171439187123903,
+      "learning_rate": 4.80101795599701e-06,
+      "loss": 0.5289,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15414602452692278,
+      "grad_norm": 16.280448607344095,
+      "learning_rate": 4.800760771670974e-06,
+      "loss": 0.4721,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15422723950296435,
+      "grad_norm": 3.7231117337427055,
+      "learning_rate": 4.800503428144469e-06,
+      "loss": 0.6052,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15430845447900593,
+      "grad_norm": 3.648455745421052,
+      "learning_rate": 4.800245925435302e-06,
+      "loss": 0.6106,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1543896694550475,
+      "grad_norm": 6.0421687476954995,
+      "learning_rate": 4.7999882635612916e-06,
+      "loss": 0.5272,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15447088443108908,
+      "grad_norm": 4.462318149762371,
+      "learning_rate": 4.799730442540265e-06,
+      "loss": 0.4801,
+      "step": 1902
+    },
+    {
+      "epoch": 0.1545520994071307,
+      "grad_norm": 5.097311868178371,
+      "learning_rate": 4.7994724623900636e-06,
+      "loss": 0.6126,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15463331438317227,
+      "grad_norm": 10.491382918494583,
+      "learning_rate": 4.799214323128537e-06,
+      "loss": 0.7263,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15471452935921384,
+      "grad_norm": 5.632700323452676,
+      "learning_rate": 4.798956024773548e-06,
+      "loss": 0.427,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15479574433525542,
+      "grad_norm": 4.125424526878242,
+      "learning_rate": 4.798697567342969e-06,
+      "loss": 0.8174,
+      "step": 1906
+    },
+    {
+      "epoch": 0.154876959311297,
+      "grad_norm": 3.2529153125095682,
+      "learning_rate": 4.798438950854685e-06,
+      "loss": 0.5607,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15495817428733857,
+      "grad_norm": 6.639548071036903,
+      "learning_rate": 4.798180175326589e-06,
+      "loss": 0.6207,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15503938926338018,
+      "grad_norm": 3.419026777929521,
+      "learning_rate": 4.797921240776587e-06,
+      "loss": 0.6548,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15512060423942176,
+      "grad_norm": 6.3759242855440315,
+      "learning_rate": 4.797662147222598e-06,
+      "loss": 0.722,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15520181921546333,
+      "grad_norm": 4.2074330009622525,
+      "learning_rate": 4.797402894682548e-06,
+      "loss": 0.7711,
+      "step": 1911
+    },
+    {
+      "epoch": 0.1552830341915049,
+      "grad_norm": 5.86918562130916,
+      "learning_rate": 4.797143483174377e-06,
+      "loss": 0.5241,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1553642491675465,
+      "grad_norm": 4.110856189622478,
+      "learning_rate": 4.796883912716034e-06,
+      "loss": 0.6855,
+      "step": 1913
+    },
+    {
+      "epoch": 0.1554454641435881,
+      "grad_norm": 4.190357484842863,
+      "learning_rate": 4.79662418332548e-06,
+      "loss": 0.616,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15552667911962967,
+      "grad_norm": 6.818021899205355,
+      "learning_rate": 4.796364295020688e-06,
+      "loss": 0.554,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15560789409567125,
+      "grad_norm": 4.400441609364029,
+      "learning_rate": 4.7961042478196394e-06,
+      "loss": 0.5425,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15568910907171282,
+      "grad_norm": 3.755867976669089,
+      "learning_rate": 4.7958440417403295e-06,
+      "loss": 0.5295,
+      "step": 1917
+    },
+    {
+      "epoch": 0.1557703240477544,
+      "grad_norm": 3.799331044287829,
+      "learning_rate": 4.795583676800762e-06,
+      "loss": 0.4355,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15585153902379598,
+      "grad_norm": 5.47739825987387,
+      "learning_rate": 4.795323153018953e-06,
+      "loss": 0.6229,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15593275399983758,
+      "grad_norm": 4.656969062795371,
+      "learning_rate": 4.795062470412931e-06,
+      "loss": 0.6301,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15601396897587916,
+      "grad_norm": 4.914269261314513,
+      "learning_rate": 4.794801629000732e-06,
+      "loss": 0.6974,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15609518395192074,
+      "grad_norm": 6.004759949716761,
+      "learning_rate": 4.794540628800405e-06,
+      "loss": 0.6272,
+      "step": 1922
+    },
+    {
+      "epoch": 0.1561763989279623,
+      "grad_norm": 5.879866804838136,
+      "learning_rate": 4.79427946983001e-06,
+      "loss": 0.624,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1562576139040039,
+      "grad_norm": 4.9010570268519835,
+      "learning_rate": 4.794018152107618e-06,
+      "loss": 0.5972,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1563388288800455,
+      "grad_norm": 6.1005563812014625,
+      "learning_rate": 4.793756675651311e-06,
+      "loss": 0.5372,
+      "step": 1925
+    },
+    {
+      "epoch": 0.15642004385608707,
+      "grad_norm": 7.657069975595423,
+      "learning_rate": 4.7934950404791815e-06,
+      "loss": 0.5059,
+      "step": 1926
+    },
+    {
+      "epoch": 0.15650125883212865,
+      "grad_norm": 3.7164386302842325,
+      "learning_rate": 4.793233246609333e-06,
+      "loss": 0.5145,
+      "step": 1927
+    },
+    {
+      "epoch": 0.15658247380817022,
+      "grad_norm": 6.7098122658626025,
+      "learning_rate": 4.792971294059882e-06,
+      "loss": 0.5909,
+      "step": 1928
+    },
+    {
+      "epoch": 0.1566636887842118,
+      "grad_norm": 4.5647090456408765,
+      "learning_rate": 4.792709182848951e-06,
+      "loss": 0.5851,
+      "step": 1929
+    },
+    {
+      "epoch": 0.15674490376025338,
+      "grad_norm": 5.659279097889884,
+      "learning_rate": 4.792446912994679e-06,
+      "loss": 0.509,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15682611873629498,
+      "grad_norm": 4.660823647963301,
+      "learning_rate": 4.792184484515214e-06,
+      "loss": 0.6214,
+      "step": 1931
+    },
+    {
+      "epoch": 0.15690733371233656,
+      "grad_norm": 5.379888896939118,
+      "learning_rate": 4.791921897428714e-06,
+      "loss": 0.6083,
+      "step": 1932
+    },
+    {
+      "epoch": 0.15698854868837814,
+      "grad_norm": 3.7091543807048555,
+      "learning_rate": 4.791659151753348e-06,
+      "loss": 0.5366,
+      "step": 1933
+    },
+    {
+      "epoch": 0.15706976366441971,
+      "grad_norm": 4.46536923401703,
+      "learning_rate": 4.791396247507297e-06,
+      "loss": 0.5548,
+      "step": 1934
+    },
+    {
+      "epoch": 0.1571509786404613,
+      "grad_norm": 3.0478693770891,
+      "learning_rate": 4.791133184708753e-06,
+      "loss": 0.6562,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1572321936165029,
+      "grad_norm": 5.052248542138617,
+      "learning_rate": 4.790869963375918e-06,
+      "loss": 0.6266,
+      "step": 1936
+    },
+    {
+      "epoch": 0.15731340859254447,
+      "grad_norm": 6.299780666458338,
+      "learning_rate": 4.790606583527006e-06,
+      "loss": 0.6602,
+      "step": 1937
+    },
+    {
+      "epoch": 0.15739462356858605,
+      "grad_norm": 7.111506262099773,
+      "learning_rate": 4.790343045180242e-06,
+      "loss": 0.6822,
+      "step": 1938
+    },
+    {
+      "epoch": 0.15747583854462763,
+      "grad_norm": 3.5717735290621015,
+      "learning_rate": 4.790079348353859e-06,
+      "loss": 0.5579,
+      "step": 1939
+    },
+    {
+      "epoch": 0.1575570535206692,
+      "grad_norm": 4.539252075565922,
+      "learning_rate": 4.789815493066106e-06,
+      "loss": 0.5513,
+      "step": 1940
+    },
+    {
+      "epoch": 0.15763826849671078,
+      "grad_norm": 3.19405646131031,
+      "learning_rate": 4.78955147933524e-06,
+      "loss": 0.6305,
+      "step": 1941
+    },
+    {
+      "epoch": 0.1577194834727524,
+      "grad_norm": 3.9600914109632805,
+      "learning_rate": 4.7892873071795285e-06,
+      "loss": 0.5575,
+      "step": 1942
+    },
+    {
+      "epoch": 0.15780069844879396,
+      "grad_norm": 5.033361013107082,
+      "learning_rate": 4.789022976617251e-06,
+      "loss": 0.6199,
+      "step": 1943
+    },
+    {
+      "epoch": 0.15788191342483554,
+      "grad_norm": 15.306333789764619,
+      "learning_rate": 4.7887584876666984e-06,
+      "loss": 0.458,
+      "step": 1944
+    },
+    {
+      "epoch": 0.15796312840087712,
+      "grad_norm": 4.337274209547185,
+      "learning_rate": 4.788493840346172e-06,
+      "loss": 0.691,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1580443433769187,
+      "grad_norm": 4.2820007012074415,
+      "learning_rate": 4.788229034673983e-06,
+      "loss": 0.7317,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1581255583529603,
+      "grad_norm": 4.361412328827347,
+      "learning_rate": 4.787964070668455e-06,
+      "loss": 0.5208,
+      "step": 1947
+    },
+    {
+      "epoch": 0.15820677332900188,
+      "grad_norm": 4.543017266735986,
+      "learning_rate": 4.787698948347922e-06,
+      "loss": 0.7488,
+      "step": 1948
+    },
+    {
+      "epoch": 0.15828798830504345,
+      "grad_norm": 6.9360624754566444,
+      "learning_rate": 4.78743366773073e-06,
+      "loss": 0.478,
+      "step": 1949
+    },
+    {
+      "epoch": 0.15836920328108503,
+      "grad_norm": 4.321466474176901,
+      "learning_rate": 4.787168228835234e-06,
+      "loss": 0.8373,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1584504182571266,
+      "grad_norm": 10.29187836219349,
+      "learning_rate": 4.7869026316798005e-06,
+      "loss": 0.5927,
+      "step": 1951
+    },
+    {
+      "epoch": 0.15853163323316818,
+      "grad_norm": 5.206315709924438,
+      "learning_rate": 4.7866368762828095e-06,
+      "loss": 0.7284,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1586128482092098,
+      "grad_norm": 4.695971985775556,
+      "learning_rate": 4.786370962662647e-06,
+      "loss": 0.6129,
+      "step": 1953
+    },
+    {
+      "epoch": 0.15869406318525137,
+      "grad_norm": 3.810668859896304,
+      "learning_rate": 4.786104890837715e-06,
+      "loss": 0.7817,
+      "step": 1954
+    },
+    {
+      "epoch": 0.15877527816129294,
+      "grad_norm": 5.945278033090391,
+      "learning_rate": 4.785838660826424e-06,
+      "loss": 0.4929,
+      "step": 1955
+    },
+    {
+      "epoch": 0.15885649313733452,
+      "grad_norm": 5.613790832997751,
+      "learning_rate": 4.785572272647196e-06,
+      "loss": 0.5365,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1589377081133761,
+      "grad_norm": 12.194006996853599,
+      "learning_rate": 4.785305726318461e-06,
+      "loss": 0.5274,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1590189230894177,
+      "grad_norm": 6.67246997329038,
+      "learning_rate": 4.785039021858665e-06,
+      "loss": 0.5129,
+      "step": 1958
+    },
+    {
+      "epoch": 0.15910013806545928,
+      "grad_norm": 4.361772829570909,
+      "learning_rate": 4.784772159286263e-06,
+      "loss": 0.5762,
+      "step": 1959
+    },
+    {
+      "epoch": 0.15918135304150086,
+      "grad_norm": 3.6171339798903115,
+      "learning_rate": 4.784505138619719e-06,
+      "loss": 0.5687,
+      "step": 1960
+    },
+    {
+      "epoch": 0.15926256801754243,
+      "grad_norm": 3.3498401020879696,
+      "learning_rate": 4.78423795987751e-06,
+      "loss": 0.6731,
+      "step": 1961
+    },
+    {
+      "epoch": 0.159343782993584,
+      "grad_norm": 4.218304560758474,
+      "learning_rate": 4.783970623078124e-06,
+      "loss": 0.5832,
+      "step": 1962
+    },
+    {
+      "epoch": 0.1594249979696256,
+      "grad_norm": 3.2954022912979473,
+      "learning_rate": 4.783703128240058e-06,
+      "loss": 0.626,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1595062129456672,
+      "grad_norm": 5.637076756015832,
+      "learning_rate": 4.783435475381822e-06,
+      "loss": 0.549,
+      "step": 1964
+    },
+    {
+      "epoch": 0.15958742792170877,
+      "grad_norm": 4.690580073274557,
+      "learning_rate": 4.7831676645219364e-06,
+      "loss": 0.6577,
+      "step": 1965
+    },
+    {
+      "epoch": 0.15966864289775035,
+      "grad_norm": 5.3749569715672125,
+      "learning_rate": 4.782899695678931e-06,
+      "loss": 0.5685,
+      "step": 1966
+    },
+    {
+      "epoch": 0.15974985787379192,
+      "grad_norm": 6.809233492187388,
+      "learning_rate": 4.782631568871349e-06,
+      "loss": 0.6891,
+      "step": 1967
+    },
+    {
+      "epoch": 0.1598310728498335,
+      "grad_norm": 4.530421811396376,
+      "learning_rate": 4.782363284117744e-06,
+      "loss": 0.3833,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1599122878258751,
+      "grad_norm": 4.835132250782642,
+      "learning_rate": 4.782094841436677e-06,
+      "loss": 0.4926,
+      "step": 1969
+    },
+    {
+      "epoch": 0.15999350280191668,
+      "grad_norm": 3.8849675683937184,
+      "learning_rate": 4.781826240846726e-06,
+      "loss": 0.7052,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16007471777795826,
+      "grad_norm": 4.260063677720038,
+      "learning_rate": 4.781557482366477e-06,
+      "loss": 0.6791,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16015593275399984,
+      "grad_norm": 9.298148375842267,
+      "learning_rate": 4.781288566014524e-06,
+      "loss": 0.6356,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1602371477300414,
+      "grad_norm": 4.484161226398304,
+      "learning_rate": 4.781019491809475e-06,
+      "loss": 0.5682,
+      "step": 1973
+    },
+    {
+      "epoch": 0.160318362706083,
+      "grad_norm": 3.7283315441063984,
+      "learning_rate": 4.78075025976995e-06,
+      "loss": 0.6072,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1603995776821246,
+      "grad_norm": 6.059353678550269,
+      "learning_rate": 4.780480869914578e-06,
+      "loss": 0.5845,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16048079265816617,
+      "grad_norm": 4.554935916876435,
+      "learning_rate": 4.780211322261998e-06,
+      "loss": 0.5095,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16056200763420775,
+      "grad_norm": 6.981060106812001,
+      "learning_rate": 4.779941616830863e-06,
+      "loss": 0.5098,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16064322261024933,
+      "grad_norm": 4.9440805505487,
+      "learning_rate": 4.779671753639835e-06,
+      "loss": 0.7675,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1607244375862909,
+      "grad_norm": 8.280400518698295,
+      "learning_rate": 4.779401732707586e-06,
+      "loss": 0.5639,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1608056525623325,
+      "grad_norm": 6.056327257248986,
+      "learning_rate": 4.779131554052801e-06,
+      "loss": 0.5981,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16088686753837408,
+      "grad_norm": 5.385241894979969,
+      "learning_rate": 4.778861217694174e-06,
+      "loss": 0.6512,
+      "step": 1981
+    },
+    {
+      "epoch": 0.16096808251441566,
+      "grad_norm": 4.550144423102369,
+      "learning_rate": 4.778590723650413e-06,
+      "loss": 0.5836,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16104929749045724,
+      "grad_norm": 4.3719124593201055,
+      "learning_rate": 4.778320071940231e-06,
+      "loss": 0.7309,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16113051246649882,
+      "grad_norm": 4.283507192707898,
+      "learning_rate": 4.77804926258236e-06,
+      "loss": 0.5897,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1612117274425404,
+      "grad_norm": 3.9610170071985524,
+      "learning_rate": 4.777778295595535e-06,
+      "loss": 0.6077,
+      "step": 1985
+    },
+    {
+      "epoch": 0.161292942418582,
+      "grad_norm": 6.219189683636367,
+      "learning_rate": 4.777507170998508e-06,
+      "loss": 0.7406,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16137415739462357,
+      "grad_norm": 4.465198885140077,
+      "learning_rate": 4.777235888810037e-06,
+      "loss": 0.5908,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16145537237066515,
+      "grad_norm": 6.450650746265666,
+      "learning_rate": 4.776964449048895e-06,
+      "loss": 0.5436,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16153658734670673,
+      "grad_norm": 9.863744897399046,
+      "learning_rate": 4.776692851733864e-06,
+      "loss": 0.4796,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1616178023227483,
+      "grad_norm": 3.703087572235551,
+      "learning_rate": 4.776421096883737e-06,
+      "loss": 0.737,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1616990172987899,
+      "grad_norm": 5.885859853270686,
+      "learning_rate": 4.776149184517318e-06,
+      "loss": 0.7037,
+      "step": 1991
+    },
+    {
+      "epoch": 0.1617802322748315,
+      "grad_norm": 4.7082939980756136,
+      "learning_rate": 4.775877114653422e-06,
+      "loss": 0.5807,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16186144725087306,
+      "grad_norm": 4.719665547064216,
+      "learning_rate": 4.775604887310874e-06,
+      "loss": 0.5684,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16194266222691464,
+      "grad_norm": 4.372013687616748,
+      "learning_rate": 4.775332502508511e-06,
+      "loss": 0.5176,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16202387720295622,
+      "grad_norm": 3.574773949292872,
+      "learning_rate": 4.775059960265181e-06,
+      "loss": 0.6306,
+      "step": 1995
+    },
+    {
+      "epoch": 0.1621050921789978,
+      "grad_norm": 3.0419771807588134,
+      "learning_rate": 4.774787260599744e-06,
+      "loss": 0.5673,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1621863071550394,
+      "grad_norm": 4.11991510271793,
+      "learning_rate": 4.7745144035310656e-06,
+      "loss": 0.5221,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16226752213108098,
+      "grad_norm": 8.878399802384537,
+      "learning_rate": 4.77424138907803e-06,
+      "loss": 0.5639,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16234873710712255,
+      "grad_norm": 5.661428317678857,
+      "learning_rate": 4.773968217259525e-06,
+      "loss": 0.6172,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16242995208316413,
+      "grad_norm": 4.587434812251549,
+      "learning_rate": 4.773694888094454e-06,
+      "loss": 0.5641,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1625111670592057,
+      "grad_norm": 6.782537400712025,
+      "learning_rate": 4.773421401601731e-06,
+      "loss": 0.6165,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1625923820352473,
+      "grad_norm": 5.248837296847351,
+      "learning_rate": 4.773147757800279e-06,
+      "loss": 0.5165,
+      "step": 2002
+    },
+    {
+      "epoch": 0.1626735970112889,
+      "grad_norm": 6.771172761964919,
+      "learning_rate": 4.772873956709032e-06,
+      "loss": 0.5348,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16275481198733047,
+      "grad_norm": 3.7091479668688407,
+      "learning_rate": 4.772599998346937e-06,
+      "loss": 0.7519,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16283602696337204,
+      "grad_norm": 4.478594715066707,
+      "learning_rate": 4.772325882732949e-06,
+      "loss": 0.5111,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16291724193941362,
+      "grad_norm": 5.771923023591905,
+      "learning_rate": 4.772051609886036e-06,
+      "loss": 0.4556,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1629984569154552,
+      "grad_norm": 4.808968764339891,
+      "learning_rate": 4.771777179825176e-06,
+      "loss": 0.5635,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1630796718914968,
+      "grad_norm": 6.317225738730894,
+      "learning_rate": 4.7715025925693595e-06,
+      "loss": 0.8752,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16316088686753838,
+      "grad_norm": 4.920049883294183,
+      "learning_rate": 4.771227848137585e-06,
+      "loss": 0.6393,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16324210184357996,
+      "grad_norm": 4.256612076091997,
+      "learning_rate": 4.770952946548864e-06,
+      "loss": 0.6003,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16332331681962153,
+      "grad_norm": 4.5734192613631315,
+      "learning_rate": 4.770677887822217e-06,
+      "loss": 0.4413,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1634045317956631,
+      "grad_norm": 4.163070055535816,
+      "learning_rate": 4.770402671976677e-06,
+      "loss": 0.6597,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16348574677170472,
+      "grad_norm": 3.0742054711235887,
+      "learning_rate": 4.77012729903129e-06,
+      "loss": 0.4076,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1635669617477463,
+      "grad_norm": 5.55441667151701,
+      "learning_rate": 4.769851769005107e-06,
+      "loss": 0.5314,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16364817672378787,
+      "grad_norm": 4.576121873555356,
+      "learning_rate": 4.769576081917195e-06,
+      "loss": 0.6124,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16372939169982945,
+      "grad_norm": 6.382821144731639,
+      "learning_rate": 4.7693002377866295e-06,
+      "loss": 0.6729,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16381060667587102,
+      "grad_norm": 9.489241945586448,
+      "learning_rate": 4.769024236632498e-06,
+      "loss": 0.4966,
+      "step": 2017
+    },
+    {
+      "epoch": 0.1638918216519126,
+      "grad_norm": 10.574632939063292,
+      "learning_rate": 4.768748078473898e-06,
+      "loss": 0.6579,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1639730366279542,
+      "grad_norm": 3.323125086340652,
+      "learning_rate": 4.768471763329938e-06,
+      "loss": 0.9138,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16405425160399578,
+      "grad_norm": 14.315467246083593,
+      "learning_rate": 4.768195291219738e-06,
+      "loss": 0.4129,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16413546658003736,
+      "grad_norm": 3.5370073104621613,
+      "learning_rate": 4.767918662162428e-06,
+      "loss": 0.6045,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16421668155607894,
+      "grad_norm": 4.952315377512936,
+      "learning_rate": 4.767641876177149e-06,
+      "loss": 0.5661,
+      "step": 2022
+    },
+    {
+      "epoch": 0.1642978965321205,
+      "grad_norm": 2.9834860448273273,
+      "learning_rate": 4.767364933283053e-06,
+      "loss": 0.5964,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16437911150816212,
+      "grad_norm": 38.67374390447997,
+      "learning_rate": 4.767087833499305e-06,
+      "loss": 0.545,
+      "step": 2024
+    },
+    {
+      "epoch": 0.1644603264842037,
+      "grad_norm": 8.792176852895958,
+      "learning_rate": 4.7668105768450755e-06,
+      "loss": 0.6039,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16454154146024527,
+      "grad_norm": 8.176394720323723,
+      "learning_rate": 4.766533163339553e-06,
+      "loss": 0.5289,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16462275643628685,
+      "grad_norm": 6.977860465026282,
+      "learning_rate": 4.766255593001929e-06,
+      "loss": 0.4951,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16470397141232843,
+      "grad_norm": 5.399038234082669,
+      "learning_rate": 4.765977865851413e-06,
+      "loss": 0.4714,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16478518638837,
+      "grad_norm": 6.62984701521281,
+      "learning_rate": 4.765699981907221e-06,
+      "loss": 0.5556,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1648664013644116,
+      "grad_norm": 6.225819915723398,
+      "learning_rate": 4.765421941188582e-06,
+      "loss": 0.8441,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16494761634045318,
+      "grad_norm": 5.344504032565665,
+      "learning_rate": 4.765143743714734e-06,
+      "loss": 0.411,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16502883131649476,
+      "grad_norm": 6.407678819914006,
+      "learning_rate": 4.764865389504927e-06,
+      "loss": 0.5036,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16511004629253634,
+      "grad_norm": 4.502779183775894,
+      "learning_rate": 4.764586878578421e-06,
+      "loss": 0.5595,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16519126126857792,
+      "grad_norm": 5.385105120731812,
+      "learning_rate": 4.7643082109544894e-06,
+      "loss": 0.5466,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16527247624461952,
+      "grad_norm": 5.3322559203413435,
+      "learning_rate": 4.764029386652412e-06,
+      "loss": 0.5369,
+      "step": 2035
+    },
+    {
+      "epoch": 0.1653536912206611,
+      "grad_norm": 5.983847101773067,
+      "learning_rate": 4.763750405691483e-06,
+      "loss": 0.5569,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16543490619670267,
+      "grad_norm": 4.212501542827999,
+      "learning_rate": 4.7634712680910075e-06,
+      "loss": 0.6493,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16551612117274425,
+      "grad_norm": 5.116892931497582,
+      "learning_rate": 4.7631919738703e-06,
+      "loss": 0.511,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16559733614878583,
+      "grad_norm": 6.147176716463638,
+      "learning_rate": 4.762912523048685e-06,
+      "loss": 0.5332,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1656785511248274,
+      "grad_norm": 4.334801382262547,
+      "learning_rate": 4.7626329156455e-06,
+      "loss": 0.843,
+      "step": 2040
+    },
+    {
+      "epoch": 0.165759766100869,
+      "grad_norm": 6.491344666003298,
+      "learning_rate": 4.7623531516800916e-06,
+      "loss": 0.5301,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1658409810769106,
+      "grad_norm": 4.47643734321392,
+      "learning_rate": 4.762073231171819e-06,
+      "loss": 0.7063,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16592219605295216,
+      "grad_norm": 4.934657642144628,
+      "learning_rate": 4.76179315414005e-06,
+      "loss": 0.6217,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16600341102899374,
+      "grad_norm": 6.800502146070916,
+      "learning_rate": 4.761512920604165e-06,
+      "loss": 0.5184,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16608462600503532,
+      "grad_norm": 4.106268196492193,
+      "learning_rate": 4.761232530583556e-06,
+      "loss": 0.4322,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16616584098107692,
+      "grad_norm": 4.860918988738907,
+      "learning_rate": 4.760951984097622e-06,
+      "loss": 0.541,
+      "step": 2046
+    },
+    {
+      "epoch": 0.1662470559571185,
+      "grad_norm": 3.6112840346969564,
+      "learning_rate": 4.760671281165777e-06,
+      "loss": 0.7025,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16632827093316008,
+      "grad_norm": 3.363415520275136,
+      "learning_rate": 4.760390421807445e-06,
+      "loss": 0.5791,
+      "step": 2048
+    },
+    {
+      "epoch": 0.16640948590920165,
+      "grad_norm": 5.52653404100446,
+      "learning_rate": 4.760109406042057e-06,
+      "loss": 0.545,
+      "step": 2049
+    },
+    {
+      "epoch": 0.16649070088524323,
+      "grad_norm": 6.546707591746984,
+      "learning_rate": 4.759828233889061e-06,
+      "loss": 0.5705,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1665719158612848,
+      "grad_norm": 5.661580374415145,
+      "learning_rate": 4.75954690536791e-06,
+      "loss": 0.5044,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1666531308373264,
+      "grad_norm": 4.883386412845315,
+      "learning_rate": 4.759265420498073e-06,
+      "loss": 0.5467,
+      "step": 2052
+    },
+    {
+      "epoch": 0.166734345813368,
+      "grad_norm": 9.411178068543851,
+      "learning_rate": 4.758983779299025e-06,
+      "loss": 0.5892,
+      "step": 2053
+    },
+    {
+      "epoch": 0.16681556078940957,
+      "grad_norm": 4.132413410970519,
+      "learning_rate": 4.758701981790255e-06,
+      "loss": 0.986,
+      "step": 2054
+    },
+    {
+      "epoch": 0.16689677576545114,
+      "grad_norm": 34.42736555216525,
+      "learning_rate": 4.7584200279912614e-06,
+      "loss": 0.6445,
+      "step": 2055
+    },
+    {
+      "epoch": 0.16697799074149272,
+      "grad_norm": 4.276245246172677,
+      "learning_rate": 4.7581379179215545e-06,
+      "loss": 0.7808,
+      "step": 2056
+    },
+    {
+      "epoch": 0.16705920571753433,
+      "grad_norm": 3.964870477401249,
+      "learning_rate": 4.757855651600656e-06,
+      "loss": 0.7379,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1671404206935759,
+      "grad_norm": 3.606188650388298,
+      "learning_rate": 4.757573229048095e-06,
+      "loss": 0.7927,
+      "step": 2058
+    },
+    {
+      "epoch": 0.16722163566961748,
+      "grad_norm": 4.909495914927784,
+      "learning_rate": 4.757290650283414e-06,
+      "loss": 0.5828,
+      "step": 2059
+    },
+    {
+      "epoch": 0.16730285064565906,
+      "grad_norm": 4.647983652975215,
+      "learning_rate": 4.757007915326167e-06,
+      "loss": 0.5334,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16738406562170063,
+      "grad_norm": 6.091135267313866,
+      "learning_rate": 4.756725024195918e-06,
+      "loss": 0.7908,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1674652805977422,
+      "grad_norm": 4.236635859135639,
+      "learning_rate": 4.75644197691224e-06,
+      "loss": 0.4387,
+      "step": 2062
+    },
+    {
+      "epoch": 0.16754649557378382,
+      "grad_norm": 5.254900085055447,
+      "learning_rate": 4.7561587734947195e-06,
+      "loss": 0.5377,
+      "step": 2063
+    },
+    {
+      "epoch": 0.1676277105498254,
+      "grad_norm": 6.08722014972973,
+      "learning_rate": 4.755875413962953e-06,
+      "loss": 0.4746,
+      "step": 2064
+    },
+    {
+      "epoch": 0.16770892552586697,
+      "grad_norm": 6.117784644667623,
+      "learning_rate": 4.7555918983365456e-06,
+      "loss": 0.4134,
+      "step": 2065
+    },
+    {
+      "epoch": 0.16779014050190855,
+      "grad_norm": 3.931238911985374,
+      "learning_rate": 4.755308226635117e-06,
+      "loss": 0.6176,
+      "step": 2066
+    },
+    {
+      "epoch": 0.16787135547795012,
+      "grad_norm": 5.993875492451563,
+      "learning_rate": 4.755024398878296e-06,
+      "loss": 0.5139,
+      "step": 2067
+    },
+    {
+      "epoch": 0.16795257045399173,
+      "grad_norm": 4.120681576998816,
+      "learning_rate": 4.75474041508572e-06,
+      "loss": 0.7734,
+      "step": 2068
+    },
+    {
+      "epoch": 0.1680337854300333,
+      "grad_norm": 4.2419734083458716,
+      "learning_rate": 4.7544562752770415e-06,
+      "loss": 0.6459,
+      "step": 2069
+    },
+    {
+      "epoch": 0.16811500040607488,
+      "grad_norm": 8.463809226446845,
+      "learning_rate": 4.75417197947192e-06,
+      "loss": 0.4706,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16819621538211646,
+      "grad_norm": 3.8343323335367128,
+      "learning_rate": 4.753887527690027e-06,
+      "loss": 0.4932,
+      "step": 2071
+    },
+    {
+      "epoch": 0.16827743035815804,
+      "grad_norm": 8.218415227607817,
+      "learning_rate": 4.753602919951046e-06,
+      "loss": 0.5316,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1683586453341996,
+      "grad_norm": 3.6050617955271216,
+      "learning_rate": 4.753318156274669e-06,
+      "loss": 0.5167,
+      "step": 2073
+    },
+    {
+      "epoch": 0.16843986031024122,
+      "grad_norm": 4.841750253634242,
+      "learning_rate": 4.753033236680602e-06,
+      "loss": 0.47,
+      "step": 2074
+    },
+    {
+      "epoch": 0.1685210752862828,
+      "grad_norm": 5.173184919410824,
+      "learning_rate": 4.75274816118856e-06,
+      "loss": 0.7099,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16860229026232437,
+      "grad_norm": 3.9701269933467622,
+      "learning_rate": 4.7524629298182655e-06,
+      "loss": 0.7438,
+      "step": 2076
+    },
+    {
+      "epoch": 0.16868350523836595,
+      "grad_norm": 5.896346024335442,
+      "learning_rate": 4.752177542589459e-06,
+      "loss": 0.5601,
+      "step": 2077
+    },
+    {
+      "epoch": 0.16876472021440753,
+      "grad_norm": 4.249527053571267,
+      "learning_rate": 4.7518919995218854e-06,
+      "loss": 0.6669,
+      "step": 2078
+    },
+    {
+      "epoch": 0.16884593519044913,
+      "grad_norm": 6.037758485869208,
+      "learning_rate": 4.7516063006353035e-06,
+      "loss": 0.5073,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1689271501664907,
+      "grad_norm": 3.7856446494128266,
+      "learning_rate": 4.7513204459494825e-06,
+      "loss": 0.5701,
+      "step": 2080
+    },
+    {
+      "epoch": 0.16900836514253229,
+      "grad_norm": 7.188036974920219,
+      "learning_rate": 4.751034435484201e-06,
+      "loss": 0.5066,
+      "step": 2081
+    },
+    {
+      "epoch": 0.16908958011857386,
+      "grad_norm": 4.317432007080439,
+      "learning_rate": 4.75074826925925e-06,
+      "loss": 0.7534,
+      "step": 2082
+    },
+    {
+      "epoch": 0.16917079509461544,
+      "grad_norm": 5.902119726898797,
+      "learning_rate": 4.750461947294431e-06,
+      "loss": 0.6171,
+      "step": 2083
+    },
+    {
+      "epoch": 0.16925201007065702,
+      "grad_norm": 3.7384712420136523,
+      "learning_rate": 4.750175469609555e-06,
+      "loss": 0.6519,
+      "step": 2084
+    },
+    {
+      "epoch": 0.16933322504669862,
+      "grad_norm": 3.71902625991903,
+      "learning_rate": 4.749888836224446e-06,
+      "loss": 0.6105,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1694144400227402,
+      "grad_norm": 4.570880922255994,
+      "learning_rate": 4.749602047158937e-06,
+      "loss": 0.8081,
+      "step": 2086
+    },
+    {
+      "epoch": 0.16949565499878178,
+      "grad_norm": 5.288276551549225,
+      "learning_rate": 4.749315102432872e-06,
+      "loss": 0.6383,
+      "step": 2087
+    },
+    {
+      "epoch": 0.16957686997482335,
+      "grad_norm": 11.77165556632558,
+      "learning_rate": 4.749028002066106e-06,
+      "loss": 0.5472,
+      "step": 2088
+    },
+    {
+      "epoch": 0.16965808495086493,
+      "grad_norm": 4.971856688210904,
+      "learning_rate": 4.748740746078505e-06,
+      "loss": 0.3578,
+      "step": 2089
+    },
+    {
+      "epoch": 0.16973929992690653,
+      "grad_norm": 3.215058873377653,
+      "learning_rate": 4.748453334489947e-06,
+      "loss": 0.6725,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1698205149029481,
+      "grad_norm": 5.921520125153694,
+      "learning_rate": 4.748165767320316e-06,
+      "loss": 0.604,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1699017298789897,
+      "grad_norm": 16.542194531090658,
+      "learning_rate": 4.747878044589513e-06,
+      "loss": 0.527,
+      "step": 2092
+    },
+    {
+      "epoch": 0.16998294485503126,
+      "grad_norm": 6.697998722104277,
+      "learning_rate": 4.747590166317447e-06,
+      "loss": 0.6809,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17006415983107284,
+      "grad_norm": 4.365123433682997,
+      "learning_rate": 4.7473021325240355e-06,
+      "loss": 0.6052,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17014537480711442,
+      "grad_norm": 4.341123013344583,
+      "learning_rate": 4.74701394322921e-06,
+      "loss": 0.7157,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17022658978315602,
+      "grad_norm": 7.815755600154752,
+      "learning_rate": 4.7467255984529124e-06,
+      "loss": 0.4511,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1703078047591976,
+      "grad_norm": 3.5475344534604276,
+      "learning_rate": 4.746437098215094e-06,
+      "loss": 0.5972,
+      "step": 2097
+    },
+    {
+      "epoch": 0.17038901973523918,
+      "grad_norm": 5.6422926548593635,
+      "learning_rate": 4.746148442535717e-06,
+      "loss": 0.6985,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17047023471128075,
+      "grad_norm": 3.7587908987149032,
+      "learning_rate": 4.745859631434757e-06,
+      "loss": 0.7605,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17055144968732233,
+      "grad_norm": 6.674895969578671,
+      "learning_rate": 4.745570664932195e-06,
+      "loss": 0.7068,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17063266466336394,
+      "grad_norm": 5.123301796152971,
+      "learning_rate": 4.745281543048027e-06,
+      "loss": 0.4672,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1707138796394055,
+      "grad_norm": 5.4550745050456655,
+      "learning_rate": 4.744992265802261e-06,
+      "loss": 0.4934,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1707950946154471,
+      "grad_norm": 5.148605782096688,
+      "learning_rate": 4.74470283321491e-06,
+      "loss": 0.5513,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17087630959148867,
+      "grad_norm": 4.960917833837893,
+      "learning_rate": 4.7444132453060046e-06,
+      "loss": 0.5779,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17095752456753024,
+      "grad_norm": 5.30615437756679,
+      "learning_rate": 4.744123502095579e-06,
+      "loss": 0.4572,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17103873954357182,
+      "grad_norm": 12.9603922902617,
+      "learning_rate": 4.743833603603685e-06,
+      "loss": 0.6257,
+      "step": 2106
+    },
+    {
+      "epoch": 0.17111995451961343,
+      "grad_norm": 4.802287043719077,
+      "learning_rate": 4.743543549850381e-06,
+      "loss": 0.695,
+      "step": 2107
+    },
+    {
+      "epoch": 0.171201169495655,
+      "grad_norm": 10.403228449617778,
+      "learning_rate": 4.743253340855737e-06,
+      "loss": 0.6116,
+      "step": 2108
+    },
+    {
+      "epoch": 0.17128238447169658,
+      "grad_norm": 6.756935813007018,
+      "learning_rate": 4.742962976639835e-06,
+      "loss": 0.7586,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17136359944773816,
+      "grad_norm": 10.07886788901266,
+      "learning_rate": 4.742672457222764e-06,
+      "loss": 0.4777,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17144481442377973,
+      "grad_norm": 5.62161561654736,
+      "learning_rate": 4.742381782624629e-06,
+      "loss": 0.4269,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17152602939982134,
+      "grad_norm": 5.949206043889717,
+      "learning_rate": 4.7420909528655416e-06,
+      "loss": 0.7793,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17160724437586292,
+      "grad_norm": 4.263095903085148,
+      "learning_rate": 4.741799967965627e-06,
+      "loss": 0.5637,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1716884593519045,
+      "grad_norm": 10.633709854598653,
+      "learning_rate": 4.74150882794502e-06,
+      "loss": 0.82,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17176967432794607,
+      "grad_norm": 3.032916783451706,
+      "learning_rate": 4.741217532823864e-06,
+      "loss": 0.72,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17185088930398765,
+      "grad_norm": 3.8608167614052484,
+      "learning_rate": 4.740926082622316e-06,
+      "loss": 0.7894,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17193210428002922,
+      "grad_norm": 3.9765186741541436,
+      "learning_rate": 4.740634477360544e-06,
+      "loss": 0.5949,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17201331925607083,
+      "grad_norm": 3.6048556532254246,
+      "learning_rate": 4.740342717058723e-06,
+      "loss": 0.6829,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1720945342321124,
+      "grad_norm": 3.612675372859013,
+      "learning_rate": 4.740050801737045e-06,
+      "loss": 0.4803,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17217574920815398,
+      "grad_norm": 4.645131781522265,
+      "learning_rate": 4.739758731415705e-06,
+      "loss": 0.7209,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17225696418419556,
+      "grad_norm": 5.631469115723518,
+      "learning_rate": 4.739466506114916e-06,
+      "loss": 0.6008,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17233817916023714,
+      "grad_norm": 4.1088983111658735,
+      "learning_rate": 4.739174125854896e-06,
+      "loss": 0.6917,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17241939413627874,
+      "grad_norm": 5.169477468086094,
+      "learning_rate": 4.738881590655877e-06,
+      "loss": 0.5403,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17250060911232032,
+      "grad_norm": 4.946107948001271,
+      "learning_rate": 4.738588900538102e-06,
+      "loss": 0.608,
+      "step": 2124
+    },
+    {
+      "epoch": 0.1725818240883619,
+      "grad_norm": 3.559651178268444,
+      "learning_rate": 4.738296055521821e-06,
+      "loss": 0.6463,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17266303906440347,
+      "grad_norm": 6.129364916078834,
+      "learning_rate": 4.738003055627301e-06,
+      "loss": 0.4888,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17274425404044505,
+      "grad_norm": 4.278147231585998,
+      "learning_rate": 4.7377099008748125e-06,
+      "loss": 0.8504,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17282546901648663,
+      "grad_norm": 8.024624331958586,
+      "learning_rate": 4.737416591284643e-06,
+      "loss": 0.6727,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17290668399252823,
+      "grad_norm": 29.45093767091559,
+      "learning_rate": 4.737123126877086e-06,
+      "loss": 0.7422,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1729878989685698,
+      "grad_norm": 3.13136259957198,
+      "learning_rate": 4.736829507672449e-06,
+      "loss": 0.5307,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17306911394461139,
+      "grad_norm": 7.567136439537917,
+      "learning_rate": 4.736535733691048e-06,
+      "loss": 0.6303,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17315032892065296,
+      "grad_norm": 5.220061316823007,
+      "learning_rate": 4.73624180495321e-06,
+      "loss": 0.5315,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17323154389669454,
+      "grad_norm": 7.195772361468456,
+      "learning_rate": 4.7359477214792754e-06,
+      "loss": 0.5764,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17331275887273614,
+      "grad_norm": 7.3854852778069615,
+      "learning_rate": 4.735653483289591e-06,
+      "loss": 0.5438,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17339397384877772,
+      "grad_norm": 6.686734429776475,
+      "learning_rate": 4.7353590904045184e-06,
+      "loss": 0.6625,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1734751888248193,
+      "grad_norm": 3.9164463638901514,
+      "learning_rate": 4.735064542844428e-06,
+      "loss": 0.5811,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17355640380086088,
+      "grad_norm": 4.957310787853842,
+      "learning_rate": 4.734769840629699e-06,
+      "loss": 0.4821,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17363761877690245,
+      "grad_norm": 8.419842894201476,
+      "learning_rate": 4.734474983780724e-06,
+      "loss": 0.5227,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17371883375294403,
+      "grad_norm": 8.041916137901524,
+      "learning_rate": 4.734179972317907e-06,
+      "loss": 0.6015,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17380004872898563,
+      "grad_norm": 4.1188752273792195,
+      "learning_rate": 4.73388480626166e-06,
+      "loss": 0.6766,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1738812637050272,
+      "grad_norm": 5.065614073087699,
+      "learning_rate": 4.733589485632407e-06,
+      "loss": 0.5939,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1739624786810688,
+      "grad_norm": 3.8622027019274077,
+      "learning_rate": 4.733294010450583e-06,
+      "loss": 0.5127,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17404369365711037,
+      "grad_norm": 5.31622281327025,
+      "learning_rate": 4.732998380736632e-06,
+      "loss": 0.6016,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17412490863315194,
+      "grad_norm": 3.832733531671153,
+      "learning_rate": 4.732702596511012e-06,
+      "loss": 0.4538,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17420612360919355,
+      "grad_norm": 4.720524116716847,
+      "learning_rate": 4.732406657794188e-06,
+      "loss": 0.5959,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17428733858523512,
+      "grad_norm": 3.3574370172144734,
+      "learning_rate": 4.732110564606639e-06,
+      "loss": 0.6259,
+      "step": 2146
+    },
+    {
+      "epoch": 0.1743685535612767,
+      "grad_norm": 3.845949435304935,
+      "learning_rate": 4.7318143169688515e-06,
+      "loss": 0.5966,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17444976853731828,
+      "grad_norm": 3.7946357565371973,
+      "learning_rate": 4.731517914901324e-06,
+      "loss": 0.6552,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17453098351335986,
+      "grad_norm": 8.89191739354207,
+      "learning_rate": 4.731221358424569e-06,
+      "loss": 0.7327,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17461219848940143,
+      "grad_norm": 6.740032168112452,
+      "learning_rate": 4.730924647559103e-06,
+      "loss": 0.6989,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17469341346544304,
+      "grad_norm": 4.6320616184970875,
+      "learning_rate": 4.730627782325459e-06,
+      "loss": 0.5567,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17477462844148461,
+      "grad_norm": 4.627022975502933,
+      "learning_rate": 4.730330762744178e-06,
+      "loss": 0.5177,
+      "step": 2152
+    },
+    {
+      "epoch": 0.1748558434175262,
+      "grad_norm": 5.2547160244571876,
+      "learning_rate": 4.730033588835812e-06,
+      "loss": 0.578,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17493705839356777,
+      "grad_norm": 4.545741644009064,
+      "learning_rate": 4.729736260620924e-06,
+      "loss": 0.5025,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17501827336960935,
+      "grad_norm": 3.1745271529048074,
+      "learning_rate": 4.729438778120088e-06,
+      "loss": 0.6065,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17509948834565095,
+      "grad_norm": 3.614800398396882,
+      "learning_rate": 4.729141141353887e-06,
+      "loss": 0.5726,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17518070332169253,
+      "grad_norm": 4.093940125645144,
+      "learning_rate": 4.7288433503429165e-06,
+      "loss": 0.7513,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1752619182977341,
+      "grad_norm": 6.4087498177915885,
+      "learning_rate": 4.728545405107782e-06,
+      "loss": 0.5511,
+      "step": 2158
+    },
+    {
+      "epoch": 0.17534313327377568,
+      "grad_norm": 6.964419687177815,
+      "learning_rate": 4.7282473056691e-06,
+      "loss": 0.5388,
+      "step": 2159
+    },
+    {
+      "epoch": 0.17542434824981726,
+      "grad_norm": 7.381195247577337,
+      "learning_rate": 4.727949052047498e-06,
+      "loss": 0.5553,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17550556322585883,
+      "grad_norm": 3.447447501574935,
+      "learning_rate": 4.7276506442636125e-06,
+      "loss": 0.6407,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17558677820190044,
+      "grad_norm": 3.624265660451841,
+      "learning_rate": 4.727352082338092e-06,
+      "loss": 0.5699,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17566799317794202,
+      "grad_norm": 4.736948616378692,
+      "learning_rate": 4.727053366291595e-06,
+      "loss": 0.5561,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1757492081539836,
+      "grad_norm": 5.557189925086155,
+      "learning_rate": 4.726754496144792e-06,
+      "loss": 0.6388,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17583042313002517,
+      "grad_norm": 5.654257813296,
+      "learning_rate": 4.726455471918363e-06,
+      "loss": 0.5798,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17591163810606675,
+      "grad_norm": 4.017186015236763,
+      "learning_rate": 4.726156293632998e-06,
+      "loss": 0.6494,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17599285308210835,
+      "grad_norm": 3.2976783263541036,
+      "learning_rate": 4.725856961309401e-06,
+      "loss": 0.6894,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17607406805814993,
+      "grad_norm": 9.505301533813855,
+      "learning_rate": 4.725557474968281e-06,
+      "loss": 0.5775,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1761552830341915,
+      "grad_norm": 5.237101724122944,
+      "learning_rate": 4.725257834630362e-06,
+      "loss": 0.5132,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17623649801023308,
+      "grad_norm": 4.666530981374312,
+      "learning_rate": 4.7249580403163786e-06,
+      "loss": 0.4708,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17631771298627466,
+      "grad_norm": 3.82098458375026,
+      "learning_rate": 4.7246580920470746e-06,
+      "loss": 0.6887,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17639892796231624,
+      "grad_norm": 4.794596840480159,
+      "learning_rate": 4.7243579898432035e-06,
+      "loss": 0.5326,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17648014293835784,
+      "grad_norm": 5.163380467947997,
+      "learning_rate": 4.724057733725532e-06,
+      "loss": 0.4342,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17656135791439942,
+      "grad_norm": 6.130524877364227,
+      "learning_rate": 4.723757323714836e-06,
+      "loss": 0.5504,
+      "step": 2174
+    },
+    {
+      "epoch": 0.176642572890441,
+      "grad_norm": 3.6388040943311566,
+      "learning_rate": 4.723456759831903e-06,
+      "loss": 0.519,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17672378786648257,
+      "grad_norm": 5.217405099472009,
+      "learning_rate": 4.7231560420975294e-06,
+      "loss": 0.4867,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17680500284252415,
+      "grad_norm": 8.145512643236385,
+      "learning_rate": 4.722855170532523e-06,
+      "loss": 0.532,
+      "step": 2177
+    },
+    {
+      "epoch": 0.17688621781856576,
+      "grad_norm": 6.513554956269954,
+      "learning_rate": 4.7225541451577035e-06,
+      "loss": 0.7263,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17696743279460733,
+      "grad_norm": 5.647381362856042,
+      "learning_rate": 4.7222529659939e-06,
+      "loss": 0.5248,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1770486477706489,
+      "grad_norm": 8.634842181250088,
+      "learning_rate": 4.721951633061952e-06,
+      "loss": 0.5163,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1771298627466905,
+      "grad_norm": 5.804232556564542,
+      "learning_rate": 4.721650146382711e-06,
+      "loss": 0.5169,
+      "step": 2181
+    },
+    {
+      "epoch": 0.17721107772273206,
+      "grad_norm": 5.170134284377379,
+      "learning_rate": 4.721348505977037e-06,
+      "loss": 0.569,
+      "step": 2182
+    },
+    {
+      "epoch": 0.17729229269877364,
+      "grad_norm": 6.347534050197857,
+      "learning_rate": 4.721046711865803e-06,
+      "loss": 0.7412,
+      "step": 2183
+    },
+    {
+      "epoch": 0.17737350767481525,
+      "grad_norm": 3.947255210170048,
+      "learning_rate": 4.720744764069892e-06,
+      "loss": 0.6272,
+      "step": 2184
+    },
+    {
+      "epoch": 0.17745472265085682,
+      "grad_norm": 8.963510036415032,
+      "learning_rate": 4.7204426626101955e-06,
+      "loss": 0.5039,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1775359376268984,
+      "grad_norm": 5.0045242805343255,
+      "learning_rate": 4.720140407507619e-06,
+      "loss": 0.5961,
+      "step": 2186
+    },
+    {
+      "epoch": 0.17761715260293998,
+      "grad_norm": 80.66410418515012,
+      "learning_rate": 4.719837998783075e-06,
+      "loss": 0.6292,
+      "step": 2187
+    },
+    {
+      "epoch": 0.17769836757898155,
+      "grad_norm": 3.8541304256097804,
+      "learning_rate": 4.7195354364574915e-06,
+      "loss": 0.5465,
+      "step": 2188
+    },
+    {
+      "epoch": 0.17777958255502316,
+      "grad_norm": 3.524866314545861,
+      "learning_rate": 4.719232720551802e-06,
+      "loss": 0.6409,
+      "step": 2189
+    },
+    {
+      "epoch": 0.17786079753106473,
+      "grad_norm": 3.929607398758829,
+      "learning_rate": 4.718929851086953e-06,
+      "loss": 0.5239,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1779420125071063,
+      "grad_norm": 6.4534562177957575,
+      "learning_rate": 4.718626828083902e-06,
+      "loss": 0.515,
+      "step": 2191
+    },
+    {
+      "epoch": 0.1780232274831479,
+      "grad_norm": 4.416334886882551,
+      "learning_rate": 4.718323651563616e-06,
+      "loss": 0.4572,
+      "step": 2192
+    },
+    {
+      "epoch": 0.17810444245918947,
+      "grad_norm": 4.2150922659750085,
+      "learning_rate": 4.718020321547075e-06,
+      "loss": 0.6128,
+      "step": 2193
+    },
+    {
+      "epoch": 0.17818565743523104,
+      "grad_norm": 3.3177339549952705,
+      "learning_rate": 4.717716838055265e-06,
+      "loss": 0.642,
+      "step": 2194
+    },
+    {
+      "epoch": 0.17826687241127265,
+      "grad_norm": 4.248446599343683,
+      "learning_rate": 4.717413201109187e-06,
+      "loss": 0.8113,
+      "step": 2195
+    },
+    {
+      "epoch": 0.17834808738731422,
+      "grad_norm": 4.561435467958204,
+      "learning_rate": 4.717109410729851e-06,
+      "loss": 0.6368,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1784293023633558,
+      "grad_norm": 5.104529981932405,
+      "learning_rate": 4.716805466938278e-06,
+      "loss": 0.6438,
+      "step": 2197
+    },
+    {
+      "epoch": 0.17851051733939738,
+      "grad_norm": 3.5323774996376374,
+      "learning_rate": 4.7165013697555e-06,
+      "loss": 0.6165,
+      "step": 2198
+    },
+    {
+      "epoch": 0.17859173231543896,
+      "grad_norm": 6.640997619244458,
+      "learning_rate": 4.716197119202556e-06,
+      "loss": 0.6347,
+      "step": 2199
+    },
+    {
+      "epoch": 0.17867294729148056,
+      "grad_norm": 11.632083395111554,
+      "learning_rate": 4.715892715300501e-06,
+      "loss": 0.7224,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17875416226752214,
+      "grad_norm": 4.224735988915179,
+      "learning_rate": 4.7155881580703984e-06,
+      "loss": 0.5769,
+      "step": 2201
+    },
+    {
+      "epoch": 0.17883537724356371,
+      "grad_norm": 7.9851742695130685,
+      "learning_rate": 4.71528344753332e-06,
+      "loss": 0.4985,
+      "step": 2202
+    },
+    {
+      "epoch": 0.1789165922196053,
+      "grad_norm": 6.25379024192091,
+      "learning_rate": 4.714978583710352e-06,
+      "loss": 0.5391,
+      "step": 2203
+    },
+    {
+      "epoch": 0.17899780719564687,
+      "grad_norm": 3.486820916489094,
+      "learning_rate": 4.714673566622589e-06,
+      "loss": 0.567,
+      "step": 2204
+    },
+    {
+      "epoch": 0.17907902217168845,
+      "grad_norm": 3.4864795044696617,
+      "learning_rate": 4.714368396291135e-06,
+      "loss": 0.7785,
+      "step": 2205
+    },
+    {
+      "epoch": 0.17916023714773005,
+      "grad_norm": 3.6141664132762172,
+      "learning_rate": 4.714063072737108e-06,
+      "loss": 0.4835,
+      "step": 2206
+    },
+    {
+      "epoch": 0.17924145212377163,
+      "grad_norm": 5.839740334780792,
+      "learning_rate": 4.713757595981634e-06,
+      "loss": 0.5031,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1793226670998132,
+      "grad_norm": 8.510815914171106,
+      "learning_rate": 4.713451966045851e-06,
+      "loss": 0.6997,
+      "step": 2208
+    },
+    {
+      "epoch": 0.17940388207585478,
+      "grad_norm": 4.803667331216526,
+      "learning_rate": 4.713146182950905e-06,
+      "loss": 0.5733,
+      "step": 2209
+    },
+    {
+      "epoch": 0.17948509705189636,
+      "grad_norm": 3.5388152829208015,
+      "learning_rate": 4.7128402467179575e-06,
+      "loss": 0.5907,
+      "step": 2210
+    },
+    {
+      "epoch": 0.17956631202793796,
+      "grad_norm": 3.03616427792538,
+      "learning_rate": 4.712534157368176e-06,
+      "loss": 0.5774,
+      "step": 2211
+    },
+    {
+      "epoch": 0.17964752700397954,
+      "grad_norm": 9.760631640572191,
+      "learning_rate": 4.7122279149227405e-06,
+      "loss": 0.5262,
+      "step": 2212
+    },
+    {
+      "epoch": 0.17972874198002112,
+      "grad_norm": 4.386181821659895,
+      "learning_rate": 4.711921519402841e-06,
+      "loss": 0.6952,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1798099569560627,
+      "grad_norm": 3.3351010256653573,
+      "learning_rate": 4.711614970829679e-06,
+      "loss": 0.5091,
+      "step": 2214
+    },
+    {
+      "epoch": 0.17989117193210427,
+      "grad_norm": 6.9608846044796024,
+      "learning_rate": 4.711308269224466e-06,
+      "loss": 0.4963,
+      "step": 2215
+    },
+    {
+      "epoch": 0.17997238690814585,
+      "grad_norm": 5.215289754630284,
+      "learning_rate": 4.7110014146084235e-06,
+      "loss": 0.5736,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18005360188418745,
+      "grad_norm": 11.93663447822261,
+      "learning_rate": 4.710694407002785e-06,
+      "loss": 0.7406,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18013481686022903,
+      "grad_norm": 5.163645505705543,
+      "learning_rate": 4.710387246428794e-06,
+      "loss": 0.6448,
+      "step": 2218
+    },
+    {
+      "epoch": 0.1802160318362706,
+      "grad_norm": 4.661698490114284,
+      "learning_rate": 4.710079932907703e-06,
+      "loss": 0.7224,
+      "step": 2219
+    },
+    {
+      "epoch": 0.18029724681231218,
+      "grad_norm": 3.401258624128954,
+      "learning_rate": 4.7097724664607775e-06,
+      "loss": 0.5251,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18037846178835376,
+      "grad_norm": 5.125047047447445,
+      "learning_rate": 4.709464847109292e-06,
+      "loss": 0.6384,
+      "step": 2221
+    },
+    {
+      "epoch": 0.18045967676439537,
+      "grad_norm": 4.023853533344583,
+      "learning_rate": 4.709157074874533e-06,
+      "loss": 0.5047,
+      "step": 2222
+    },
+    {
+      "epoch": 0.18054089174043694,
+      "grad_norm": 6.949692493391,
+      "learning_rate": 4.7088491497777965e-06,
+      "loss": 0.5204,
+      "step": 2223
+    },
+    {
+      "epoch": 0.18062210671647852,
+      "grad_norm": 4.10788473501353,
+      "learning_rate": 4.708541071840388e-06,
+      "loss": 0.5865,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1807033216925201,
+      "grad_norm": 4.694036134420709,
+      "learning_rate": 4.708232841083628e-06,
+      "loss": 0.5692,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18078453666856167,
+      "grad_norm": 5.994257706825592,
+      "learning_rate": 4.70792445752884e-06,
+      "loss": 0.5018,
+      "step": 2226
+    },
+    {
+      "epoch": 0.18086575164460328,
+      "grad_norm": 8.767129488205637,
+      "learning_rate": 4.707615921197366e-06,
+      "loss": 0.5381,
+      "step": 2227
+    },
+    {
+      "epoch": 0.18094696662064486,
+      "grad_norm": 4.847865205503708,
+      "learning_rate": 4.707307232110554e-06,
+      "loss": 0.7474,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18102818159668643,
+      "grad_norm": 5.668800604259548,
+      "learning_rate": 4.706998390289763e-06,
+      "loss": 0.6027,
+      "step": 2229
+    },
+    {
+      "epoch": 0.181109396572728,
+      "grad_norm": 6.745634612012881,
+      "learning_rate": 4.706689395756363e-06,
+      "loss": 0.5229,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1811906115487696,
+      "grad_norm": 4.895045129217889,
+      "learning_rate": 4.706380248531737e-06,
+      "loss": 0.5253,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18127182652481116,
+      "grad_norm": 4.810666628926101,
+      "learning_rate": 4.706070948637274e-06,
+      "loss": 0.6069,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18135304150085277,
+      "grad_norm": 7.912604183646548,
+      "learning_rate": 4.705761496094377e-06,
+      "loss": 0.5576,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18143425647689435,
+      "grad_norm": 4.074525434845986,
+      "learning_rate": 4.705451890924459e-06,
+      "loss": 0.6926,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18151547145293592,
+      "grad_norm": 9.471081474256104,
+      "learning_rate": 4.705142133148943e-06,
+      "loss": 0.5167,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1815966864289775,
+      "grad_norm": 4.8687622437884555,
+      "learning_rate": 4.70483222278926e-06,
+      "loss": 0.6418,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18167790140501908,
+      "grad_norm": 4.3746758828868995,
+      "learning_rate": 4.704522159866857e-06,
+      "loss": 0.5135,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18175911638106068,
+      "grad_norm": 4.148684813653182,
+      "learning_rate": 4.704211944403188e-06,
+      "loss": 0.6289,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18184033135710226,
+      "grad_norm": 4.101220246406701,
+      "learning_rate": 4.703901576419717e-06,
+      "loss": 0.473,
+      "step": 2239
+    },
+    {
+      "epoch": 0.18192154633314384,
+      "grad_norm": 4.884276965609116,
+      "learning_rate": 4.703591055937922e-06,
+      "loss": 0.5761,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1820027613091854,
+      "grad_norm": 9.506027279588086,
+      "learning_rate": 4.7032803829792875e-06,
+      "loss": 0.5745,
+      "step": 2241
+    },
+    {
+      "epoch": 0.182083976285227,
+      "grad_norm": 4.5640936040307825,
+      "learning_rate": 4.702969557565312e-06,
+      "loss": 0.6804,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18216519126126857,
+      "grad_norm": 5.16553253038635,
+      "learning_rate": 4.702658579717502e-06,
+      "loss": 0.4686,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18224640623731017,
+      "grad_norm": 5.1964649031938395,
+      "learning_rate": 4.702347449457375e-06,
+      "loss": 0.5184,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18232762121335175,
+      "grad_norm": 8.556515016420585,
+      "learning_rate": 4.702036166806461e-06,
+      "loss": 0.3698,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18240883618939333,
+      "grad_norm": 7.054847431413247,
+      "learning_rate": 4.7017247317862976e-06,
+      "loss": 0.5364,
+      "step": 2246
+    },
+    {
+      "epoch": 0.1824900511654349,
+      "grad_norm": 5.019891726347235,
+      "learning_rate": 4.701413144418437e-06,
+      "loss": 0.4901,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18257126614147648,
+      "grad_norm": 4.195016853555098,
+      "learning_rate": 4.701101404724435e-06,
+      "loss": 0.4601,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18265248111751808,
+      "grad_norm": 5.378788487799258,
+      "learning_rate": 4.700789512725867e-06,
+      "loss": 0.6267,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18273369609355966,
+      "grad_norm": 3.793906030224367,
+      "learning_rate": 4.700477468444311e-06,
+      "loss": 0.6438,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18281491106960124,
+      "grad_norm": 4.868151253449179,
+      "learning_rate": 4.700165271901361e-06,
+      "loss": 0.6933,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18289612604564282,
+      "grad_norm": 7.777361957555963,
+      "learning_rate": 4.699852923118618e-06,
+      "loss": 0.4894,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1829773410216844,
+      "grad_norm": 4.106355721520342,
+      "learning_rate": 4.699540422117695e-06,
+      "loss": 0.4232,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18305855599772597,
+      "grad_norm": 5.430574051132866,
+      "learning_rate": 4.699227768920216e-06,
+      "loss": 0.5387,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18313977097376757,
+      "grad_norm": 6.575598696133923,
+      "learning_rate": 4.6989149635478145e-06,
+      "loss": 0.5371,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18322098594980915,
+      "grad_norm": 5.4676777555816765,
+      "learning_rate": 4.698602006022136e-06,
+      "loss": 0.5012,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18330220092585073,
+      "grad_norm": 4.576711281154184,
+      "learning_rate": 4.698288896364834e-06,
+      "loss": 0.6324,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1833834159018923,
+      "grad_norm": 11.976819517596883,
+      "learning_rate": 4.697975634597574e-06,
+      "loss": 0.5405,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18346463087793388,
+      "grad_norm": 6.301644428134142,
+      "learning_rate": 4.697662220742033e-06,
+      "loss": 0.5355,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1835458458539755,
+      "grad_norm": 16.567162364593425,
+      "learning_rate": 4.697348654819898e-06,
+      "loss": 0.5812,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18362706083001706,
+      "grad_norm": 5.845436104952721,
+      "learning_rate": 4.697034936852865e-06,
+      "loss": 0.6177,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18370827580605864,
+      "grad_norm": 3.4478757836782608,
+      "learning_rate": 4.6967210668626415e-06,
+      "loss": 0.6487,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18378949078210022,
+      "grad_norm": 8.26455582424995,
+      "learning_rate": 4.696407044870947e-06,
+      "loss": 0.6865,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1838707057581418,
+      "grad_norm": 3.6489257014011924,
+      "learning_rate": 4.696092870899509e-06,
+      "loss": 0.4881,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18395192073418337,
+      "grad_norm": 7.476743860852032,
+      "learning_rate": 4.695778544970066e-06,
+      "loss": 0.5365,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18403313571022498,
+      "grad_norm": 5.335228274279977,
+      "learning_rate": 4.695464067104371e-06,
+      "loss": 0.5978,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18411435068626655,
+      "grad_norm": 7.0816544622030175,
+      "learning_rate": 4.6951494373241805e-06,
+      "loss": 0.769,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18419556566230813,
+      "grad_norm": 3.851216245811756,
+      "learning_rate": 4.694834655651266e-06,
+      "loss": 0.627,
+      "step": 2268
+    },
+    {
+      "epoch": 0.1842767806383497,
+      "grad_norm": 4.805289628209942,
+      "learning_rate": 4.6945197221074104e-06,
+      "loss": 0.8659,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18435799561439128,
+      "grad_norm": 7.794985897939018,
+      "learning_rate": 4.694204636714403e-06,
+      "loss": 0.5485,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1844392105904329,
+      "grad_norm": 20.84411786544806,
+      "learning_rate": 4.693889399494049e-06,
+      "loss": 0.5583,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18452042556647447,
+      "grad_norm": 4.8093342027022326,
+      "learning_rate": 4.693574010468159e-06,
+      "loss": 0.7422,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18460164054251604,
+      "grad_norm": 4.897421838824223,
+      "learning_rate": 4.693258469658557e-06,
+      "loss": 0.6693,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18468285551855762,
+      "grad_norm": 5.418227314133965,
+      "learning_rate": 4.692942777087076e-06,
+      "loss": 0.5361,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1847640704945992,
+      "grad_norm": 4.1871350066526745,
+      "learning_rate": 4.692626932775561e-06,
+      "loss": 0.6755,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18484528547064077,
+      "grad_norm": 10.739798887484266,
+      "learning_rate": 4.6923109367458665e-06,
+      "loss": 0.7373,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18492650044668238,
+      "grad_norm": 4.134435714566178,
+      "learning_rate": 4.6919947890198585e-06,
+      "loss": 0.5737,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18500771542272396,
+      "grad_norm": 4.222974710982374,
+      "learning_rate": 4.691678489619411e-06,
+      "loss": 0.7376,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18508893039876553,
+      "grad_norm": 4.6176435172773065,
+      "learning_rate": 4.691362038566411e-06,
+      "loss": 0.6068,
+      "step": 2279
+    },
+    {
+      "epoch": 0.1851701453748071,
+      "grad_norm": 3.8516607452654577,
+      "learning_rate": 4.691045435882758e-06,
+      "loss": 0.5776,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1852513603508487,
+      "grad_norm": 4.380266220174846,
+      "learning_rate": 4.690728681590354e-06,
+      "loss": 0.4589,
+      "step": 2281
+    },
+    {
+      "epoch": 0.1853325753268903,
+      "grad_norm": 20.10704260314573,
+      "learning_rate": 4.6904117757111215e-06,
+      "loss": 0.4806,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18541379030293187,
+      "grad_norm": 4.622570996308856,
+      "learning_rate": 4.6900947182669855e-06,
+      "loss": 0.5437,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18549500527897345,
+      "grad_norm": 3.4999814335004125,
+      "learning_rate": 4.689777509279886e-06,
+      "loss": 0.6253,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18557622025501502,
+      "grad_norm": 8.159646406598347,
+      "learning_rate": 4.689460148771773e-06,
+      "loss": 0.6941,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1856574352310566,
+      "grad_norm": 4.119601877397449,
+      "learning_rate": 4.6891426367646046e-06,
+      "loss": 0.5396,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18573865020709818,
+      "grad_norm": 5.783579437576225,
+      "learning_rate": 4.6888249732803516e-06,
+      "loss": 0.3815,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18581986518313978,
+      "grad_norm": 4.479086903423486,
+      "learning_rate": 4.688507158340994e-06,
+      "loss": 0.5961,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18590108015918136,
+      "grad_norm": 5.521280967881707,
+      "learning_rate": 4.688189191968524e-06,
+      "loss": 0.5213,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18598229513522294,
+      "grad_norm": 5.560685189930327,
+      "learning_rate": 4.687871074184944e-06,
+      "loss": 0.5069,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1860635101112645,
+      "grad_norm": 8.713540857848809,
+      "learning_rate": 4.687552805012263e-06,
+      "loss": 0.4698,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1861447250873061,
+      "grad_norm": 7.120678804920857,
+      "learning_rate": 4.687234384472506e-06,
+      "loss": 0.5941,
+      "step": 2292
+    },
+    {
+      "epoch": 0.1862259400633477,
+      "grad_norm": 3.7747766229799105,
+      "learning_rate": 4.686915812587706e-06,
+      "loss": 0.532,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18630715503938927,
+      "grad_norm": 5.78663404222734,
+      "learning_rate": 4.686597089379905e-06,
+      "loss": 0.6332,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18638837001543085,
+      "grad_norm": 5.344791467556041,
+      "learning_rate": 4.6862782148711584e-06,
+      "loss": 0.551,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18646958499147243,
+      "grad_norm": 3.2548480768417183,
+      "learning_rate": 4.685959189083531e-06,
+      "loss": 0.7081,
+      "step": 2296
+    },
+    {
+      "epoch": 0.186550799967514,
+      "grad_norm": 4.522340647304046,
+      "learning_rate": 4.685640012039095e-06,
+      "loss": 0.6577,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18663201494355558,
+      "grad_norm": 4.317215335517002,
+      "learning_rate": 4.685320683759939e-06,
+      "loss": 0.5544,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18671322991959718,
+      "grad_norm": 5.363458185770196,
+      "learning_rate": 4.685001204268156e-06,
+      "loss": 0.4849,
+      "step": 2299
+    },
+    {
+      "epoch": 0.18679444489563876,
+      "grad_norm": 25.593738509492916,
+      "learning_rate": 4.684681573585854e-06,
+      "loss": 0.5735,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18687565987168034,
+      "grad_norm": 6.172329696323195,
+      "learning_rate": 4.684361791735149e-06,
+      "loss": 0.5465,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18695687484772192,
+      "grad_norm": 4.8669050067958874,
+      "learning_rate": 4.684041858738169e-06,
+      "loss": 0.6078,
+      "step": 2302
+    },
+    {
+      "epoch": 0.1870380898237635,
+      "grad_norm": 12.729610338065322,
+      "learning_rate": 4.683721774617052e-06,
+      "loss": 0.5647,
+      "step": 2303
+    },
+    {
+      "epoch": 0.1871193047998051,
+      "grad_norm": 5.8134799742582475,
+      "learning_rate": 4.6834015393939445e-06,
+      "loss": 0.5026,
+      "step": 2304
+    },
+    {
+      "epoch": 0.18720051977584667,
+      "grad_norm": 3.5817996920030257,
+      "learning_rate": 4.683081153091006e-06,
+      "loss": 0.4811,
+      "step": 2305
+    },
+    {
+      "epoch": 0.18728173475188825,
+      "grad_norm": 4.114121320477102,
+      "learning_rate": 4.682760615730405e-06,
+      "loss": 0.6439,
+      "step": 2306
+    },
+    {
+      "epoch": 0.18736294972792983,
+      "grad_norm": 7.202440009743923,
+      "learning_rate": 4.682439927334323e-06,
+      "loss": 0.6019,
+      "step": 2307
+    },
+    {
+      "epoch": 0.1874441647039714,
+      "grad_norm": 5.323112544773871,
+      "learning_rate": 4.682119087924948e-06,
+      "loss": 0.7116,
+      "step": 2308
+    },
+    {
+      "epoch": 0.18752537968001298,
+      "grad_norm": 3.6253125361558967,
+      "learning_rate": 4.681798097524479e-06,
+      "loss": 0.6427,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1876065946560546,
+      "grad_norm": 4.355698784301835,
+      "learning_rate": 4.681476956155131e-06,
+      "loss": 0.4505,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18768780963209616,
+      "grad_norm": 3.395486667581571,
+      "learning_rate": 4.681155663839122e-06,
+      "loss": 0.5654,
+      "step": 2311
+    },
+    {
+      "epoch": 0.18776902460813774,
+      "grad_norm": 5.363908094574731,
+      "learning_rate": 4.680834220598685e-06,
+      "loss": 0.5545,
+      "step": 2312
+    },
+    {
+      "epoch": 0.18785023958417932,
+      "grad_norm": 4.304796229403271,
+      "learning_rate": 4.6805126264560605e-06,
+      "loss": 0.6393,
+      "step": 2313
+    },
+    {
+      "epoch": 0.1879314545602209,
+      "grad_norm": 4.291548258367333,
+      "learning_rate": 4.680190881433504e-06,
+      "loss": 0.6884,
+      "step": 2314
+    },
+    {
+      "epoch": 0.1880126695362625,
+      "grad_norm": 9.784318271093905,
+      "learning_rate": 4.679868985553276e-06,
+      "loss": 0.5777,
+      "step": 2315
+    },
+    {
+      "epoch": 0.18809388451230408,
+      "grad_norm": 7.761537889560317,
+      "learning_rate": 4.6795469388376525e-06,
+      "loss": 0.47,
+      "step": 2316
+    },
+    {
+      "epoch": 0.18817509948834565,
+      "grad_norm": 6.995156273431482,
+      "learning_rate": 4.6792247413089145e-06,
+      "loss": 0.6137,
+      "step": 2317
+    },
+    {
+      "epoch": 0.18825631446438723,
+      "grad_norm": 4.313226522351422,
+      "learning_rate": 4.678902392989359e-06,
+      "loss": 0.6229,
+      "step": 2318
+    },
+    {
+      "epoch": 0.1883375294404288,
+      "grad_norm": 5.081068946857541,
+      "learning_rate": 4.678579893901288e-06,
+      "loss": 0.5712,
+      "step": 2319
+    },
+    {
+      "epoch": 0.18841874441647039,
+      "grad_norm": 4.809789843673943,
+      "learning_rate": 4.678257244067019e-06,
+      "loss": 0.5531,
+      "step": 2320
+    },
+    {
+      "epoch": 0.188499959392512,
+      "grad_norm": 3.8452865625471255,
+      "learning_rate": 4.677934443508877e-06,
+      "loss": 0.5659,
+      "step": 2321
+    },
+    {
+      "epoch": 0.18858117436855357,
+      "grad_norm": 4.917273325549435,
+      "learning_rate": 4.6776114922491985e-06,
+      "loss": 0.5518,
+      "step": 2322
+    },
+    {
+      "epoch": 0.18866238934459514,
+      "grad_norm": 4.459277191818069,
+      "learning_rate": 4.67728839031033e-06,
+      "loss": 0.5119,
+      "step": 2323
+    },
+    {
+      "epoch": 0.18874360432063672,
+      "grad_norm": 8.098290875630656,
+      "learning_rate": 4.676965137714626e-06,
+      "loss": 0.5824,
+      "step": 2324
+    },
+    {
+      "epoch": 0.1888248192966783,
+      "grad_norm": 5.792865775052247,
+      "learning_rate": 4.676641734484457e-06,
+      "loss": 0.5991,
+      "step": 2325
+    },
+    {
+      "epoch": 0.1889060342727199,
+      "grad_norm": 6.158616589247779,
+      "learning_rate": 4.6763181806422e-06,
+      "loss": 0.6869,
+      "step": 2326
+    },
+    {
+      "epoch": 0.18898724924876148,
+      "grad_norm": 6.1672909888851395,
+      "learning_rate": 4.675994476210243e-06,
+      "loss": 0.4876,
+      "step": 2327
+    },
+    {
+      "epoch": 0.18906846422480306,
+      "grad_norm": 4.374487244986078,
+      "learning_rate": 4.675670621210985e-06,
+      "loss": 0.4851,
+      "step": 2328
+    },
+    {
+      "epoch": 0.18914967920084463,
+      "grad_norm": 4.378802636321053,
+      "learning_rate": 4.675346615666834e-06,
+      "loss": 0.5938,
+      "step": 2329
+    },
+    {
+      "epoch": 0.1892308941768862,
+      "grad_norm": 6.805865659937306,
+      "learning_rate": 4.675022459600209e-06,
+      "loss": 0.5517,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1893121091529278,
+      "grad_norm": 10.069257270167085,
+      "learning_rate": 4.674698153033542e-06,
+      "loss": 0.5793,
+      "step": 2331
+    },
+    {
+      "epoch": 0.1893933241289694,
+      "grad_norm": 5.002867430656955,
+      "learning_rate": 4.674373695989272e-06,
+      "loss": 0.4501,
+      "step": 2332
+    },
+    {
+      "epoch": 0.18947453910501097,
+      "grad_norm": 6.5558875829671726,
+      "learning_rate": 4.67404908848985e-06,
+      "loss": 0.6941,
+      "step": 2333
+    },
+    {
+      "epoch": 0.18955575408105255,
+      "grad_norm": 3.310181376592309,
+      "learning_rate": 4.673724330557737e-06,
+      "loss": 0.5876,
+      "step": 2334
+    },
+    {
+      "epoch": 0.18963696905709412,
+      "grad_norm": 4.3922451080442615,
+      "learning_rate": 4.673399422215405e-06,
+      "loss": 0.6729,
+      "step": 2335
+    },
+    {
+      "epoch": 0.1897181840331357,
+      "grad_norm": 3.2619584312724124,
+      "learning_rate": 4.673074363485336e-06,
+      "loss": 0.6935,
+      "step": 2336
+    },
+    {
+      "epoch": 0.1897993990091773,
+      "grad_norm": 4.553167577761572,
+      "learning_rate": 4.672749154390021e-06,
+      "loss": 0.4633,
+      "step": 2337
+    },
+    {
+      "epoch": 0.18988061398521888,
+      "grad_norm": 3.149250992164633,
+      "learning_rate": 4.6724237949519635e-06,
+      "loss": 0.6547,
+      "step": 2338
+    },
+    {
+      "epoch": 0.18996182896126046,
+      "grad_norm": 3.8540236289067646,
+      "learning_rate": 4.672098285193677e-06,
+      "loss": 0.5031,
+      "step": 2339
+    },
+    {
+      "epoch": 0.19004304393730204,
+      "grad_norm": 5.029855666317853,
+      "learning_rate": 4.671772625137685e-06,
+      "loss": 0.5433,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1901242589133436,
+      "grad_norm": 4.4089431032158,
+      "learning_rate": 4.6714468148065215e-06,
+      "loss": 0.6162,
+      "step": 2341
+    },
+    {
+      "epoch": 0.1902054738893852,
+      "grad_norm": 4.951136934389378,
+      "learning_rate": 4.67112085422273e-06,
+      "loss": 0.5549,
+      "step": 2342
+    },
+    {
+      "epoch": 0.1902866888654268,
+      "grad_norm": 4.969440488977865,
+      "learning_rate": 4.6707947434088665e-06,
+      "loss": 0.5311,
+      "step": 2343
+    },
+    {
+      "epoch": 0.19036790384146837,
+      "grad_norm": 7.467675141875417,
+      "learning_rate": 4.670468482387495e-06,
+      "loss": 0.5161,
+      "step": 2344
+    },
+    {
+      "epoch": 0.19044911881750995,
+      "grad_norm": 5.817463721744332,
+      "learning_rate": 4.670142071181192e-06,
+      "loss": 0.5573,
+      "step": 2345
+    },
+    {
+      "epoch": 0.19053033379355153,
+      "grad_norm": 3.588261641157244,
+      "learning_rate": 4.6698155098125435e-06,
+      "loss": 0.4921,
+      "step": 2346
+    },
+    {
+      "epoch": 0.1906115487695931,
+      "grad_norm": 5.247075946408668,
+      "learning_rate": 4.6694887983041434e-06,
+      "loss": 0.6335,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1906927637456347,
+      "grad_norm": 5.98092813691066,
+      "learning_rate": 4.669161936678602e-06,
+      "loss": 0.6301,
+      "step": 2348
+    },
+    {
+      "epoch": 0.19077397872167629,
+      "grad_norm": 4.882612458777484,
+      "learning_rate": 4.668834924958534e-06,
+      "loss": 0.605,
+      "step": 2349
+    },
+    {
+      "epoch": 0.19085519369771786,
+      "grad_norm": 4.135928802898827,
+      "learning_rate": 4.668507763166568e-06,
+      "loss": 0.7428,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19093640867375944,
+      "grad_norm": 6.436181992978232,
+      "learning_rate": 4.668180451325341e-06,
+      "loss": 0.6098,
+      "step": 2351
+    },
+    {
+      "epoch": 0.19101762364980102,
+      "grad_norm": 8.974548488637932,
+      "learning_rate": 4.667852989457502e-06,
+      "loss": 0.5165,
+      "step": 2352
+    },
+    {
+      "epoch": 0.1910988386258426,
+      "grad_norm": 4.064278950992003,
+      "learning_rate": 4.6675253775857096e-06,
+      "loss": 0.5658,
+      "step": 2353
+    },
+    {
+      "epoch": 0.1911800536018842,
+      "grad_norm": 4.497922853159247,
+      "learning_rate": 4.667197615732633e-06,
+      "loss": 0.6748,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19126126857792577,
+      "grad_norm": 4.494597527974153,
+      "learning_rate": 4.66686970392095e-06,
+      "loss": 0.5488,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19134248355396735,
+      "grad_norm": 5.505569141881063,
+      "learning_rate": 4.666541642173352e-06,
+      "loss": 0.5822,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19142369853000893,
+      "grad_norm": 3.730344521450571,
+      "learning_rate": 4.666213430512538e-06,
+      "loss": 0.634,
+      "step": 2357
+    },
+    {
+      "epoch": 0.1915049135060505,
+      "grad_norm": 5.750851724066535,
+      "learning_rate": 4.66588506896122e-06,
+      "loss": 0.5967,
+      "step": 2358
+    },
+    {
+      "epoch": 0.1915861284820921,
+      "grad_norm": 5.307685486956423,
+      "learning_rate": 4.665556557542118e-06,
+      "loss": 0.7144,
+      "step": 2359
+    },
+    {
+      "epoch": 0.1916673434581337,
+      "grad_norm": 5.143047388945135,
+      "learning_rate": 4.6652278962779625e-06,
+      "loss": 0.6383,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19174855843417526,
+      "grad_norm": 4.999760195133255,
+      "learning_rate": 4.664899085191496e-06,
+      "loss": 0.6477,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19182977341021684,
+      "grad_norm": 7.787092748476295,
+      "learning_rate": 4.664570124305472e-06,
+      "loss": 0.5304,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19191098838625842,
+      "grad_norm": 4.318743886643792,
+      "learning_rate": 4.66424101364265e-06,
+      "loss": 0.4677,
+      "step": 2363
+    },
+    {
+      "epoch": 0.1919922033623,
+      "grad_norm": 4.9984105511208545,
+      "learning_rate": 4.663911753225803e-06,
+      "loss": 0.5786,
+      "step": 2364
+    },
+    {
+      "epoch": 0.1920734183383416,
+      "grad_norm": 3.7096978585490556,
+      "learning_rate": 4.663582343077716e-06,
+      "loss": 0.4654,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19215463331438318,
+      "grad_norm": 5.0966976399052815,
+      "learning_rate": 4.663252783221182e-06,
+      "loss": 0.4797,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19223584829042475,
+      "grad_norm": 11.273482637934793,
+      "learning_rate": 4.662923073679003e-06,
+      "loss": 0.5612,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19231706326646633,
+      "grad_norm": 5.498434540946209,
+      "learning_rate": 4.662593214473995e-06,
+      "loss": 0.6027,
+      "step": 2368
+    },
+    {
+      "epoch": 0.1923982782425079,
+      "grad_norm": 3.859479042229787,
+      "learning_rate": 4.662263205628983e-06,
+      "loss": 0.7433,
+      "step": 2369
+    },
+    {
+      "epoch": 0.1924794932185495,
+      "grad_norm": 4.418822540532742,
+      "learning_rate": 4.661933047166799e-06,
+      "loss": 0.4785,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1925607081945911,
+      "grad_norm": 5.051803928696753,
+      "learning_rate": 4.661602739110291e-06,
+      "loss": 0.5654,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19264192317063267,
+      "grad_norm": 4.0841298739845895,
+      "learning_rate": 4.661272281482313e-06,
+      "loss": 0.7034,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19272313814667424,
+      "grad_norm": 7.928386324620138,
+      "learning_rate": 4.660941674305732e-06,
+      "loss": 0.57,
+      "step": 2373
+    },
+    {
+      "epoch": 0.19280435312271582,
+      "grad_norm": 6.641168636182495,
+      "learning_rate": 4.660610917603423e-06,
+      "loss": 0.4108,
+      "step": 2374
+    },
+    {
+      "epoch": 0.1928855680987574,
+      "grad_norm": 7.233453638420033,
+      "learning_rate": 4.6602800113982746e-06,
+      "loss": 0.5886,
+      "step": 2375
+    },
+    {
+      "epoch": 0.192966783074799,
+      "grad_norm": 5.3241909788384385,
+      "learning_rate": 4.659948955713181e-06,
+      "loss": 0.6516,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19304799805084058,
+      "grad_norm": 6.517892337897202,
+      "learning_rate": 4.659617750571052e-06,
+      "loss": 0.5341,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19312921302688216,
+      "grad_norm": 8.78691467540056,
+      "learning_rate": 4.659286395994806e-06,
+      "loss": 0.7056,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19321042800292373,
+      "grad_norm": 5.510965559445886,
+      "learning_rate": 4.658954892007367e-06,
+      "loss": 0.4839,
+      "step": 2379
+    },
+    {
+      "epoch": 0.1932916429789653,
+      "grad_norm": 5.065066805659556,
+      "learning_rate": 4.658623238631675e-06,
+      "loss": 0.5688,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19337285795500692,
+      "grad_norm": 4.049904522221411,
+      "learning_rate": 4.658291435890681e-06,
+      "loss": 0.5662,
+      "step": 2381
+    },
+    {
+      "epoch": 0.1934540729310485,
+      "grad_norm": 5.778500245146091,
+      "learning_rate": 4.657959483807342e-06,
+      "loss": 0.52,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19353528790709007,
+      "grad_norm": 4.275606455147308,
+      "learning_rate": 4.657627382404627e-06,
+      "loss": 0.4357,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19361650288313165,
+      "grad_norm": 4.9185646028031345,
+      "learning_rate": 4.657295131705516e-06,
+      "loss": 0.6067,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19369771785917322,
+      "grad_norm": 4.858322104090867,
+      "learning_rate": 4.6569627317329995e-06,
+      "loss": 0.58,
+      "step": 2385
+    },
+    {
+      "epoch": 0.1937789328352148,
+      "grad_norm": 5.010825230384572,
+      "learning_rate": 4.656630182510078e-06,
+      "loss": 0.5096,
+      "step": 2386
+    },
+    {
+      "epoch": 0.1938601478112564,
+      "grad_norm": 4.971344756018251,
+      "learning_rate": 4.656297484059761e-06,
+      "loss": 0.5781,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19394136278729798,
+      "grad_norm": 5.7140608376861906,
+      "learning_rate": 4.655964636405071e-06,
+      "loss": 0.5301,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19402257776333956,
+      "grad_norm": 5.25952348586539,
+      "learning_rate": 4.655631639569037e-06,
+      "loss": 0.5345,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19410379273938114,
+      "grad_norm": 4.858225371501829,
+      "learning_rate": 4.655298493574704e-06,
+      "loss": 0.5165,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19418500771542271,
+      "grad_norm": 7.9815798526373145,
+      "learning_rate": 4.65496519844512e-06,
+      "loss": 0.5336,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19426622269146432,
+      "grad_norm": 6.221631927317744,
+      "learning_rate": 4.654631754203351e-06,
+      "loss": 0.5233,
+      "step": 2392
+    },
+    {
+      "epoch": 0.1943474376675059,
+      "grad_norm": 5.199395166051189,
+      "learning_rate": 4.6542981608724665e-06,
+      "loss": 0.5399,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19442865264354747,
+      "grad_norm": 5.85510080538778,
+      "learning_rate": 4.6539644184755515e-06,
+      "loss": 0.6658,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19450986761958905,
+      "grad_norm": 5.187441178217061,
+      "learning_rate": 4.6536305270356975e-06,
+      "loss": 0.5119,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19459108259563063,
+      "grad_norm": 4.41810789985307,
+      "learning_rate": 4.65329648657601e-06,
+      "loss": 0.3997,
+      "step": 2396
+    },
+    {
+      "epoch": 0.1946722975716722,
+      "grad_norm": 5.73516894480618,
+      "learning_rate": 4.652962297119601e-06,
+      "loss": 0.7692,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1947535125477138,
+      "grad_norm": 7.959466481454268,
+      "learning_rate": 4.652627958689596e-06,
+      "loss": 0.4979,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19483472752375539,
+      "grad_norm": 5.346996268227039,
+      "learning_rate": 4.65229347130913e-06,
+      "loss": 0.6685,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19491594249979696,
+      "grad_norm": 14.105527004476242,
+      "learning_rate": 4.651958835001345e-06,
+      "loss": 0.5679,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19499715747583854,
+      "grad_norm": 5.030593581231049,
+      "learning_rate": 4.651624049789397e-06,
+      "loss": 0.4162,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19507837245188012,
+      "grad_norm": 10.896072520758361,
+      "learning_rate": 4.651289115696454e-06,
+      "loss": 0.479,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19515958742792172,
+      "grad_norm": 4.958720750747442,
+      "learning_rate": 4.650954032745689e-06,
+      "loss": 0.5053,
+      "step": 2403
+    },
+    {
+      "epoch": 0.1952408024039633,
+      "grad_norm": 3.974794068624415,
+      "learning_rate": 4.6506188009602885e-06,
+      "loss": 0.574,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19532201738000488,
+      "grad_norm": 8.24891618792213,
+      "learning_rate": 4.65028342036345e-06,
+      "loss": 0.514,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19540323235604645,
+      "grad_norm": 4.335525878369569,
+      "learning_rate": 4.6499478909783764e-06,
+      "loss": 0.427,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19548444733208803,
+      "grad_norm": 6.424849313389938,
+      "learning_rate": 4.649612212828289e-06,
+      "loss": 0.4575,
+      "step": 2407
+    },
+    {
+      "epoch": 0.1955656623081296,
+      "grad_norm": 4.535461924057379,
+      "learning_rate": 4.6492763859364134e-06,
+      "loss": 0.6179,
+      "step": 2408
+    },
+    {
+      "epoch": 0.1956468772841712,
+      "grad_norm": 7.409680541443817,
+      "learning_rate": 4.648940410325987e-06,
+      "loss": 0.6637,
+      "step": 2409
+    },
+    {
+      "epoch": 0.1957280922602128,
+      "grad_norm": 4.6863434003973845,
+      "learning_rate": 4.648604286020256e-06,
+      "loss": 0.5738,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19580930723625437,
+      "grad_norm": 5.544427789484983,
+      "learning_rate": 4.64826801304248e-06,
+      "loss": 0.4905,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19589052221229594,
+      "grad_norm": 6.119211167714467,
+      "learning_rate": 4.647931591415929e-06,
+      "loss": 0.7071,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19597173718833752,
+      "grad_norm": 3.1340827150045714,
+      "learning_rate": 4.647595021163878e-06,
+      "loss": 0.6431,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19605295216437912,
+      "grad_norm": 5.376283662779039,
+      "learning_rate": 4.647258302309618e-06,
+      "loss": 0.6288,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1961341671404207,
+      "grad_norm": 6.177768802745785,
+      "learning_rate": 4.646921434876447e-06,
+      "loss": 0.4535,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19621538211646228,
+      "grad_norm": 12.088856729370683,
+      "learning_rate": 4.646584418887675e-06,
+      "loss": 0.6907,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19629659709250386,
+      "grad_norm": 8.938823511128325,
+      "learning_rate": 4.646247254366622e-06,
+      "loss": 0.5471,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19637781206854543,
+      "grad_norm": 6.009413684644707,
+      "learning_rate": 4.645909941336619e-06,
+      "loss": 0.6261,
+      "step": 2418
+    },
+    {
+      "epoch": 0.196459027044587,
+      "grad_norm": 5.0207649299589034,
+      "learning_rate": 4.645572479821004e-06,
+      "loss": 0.5226,
+      "step": 2419
+    },
+    {
+      "epoch": 0.1965402420206286,
+      "grad_norm": 5.8465718027493025,
+      "learning_rate": 4.645234869843129e-06,
+      "loss": 0.7304,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1966214569966702,
+      "grad_norm": 4.44988664177105,
+      "learning_rate": 4.644897111426355e-06,
+      "loss": 0.4947,
+      "step": 2421
+    },
+    {
+      "epoch": 0.19670267197271177,
+      "grad_norm": 4.080957228552931,
+      "learning_rate": 4.6445592045940515e-06,
+      "loss": 0.6638,
+      "step": 2422
+    },
+    {
+      "epoch": 0.19678388694875334,
+      "grad_norm": 6.67424212054808,
+      "learning_rate": 4.644221149369602e-06,
+      "loss": 0.5218,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19686510192479492,
+      "grad_norm": 4.3515760143355084,
+      "learning_rate": 4.643882945776397e-06,
+      "loss": 0.5116,
+      "step": 2424
+    },
+    {
+      "epoch": 0.19694631690083653,
+      "grad_norm": 4.665310299217762,
+      "learning_rate": 4.6435445938378375e-06,
+      "loss": 0.5858,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1970275318768781,
+      "grad_norm": 3.4118287749102367,
+      "learning_rate": 4.643206093577338e-06,
+      "loss": 0.3923,
+      "step": 2426
+    },
+    {
+      "epoch": 0.19710874685291968,
+      "grad_norm": 4.722707242292216,
+      "learning_rate": 4.642867445018318e-06,
+      "loss": 0.5732,
+      "step": 2427
+    },
+    {
+      "epoch": 0.19718996182896126,
+      "grad_norm": 4.984185032243032,
+      "learning_rate": 4.642528648184213e-06,
+      "loss": 0.4799,
+      "step": 2428
+    },
+    {
+      "epoch": 0.19727117680500283,
+      "grad_norm": 4.175601292098027,
+      "learning_rate": 4.642189703098466e-06,
+      "loss": 0.5171,
+      "step": 2429
+    },
+    {
+      "epoch": 0.1973523917810444,
+      "grad_norm": 8.539390583313239,
+      "learning_rate": 4.6418506097845264e-06,
+      "loss": 0.7051,
+      "step": 2430
+    },
+    {
+      "epoch": 0.19743360675708602,
+      "grad_norm": 6.857246634219236,
+      "learning_rate": 4.641511368265861e-06,
+      "loss": 0.491,
+      "step": 2431
+    },
+    {
+      "epoch": 0.1975148217331276,
+      "grad_norm": 8.537762993370494,
+      "learning_rate": 4.641171978565943e-06,
+      "loss": 0.7512,
+      "step": 2432
+    },
+    {
+      "epoch": 0.19759603670916917,
+      "grad_norm": 4.408249935592557,
+      "learning_rate": 4.640832440708256e-06,
+      "loss": 0.6716,
+      "step": 2433
+    },
+    {
+      "epoch": 0.19767725168521075,
+      "grad_norm": 5.109570036404427,
+      "learning_rate": 4.640492754716294e-06,
+      "loss": 0.5133,
+      "step": 2434
+    },
+    {
+      "epoch": 0.19775846666125232,
+      "grad_norm": 4.736765394511924,
+      "learning_rate": 4.640152920613562e-06,
+      "loss": 0.9746,
+      "step": 2435
+    },
+    {
+      "epoch": 0.19783968163729393,
+      "grad_norm": 5.063603609846527,
+      "learning_rate": 4.639812938423574e-06,
+      "loss": 0.4769,
+      "step": 2436
+    },
+    {
+      "epoch": 0.1979208966133355,
+      "grad_norm": 5.272415199098991,
+      "learning_rate": 4.639472808169857e-06,
+      "loss": 0.5295,
+      "step": 2437
+    },
+    {
+      "epoch": 0.19800211158937708,
+      "grad_norm": 8.257542098149266,
+      "learning_rate": 4.639132529875943e-06,
+      "loss": 0.6467,
+      "step": 2438
+    },
+    {
+      "epoch": 0.19808332656541866,
+      "grad_norm": 5.3139070377916315,
+      "learning_rate": 4.63879210356538e-06,
+      "loss": 0.5433,
+      "step": 2439
+    },
+    {
+      "epoch": 0.19816454154146024,
+      "grad_norm": 6.841621560493213,
+      "learning_rate": 4.6384515292617226e-06,
+      "loss": 0.5335,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19824575651750181,
+      "grad_norm": 8.402046267183342,
+      "learning_rate": 4.6381108069885376e-06,
+      "loss": 0.5433,
+      "step": 2441
+    },
+    {
+      "epoch": 0.19832697149354342,
+      "grad_norm": 3.6536735919997674,
+      "learning_rate": 4.6377699367694e-06,
+      "loss": 0.5768,
+      "step": 2442
+    },
+    {
+      "epoch": 0.198408186469585,
+      "grad_norm": 6.961987557428991,
+      "learning_rate": 4.637428918627896e-06,
+      "loss": 0.5184,
+      "step": 2443
+    },
+    {
+      "epoch": 0.19848940144562657,
+      "grad_norm": 5.119773267513261,
+      "learning_rate": 4.637087752587624e-06,
+      "loss": 0.6723,
+      "step": 2444
+    },
+    {
+      "epoch": 0.19857061642166815,
+      "grad_norm": 5.37579569573017,
+      "learning_rate": 4.636746438672189e-06,
+      "loss": 0.4536,
+      "step": 2445
+    },
+    {
+      "epoch": 0.19865183139770973,
+      "grad_norm": 4.8653070857140825,
+      "learning_rate": 4.63640497690521e-06,
+      "loss": 0.7141,
+      "step": 2446
+    },
+    {
+      "epoch": 0.19873304637375133,
+      "grad_norm": 5.213638271403077,
+      "learning_rate": 4.636063367310313e-06,
+      "loss": 0.4026,
+      "step": 2447
+    },
+    {
+      "epoch": 0.1988142613497929,
+      "grad_norm": 4.022917999876865,
+      "learning_rate": 4.635721609911137e-06,
+      "loss": 0.8184,
+      "step": 2448
+    },
+    {
+      "epoch": 0.19889547632583449,
+      "grad_norm": 5.721442337840194,
+      "learning_rate": 4.635379704731327e-06,
+      "loss": 0.5173,
+      "step": 2449
+    },
+    {
+      "epoch": 0.19897669130187606,
+      "grad_norm": 4.746208323114987,
+      "learning_rate": 4.635037651794544e-06,
+      "loss": 0.6091,
+      "step": 2450
+    },
+    {
+      "epoch": 0.19905790627791764,
+      "grad_norm": 3.7466714374073393,
+      "learning_rate": 4.634695451124454e-06,
+      "loss": 0.5977,
+      "step": 2451
+    },
+    {
+      "epoch": 0.19913912125395922,
+      "grad_norm": 5.06981488833001,
+      "learning_rate": 4.634353102744737e-06,
+      "loss": 0.4463,
+      "step": 2452
+    },
+    {
+      "epoch": 0.19922033623000082,
+      "grad_norm": 4.625187502181156,
+      "learning_rate": 4.634010606679081e-06,
+      "loss": 0.5105,
+      "step": 2453
+    },
+    {
+      "epoch": 0.1993015512060424,
+      "grad_norm": 4.590440891315889,
+      "learning_rate": 4.633667962951186e-06,
+      "loss": 0.5703,
+      "step": 2454
+    },
+    {
+      "epoch": 0.19938276618208398,
+      "grad_norm": 8.282884396118154,
+      "learning_rate": 4.6333251715847595e-06,
+      "loss": 0.5573,
+      "step": 2455
+    },
+    {
+      "epoch": 0.19946398115812555,
+      "grad_norm": 5.99824415434856,
+      "learning_rate": 4.6329822326035214e-06,
+      "loss": 0.6637,
+      "step": 2456
+    },
+    {
+      "epoch": 0.19954519613416713,
+      "grad_norm": 4.69181027886392,
+      "learning_rate": 4.632639146031201e-06,
+      "loss": 0.548,
+      "step": 2457
+    },
+    {
+      "epoch": 0.19962641111020873,
+      "grad_norm": 4.774146261889797,
+      "learning_rate": 4.63229591189154e-06,
+      "loss": 0.5267,
+      "step": 2458
+    },
+    {
+      "epoch": 0.1997076260862503,
+      "grad_norm": 4.0388362634362505,
+      "learning_rate": 4.631952530208286e-06,
+      "loss": 0.6004,
+      "step": 2459
+    },
+    {
+      "epoch": 0.1997888410622919,
+      "grad_norm": 11.372135655939333,
+      "learning_rate": 4.6316090010052006e-06,
+      "loss": 0.5377,
+      "step": 2460
+    },
+    {
+      "epoch": 0.19987005603833347,
+      "grad_norm": 8.258738666338818,
+      "learning_rate": 4.631265324306053e-06,
+      "loss": 0.6255,
+      "step": 2461
+    },
+    {
+      "epoch": 0.19995127101437504,
+      "grad_norm": 4.838265403873367,
+      "learning_rate": 4.630921500134625e-06,
+      "loss": 0.513,
+      "step": 2462
+    },
+    {
+      "epoch": 0.20003248599041662,
+      "grad_norm": 6.210942277246636,
+      "learning_rate": 4.630577528514707e-06,
+      "loss": 0.5705,
+      "step": 2463
+    },
+    {
+      "epoch": 0.20011370096645822,
+      "grad_norm": 4.867014588828138,
+      "learning_rate": 4.6302334094701e-06,
+      "loss": 0.7147,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2001949159424998,
+      "grad_norm": 3.4107828105330116,
+      "learning_rate": 4.629889143024615e-06,
+      "loss": 0.4722,
+      "step": 2465
+    },
+    {
+      "epoch": 0.20027613091854138,
+      "grad_norm": 6.2252845176976574,
+      "learning_rate": 4.6295447292020735e-06,
+      "loss": 0.5165,
+      "step": 2466
+    },
+    {
+      "epoch": 0.20035734589458296,
+      "grad_norm": 5.732379914152118,
+      "learning_rate": 4.629200168026307e-06,
+      "loss": 0.4561,
+      "step": 2467
+    },
+    {
+      "epoch": 0.20043856087062453,
+      "grad_norm": 6.463671611218234,
+      "learning_rate": 4.6288554595211575e-06,
+      "loss": 0.5683,
+      "step": 2468
+    },
+    {
+      "epoch": 0.20051977584666614,
+      "grad_norm": 6.359868474828741,
+      "learning_rate": 4.628510603710478e-06,
+      "loss": 0.5942,
+      "step": 2469
+    },
+    {
+      "epoch": 0.20060099082270771,
+      "grad_norm": 5.920271897280425,
+      "learning_rate": 4.628165600618129e-06,
+      "loss": 0.5336,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2006822057987493,
+      "grad_norm": 4.578491403510653,
+      "learning_rate": 4.627820450267984e-06,
+      "loss": 0.4962,
+      "step": 2471
+    },
+    {
+      "epoch": 0.20076342077479087,
+      "grad_norm": 3.829851338031233,
+      "learning_rate": 4.627475152683924e-06,
+      "loss": 0.554,
+      "step": 2472
+    },
+    {
+      "epoch": 0.20084463575083245,
+      "grad_norm": 5.016784542180406,
+      "learning_rate": 4.627129707889843e-06,
+      "loss": 0.7465,
+      "step": 2473
+    },
+    {
+      "epoch": 0.20092585072687402,
+      "grad_norm": 7.808176500440273,
+      "learning_rate": 4.626784115909645e-06,
+      "loss": 0.5234,
+      "step": 2474
+    },
+    {
+      "epoch": 0.20100706570291563,
+      "grad_norm": 3.777859086820638,
+      "learning_rate": 4.626438376767241e-06,
+      "loss": 0.7113,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2010882806789572,
+      "grad_norm": 6.167573513223816,
+      "learning_rate": 4.626092490486557e-06,
+      "loss": 0.5625,
+      "step": 2476
+    },
+    {
+      "epoch": 0.20116949565499878,
+      "grad_norm": 4.609400429566531,
+      "learning_rate": 4.6257464570915235e-06,
+      "loss": 0.39,
+      "step": 2477
+    },
+    {
+      "epoch": 0.20125071063104036,
+      "grad_norm": 4.207144293674372,
+      "learning_rate": 4.625400276606086e-06,
+      "loss": 0.4698,
+      "step": 2478
+    },
+    {
+      "epoch": 0.20133192560708194,
+      "grad_norm": 4.988407235613174,
+      "learning_rate": 4.625053949054198e-06,
+      "loss": 0.6785,
+      "step": 2479
+    },
+    {
+      "epoch": 0.20141314058312354,
+      "grad_norm": 3.9367620435534314,
+      "learning_rate": 4.6247074744598234e-06,
+      "loss": 0.7332,
+      "step": 2480
+    },
+    {
+      "epoch": 0.20149435555916512,
+      "grad_norm": 3.9271827060994564,
+      "learning_rate": 4.6243608528469356e-06,
+      "loss": 0.489,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2015755705352067,
+      "grad_norm": 5.02867311978453,
+      "learning_rate": 4.6240140842395205e-06,
+      "loss": 0.5228,
+      "step": 2482
+    },
+    {
+      "epoch": 0.20165678551124827,
+      "grad_norm": 3.9679711814084406,
+      "learning_rate": 4.623667168661572e-06,
+      "loss": 0.8265,
+      "step": 2483
+    },
+    {
+      "epoch": 0.20173800048728985,
+      "grad_norm": 4.395768395942478,
+      "learning_rate": 4.623320106137095e-06,
+      "loss": 0.5777,
+      "step": 2484
+    },
+    {
+      "epoch": 0.20181921546333143,
+      "grad_norm": 4.203898879590677,
+      "learning_rate": 4.6229728966901036e-06,
+      "loss": 0.7345,
+      "step": 2485
+    },
+    {
+      "epoch": 0.20190043043937303,
+      "grad_norm": 3.5440873158516673,
+      "learning_rate": 4.622625540344623e-06,
+      "loss": 0.6419,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2019816454154146,
+      "grad_norm": 2.545901550496782,
+      "learning_rate": 4.62227803712469e-06,
+      "loss": 0.6928,
+      "step": 2487
+    },
+    {
+      "epoch": 0.20206286039145618,
+      "grad_norm": 4.89784077750463,
+      "learning_rate": 4.621930387054349e-06,
+      "loss": 0.5345,
+      "step": 2488
+    },
+    {
+      "epoch": 0.20214407536749776,
+      "grad_norm": 5.15304789976238,
+      "learning_rate": 4.621582590157654e-06,
+      "loss": 0.5168,
+      "step": 2489
+    },
+    {
+      "epoch": 0.20222529034353934,
+      "grad_norm": 3.6944859365300062,
+      "learning_rate": 4.621234646458673e-06,
+      "loss": 0.6469,
+      "step": 2490
+    },
+    {
+      "epoch": 0.20230650531958094,
+      "grad_norm": 4.253884701651634,
+      "learning_rate": 4.6208865559814805e-06,
+      "loss": 0.6652,
+      "step": 2491
+    },
+    {
+      "epoch": 0.20238772029562252,
+      "grad_norm": 5.40192379256261,
+      "learning_rate": 4.620538318750163e-06,
+      "loss": 0.5256,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2024689352716641,
+      "grad_norm": 3.4100834853157704,
+      "learning_rate": 4.620189934788817e-06,
+      "loss": 0.6897,
+      "step": 2493
+    },
+    {
+      "epoch": 0.20255015024770567,
+      "grad_norm": 3.8199295718971102,
+      "learning_rate": 4.6198414041215484e-06,
+      "loss": 0.5001,
+      "step": 2494
+    },
+    {
+      "epoch": 0.20263136522374725,
+      "grad_norm": 3.145895750195355,
+      "learning_rate": 4.619492726772473e-06,
+      "loss": 0.6773,
+      "step": 2495
+    },
+    {
+      "epoch": 0.20271258019978883,
+      "grad_norm": 4.373203705255476,
+      "learning_rate": 4.619143902765719e-06,
+      "loss": 0.6648,
+      "step": 2496
+    },
+    {
+      "epoch": 0.20279379517583043,
+      "grad_norm": 6.177848983200393,
+      "learning_rate": 4.618794932125422e-06,
+      "loss": 0.4791,
+      "step": 2497
+    },
+    {
+      "epoch": 0.202875010151872,
+      "grad_norm": 3.952402300031322,
+      "learning_rate": 4.61844581487573e-06,
+      "loss": 0.8067,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2029562251279136,
+      "grad_norm": 3.670343152102567,
+      "learning_rate": 4.618096551040798e-06,
+      "loss": 0.502,
+      "step": 2499
+    },
+    {
+      "epoch": 0.20303744010395516,
+      "grad_norm": 4.359414649768362,
+      "learning_rate": 4.617747140644796e-06,
+      "loss": 0.5255,
+      "step": 2500
+    },
+    {
+      "epoch": 0.20311865507999674,
+      "grad_norm": 6.653021158021161,
+      "learning_rate": 4.617397583711899e-06,
+      "loss": 0.6263,
+      "step": 2501
+    },
+    {
+      "epoch": 0.20319987005603835,
+      "grad_norm": 4.748016005088914,
+      "learning_rate": 4.617047880266295e-06,
+      "loss": 0.6699,
+      "step": 2502
+    },
+    {
+      "epoch": 0.20328108503207992,
+      "grad_norm": 3.35437246059798,
+      "learning_rate": 4.616698030332183e-06,
+      "loss": 0.4972,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2033623000081215,
+      "grad_norm": 14.870459669223077,
+      "learning_rate": 4.616348033933769e-06,
+      "loss": 0.5515,
+      "step": 2504
+    },
+    {
+      "epoch": 0.20344351498416308,
+      "grad_norm": 5.088873364458127,
+      "learning_rate": 4.615997891095272e-06,
+      "loss": 0.6022,
+      "step": 2505
+    },
+    {
+      "epoch": 0.20352472996020465,
+      "grad_norm": 3.7990841422951447,
+      "learning_rate": 4.6156476018409204e-06,
+      "loss": 0.6326,
+      "step": 2506
+    },
+    {
+      "epoch": 0.20360594493624623,
+      "grad_norm": 4.333396061155222,
+      "learning_rate": 4.61529716619495e-06,
+      "loss": 0.6267,
+      "step": 2507
+    },
+    {
+      "epoch": 0.20368715991228784,
+      "grad_norm": 4.685302979204798,
+      "learning_rate": 4.614946584181612e-06,
+      "loss": 0.6191,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2037683748883294,
+      "grad_norm": 4.124779647332827,
+      "learning_rate": 4.614595855825164e-06,
+      "loss": 0.4998,
+      "step": 2509
+    },
+    {
+      "epoch": 0.203849589864371,
+      "grad_norm": 4.501692024239236,
+      "learning_rate": 4.6142449811498725e-06,
+      "loss": 0.5782,
+      "step": 2510
+    },
+    {
+      "epoch": 0.20393080484041257,
+      "grad_norm": 4.707923879718902,
+      "learning_rate": 4.613893960180018e-06,
+      "loss": 0.5297,
+      "step": 2511
+    },
+    {
+      "epoch": 0.20401201981645414,
+      "grad_norm": 7.3215672264551035,
+      "learning_rate": 4.613542792939891e-06,
+      "loss": 0.7369,
+      "step": 2512
+    },
+    {
+      "epoch": 0.20409323479249575,
+      "grad_norm": 5.2423338527201615,
+      "learning_rate": 4.613191479453787e-06,
+      "loss": 0.8081,
+      "step": 2513
+    },
+    {
+      "epoch": 0.20417444976853732,
+      "grad_norm": 5.115816586092132,
+      "learning_rate": 4.612840019746016e-06,
+      "loss": 0.67,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2042556647445789,
+      "grad_norm": 3.4887642989488348,
+      "learning_rate": 4.612488413840899e-06,
+      "loss": 0.8183,
+      "step": 2515
+    },
+    {
+      "epoch": 0.20433687972062048,
+      "grad_norm": 5.202669123420781,
+      "learning_rate": 4.6121366617627635e-06,
+      "loss": 0.4956,
+      "step": 2516
+    },
+    {
+      "epoch": 0.20441809469666206,
+      "grad_norm": 4.783952508433326,
+      "learning_rate": 4.6117847635359494e-06,
+      "loss": 0.5427,
+      "step": 2517
+    },
+    {
+      "epoch": 0.20449930967270363,
+      "grad_norm": 3.8836746569798,
+      "learning_rate": 4.611432719184806e-06,
+      "loss": 0.7236,
+      "step": 2518
+    },
+    {
+      "epoch": 0.20458052464874524,
+      "grad_norm": 4.277705470777233,
+      "learning_rate": 4.611080528733693e-06,
+      "loss": 0.5379,
+      "step": 2519
+    },
+    {
+      "epoch": 0.20466173962478681,
+      "grad_norm": 4.597920469594989,
+      "learning_rate": 4.6107281922069805e-06,
+      "loss": 0.6815,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2047429546008284,
+      "grad_norm": 5.510027910722754,
+      "learning_rate": 4.610375709629047e-06,
+      "loss": 0.5213,
+      "step": 2521
+    },
+    {
+      "epoch": 0.20482416957686997,
+      "grad_norm": 26.571846192694977,
+      "learning_rate": 4.610023081024284e-06,
+      "loss": 0.6494,
+      "step": 2522
+    },
+    {
+      "epoch": 0.20490538455291155,
+      "grad_norm": 7.215116215564211,
+      "learning_rate": 4.6096703064170915e-06,
+      "loss": 0.5739,
+      "step": 2523
+    },
+    {
+      "epoch": 0.20498659952895315,
+      "grad_norm": 5.568682872058116,
+      "learning_rate": 4.609317385831879e-06,
+      "loss": 0.7281,
+      "step": 2524
+    },
+    {
+      "epoch": 0.20506781450499473,
+      "grad_norm": 7.457927768406811,
+      "learning_rate": 4.608964319293066e-06,
+      "loss": 0.6219,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2051490294810363,
+      "grad_norm": 4.3162604433484155,
+      "learning_rate": 4.6086111068250834e-06,
+      "loss": 0.5108,
+      "step": 2526
+    },
+    {
+      "epoch": 0.20523024445707788,
+      "grad_norm": 5.339053755707136,
+      "learning_rate": 4.608257748452372e-06,
+      "loss": 0.4834,
+      "step": 2527
+    },
+    {
+      "epoch": 0.20531145943311946,
+      "grad_norm": 3.319564859771241,
+      "learning_rate": 4.607904244199384e-06,
+      "loss": 0.6459,
+      "step": 2528
+    },
+    {
+      "epoch": 0.20539267440916104,
+      "grad_norm": 3.3949377280505657,
+      "learning_rate": 4.6075505940905765e-06,
+      "loss": 0.8157,
+      "step": 2529
+    },
+    {
+      "epoch": 0.20547388938520264,
+      "grad_norm": 4.846601443639954,
+      "learning_rate": 4.607196798150423e-06,
+      "loss": 0.6862,
+      "step": 2530
+    },
+    {
+      "epoch": 0.20555510436124422,
+      "grad_norm": 3.634653145365433,
+      "learning_rate": 4.606842856403402e-06,
+      "loss": 0.4921,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2056363193372858,
+      "grad_norm": 3.453187944742966,
+      "learning_rate": 4.6064887688740065e-06,
+      "loss": 0.7135,
+      "step": 2532
+    },
+    {
+      "epoch": 0.20571753431332737,
+      "grad_norm": 6.8028434368171,
+      "learning_rate": 4.606134535586737e-06,
+      "loss": 0.7577,
+      "step": 2533
+    },
+    {
+      "epoch": 0.20579874928936895,
+      "grad_norm": 5.81957789301719,
+      "learning_rate": 4.605780156566103e-06,
+      "loss": 0.6454,
+      "step": 2534
+    },
+    {
+      "epoch": 0.20587996426541055,
+      "grad_norm": 3.726486607180334,
+      "learning_rate": 4.6054256318366275e-06,
+      "loss": 0.5258,
+      "step": 2535
+    },
+    {
+      "epoch": 0.20596117924145213,
+      "grad_norm": 6.553307065329235,
+      "learning_rate": 4.6050709614228416e-06,
+      "loss": 0.6497,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2060423942174937,
+      "grad_norm": 6.182500794690838,
+      "learning_rate": 4.604716145349285e-06,
+      "loss": 0.5451,
+      "step": 2537
+    },
+    {
+      "epoch": 0.20612360919353528,
+      "grad_norm": 3.621812427485916,
+      "learning_rate": 4.604361183640511e-06,
+      "loss": 0.5455,
+      "step": 2538
+    },
+    {
+      "epoch": 0.20620482416957686,
+      "grad_norm": 6.106652776271362,
+      "learning_rate": 4.60400607632108e-06,
+      "loss": 0.5335,
+      "step": 2539
+    },
+    {
+      "epoch": 0.20628603914561844,
+      "grad_norm": 8.175917425138248,
+      "learning_rate": 4.603650823415563e-06,
+      "loss": 0.5764,
+      "step": 2540
+    },
+    {
+      "epoch": 0.20636725412166004,
+      "grad_norm": 4.015447399797649,
+      "learning_rate": 4.603295424948544e-06,
+      "loss": 0.514,
+      "step": 2541
+    },
+    {
+      "epoch": 0.20644846909770162,
+      "grad_norm": 4.883808409792754,
+      "learning_rate": 4.602939880944612e-06,
+      "loss": 0.5623,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2065296840737432,
+      "grad_norm": 3.0577448372785745,
+      "learning_rate": 4.6025841914283705e-06,
+      "loss": 0.5176,
+      "step": 2543
+    },
+    {
+      "epoch": 0.20661089904978477,
+      "grad_norm": 3.73178717833002,
+      "learning_rate": 4.602228356424431e-06,
+      "loss": 0.5816,
+      "step": 2544
+    },
+    {
+      "epoch": 0.20669211402582635,
+      "grad_norm": 3.105000510085083,
+      "learning_rate": 4.601872375957414e-06,
+      "loss": 0.7313,
+      "step": 2545
+    },
+    {
+      "epoch": 0.20677332900186796,
+      "grad_norm": 5.053370814100823,
+      "learning_rate": 4.601516250051954e-06,
+      "loss": 0.4578,
+      "step": 2546
+    },
+    {
+      "epoch": 0.20685454397790953,
+      "grad_norm": 3.758806458666759,
+      "learning_rate": 4.601159978732691e-06,
+      "loss": 0.5377,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2069357589539511,
+      "grad_norm": 3.486408365033022,
+      "learning_rate": 4.600803562024277e-06,
+      "loss": 0.6031,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2070169739299927,
+      "grad_norm": 4.0286968600378925,
+      "learning_rate": 4.6004469999513755e-06,
+      "loss": 0.6561,
+      "step": 2549
+    },
+    {
+      "epoch": 0.20709818890603426,
+      "grad_norm": 4.653795014118944,
+      "learning_rate": 4.600090292538658e-06,
+      "loss": 0.5098,
+      "step": 2550
+    },
+    {
+      "epoch": 0.20717940388207584,
+      "grad_norm": 4.008738162793329,
+      "learning_rate": 4.599733439810807e-06,
+      "loss": 0.596,
+      "step": 2551
+    },
+    {
+      "epoch": 0.20726061885811745,
+      "grad_norm": 6.379187521913423,
+      "learning_rate": 4.5993764417925145e-06,
+      "loss": 0.4166,
+      "step": 2552
+    },
+    {
+      "epoch": 0.20734183383415902,
+      "grad_norm": 5.162913747178339,
+      "learning_rate": 4.599019298508482e-06,
+      "loss": 0.6534,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2074230488102006,
+      "grad_norm": 4.626210052826402,
+      "learning_rate": 4.598662009983424e-06,
+      "loss": 0.459,
+      "step": 2554
+    },
+    {
+      "epoch": 0.20750426378624218,
+      "grad_norm": 4.343058752050401,
+      "learning_rate": 4.598304576242063e-06,
+      "loss": 0.5086,
+      "step": 2555
+    },
+    {
+      "epoch": 0.20758547876228375,
+      "grad_norm": 7.300206881482361,
+      "learning_rate": 4.597946997309129e-06,
+      "loss": 0.5035,
+      "step": 2556
+    },
+    {
+      "epoch": 0.20766669373832536,
+      "grad_norm": 3.0899015840363075,
+      "learning_rate": 4.597589273209366e-06,
+      "loss": 0.5187,
+      "step": 2557
+    },
+    {
+      "epoch": 0.20774790871436694,
+      "grad_norm": 6.052813956883026,
+      "learning_rate": 4.597231403967527e-06,
+      "loss": 0.5379,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2078291236904085,
+      "grad_norm": 5.517294439794208,
+      "learning_rate": 4.5968733896083745e-06,
+      "loss": 0.8438,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2079103386664501,
+      "grad_norm": 4.128375503613039,
+      "learning_rate": 4.59651523015668e-06,
+      "loss": 0.6151,
+      "step": 2560
+    },
+    {
+      "epoch": 0.20799155364249167,
+      "grad_norm": 6.118250286731413,
+      "learning_rate": 4.5961569256372285e-06,
+      "loss": 0.5982,
+      "step": 2561
+    },
+    {
+      "epoch": 0.20807276861853324,
+      "grad_norm": 6.553871079659764,
+      "learning_rate": 4.595798476074811e-06,
+      "loss": 0.4833,
+      "step": 2562
+    },
+    {
+      "epoch": 0.20815398359457485,
+      "grad_norm": 5.380395025356601,
+      "learning_rate": 4.59543988149423e-06,
+      "loss": 0.506,
+      "step": 2563
+    },
+    {
+      "epoch": 0.20823519857061643,
+      "grad_norm": 5.37381105799694,
+      "learning_rate": 4.595081141920301e-06,
+      "loss": 0.4393,
+      "step": 2564
+    },
+    {
+      "epoch": 0.208316413546658,
+      "grad_norm": 4.617124660591345,
+      "learning_rate": 4.594722257377844e-06,
+      "loss": 0.6313,
+      "step": 2565
+    },
+    {
+      "epoch": 0.20839762852269958,
+      "grad_norm": 3.1785490526390974,
+      "learning_rate": 4.594363227891693e-06,
+      "loss": 0.6085,
+      "step": 2566
+    },
+    {
+      "epoch": 0.20847884349874116,
+      "grad_norm": 4.209871969565939,
+      "learning_rate": 4.5940040534866905e-06,
+      "loss": 0.8759,
+      "step": 2567
+    },
+    {
+      "epoch": 0.20856005847478276,
+      "grad_norm": 5.084178440847334,
+      "learning_rate": 4.59364473418769e-06,
+      "loss": 0.4414,
+      "step": 2568
+    },
+    {
+      "epoch": 0.20864127345082434,
+      "grad_norm": 4.799592885282473,
+      "learning_rate": 4.593285270019555e-06,
+      "loss": 0.4927,
+      "step": 2569
+    },
+    {
+      "epoch": 0.20872248842686592,
+      "grad_norm": 3.748127121013163,
+      "learning_rate": 4.592925661007157e-06,
+      "loss": 0.7891,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2088037034029075,
+      "grad_norm": 7.646211226477787,
+      "learning_rate": 4.592565907175381e-06,
+      "loss": 0.4914,
+      "step": 2571
+    },
+    {
+      "epoch": 0.20888491837894907,
+      "grad_norm": 4.10621409423765,
+      "learning_rate": 4.592206008549118e-06,
+      "loss": 0.5923,
+      "step": 2572
+    },
+    {
+      "epoch": 0.20896613335499067,
+      "grad_norm": 5.917514515415448,
+      "learning_rate": 4.591845965153272e-06,
+      "loss": 0.6719,
+      "step": 2573
+    },
+    {
+      "epoch": 0.20904734833103225,
+      "grad_norm": 4.0238274192183985,
+      "learning_rate": 4.591485777012757e-06,
+      "loss": 0.636,
+      "step": 2574
+    },
+    {
+      "epoch": 0.20912856330707383,
+      "grad_norm": 5.0280430135245116,
+      "learning_rate": 4.591125444152495e-06,
+      "loss": 0.7144,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2092097782831154,
+      "grad_norm": 5.449721642811628,
+      "learning_rate": 4.590764966597419e-06,
+      "loss": 0.5855,
+      "step": 2576
+    },
+    {
+      "epoch": 0.20929099325915698,
+      "grad_norm": 5.452281756916645,
+      "learning_rate": 4.590404344372472e-06,
+      "loss": 0.5043,
+      "step": 2577
+    },
+    {
+      "epoch": 0.20937220823519856,
+      "grad_norm": 7.9224346525989375,
+      "learning_rate": 4.590043577502609e-06,
+      "loss": 0.5569,
+      "step": 2578
+    },
+    {
+      "epoch": 0.20945342321124016,
+      "grad_norm": 5.570197940829358,
+      "learning_rate": 4.589682666012791e-06,
+      "loss": 0.5516,
+      "step": 2579
+    },
+    {
+      "epoch": 0.20953463818728174,
+      "grad_norm": 4.7228930255767425,
+      "learning_rate": 4.5893216099279925e-06,
+      "loss": 0.7413,
+      "step": 2580
+    },
+    {
+      "epoch": 0.20961585316332332,
+      "grad_norm": 4.327696809821123,
+      "learning_rate": 4.588960409273196e-06,
+      "loss": 0.6607,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2096970681393649,
+      "grad_norm": 3.271070383423843,
+      "learning_rate": 4.588599064073395e-06,
+      "loss": 0.6295,
+      "step": 2582
+    },
+    {
+      "epoch": 0.20977828311540647,
+      "grad_norm": 4.928203939520188,
+      "learning_rate": 4.588237574353592e-06,
+      "loss": 0.4415,
+      "step": 2583
+    },
+    {
+      "epoch": 0.20985949809144808,
+      "grad_norm": 4.524773784130613,
+      "learning_rate": 4.587875940138801e-06,
+      "loss": 0.4493,
+      "step": 2584
+    },
+    {
+      "epoch": 0.20994071306748965,
+      "grad_norm": 5.4612869658664005,
+      "learning_rate": 4.587514161454045e-06,
+      "loss": 0.6702,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21002192804353123,
+      "grad_norm": 5.877089399039422,
+      "learning_rate": 4.587152238324357e-06,
+      "loss": 0.5599,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2101031430195728,
+      "grad_norm": 5.615645343923605,
+      "learning_rate": 4.58679017077478e-06,
+      "loss": 0.6475,
+      "step": 2587
+    },
+    {
+      "epoch": 0.21018435799561438,
+      "grad_norm": 4.466358103769594,
+      "learning_rate": 4.586427958830367e-06,
+      "loss": 0.7045,
+      "step": 2588
+    },
+    {
+      "epoch": 0.21026557297165596,
+      "grad_norm": 3.905668647886352,
+      "learning_rate": 4.586065602516182e-06,
+      "loss": 0.5548,
+      "step": 2589
+    },
+    {
+      "epoch": 0.21034678794769757,
+      "grad_norm": 5.834395303055314,
+      "learning_rate": 4.585703101857298e-06,
+      "loss": 0.4948,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21042800292373914,
+      "grad_norm": 5.414482998853449,
+      "learning_rate": 4.585340456878798e-06,
+      "loss": 0.4951,
+      "step": 2591
+    },
+    {
+      "epoch": 0.21050921789978072,
+      "grad_norm": 3.1815206113624535,
+      "learning_rate": 4.584977667605774e-06,
+      "loss": 0.5363,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2105904328758223,
+      "grad_norm": 3.1419631204303915,
+      "learning_rate": 4.5846147340633305e-06,
+      "loss": 0.5075,
+      "step": 2593
+    },
+    {
+      "epoch": 0.21067164785186387,
+      "grad_norm": 3.5141257327029205,
+      "learning_rate": 4.58425165627658e-06,
+      "loss": 0.543,
+      "step": 2594
+    },
+    {
+      "epoch": 0.21075286282790548,
+      "grad_norm": 4.156860647624618,
+      "learning_rate": 4.583888434270645e-06,
+      "loss": 0.5039,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21083407780394706,
+      "grad_norm": 4.750813630585856,
+      "learning_rate": 4.58352506807066e-06,
+      "loss": 0.5138,
+      "step": 2596
+    },
+    {
+      "epoch": 0.21091529277998863,
+      "grad_norm": 9.312381581764441,
+      "learning_rate": 4.583161557701767e-06,
+      "loss": 0.5763,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2109965077560302,
+      "grad_norm": 8.915059751168238,
+      "learning_rate": 4.582797903189119e-06,
+      "loss": 0.5809,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2110777227320718,
+      "grad_norm": 3.648070795784496,
+      "learning_rate": 4.582434104557879e-06,
+      "loss": 0.482,
+      "step": 2599
+    },
+    {
+      "epoch": 0.21115893770811336,
+      "grad_norm": 4.2519549882216845,
+      "learning_rate": 4.582070161833221e-06,
+      "loss": 0.5429,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21124015268415497,
+      "grad_norm": 4.546548576906945,
+      "learning_rate": 4.581706075040326e-06,
+      "loss": 0.5398,
+      "step": 2601
+    },
+    {
+      "epoch": 0.21132136766019655,
+      "grad_norm": 5.001987184534323,
+      "learning_rate": 4.5813418442043885e-06,
+      "loss": 0.5318,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21140258263623812,
+      "grad_norm": 6.362483259653969,
+      "learning_rate": 4.58097746935061e-06,
+      "loss": 0.5363,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2114837976122797,
+      "grad_norm": 4.0989340540279064,
+      "learning_rate": 4.580612950504204e-06,
+      "loss": 0.5545,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21156501258832128,
+      "grad_norm": 5.132444357319185,
+      "learning_rate": 4.580248287690394e-06,
+      "loss": 0.4948,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21164622756436288,
+      "grad_norm": 7.374853871968238,
+      "learning_rate": 4.579883480934413e-06,
+      "loss": 0.5281,
+      "step": 2606
+    },
+    {
+      "epoch": 0.21172744254040446,
+      "grad_norm": 8.014579215103465,
+      "learning_rate": 4.579518530261501e-06,
+      "loss": 0.6517,
+      "step": 2607
+    },
+    {
+      "epoch": 0.21180865751644604,
+      "grad_norm": 4.7524183309310315,
+      "learning_rate": 4.579153435696913e-06,
+      "loss": 0.5649,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2118898724924876,
+      "grad_norm": 3.2946505830547697,
+      "learning_rate": 4.578788197265911e-06,
+      "loss": 0.4996,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2119710874685292,
+      "grad_norm": 4.2400296703901645,
+      "learning_rate": 4.578422814993768e-06,
+      "loss": 0.5894,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21205230244457077,
+      "grad_norm": 5.74324372003261,
+      "learning_rate": 4.578057288905766e-06,
+      "loss": 0.6345,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21213351742061237,
+      "grad_norm": 3.66763262733531,
+      "learning_rate": 4.577691619027197e-06,
+      "loss": 0.5196,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21221473239665395,
+      "grad_norm": 7.157752603621458,
+      "learning_rate": 4.577325805383364e-06,
+      "loss": 0.4639,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21229594737269553,
+      "grad_norm": 6.775920682174733,
+      "learning_rate": 4.57695984799958e-06,
+      "loss": 0.5971,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2123771623487371,
+      "grad_norm": 5.5027350147320275,
+      "learning_rate": 4.576593746901166e-06,
+      "loss": 0.6259,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21245837732477868,
+      "grad_norm": 4.893766528880051,
+      "learning_rate": 4.576227502113455e-06,
+      "loss": 0.5943,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21253959230082028,
+      "grad_norm": 4.639805956669454,
+      "learning_rate": 4.575861113661791e-06,
+      "loss": 0.7361,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21262080727686186,
+      "grad_norm": 4.20860766014754,
+      "learning_rate": 4.575494581571521e-06,
+      "loss": 0.3826,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21270202225290344,
+      "grad_norm": 5.655888239003785,
+      "learning_rate": 4.575127905868013e-06,
+      "loss": 0.5891,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21278323722894502,
+      "grad_norm": 5.032532015247516,
+      "learning_rate": 4.574761086576635e-06,
+      "loss": 0.4914,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2128644522049866,
+      "grad_norm": 11.837098472717457,
+      "learning_rate": 4.57439412372277e-06,
+      "loss": 0.4658,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21294566718102817,
+      "grad_norm": 14.561447419694415,
+      "learning_rate": 4.574027017331812e-06,
+      "loss": 0.5611,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21302688215706977,
+      "grad_norm": 3.633636275262492,
+      "learning_rate": 4.57365976742916e-06,
+      "loss": 0.5933,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21310809713311135,
+      "grad_norm": 6.065413078817048,
+      "learning_rate": 4.573292374040227e-06,
+      "loss": 0.5984,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21318931210915293,
+      "grad_norm": 12.083369823179202,
+      "learning_rate": 4.572924837190434e-06,
+      "loss": 0.4037,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2132705270851945,
+      "grad_norm": 4.512701432450698,
+      "learning_rate": 4.572557156905213e-06,
+      "loss": 0.6599,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21335174206123608,
+      "grad_norm": 4.466963402457013,
+      "learning_rate": 4.572189333210007e-06,
+      "loss": 0.6408,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2134329570372777,
+      "grad_norm": 4.5137793929351115,
+      "learning_rate": 4.571821366130265e-06,
+      "loss": 0.4764,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21351417201331926,
+      "grad_norm": 3.8656056596236184,
+      "learning_rate": 4.571453255691449e-06,
+      "loss": 0.4507,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21359538698936084,
+      "grad_norm": 4.382341801906267,
+      "learning_rate": 4.571085001919031e-06,
+      "loss": 0.6934,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21367660196540242,
+      "grad_norm": 8.589875710292333,
+      "learning_rate": 4.570716604838492e-06,
+      "loss": 0.6249,
+      "step": 2631
+    },
+    {
+      "epoch": 0.213757816941444,
+      "grad_norm": 4.007630827099167,
+      "learning_rate": 4.570348064475323e-06,
+      "loss": 0.6744,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21383903191748557,
+      "grad_norm": 5.484359947965143,
+      "learning_rate": 4.569979380855025e-06,
+      "loss": 0.5493,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21392024689352718,
+      "grad_norm": 4.493010697403236,
+      "learning_rate": 4.56961055400311e-06,
+      "loss": 0.5519,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21400146186956875,
+      "grad_norm": 4.045159603215263,
+      "learning_rate": 4.5692415839450965e-06,
+      "loss": 0.59,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21408267684561033,
+      "grad_norm": 4.821023174942057,
+      "learning_rate": 4.568872470706518e-06,
+      "loss": 0.3977,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2141638918216519,
+      "grad_norm": 5.675194155783661,
+      "learning_rate": 4.568503214312913e-06,
+      "loss": 0.5558,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21424510679769349,
+      "grad_norm": 4.7601824577948015,
+      "learning_rate": 4.568133814789833e-06,
+      "loss": 0.5638,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2143263217737351,
+      "grad_norm": 5.107493849030825,
+      "learning_rate": 4.567764272162839e-06,
+      "loss": 0.5493,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21440753674977667,
+      "grad_norm": 4.8431041110711766,
+      "learning_rate": 4.567394586457501e-06,
+      "loss": 0.5961,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21448875172581824,
+      "grad_norm": 6.24655286207524,
+      "learning_rate": 4.567024757699399e-06,
+      "loss": 0.5455,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21456996670185982,
+      "grad_norm": 4.819548308107245,
+      "learning_rate": 4.566654785914123e-06,
+      "loss": 0.5451,
+      "step": 2642
+    },
+    {
+      "epoch": 0.2146511816779014,
+      "grad_norm": 8.148594207313314,
+      "learning_rate": 4.566284671127273e-06,
+      "loss": 0.4268,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21473239665394298,
+      "grad_norm": 5.200028510658096,
+      "learning_rate": 4.56591441336446e-06,
+      "loss": 0.5667,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21481361162998458,
+      "grad_norm": 3.953005099419873,
+      "learning_rate": 4.565544012651304e-06,
+      "loss": 0.9163,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21489482660602616,
+      "grad_norm": 6.166119347251547,
+      "learning_rate": 4.565173469013432e-06,
+      "loss": 0.4929,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21497604158206773,
+      "grad_norm": 17.196433928853956,
+      "learning_rate": 4.564802782476487e-06,
+      "loss": 0.5905,
+      "step": 2647
+    },
+    {
+      "epoch": 0.2150572565581093,
+      "grad_norm": 6.178896351260531,
+      "learning_rate": 4.564431953066118e-06,
+      "loss": 0.7147,
+      "step": 2648
+    },
+    {
+      "epoch": 0.2151384715341509,
+      "grad_norm": 6.440122516675341,
+      "learning_rate": 4.564060980807983e-06,
+      "loss": 0.7101,
+      "step": 2649
+    },
+    {
+      "epoch": 0.2152196865101925,
+      "grad_norm": 5.2000555006994125,
+      "learning_rate": 4.563689865727752e-06,
+      "loss": 0.6096,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21530090148623407,
+      "grad_norm": 4.6419662478035155,
+      "learning_rate": 4.563318607851104e-06,
+      "loss": 0.5928,
+      "step": 2651
+    },
+    {
+      "epoch": 0.21538211646227565,
+      "grad_norm": 8.141047545711643,
+      "learning_rate": 4.562947207203728e-06,
+      "loss": 0.4954,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21546333143831722,
+      "grad_norm": 3.9486721093318153,
+      "learning_rate": 4.562575663811324e-06,
+      "loss": 0.6568,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2155445464143588,
+      "grad_norm": 5.149290757056804,
+      "learning_rate": 4.5622039776996006e-06,
+      "loss": 0.5941,
+      "step": 2654
+    },
+    {
+      "epoch": 0.21562576139040038,
+      "grad_norm": 6.391781782039198,
+      "learning_rate": 4.561832148894275e-06,
+      "loss": 0.4204,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21570697636644198,
+      "grad_norm": 9.968941914044036,
+      "learning_rate": 4.561460177421078e-06,
+      "loss": 0.5599,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21578819134248356,
+      "grad_norm": 5.522046324373681,
+      "learning_rate": 4.561088063305745e-06,
+      "loss": 0.6469,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21586940631852514,
+      "grad_norm": 3.8464087079545277,
+      "learning_rate": 4.560715806574028e-06,
+      "loss": 0.6275,
+      "step": 2658
+    },
+    {
+      "epoch": 0.2159506212945667,
+      "grad_norm": 5.078649015582283,
+      "learning_rate": 4.560343407251682e-06,
+      "loss": 0.5032,
+      "step": 2659
+    },
+    {
+      "epoch": 0.2160318362706083,
+      "grad_norm": 4.334473048243304,
+      "learning_rate": 4.559970865364477e-06,
+      "loss": 0.6853,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2161130512466499,
+      "grad_norm": 6.419291069877329,
+      "learning_rate": 4.55959818093819e-06,
+      "loss": 0.5029,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21619426622269147,
+      "grad_norm": 5.8025560914497385,
+      "learning_rate": 4.559225353998609e-06,
+      "loss": 0.4812,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21627548119873305,
+      "grad_norm": 4.30038594406105,
+      "learning_rate": 4.558852384571533e-06,
+      "loss": 0.5198,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21635669617477463,
+      "grad_norm": 4.101496826162754,
+      "learning_rate": 4.558479272682768e-06,
+      "loss": 0.6267,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2164379111508162,
+      "grad_norm": 5.736035126104208,
+      "learning_rate": 4.558106018358131e-06,
+      "loss": 0.5187,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21651912612685778,
+      "grad_norm": 5.345843996754261,
+      "learning_rate": 4.557732621623449e-06,
+      "loss": 0.5716,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21660034110289939,
+      "grad_norm": 4.853630853027857,
+      "learning_rate": 4.557359082504562e-06,
+      "loss": 0.4583,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21668155607894096,
+      "grad_norm": 3.473558582520133,
+      "learning_rate": 4.556985401027314e-06,
+      "loss": 0.4872,
+      "step": 2668
+    },
+    {
+      "epoch": 0.21676277105498254,
+      "grad_norm": 6.011752197100141,
+      "learning_rate": 4.556611577217563e-06,
+      "loss": 0.5814,
+      "step": 2669
+    },
+    {
+      "epoch": 0.21684398603102412,
+      "grad_norm": 4.738766968848329,
+      "learning_rate": 4.5562376111011745e-06,
+      "loss": 0.4709,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2169252010070657,
+      "grad_norm": 9.10744576822217,
+      "learning_rate": 4.5558635027040265e-06,
+      "loss": 0.542,
+      "step": 2671
+    },
+    {
+      "epoch": 0.2170064159831073,
+      "grad_norm": 3.811056086790161,
+      "learning_rate": 4.555489252052005e-06,
+      "loss": 0.6779,
+      "step": 2672
+    },
+    {
+      "epoch": 0.21708763095914888,
+      "grad_norm": 5.148351495523538,
+      "learning_rate": 4.5551148591710045e-06,
+      "loss": 0.491,
+      "step": 2673
+    },
+    {
+      "epoch": 0.21716884593519045,
+      "grad_norm": 5.581434436680359,
+      "learning_rate": 4.5547403240869335e-06,
+      "loss": 0.6449,
+      "step": 2674
+    },
+    {
+      "epoch": 0.21725006091123203,
+      "grad_norm": 3.742718625839809,
+      "learning_rate": 4.554365646825706e-06,
+      "loss": 0.7014,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2173312758872736,
+      "grad_norm": 6.612469532237822,
+      "learning_rate": 4.5539908274132485e-06,
+      "loss": 0.596,
+      "step": 2676
+    },
+    {
+      "epoch": 0.21741249086331518,
+      "grad_norm": 4.306520677998056,
+      "learning_rate": 4.553615865875496e-06,
+      "loss": 0.7381,
+      "step": 2677
+    },
+    {
+      "epoch": 0.2174937058393568,
+      "grad_norm": 4.614381185208202,
+      "learning_rate": 4.553240762238394e-06,
+      "loss": 0.6964,
+      "step": 2678
+    },
+    {
+      "epoch": 0.21757492081539836,
+      "grad_norm": 7.406343400623716,
+      "learning_rate": 4.552865516527899e-06,
+      "loss": 0.5004,
+      "step": 2679
+    },
+    {
+      "epoch": 0.21765613579143994,
+      "grad_norm": 3.9606433285981266,
+      "learning_rate": 4.552490128769975e-06,
+      "loss": 0.6269,
+      "step": 2680
+    },
+    {
+      "epoch": 0.21773735076748152,
+      "grad_norm": 4.11495963139638,
+      "learning_rate": 4.5521145989905955e-06,
+      "loss": 0.5763,
+      "step": 2681
+    },
+    {
+      "epoch": 0.2178185657435231,
+      "grad_norm": 4.588404789142679,
+      "learning_rate": 4.551738927215747e-06,
+      "loss": 0.4674,
+      "step": 2682
+    },
+    {
+      "epoch": 0.2178997807195647,
+      "grad_norm": 7.297594706002203,
+      "learning_rate": 4.5513631134714235e-06,
+      "loss": 0.5801,
+      "step": 2683
+    },
+    {
+      "epoch": 0.21798099569560628,
+      "grad_norm": 4.3047090539293045,
+      "learning_rate": 4.550987157783629e-06,
+      "loss": 0.4813,
+      "step": 2684
+    },
+    {
+      "epoch": 0.21806221067164785,
+      "grad_norm": 5.768465577812249,
+      "learning_rate": 4.550611060178378e-06,
+      "loss": 0.567,
+      "step": 2685
+    },
+    {
+      "epoch": 0.21814342564768943,
+      "grad_norm": 5.651861901253761,
+      "learning_rate": 4.550234820681695e-06,
+      "loss": 0.6873,
+      "step": 2686
+    },
+    {
+      "epoch": 0.218224640623731,
+      "grad_norm": 9.183694716367905,
+      "learning_rate": 4.549858439319612e-06,
+      "loss": 0.6324,
+      "step": 2687
+    },
+    {
+      "epoch": 0.21830585559977259,
+      "grad_norm": 7.585915804173281,
+      "learning_rate": 4.549481916118174e-06,
+      "loss": 0.5873,
+      "step": 2688
+    },
+    {
+      "epoch": 0.2183870705758142,
+      "grad_norm": 3.760740144941256,
+      "learning_rate": 4.5491052511034345e-06,
+      "loss": 0.6382,
+      "step": 2689
+    },
+    {
+      "epoch": 0.21846828555185577,
+      "grad_norm": 4.945745347505512,
+      "learning_rate": 4.548728444301456e-06,
+      "loss": 0.5818,
+      "step": 2690
+    },
+    {
+      "epoch": 0.21854950052789734,
+      "grad_norm": 4.9769795576864215,
+      "learning_rate": 4.548351495738312e-06,
+      "loss": 0.7123,
+      "step": 2691
+    },
+    {
+      "epoch": 0.21863071550393892,
+      "grad_norm": 4.481119259930459,
+      "learning_rate": 4.547974405440085e-06,
+      "loss": 0.4976,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2187119304799805,
+      "grad_norm": 5.7374511609884955,
+      "learning_rate": 4.547597173432869e-06,
+      "loss": 0.5487,
+      "step": 2693
+    },
+    {
+      "epoch": 0.2187931454560221,
+      "grad_norm": 5.502760303179846,
+      "learning_rate": 4.547219799742765e-06,
+      "loss": 0.7687,
+      "step": 2694
+    },
+    {
+      "epoch": 0.21887436043206368,
+      "grad_norm": 10.421957536535567,
+      "learning_rate": 4.5468422843958845e-06,
+      "loss": 0.5052,
+      "step": 2695
+    },
+    {
+      "epoch": 0.21895557540810526,
+      "grad_norm": 3.6707221757502917,
+      "learning_rate": 4.546464627418351e-06,
+      "loss": 0.5888,
+      "step": 2696
+    },
+    {
+      "epoch": 0.21903679038414683,
+      "grad_norm": 3.304249792857001,
+      "learning_rate": 4.546086828836297e-06,
+      "loss": 0.6277,
+      "step": 2697
+    },
+    {
+      "epoch": 0.2191180053601884,
+      "grad_norm": 4.76062394973188,
+      "learning_rate": 4.545708888675862e-06,
+      "loss": 0.6074,
+      "step": 2698
+    },
+    {
+      "epoch": 0.21919922033623,
+      "grad_norm": 4.462680720514244,
+      "learning_rate": 4.5453308069632e-06,
+      "loss": 0.5367,
+      "step": 2699
+    },
+    {
+      "epoch": 0.2192804353122716,
+      "grad_norm": 7.372740705732592,
+      "learning_rate": 4.54495258372447e-06,
+      "loss": 0.5074,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21936165028831317,
+      "grad_norm": 7.860054087415267,
+      "learning_rate": 4.544574218985845e-06,
+      "loss": 0.4761,
+      "step": 2701
+    },
+    {
+      "epoch": 0.21944286526435475,
+      "grad_norm": 8.081036919499534,
+      "learning_rate": 4.544195712773504e-06,
+      "loss": 0.5659,
+      "step": 2702
+    },
+    {
+      "epoch": 0.21952408024039632,
+      "grad_norm": 5.8269057668228985,
+      "learning_rate": 4.543817065113638e-06,
+      "loss": 0.462,
+      "step": 2703
+    },
+    {
+      "epoch": 0.2196052952164379,
+      "grad_norm": 4.2668146590202305,
+      "learning_rate": 4.543438276032448e-06,
+      "loss": 0.4101,
+      "step": 2704
+    },
+    {
+      "epoch": 0.2196865101924795,
+      "grad_norm": 5.140738961702395,
+      "learning_rate": 4.543059345556145e-06,
+      "loss": 0.48,
+      "step": 2705
+    },
+    {
+      "epoch": 0.21976772516852108,
+      "grad_norm": 5.170374218467899,
+      "learning_rate": 4.542680273710947e-06,
+      "loss": 0.5453,
+      "step": 2706
+    },
+    {
+      "epoch": 0.21984894014456266,
+      "grad_norm": 5.753753477369303,
+      "learning_rate": 4.542301060523086e-06,
+      "loss": 0.5962,
+      "step": 2707
+    },
+    {
+      "epoch": 0.21993015512060424,
+      "grad_norm": 4.532213136821538,
+      "learning_rate": 4.541921706018799e-06,
+      "loss": 0.561,
+      "step": 2708
+    },
+    {
+      "epoch": 0.22001137009664581,
+      "grad_norm": 5.5635492887810125,
+      "learning_rate": 4.541542210224337e-06,
+      "loss": 0.6049,
+      "step": 2709
+    },
+    {
+      "epoch": 0.2200925850726874,
+      "grad_norm": 4.789661183080903,
+      "learning_rate": 4.5411625731659595e-06,
+      "loss": 0.5815,
+      "step": 2710
+    },
+    {
+      "epoch": 0.220173800048729,
+      "grad_norm": 3.231674262508777,
+      "learning_rate": 4.540782794869933e-06,
+      "loss": 0.5184,
+      "step": 2711
+    },
+    {
+      "epoch": 0.22025501502477057,
+      "grad_norm": 6.341088144720809,
+      "learning_rate": 4.5404028753625396e-06,
+      "loss": 0.4839,
+      "step": 2712
+    },
+    {
+      "epoch": 0.22033623000081215,
+      "grad_norm": 4.816690291787395,
+      "learning_rate": 4.5400228146700654e-06,
+      "loss": 0.6509,
+      "step": 2713
+    },
+    {
+      "epoch": 0.22041744497685373,
+      "grad_norm": 3.6214196801416474,
+      "learning_rate": 4.539642612818809e-06,
+      "loss": 0.627,
+      "step": 2714
+    },
+    {
+      "epoch": 0.2204986599528953,
+      "grad_norm": 7.585870445781305,
+      "learning_rate": 4.539262269835078e-06,
+      "loss": 0.4846,
+      "step": 2715
+    },
+    {
+      "epoch": 0.2205798749289369,
+      "grad_norm": 5.185206149837472,
+      "learning_rate": 4.538881785745191e-06,
+      "loss": 0.5169,
+      "step": 2716
+    },
+    {
+      "epoch": 0.22066108990497849,
+      "grad_norm": 9.120252331770077,
+      "learning_rate": 4.538501160575475e-06,
+      "loss": 0.5466,
+      "step": 2717
+    },
+    {
+      "epoch": 0.22074230488102006,
+      "grad_norm": 4.271633224077144,
+      "learning_rate": 4.538120394352267e-06,
+      "loss": 0.5654,
+      "step": 2718
+    },
+    {
+      "epoch": 0.22082351985706164,
+      "grad_norm": 6.425265204493728,
+      "learning_rate": 4.5377394871019145e-06,
+      "loss": 0.5984,
+      "step": 2719
+    },
+    {
+      "epoch": 0.22090473483310322,
+      "grad_norm": 5.662199724686198,
+      "learning_rate": 4.5373584388507745e-06,
+      "loss": 0.5098,
+      "step": 2720
+    },
+    {
+      "epoch": 0.2209859498091448,
+      "grad_norm": 7.194165344057847,
+      "learning_rate": 4.536977249625213e-06,
+      "loss": 0.49,
+      "step": 2721
+    },
+    {
+      "epoch": 0.2210671647851864,
+      "grad_norm": 6.981685701050804,
+      "learning_rate": 4.536595919451606e-06,
+      "loss": 0.6383,
+      "step": 2722
+    },
+    {
+      "epoch": 0.22114837976122798,
+      "grad_norm": 4.116671598914044,
+      "learning_rate": 4.53621444835634e-06,
+      "loss": 0.5659,
+      "step": 2723
+    },
+    {
+      "epoch": 0.22122959473726955,
+      "grad_norm": 5.134961603692183,
+      "learning_rate": 4.535832836365811e-06,
+      "loss": 0.4805,
+      "step": 2724
+    },
+    {
+      "epoch": 0.22131080971331113,
+      "grad_norm": 4.576284219113115,
+      "learning_rate": 4.535451083506424e-06,
+      "loss": 0.7364,
+      "step": 2725
+    },
+    {
+      "epoch": 0.2213920246893527,
+      "grad_norm": 6.0501880740602365,
+      "learning_rate": 4.535069189804594e-06,
+      "loss": 0.5578,
+      "step": 2726
+    },
+    {
+      "epoch": 0.2214732396653943,
+      "grad_norm": 4.658417802807258,
+      "learning_rate": 4.534687155286747e-06,
+      "loss": 0.5017,
+      "step": 2727
+    },
+    {
+      "epoch": 0.2215544546414359,
+      "grad_norm": 3.5188270873838117,
+      "learning_rate": 4.534304979979317e-06,
+      "loss": 0.5166,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22163566961747747,
+      "grad_norm": 3.6771178180337865,
+      "learning_rate": 4.53392266390875e-06,
+      "loss": 0.717,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22171688459351904,
+      "grad_norm": 4.874076485337167,
+      "learning_rate": 4.533540207101498e-06,
+      "loss": 0.676,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22179809956956062,
+      "grad_norm": 3.3053339197225307,
+      "learning_rate": 4.533157609584026e-06,
+      "loss": 0.7047,
+      "step": 2731
+    },
+    {
+      "epoch": 0.2218793145456022,
+      "grad_norm": 4.0334080892398845,
+      "learning_rate": 4.532774871382807e-06,
+      "loss": 0.6683,
+      "step": 2732
+    },
+    {
+      "epoch": 0.2219605295216438,
+      "grad_norm": 7.9679246058681885,
+      "learning_rate": 4.532391992524327e-06,
+      "loss": 0.5633,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22204174449768538,
+      "grad_norm": 5.020060727741971,
+      "learning_rate": 4.532008973035076e-06,
+      "loss": 0.6868,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22212295947372696,
+      "grad_norm": 3.77106126429378,
+      "learning_rate": 4.531625812941559e-06,
+      "loss": 0.5032,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22220417444976853,
+      "grad_norm": 3.682569403476453,
+      "learning_rate": 4.531242512270287e-06,
+      "loss": 0.7004,
+      "step": 2736
+    },
+    {
+      "epoch": 0.2222853894258101,
+      "grad_norm": 4.487348592439573,
+      "learning_rate": 4.530859071047785e-06,
+      "loss": 0.5239,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22236660440185171,
+      "grad_norm": 4.188000337446031,
+      "learning_rate": 4.530475489300583e-06,
+      "loss": 0.4732,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2224478193778933,
+      "grad_norm": 8.498349813305607,
+      "learning_rate": 4.530091767055223e-06,
+      "loss": 0.4986,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22252903435393487,
+      "grad_norm": 5.52528756591483,
+      "learning_rate": 4.5297079043382566e-06,
+      "loss": 0.6785,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22261024932997645,
+      "grad_norm": 5.752497762609521,
+      "learning_rate": 4.529323901176245e-06,
+      "loss": 0.4531,
+      "step": 2741
+    },
+    {
+      "epoch": 0.22269146430601802,
+      "grad_norm": 3.092790657666851,
+      "learning_rate": 4.52893975759576e-06,
+      "loss": 0.7052,
+      "step": 2742
+    },
+    {
+      "epoch": 0.2227726792820596,
+      "grad_norm": 4.559669547606855,
+      "learning_rate": 4.528555473623381e-06,
+      "loss": 0.5464,
+      "step": 2743
+    },
+    {
+      "epoch": 0.2228538942581012,
+      "grad_norm": 3.5783978557136127,
+      "learning_rate": 4.5281710492857e-06,
+      "loss": 0.6876,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22293510923414278,
+      "grad_norm": 4.328813594256833,
+      "learning_rate": 4.527786484609316e-06,
+      "loss": 0.564,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22301632421018436,
+      "grad_norm": 3.952478916018802,
+      "learning_rate": 4.52740177962084e-06,
+      "loss": 0.5783,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22309753918622593,
+      "grad_norm": 4.924032799834654,
+      "learning_rate": 4.52701693434689e-06,
+      "loss": 0.7729,
+      "step": 2747
+    },
+    {
+      "epoch": 0.2231787541622675,
+      "grad_norm": 5.139762166088582,
+      "learning_rate": 4.526631948814096e-06,
+      "loss": 0.5408,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22325996913830912,
+      "grad_norm": 4.468079919633165,
+      "learning_rate": 4.5262468230490975e-06,
+      "loss": 0.6876,
+      "step": 2749
+    },
+    {
+      "epoch": 0.2233411841143507,
+      "grad_norm": 4.463738118530915,
+      "learning_rate": 4.525861557078542e-06,
+      "loss": 0.7465,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22342239909039227,
+      "grad_norm": 7.852415705301734,
+      "learning_rate": 4.525476150929089e-06,
+      "loss": 0.5134,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22350361406643385,
+      "grad_norm": 4.963966987188751,
+      "learning_rate": 4.525090604627406e-06,
+      "loss": 0.5476,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22358482904247542,
+      "grad_norm": 6.07590271752492,
+      "learning_rate": 4.52470491820017e-06,
+      "loss": 0.4523,
+      "step": 2753
+    },
+    {
+      "epoch": 0.223666044018517,
+      "grad_norm": 4.302538032685536,
+      "learning_rate": 4.52431909167407e-06,
+      "loss": 0.6684,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2237472589945586,
+      "grad_norm": 4.256570317991106,
+      "learning_rate": 4.5239331250758025e-06,
+      "loss": 0.5804,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22382847397060018,
+      "grad_norm": 3.987772037631392,
+      "learning_rate": 4.523547018432074e-06,
+      "loss": 0.5361,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22390968894664176,
+      "grad_norm": 4.25893742048742,
+      "learning_rate": 4.523160771769602e-06,
+      "loss": 0.5403,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22399090392268334,
+      "grad_norm": 4.754148515001833,
+      "learning_rate": 4.52277438511511e-06,
+      "loss": 0.6913,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22407211889872491,
+      "grad_norm": 11.158110474612203,
+      "learning_rate": 4.522387858495337e-06,
+      "loss": 0.4877,
+      "step": 2759
+    },
+    {
+      "epoch": 0.22415333387476652,
+      "grad_norm": 5.122504334003654,
+      "learning_rate": 4.522001191937028e-06,
+      "loss": 0.4932,
+      "step": 2760
+    },
+    {
+      "epoch": 0.2242345488508081,
+      "grad_norm": 4.861511116519465,
+      "learning_rate": 4.521614385466938e-06,
+      "loss": 0.5527,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22431576382684967,
+      "grad_norm": 3.6596485107845003,
+      "learning_rate": 4.521227439111831e-06,
+      "loss": 0.7121,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22439697880289125,
+      "grad_norm": 4.522445266045689,
+      "learning_rate": 4.520840352898483e-06,
+      "loss": 0.5672,
+      "step": 2763
+    },
+    {
+      "epoch": 0.22447819377893283,
+      "grad_norm": 7.737848493782743,
+      "learning_rate": 4.520453126853677e-06,
+      "loss": 0.5862,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2245594087549744,
+      "grad_norm": 6.141302741029408,
+      "learning_rate": 4.520065761004209e-06,
+      "loss": 0.5703,
+      "step": 2765
+    },
+    {
+      "epoch": 0.224640623731016,
+      "grad_norm": 6.569966061095918,
+      "learning_rate": 4.51967825537688e-06,
+      "loss": 0.5038,
+      "step": 2766
+    },
+    {
+      "epoch": 0.2247218387070576,
+      "grad_norm": 5.8858674844088235,
+      "learning_rate": 4.5192906099985055e-06,
+      "loss": 0.5216,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22480305368309916,
+      "grad_norm": 4.116267752952551,
+      "learning_rate": 4.518902824895908e-06,
+      "loss": 0.4604,
+      "step": 2768
+    },
+    {
+      "epoch": 0.22488426865914074,
+      "grad_norm": 6.293436608006783,
+      "learning_rate": 4.518514900095919e-06,
+      "loss": 0.466,
+      "step": 2769
+    },
+    {
+      "epoch": 0.22496548363518232,
+      "grad_norm": 8.339236282987327,
+      "learning_rate": 4.518126835625382e-06,
+      "loss": 0.5357,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22504669861122392,
+      "grad_norm": 4.726236800961924,
+      "learning_rate": 4.51773863151115e-06,
+      "loss": 0.6255,
+      "step": 2771
+    },
+    {
+      "epoch": 0.2251279135872655,
+      "grad_norm": 4.1928620515740365,
+      "learning_rate": 4.517350287780081e-06,
+      "loss": 0.553,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22520912856330708,
+      "grad_norm": 6.445972691759003,
+      "learning_rate": 4.51696180445905e-06,
+      "loss": 0.4875,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22529034353934865,
+      "grad_norm": 14.794833956873413,
+      "learning_rate": 4.516573181574937e-06,
+      "loss": 0.5604,
+      "step": 2774
+    },
+    {
+      "epoch": 0.22537155851539023,
+      "grad_norm": 5.562154003766993,
+      "learning_rate": 4.516184419154633e-06,
+      "loss": 0.5572,
+      "step": 2775
+    },
+    {
+      "epoch": 0.2254527734914318,
+      "grad_norm": 3.647681045658577,
+      "learning_rate": 4.515795517225037e-06,
+      "loss": 0.617,
+      "step": 2776
+    },
+    {
+      "epoch": 0.2255339884674734,
+      "grad_norm": 5.695037511735684,
+      "learning_rate": 4.51540647581306e-06,
+      "loss": 0.412,
+      "step": 2777
+    },
+    {
+      "epoch": 0.225615203443515,
+      "grad_norm": 4.475228081800305,
+      "learning_rate": 4.51501729494562e-06,
+      "loss": 0.6703,
+      "step": 2778
+    },
+    {
+      "epoch": 0.22569641841955657,
+      "grad_norm": 4.611656081080196,
+      "learning_rate": 4.514627974649649e-06,
+      "loss": 0.6964,
+      "step": 2779
+    },
+    {
+      "epoch": 0.22577763339559814,
+      "grad_norm": 3.968385112319059,
+      "learning_rate": 4.514238514952084e-06,
+      "loss": 0.486,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22585884837163972,
+      "grad_norm": 4.069304366874197,
+      "learning_rate": 4.513848915879874e-06,
+      "loss": 0.501,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22594006334768132,
+      "grad_norm": 5.496815740257597,
+      "learning_rate": 4.513459177459977e-06,
+      "loss": 0.6377,
+      "step": 2782
+    },
+    {
+      "epoch": 0.2260212783237229,
+      "grad_norm": 6.857579000068127,
+      "learning_rate": 4.513069299719361e-06,
+      "loss": 0.5332,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22610249329976448,
+      "grad_norm": 3.9422493886680354,
+      "learning_rate": 4.512679282685003e-06,
+      "loss": 0.7389,
+      "step": 2784
+    },
+    {
+      "epoch": 0.22618370827580606,
+      "grad_norm": 3.4503956109068135,
+      "learning_rate": 4.512289126383892e-06,
+      "loss": 0.4416,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22626492325184763,
+      "grad_norm": 4.6592622473563665,
+      "learning_rate": 4.511898830843022e-06,
+      "loss": 0.5942,
+      "step": 2786
+    },
+    {
+      "epoch": 0.2263461382278892,
+      "grad_norm": 5.905176155853469,
+      "learning_rate": 4.511508396089401e-06,
+      "loss": 0.5971,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22642735320393081,
+      "grad_norm": 3.3680130138338744,
+      "learning_rate": 4.5111178221500455e-06,
+      "loss": 0.5056,
+      "step": 2788
+    },
+    {
+      "epoch": 0.2265085681799724,
+      "grad_norm": 3.6474870201796006,
+      "learning_rate": 4.51072710905198e-06,
+      "loss": 0.623,
+      "step": 2789
+    },
+    {
+      "epoch": 0.22658978315601397,
+      "grad_norm": 5.380202578661765,
+      "learning_rate": 4.5103362568222395e-06,
+      "loss": 0.5094,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22667099813205555,
+      "grad_norm": 3.51083847592959,
+      "learning_rate": 4.509945265487871e-06,
+      "loss": 0.5929,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22675221310809712,
+      "grad_norm": 4.91331163188237,
+      "learning_rate": 4.5095541350759265e-06,
+      "loss": 0.5545,
+      "step": 2792
+    },
+    {
+      "epoch": 0.22683342808413873,
+      "grad_norm": 3.869643139237871,
+      "learning_rate": 4.5091628656134715e-06,
+      "loss": 0.5104,
+      "step": 2793
+    },
+    {
+      "epoch": 0.2269146430601803,
+      "grad_norm": 11.289546788075414,
+      "learning_rate": 4.508771457127579e-06,
+      "loss": 0.4783,
+      "step": 2794
+    },
+    {
+      "epoch": 0.22699585803622188,
+      "grad_norm": 4.701834247380548,
+      "learning_rate": 4.508379909645334e-06,
+      "loss": 0.6242,
+      "step": 2795
+    },
+    {
+      "epoch": 0.22707707301226346,
+      "grad_norm": 3.8081500149721608,
+      "learning_rate": 4.5079882231938274e-06,
+      "loss": 0.6682,
+      "step": 2796
+    },
+    {
+      "epoch": 0.22715828798830504,
+      "grad_norm": 8.196015374000737,
+      "learning_rate": 4.5075963978001634e-06,
+      "loss": 0.5618,
+      "step": 2797
+    },
+    {
+      "epoch": 0.2272395029643466,
+      "grad_norm": 3.4819053096272743,
+      "learning_rate": 4.5072044334914546e-06,
+      "loss": 0.4528,
+      "step": 2798
+    },
+    {
+      "epoch": 0.22732071794038822,
+      "grad_norm": 5.156427328355886,
+      "learning_rate": 4.506812330294821e-06,
+      "loss": 0.5095,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2274019329164298,
+      "grad_norm": 3.7634245699967477,
+      "learning_rate": 4.506420088237395e-06,
+      "loss": 0.6707,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22748314789247137,
+      "grad_norm": 3.431587308735926,
+      "learning_rate": 4.5060277073463174e-06,
+      "loss": 0.566,
+      "step": 2801
+    },
+    {
+      "epoch": 0.22756436286851295,
+      "grad_norm": 4.437013694098261,
+      "learning_rate": 4.50563518764874e-06,
+      "loss": 0.7123,
+      "step": 2802
+    },
+    {
+      "epoch": 0.22764557784455453,
+      "grad_norm": 4.567183665178771,
+      "learning_rate": 4.505242529171822e-06,
+      "loss": 0.5152,
+      "step": 2803
+    },
+    {
+      "epoch": 0.22772679282059613,
+      "grad_norm": 3.9017580026940664,
+      "learning_rate": 4.504849731942734e-06,
+      "loss": 0.5201,
+      "step": 2804
+    },
+    {
+      "epoch": 0.2278080077966377,
+      "grad_norm": 4.536121689617816,
+      "learning_rate": 4.504456795988654e-06,
+      "loss": 0.6886,
+      "step": 2805
+    },
+    {
+      "epoch": 0.22788922277267928,
+      "grad_norm": 10.84869509304053,
+      "learning_rate": 4.504063721336773e-06,
+      "loss": 0.5154,
+      "step": 2806
+    },
+    {
+      "epoch": 0.22797043774872086,
+      "grad_norm": 5.404245994389873,
+      "learning_rate": 4.503670508014289e-06,
+      "loss": 0.5609,
+      "step": 2807
+    },
+    {
+      "epoch": 0.22805165272476244,
+      "grad_norm": 6.579796252814421,
+      "learning_rate": 4.50327715604841e-06,
+      "loss": 0.523,
+      "step": 2808
+    },
+    {
+      "epoch": 0.22813286770080402,
+      "grad_norm": 3.973362927458198,
+      "learning_rate": 4.5028836654663535e-06,
+      "loss": 0.5637,
+      "step": 2809
+    },
+    {
+      "epoch": 0.22821408267684562,
+      "grad_norm": 5.793419725047325,
+      "learning_rate": 4.502490036295348e-06,
+      "loss": 0.5813,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2282952976528872,
+      "grad_norm": 4.201946994996218,
+      "learning_rate": 4.50209626856263e-06,
+      "loss": 0.3738,
+      "step": 2811
+    },
+    {
+      "epoch": 0.22837651262892877,
+      "grad_norm": 4.182121020616893,
+      "learning_rate": 4.501702362295446e-06,
+      "loss": 0.5465,
+      "step": 2812
+    },
+    {
+      "epoch": 0.22845772760497035,
+      "grad_norm": 3.5727363113753645,
+      "learning_rate": 4.501308317521052e-06,
+      "loss": 0.5189,
+      "step": 2813
+    },
+    {
+      "epoch": 0.22853894258101193,
+      "grad_norm": 5.356568404800836,
+      "learning_rate": 4.500914134266715e-06,
+      "loss": 0.8021,
+      "step": 2814
+    },
+    {
+      "epoch": 0.22862015755705353,
+      "grad_norm": 4.550631173863838,
+      "learning_rate": 4.500519812559709e-06,
+      "loss": 0.583,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2287013725330951,
+      "grad_norm": 4.328142108387039,
+      "learning_rate": 4.50012535242732e-06,
+      "loss": 0.4888,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2287825875091367,
+      "grad_norm": 4.774451426228126,
+      "learning_rate": 4.499730753896841e-06,
+      "loss": 0.4892,
+      "step": 2817
+    },
+    {
+      "epoch": 0.22886380248517826,
+      "grad_norm": 4.5207191018836355,
+      "learning_rate": 4.4993360169955784e-06,
+      "loss": 0.4906,
+      "step": 2818
+    },
+    {
+      "epoch": 0.22894501746121984,
+      "grad_norm": 3.12349199747274,
+      "learning_rate": 4.498941141750845e-06,
+      "loss": 0.6733,
+      "step": 2819
+    },
+    {
+      "epoch": 0.22902623243726142,
+      "grad_norm": 3.911794379996745,
+      "learning_rate": 4.498546128189963e-06,
+      "loss": 0.5263,
+      "step": 2820
+    },
+    {
+      "epoch": 0.22910744741330302,
+      "grad_norm": 5.139451559263501,
+      "learning_rate": 4.498150976340266e-06,
+      "loss": 0.5766,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2291886623893446,
+      "grad_norm": 4.27868628115041,
+      "learning_rate": 4.497755686229097e-06,
+      "loss": 0.529,
+      "step": 2822
+    },
+    {
+      "epoch": 0.22926987736538618,
+      "grad_norm": 3.9641320918354372,
+      "learning_rate": 4.497360257883808e-06,
+      "loss": 0.5311,
+      "step": 2823
+    },
+    {
+      "epoch": 0.22935109234142775,
+      "grad_norm": 3.5108973371563055,
+      "learning_rate": 4.496964691331759e-06,
+      "loss": 0.5227,
+      "step": 2824
+    },
+    {
+      "epoch": 0.22943230731746933,
+      "grad_norm": 5.403815218671819,
+      "learning_rate": 4.496568986600323e-06,
+      "loss": 0.9042,
+      "step": 2825
+    },
+    {
+      "epoch": 0.22951352229351094,
+      "grad_norm": 4.12742663870129,
+      "learning_rate": 4.4961731437168795e-06,
+      "loss": 0.7359,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2295947372695525,
+      "grad_norm": 3.568114706661449,
+      "learning_rate": 4.4957771627088185e-06,
+      "loss": 0.6484,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2296759522455941,
+      "grad_norm": 5.842995081486275,
+      "learning_rate": 4.495381043603541e-06,
+      "loss": 0.6022,
+      "step": 2828
+    },
+    {
+      "epoch": 0.22975716722163567,
+      "grad_norm": 5.602774830593073,
+      "learning_rate": 4.494984786428455e-06,
+      "loss": 0.6084,
+      "step": 2829
+    },
+    {
+      "epoch": 0.22983838219767724,
+      "grad_norm": 5.779320891496907,
+      "learning_rate": 4.494588391210981e-06,
+      "loss": 0.5428,
+      "step": 2830
+    },
+    {
+      "epoch": 0.22991959717371882,
+      "grad_norm": 5.734619447546111,
+      "learning_rate": 4.494191857978546e-06,
+      "loss": 0.5494,
+      "step": 2831
+    },
+    {
+      "epoch": 0.23000081214976043,
+      "grad_norm": 3.540278577269194,
+      "learning_rate": 4.493795186758589e-06,
+      "loss": 0.6195,
+      "step": 2832
+    },
+    {
+      "epoch": 0.230082027125802,
+      "grad_norm": 5.458025507421001,
+      "learning_rate": 4.493398377578557e-06,
+      "loss": 0.5911,
+      "step": 2833
+    },
+    {
+      "epoch": 0.23016324210184358,
+      "grad_norm": 3.0958255282858596,
+      "learning_rate": 4.4930014304659066e-06,
+      "loss": 0.6099,
+      "step": 2834
+    },
+    {
+      "epoch": 0.23024445707788516,
+      "grad_norm": 4.075827299092992,
+      "learning_rate": 4.492604345448106e-06,
+      "loss": 0.5688,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23032567205392673,
+      "grad_norm": 5.7956480372939065,
+      "learning_rate": 4.492207122552629e-06,
+      "loss": 0.6,
+      "step": 2836
+    },
+    {
+      "epoch": 0.23040688702996834,
+      "grad_norm": 5.13610708333408,
+      "learning_rate": 4.491809761806964e-06,
+      "loss": 0.5496,
+      "step": 2837
+    },
+    {
+      "epoch": 0.23048810200600992,
+      "grad_norm": 5.910698565673017,
+      "learning_rate": 4.491412263238605e-06,
+      "loss": 0.4669,
+      "step": 2838
+    },
+    {
+      "epoch": 0.2305693169820515,
+      "grad_norm": 3.4340630962930305,
+      "learning_rate": 4.4910146268750555e-06,
+      "loss": 0.5895,
+      "step": 2839
+    },
+    {
+      "epoch": 0.23065053195809307,
+      "grad_norm": 4.029802742225141,
+      "learning_rate": 4.490616852743832e-06,
+      "loss": 0.5887,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23073174693413465,
+      "grad_norm": 5.37046614141328,
+      "learning_rate": 4.490218940872457e-06,
+      "loss": 0.6715,
+      "step": 2841
+    },
+    {
+      "epoch": 0.23081296191017622,
+      "grad_norm": 6.574976668750347,
+      "learning_rate": 4.489820891288466e-06,
+      "loss": 0.6401,
+      "step": 2842
+    },
+    {
+      "epoch": 0.23089417688621783,
+      "grad_norm": 4.12936501875235,
+      "learning_rate": 4.489422704019399e-06,
+      "loss": 0.6287,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2309753918622594,
+      "grad_norm": 4.078499752253989,
+      "learning_rate": 4.489024379092809e-06,
+      "loss": 0.5193,
+      "step": 2844
+    },
+    {
+      "epoch": 0.23105660683830098,
+      "grad_norm": 8.139115119863513,
+      "learning_rate": 4.48862591653626e-06,
+      "loss": 0.415,
+      "step": 2845
+    },
+    {
+      "epoch": 0.23113782181434256,
+      "grad_norm": 6.159373760750005,
+      "learning_rate": 4.488227316377322e-06,
+      "loss": 0.5214,
+      "step": 2846
+    },
+    {
+      "epoch": 0.23121903679038414,
+      "grad_norm": 3.4411255569736956,
+      "learning_rate": 4.487828578643576e-06,
+      "loss": 0.541,
+      "step": 2847
+    },
+    {
+      "epoch": 0.23130025176642574,
+      "grad_norm": 4.121631308793716,
+      "learning_rate": 4.4874297033626126e-06,
+      "loss": 0.6158,
+      "step": 2848
+    },
+    {
+      "epoch": 0.23138146674246732,
+      "grad_norm": 3.8683262343429647,
+      "learning_rate": 4.487030690562032e-06,
+      "loss": 0.5957,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2314626817185089,
+      "grad_norm": 4.36066324441673,
+      "learning_rate": 4.486631540269445e-06,
+      "loss": 0.5908,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23154389669455047,
+      "grad_norm": 23.856582803005605,
+      "learning_rate": 4.486232252512468e-06,
+      "loss": 0.6196,
+      "step": 2851
+    },
+    {
+      "epoch": 0.23162511167059205,
+      "grad_norm": 4.565500746372573,
+      "learning_rate": 4.485832827318733e-06,
+      "loss": 0.517,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23170632664663363,
+      "grad_norm": 5.415515197041984,
+      "learning_rate": 4.485433264715874e-06,
+      "loss": 0.7045,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23178754162267523,
+      "grad_norm": 7.611491735007953,
+      "learning_rate": 4.485033564731542e-06,
+      "loss": 0.632,
+      "step": 2854
+    },
+    {
+      "epoch": 0.2318687565987168,
+      "grad_norm": 5.094950046339584,
+      "learning_rate": 4.484633727393393e-06,
+      "loss": 0.5768,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23194997157475838,
+      "grad_norm": 46.65401837853409,
+      "learning_rate": 4.484233752729093e-06,
+      "loss": 0.5038,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23203118655079996,
+      "grad_norm": 25.235913155436222,
+      "learning_rate": 4.483833640766319e-06,
+      "loss": 0.6252,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23211240152684154,
+      "grad_norm": 7.36939234328555,
+      "learning_rate": 4.4834333915327564e-06,
+      "loss": 0.6505,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23219361650288314,
+      "grad_norm": 4.830643690141342,
+      "learning_rate": 4.483033005056101e-06,
+      "loss": 0.5895,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23227483147892472,
+      "grad_norm": 9.183888328065596,
+      "learning_rate": 4.482632481364055e-06,
+      "loss": 0.4614,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2323560464549663,
+      "grad_norm": 4.613246918041221,
+      "learning_rate": 4.482231820484336e-06,
+      "loss": 0.6206,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23243726143100787,
+      "grad_norm": 7.324346436954002,
+      "learning_rate": 4.4818310224446645e-06,
+      "loss": 0.4812,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23251847640704945,
+      "grad_norm": 6.5981856748393755,
+      "learning_rate": 4.481430087272776e-06,
+      "loss": 0.6606,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23259969138309103,
+      "grad_norm": 4.619941208113225,
+      "learning_rate": 4.481029014996412e-06,
+      "loss": 0.4754,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23268090635913263,
+      "grad_norm": 5.7207154634398085,
+      "learning_rate": 4.480627805643324e-06,
+      "loss": 0.5482,
+      "step": 2865
+    },
+    {
+      "epoch": 0.2327621213351742,
+      "grad_norm": 7.298594910122886,
+      "learning_rate": 4.480226459241275e-06,
+      "loss": 0.6597,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2328433363112158,
+      "grad_norm": 9.363398730309251,
+      "learning_rate": 4.479824975818034e-06,
+      "loss": 0.6121,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23292455128725736,
+      "grad_norm": 4.781902048228461,
+      "learning_rate": 4.4794233554013835e-06,
+      "loss": 0.5151,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23300576626329894,
+      "grad_norm": 2.949065575080919,
+      "learning_rate": 4.479021598019113e-06,
+      "loss": 0.5063,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23308698123934055,
+      "grad_norm": 4.021888781896201,
+      "learning_rate": 4.4786197036990205e-06,
+      "loss": 0.5932,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23316819621538212,
+      "grad_norm": 5.096647466230962,
+      "learning_rate": 4.478217672468918e-06,
+      "loss": 0.6553,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2332494111914237,
+      "grad_norm": 3.3544437426357434,
+      "learning_rate": 4.47781550435662e-06,
+      "loss": 0.7021,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23333062616746528,
+      "grad_norm": 3.955029448497719,
+      "learning_rate": 4.4774131993899585e-06,
+      "loss": 0.5843,
+      "step": 2873
+    },
+    {
+      "epoch": 0.23341184114350685,
+      "grad_norm": 5.718278060740205,
+      "learning_rate": 4.477010757596768e-06,
+      "loss": 0.5778,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23349305611954843,
+      "grad_norm": 3.558212336992993,
+      "learning_rate": 4.4766081790048965e-06,
+      "loss": 0.4537,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23357427109559004,
+      "grad_norm": 3.821349848295602,
+      "learning_rate": 4.4762054636422005e-06,
+      "loss": 0.5913,
+      "step": 2876
+    },
+    {
+      "epoch": 0.2336554860716316,
+      "grad_norm": 5.178508133946879,
+      "learning_rate": 4.475802611536545e-06,
+      "loss": 0.5516,
+      "step": 2877
+    },
+    {
+      "epoch": 0.2337367010476732,
+      "grad_norm": 4.64155733262824,
+      "learning_rate": 4.475399622715805e-06,
+      "loss": 0.5186,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23381791602371477,
+      "grad_norm": 5.700083589065183,
+      "learning_rate": 4.474996497207866e-06,
+      "loss": 0.6616,
+      "step": 2879
+    },
+    {
+      "epoch": 0.23389913099975634,
+      "grad_norm": 5.357424859826312,
+      "learning_rate": 4.4745932350406225e-06,
+      "loss": 0.6395,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23398034597579795,
+      "grad_norm": 5.301441690670492,
+      "learning_rate": 4.474189836241976e-06,
+      "loss": 0.5116,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23406156095183953,
+      "grad_norm": 3.7935528708576784,
+      "learning_rate": 4.473786300839843e-06,
+      "loss": 0.537,
+      "step": 2882
+    },
+    {
+      "epoch": 0.2341427759278811,
+      "grad_norm": 4.6354063500549945,
+      "learning_rate": 4.4733826288621435e-06,
+      "loss": 0.4027,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23422399090392268,
+      "grad_norm": 4.258268978553181,
+      "learning_rate": 4.47297882033681e-06,
+      "loss": 0.5345,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23430520587996426,
+      "grad_norm": 4.645513866526441,
+      "learning_rate": 4.472574875291784e-06,
+      "loss": 0.3998,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23438642085600583,
+      "grad_norm": 5.592577886479116,
+      "learning_rate": 4.472170793755016e-06,
+      "loss": 0.5404,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23446763583204744,
+      "grad_norm": 4.864175858056895,
+      "learning_rate": 4.471766575754467e-06,
+      "loss": 0.4769,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23454885080808902,
+      "grad_norm": 5.3559791822991984,
+      "learning_rate": 4.471362221318106e-06,
+      "loss": 0.5522,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2346300657841306,
+      "grad_norm": 3.985841294995745,
+      "learning_rate": 4.470957730473913e-06,
+      "loss": 0.5913,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23471128076017217,
+      "grad_norm": 7.9982149419628765,
+      "learning_rate": 4.470553103249876e-06,
+      "loss": 0.5101,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23479249573621375,
+      "grad_norm": 3.748824385107628,
+      "learning_rate": 4.470148339673993e-06,
+      "loss": 0.5213,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23487371071225535,
+      "grad_norm": 4.816022996866285,
+      "learning_rate": 4.469743439774272e-06,
+      "loss": 0.5209,
+      "step": 2892
+    },
+    {
+      "epoch": 0.23495492568829693,
+      "grad_norm": 5.863963017632638,
+      "learning_rate": 4.46933840357873e-06,
+      "loss": 0.5256,
+      "step": 2893
+    },
+    {
+      "epoch": 0.2350361406643385,
+      "grad_norm": 4.876008648396625,
+      "learning_rate": 4.468933231115393e-06,
+      "loss": 0.53,
+      "step": 2894
+    },
+    {
+      "epoch": 0.23511735564038008,
+      "grad_norm": 6.008481144166018,
+      "learning_rate": 4.468527922412297e-06,
+      "loss": 0.5812,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23519857061642166,
+      "grad_norm": 6.418455057806858,
+      "learning_rate": 4.468122477497486e-06,
+      "loss": 0.5318,
+      "step": 2896
+    },
+    {
+      "epoch": 0.23527978559246324,
+      "grad_norm": 2.5240378044740415,
+      "learning_rate": 4.467716896399017e-06,
+      "loss": 0.5948,
+      "step": 2897
+    },
+    {
+      "epoch": 0.23536100056850484,
+      "grad_norm": 4.361122855517343,
+      "learning_rate": 4.4673111791449515e-06,
+      "loss": 0.5233,
+      "step": 2898
+    },
+    {
+      "epoch": 0.23544221554454642,
+      "grad_norm": 4.528523763290337,
+      "learning_rate": 4.466905325763365e-06,
+      "loss": 0.6362,
+      "step": 2899
+    },
+    {
+      "epoch": 0.235523430520588,
+      "grad_norm": 6.538860668916424,
+      "learning_rate": 4.4664993362823394e-06,
+      "loss": 0.5745,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23560464549662957,
+      "grad_norm": 5.847755183433358,
+      "learning_rate": 4.466093210729967e-06,
+      "loss": 0.6584,
+      "step": 2901
+    },
+    {
+      "epoch": 0.23568586047267115,
+      "grad_norm": 4.179474365670219,
+      "learning_rate": 4.465686949134351e-06,
+      "loss": 0.6396,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23576707544871275,
+      "grad_norm": 5.774986352129162,
+      "learning_rate": 4.465280551523601e-06,
+      "loss": 0.5861,
+      "step": 2903
+    },
+    {
+      "epoch": 0.23584829042475433,
+      "grad_norm": 3.2478970012240325,
+      "learning_rate": 4.464874017925837e-06,
+      "loss": 0.6337,
+      "step": 2904
+    },
+    {
+      "epoch": 0.2359295054007959,
+      "grad_norm": 6.068452434980089,
+      "learning_rate": 4.46446734836919e-06,
+      "loss": 0.6407,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23601072037683749,
+      "grad_norm": 3.2686213630532026,
+      "learning_rate": 4.4640605428818e-06,
+      "loss": 0.6955,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23609193535287906,
+      "grad_norm": 4.07905119935594,
+      "learning_rate": 4.463653601491815e-06,
+      "loss": 0.4519,
+      "step": 2907
+    },
+    {
+      "epoch": 0.23617315032892067,
+      "grad_norm": 6.169592515743247,
+      "learning_rate": 4.463246524227393e-06,
+      "loss": 0.523,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23625436530496224,
+      "grad_norm": 4.321690950129896,
+      "learning_rate": 4.462839311116702e-06,
+      "loss": 0.522,
+      "step": 2909
+    },
+    {
+      "epoch": 0.23633558028100382,
+      "grad_norm": 7.237413918755258,
+      "learning_rate": 4.462431962187919e-06,
+      "loss": 0.747,
+      "step": 2910
+    },
+    {
+      "epoch": 0.2364167952570454,
+      "grad_norm": 3.734173317333828,
+      "learning_rate": 4.46202447746923e-06,
+      "loss": 0.6299,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23649801023308697,
+      "grad_norm": 5.880101867809025,
+      "learning_rate": 4.461616856988831e-06,
+      "loss": 0.6253,
+      "step": 2912
+    },
+    {
+      "epoch": 0.23657922520912855,
+      "grad_norm": 4.370253406800247,
+      "learning_rate": 4.461209100774928e-06,
+      "loss": 0.5456,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23666044018517016,
+      "grad_norm": 4.39926590709237,
+      "learning_rate": 4.460801208855734e-06,
+      "loss": 0.427,
+      "step": 2914
+    },
+    {
+      "epoch": 0.23674165516121173,
+      "grad_norm": 6.280728294974258,
+      "learning_rate": 4.4603931812594735e-06,
+      "loss": 0.7197,
+      "step": 2915
+    },
+    {
+      "epoch": 0.2368228701372533,
+      "grad_norm": 6.982111134101197,
+      "learning_rate": 4.45998501801438e-06,
+      "loss": 0.4599,
+      "step": 2916
+    },
+    {
+      "epoch": 0.2369040851132949,
+      "grad_norm": 7.601832967298451,
+      "learning_rate": 4.459576719148697e-06,
+      "loss": 0.4797,
+      "step": 2917
+    },
+    {
+      "epoch": 0.23698530008933646,
+      "grad_norm": 5.8714956548009924,
+      "learning_rate": 4.459168284690676e-06,
+      "loss": 0.5447,
+      "step": 2918
+    },
+    {
+      "epoch": 0.23706651506537807,
+      "grad_norm": 7.1059975498172685,
+      "learning_rate": 4.458759714668578e-06,
+      "loss": 0.6044,
+      "step": 2919
+    },
+    {
+      "epoch": 0.23714773004141965,
+      "grad_norm": 4.612443635513808,
+      "learning_rate": 4.458351009110675e-06,
+      "loss": 0.5054,
+      "step": 2920
+    },
+    {
+      "epoch": 0.23722894501746122,
+      "grad_norm": 4.117158983616684,
+      "learning_rate": 4.457942168045246e-06,
+      "loss": 0.6243,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2373101599935028,
+      "grad_norm": 4.243517625792392,
+      "learning_rate": 4.457533191500581e-06,
+      "loss": 0.6199,
+      "step": 2922
+    },
+    {
+      "epoch": 0.23739137496954438,
+      "grad_norm": 6.196728526337707,
+      "learning_rate": 4.45712407950498e-06,
+      "loss": 0.4745,
+      "step": 2923
+    },
+    {
+      "epoch": 0.23747258994558595,
+      "grad_norm": 4.089855406887254,
+      "learning_rate": 4.45671483208675e-06,
+      "loss": 0.5762,
+      "step": 2924
+    },
+    {
+      "epoch": 0.23755380492162756,
+      "grad_norm": 8.680736194237051,
+      "learning_rate": 4.45630544927421e-06,
+      "loss": 0.6403,
+      "step": 2925
+    },
+    {
+      "epoch": 0.23763501989766914,
+      "grad_norm": 3.4364871224606452,
+      "learning_rate": 4.4558959310956865e-06,
+      "loss": 0.5683,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2377162348737107,
+      "grad_norm": 4.94476918130175,
+      "learning_rate": 4.4554862775795146e-06,
+      "loss": 0.4727,
+      "step": 2927
+    },
+    {
+      "epoch": 0.2377974498497523,
+      "grad_norm": 3.68514712087398,
+      "learning_rate": 4.455076488754043e-06,
+      "loss": 0.5719,
+      "step": 2928
+    },
+    {
+      "epoch": 0.23787866482579387,
+      "grad_norm": 4.764481615544404,
+      "learning_rate": 4.4546665646476254e-06,
+      "loss": 0.575,
+      "step": 2929
+    },
+    {
+      "epoch": 0.23795987980183547,
+      "grad_norm": 4.8777757605766405,
+      "learning_rate": 4.4542565052886256e-06,
+      "loss": 0.6751,
+      "step": 2930
+    },
+    {
+      "epoch": 0.23804109477787705,
+      "grad_norm": 4.182540781833306,
+      "learning_rate": 4.45384631070542e-06,
+      "loss": 0.6563,
+      "step": 2931
+    },
+    {
+      "epoch": 0.23812230975391863,
+      "grad_norm": 3.5345505370213264,
+      "learning_rate": 4.453435980926388e-06,
+      "loss": 0.5779,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2382035247299602,
+      "grad_norm": 8.12041102050761,
+      "learning_rate": 4.453025515979926e-06,
+      "loss": 0.5615,
+      "step": 2933
+    },
+    {
+      "epoch": 0.23828473970600178,
+      "grad_norm": 6.278197290858452,
+      "learning_rate": 4.452614915894434e-06,
+      "loss": 0.4717,
+      "step": 2934
+    },
+    {
+      "epoch": 0.23836595468204336,
+      "grad_norm": 3.470804618193489,
+      "learning_rate": 4.452204180698325e-06,
+      "loss": 0.5222,
+      "step": 2935
+    },
+    {
+      "epoch": 0.23844716965808496,
+      "grad_norm": 4.736573933269891,
+      "learning_rate": 4.451793310420017e-06,
+      "loss": 0.6273,
+      "step": 2936
+    },
+    {
+      "epoch": 0.23852838463412654,
+      "grad_norm": 6.791656553293532,
+      "learning_rate": 4.451382305087943e-06,
+      "loss": 0.6002,
+      "step": 2937
+    },
+    {
+      "epoch": 0.23860959961016812,
+      "grad_norm": 4.478548259706515,
+      "learning_rate": 4.450971164730541e-06,
+      "loss": 0.4814,
+      "step": 2938
+    },
+    {
+      "epoch": 0.2386908145862097,
+      "grad_norm": 3.681202303319773,
+      "learning_rate": 4.4505598893762595e-06,
+      "loss": 0.5895,
+      "step": 2939
+    },
+    {
+      "epoch": 0.23877202956225127,
+      "grad_norm": 4.1452045492879055,
+      "learning_rate": 4.4501484790535555e-06,
+      "loss": 0.6095,
+      "step": 2940
+    },
+    {
+      "epoch": 0.23885324453829287,
+      "grad_norm": 6.265763596141227,
+      "learning_rate": 4.449736933790899e-06,
+      "loss": 0.6445,
+      "step": 2941
+    },
+    {
+      "epoch": 0.23893445951433445,
+      "grad_norm": 3.195443239329471,
+      "learning_rate": 4.449325253616765e-06,
+      "loss": 0.6796,
+      "step": 2942
+    },
+    {
+      "epoch": 0.23901567449037603,
+      "grad_norm": 4.681861481447409,
+      "learning_rate": 4.448913438559641e-06,
+      "loss": 0.5013,
+      "step": 2943
+    },
+    {
+      "epoch": 0.2390968894664176,
+      "grad_norm": 3.9893399361235873,
+      "learning_rate": 4.448501488648021e-06,
+      "loss": 0.4841,
+      "step": 2944
+    },
+    {
+      "epoch": 0.23917810444245918,
+      "grad_norm": 8.26472685534627,
+      "learning_rate": 4.448089403910411e-06,
+      "loss": 0.65,
+      "step": 2945
+    },
+    {
+      "epoch": 0.23925931941850076,
+      "grad_norm": 7.279481691054026,
+      "learning_rate": 4.447677184375323e-06,
+      "loss": 0.6863,
+      "step": 2946
+    },
+    {
+      "epoch": 0.23934053439454236,
+      "grad_norm": 5.107617453615278,
+      "learning_rate": 4.447264830071282e-06,
+      "loss": 0.4314,
+      "step": 2947
+    },
+    {
+      "epoch": 0.23942174937058394,
+      "grad_norm": 10.45135842568149,
+      "learning_rate": 4.446852341026822e-06,
+      "loss": 0.5368,
+      "step": 2948
+    },
+    {
+      "epoch": 0.23950296434662552,
+      "grad_norm": 3.384814108308006,
+      "learning_rate": 4.4464397172704825e-06,
+      "loss": 0.5992,
+      "step": 2949
+    },
+    {
+      "epoch": 0.2395841793226671,
+      "grad_norm": 3.6038417485455723,
+      "learning_rate": 4.446026958830816e-06,
+      "loss": 0.6577,
+      "step": 2950
+    },
+    {
+      "epoch": 0.23966539429870867,
+      "grad_norm": 5.105585351046193,
+      "learning_rate": 4.4456140657363824e-06,
+      "loss": 0.5815,
+      "step": 2951
+    },
+    {
+      "epoch": 0.23974660927475028,
+      "grad_norm": 6.436128437554316,
+      "learning_rate": 4.445201038015753e-06,
+      "loss": 0.6109,
+      "step": 2952
+    },
+    {
+      "epoch": 0.23982782425079185,
+      "grad_norm": 6.0286114195287,
+      "learning_rate": 4.4447878756975074e-06,
+      "loss": 0.5737,
+      "step": 2953
+    },
+    {
+      "epoch": 0.23990903922683343,
+      "grad_norm": 6.340347602254372,
+      "learning_rate": 4.444374578810233e-06,
+      "loss": 0.6333,
+      "step": 2954
+    },
+    {
+      "epoch": 0.239990254202875,
+      "grad_norm": 3.611889348535864,
+      "learning_rate": 4.443961147382528e-06,
+      "loss": 0.5584,
+      "step": 2955
+    },
+    {
+      "epoch": 0.24007146917891659,
+      "grad_norm": 4.508170768419845,
+      "learning_rate": 4.4435475814429995e-06,
+      "loss": 0.5662,
+      "step": 2956
+    },
+    {
+      "epoch": 0.24015268415495816,
+      "grad_norm": 3.857621418576937,
+      "learning_rate": 4.4431338810202655e-06,
+      "loss": 0.5413,
+      "step": 2957
+    },
+    {
+      "epoch": 0.24023389913099977,
+      "grad_norm": 5.167525653044904,
+      "learning_rate": 4.4427200461429494e-06,
+      "loss": 0.5279,
+      "step": 2958
+    },
+    {
+      "epoch": 0.24031511410704134,
+      "grad_norm": 9.873321863622351,
+      "learning_rate": 4.442306076839689e-06,
+      "loss": 0.4993,
+      "step": 2959
+    },
+    {
+      "epoch": 0.24039632908308292,
+      "grad_norm": 9.73450941091562,
+      "learning_rate": 4.441891973139127e-06,
+      "loss": 0.5868,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2404775440591245,
+      "grad_norm": 3.7053040122979297,
+      "learning_rate": 4.441477735069918e-06,
+      "loss": 0.6352,
+      "step": 2961
+    },
+    {
+      "epoch": 0.24055875903516608,
+      "grad_norm": 5.920625186279193,
+      "learning_rate": 4.441063362660726e-06,
+      "loss": 0.6923,
+      "step": 2962
+    },
+    {
+      "epoch": 0.24063997401120768,
+      "grad_norm": 6.875957645683971,
+      "learning_rate": 4.44064885594022e-06,
+      "loss": 0.5829,
+      "step": 2963
+    },
+    {
+      "epoch": 0.24072118898724926,
+      "grad_norm": 6.3088089208901215,
+      "learning_rate": 4.440234214937086e-06,
+      "loss": 0.4949,
+      "step": 2964
+    },
+    {
+      "epoch": 0.24080240396329083,
+      "grad_norm": 6.283915304694119,
+      "learning_rate": 4.439819439680012e-06,
+      "loss": 0.5373,
+      "step": 2965
+    },
+    {
+      "epoch": 0.2408836189393324,
+      "grad_norm": 5.544344295999567,
+      "learning_rate": 4.439404530197699e-06,
+      "loss": 0.5615,
+      "step": 2966
+    },
+    {
+      "epoch": 0.240964833915374,
+      "grad_norm": 5.522735082552348,
+      "learning_rate": 4.438989486518856e-06,
+      "loss": 0.496,
+      "step": 2967
+    },
+    {
+      "epoch": 0.24104604889141557,
+      "grad_norm": 6.134276115094066,
+      "learning_rate": 4.438574308672203e-06,
+      "loss": 0.3989,
+      "step": 2968
+    },
+    {
+      "epoch": 0.24112726386745717,
+      "grad_norm": 18.34187318276363,
+      "learning_rate": 4.438158996686468e-06,
+      "loss": 0.4992,
+      "step": 2969
+    },
+    {
+      "epoch": 0.24120847884349875,
+      "grad_norm": 5.7407597574290605,
+      "learning_rate": 4.4377435505903876e-06,
+      "loss": 0.7617,
+      "step": 2970
+    },
+    {
+      "epoch": 0.24128969381954032,
+      "grad_norm": 5.129242694583421,
+      "learning_rate": 4.4373279704127095e-06,
+      "loss": 0.6338,
+      "step": 2971
+    },
+    {
+      "epoch": 0.2413709087955819,
+      "grad_norm": 5.956882401490225,
+      "learning_rate": 4.4369122561821885e-06,
+      "loss": 0.4831,
+      "step": 2972
+    },
+    {
+      "epoch": 0.24145212377162348,
+      "grad_norm": 4.390822342695221,
+      "learning_rate": 4.436496407927591e-06,
+      "loss": 0.5962,
+      "step": 2973
+    },
+    {
+      "epoch": 0.24153333874766508,
+      "grad_norm": 8.138342846156773,
+      "learning_rate": 4.436080425677689e-06,
+      "loss": 0.5129,
+      "step": 2974
+    },
+    {
+      "epoch": 0.24161455372370666,
+      "grad_norm": 3.692323205845064,
+      "learning_rate": 4.43566430946127e-06,
+      "loss": 0.6631,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24169576869974824,
+      "grad_norm": 4.598524784324876,
+      "learning_rate": 4.435248059307124e-06,
+      "loss": 0.5203,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2417769836757898,
+      "grad_norm": 36.50302303476803,
+      "learning_rate": 4.434831675244056e-06,
+      "loss": 0.5976,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2418581986518314,
+      "grad_norm": 4.225579732329154,
+      "learning_rate": 4.434415157300875e-06,
+      "loss": 0.7187,
+      "step": 2978
+    },
+    {
+      "epoch": 0.24193941362787297,
+      "grad_norm": 6.144364914404981,
+      "learning_rate": 4.433998505506402e-06,
+      "loss": 0.5036,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24202062860391457,
+      "grad_norm": 5.890869286549283,
+      "learning_rate": 4.433581719889469e-06,
+      "loss": 0.5174,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24210184357995615,
+      "grad_norm": 4.666862090401287,
+      "learning_rate": 4.433164800478914e-06,
+      "loss": 0.5758,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24218305855599773,
+      "grad_norm": 4.488080886068684,
+      "learning_rate": 4.432747747303586e-06,
+      "loss": 0.5223,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2422642735320393,
+      "grad_norm": 4.722733122087884,
+      "learning_rate": 4.432330560392343e-06,
+      "loss": 0.5231,
+      "step": 2983
+    },
+    {
+      "epoch": 0.24234548850808088,
+      "grad_norm": 10.302897259600046,
+      "learning_rate": 4.431913239774052e-06,
+      "loss": 0.6269,
+      "step": 2984
+    },
+    {
+      "epoch": 0.24242670348412249,
+      "grad_norm": 4.986694907260506,
+      "learning_rate": 4.4314957854775895e-06,
+      "loss": 0.4772,
+      "step": 2985
+    },
+    {
+      "epoch": 0.24250791846016406,
+      "grad_norm": 28.644633867544297,
+      "learning_rate": 4.43107819753184e-06,
+      "loss": 0.4776,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24258913343620564,
+      "grad_norm": 4.650389954451338,
+      "learning_rate": 4.4306604759657e-06,
+      "loss": 0.6489,
+      "step": 2987
+    },
+    {
+      "epoch": 0.24267034841224722,
+      "grad_norm": 8.110031601217097,
+      "learning_rate": 4.430242620808073e-06,
+      "loss": 0.4797,
+      "step": 2988
+    },
+    {
+      "epoch": 0.2427515633882888,
+      "grad_norm": 4.435012545474071,
+      "learning_rate": 4.429824632087873e-06,
+      "loss": 0.5772,
+      "step": 2989
+    },
+    {
+      "epoch": 0.24283277836433037,
+      "grad_norm": 6.6461768049947505,
+      "learning_rate": 4.42940650983402e-06,
+      "loss": 0.6512,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24291399334037198,
+      "grad_norm": 4.2488570760327935,
+      "learning_rate": 4.428988254075449e-06,
+      "loss": 0.4932,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24299520831641355,
+      "grad_norm": 3.4583725498967657,
+      "learning_rate": 4.4285698648411005e-06,
+      "loss": 0.6507,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24307642329245513,
+      "grad_norm": 8.442412186418991,
+      "learning_rate": 4.428151342159923e-06,
+      "loss": 0.7045,
+      "step": 2993
+    },
+    {
+      "epoch": 0.2431576382684967,
+      "grad_norm": 4.275956058237837,
+      "learning_rate": 4.427732686060877e-06,
+      "loss": 0.9022,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24323885324453828,
+      "grad_norm": 3.5996383051852328,
+      "learning_rate": 4.427313896572933e-06,
+      "loss": 0.5063,
+      "step": 2995
+    },
+    {
+      "epoch": 0.2433200682205799,
+      "grad_norm": 8.226615556870824,
+      "learning_rate": 4.426894973725066e-06,
+      "loss": 0.5054,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24340128319662147,
+      "grad_norm": 4.428674996775757,
+      "learning_rate": 4.426475917546266e-06,
+      "loss": 0.5189,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24348249817266304,
+      "grad_norm": 3.292995498057493,
+      "learning_rate": 4.426056728065527e-06,
+      "loss": 0.5175,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24356371314870462,
+      "grad_norm": 5.396362723485166,
+      "learning_rate": 4.425637405311857e-06,
+      "loss": 0.5239,
+      "step": 2999
+    },
+    {
+      "epoch": 0.2436449281247462,
+      "grad_norm": 4.409324480057902,
+      "learning_rate": 4.425217949314269e-06,
+      "loss": 0.5758,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24372614310078777,
+      "grad_norm": 4.772191402586126,
+      "learning_rate": 4.424798360101788e-06,
+      "loss": 0.5178,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24380735807682938,
+      "grad_norm": 4.70270174443938,
+      "learning_rate": 4.424378637703448e-06,
+      "loss": 0.5634,
+      "step": 3002
+    },
+    {
+      "epoch": 0.24388857305287096,
+      "grad_norm": 4.384411856516953,
+      "learning_rate": 4.423958782148291e-06,
+      "loss": 0.6138,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24396978802891253,
+      "grad_norm": 4.849333757605469,
+      "learning_rate": 4.423538793465368e-06,
+      "loss": 0.5399,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2440510030049541,
+      "grad_norm": 7.683916564358679,
+      "learning_rate": 4.423118671683741e-06,
+      "loss": 0.5882,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24413221798099569,
+      "grad_norm": 6.414015664183622,
+      "learning_rate": 4.42269841683248e-06,
+      "loss": 0.501,
+      "step": 3006
+    },
+    {
+      "epoch": 0.2442134329570373,
+      "grad_norm": 3.445748481334447,
+      "learning_rate": 4.422278028940664e-06,
+      "loss": 0.7989,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24429464793307887,
+      "grad_norm": 3.446395382611435,
+      "learning_rate": 4.4218575080373825e-06,
+      "loss": 0.6564,
+      "step": 3008
+    },
+    {
+      "epoch": 0.24437586290912044,
+      "grad_norm": 7.075485495123882,
+      "learning_rate": 4.421436854151731e-06,
+      "loss": 0.5491,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24445707788516202,
+      "grad_norm": 13.379409500342488,
+      "learning_rate": 4.421016067312821e-06,
+      "loss": 0.5192,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2445382928612036,
+      "grad_norm": 6.268741236760656,
+      "learning_rate": 4.420595147549764e-06,
+      "loss": 0.5197,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24461950783724518,
+      "grad_norm": 4.324676469837308,
+      "learning_rate": 4.420174094891688e-06,
+      "loss": 0.6056,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24470072281328678,
+      "grad_norm": 4.740074003918016,
+      "learning_rate": 4.419752909367727e-06,
+      "loss": 0.4472,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24478193778932836,
+      "grad_norm": 4.345165270453679,
+      "learning_rate": 4.419331591007025e-06,
+      "loss": 0.662,
+      "step": 3014
+    },
+    {
+      "epoch": 0.24486315276536993,
+      "grad_norm": 3.2912342359960913,
+      "learning_rate": 4.418910139838734e-06,
+      "loss": 0.6347,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2449443677414115,
+      "grad_norm": 4.197613659900217,
+      "learning_rate": 4.418488555892018e-06,
+      "loss": 0.709,
+      "step": 3016
+    },
+    {
+      "epoch": 0.2450255827174531,
+      "grad_norm": 6.165218832788323,
+      "learning_rate": 4.418066839196047e-06,
+      "loss": 0.5097,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2451067976934947,
+      "grad_norm": 5.754235502940853,
+      "learning_rate": 4.4176449897800025e-06,
+      "loss": 0.4602,
+      "step": 3018
+    },
+    {
+      "epoch": 0.24518801266953627,
+      "grad_norm": 6.807261983784002,
+      "learning_rate": 4.417223007673073e-06,
+      "loss": 0.6116,
+      "step": 3019
+    },
+    {
+      "epoch": 0.24526922764557785,
+      "grad_norm": 5.874117891239969,
+      "learning_rate": 4.4168008929044585e-06,
+      "loss": 0.4809,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24535044262161942,
+      "grad_norm": 4.771234322125993,
+      "learning_rate": 4.416378645503366e-06,
+      "loss": 0.5298,
+      "step": 3021
+    },
+    {
+      "epoch": 0.245431657597661,
+      "grad_norm": 5.033152836788516,
+      "learning_rate": 4.415956265499014e-06,
+      "loss": 0.5011,
+      "step": 3022
+    },
+    {
+      "epoch": 0.24551287257370258,
+      "grad_norm": 4.892045552702636,
+      "learning_rate": 4.415533752920629e-06,
+      "loss": 0.5847,
+      "step": 3023
+    },
+    {
+      "epoch": 0.24559408754974418,
+      "grad_norm": 3.158494131947643,
+      "learning_rate": 4.415111107797445e-06,
+      "loss": 0.5626,
+      "step": 3024
+    },
+    {
+      "epoch": 0.24567530252578576,
+      "grad_norm": 19.734315350148133,
+      "learning_rate": 4.414688330158709e-06,
+      "loss": 0.6228,
+      "step": 3025
+    },
+    {
+      "epoch": 0.24575651750182734,
+      "grad_norm": 3.975333547273237,
+      "learning_rate": 4.4142654200336735e-06,
+      "loss": 0.6962,
+      "step": 3026
+    },
+    {
+      "epoch": 0.24583773247786891,
+      "grad_norm": 5.615922654297775,
+      "learning_rate": 4.413842377451602e-06,
+      "loss": 0.541,
+      "step": 3027
+    },
+    {
+      "epoch": 0.2459189474539105,
+      "grad_norm": 9.003336691465453,
+      "learning_rate": 4.4134192024417674e-06,
+      "loss": 0.4546,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2460001624299521,
+      "grad_norm": 3.891965066900615,
+      "learning_rate": 4.412995895033449e-06,
+      "loss": 0.4849,
+      "step": 3029
+    },
+    {
+      "epoch": 0.24608137740599367,
+      "grad_norm": 4.878722949580337,
+      "learning_rate": 4.412572455255942e-06,
+      "loss": 0.615,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24616259238203525,
+      "grad_norm": 7.419504517267929,
+      "learning_rate": 4.412148883138541e-06,
+      "loss": 0.5744,
+      "step": 3031
+    },
+    {
+      "epoch": 0.24624380735807683,
+      "grad_norm": 15.433110249433737,
+      "learning_rate": 4.4117251787105566e-06,
+      "loss": 0.4246,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2463250223341184,
+      "grad_norm": 4.626381274322953,
+      "learning_rate": 4.411301342001309e-06,
+      "loss": 0.4779,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24640623731015998,
+      "grad_norm": 3.140055018404333,
+      "learning_rate": 4.4108773730401235e-06,
+      "loss": 0.5733,
+      "step": 3034
+    },
+    {
+      "epoch": 0.24648745228620159,
+      "grad_norm": 5.288841429027363,
+      "learning_rate": 4.410453271856337e-06,
+      "loss": 0.5525,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24656866726224316,
+      "grad_norm": 5.257436077679411,
+      "learning_rate": 4.410029038479295e-06,
+      "loss": 0.5659,
+      "step": 3036
+    },
+    {
+      "epoch": 0.24664988223828474,
+      "grad_norm": 5.634845248429157,
+      "learning_rate": 4.409604672938352e-06,
+      "loss": 0.5136,
+      "step": 3037
+    },
+    {
+      "epoch": 0.24673109721432632,
+      "grad_norm": 6.3457504654963195,
+      "learning_rate": 4.409180175262872e-06,
+      "loss": 0.5139,
+      "step": 3038
+    },
+    {
+      "epoch": 0.2468123121903679,
+      "grad_norm": 6.705941236376485,
+      "learning_rate": 4.408755545482229e-06,
+      "loss": 0.6184,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2468935271664095,
+      "grad_norm": 4.672344302149794,
+      "learning_rate": 4.408330783625803e-06,
+      "loss": 0.6296,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24697474214245108,
+      "grad_norm": 7.4047101228444685,
+      "learning_rate": 4.407905889722987e-06,
+      "loss": 0.3766,
+      "step": 3041
+    },
+    {
+      "epoch": 0.24705595711849265,
+      "grad_norm": 4.2484417683054545,
+      "learning_rate": 4.407480863803181e-06,
+      "loss": 0.5838,
+      "step": 3042
+    },
+    {
+      "epoch": 0.24713717209453423,
+      "grad_norm": 5.209047094212918,
+      "learning_rate": 4.407055705895794e-06,
+      "loss": 0.6218,
+      "step": 3043
+    },
+    {
+      "epoch": 0.2472183870705758,
+      "grad_norm": 4.0255565337154025,
+      "learning_rate": 4.4066304160302455e-06,
+      "loss": 0.5566,
+      "step": 3044
+    },
+    {
+      "epoch": 0.24729960204661738,
+      "grad_norm": 4.333677201999764,
+      "learning_rate": 4.4062049942359634e-06,
+      "loss": 0.5295,
+      "step": 3045
+    },
+    {
+      "epoch": 0.247380817022659,
+      "grad_norm": 4.229077150168856,
+      "learning_rate": 4.405779440542383e-06,
+      "loss": 0.7309,
+      "step": 3046
+    },
+    {
+      "epoch": 0.24746203199870057,
+      "grad_norm": 7.678488339873706,
+      "learning_rate": 4.405353754978952e-06,
+      "loss": 0.5675,
+      "step": 3047
+    },
+    {
+      "epoch": 0.24754324697474214,
+      "grad_norm": 5.135214318233568,
+      "learning_rate": 4.404927937575125e-06,
+      "loss": 0.6678,
+      "step": 3048
+    },
+    {
+      "epoch": 0.24762446195078372,
+      "grad_norm": 5.25362634341942,
+      "learning_rate": 4.4045019883603676e-06,
+      "loss": 0.544,
+      "step": 3049
+    },
+    {
+      "epoch": 0.2477056769268253,
+      "grad_norm": 5.0585319314475194,
+      "learning_rate": 4.40407590736415e-06,
+      "loss": 0.6509,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2477868919028669,
+      "grad_norm": 6.248691395394384,
+      "learning_rate": 4.403649694615959e-06,
+      "loss": 0.5899,
+      "step": 3051
+    },
+    {
+      "epoch": 0.24786810687890848,
+      "grad_norm": 10.336963003211816,
+      "learning_rate": 4.403223350145283e-06,
+      "loss": 0.4696,
+      "step": 3052
+    },
+    {
+      "epoch": 0.24794932185495006,
+      "grad_norm": 4.541662859187703,
+      "learning_rate": 4.402796873981623e-06,
+      "loss": 0.4006,
+      "step": 3053
+    },
+    {
+      "epoch": 0.24803053683099163,
+      "grad_norm": 6.296840627991974,
+      "learning_rate": 4.402370266154491e-06,
+      "loss": 0.5062,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2481117518070332,
+      "grad_norm": 5.894042406789113,
+      "learning_rate": 4.401943526693404e-06,
+      "loss": 0.7281,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2481929667830748,
+      "grad_norm": 5.656446562333912,
+      "learning_rate": 4.401516655627891e-06,
+      "loss": 0.6206,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2482741817591164,
+      "grad_norm": 5.686281684309384,
+      "learning_rate": 4.401089652987489e-06,
+      "loss": 0.4256,
+      "step": 3057
+    },
+    {
+      "epoch": 0.24835539673515797,
+      "grad_norm": 6.0150133514660284,
+      "learning_rate": 4.4006625188017445e-06,
+      "loss": 0.502,
+      "step": 3058
+    },
+    {
+      "epoch": 0.24843661171119955,
+      "grad_norm": 5.249632475937014,
+      "learning_rate": 4.400235253100214e-06,
+      "loss": 0.5524,
+      "step": 3059
+    },
+    {
+      "epoch": 0.24851782668724112,
+      "grad_norm": 4.835925595719573,
+      "learning_rate": 4.399807855912459e-06,
+      "loss": 0.6269,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2485990416632827,
+      "grad_norm": 4.606327993826141,
+      "learning_rate": 4.3993803272680555e-06,
+      "loss": 0.6161,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2486802566393243,
+      "grad_norm": 7.236306200132777,
+      "learning_rate": 4.398952667196585e-06,
+      "loss": 0.479,
+      "step": 3062
+    },
+    {
+      "epoch": 0.24876147161536588,
+      "grad_norm": 5.858172893023612,
+      "learning_rate": 4.398524875727641e-06,
+      "loss": 0.568,
+      "step": 3063
+    },
+    {
+      "epoch": 0.24884268659140746,
+      "grad_norm": 4.056227046494207,
+      "learning_rate": 4.398096952890823e-06,
+      "loss": 0.631,
+      "step": 3064
+    },
+    {
+      "epoch": 0.24892390156744904,
+      "grad_norm": 8.0717170292666,
+      "learning_rate": 4.397668898715743e-06,
+      "loss": 0.5397,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2490051165434906,
+      "grad_norm": 6.347095752035782,
+      "learning_rate": 4.397240713232016e-06,
+      "loss": 0.5774,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2490863315195322,
+      "grad_norm": 4.659148068120233,
+      "learning_rate": 4.3968123964692745e-06,
+      "loss": 0.4825,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2491675464955738,
+      "grad_norm": 5.987640448722061,
+      "learning_rate": 4.396383948457153e-06,
+      "loss": 0.6587,
+      "step": 3068
+    },
+    {
+      "epoch": 0.24924876147161537,
+      "grad_norm": 3.9258521372891573,
+      "learning_rate": 4.395955369225299e-06,
+      "loss": 0.8384,
+      "step": 3069
+    },
+    {
+      "epoch": 0.24932997644765695,
+      "grad_norm": 5.296577196209221,
+      "learning_rate": 4.395526658803367e-06,
+      "loss": 0.5995,
+      "step": 3070
+    },
+    {
+      "epoch": 0.24941119142369853,
+      "grad_norm": 6.077098107367331,
+      "learning_rate": 4.395097817221023e-06,
+      "loss": 0.5141,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2494924063997401,
+      "grad_norm": 9.074255129254627,
+      "learning_rate": 4.39466884450794e-06,
+      "loss": 0.5545,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2495736213757817,
+      "grad_norm": 3.837360404731743,
+      "learning_rate": 4.3942397406937996e-06,
+      "loss": 0.7089,
+      "step": 3073
+    },
+    {
+      "epoch": 0.24965483635182328,
+      "grad_norm": 24.0705694528878,
+      "learning_rate": 4.393810505808294e-06,
+      "loss": 0.685,
+      "step": 3074
+    },
+    {
+      "epoch": 0.24973605132786486,
+      "grad_norm": 29.345212216374478,
+      "learning_rate": 4.393381139881125e-06,
+      "loss": 0.5362,
+      "step": 3075
+    },
+    {
+      "epoch": 0.24981726630390644,
+      "grad_norm": 2.6160968338209645,
+      "learning_rate": 4.392951642942001e-06,
+      "loss": 0.5189,
+      "step": 3076
+    },
+    {
+      "epoch": 0.24989848127994801,
+      "grad_norm": 3.2671386577254586,
+      "learning_rate": 4.392522015020643e-06,
+      "loss": 0.526,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2499796962559896,
+      "grad_norm": 3.4560985141832905,
+      "learning_rate": 4.392092256146776e-06,
+      "loss": 0.484,
+      "step": 3078
+    },
+    {
+      "epoch": 0.25006091123203117,
+      "grad_norm": 4.1989067824587405,
+      "learning_rate": 4.391662366350139e-06,
+      "loss": 0.482,
+      "step": 3079
+    },
+    {
+      "epoch": 0.25014212620807275,
+      "grad_norm": 5.326290581465668,
+      "learning_rate": 4.3912323456604785e-06,
+      "loss": 0.6224,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2502233411841144,
+      "grad_norm": 7.502364694548779,
+      "learning_rate": 4.390802194107548e-06,
+      "loss": 0.5254,
+      "step": 3081
+    },
+    {
+      "epoch": 0.25030455616015596,
+      "grad_norm": 4.007678084848024,
+      "learning_rate": 4.390371911721113e-06,
+      "loss": 0.6956,
+      "step": 3082
+    },
+    {
+      "epoch": 0.25038577113619753,
+      "grad_norm": 6.194239862786131,
+      "learning_rate": 4.389941498530946e-06,
+      "loss": 0.6204,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2504669861122391,
+      "grad_norm": 5.28161752223026,
+      "learning_rate": 4.38951095456683e-06,
+      "loss": 0.6249,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2505482010882807,
+      "grad_norm": 6.6004264867031415,
+      "learning_rate": 4.389080279858556e-06,
+      "loss": 0.6299,
+      "step": 3085
+    },
+    {
+      "epoch": 0.25062941606432226,
+      "grad_norm": 4.08243439857223,
+      "learning_rate": 4.388649474435925e-06,
+      "loss": 0.5395,
+      "step": 3086
+    },
+    {
+      "epoch": 0.25071063104036384,
+      "grad_norm": 10.991487088043705,
+      "learning_rate": 4.388218538328746e-06,
+      "loss": 0.4487,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2507918460164054,
+      "grad_norm": 5.374906431529036,
+      "learning_rate": 4.387787471566837e-06,
+      "loss": 0.6908,
+      "step": 3088
+    },
+    {
+      "epoch": 0.250873060992447,
+      "grad_norm": 6.4091647653829344,
+      "learning_rate": 4.387356274180025e-06,
+      "loss": 0.559,
+      "step": 3089
+    },
+    {
+      "epoch": 0.25095427596848857,
+      "grad_norm": 3.189931420557015,
+      "learning_rate": 4.386924946198148e-06,
+      "loss": 0.5294,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25103549094453015,
+      "grad_norm": 3.881260853633071,
+      "learning_rate": 4.386493487651052e-06,
+      "loss": 0.7014,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2511167059205718,
+      "grad_norm": 9.67421505771052,
+      "learning_rate": 4.38606189856859e-06,
+      "loss": 0.506,
+      "step": 3092
+    },
+    {
+      "epoch": 0.25119792089661336,
+      "grad_norm": 12.331043817036711,
+      "learning_rate": 4.385630178980627e-06,
+      "loss": 0.539,
+      "step": 3093
+    },
+    {
+      "epoch": 0.25127913587265494,
+      "grad_norm": 4.246731275050449,
+      "learning_rate": 4.385198328917034e-06,
+      "loss": 0.5523,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2513603508486965,
+      "grad_norm": 5.769645319972767,
+      "learning_rate": 4.384766348407695e-06,
+      "loss": 0.5281,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2514415658247381,
+      "grad_norm": 4.286081820286076,
+      "learning_rate": 4.3843342374825e-06,
+      "loss": 0.5954,
+      "step": 3096
+    },
+    {
+      "epoch": 0.25152278080077967,
+      "grad_norm": 3.6620430541424187,
+      "learning_rate": 4.383901996171348e-06,
+      "loss": 0.809,
+      "step": 3097
+    },
+    {
+      "epoch": 0.25160399577682124,
+      "grad_norm": 4.730554157695557,
+      "learning_rate": 4.383469624504148e-06,
+      "loss": 0.5542,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2516852107528628,
+      "grad_norm": 5.354061606302798,
+      "learning_rate": 4.3830371225108185e-06,
+      "loss": 0.4754,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2517664257289044,
+      "grad_norm": 7.122075601960895,
+      "learning_rate": 4.382604490221286e-06,
+      "loss": 0.4081,
+      "step": 3100
+    },
+    {
+      "epoch": 0.251847640704946,
+      "grad_norm": 4.487864560668316,
+      "learning_rate": 4.382171727665486e-06,
+      "loss": 0.71,
+      "step": 3101
+    },
+    {
+      "epoch": 0.25192885568098755,
+      "grad_norm": 5.423045386515985,
+      "learning_rate": 4.381738834873364e-06,
+      "loss": 0.518,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2520100706570292,
+      "grad_norm": 4.117696423281904,
+      "learning_rate": 4.381305811874873e-06,
+      "loss": 0.3472,
+      "step": 3103
+    },
+    {
+      "epoch": 0.25209128563307076,
+      "grad_norm": 4.590924951658724,
+      "learning_rate": 4.3808726586999766e-06,
+      "loss": 0.6369,
+      "step": 3104
+    },
+    {
+      "epoch": 0.25217250060911234,
+      "grad_norm": 6.69015366947571,
+      "learning_rate": 4.380439375378646e-06,
+      "loss": 0.5335,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2522537155851539,
+      "grad_norm": 9.103381660915732,
+      "learning_rate": 4.380005961940864e-06,
+      "loss": 0.4577,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2523349305611955,
+      "grad_norm": 7.434831084210403,
+      "learning_rate": 4.379572418416619e-06,
+      "loss": 0.5355,
+      "step": 3107
+    },
+    {
+      "epoch": 0.25241614553723707,
+      "grad_norm": 8.173889824189096,
+      "learning_rate": 4.37913874483591e-06,
+      "loss": 0.4481,
+      "step": 3108
+    },
+    {
+      "epoch": 0.25249736051327865,
+      "grad_norm": 10.499303258759314,
+      "learning_rate": 4.378704941228746e-06,
+      "loss": 0.5386,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2525785754893202,
+      "grad_norm": 5.028078700952843,
+      "learning_rate": 4.378271007625141e-06,
+      "loss": 0.4977,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2526597904653618,
+      "grad_norm": 4.3897095234161245,
+      "learning_rate": 4.377836944055124e-06,
+      "loss": 0.5747,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2527410054414034,
+      "grad_norm": 5.593416675150363,
+      "learning_rate": 4.377402750548729e-06,
+      "loss": 0.6102,
+      "step": 3112
+    },
+    {
+      "epoch": 0.25282222041744495,
+      "grad_norm": 5.376961891531233,
+      "learning_rate": 4.376968427135999e-06,
+      "loss": 0.6942,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2529034353934866,
+      "grad_norm": 7.126459999216967,
+      "learning_rate": 4.376533973846988e-06,
+      "loss": 0.7788,
+      "step": 3114
+    },
+    {
+      "epoch": 0.25298465036952816,
+      "grad_norm": 4.854425202352291,
+      "learning_rate": 4.376099390711758e-06,
+      "loss": 0.4626,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25306586534556974,
+      "grad_norm": 8.756831124179643,
+      "learning_rate": 4.375664677760378e-06,
+      "loss": 0.6323,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2531470803216113,
+      "grad_norm": 4.267269088423493,
+      "learning_rate": 4.375229835022929e-06,
+      "loss": 0.6067,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2532282952976529,
+      "grad_norm": 5.198815978500407,
+      "learning_rate": 4.374794862529501e-06,
+      "loss": 0.5796,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25330951027369447,
+      "grad_norm": 4.95003494031335,
+      "learning_rate": 4.374359760310191e-06,
+      "loss": 0.4924,
+      "step": 3119
+    },
+    {
+      "epoch": 0.25339072524973605,
+      "grad_norm": 4.604098155034568,
+      "learning_rate": 4.373924528395105e-06,
+      "loss": 0.6536,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2534719402257776,
+      "grad_norm": 5.033107338044671,
+      "learning_rate": 4.373489166814358e-06,
+      "loss": 0.6378,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2535531552018192,
+      "grad_norm": 5.806258612021515,
+      "learning_rate": 4.3730536755980776e-06,
+      "loss": 0.4952,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2536343701778608,
+      "grad_norm": 4.217899119753335,
+      "learning_rate": 4.372618054776395e-06,
+      "loss": 0.7578,
+      "step": 3123
+    },
+    {
+      "epoch": 0.25371558515390236,
+      "grad_norm": 4.5310110282850085,
+      "learning_rate": 4.372182304379455e-06,
+      "loss": 0.5417,
+      "step": 3124
+    },
+    {
+      "epoch": 0.253796800129944,
+      "grad_norm": 5.792695550840339,
+      "learning_rate": 4.371746424437406e-06,
+      "loss": 0.5578,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25387801510598557,
+      "grad_norm": 9.674533837851303,
+      "learning_rate": 4.371310414980412e-06,
+      "loss": 0.5975,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25395923008202714,
+      "grad_norm": 4.234861474964158,
+      "learning_rate": 4.37087427603864e-06,
+      "loss": 0.5482,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2540404450580687,
+      "grad_norm": 5.754506563667726,
+      "learning_rate": 4.37043800764227e-06,
+      "loss": 0.5447,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2541216600341103,
+      "grad_norm": 10.678832172560531,
+      "learning_rate": 4.37000160982149e-06,
+      "loss": 0.7462,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2542028750101519,
+      "grad_norm": 4.948964860269852,
+      "learning_rate": 4.369565082606495e-06,
+      "loss": 0.7041,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25428408998619345,
+      "grad_norm": 3.6881446493621657,
+      "learning_rate": 4.369128426027489e-06,
+      "loss": 0.6471,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25436530496223503,
+      "grad_norm": 4.0215825826282625,
+      "learning_rate": 4.36869164011469e-06,
+      "loss": 0.6399,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2544465199382766,
+      "grad_norm": 5.583956863498479,
+      "learning_rate": 4.368254724898319e-06,
+      "loss": 0.7778,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2545277349143182,
+      "grad_norm": 4.313157219895635,
+      "learning_rate": 4.367817680408609e-06,
+      "loss": 0.5685,
+      "step": 3134
+    },
+    {
+      "epoch": 0.25460894989035976,
+      "grad_norm": 5.121719815553165,
+      "learning_rate": 4.3673805066758e-06,
+      "loss": 0.6546,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2546901648664014,
+      "grad_norm": 6.598529582533699,
+      "learning_rate": 4.366943203730144e-06,
+      "loss": 0.5776,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25477137984244297,
+      "grad_norm": 8.179382375263176,
+      "learning_rate": 4.366505771601898e-06,
+      "loss": 0.655,
+      "step": 3137
+    },
+    {
+      "epoch": 0.25485259481848455,
+      "grad_norm": 3.5740269182727085,
+      "learning_rate": 4.366068210321331e-06,
+      "loss": 0.5393,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2549338097945261,
+      "grad_norm": 9.451730618946357,
+      "learning_rate": 4.3656305199187195e-06,
+      "loss": 0.5082,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2550150247705677,
+      "grad_norm": 6.028800369548608,
+      "learning_rate": 4.365192700424351e-06,
+      "loss": 0.6664,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2550962397466093,
+      "grad_norm": 5.858773502768749,
+      "learning_rate": 4.364754751868519e-06,
+      "loss": 0.5164,
+      "step": 3141
+    },
+    {
+      "epoch": 0.25517745472265085,
+      "grad_norm": 7.052601898082844,
+      "learning_rate": 4.364316674281526e-06,
+      "loss": 0.544,
+      "step": 3142
+    },
+    {
+      "epoch": 0.25525866969869243,
+      "grad_norm": 5.076254392341973,
+      "learning_rate": 4.363878467693686e-06,
+      "loss": 0.5765,
+      "step": 3143
+    },
+    {
+      "epoch": 0.255339884674734,
+      "grad_norm": 4.671237955123445,
+      "learning_rate": 4.363440132135322e-06,
+      "loss": 0.6367,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2554210996507756,
+      "grad_norm": 4.135246305916654,
+      "learning_rate": 4.363001667636762e-06,
+      "loss": 0.725,
+      "step": 3145
+    },
+    {
+      "epoch": 0.25550231462681716,
+      "grad_norm": 5.579014444017926,
+      "learning_rate": 4.362563074228346e-06,
+      "loss": 0.5253,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2555835296028588,
+      "grad_norm": 4.386851280267495,
+      "learning_rate": 4.3621243519404235e-06,
+      "loss": 0.5849,
+      "step": 3147
+    },
+    {
+      "epoch": 0.25566474457890037,
+      "grad_norm": 4.068805050530992,
+      "learning_rate": 4.36168550080335e-06,
+      "loss": 0.5252,
+      "step": 3148
+    },
+    {
+      "epoch": 0.25574595955494195,
+      "grad_norm": 5.079228239353849,
+      "learning_rate": 4.361246520847493e-06,
+      "loss": 0.4509,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2558271745309835,
+      "grad_norm": 4.417633075136854,
+      "learning_rate": 4.360807412103228e-06,
+      "loss": 0.5594,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2559083895070251,
+      "grad_norm": 5.457155176446224,
+      "learning_rate": 4.3603681746009374e-06,
+      "loss": 0.5581,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2559896044830667,
+      "grad_norm": 4.311137196331857,
+      "learning_rate": 4.3599288083710155e-06,
+      "loss": 0.5777,
+      "step": 3152
+    },
+    {
+      "epoch": 0.25607081945910826,
+      "grad_norm": 5.779855459554121,
+      "learning_rate": 4.359489313443864e-06,
+      "loss": 0.6154,
+      "step": 3153
+    },
+    {
+      "epoch": 0.25615203443514983,
+      "grad_norm": 5.203623644425137,
+      "learning_rate": 4.359049689849893e-06,
+      "loss": 0.5689,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2562332494111914,
+      "grad_norm": 6.2014101236501995,
+      "learning_rate": 4.358609937619522e-06,
+      "loss": 0.6593,
+      "step": 3155
+    },
+    {
+      "epoch": 0.256314464387233,
+      "grad_norm": 6.085891873335906,
+      "learning_rate": 4.358170056783179e-06,
+      "loss": 0.425,
+      "step": 3156
+    },
+    {
+      "epoch": 0.25639567936327456,
+      "grad_norm": 3.58366241946543,
+      "learning_rate": 4.357730047371304e-06,
+      "loss": 0.538,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2564768943393162,
+      "grad_norm": 5.44533166154409,
+      "learning_rate": 4.357289909414341e-06,
+      "loss": 0.5894,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2565581093153578,
+      "grad_norm": 3.8173019501608625,
+      "learning_rate": 4.356849642942746e-06,
+      "loss": 0.5385,
+      "step": 3159
+    },
+    {
+      "epoch": 0.25663932429139935,
+      "grad_norm": 6.388033240415221,
+      "learning_rate": 4.356409247986982e-06,
+      "loss": 0.5628,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25672053926744093,
+      "grad_norm": 10.182238415534854,
+      "learning_rate": 4.355968724577523e-06,
+      "loss": 0.5962,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2568017542434825,
+      "grad_norm": 3.7342682561919998,
+      "learning_rate": 4.355528072744851e-06,
+      "loss": 0.6925,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2568829692195241,
+      "grad_norm": 3.771086888643871,
+      "learning_rate": 4.355087292519458e-06,
+      "loss": 0.7971,
+      "step": 3163
+    },
+    {
+      "epoch": 0.25696418419556566,
+      "grad_norm": 4.185200079768148,
+      "learning_rate": 4.354646383931841e-06,
+      "loss": 0.6958,
+      "step": 3164
+    },
+    {
+      "epoch": 0.25704539917160724,
+      "grad_norm": 4.773930727509564,
+      "learning_rate": 4.3542053470125104e-06,
+      "loss": 0.5607,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2571266141476488,
+      "grad_norm": 4.658976115697843,
+      "learning_rate": 4.353764181791983e-06,
+      "loss": 0.5175,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2572078291236904,
+      "grad_norm": 6.33815354108174,
+      "learning_rate": 4.353322888300785e-06,
+      "loss": 0.4758,
+      "step": 3167
+    },
+    {
+      "epoch": 0.25728904409973197,
+      "grad_norm": 3.6435625688001423,
+      "learning_rate": 4.3528814665694515e-06,
+      "loss": 0.6159,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2573702590757736,
+      "grad_norm": 3.901416738242754,
+      "learning_rate": 4.352439916628527e-06,
+      "loss": 0.6253,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2574514740518152,
+      "grad_norm": 5.97609923012503,
+      "learning_rate": 4.351998238508563e-06,
+      "loss": 0.4725,
+      "step": 3170
+    },
+    {
+      "epoch": 0.25753268902785675,
+      "grad_norm": 3.836440857096687,
+      "learning_rate": 4.351556432240124e-06,
+      "loss": 0.4923,
+      "step": 3171
+    },
+    {
+      "epoch": 0.25761390400389833,
+      "grad_norm": 6.0800255854901595,
+      "learning_rate": 4.351114497853779e-06,
+      "loss": 0.6401,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2576951189799399,
+      "grad_norm": 5.622546507114739,
+      "learning_rate": 4.350672435380107e-06,
+      "loss": 0.7105,
+      "step": 3173
+    },
+    {
+      "epoch": 0.2577763339559815,
+      "grad_norm": 5.822121836191669,
+      "learning_rate": 4.350230244849697e-06,
+      "loss": 0.6403,
+      "step": 3174
+    },
+    {
+      "epoch": 0.25785754893202306,
+      "grad_norm": 3.5905578627539123,
+      "learning_rate": 4.349787926293146e-06,
+      "loss": 0.6267,
+      "step": 3175
+    },
+    {
+      "epoch": 0.25793876390806464,
+      "grad_norm": 3.4379879993140126,
+      "learning_rate": 4.349345479741062e-06,
+      "loss": 0.7763,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2580199788841062,
+      "grad_norm": 5.296351841281579,
+      "learning_rate": 4.348902905224057e-06,
+      "loss": 0.5434,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2581011938601478,
+      "grad_norm": 4.502659862665315,
+      "learning_rate": 4.348460202772756e-06,
+      "loss": 0.6596,
+      "step": 3178
+    },
+    {
+      "epoch": 0.25818240883618937,
+      "grad_norm": 5.66077638544368,
+      "learning_rate": 4.348017372417792e-06,
+      "loss": 0.4524,
+      "step": 3179
+    },
+    {
+      "epoch": 0.258263623812231,
+      "grad_norm": 6.837754388761909,
+      "learning_rate": 4.347574414189807e-06,
+      "loss": 0.539,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2583448387882726,
+      "grad_norm": 4.065776898335388,
+      "learning_rate": 4.347131328119451e-06,
+      "loss": 0.4425,
+      "step": 3181
+    },
+    {
+      "epoch": 0.25842605376431416,
+      "grad_norm": 6.317177110076131,
+      "learning_rate": 4.346688114237381e-06,
+      "loss": 0.5503,
+      "step": 3182
+    },
+    {
+      "epoch": 0.25850726874035573,
+      "grad_norm": 3.9597286603343207,
+      "learning_rate": 4.346244772574268e-06,
+      "loss": 0.565,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2585884837163973,
+      "grad_norm": 4.617361547979039,
+      "learning_rate": 4.345801303160788e-06,
+      "loss": 0.6364,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2586696986924389,
+      "grad_norm": 3.869309442994426,
+      "learning_rate": 4.3453577060276264e-06,
+      "loss": 0.7814,
+      "step": 3185
+    },
+    {
+      "epoch": 0.25875091366848046,
+      "grad_norm": 5.268139660706938,
+      "learning_rate": 4.344913981205479e-06,
+      "loss": 0.6746,
+      "step": 3186
+    },
+    {
+      "epoch": 0.25883212864452204,
+      "grad_norm": 8.40326725980448,
+      "learning_rate": 4.344470128725047e-06,
+      "loss": 0.5914,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2589133436205636,
+      "grad_norm": 6.871592559950185,
+      "learning_rate": 4.344026148617043e-06,
+      "loss": 0.5718,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2589945585966052,
+      "grad_norm": 5.20815182160167,
+      "learning_rate": 4.343582040912191e-06,
+      "loss": 0.5049,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2590757735726468,
+      "grad_norm": 5.335316024532678,
+      "learning_rate": 4.343137805641217e-06,
+      "loss": 0.6226,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2591569885486884,
+      "grad_norm": 3.8793130716336335,
+      "learning_rate": 4.3426934428348624e-06,
+      "loss": 0.4488,
+      "step": 3191
+    },
+    {
+      "epoch": 0.25923820352473,
+      "grad_norm": 5.127013360030541,
+      "learning_rate": 4.342248952523874e-06,
+      "loss": 0.4591,
+      "step": 3192
+    },
+    {
+      "epoch": 0.25931941850077156,
+      "grad_norm": 4.532500566508718,
+      "learning_rate": 4.341804334739008e-06,
+      "loss": 0.4925,
+      "step": 3193
+    },
+    {
+      "epoch": 0.25940063347681314,
+      "grad_norm": 5.195487587147855,
+      "learning_rate": 4.34135958951103e-06,
+      "loss": 0.5041,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2594818484528547,
+      "grad_norm": 3.983184952506586,
+      "learning_rate": 4.3409147168707124e-06,
+      "loss": 0.5198,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2595630634288963,
+      "grad_norm": 4.96090598003835,
+      "learning_rate": 4.34046971684884e-06,
+      "loss": 0.4985,
+      "step": 3196
+    },
+    {
+      "epoch": 0.25964427840493787,
+      "grad_norm": 4.583631169515384,
+      "learning_rate": 4.340024589476204e-06,
+      "loss": 0.634,
+      "step": 3197
+    },
+    {
+      "epoch": 0.25972549338097944,
+      "grad_norm": 5.545926672500218,
+      "learning_rate": 4.3395793347836034e-06,
+      "loss": 0.4752,
+      "step": 3198
+    },
+    {
+      "epoch": 0.259806708357021,
+      "grad_norm": 6.711460341815773,
+      "learning_rate": 4.33913395280185e-06,
+      "loss": 0.7308,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2598879233330626,
+      "grad_norm": 3.7919381803697276,
+      "learning_rate": 4.33868844356176e-06,
+      "loss": 0.5049,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2599691383091042,
+      "grad_norm": 6.205008626331162,
+      "learning_rate": 4.338242807094161e-06,
+      "loss": 0.6063,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2600503532851458,
+      "grad_norm": 4.694118634605663,
+      "learning_rate": 4.3377970434298885e-06,
+      "loss": 0.6166,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2601315682611874,
+      "grad_norm": 5.435770883892459,
+      "learning_rate": 4.337351152599787e-06,
+      "loss": 0.5616,
+      "step": 3203
+    },
+    {
+      "epoch": 0.26021278323722896,
+      "grad_norm": 3.8614175476698485,
+      "learning_rate": 4.33690513463471e-06,
+      "loss": 0.5997,
+      "step": 3204
+    },
+    {
+      "epoch": 0.26029399821327054,
+      "grad_norm": 5.804567048793658,
+      "learning_rate": 4.336458989565519e-06,
+      "loss": 0.5467,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2603752131893121,
+      "grad_norm": 6.725056494128386,
+      "learning_rate": 4.336012717423085e-06,
+      "loss": 0.4595,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2604564281653537,
+      "grad_norm": 5.3233685676327065,
+      "learning_rate": 4.335566318238289e-06,
+      "loss": 0.5616,
+      "step": 3207
+    },
+    {
+      "epoch": 0.26053764314139527,
+      "grad_norm": 3.4875685060457218,
+      "learning_rate": 4.335119792042017e-06,
+      "loss": 0.5414,
+      "step": 3208
+    },
+    {
+      "epoch": 0.26061885811743685,
+      "grad_norm": 4.305522238103512,
+      "learning_rate": 4.334673138865169e-06,
+      "loss": 0.5734,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2607000730934784,
+      "grad_norm": 6.18157123882514,
+      "learning_rate": 4.334226358738649e-06,
+      "loss": 0.462,
+      "step": 3210
+    },
+    {
+      "epoch": 0.26078128806952,
+      "grad_norm": 14.713830091090161,
+      "learning_rate": 4.333779451693373e-06,
+      "loss": 0.6092,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2608625030455616,
+      "grad_norm": 11.006487008864882,
+      "learning_rate": 4.333332417760263e-06,
+      "loss": 0.4972,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2609437180216032,
+      "grad_norm": 4.400134883259066,
+      "learning_rate": 4.332885256970253e-06,
+      "loss": 0.5909,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2610249329976448,
+      "grad_norm": 5.443017416295305,
+      "learning_rate": 4.332437969354284e-06,
+      "loss": 0.4379,
+      "step": 3214
+    },
+    {
+      "epoch": 0.26110614797368636,
+      "grad_norm": 4.4738174242506235,
+      "learning_rate": 4.331990554943305e-06,
+      "loss": 0.6238,
+      "step": 3215
+    },
+    {
+      "epoch": 0.26118736294972794,
+      "grad_norm": 5.0416535662280655,
+      "learning_rate": 4.331543013768276e-06,
+      "loss": 0.3969,
+      "step": 3216
+    },
+    {
+      "epoch": 0.2612685779257695,
+      "grad_norm": 4.237641745664451,
+      "learning_rate": 4.331095345860162e-06,
+      "loss": 0.4619,
+      "step": 3217
+    },
+    {
+      "epoch": 0.2613497929018111,
+      "grad_norm": 5.09656551194579,
+      "learning_rate": 4.330647551249942e-06,
+      "loss": 0.5224,
+      "step": 3218
+    },
+    {
+      "epoch": 0.2614310078778527,
+      "grad_norm": 4.113097985628701,
+      "learning_rate": 4.330199629968601e-06,
+      "loss": 0.7045,
+      "step": 3219
+    },
+    {
+      "epoch": 0.26151222285389425,
+      "grad_norm": 4.391696991616136,
+      "learning_rate": 4.329751582047132e-06,
+      "loss": 0.4772,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2615934378299358,
+      "grad_norm": 6.284634192395287,
+      "learning_rate": 4.3293034075165355e-06,
+      "loss": 0.7468,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2616746528059774,
+      "grad_norm": 5.31239683535975,
+      "learning_rate": 4.328855106407826e-06,
+      "loss": 0.5164,
+      "step": 3222
+    },
+    {
+      "epoch": 0.261755867782019,
+      "grad_norm": 4.2612158979297945,
+      "learning_rate": 4.328406678752022e-06,
+      "loss": 0.6003,
+      "step": 3223
+    },
+    {
+      "epoch": 0.2618370827580606,
+      "grad_norm": 5.243856895565067,
+      "learning_rate": 4.3279581245801515e-06,
+      "loss": 0.5995,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2619182977341022,
+      "grad_norm": 5.3541602152619046,
+      "learning_rate": 4.327509443923254e-06,
+      "loss": 0.4661,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26199951271014377,
+      "grad_norm": 5.615300922143262,
+      "learning_rate": 4.327060636812375e-06,
+      "loss": 0.634,
+      "step": 3226
+    },
+    {
+      "epoch": 0.26208072768618534,
+      "grad_norm": 5.152013169697966,
+      "learning_rate": 4.32661170327857e-06,
+      "loss": 0.6963,
+      "step": 3227
+    },
+    {
+      "epoch": 0.2621619426622269,
+      "grad_norm": 5.2073960016961545,
+      "learning_rate": 4.326162643352901e-06,
+      "loss": 0.7672,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2622431576382685,
+      "grad_norm": 5.909318364895728,
+      "learning_rate": 4.325713457066443e-06,
+      "loss": 0.6173,
+      "step": 3229
+    },
+    {
+      "epoch": 0.2623243726143101,
+      "grad_norm": 16.26621162135915,
+      "learning_rate": 4.325264144450276e-06,
+      "loss": 0.3936,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26240558759035165,
+      "grad_norm": 4.238425289228819,
+      "learning_rate": 4.324814705535491e-06,
+      "loss": 0.5425,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26248680256639323,
+      "grad_norm": 7.639214601519878,
+      "learning_rate": 4.324365140353185e-06,
+      "loss": 0.6787,
+      "step": 3232
+    },
+    {
+      "epoch": 0.2625680175424348,
+      "grad_norm": 4.776323919697988,
+      "learning_rate": 4.323915448934466e-06,
+      "loss": 0.7179,
+      "step": 3233
+    },
+    {
+      "epoch": 0.2626492325184764,
+      "grad_norm": 4.829079791949981,
+      "learning_rate": 4.323465631310452e-06,
+      "loss": 0.5759,
+      "step": 3234
+    },
+    {
+      "epoch": 0.262730447494518,
+      "grad_norm": 5.074740707653109,
+      "learning_rate": 4.323015687512267e-06,
+      "loss": 0.5894,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2628116624705596,
+      "grad_norm": 9.615050714429785,
+      "learning_rate": 4.322565617571044e-06,
+      "loss": 0.5466,
+      "step": 3236
+    },
+    {
+      "epoch": 0.26289287744660117,
+      "grad_norm": 5.123025702283434,
+      "learning_rate": 4.322115421517926e-06,
+      "loss": 0.5913,
+      "step": 3237
+    },
+    {
+      "epoch": 0.26297409242264275,
+      "grad_norm": 4.751136267011921,
+      "learning_rate": 4.321665099384064e-06,
+      "loss": 0.4789,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2630553073986843,
+      "grad_norm": 10.947055513952483,
+      "learning_rate": 4.321214651200619e-06,
+      "loss": 0.5118,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2631365223747259,
+      "grad_norm": 5.687786414091623,
+      "learning_rate": 4.320764076998759e-06,
+      "loss": 0.6006,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2632177373507675,
+      "grad_norm": 4.846530548806665,
+      "learning_rate": 4.32031337680966e-06,
+      "loss": 0.6819,
+      "step": 3241
+    },
+    {
+      "epoch": 0.26329895232680905,
+      "grad_norm": 4.29992728461369,
+      "learning_rate": 4.31986255066451e-06,
+      "loss": 0.6645,
+      "step": 3242
+    },
+    {
+      "epoch": 0.26338016730285063,
+      "grad_norm": 4.728540495274627,
+      "learning_rate": 4.319411598594503e-06,
+      "loss": 0.65,
+      "step": 3243
+    },
+    {
+      "epoch": 0.2634613822788922,
+      "grad_norm": 4.044743720033708,
+      "learning_rate": 4.318960520630842e-06,
+      "loss": 0.5344,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2635425972549338,
+      "grad_norm": 4.062414651992789,
+      "learning_rate": 4.3185093168047395e-06,
+      "loss": 0.6314,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2636238122309754,
+      "grad_norm": 5.77814635308223,
+      "learning_rate": 4.318057987147418e-06,
+      "loss": 0.5496,
+      "step": 3246
+    },
+    {
+      "epoch": 0.263705027207017,
+      "grad_norm": 8.868723303386513,
+      "learning_rate": 4.317606531690104e-06,
+      "loss": 0.5484,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2637862421830586,
+      "grad_norm": 5.931221128514847,
+      "learning_rate": 4.317154950464039e-06,
+      "loss": 0.5311,
+      "step": 3248
+    },
+    {
+      "epoch": 0.26386745715910015,
+      "grad_norm": 5.582327237175479,
+      "learning_rate": 4.316703243500467e-06,
+      "loss": 0.532,
+      "step": 3249
+    },
+    {
+      "epoch": 0.2639486721351417,
+      "grad_norm": 3.5277983786023235,
+      "learning_rate": 4.3162514108306465e-06,
+      "loss": 0.7615,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2640298871111833,
+      "grad_norm": 6.701677321017773,
+      "learning_rate": 4.315799452485841e-06,
+      "loss": 0.4753,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2641111020872249,
+      "grad_norm": 4.478895629503352,
+      "learning_rate": 4.3153473684973226e-06,
+      "loss": 0.5828,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26419231706326646,
+      "grad_norm": 6.992879628361261,
+      "learning_rate": 4.314895158896374e-06,
+      "loss": 0.5683,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26427353203930803,
+      "grad_norm": 4.622185359910068,
+      "learning_rate": 4.314442823714286e-06,
+      "loss": 0.4762,
+      "step": 3254
+    },
+    {
+      "epoch": 0.2643547470153496,
+      "grad_norm": 3.3396459737640587,
+      "learning_rate": 4.313990362982357e-06,
+      "loss": 0.7585,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2644359619913912,
+      "grad_norm": 9.71525006135501,
+      "learning_rate": 4.313537776731895e-06,
+      "loss": 0.4599,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2645171769674328,
+      "grad_norm": 5.997259305627276,
+      "learning_rate": 4.313085064994218e-06,
+      "loss": 0.59,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2645983919434744,
+      "grad_norm": 3.4896157532430285,
+      "learning_rate": 4.3126322278006496e-06,
+      "loss": 0.6535,
+      "step": 3258
+    },
+    {
+      "epoch": 0.264679606919516,
+      "grad_norm": 3.163990271513905,
+      "learning_rate": 4.312179265182523e-06,
+      "loss": 0.4253,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26476082189555755,
+      "grad_norm": 4.468139415385732,
+      "learning_rate": 4.311726177171184e-06,
+      "loss": 0.5902,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26484203687159913,
+      "grad_norm": 3.7893680396692053,
+      "learning_rate": 4.311272963797981e-06,
+      "loss": 0.5972,
+      "step": 3261
+    },
+    {
+      "epoch": 0.2649232518476407,
+      "grad_norm": 5.50109202559223,
+      "learning_rate": 4.3108196250942746e-06,
+      "loss": 0.4875,
+      "step": 3262
+    },
+    {
+      "epoch": 0.2650044668236823,
+      "grad_norm": 10.031595831871455,
+      "learning_rate": 4.310366161091435e-06,
+      "loss": 0.7109,
+      "step": 3263
+    },
+    {
+      "epoch": 0.26508568179972386,
+      "grad_norm": 5.5919773594492,
+      "learning_rate": 4.309912571820837e-06,
+      "loss": 0.521,
+      "step": 3264
+    },
+    {
+      "epoch": 0.26516689677576544,
+      "grad_norm": 4.444420349163222,
+      "learning_rate": 4.309458857313868e-06,
+      "loss": 0.6694,
+      "step": 3265
+    },
+    {
+      "epoch": 0.265248111751807,
+      "grad_norm": 3.8949106613348947,
+      "learning_rate": 4.309005017601924e-06,
+      "loss": 0.6129,
+      "step": 3266
+    },
+    {
+      "epoch": 0.2653293267278486,
+      "grad_norm": 3.3059067568093847,
+      "learning_rate": 4.308551052716406e-06,
+      "loss": 0.6001,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2654105417038902,
+      "grad_norm": 5.114398735836412,
+      "learning_rate": 4.308096962688726e-06,
+      "loss": 0.5416,
+      "step": 3268
+    },
+    {
+      "epoch": 0.2654917566799318,
+      "grad_norm": 3.1154319915363855,
+      "learning_rate": 4.307642747550306e-06,
+      "loss": 0.5931,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2655729716559734,
+      "grad_norm": 4.307648831780373,
+      "learning_rate": 4.307188407332574e-06,
+      "loss": 0.5852,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26565418663201495,
+      "grad_norm": 4.424486660650266,
+      "learning_rate": 4.306733942066969e-06,
+      "loss": 0.5399,
+      "step": 3271
+    },
+    {
+      "epoch": 0.26573540160805653,
+      "grad_norm": 6.036748818342054,
+      "learning_rate": 4.306279351784938e-06,
+      "loss": 0.4185,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2658166165840981,
+      "grad_norm": 3.6762274889278195,
+      "learning_rate": 4.305824636517935e-06,
+      "loss": 0.4484,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2658978315601397,
+      "grad_norm": 4.098602260190931,
+      "learning_rate": 4.305369796297424e-06,
+      "loss": 0.4614,
+      "step": 3274
+    },
+    {
+      "epoch": 0.26597904653618126,
+      "grad_norm": 5.298237691361923,
+      "learning_rate": 4.3049148311548785e-06,
+      "loss": 0.5971,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26606026151222284,
+      "grad_norm": 5.05302699787098,
+      "learning_rate": 4.304459741121778e-06,
+      "loss": 0.4597,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2661414764882644,
+      "grad_norm": 4.995834340661036,
+      "learning_rate": 4.304004526229614e-06,
+      "loss": 0.5854,
+      "step": 3277
+    },
+    {
+      "epoch": 0.266222691464306,
+      "grad_norm": 2.5777374640004838,
+      "learning_rate": 4.303549186509884e-06,
+      "loss": 0.498,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2663039064403476,
+      "grad_norm": 3.390125198708796,
+      "learning_rate": 4.303093721994096e-06,
+      "loss": 0.4985,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2663851214163892,
+      "grad_norm": 5.071175646539791,
+      "learning_rate": 4.302638132713766e-06,
+      "loss": 0.5427,
+      "step": 3280
+    },
+    {
+      "epoch": 0.2664663363924308,
+      "grad_norm": 4.587539332384161,
+      "learning_rate": 4.302182418700415e-06,
+      "loss": 0.4533,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26654755136847236,
+      "grad_norm": 10.760547007231233,
+      "learning_rate": 4.301726579985581e-06,
+      "loss": 0.5711,
+      "step": 3282
+    },
+    {
+      "epoch": 0.26662876634451393,
+      "grad_norm": 6.241807773314562,
+      "learning_rate": 4.301270616600802e-06,
+      "loss": 0.7594,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2667099813205555,
+      "grad_norm": 5.23430106311079,
+      "learning_rate": 4.30081452857763e-06,
+      "loss": 0.5922,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2667911962965971,
+      "grad_norm": 5.314419566343465,
+      "learning_rate": 4.300358315947622e-06,
+      "loss": 0.6298,
+      "step": 3285
+    },
+    {
+      "epoch": 0.26687241127263867,
+      "grad_norm": 5.02528385387479,
+      "learning_rate": 4.299901978742349e-06,
+      "loss": 0.5465,
+      "step": 3286
+    },
+    {
+      "epoch": 0.26695362624868024,
+      "grad_norm": 4.18489447765815,
+      "learning_rate": 4.2994455169933835e-06,
+      "loss": 0.5514,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2670348412247218,
+      "grad_norm": 4.050947136506907,
+      "learning_rate": 4.298988930732312e-06,
+      "loss": 0.6557,
+      "step": 3288
+    },
+    {
+      "epoch": 0.2671160562007634,
+      "grad_norm": 4.82281153344797,
+      "learning_rate": 4.2985322199907275e-06,
+      "loss": 0.4805,
+      "step": 3289
+    },
+    {
+      "epoch": 0.26719727117680503,
+      "grad_norm": 3.2786394296234787,
+      "learning_rate": 4.298075384800232e-06,
+      "loss": 0.7165,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2672784861528466,
+      "grad_norm": 4.134763536721899,
+      "learning_rate": 4.297618425192436e-06,
+      "loss": 0.6281,
+      "step": 3291
+    },
+    {
+      "epoch": 0.2673597011288882,
+      "grad_norm": 5.393028330188706,
+      "learning_rate": 4.297161341198957e-06,
+      "loss": 0.4522,
+      "step": 3292
+    },
+    {
+      "epoch": 0.26744091610492976,
+      "grad_norm": 3.3337915433920178,
+      "learning_rate": 4.296704132851427e-06,
+      "loss": 0.5315,
+      "step": 3293
+    },
+    {
+      "epoch": 0.26752213108097134,
+      "grad_norm": 4.494608257605046,
+      "learning_rate": 4.296246800181479e-06,
+      "loss": 0.594,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2676033460570129,
+      "grad_norm": 6.986279435908919,
+      "learning_rate": 4.29578934322076e-06,
+      "loss": 0.4113,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2676845610330545,
+      "grad_norm": 4.754454279458852,
+      "learning_rate": 4.295331762000921e-06,
+      "loss": 0.656,
+      "step": 3296
+    },
+    {
+      "epoch": 0.26776577600909607,
+      "grad_norm": 4.557941560884389,
+      "learning_rate": 4.294874056553626e-06,
+      "loss": 0.6003,
+      "step": 3297
+    },
+    {
+      "epoch": 0.26784699098513765,
+      "grad_norm": 3.5253434081641593,
+      "learning_rate": 4.294416226910546e-06,
+      "loss": 0.6373,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2679282059611792,
+      "grad_norm": 5.390157405133369,
+      "learning_rate": 4.2939582731033605e-06,
+      "loss": 0.5665,
+      "step": 3299
+    },
+    {
+      "epoch": 0.2680094209372208,
+      "grad_norm": 4.444716348064678,
+      "learning_rate": 4.293500195163756e-06,
+      "loss": 0.4936,
+      "step": 3300
+    },
+    {
+      "epoch": 0.26809063591326243,
+      "grad_norm": 4.929185123300809,
+      "learning_rate": 4.29304199312343e-06,
+      "loss": 0.6039,
+      "step": 3301
+    },
+    {
+      "epoch": 0.268171850889304,
+      "grad_norm": 5.005771469258986,
+      "learning_rate": 4.292583667014087e-06,
+      "loss": 0.6245,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2682530658653456,
+      "grad_norm": 6.794918253939256,
+      "learning_rate": 4.292125216867443e-06,
+      "loss": 0.4753,
+      "step": 3303
+    },
+    {
+      "epoch": 0.26833428084138716,
+      "grad_norm": 4.918313944683376,
+      "learning_rate": 4.2916666427152175e-06,
+      "loss": 0.6796,
+      "step": 3304
+    },
+    {
+      "epoch": 0.26841549581742874,
+      "grad_norm": 5.184653793095706,
+      "learning_rate": 4.291207944589143e-06,
+      "loss": 0.5299,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2684967107934703,
+      "grad_norm": 5.158508970573285,
+      "learning_rate": 4.290749122520959e-06,
+      "loss": 0.5115,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2685779257695119,
+      "grad_norm": 5.019559380014884,
+      "learning_rate": 4.290290176542412e-06,
+      "loss": 0.6411,
+      "step": 3307
+    },
+    {
+      "epoch": 0.26865914074555347,
+      "grad_norm": 5.867738157836222,
+      "learning_rate": 4.289831106685261e-06,
+      "loss": 0.5529,
+      "step": 3308
+    },
+    {
+      "epoch": 0.26874035572159505,
+      "grad_norm": 4.019129715439323,
+      "learning_rate": 4.289371912981268e-06,
+      "loss": 0.5655,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2688215706976366,
+      "grad_norm": 4.959547313791584,
+      "learning_rate": 4.28891259546221e-06,
+      "loss": 0.7881,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2689027856736782,
+      "grad_norm": 3.3894537647637466,
+      "learning_rate": 4.288453154159869e-06,
+      "loss": 0.6476,
+      "step": 3311
+    },
+    {
+      "epoch": 0.26898400064971983,
+      "grad_norm": 4.576484804759739,
+      "learning_rate": 4.287993589106034e-06,
+      "loss": 0.4195,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2690652156257614,
+      "grad_norm": 5.5159598677639075,
+      "learning_rate": 4.287533900332506e-06,
+      "loss": 0.5731,
+      "step": 3313
+    },
+    {
+      "epoch": 0.269146430601803,
+      "grad_norm": 3.476369390277886,
+      "learning_rate": 4.287074087871092e-06,
+      "loss": 0.7422,
+      "step": 3314
+    },
+    {
+      "epoch": 0.26922764557784457,
+      "grad_norm": 4.3872226024253305,
+      "learning_rate": 4.2866141517536085e-06,
+      "loss": 0.5666,
+      "step": 3315
+    },
+    {
+      "epoch": 0.26930886055388614,
+      "grad_norm": 4.1888189167131795,
+      "learning_rate": 4.286154092011882e-06,
+      "loss": 0.6373,
+      "step": 3316
+    },
+    {
+      "epoch": 0.2693900755299277,
+      "grad_norm": 6.277380300582101,
+      "learning_rate": 4.285693908677746e-06,
+      "loss": 0.5566,
+      "step": 3317
+    },
+    {
+      "epoch": 0.2694712905059693,
+      "grad_norm": 3.4248947125674802,
+      "learning_rate": 4.285233601783041e-06,
+      "loss": 0.7619,
+      "step": 3318
+    },
+    {
+      "epoch": 0.2695525054820109,
+      "grad_norm": 9.083452534210942,
+      "learning_rate": 4.28477317135962e-06,
+      "loss": 0.441,
+      "step": 3319
+    },
+    {
+      "epoch": 0.26963372045805245,
+      "grad_norm": 5.194205565045371,
+      "learning_rate": 4.28431261743934e-06,
+      "loss": 0.5573,
+      "step": 3320
+    },
+    {
+      "epoch": 0.269714935434094,
+      "grad_norm": 4.678152409953697,
+      "learning_rate": 4.2838519400540715e-06,
+      "loss": 0.5867,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2697961504101356,
+      "grad_norm": 6.079974041201289,
+      "learning_rate": 4.283391139235688e-06,
+      "loss": 0.5913,
+      "step": 3322
+    },
+    {
+      "epoch": 0.26987736538617724,
+      "grad_norm": 4.43636337967898,
+      "learning_rate": 4.282930215016078e-06,
+      "loss": 0.4492,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2699585803622188,
+      "grad_norm": 5.658242363279942,
+      "learning_rate": 4.282469167427132e-06,
+      "loss": 0.7522,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2700397953382604,
+      "grad_norm": 6.996911479616853,
+      "learning_rate": 4.2820079965007545e-06,
+      "loss": 0.5871,
+      "step": 3325
+    },
+    {
+      "epoch": 0.27012101031430197,
+      "grad_norm": 4.53448881154164,
+      "learning_rate": 4.281546702268853e-06,
+      "loss": 0.3827,
+      "step": 3326
+    },
+    {
+      "epoch": 0.27020222529034355,
+      "grad_norm": 6.424034548623486,
+      "learning_rate": 4.28108528476335e-06,
+      "loss": 0.3635,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2702834402663851,
+      "grad_norm": 4.466125493504516,
+      "learning_rate": 4.280623744016171e-06,
+      "loss": 0.5539,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2703646552424267,
+      "grad_norm": 3.693236745360508,
+      "learning_rate": 4.280162080059252e-06,
+      "loss": 0.5083,
+      "step": 3329
+    },
+    {
+      "epoch": 0.2704458702184683,
+      "grad_norm": 7.92497787565039,
+      "learning_rate": 4.279700292924539e-06,
+      "loss": 0.6099,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27052708519450985,
+      "grad_norm": 6.126906391884235,
+      "learning_rate": 4.279238382643985e-06,
+      "loss": 0.4806,
+      "step": 3331
+    },
+    {
+      "epoch": 0.27060830017055143,
+      "grad_norm": 10.147235977729743,
+      "learning_rate": 4.278776349249551e-06,
+      "loss": 0.5868,
+      "step": 3332
+    },
+    {
+      "epoch": 0.270689515146593,
+      "grad_norm": 4.694547184635383,
+      "learning_rate": 4.278314192773208e-06,
+      "loss": 0.5579,
+      "step": 3333
+    },
+    {
+      "epoch": 0.27077073012263464,
+      "grad_norm": 5.790255949685903,
+      "learning_rate": 4.277851913246934e-06,
+      "loss": 0.5259,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2708519450986762,
+      "grad_norm": 4.610604823465653,
+      "learning_rate": 4.277389510702717e-06,
+      "loss": 0.5614,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2709331600747178,
+      "grad_norm": 6.063379795121963,
+      "learning_rate": 4.276926985172553e-06,
+      "loss": 0.5133,
+      "step": 3336
+    },
+    {
+      "epoch": 0.27101437505075937,
+      "grad_norm": 4.576106292327785,
+      "learning_rate": 4.276464336688445e-06,
+      "loss": 0.4748,
+      "step": 3337
+    },
+    {
+      "epoch": 0.27109559002680095,
+      "grad_norm": 5.762207004631984,
+      "learning_rate": 4.2760015652824074e-06,
+      "loss": 0.7092,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2711768050028425,
+      "grad_norm": 4.920850361695753,
+      "learning_rate": 4.27553867098646e-06,
+      "loss": 0.4025,
+      "step": 3339
+    },
+    {
+      "epoch": 0.2712580199788841,
+      "grad_norm": 4.965952642074921,
+      "learning_rate": 4.275075653832635e-06,
+      "loss": 0.5994,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2713392349549257,
+      "grad_norm": 6.661938259420609,
+      "learning_rate": 4.274612513852968e-06,
+      "loss": 0.6148,
+      "step": 3341
+    },
+    {
+      "epoch": 0.27142044993096726,
+      "grad_norm": 4.176847484212596,
+      "learning_rate": 4.274149251079507e-06,
+      "loss": 0.4974,
+      "step": 3342
+    },
+    {
+      "epoch": 0.27150166490700883,
+      "grad_norm": 9.324003992934651,
+      "learning_rate": 4.273685865544308e-06,
+      "loss": 0.518,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2715828798830504,
+      "grad_norm": 10.84005640144426,
+      "learning_rate": 4.273222357279434e-06,
+      "loss": 0.5259,
+      "step": 3344
+    },
+    {
+      "epoch": 0.27166409485909204,
+      "grad_norm": 4.583615144085001,
+      "learning_rate": 4.272758726316958e-06,
+      "loss": 0.5688,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2717453098351336,
+      "grad_norm": 5.704188407998161,
+      "learning_rate": 4.272294972688959e-06,
+      "loss": 0.7256,
+      "step": 3346
+    },
+    {
+      "epoch": 0.2718265248111752,
+      "grad_norm": 5.611046782367686,
+      "learning_rate": 4.2718310964275285e-06,
+      "loss": 0.4029,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2719077397872168,
+      "grad_norm": 3.432494651225952,
+      "learning_rate": 4.271367097564763e-06,
+      "loss": 0.5548,
+      "step": 3348
+    },
+    {
+      "epoch": 0.27198895476325835,
+      "grad_norm": 5.993392392105414,
+      "learning_rate": 4.27090297613277e-06,
+      "loss": 0.4336,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2720701697392999,
+      "grad_norm": 5.973550784862252,
+      "learning_rate": 4.270438732163663e-06,
+      "loss": 0.5539,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2721513847153415,
+      "grad_norm": 7.021609581750342,
+      "learning_rate": 4.269974365689565e-06,
+      "loss": 0.5499,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2722325996913831,
+      "grad_norm": 4.0778999962904106,
+      "learning_rate": 4.269509876742609e-06,
+      "loss": 0.5756,
+      "step": 3352
+    },
+    {
+      "epoch": 0.27231381466742466,
+      "grad_norm": 4.339699329280206,
+      "learning_rate": 4.269045265354935e-06,
+      "loss": 0.6475,
+      "step": 3353
+    },
+    {
+      "epoch": 0.27239502964346624,
+      "grad_norm": 3.557598291542795,
+      "learning_rate": 4.26858053155869e-06,
+      "loss": 0.6789,
+      "step": 3354
+    },
+    {
+      "epoch": 0.2724762446195078,
+      "grad_norm": 5.638513919453793,
+      "learning_rate": 4.268115675386033e-06,
+      "loss": 0.5661,
+      "step": 3355
+    },
+    {
+      "epoch": 0.27255745959554945,
+      "grad_norm": 7.007916663451633,
+      "learning_rate": 4.267650696869129e-06,
+      "loss": 0.6213,
+      "step": 3356
+    },
+    {
+      "epoch": 0.272638674571591,
+      "grad_norm": 6.707656670971426,
+      "learning_rate": 4.267185596040152e-06,
+      "loss": 0.6185,
+      "step": 3357
+    },
+    {
+      "epoch": 0.2727198895476326,
+      "grad_norm": 5.468221363182492,
+      "learning_rate": 4.266720372931285e-06,
+      "loss": 0.6918,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2728011045236742,
+      "grad_norm": 3.4031280276219826,
+      "learning_rate": 4.2662550275747175e-06,
+      "loss": 0.4968,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27288231949971575,
+      "grad_norm": 3.7289743320117914,
+      "learning_rate": 4.26578956000265e-06,
+      "loss": 0.5128,
+      "step": 3360
+    },
+    {
+      "epoch": 0.27296353447575733,
+      "grad_norm": 4.058327224361656,
+      "learning_rate": 4.26532397024729e-06,
+      "loss": 0.5662,
+      "step": 3361
+    },
+    {
+      "epoch": 0.2730447494517989,
+      "grad_norm": 6.483948262408605,
+      "learning_rate": 4.264858258340854e-06,
+      "loss": 0.5985,
+      "step": 3362
+    },
+    {
+      "epoch": 0.2731259644278405,
+      "grad_norm": 6.7970175715192065,
+      "learning_rate": 4.264392424315568e-06,
+      "loss": 0.5611,
+      "step": 3363
+    },
+    {
+      "epoch": 0.27320717940388206,
+      "grad_norm": 4.719015734535453,
+      "learning_rate": 4.263926468203663e-06,
+      "loss": 0.5915,
+      "step": 3364
+    },
+    {
+      "epoch": 0.27328839437992364,
+      "grad_norm": 5.439068319982008,
+      "learning_rate": 4.2634603900373825e-06,
+      "loss": 0.6571,
+      "step": 3365
+    },
+    {
+      "epoch": 0.2733696093559652,
+      "grad_norm": 9.218141495545822,
+      "learning_rate": 4.262994189848976e-06,
+      "loss": 0.5393,
+      "step": 3366
+    },
+    {
+      "epoch": 0.27345082433200685,
+      "grad_norm": 4.307338467347387,
+      "learning_rate": 4.262527867670702e-06,
+      "loss": 0.5521,
+      "step": 3367
+    },
+    {
+      "epoch": 0.2735320393080484,
+      "grad_norm": 4.015516448961165,
+      "learning_rate": 4.2620614235348265e-06,
+      "loss": 0.4343,
+      "step": 3368
+    },
+    {
+      "epoch": 0.27361325428409,
+      "grad_norm": 5.503396831053377,
+      "learning_rate": 4.261594857473628e-06,
+      "loss": 0.4884,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2736944692601316,
+      "grad_norm": 5.761220256833325,
+      "learning_rate": 4.261128169519388e-06,
+      "loss": 0.5385,
+      "step": 3370
+    },
+    {
+      "epoch": 0.27377568423617316,
+      "grad_norm": 7.751355969511051,
+      "learning_rate": 4.2606613597043975e-06,
+      "loss": 0.5659,
+      "step": 3371
+    },
+    {
+      "epoch": 0.27385689921221473,
+      "grad_norm": 4.671040413073545,
+      "learning_rate": 4.260194428060961e-06,
+      "loss": 0.4999,
+      "step": 3372
+    },
+    {
+      "epoch": 0.2739381141882563,
+      "grad_norm": 7.901394405330174,
+      "learning_rate": 4.2597273746213855e-06,
+      "loss": 0.6527,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2740193291642979,
+      "grad_norm": 6.603352470726141,
+      "learning_rate": 4.259260199417988e-06,
+      "loss": 0.5473,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27410054414033946,
+      "grad_norm": 5.40024658636952,
+      "learning_rate": 4.2587929024830964e-06,
+      "loss": 0.5654,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27418175911638104,
+      "grad_norm": 9.759841121522395,
+      "learning_rate": 4.258325483849044e-06,
+      "loss": 0.6142,
+      "step": 3376
+    },
+    {
+      "epoch": 0.2742629740924226,
+      "grad_norm": 5.125699081398096,
+      "learning_rate": 4.257857943548173e-06,
+      "loss": 0.5213,
+      "step": 3377
+    },
+    {
+      "epoch": 0.27434418906846425,
+      "grad_norm": 4.34028312916136,
+      "learning_rate": 4.257390281612837e-06,
+      "loss": 0.4747,
+      "step": 3378
+    },
+    {
+      "epoch": 0.2744254040445058,
+      "grad_norm": 4.263312229046128,
+      "learning_rate": 4.256922498075394e-06,
+      "loss": 0.5938,
+      "step": 3379
+    },
+    {
+      "epoch": 0.2745066190205474,
+      "grad_norm": 4.797086281941267,
+      "learning_rate": 4.256454592968212e-06,
+      "loss": 0.4205,
+      "step": 3380
+    },
+    {
+      "epoch": 0.274587833996589,
+      "grad_norm": 7.033829432880135,
+      "learning_rate": 4.255986566323668e-06,
+      "loss": 0.6218,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27466904897263056,
+      "grad_norm": 7.8902900782976095,
+      "learning_rate": 4.255518418174148e-06,
+      "loss": 0.5008,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27475026394867214,
+      "grad_norm": 3.2994006635611655,
+      "learning_rate": 4.2550501485520445e-06,
+      "loss": 0.694,
+      "step": 3383
+    },
+    {
+      "epoch": 0.2748314789247137,
+      "grad_norm": 5.401884002359012,
+      "learning_rate": 4.254581757489758e-06,
+      "loss": 0.5462,
+      "step": 3384
+    },
+    {
+      "epoch": 0.2749126939007553,
+      "grad_norm": 11.938262897785561,
+      "learning_rate": 4.254113245019701e-06,
+      "loss": 0.5148,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27499390887679687,
+      "grad_norm": 6.8820894517380555,
+      "learning_rate": 4.25364461117429e-06,
+      "loss": 0.6357,
+      "step": 3386
+    },
+    {
+      "epoch": 0.27507512385283844,
+      "grad_norm": 3.919541658984626,
+      "learning_rate": 4.2531758559859535e-06,
+      "loss": 0.609,
+      "step": 3387
+    },
+    {
+      "epoch": 0.27515633882888,
+      "grad_norm": 3.9922271367677475,
+      "learning_rate": 4.252706979487127e-06,
+      "loss": 0.5394,
+      "step": 3388
+    },
+    {
+      "epoch": 0.27523755380492165,
+      "grad_norm": 4.307590756579932,
+      "learning_rate": 4.2522379817102525e-06,
+      "loss": 0.4479,
+      "step": 3389
+    },
+    {
+      "epoch": 0.27531876878096323,
+      "grad_norm": 4.6485314568936715,
+      "learning_rate": 4.251768862687783e-06,
+      "loss": 0.5758,
+      "step": 3390
+    },
+    {
+      "epoch": 0.2753999837570048,
+      "grad_norm": 3.219961667439573,
+      "learning_rate": 4.25129962245218e-06,
+      "loss": 0.5229,
+      "step": 3391
+    },
+    {
+      "epoch": 0.2754811987330464,
+      "grad_norm": 3.6351810295222955,
+      "learning_rate": 4.250830261035911e-06,
+      "loss": 0.5625,
+      "step": 3392
+    },
+    {
+      "epoch": 0.27556241370908796,
+      "grad_norm": 5.369015436096168,
+      "learning_rate": 4.250360778471455e-06,
+      "loss": 0.6667,
+      "step": 3393
+    },
+    {
+      "epoch": 0.27564362868512954,
+      "grad_norm": 4.427072372118147,
+      "learning_rate": 4.249891174791297e-06,
+      "loss": 0.6424,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2757248436611711,
+      "grad_norm": 5.994983906902898,
+      "learning_rate": 4.249421450027929e-06,
+      "loss": 0.6288,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2758060586372127,
+      "grad_norm": 5.740886381814397,
+      "learning_rate": 4.248951604213858e-06,
+      "loss": 0.6017,
+      "step": 3396
+    },
+    {
+      "epoch": 0.27588727361325427,
+      "grad_norm": 6.578063566695588,
+      "learning_rate": 4.24848163738159e-06,
+      "loss": 0.5481,
+      "step": 3397
+    },
+    {
+      "epoch": 0.27596848858929585,
+      "grad_norm": 4.114106520635557,
+      "learning_rate": 4.248011549563647e-06,
+      "loss": 0.601,
+      "step": 3398
+    },
+    {
+      "epoch": 0.2760497035653374,
+      "grad_norm": 5.445436802534245,
+      "learning_rate": 4.247541340792557e-06,
+      "loss": 0.5067,
+      "step": 3399
+    },
+    {
+      "epoch": 0.27613091854137906,
+      "grad_norm": 5.480238969386964,
+      "learning_rate": 4.247071011100855e-06,
+      "loss": 0.4474,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27621213351742063,
+      "grad_norm": 3.9075643825539865,
+      "learning_rate": 4.246600560521084e-06,
+      "loss": 0.5349,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2762933484934622,
+      "grad_norm": 4.585770407486754,
+      "learning_rate": 4.246129989085798e-06,
+      "loss": 0.6055,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2763745634695038,
+      "grad_norm": 6.06730610156998,
+      "learning_rate": 4.245659296827559e-06,
+      "loss": 0.5557,
+      "step": 3403
+    },
+    {
+      "epoch": 0.27645577844554536,
+      "grad_norm": 4.632787296567119,
+      "learning_rate": 4.245188483778935e-06,
+      "loss": 0.5214,
+      "step": 3404
+    },
+    {
+      "epoch": 0.27653699342158694,
+      "grad_norm": 8.317514548088848,
+      "learning_rate": 4.244717549972504e-06,
+      "loss": 0.5625,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2766182083976285,
+      "grad_norm": 3.6101896303329255,
+      "learning_rate": 4.2442464954408524e-06,
+      "loss": 0.5844,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2766994233736701,
+      "grad_norm": 4.099925210417941,
+      "learning_rate": 4.243775320216575e-06,
+      "loss": 0.5043,
+      "step": 3407
+    },
+    {
+      "epoch": 0.27678063834971167,
+      "grad_norm": 5.864792263094429,
+      "learning_rate": 4.243304024332273e-06,
+      "loss": 0.5581,
+      "step": 3408
+    },
+    {
+      "epoch": 0.27686185332575325,
+      "grad_norm": 4.148641967222804,
+      "learning_rate": 4.24283260782056e-06,
+      "loss": 0.7032,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2769430683017948,
+      "grad_norm": 5.951997919915405,
+      "learning_rate": 4.2423610707140545e-06,
+      "loss": 0.4561,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27702428327783646,
+      "grad_norm": 3.9198889092781233,
+      "learning_rate": 4.241889413045384e-06,
+      "loss": 0.6652,
+      "step": 3411
+    },
+    {
+      "epoch": 0.27710549825387804,
+      "grad_norm": 4.4203783957213725,
+      "learning_rate": 4.2414176348471845e-06,
+      "loss": 0.7011,
+      "step": 3412
+    },
+    {
+      "epoch": 0.2771867132299196,
+      "grad_norm": 2.6236418539062303,
+      "learning_rate": 4.240945736152101e-06,
+      "loss": 0.5444,
+      "step": 3413
+    },
+    {
+      "epoch": 0.2772679282059612,
+      "grad_norm": 3.283626852296422,
+      "learning_rate": 4.240473716992786e-06,
+      "loss": 0.5089,
+      "step": 3414
+    },
+    {
+      "epoch": 0.27734914318200277,
+      "grad_norm": 4.388557235644921,
+      "learning_rate": 4.240001577401903e-06,
+      "loss": 0.6291,
+      "step": 3415
+    },
+    {
+      "epoch": 0.27743035815804434,
+      "grad_norm": 5.882384300446532,
+      "learning_rate": 4.239529317412118e-06,
+      "loss": 0.3563,
+      "step": 3416
+    },
+    {
+      "epoch": 0.2775115731340859,
+      "grad_norm": 5.663696285420261,
+      "learning_rate": 4.239056937056111e-06,
+      "loss": 0.4948,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2775927881101275,
+      "grad_norm": 5.913404763781138,
+      "learning_rate": 4.238584436366568e-06,
+      "loss": 0.7371,
+      "step": 3418
+    },
+    {
+      "epoch": 0.2776740030861691,
+      "grad_norm": 2.9595554999598455,
+      "learning_rate": 4.238111815376182e-06,
+      "loss": 0.5524,
+      "step": 3419
+    },
+    {
+      "epoch": 0.27775521806221065,
+      "grad_norm": 6.622001053281284,
+      "learning_rate": 4.23763907411766e-06,
+      "loss": 0.436,
+      "step": 3420
+    },
+    {
+      "epoch": 0.27783643303825223,
+      "grad_norm": 4.194004633417891,
+      "learning_rate": 4.237166212623708e-06,
+      "loss": 0.4323,
+      "step": 3421
+    },
+    {
+      "epoch": 0.27791764801429386,
+      "grad_norm": 5.700219795773494,
+      "learning_rate": 4.236693230927048e-06,
+      "loss": 0.496,
+      "step": 3422
+    },
+    {
+      "epoch": 0.27799886299033544,
+      "grad_norm": 4.15322324920816,
+      "learning_rate": 4.2362201290604085e-06,
+      "loss": 0.5907,
+      "step": 3423
+    },
+    {
+      "epoch": 0.278080077966377,
+      "grad_norm": 5.540030074318849,
+      "learning_rate": 4.235746907056525e-06,
+      "loss": 0.652,
+      "step": 3424
+    },
+    {
+      "epoch": 0.2781612929424186,
+      "grad_norm": 6.39998859259988,
+      "learning_rate": 4.235273564948142e-06,
+      "loss": 0.4876,
+      "step": 3425
+    },
+    {
+      "epoch": 0.27824250791846017,
+      "grad_norm": 3.7525190738960816,
+      "learning_rate": 4.234800102768012e-06,
+      "loss": 0.7998,
+      "step": 3426
+    },
+    {
+      "epoch": 0.27832372289450175,
+      "grad_norm": 4.294740550636778,
+      "learning_rate": 4.234326520548895e-06,
+      "loss": 0.6744,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2784049378705433,
+      "grad_norm": 5.52457863696226,
+      "learning_rate": 4.233852818323563e-06,
+      "loss": 0.4104,
+      "step": 3428
+    },
+    {
+      "epoch": 0.2784861528465849,
+      "grad_norm": 5.182650646362298,
+      "learning_rate": 4.233378996124792e-06,
+      "loss": 0.5443,
+      "step": 3429
+    },
+    {
+      "epoch": 0.2785673678226265,
+      "grad_norm": 8.063914556948795,
+      "learning_rate": 4.232905053985368e-06,
+      "loss": 0.6733,
+      "step": 3430
+    },
+    {
+      "epoch": 0.27864858279866805,
+      "grad_norm": 3.9395695955779186,
+      "learning_rate": 4.232430991938085e-06,
+      "loss": 0.5109,
+      "step": 3431
+    },
+    {
+      "epoch": 0.27872979777470963,
+      "grad_norm": 2.5171709827838993,
+      "learning_rate": 4.231956810015747e-06,
+      "loss": 0.672,
+      "step": 3432
+    },
+    {
+      "epoch": 0.27881101275075126,
+      "grad_norm": 5.452147154881219,
+      "learning_rate": 4.231482508251164e-06,
+      "loss": 0.5055,
+      "step": 3433
+    },
+    {
+      "epoch": 0.27889222772679284,
+      "grad_norm": 10.095151440140139,
+      "learning_rate": 4.231008086677154e-06,
+      "loss": 0.4714,
+      "step": 3434
+    },
+    {
+      "epoch": 0.2789734427028344,
+      "grad_norm": 4.246691334687768,
+      "learning_rate": 4.230533545326547e-06,
+      "loss": 0.5819,
+      "step": 3435
+    },
+    {
+      "epoch": 0.279054657678876,
+      "grad_norm": 7.118049034428091,
+      "learning_rate": 4.230058884232177e-06,
+      "loss": 0.7539,
+      "step": 3436
+    },
+    {
+      "epoch": 0.27913587265491757,
+      "grad_norm": 4.493070103171718,
+      "learning_rate": 4.229584103426888e-06,
+      "loss": 0.5447,
+      "step": 3437
+    },
+    {
+      "epoch": 0.27921708763095915,
+      "grad_norm": 3.9508050807048236,
+      "learning_rate": 4.229109202943533e-06,
+      "loss": 0.5914,
+      "step": 3438
+    },
+    {
+      "epoch": 0.2792983026070007,
+      "grad_norm": 6.3185262215803775,
+      "learning_rate": 4.228634182814972e-06,
+      "loss": 0.5831,
+      "step": 3439
+    },
+    {
+      "epoch": 0.2793795175830423,
+      "grad_norm": 4.369461605720211,
+      "learning_rate": 4.228159043074075e-06,
+      "loss": 0.6527,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2794607325590839,
+      "grad_norm": 6.77973389079385,
+      "learning_rate": 4.227683783753717e-06,
+      "loss": 0.5189,
+      "step": 3441
+    },
+    {
+      "epoch": 0.27954194753512546,
+      "grad_norm": 4.379921845032675,
+      "learning_rate": 4.227208404886787e-06,
+      "loss": 0.5284,
+      "step": 3442
+    },
+    {
+      "epoch": 0.27962316251116703,
+      "grad_norm": 4.507924635404893,
+      "learning_rate": 4.2267329065061745e-06,
+      "loss": 0.5291,
+      "step": 3443
+    },
+    {
+      "epoch": 0.27970437748720867,
+      "grad_norm": 6.075130482555993,
+      "learning_rate": 4.226257288644784e-06,
+      "loss": 0.5183,
+      "step": 3444
+    },
+    {
+      "epoch": 0.27978559246325024,
+      "grad_norm": 4.72530804483612,
+      "learning_rate": 4.225781551335526e-06,
+      "loss": 0.5276,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2798668074392918,
+      "grad_norm": 5.2670855376913535,
+      "learning_rate": 4.225305694611318e-06,
+      "loss": 0.5282,
+      "step": 3446
+    },
+    {
+      "epoch": 0.2799480224153334,
+      "grad_norm": 4.900878051447264,
+      "learning_rate": 4.224829718505087e-06,
+      "loss": 0.5453,
+      "step": 3447
+    },
+    {
+      "epoch": 0.280029237391375,
+      "grad_norm": 7.178013534818383,
+      "learning_rate": 4.224353623049767e-06,
+      "loss": 0.6766,
+      "step": 3448
+    },
+    {
+      "epoch": 0.28011045236741655,
+      "grad_norm": 4.664935512001858,
+      "learning_rate": 4.2238774082783025e-06,
+      "loss": 0.6089,
+      "step": 3449
+    },
+    {
+      "epoch": 0.28019166734345813,
+      "grad_norm": 4.852476116024416,
+      "learning_rate": 4.223401074223646e-06,
+      "loss": 0.6572,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2802728823194997,
+      "grad_norm": 8.563098982741186,
+      "learning_rate": 4.222924620918755e-06,
+      "loss": 0.5789,
+      "step": 3451
+    },
+    {
+      "epoch": 0.2803540972955413,
+      "grad_norm": 4.014234539171271,
+      "learning_rate": 4.222448048396599e-06,
+      "loss": 0.6374,
+      "step": 3452
+    },
+    {
+      "epoch": 0.28043531227158286,
+      "grad_norm": 4.822310015531145,
+      "learning_rate": 4.221971356690154e-06,
+      "loss": 0.5382,
+      "step": 3453
+    },
+    {
+      "epoch": 0.28051652724762444,
+      "grad_norm": 4.267417151682677,
+      "learning_rate": 4.221494545832405e-06,
+      "loss": 0.4697,
+      "step": 3454
+    },
+    {
+      "epoch": 0.28059774222366607,
+      "grad_norm": 8.796169733126867,
+      "learning_rate": 4.221017615856344e-06,
+      "loss": 0.467,
+      "step": 3455
+    },
+    {
+      "epoch": 0.28067895719970765,
+      "grad_norm": 5.742915475497312,
+      "learning_rate": 4.220540566794972e-06,
+      "loss": 0.5189,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2807601721757492,
+      "grad_norm": 4.702000191158769,
+      "learning_rate": 4.220063398681299e-06,
+      "loss": 0.6394,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2808413871517908,
+      "grad_norm": 7.827322462489466,
+      "learning_rate": 4.219586111548342e-06,
+      "loss": 0.427,
+      "step": 3458
+    },
+    {
+      "epoch": 0.2809226021278324,
+      "grad_norm": 4.5952790790019495,
+      "learning_rate": 4.219108705429127e-06,
+      "loss": 0.5275,
+      "step": 3459
+    },
+    {
+      "epoch": 0.28100381710387395,
+      "grad_norm": 4.260809941886097,
+      "learning_rate": 4.218631180356688e-06,
+      "loss": 0.6774,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28108503207991553,
+      "grad_norm": 5.933145706552524,
+      "learning_rate": 4.218153536364067e-06,
+      "loss": 0.5556,
+      "step": 3461
+    },
+    {
+      "epoch": 0.2811662470559571,
+      "grad_norm": 5.5215075629008545,
+      "learning_rate": 4.217675773484314e-06,
+      "loss": 0.6402,
+      "step": 3462
+    },
+    {
+      "epoch": 0.2812474620319987,
+      "grad_norm": 4.367099875870321,
+      "learning_rate": 4.217197891750488e-06,
+      "loss": 0.553,
+      "step": 3463
+    },
+    {
+      "epoch": 0.28132867700804026,
+      "grad_norm": 8.120031592132388,
+      "learning_rate": 4.216719891195657e-06,
+      "loss": 0.7601,
+      "step": 3464
+    },
+    {
+      "epoch": 0.28140989198408184,
+      "grad_norm": 13.915399996993806,
+      "learning_rate": 4.216241771852895e-06,
+      "loss": 0.3699,
+      "step": 3465
+    },
+    {
+      "epoch": 0.28149110696012347,
+      "grad_norm": 4.475259050725444,
+      "learning_rate": 4.215763533755285e-06,
+      "loss": 0.6308,
+      "step": 3466
+    },
+    {
+      "epoch": 0.28157232193616505,
+      "grad_norm": 4.246073015911453,
+      "learning_rate": 4.215285176935919e-06,
+      "loss": 0.5941,
+      "step": 3467
+    },
+    {
+      "epoch": 0.2816535369122066,
+      "grad_norm": 4.463656799978896,
+      "learning_rate": 4.214806701427896e-06,
+      "loss": 0.5504,
+      "step": 3468
+    },
+    {
+      "epoch": 0.2817347518882482,
+      "grad_norm": 2.9321912614363517,
+      "learning_rate": 4.214328107264326e-06,
+      "loss": 0.5993,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2818159668642898,
+      "grad_norm": 5.912480103116702,
+      "learning_rate": 4.213849394478323e-06,
+      "loss": 0.8218,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28189718184033136,
+      "grad_norm": 3.0648081596549996,
+      "learning_rate": 4.213370563103013e-06,
+      "loss": 0.5549,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28197839681637293,
+      "grad_norm": 7.610704699641839,
+      "learning_rate": 4.212891613171528e-06,
+      "loss": 0.539,
+      "step": 3472
+    },
+    {
+      "epoch": 0.2820596117924145,
+      "grad_norm": 4.821090407393734,
+      "learning_rate": 4.212412544717009e-06,
+      "loss": 0.5433,
+      "step": 3473
+    },
+    {
+      "epoch": 0.2821408267684561,
+      "grad_norm": 4.480177003679071,
+      "learning_rate": 4.211933357772604e-06,
+      "loss": 0.5649,
+      "step": 3474
+    },
+    {
+      "epoch": 0.28222204174449766,
+      "grad_norm": 6.0539202446851546,
+      "learning_rate": 4.211454052371471e-06,
+      "loss": 0.5074,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28230325672053924,
+      "grad_norm": 3.2128791712651066,
+      "learning_rate": 4.210974628546776e-06,
+      "loss": 0.6066,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2823844716965809,
+      "grad_norm": 5.748988331106903,
+      "learning_rate": 4.210495086331691e-06,
+      "loss": 0.5114,
+      "step": 3477
+    },
+    {
+      "epoch": 0.28246568667262245,
+      "grad_norm": 2.985462121611765,
+      "learning_rate": 4.2100154257594e-06,
+      "loss": 0.6491,
+      "step": 3478
+    },
+    {
+      "epoch": 0.28254690164866403,
+      "grad_norm": 4.631989287533722,
+      "learning_rate": 4.20953564686309e-06,
+      "loss": 0.5046,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2826281166247056,
+      "grad_norm": 7.645616919092413,
+      "learning_rate": 4.2090557496759615e-06,
+      "loss": 0.5868,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2827093316007472,
+      "grad_norm": 13.216352857254574,
+      "learning_rate": 4.208575734231221e-06,
+      "loss": 0.539,
+      "step": 3481
+    },
+    {
+      "epoch": 0.28279054657678876,
+      "grad_norm": 4.282296217606311,
+      "learning_rate": 4.208095600562081e-06,
+      "loss": 0.6534,
+      "step": 3482
+    },
+    {
+      "epoch": 0.28287176155283034,
+      "grad_norm": 3.613475303783212,
+      "learning_rate": 4.2076153487017655e-06,
+      "loss": 0.5671,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2829529765288719,
+      "grad_norm": 4.958757945118231,
+      "learning_rate": 4.207134978683506e-06,
+      "loss": 0.5416,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2830341915049135,
+      "grad_norm": 4.199726726733855,
+      "learning_rate": 4.206654490540541e-06,
+      "loss": 0.6141,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28311540648095507,
+      "grad_norm": 4.600582479141581,
+      "learning_rate": 4.206173884306116e-06,
+      "loss": 0.6657,
+      "step": 3486
+    },
+    {
+      "epoch": 0.28319662145699664,
+      "grad_norm": 7.6947988469297375,
+      "learning_rate": 4.20569316001349e-06,
+      "loss": 0.5541,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2832778364330383,
+      "grad_norm": 5.367699544702637,
+      "learning_rate": 4.205212317695924e-06,
+      "loss": 0.6524,
+      "step": 3488
+    },
+    {
+      "epoch": 0.28335905140907985,
+      "grad_norm": 6.029688715188512,
+      "learning_rate": 4.204731357386689e-06,
+      "loss": 0.5518,
+      "step": 3489
+    },
+    {
+      "epoch": 0.28344026638512143,
+      "grad_norm": 4.352090436645465,
+      "learning_rate": 4.204250279119068e-06,
+      "loss": 0.5141,
+      "step": 3490
+    },
+    {
+      "epoch": 0.283521481361163,
+      "grad_norm": 4.073146574157928,
+      "learning_rate": 4.203769082926346e-06,
+      "loss": 0.5047,
+      "step": 3491
+    },
+    {
+      "epoch": 0.2836026963372046,
+      "grad_norm": 4.576078022715138,
+      "learning_rate": 4.203287768841822e-06,
+      "loss": 0.5063,
+      "step": 3492
+    },
+    {
+      "epoch": 0.28368391131324616,
+      "grad_norm": 5.141508394238785,
+      "learning_rate": 4.202806336898798e-06,
+      "loss": 0.5,
+      "step": 3493
+    },
+    {
+      "epoch": 0.28376512628928774,
+      "grad_norm": 4.33832646184562,
+      "learning_rate": 4.202324787130587e-06,
+      "loss": 0.5375,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2838463412653293,
+      "grad_norm": 7.671372488757828,
+      "learning_rate": 4.201843119570511e-06,
+      "loss": 0.6192,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2839275562413709,
+      "grad_norm": 6.000625361865301,
+      "learning_rate": 4.201361334251898e-06,
+      "loss": 0.5039,
+      "step": 3496
+    },
+    {
+      "epoch": 0.28400877121741247,
+      "grad_norm": 10.56462611054689,
+      "learning_rate": 4.200879431208084e-06,
+      "loss": 0.4379,
+      "step": 3497
+    },
+    {
+      "epoch": 0.28408998619345405,
+      "grad_norm": 5.2118113489010325,
+      "learning_rate": 4.200397410472416e-06,
+      "loss": 0.4859,
+      "step": 3498
+    },
+    {
+      "epoch": 0.2841712011694957,
+      "grad_norm": 3.7680592065666447,
+      "learning_rate": 4.199915272078247e-06,
+      "loss": 0.5509,
+      "step": 3499
+    },
+    {
+      "epoch": 0.28425241614553726,
+      "grad_norm": 4.998842782205934,
+      "learning_rate": 4.199433016058936e-06,
+      "loss": 0.5495,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28433363112157883,
+      "grad_norm": 3.6541213589146166,
+      "learning_rate": 4.198950642447856e-06,
+      "loss": 0.7963,
+      "step": 3501
+    },
+    {
+      "epoch": 0.2844148460976204,
+      "grad_norm": 4.6320107487602655,
+      "learning_rate": 4.198468151278382e-06,
+      "loss": 0.5354,
+      "step": 3502
+    },
+    {
+      "epoch": 0.284496061073662,
+      "grad_norm": 16.856109294832514,
+      "learning_rate": 4.197985542583902e-06,
+      "loss": 0.5639,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28457727604970356,
+      "grad_norm": 4.341047779732985,
+      "learning_rate": 4.197502816397809e-06,
+      "loss": 0.5327,
+      "step": 3504
+    },
+    {
+      "epoch": 0.28465849102574514,
+      "grad_norm": 6.54679516358405,
+      "learning_rate": 4.197019972753504e-06,
+      "loss": 0.5685,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2847397060017867,
+      "grad_norm": 8.538628468801688,
+      "learning_rate": 4.1965370116843985e-06,
+      "loss": 0.6608,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2848209209778283,
+      "grad_norm": 6.781146736308602,
+      "learning_rate": 4.1960539332239115e-06,
+      "loss": 0.5363,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2849021359538699,
+      "grad_norm": 6.568218511710609,
+      "learning_rate": 4.195570737405468e-06,
+      "loss": 0.4654,
+      "step": 3508
+    },
+    {
+      "epoch": 0.28498335092991145,
+      "grad_norm": 6.14037981907578,
+      "learning_rate": 4.195087424262503e-06,
+      "loss": 0.6075,
+      "step": 3509
+    },
+    {
+      "epoch": 0.2850645659059531,
+      "grad_norm": 3.9057111334726016,
+      "learning_rate": 4.194603993828459e-06,
+      "loss": 0.4975,
+      "step": 3510
+    },
+    {
+      "epoch": 0.28514578088199466,
+      "grad_norm": 6.56859758615991,
+      "learning_rate": 4.194120446136788e-06,
+      "loss": 0.6143,
+      "step": 3511
+    },
+    {
+      "epoch": 0.28522699585803624,
+      "grad_norm": 7.248281271806781,
+      "learning_rate": 4.193636781220948e-06,
+      "loss": 0.6135,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2853082108340778,
+      "grad_norm": 3.6814056969521265,
+      "learning_rate": 4.1931529991144056e-06,
+      "loss": 0.644,
+      "step": 3513
+    },
+    {
+      "epoch": 0.2853894258101194,
+      "grad_norm": 4.124004556862366,
+      "learning_rate": 4.192669099850637e-06,
+      "loss": 0.4091,
+      "step": 3514
+    },
+    {
+      "epoch": 0.28547064078616097,
+      "grad_norm": 9.237020510741628,
+      "learning_rate": 4.192185083463125e-06,
+      "loss": 0.6916,
+      "step": 3515
+    },
+    {
+      "epoch": 0.28555185576220254,
+      "grad_norm": 4.325518645446921,
+      "learning_rate": 4.19170094998536e-06,
+      "loss": 0.641,
+      "step": 3516
+    },
+    {
+      "epoch": 0.2856330707382441,
+      "grad_norm": 3.772328467417333,
+      "learning_rate": 4.191216699450844e-06,
+      "loss": 0.5248,
+      "step": 3517
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 4.5828680511801,
+      "learning_rate": 4.190732331893083e-06,
+      "loss": 0.5488,
+      "step": 3518
+    },
+    {
+      "epoch": 0.2857955006903273,
+      "grad_norm": 6.7144709613110045,
+      "learning_rate": 4.190247847345591e-06,
+      "loss": 0.6085,
+      "step": 3519
+    },
+    {
+      "epoch": 0.28587671566636885,
+      "grad_norm": 6.61578411126398,
+      "learning_rate": 4.189763245841895e-06,
+      "loss": 0.4582,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2859579306424105,
+      "grad_norm": 7.697721977409483,
+      "learning_rate": 4.189278527415524e-06,
+      "loss": 0.4666,
+      "step": 3521
+    },
+    {
+      "epoch": 0.28603914561845206,
+      "grad_norm": 9.805505848445298,
+      "learning_rate": 4.188793692100021e-06,
+      "loss": 0.5197,
+      "step": 3522
+    },
+    {
+      "epoch": 0.28612036059449364,
+      "grad_norm": 4.38159024051503,
+      "learning_rate": 4.1883087399289315e-06,
+      "loss": 0.6191,
+      "step": 3523
+    },
+    {
+      "epoch": 0.2862015755705352,
+      "grad_norm": 5.443162368213296,
+      "learning_rate": 4.187823670935812e-06,
+      "loss": 0.4839,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2862827905465768,
+      "grad_norm": 9.060902521439406,
+      "learning_rate": 4.187338485154228e-06,
+      "loss": 0.5212,
+      "step": 3525
+    },
+    {
+      "epoch": 0.28636400552261837,
+      "grad_norm": 4.5276722061450405,
+      "learning_rate": 4.186853182617751e-06,
+      "loss": 0.5874,
+      "step": 3526
+    },
+    {
+      "epoch": 0.28644522049865995,
+      "grad_norm": 4.787775939104659,
+      "learning_rate": 4.1863677633599605e-06,
+      "loss": 0.6216,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2865264354747015,
+      "grad_norm": 4.032333523480888,
+      "learning_rate": 4.1858822274144465e-06,
+      "loss": 0.8107,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2866076504507431,
+      "grad_norm": 4.164487711997063,
+      "learning_rate": 4.185396574814804e-06,
+      "loss": 0.5834,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2866888654267847,
+      "grad_norm": 4.736119308914805,
+      "learning_rate": 4.184910805594639e-06,
+      "loss": 0.5395,
+      "step": 3530
+    },
+    {
+      "epoch": 0.28677008040282626,
+      "grad_norm": 4.206365573240851,
+      "learning_rate": 4.184424919787563e-06,
+      "loss": 0.634,
+      "step": 3531
+    },
+    {
+      "epoch": 0.2868512953788679,
+      "grad_norm": 3.609471547339424,
+      "learning_rate": 4.183938917427198e-06,
+      "loss": 0.512,
+      "step": 3532
+    },
+    {
+      "epoch": 0.28693251035490946,
+      "grad_norm": 3.3756577667802303,
+      "learning_rate": 4.183452798547171e-06,
+      "loss": 0.6594,
+      "step": 3533
+    },
+    {
+      "epoch": 0.28701372533095104,
+      "grad_norm": 4.577782600286523,
+      "learning_rate": 4.1829665631811214e-06,
+      "loss": 0.5162,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2870949403069926,
+      "grad_norm": 5.423760632233463,
+      "learning_rate": 4.182480211362691e-06,
+      "loss": 0.6164,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2871761552830342,
+      "grad_norm": 8.37175609091639,
+      "learning_rate": 4.181993743125535e-06,
+      "loss": 0.5202,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2872573702590758,
+      "grad_norm": 4.332707609863406,
+      "learning_rate": 4.181507158503314e-06,
+      "loss": 0.7249,
+      "step": 3537
+    },
+    {
+      "epoch": 0.28733858523511735,
+      "grad_norm": 4.538954914725168,
+      "learning_rate": 4.1810204575296966e-06,
+      "loss": 0.5169,
+      "step": 3538
+    },
+    {
+      "epoch": 0.2874198002111589,
+      "grad_norm": 4.191746753570993,
+      "learning_rate": 4.180533640238361e-06,
+      "loss": 0.6053,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2875010151872005,
+      "grad_norm": 4.85627080590084,
+      "learning_rate": 4.180046706662991e-06,
+      "loss": 0.5235,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2875822301632421,
+      "grad_norm": 6.5813372383035915,
+      "learning_rate": 4.17955965683728e-06,
+      "loss": 0.5635,
+      "step": 3541
+    },
+    {
+      "epoch": 0.28766344513928366,
+      "grad_norm": 6.353502644005653,
+      "learning_rate": 4.17907249079493e-06,
+      "loss": 0.5271,
+      "step": 3542
+    },
+    {
+      "epoch": 0.2877446601153253,
+      "grad_norm": 3.316222423088927,
+      "learning_rate": 4.17858520856965e-06,
+      "loss": 0.5198,
+      "step": 3543
+    },
+    {
+      "epoch": 0.28782587509136687,
+      "grad_norm": 3.6027797950419846,
+      "learning_rate": 4.178097810195157e-06,
+      "loss": 0.5364,
+      "step": 3544
+    },
+    {
+      "epoch": 0.28790709006740844,
+      "grad_norm": 6.196439592693107,
+      "learning_rate": 4.177610295705178e-06,
+      "loss": 0.5973,
+      "step": 3545
+    },
+    {
+      "epoch": 0.28798830504345,
+      "grad_norm": 3.281604318639601,
+      "learning_rate": 4.177122665133444e-06,
+      "loss": 0.6097,
+      "step": 3546
+    },
+    {
+      "epoch": 0.2880695200194916,
+      "grad_norm": 5.0788056380548525,
+      "learning_rate": 4.176634918513698e-06,
+      "loss": 0.479,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2881507349955332,
+      "grad_norm": 10.492165661109485,
+      "learning_rate": 4.176147055879689e-06,
+      "loss": 0.6935,
+      "step": 3548
+    },
+    {
+      "epoch": 0.28823194997157475,
+      "grad_norm": 5.566980762163421,
+      "learning_rate": 4.175659077265175e-06,
+      "loss": 0.5723,
+      "step": 3549
+    },
+    {
+      "epoch": 0.28831316494761633,
+      "grad_norm": 4.510747031993934,
+      "learning_rate": 4.175170982703921e-06,
+      "loss": 0.4667,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2883943799236579,
+      "grad_norm": 5.795017468211081,
+      "learning_rate": 4.1746827722297e-06,
+      "loss": 0.6429,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2884755948996995,
+      "grad_norm": 4.90772281384199,
+      "learning_rate": 4.174194445876295e-06,
+      "loss": 0.6138,
+      "step": 3552
+    },
+    {
+      "epoch": 0.28855680987574106,
+      "grad_norm": 9.38458036391076,
+      "learning_rate": 4.1737060036774945e-06,
+      "loss": 0.5942,
+      "step": 3553
+    },
+    {
+      "epoch": 0.2886380248517827,
+      "grad_norm": 3.744621239251605,
+      "learning_rate": 4.173217445667097e-06,
+      "loss": 0.4725,
+      "step": 3554
+    },
+    {
+      "epoch": 0.28871923982782427,
+      "grad_norm": 3.5760812252375858,
+      "learning_rate": 4.172728771878908e-06,
+      "loss": 0.4955,
+      "step": 3555
+    },
+    {
+      "epoch": 0.28880045480386585,
+      "grad_norm": 5.300448357602322,
+      "learning_rate": 4.17223998234674e-06,
+      "loss": 0.5496,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2888816697799074,
+      "grad_norm": 4.301151535524972,
+      "learning_rate": 4.171751077104415e-06,
+      "loss": 0.6269,
+      "step": 3557
+    },
+    {
+      "epoch": 0.288962884755949,
+      "grad_norm": 4.150238159594849,
+      "learning_rate": 4.171262056185764e-06,
+      "loss": 0.6023,
+      "step": 3558
+    },
+    {
+      "epoch": 0.2890440997319906,
+      "grad_norm": 6.588110479356063,
+      "learning_rate": 4.170772919624624e-06,
+      "loss": 0.5044,
+      "step": 3559
+    },
+    {
+      "epoch": 0.28912531470803216,
+      "grad_norm": 9.52780556407402,
+      "learning_rate": 4.170283667454839e-06,
+      "loss": 0.4627,
+      "step": 3560
+    },
+    {
+      "epoch": 0.28920652968407373,
+      "grad_norm": 3.112608140290074,
+      "learning_rate": 4.169794299710266e-06,
+      "loss": 0.5403,
+      "step": 3561
+    },
+    {
+      "epoch": 0.2892877446601153,
+      "grad_norm": 5.782425679496101,
+      "learning_rate": 4.169304816424763e-06,
+      "loss": 0.6422,
+      "step": 3562
+    },
+    {
+      "epoch": 0.2893689596361569,
+      "grad_norm": 6.327129043676947,
+      "learning_rate": 4.168815217632202e-06,
+      "loss": 0.5983,
+      "step": 3563
+    },
+    {
+      "epoch": 0.28945017461219846,
+      "grad_norm": 5.64892446517355,
+      "learning_rate": 4.168325503366461e-06,
+      "loss": 0.6639,
+      "step": 3564
+    },
+    {
+      "epoch": 0.2895313895882401,
+      "grad_norm": 4.212091644362981,
+      "learning_rate": 4.167835673661422e-06,
+      "loss": 0.5173,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2896126045642817,
+      "grad_norm": 4.627484710444349,
+      "learning_rate": 4.167345728550984e-06,
+      "loss": 0.5776,
+      "step": 3566
+    },
+    {
+      "epoch": 0.28969381954032325,
+      "grad_norm": 6.431815623240765,
+      "learning_rate": 4.166855668069045e-06,
+      "loss": 0.5357,
+      "step": 3567
+    },
+    {
+      "epoch": 0.2897750345163648,
+      "grad_norm": 5.104463737375422,
+      "learning_rate": 4.166365492249514e-06,
+      "loss": 0.4888,
+      "step": 3568
+    },
+    {
+      "epoch": 0.2898562494924064,
+      "grad_norm": 2.6553169339995755,
+      "learning_rate": 4.1658752011263125e-06,
+      "loss": 0.4652,
+      "step": 3569
+    },
+    {
+      "epoch": 0.289937464468448,
+      "grad_norm": 4.522379342995965,
+      "learning_rate": 4.1653847947333625e-06,
+      "loss": 0.6268,
+      "step": 3570
+    },
+    {
+      "epoch": 0.29001867944448956,
+      "grad_norm": 4.206423587982746,
+      "learning_rate": 4.164894273104599e-06,
+      "loss": 0.7023,
+      "step": 3571
+    },
+    {
+      "epoch": 0.29009989442053113,
+      "grad_norm": 3.550921658943635,
+      "learning_rate": 4.164403636273963e-06,
+      "loss": 0.6467,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2901811093965727,
+      "grad_norm": 4.41960430630749,
+      "learning_rate": 4.163912884275403e-06,
+      "loss": 0.6449,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2902623243726143,
+      "grad_norm": 5.335473970114484,
+      "learning_rate": 4.163422017142879e-06,
+      "loss": 0.532,
+      "step": 3574
+    },
+    {
+      "epoch": 0.29034353934865587,
+      "grad_norm": 9.485322332094816,
+      "learning_rate": 4.162931034910354e-06,
+      "loss": 0.5057,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2904247543246975,
+      "grad_norm": 5.472990906211878,
+      "learning_rate": 4.162439937611803e-06,
+      "loss": 0.5523,
+      "step": 3576
+    },
+    {
+      "epoch": 0.2905059693007391,
+      "grad_norm": 4.626244114541653,
+      "learning_rate": 4.161948725281206e-06,
+      "loss": 0.5859,
+      "step": 3577
+    },
+    {
+      "epoch": 0.29058718427678065,
+      "grad_norm": 4.102620389372604,
+      "learning_rate": 4.161457397952553e-06,
+      "loss": 0.5745,
+      "step": 3578
+    },
+    {
+      "epoch": 0.29066839925282223,
+      "grad_norm": 5.532218377747797,
+      "learning_rate": 4.160965955659843e-06,
+      "loss": 0.4707,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2907496142288638,
+      "grad_norm": 4.2544863217335545,
+      "learning_rate": 4.160474398437077e-06,
+      "loss": 0.5238,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2908308292049054,
+      "grad_norm": 5.496605956094565,
+      "learning_rate": 4.159982726318271e-06,
+      "loss": 0.8256,
+      "step": 3581
+    },
+    {
+      "epoch": 0.29091204418094696,
+      "grad_norm": 4.300837180788803,
+      "learning_rate": 4.159490939337447e-06,
+      "loss": 0.6179,
+      "step": 3582
+    },
+    {
+      "epoch": 0.29099325915698854,
+      "grad_norm": 3.957326025756882,
+      "learning_rate": 4.158999037528632e-06,
+      "loss": 0.5216,
+      "step": 3583
+    },
+    {
+      "epoch": 0.2910744741330301,
+      "grad_norm": 5.327052291363593,
+      "learning_rate": 4.1585070209258635e-06,
+      "loss": 0.859,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2911556891090717,
+      "grad_norm": 6.708637795707786,
+      "learning_rate": 4.158014889563187e-06,
+      "loss": 0.4936,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29123690408511327,
+      "grad_norm": 3.6081169619085487,
+      "learning_rate": 4.157522643474654e-06,
+      "loss": 0.4951,
+      "step": 3586
+    },
+    {
+      "epoch": 0.2913181190611549,
+      "grad_norm": 5.692407639431199,
+      "learning_rate": 4.157030282694328e-06,
+      "loss": 0.4514,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2913993340371965,
+      "grad_norm": 5.96319840346763,
+      "learning_rate": 4.156537807256275e-06,
+      "loss": 0.6435,
+      "step": 3588
+    },
+    {
+      "epoch": 0.29148054901323805,
+      "grad_norm": 4.057628153052015,
+      "learning_rate": 4.156045217194573e-06,
+      "loss": 0.5785,
+      "step": 3589
+    },
+    {
+      "epoch": 0.29156176398927963,
+      "grad_norm": 4.725549294999352,
+      "learning_rate": 4.1555525125433074e-06,
+      "loss": 0.561,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2916429789653212,
+      "grad_norm": 4.897497425355875,
+      "learning_rate": 4.155059693336569e-06,
+      "loss": 0.4877,
+      "step": 3591
+    },
+    {
+      "epoch": 0.2917241939413628,
+      "grad_norm": 6.7306605891617,
+      "learning_rate": 4.1545667596084596e-06,
+      "loss": 0.6536,
+      "step": 3592
+    },
+    {
+      "epoch": 0.29180540891740436,
+      "grad_norm": 4.458589616639652,
+      "learning_rate": 4.154073711393087e-06,
+      "loss": 0.6075,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29188662389344594,
+      "grad_norm": 16.172763248388275,
+      "learning_rate": 4.153580548724567e-06,
+      "loss": 0.5503,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2919678388694875,
+      "grad_norm": 5.597703145374191,
+      "learning_rate": 4.153087271637025e-06,
+      "loss": 0.5837,
+      "step": 3595
+    },
+    {
+      "epoch": 0.2920490538455291,
+      "grad_norm": 11.754470585418524,
+      "learning_rate": 4.1525938801645926e-06,
+      "loss": 0.5668,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29213026882157067,
+      "grad_norm": 6.254803456678843,
+      "learning_rate": 4.152100374341409e-06,
+      "loss": 0.5732,
+      "step": 3597
+    },
+    {
+      "epoch": 0.2922114837976123,
+      "grad_norm": 5.190101012718226,
+      "learning_rate": 4.151606754201625e-06,
+      "loss": 0.6051,
+      "step": 3598
+    },
+    {
+      "epoch": 0.2922926987736539,
+      "grad_norm": 10.669928997016616,
+      "learning_rate": 4.151113019779393e-06,
+      "loss": 0.639,
+      "step": 3599
+    },
+    {
+      "epoch": 0.29237391374969546,
+      "grad_norm": 5.619799524172815,
+      "learning_rate": 4.150619171108879e-06,
+      "loss": 0.5745,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29245512872573703,
+      "grad_norm": 4.562644028883115,
+      "learning_rate": 4.150125208224255e-06,
+      "loss": 0.6914,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2925363437017786,
+      "grad_norm": 4.491512432892291,
+      "learning_rate": 4.149631131159698e-06,
+      "loss": 0.4882,
+      "step": 3602
+    },
+    {
+      "epoch": 0.2926175586778202,
+      "grad_norm": 4.214753071853409,
+      "learning_rate": 4.149136939949399e-06,
+      "loss": 0.5967,
+      "step": 3603
+    },
+    {
+      "epoch": 0.29269877365386177,
+      "grad_norm": 6.53315802645859,
+      "learning_rate": 4.14864263462755e-06,
+      "loss": 0.5239,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29277998862990334,
+      "grad_norm": 5.806026900388708,
+      "learning_rate": 4.148148215228357e-06,
+      "loss": 0.6479,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2928612036059449,
+      "grad_norm": 4.392767448225316,
+      "learning_rate": 4.147653681786031e-06,
+      "loss": 0.5045,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2929424185819865,
+      "grad_norm": 7.680285801817868,
+      "learning_rate": 4.147159034334789e-06,
+      "loss": 0.5433,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2930236335580281,
+      "grad_norm": 5.275584522954317,
+      "learning_rate": 4.146664272908859e-06,
+      "loss": 0.5867,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2931048485340697,
+      "grad_norm": 5.37369190885388,
+      "learning_rate": 4.146169397542478e-06,
+      "loss": 0.5683,
+      "step": 3609
+    },
+    {
+      "epoch": 0.2931860635101113,
+      "grad_norm": 3.8049589428305413,
+      "learning_rate": 4.145674408269885e-06,
+      "loss": 0.6545,
+      "step": 3610
+    },
+    {
+      "epoch": 0.29326727848615286,
+      "grad_norm": 5.091756495748853,
+      "learning_rate": 4.145179305125333e-06,
+      "loss": 0.5765,
+      "step": 3611
+    },
+    {
+      "epoch": 0.29334849346219444,
+      "grad_norm": 13.084271350497739,
+      "learning_rate": 4.14468408814308e-06,
+      "loss": 0.5044,
+      "step": 3612
+    },
+    {
+      "epoch": 0.293429708438236,
+      "grad_norm": 5.905329389388773,
+      "learning_rate": 4.1441887573573935e-06,
+      "loss": 0.5513,
+      "step": 3613
+    },
+    {
+      "epoch": 0.2935109234142776,
+      "grad_norm": 4.255974685046024,
+      "learning_rate": 4.143693312802546e-06,
+      "loss": 0.4885,
+      "step": 3614
+    },
+    {
+      "epoch": 0.29359213839031917,
+      "grad_norm": 7.189955017278807,
+      "learning_rate": 4.143197754512821e-06,
+      "loss": 0.5293,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29367335336636075,
+      "grad_norm": 4.434545028745304,
+      "learning_rate": 4.142702082522507e-06,
+      "loss": 0.4807,
+      "step": 3616
+    },
+    {
+      "epoch": 0.2937545683424023,
+      "grad_norm": 5.613891388134908,
+      "learning_rate": 4.142206296865904e-06,
+      "loss": 0.5229,
+      "step": 3617
+    },
+    {
+      "epoch": 0.2938357833184439,
+      "grad_norm": 3.159086352727405,
+      "learning_rate": 4.141710397577315e-06,
+      "loss": 0.6718,
+      "step": 3618
+    },
+    {
+      "epoch": 0.2939169982944855,
+      "grad_norm": 3.5367459467799858,
+      "learning_rate": 4.141214384691056e-06,
+      "loss": 0.5547,
+      "step": 3619
+    },
+    {
+      "epoch": 0.2939982132705271,
+      "grad_norm": 7.18465729456046,
+      "learning_rate": 4.1407182582414476e-06,
+      "loss": 0.5301,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2940794282465687,
+      "grad_norm": 5.6826492806068405,
+      "learning_rate": 4.140222018262818e-06,
+      "loss": 0.6391,
+      "step": 3621
+    },
+    {
+      "epoch": 0.29416064322261026,
+      "grad_norm": 3.357835141229998,
+      "learning_rate": 4.139725664789507e-06,
+      "loss": 0.7172,
+      "step": 3622
+    },
+    {
+      "epoch": 0.29424185819865184,
+      "grad_norm": 17.104412542873206,
+      "learning_rate": 4.139229197855857e-06,
+      "loss": 0.441,
+      "step": 3623
+    },
+    {
+      "epoch": 0.2943230731746934,
+      "grad_norm": 4.059380496052101,
+      "learning_rate": 4.138732617496223e-06,
+      "loss": 0.5264,
+      "step": 3624
+    },
+    {
+      "epoch": 0.294404288150735,
+      "grad_norm": 4.907343820967712,
+      "learning_rate": 4.138235923744964e-06,
+      "loss": 0.4865,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29448550312677657,
+      "grad_norm": 4.767491409739287,
+      "learning_rate": 4.13773911663645e-06,
+      "loss": 0.5446,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29456671810281815,
+      "grad_norm": 11.043074971749009,
+      "learning_rate": 4.137242196205056e-06,
+      "loss": 0.514,
+      "step": 3627
+    },
+    {
+      "epoch": 0.2946479330788597,
+      "grad_norm": 8.751284247365856,
+      "learning_rate": 4.136745162485168e-06,
+      "loss": 0.5782,
+      "step": 3628
+    },
+    {
+      "epoch": 0.2947291480549013,
+      "grad_norm": 4.259342421928949,
+      "learning_rate": 4.1362480155111764e-06,
+      "loss": 0.4735,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2948103630309429,
+      "grad_norm": 5.982588574242516,
+      "learning_rate": 4.135750755317481e-06,
+      "loss": 0.6233,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2948915780069845,
+      "grad_norm": 4.717722006131577,
+      "learning_rate": 4.135253381938492e-06,
+      "loss": 0.6496,
+      "step": 3631
+    },
+    {
+      "epoch": 0.2949727929830261,
+      "grad_norm": 3.9506979794735337,
+      "learning_rate": 4.134755895408623e-06,
+      "loss": 0.6076,
+      "step": 3632
+    },
+    {
+      "epoch": 0.29505400795906767,
+      "grad_norm": 6.244712141178371,
+      "learning_rate": 4.134258295762297e-06,
+      "loss": 0.5481,
+      "step": 3633
+    },
+    {
+      "epoch": 0.29513522293510924,
+      "grad_norm": 5.110391491927889,
+      "learning_rate": 4.1337605830339465e-06,
+      "loss": 0.5417,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2952164379111508,
+      "grad_norm": 4.8675201483594615,
+      "learning_rate": 4.133262757258011e-06,
+      "loss": 0.6065,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2952976528871924,
+      "grad_norm": 11.498154785191828,
+      "learning_rate": 4.132764818468936e-06,
+      "loss": 0.5392,
+      "step": 3636
+    },
+    {
+      "epoch": 0.295378867863234,
+      "grad_norm": 6.065167654609438,
+      "learning_rate": 4.1322667667011774e-06,
+      "loss": 0.6767,
+      "step": 3637
+    },
+    {
+      "epoch": 0.29546008283927555,
+      "grad_norm": 9.853858235216926,
+      "learning_rate": 4.131768601989196e-06,
+      "loss": 0.5793,
+      "step": 3638
+    },
+    {
+      "epoch": 0.2955412978153171,
+      "grad_norm": 5.19821566952064,
+      "learning_rate": 4.131270324367464e-06,
+      "loss": 0.7266,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2956225127913587,
+      "grad_norm": 6.474598040312163,
+      "learning_rate": 4.130771933870459e-06,
+      "loss": 0.6649,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2957037277674003,
+      "grad_norm": 6.622824718236689,
+      "learning_rate": 4.130273430532667e-06,
+      "loss": 0.4317,
+      "step": 3641
+    },
+    {
+      "epoch": 0.2957849427434419,
+      "grad_norm": 8.781051249556342,
+      "learning_rate": 4.129774814388582e-06,
+      "loss": 0.4864,
+      "step": 3642
+    },
+    {
+      "epoch": 0.2958661577194835,
+      "grad_norm": 4.776911195727594,
+      "learning_rate": 4.1292760854727045e-06,
+      "loss": 0.5531,
+      "step": 3643
+    },
+    {
+      "epoch": 0.29594737269552507,
+      "grad_norm": 4.828770110545079,
+      "learning_rate": 4.128777243819546e-06,
+      "loss": 0.5435,
+      "step": 3644
+    },
+    {
+      "epoch": 0.29602858767156665,
+      "grad_norm": 4.550741300328438,
+      "learning_rate": 4.128278289463621e-06,
+      "loss": 0.4474,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2961098026476082,
+      "grad_norm": 3.03442154965047,
+      "learning_rate": 4.127779222439457e-06,
+      "loss": 0.5896,
+      "step": 3646
+    },
+    {
+      "epoch": 0.2961910176236498,
+      "grad_norm": 4.7725701865486085,
+      "learning_rate": 4.127280042781585e-06,
+      "loss": 0.6183,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2962722325996914,
+      "grad_norm": 5.751607562129402,
+      "learning_rate": 4.126780750524546e-06,
+      "loss": 0.4919,
+      "step": 3648
+    },
+    {
+      "epoch": 0.29635344757573295,
+      "grad_norm": 3.5887135989563057,
+      "learning_rate": 4.126281345702889e-06,
+      "loss": 0.5275,
+      "step": 3649
+    },
+    {
+      "epoch": 0.29643466255177453,
+      "grad_norm": 6.464495222017677,
+      "learning_rate": 4.125781828351171e-06,
+      "loss": 0.8401,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2965158775278161,
+      "grad_norm": 5.980573147403461,
+      "learning_rate": 4.125282198503953e-06,
+      "loss": 0.5954,
+      "step": 3651
+    },
+    {
+      "epoch": 0.2965970925038577,
+      "grad_norm": 5.228111544471104,
+      "learning_rate": 4.124782456195809e-06,
+      "loss": 0.5105,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2966783074798993,
+      "grad_norm": 7.53149544733575,
+      "learning_rate": 4.124282601461319e-06,
+      "loss": 0.4924,
+      "step": 3653
+    },
+    {
+      "epoch": 0.2967595224559409,
+      "grad_norm": 5.556355856990694,
+      "learning_rate": 4.123782634335068e-06,
+      "loss": 0.4124,
+      "step": 3654
+    },
+    {
+      "epoch": 0.29684073743198247,
+      "grad_norm": 4.358048092574572,
+      "learning_rate": 4.123282554851654e-06,
+      "loss": 0.5824,
+      "step": 3655
+    },
+    {
+      "epoch": 0.29692195240802405,
+      "grad_norm": 3.5786979265887187,
+      "learning_rate": 4.122782363045677e-06,
+      "loss": 0.4748,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2970031673840656,
+      "grad_norm": 4.524617536184045,
+      "learning_rate": 4.12228205895175e-06,
+      "loss": 0.4482,
+      "step": 3657
+    },
+    {
+      "epoch": 0.2970843823601072,
+      "grad_norm": 6.2371920872247735,
+      "learning_rate": 4.12178164260449e-06,
+      "loss": 0.565,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2971655973361488,
+      "grad_norm": 3.54005929501434,
+      "learning_rate": 4.121281114038524e-06,
+      "loss": 0.4706,
+      "step": 3659
+    },
+    {
+      "epoch": 0.29724681231219036,
+      "grad_norm": 6.2010580244373905,
+      "learning_rate": 4.120780473288485e-06,
+      "loss": 0.4807,
+      "step": 3660
+    },
+    {
+      "epoch": 0.29732802728823193,
+      "grad_norm": 3.39391672989296,
+      "learning_rate": 4.120279720389015e-06,
+      "loss": 0.5279,
+      "step": 3661
+    },
+    {
+      "epoch": 0.2974092422642735,
+      "grad_norm": 7.485601794086229,
+      "learning_rate": 4.119778855374763e-06,
+      "loss": 0.6333,
+      "step": 3662
+    },
+    {
+      "epoch": 0.2974904572403151,
+      "grad_norm": 4.283931201224972,
+      "learning_rate": 4.1192778782803875e-06,
+      "loss": 0.5886,
+      "step": 3663
+    },
+    {
+      "epoch": 0.2975716722163567,
+      "grad_norm": 6.09549459673354,
+      "learning_rate": 4.118776789140551e-06,
+      "loss": 0.4158,
+      "step": 3664
+    },
+    {
+      "epoch": 0.2976528871923983,
+      "grad_norm": 3.9183720909856645,
+      "learning_rate": 4.1182755879899305e-06,
+      "loss": 0.5857,
+      "step": 3665
+    },
+    {
+      "epoch": 0.2977341021684399,
+      "grad_norm": 4.554155011331954,
+      "learning_rate": 4.117774274863203e-06,
+      "loss": 0.5828,
+      "step": 3666
+    },
+    {
+      "epoch": 0.29781531714448145,
+      "grad_norm": 6.026037573474456,
+      "learning_rate": 4.117272849795057e-06,
+      "loss": 0.7132,
+      "step": 3667
+    },
+    {
+      "epoch": 0.297896532120523,
+      "grad_norm": 5.369767546922633,
+      "learning_rate": 4.116771312820189e-06,
+      "loss": 0.6505,
+      "step": 3668
+    },
+    {
+      "epoch": 0.2979777470965646,
+      "grad_norm": 3.49258490401417,
+      "learning_rate": 4.116269663973304e-06,
+      "loss": 0.537,
+      "step": 3669
+    },
+    {
+      "epoch": 0.2980589620726062,
+      "grad_norm": 8.038084077662122,
+      "learning_rate": 4.115767903289112e-06,
+      "loss": 0.6225,
+      "step": 3670
+    },
+    {
+      "epoch": 0.29814017704864776,
+      "grad_norm": 10.562283989123358,
+      "learning_rate": 4.115266030802332e-06,
+      "loss": 0.4825,
+      "step": 3671
+    },
+    {
+      "epoch": 0.29822139202468934,
+      "grad_norm": 5.676004986730943,
+      "learning_rate": 4.114764046547691e-06,
+      "loss": 0.59,
+      "step": 3672
+    },
+    {
+      "epoch": 0.2983026070007309,
+      "grad_norm": 8.432493876901468,
+      "learning_rate": 4.114261950559924e-06,
+      "loss": 0.5298,
+      "step": 3673
+    },
+    {
+      "epoch": 0.2983838219767725,
+      "grad_norm": 4.1760389034352015,
+      "learning_rate": 4.113759742873774e-06,
+      "loss": 0.6273,
+      "step": 3674
+    },
+    {
+      "epoch": 0.2984650369528141,
+      "grad_norm": 6.694938489409981,
+      "learning_rate": 4.11325742352399e-06,
+      "loss": 0.4842,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2985462519288557,
+      "grad_norm": 4.346992428927781,
+      "learning_rate": 4.112754992545331e-06,
+      "loss": 0.5768,
+      "step": 3676
+    },
+    {
+      "epoch": 0.2986274669048973,
+      "grad_norm": 4.528058186506179,
+      "learning_rate": 4.112252449972562e-06,
+      "loss": 0.5956,
+      "step": 3677
+    },
+    {
+      "epoch": 0.29870868188093885,
+      "grad_norm": 4.764966099793393,
+      "learning_rate": 4.111749795840455e-06,
+      "loss": 0.5005,
+      "step": 3678
+    },
+    {
+      "epoch": 0.29878989685698043,
+      "grad_norm": 10.20025692058819,
+      "learning_rate": 4.111247030183793e-06,
+      "loss": 0.5522,
+      "step": 3679
+    },
+    {
+      "epoch": 0.298871111833022,
+      "grad_norm": 31.38171024066769,
+      "learning_rate": 4.110744153037363e-06,
+      "loss": 0.505,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2989523268090636,
+      "grad_norm": 4.523115084329832,
+      "learning_rate": 4.110241164435964e-06,
+      "loss": 0.4673,
+      "step": 3681
+    },
+    {
+      "epoch": 0.29903354178510516,
+      "grad_norm": 3.7531700099140393,
+      "learning_rate": 4.109738064414397e-06,
+      "loss": 0.4812,
+      "step": 3682
+    },
+    {
+      "epoch": 0.29911475676114674,
+      "grad_norm": 5.06537300558853,
+      "learning_rate": 4.1092348530074764e-06,
+      "loss": 0.5328,
+      "step": 3683
+    },
+    {
+      "epoch": 0.2991959717371883,
+      "grad_norm": 6.176828372115656,
+      "learning_rate": 4.10873153025002e-06,
+      "loss": 0.4383,
+      "step": 3684
+    },
+    {
+      "epoch": 0.2992771867132299,
+      "grad_norm": 5.037654343199056,
+      "learning_rate": 4.108228096176856e-06,
+      "loss": 0.6148,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2993584016892715,
+      "grad_norm": 6.942767647531231,
+      "learning_rate": 4.10772455082282e-06,
+      "loss": 0.558,
+      "step": 3686
+    },
+    {
+      "epoch": 0.2994396166653131,
+      "grad_norm": 4.277595045669092,
+      "learning_rate": 4.107220894222753e-06,
+      "loss": 0.7493,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2995208316413547,
+      "grad_norm": 6.128933496939369,
+      "learning_rate": 4.106717126411506e-06,
+      "loss": 0.4093,
+      "step": 3688
+    },
+    {
+      "epoch": 0.29960204661739626,
+      "grad_norm": 3.4094618806546713,
+      "learning_rate": 4.106213247423938e-06,
+      "loss": 0.5386,
+      "step": 3689
+    },
+    {
+      "epoch": 0.29968326159343783,
+      "grad_norm": 5.429499060264869,
+      "learning_rate": 4.105709257294914e-06,
+      "loss": 0.5606,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2997644765694794,
+      "grad_norm": 8.126911051862221,
+      "learning_rate": 4.105205156059307e-06,
+      "loss": 0.4964,
+      "step": 3691
+    },
+    {
+      "epoch": 0.299845691545521,
+      "grad_norm": 5.576298258657682,
+      "learning_rate": 4.104700943751999e-06,
+      "loss": 0.5032,
+      "step": 3692
+    },
+    {
+      "epoch": 0.29992690652156256,
+      "grad_norm": 5.428256954230044,
+      "learning_rate": 4.104196620407878e-06,
+      "loss": 0.5164,
+      "step": 3693
+    },
+    {
+      "epoch": 0.30000812149760414,
+      "grad_norm": 5.490233849085779,
+      "learning_rate": 4.1036921860618415e-06,
+      "loss": 0.4619,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3000893364736457,
+      "grad_norm": 3.6578438271498683,
+      "learning_rate": 4.103187640748792e-06,
+      "loss": 0.4709,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3001705514496873,
+      "grad_norm": 4.119105065976182,
+      "learning_rate": 4.102682984503644e-06,
+      "loss": 0.5943,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3002517664257289,
+      "grad_norm": 4.816227234767905,
+      "learning_rate": 4.102178217361315e-06,
+      "loss": 0.5706,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3003329814017705,
+      "grad_norm": 3.586911969028518,
+      "learning_rate": 4.101673339356733e-06,
+      "loss": 0.4774,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3004141963778121,
+      "grad_norm": 9.66099144536593,
+      "learning_rate": 4.101168350524832e-06,
+      "loss": 0.4695,
+      "step": 3699
+    },
+    {
+      "epoch": 0.30049541135385366,
+      "grad_norm": 4.619743032013704,
+      "learning_rate": 4.100663250900556e-06,
+      "loss": 0.4937,
+      "step": 3700
+    },
+    {
+      "epoch": 0.30057662632989524,
+      "grad_norm": 5.0528198312806705,
+      "learning_rate": 4.100158040518854e-06,
+      "loss": 0.4896,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3006578413059368,
+      "grad_norm": 6.192798320269057,
+      "learning_rate": 4.099652719414684e-06,
+      "loss": 0.4393,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3007390562819784,
+      "grad_norm": 4.544818358602807,
+      "learning_rate": 4.099147287623012e-06,
+      "loss": 0.4435,
+      "step": 3703
+    },
+    {
+      "epoch": 0.30082027125801997,
+      "grad_norm": 3.843568388195506,
+      "learning_rate": 4.098641745178812e-06,
+      "loss": 0.5758,
+      "step": 3704
+    },
+    {
+      "epoch": 0.30090148623406154,
+      "grad_norm": 7.032873685349521,
+      "learning_rate": 4.098136092117063e-06,
+      "loss": 0.5837,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3009827012101031,
+      "grad_norm": 4.039586791406998,
+      "learning_rate": 4.097630328472755e-06,
+      "loss": 0.4516,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3010639161861447,
+      "grad_norm": 5.893186542446594,
+      "learning_rate": 4.097124454280883e-06,
+      "loss": 0.4981,
+      "step": 3707
+    },
+    {
+      "epoch": 0.30114513116218633,
+      "grad_norm": 4.234572464495813,
+      "learning_rate": 4.096618469576451e-06,
+      "loss": 0.721,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3012263461382279,
+      "grad_norm": 4.138584167113376,
+      "learning_rate": 4.0961123743944715e-06,
+      "loss": 0.4969,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3013075611142695,
+      "grad_norm": 4.733715325450099,
+      "learning_rate": 4.095606168769964e-06,
+      "loss": 0.65,
+      "step": 3710
+    },
+    {
+      "epoch": 0.30138877609031106,
+      "grad_norm": 5.217622232509772,
+      "learning_rate": 4.095099852737953e-06,
+      "loss": 0.6026,
+      "step": 3711
+    },
+    {
+      "epoch": 0.30146999106635264,
+      "grad_norm": 6.573735753701009,
+      "learning_rate": 4.094593426333474e-06,
+      "loss": 0.6201,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3015512060423942,
+      "grad_norm": 7.983639715927653,
+      "learning_rate": 4.09408688959157e-06,
+      "loss": 0.6581,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3016324210184358,
+      "grad_norm": 5.652732617432702,
+      "learning_rate": 4.093580242547289e-06,
+      "loss": 0.7463,
+      "step": 3714
+    },
+    {
+      "epoch": 0.30171363599447737,
+      "grad_norm": 5.775875513388475,
+      "learning_rate": 4.09307348523569e-06,
+      "loss": 0.4833,
+      "step": 3715
+    },
+    {
+      "epoch": 0.30179485097051895,
+      "grad_norm": 5.373772908608717,
+      "learning_rate": 4.092566617691837e-06,
+      "loss": 0.4648,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3018760659465605,
+      "grad_norm": 4.681484734835625,
+      "learning_rate": 4.092059639950802e-06,
+      "loss": 0.543,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3019572809226021,
+      "grad_norm": 3.293957563526287,
+      "learning_rate": 4.0915525520476665e-06,
+      "loss": 0.7871,
+      "step": 3718
+    },
+    {
+      "epoch": 0.30203849589864373,
+      "grad_norm": 9.736689094555402,
+      "learning_rate": 4.091045354017517e-06,
+      "loss": 0.54,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3021197108746853,
+      "grad_norm": 7.004070985864832,
+      "learning_rate": 4.090538045895449e-06,
+      "loss": 0.4686,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3022009258507269,
+      "grad_norm": 6.304287102750301,
+      "learning_rate": 4.090030627716567e-06,
+      "loss": 0.6621,
+      "step": 3721
+    },
+    {
+      "epoch": 0.30228214082676846,
+      "grad_norm": 5.678988794221106,
+      "learning_rate": 4.08952309951598e-06,
+      "loss": 0.6359,
+      "step": 3722
+    },
+    {
+      "epoch": 0.30236335580281004,
+      "grad_norm": 7.331654562862624,
+      "learning_rate": 4.0890154613288066e-06,
+      "loss": 0.6547,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3024445707788516,
+      "grad_norm": 3.7966525875405277,
+      "learning_rate": 4.088507713190174e-06,
+      "loss": 0.5675,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3025257857548932,
+      "grad_norm": 4.688025131548344,
+      "learning_rate": 4.087999855135215e-06,
+      "loss": 0.6325,
+      "step": 3725
+    },
+    {
+      "epoch": 0.30260700073093477,
+      "grad_norm": 4.823758976025031,
+      "learning_rate": 4.087491887199069e-06,
+      "loss": 0.5405,
+      "step": 3726
+    },
+    {
+      "epoch": 0.30268821570697635,
+      "grad_norm": 3.9620946246180884,
+      "learning_rate": 4.086983809416887e-06,
+      "loss": 0.4993,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3027694306830179,
+      "grad_norm": 4.5032403641906305,
+      "learning_rate": 4.086475621823824e-06,
+      "loss": 0.4079,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3028506456590595,
+      "grad_norm": 20.477342042531873,
+      "learning_rate": 4.085967324455045e-06,
+      "loss": 0.5201,
+      "step": 3729
+    },
+    {
+      "epoch": 0.30293186063510114,
+      "grad_norm": 6.231510482092364,
+      "learning_rate": 4.085458917345721e-06,
+      "loss": 0.7878,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3030130756111427,
+      "grad_norm": 4.154775199567728,
+      "learning_rate": 4.084950400531029e-06,
+      "loss": 0.5936,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3030942905871843,
+      "grad_norm": 4.694151372288833,
+      "learning_rate": 4.0844417740461586e-06,
+      "loss": 0.4481,
+      "step": 3732
+    },
+    {
+      "epoch": 0.30317550556322587,
+      "grad_norm": 4.709650246764357,
+      "learning_rate": 4.083933037926303e-06,
+      "loss": 0.6051,
+      "step": 3733
+    },
+    {
+      "epoch": 0.30325672053926744,
+      "grad_norm": 5.029331738383321,
+      "learning_rate": 4.0834241922066644e-06,
+      "loss": 0.5312,
+      "step": 3734
+    },
+    {
+      "epoch": 0.303337935515309,
+      "grad_norm": 4.293546250505703,
+      "learning_rate": 4.082915236922451e-06,
+      "loss": 0.5698,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3034191504913506,
+      "grad_norm": 3.9815078810069218,
+      "learning_rate": 4.082406172108882e-06,
+      "loss": 0.6138,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3035003654673922,
+      "grad_norm": 8.25343748834109,
+      "learning_rate": 4.0818969978011795e-06,
+      "loss": 0.5962,
+      "step": 3737
+    },
+    {
+      "epoch": 0.30358158044343375,
+      "grad_norm": 8.293232016506174,
+      "learning_rate": 4.081387714034577e-06,
+      "loss": 0.5964,
+      "step": 3738
+    },
+    {
+      "epoch": 0.30366279541947533,
+      "grad_norm": 9.876486919775553,
+      "learning_rate": 4.080878320844315e-06,
+      "loss": 0.5288,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3037440103955169,
+      "grad_norm": 4.90237518330183,
+      "learning_rate": 4.080368818265639e-06,
+      "loss": 0.542,
+      "step": 3740
+    },
+    {
+      "epoch": 0.30382522537155854,
+      "grad_norm": 5.433619699450851,
+      "learning_rate": 4.079859206333805e-06,
+      "loss": 0.5346,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3039064403476001,
+      "grad_norm": 6.417127497404236,
+      "learning_rate": 4.079349485084074e-06,
+      "loss": 0.5068,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3039876553236417,
+      "grad_norm": 9.395212073819222,
+      "learning_rate": 4.078839654551718e-06,
+      "loss": 0.5981,
+      "step": 3743
+    },
+    {
+      "epoch": 0.30406887029968327,
+      "grad_norm": 6.080896733160445,
+      "learning_rate": 4.078329714772015e-06,
+      "loss": 0.5716,
+      "step": 3744
+    },
+    {
+      "epoch": 0.30415008527572485,
+      "grad_norm": 3.9795202208766463,
+      "learning_rate": 4.0778196657802484e-06,
+      "loss": 0.589,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3042313002517664,
+      "grad_norm": 5.978058004337624,
+      "learning_rate": 4.077309507611711e-06,
+      "loss": 0.559,
+      "step": 3746
+    },
+    {
+      "epoch": 0.304312515227808,
+      "grad_norm": 10.73728162343595,
+      "learning_rate": 4.076799240301703e-06,
+      "loss": 0.5419,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3043937302038496,
+      "grad_norm": 5.705981837585083,
+      "learning_rate": 4.076288863885533e-06,
+      "loss": 0.5277,
+      "step": 3748
+    },
+    {
+      "epoch": 0.30447494517989115,
+      "grad_norm": 4.148490626037343,
+      "learning_rate": 4.0757783783985164e-06,
+      "loss": 0.6104,
+      "step": 3749
+    },
+    {
+      "epoch": 0.30455616015593273,
+      "grad_norm": 7.3071974055301006,
+      "learning_rate": 4.0752677838759755e-06,
+      "loss": 0.6195,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3046373751319743,
+      "grad_norm": 3.536531811654719,
+      "learning_rate": 4.074757080353241e-06,
+      "loss": 0.6436,
+      "step": 3751
+    },
+    {
+      "epoch": 0.30471859010801594,
+      "grad_norm": 4.25870092662668,
+      "learning_rate": 4.074246267865652e-06,
+      "loss": 0.6274,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3047998050840575,
+      "grad_norm": 3.6079741565893895,
+      "learning_rate": 4.073735346448551e-06,
+      "loss": 0.6034,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3048810200600991,
+      "grad_norm": 7.935117748318247,
+      "learning_rate": 4.073224316137293e-06,
+      "loss": 0.5586,
+      "step": 3754
+    },
+    {
+      "epoch": 0.30496223503614067,
+      "grad_norm": 6.813447251757785,
+      "learning_rate": 4.072713176967239e-06,
+      "loss": 0.5641,
+      "step": 3755
+    },
+    {
+      "epoch": 0.30504345001218225,
+      "grad_norm": 6.715619033709069,
+      "learning_rate": 4.072201928973757e-06,
+      "loss": 0.5836,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3051246649882238,
+      "grad_norm": 3.5300282565737944,
+      "learning_rate": 4.071690572192222e-06,
+      "loss": 0.6304,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3052058799642654,
+      "grad_norm": 9.284044398768938,
+      "learning_rate": 4.071179106658017e-06,
+      "loss": 0.5776,
+      "step": 3758
+    },
+    {
+      "epoch": 0.305287094940307,
+      "grad_norm": 6.469344755338421,
+      "learning_rate": 4.070667532406534e-06,
+      "loss": 0.518,
+      "step": 3759
+    },
+    {
+      "epoch": 0.30536830991634856,
+      "grad_norm": 9.204050268328828,
+      "learning_rate": 4.070155849473169e-06,
+      "loss": 0.4228,
+      "step": 3760
+    },
+    {
+      "epoch": 0.30544952489239013,
+      "grad_norm": 7.33932480094826,
+      "learning_rate": 4.06964405789333e-06,
+      "loss": 0.5276,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3055307398684317,
+      "grad_norm": 4.5551115854633695,
+      "learning_rate": 4.06913215770243e-06,
+      "loss": 0.5275,
+      "step": 3762
+    },
+    {
+      "epoch": 0.30561195484447334,
+      "grad_norm": 7.628505640253093,
+      "learning_rate": 4.068620148935889e-06,
+      "loss": 0.5706,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3056931698205149,
+      "grad_norm": 7.296776756806712,
+      "learning_rate": 4.0681080316291355e-06,
+      "loss": 0.4332,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3057743847965565,
+      "grad_norm": 6.211933336592377,
+      "learning_rate": 4.067595805817604e-06,
+      "loss": 0.5385,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3058555997725981,
+      "grad_norm": 5.2682390368881125,
+      "learning_rate": 4.0670834715367405e-06,
+      "loss": 0.6172,
+      "step": 3766
+    },
+    {
+      "epoch": 0.30593681474863965,
+      "grad_norm": 5.7266015459873465,
+      "learning_rate": 4.066571028821994e-06,
+      "loss": 0.663,
+      "step": 3767
+    },
+    {
+      "epoch": 0.30601802972468123,
+      "grad_norm": 4.566744030336992,
+      "learning_rate": 4.066058477708824e-06,
+      "loss": 0.6204,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3060992447007228,
+      "grad_norm": 5.241528003933075,
+      "learning_rate": 4.065545818232695e-06,
+      "loss": 0.5374,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3061804596767644,
+      "grad_norm": 4.210043089768735,
+      "learning_rate": 4.06503305042908e-06,
+      "loss": 0.5627,
+      "step": 3770
+    },
+    {
+      "epoch": 0.30626167465280596,
+      "grad_norm": 24.358962625710497,
+      "learning_rate": 4.064520174333462e-06,
+      "loss": 0.4803,
+      "step": 3771
+    },
+    {
+      "epoch": 0.30634288962884754,
+      "grad_norm": 5.001904028877196,
+      "learning_rate": 4.0640071899813284e-06,
+      "loss": 0.5884,
+      "step": 3772
+    },
+    {
+      "epoch": 0.30642410460488917,
+      "grad_norm": 3.685895651696086,
+      "learning_rate": 4.0634940974081735e-06,
+      "loss": 0.5976,
+      "step": 3773
+    },
+    {
+      "epoch": 0.30650531958093075,
+      "grad_norm": 4.690707174624639,
+      "learning_rate": 4.062980896649502e-06,
+      "loss": 0.4432,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3065865345569723,
+      "grad_norm": 7.66057017392417,
+      "learning_rate": 4.062467587740825e-06,
+      "loss": 0.5297,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3066677495330139,
+      "grad_norm": 4.80151641086556,
+      "learning_rate": 4.0619541707176595e-06,
+      "loss": 0.5717,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3067489645090555,
+      "grad_norm": 5.073279099762993,
+      "learning_rate": 4.061440645615532e-06,
+      "loss": 0.4853,
+      "step": 3777
+    },
+    {
+      "epoch": 0.30683017948509705,
+      "grad_norm": 6.406852301630108,
+      "learning_rate": 4.060927012469976e-06,
+      "loss": 0.5093,
+      "step": 3778
+    },
+    {
+      "epoch": 0.30691139446113863,
+      "grad_norm": 3.919683009512748,
+      "learning_rate": 4.060413271316531e-06,
+      "loss": 0.4647,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3069926094371802,
+      "grad_norm": 4.810874854512679,
+      "learning_rate": 4.059899422190747e-06,
+      "loss": 0.3816,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3070738244132218,
+      "grad_norm": 7.470318516960077,
+      "learning_rate": 4.059385465128179e-06,
+      "loss": 0.4818,
+      "step": 3781
+    },
+    {
+      "epoch": 0.30715503938926336,
+      "grad_norm": 4.142497626949441,
+      "learning_rate": 4.058871400164388e-06,
+      "loss": 0.6359,
+      "step": 3782
+    },
+    {
+      "epoch": 0.30723625436530494,
+      "grad_norm": 4.107423918701493,
+      "learning_rate": 4.058357227334947e-06,
+      "loss": 0.5755,
+      "step": 3783
+    },
+    {
+      "epoch": 0.30731746934134657,
+      "grad_norm": 7.190290906025693,
+      "learning_rate": 4.057842946675434e-06,
+      "loss": 0.4618,
+      "step": 3784
+    },
+    {
+      "epoch": 0.30739868431738815,
+      "grad_norm": 6.779887515845196,
+      "learning_rate": 4.057328558221434e-06,
+      "loss": 0.4605,
+      "step": 3785
+    },
+    {
+      "epoch": 0.3074798992934297,
+      "grad_norm": 6.8626494300279495,
+      "learning_rate": 4.056814062008539e-06,
+      "loss": 0.4735,
+      "step": 3786
+    },
+    {
+      "epoch": 0.3075611142694713,
+      "grad_norm": 4.43663040407582,
+      "learning_rate": 4.056299458072351e-06,
+      "loss": 0.5549,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3076423292455129,
+      "grad_norm": 9.927091358251289,
+      "learning_rate": 4.0557847464484766e-06,
+      "loss": 0.6012,
+      "step": 3788
+    },
+    {
+      "epoch": 0.30772354422155446,
+      "grad_norm": 3.3047332458402874,
+      "learning_rate": 4.055269927172532e-06,
+      "loss": 0.5754,
+      "step": 3789
+    },
+    {
+      "epoch": 0.30780475919759603,
+      "grad_norm": 4.086984576275844,
+      "learning_rate": 4.054755000280139e-06,
+      "loss": 0.6482,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3078859741736376,
+      "grad_norm": 10.124343594059361,
+      "learning_rate": 4.054239965806929e-06,
+      "loss": 0.6541,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3079671891496792,
+      "grad_norm": 9.088285597560201,
+      "learning_rate": 4.053724823788538e-06,
+      "loss": 0.5313,
+      "step": 3792
+    },
+    {
+      "epoch": 0.30804840412572077,
+      "grad_norm": 10.64301926225871,
+      "learning_rate": 4.053209574260614e-06,
+      "loss": 0.5672,
+      "step": 3793
+    },
+    {
+      "epoch": 0.30812961910176234,
+      "grad_norm": 4.299438598485095,
+      "learning_rate": 4.052694217258806e-06,
+      "loss": 0.4908,
+      "step": 3794
+    },
+    {
+      "epoch": 0.308210834077804,
+      "grad_norm": 50.64834820262253,
+      "learning_rate": 4.052178752818776e-06,
+      "loss": 0.4808,
+      "step": 3795
+    },
+    {
+      "epoch": 0.30829204905384555,
+      "grad_norm": 3.318247558859262,
+      "learning_rate": 4.051663180976192e-06,
+      "loss": 0.6804,
+      "step": 3796
+    },
+    {
+      "epoch": 0.30837326402988713,
+      "grad_norm": 15.92488834517386,
+      "learning_rate": 4.051147501766727e-06,
+      "loss": 0.5227,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3084544790059287,
+      "grad_norm": 3.980666622212895,
+      "learning_rate": 4.050631715226064e-06,
+      "loss": 0.5733,
+      "step": 3798
+    },
+    {
+      "epoch": 0.3085356939819703,
+      "grad_norm": 5.924342358250464,
+      "learning_rate": 4.050115821389894e-06,
+      "loss": 0.4722,
+      "step": 3799
+    },
+    {
+      "epoch": 0.30861690895801186,
+      "grad_norm": 6.826343924107241,
+      "learning_rate": 4.049599820293913e-06,
+      "loss": 0.5608,
+      "step": 3800
+    },
+    {
+      "epoch": 0.30869812393405344,
+      "grad_norm": 5.798254498370531,
+      "learning_rate": 4.049083711973824e-06,
+      "loss": 0.482,
+      "step": 3801
+    },
+    {
+      "epoch": 0.308779338910095,
+      "grad_norm": 4.3433768010776905,
+      "learning_rate": 4.0485674964653424e-06,
+      "loss": 0.4834,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3088605538861366,
+      "grad_norm": 4.326553027616668,
+      "learning_rate": 4.048051173804185e-06,
+      "loss": 0.5556,
+      "step": 3803
+    },
+    {
+      "epoch": 0.30894176886217817,
+      "grad_norm": 6.800606044283711,
+      "learning_rate": 4.047534744026079e-06,
+      "loss": 0.6,
+      "step": 3804
+    },
+    {
+      "epoch": 0.30902298383821974,
+      "grad_norm": 10.590880966190726,
+      "learning_rate": 4.04701820716676e-06,
+      "loss": 0.4796,
+      "step": 3805
+    },
+    {
+      "epoch": 0.3091041988142614,
+      "grad_norm": 5.6795091144428635,
+      "learning_rate": 4.046501563261968e-06,
+      "loss": 0.5129,
+      "step": 3806
+    },
+    {
+      "epoch": 0.30918541379030295,
+      "grad_norm": 3.210645271484872,
+      "learning_rate": 4.045984812347452e-06,
+      "loss": 0.5332,
+      "step": 3807
+    },
+    {
+      "epoch": 0.30926662876634453,
+      "grad_norm": 6.294410231720947,
+      "learning_rate": 4.045467954458969e-06,
+      "loss": 0.6181,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3093478437423861,
+      "grad_norm": 4.474575703083307,
+      "learning_rate": 4.044950989632283e-06,
+      "loss": 0.4792,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3094290587184277,
+      "grad_norm": 4.779056570525405,
+      "learning_rate": 4.044433917903166e-06,
+      "loss": 0.5667,
+      "step": 3810
+    },
+    {
+      "epoch": 0.30951027369446926,
+      "grad_norm": 4.782139569666095,
+      "learning_rate": 4.043916739307394e-06,
+      "loss": 0.7153,
+      "step": 3811
+    },
+    {
+      "epoch": 0.30959148867051084,
+      "grad_norm": 6.550010939029318,
+      "learning_rate": 4.0433994538807564e-06,
+      "loss": 0.5078,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3096727036465524,
+      "grad_norm": 7.70154315773568,
+      "learning_rate": 4.042882061659043e-06,
+      "loss": 0.5411,
+      "step": 3813
+    },
+    {
+      "epoch": 0.309753918622594,
+      "grad_norm": 5.468435592136346,
+      "learning_rate": 4.042364562678059e-06,
+      "loss": 0.6321,
+      "step": 3814
+    },
+    {
+      "epoch": 0.30983513359863557,
+      "grad_norm": 4.143944075441312,
+      "learning_rate": 4.041846956973608e-06,
+      "loss": 0.46,
+      "step": 3815
+    },
+    {
+      "epoch": 0.30991634857467715,
+      "grad_norm": 4.798354632196031,
+      "learning_rate": 4.041329244581509e-06,
+      "loss": 0.4921,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3099975635507188,
+      "grad_norm": 5.155312927681158,
+      "learning_rate": 4.040811425537583e-06,
+      "loss": 0.6239,
+      "step": 3817
+    },
+    {
+      "epoch": 0.31007877852676036,
+      "grad_norm": 3.701165442814568,
+      "learning_rate": 4.040293499877661e-06,
+      "loss": 0.6161,
+      "step": 3818
+    },
+    {
+      "epoch": 0.31015999350280193,
+      "grad_norm": 5.396166406448343,
+      "learning_rate": 4.039775467637581e-06,
+      "loss": 0.4338,
+      "step": 3819
+    },
+    {
+      "epoch": 0.3102412084788435,
+      "grad_norm": 4.72888620508123,
+      "learning_rate": 4.039257328853188e-06,
+      "loss": 0.7018,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3103224234548851,
+      "grad_norm": 4.775493593228451,
+      "learning_rate": 4.038739083560334e-06,
+      "loss": 0.526,
+      "step": 3821
+    },
+    {
+      "epoch": 0.31040363843092666,
+      "grad_norm": 4.103968235635027,
+      "learning_rate": 4.038220731794878e-06,
+      "loss": 0.6318,
+      "step": 3822
+    },
+    {
+      "epoch": 0.31048485340696824,
+      "grad_norm": 8.31477225841899,
+      "learning_rate": 4.03770227359269e-06,
+      "loss": 0.5303,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3105660683830098,
+      "grad_norm": 3.8725355108338766,
+      "learning_rate": 4.037183708989642e-06,
+      "loss": 0.4775,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3106472833590514,
+      "grad_norm": 7.614229033887325,
+      "learning_rate": 4.0366650380216165e-06,
+      "loss": 0.4905,
+      "step": 3825
+    },
+    {
+      "epoch": 0.310728498335093,
+      "grad_norm": 5.153358446059254,
+      "learning_rate": 4.036146260724503e-06,
+      "loss": 0.5249,
+      "step": 3826
+    },
+    {
+      "epoch": 0.31080971331113455,
+      "grad_norm": 8.578326851480215,
+      "learning_rate": 4.0356273771341984e-06,
+      "loss": 0.6979,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3108909282871762,
+      "grad_norm": 5.785574362196001,
+      "learning_rate": 4.035108387286607e-06,
+      "loss": 0.4557,
+      "step": 3828
+    },
+    {
+      "epoch": 0.31097214326321776,
+      "grad_norm": 3.55379778886365,
+      "learning_rate": 4.03458929121764e-06,
+      "loss": 0.6176,
+      "step": 3829
+    },
+    {
+      "epoch": 0.31105335823925934,
+      "grad_norm": 7.612980594268243,
+      "learning_rate": 4.0340700889632145e-06,
+      "loss": 0.7673,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3111345732153009,
+      "grad_norm": 9.252507294094109,
+      "learning_rate": 4.033550780559258e-06,
+      "loss": 0.5036,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3112157881913425,
+      "grad_norm": 3.594686390622405,
+      "learning_rate": 4.033031366041704e-06,
+      "loss": 0.4821,
+      "step": 3832
+    },
+    {
+      "epoch": 0.31129700316738407,
+      "grad_norm": 6.285694736961024,
+      "learning_rate": 4.0325118454464935e-06,
+      "loss": 0.5007,
+      "step": 3833
+    },
+    {
+      "epoch": 0.31137821814342564,
+      "grad_norm": 3.160762643660874,
+      "learning_rate": 4.031992218809573e-06,
+      "loss": 0.5703,
+      "step": 3834
+    },
+    {
+      "epoch": 0.3114594331194672,
+      "grad_norm": 8.890289793696613,
+      "learning_rate": 4.0314724861669e-06,
+      "loss": 0.5602,
+      "step": 3835
+    },
+    {
+      "epoch": 0.3115406480955088,
+      "grad_norm": 4.321571181434268,
+      "learning_rate": 4.0309526475544355e-06,
+      "loss": 0.6061,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3116218630715504,
+      "grad_norm": 7.365273193938057,
+      "learning_rate": 4.03043270300815e-06,
+      "loss": 0.5242,
+      "step": 3837
+    },
+    {
+      "epoch": 0.31170307804759195,
+      "grad_norm": 6.152202896973093,
+      "learning_rate": 4.029912652564022e-06,
+      "loss": 0.527,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3117842930236336,
+      "grad_norm": 9.939281394272458,
+      "learning_rate": 4.029392496258035e-06,
+      "loss": 0.6952,
+      "step": 3839
+    },
+    {
+      "epoch": 0.31186550799967516,
+      "grad_norm": 7.9876078180108685,
+      "learning_rate": 4.028872234126181e-06,
+      "loss": 0.4901,
+      "step": 3840
+    },
+    {
+      "epoch": 0.31194672297571674,
+      "grad_norm": 4.522557141556952,
+      "learning_rate": 4.02835186620446e-06,
+      "loss": 0.4127,
+      "step": 3841
+    },
+    {
+      "epoch": 0.3120279379517583,
+      "grad_norm": 5.5314558917789505,
+      "learning_rate": 4.027831392528879e-06,
+      "loss": 0.5045,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3121091529277999,
+      "grad_norm": 6.372253503387104,
+      "learning_rate": 4.027310813135451e-06,
+      "loss": 0.5654,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31219036790384147,
+      "grad_norm": 3.5189797244740446,
+      "learning_rate": 4.0267901280601985e-06,
+      "loss": 0.4954,
+      "step": 3844
+    },
+    {
+      "epoch": 0.31227158287988305,
+      "grad_norm": 6.001913797204444,
+      "learning_rate": 4.026269337339149e-06,
+      "loss": 0.5535,
+      "step": 3845
+    },
+    {
+      "epoch": 0.3123527978559246,
+      "grad_norm": 6.805227975140487,
+      "learning_rate": 4.025748441008339e-06,
+      "loss": 0.5175,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3124340128319662,
+      "grad_norm": 3.6933697953285423,
+      "learning_rate": 4.0252274391038125e-06,
+      "loss": 0.6538,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3125152278080078,
+      "grad_norm": 4.975606249432986,
+      "learning_rate": 4.024706331661618e-06,
+      "loss": 0.6194,
+      "step": 3848
+    },
+    {
+      "epoch": 0.31259644278404936,
+      "grad_norm": 21.413001819693548,
+      "learning_rate": 4.024185118717816e-06,
+      "loss": 0.5108,
+      "step": 3849
+    },
+    {
+      "epoch": 0.312677657760091,
+      "grad_norm": 4.688146949121658,
+      "learning_rate": 4.023663800308471e-06,
+      "loss": 0.5959,
+      "step": 3850
+    },
+    {
+      "epoch": 0.31275887273613256,
+      "grad_norm": 4.929413723785714,
+      "learning_rate": 4.023142376469653e-06,
+      "loss": 0.4974,
+      "step": 3851
+    },
+    {
+      "epoch": 0.31284008771217414,
+      "grad_norm": 7.165568666050521,
+      "learning_rate": 4.022620847237445e-06,
+      "loss": 0.5106,
+      "step": 3852
+    },
+    {
+      "epoch": 0.3129213026882157,
+      "grad_norm": 5.314358756089364,
+      "learning_rate": 4.022099212647933e-06,
+      "loss": 0.6104,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3130025176642573,
+      "grad_norm": 4.330125714880257,
+      "learning_rate": 4.021577472737209e-06,
+      "loss": 0.587,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3130837326402989,
+      "grad_norm": 4.654860553507325,
+      "learning_rate": 4.021055627541379e-06,
+      "loss": 0.4938,
+      "step": 3855
+    },
+    {
+      "epoch": 0.31316494761634045,
+      "grad_norm": 5.006258634790825,
+      "learning_rate": 4.020533677096549e-06,
+      "loss": 0.5747,
+      "step": 3856
+    },
+    {
+      "epoch": 0.313246162592382,
+      "grad_norm": 6.0280062807614545,
+      "learning_rate": 4.020011621438836e-06,
+      "loss": 0.5764,
+      "step": 3857
+    },
+    {
+      "epoch": 0.3133273775684236,
+      "grad_norm": 6.574325175703227,
+      "learning_rate": 4.019489460604364e-06,
+      "loss": 0.5299,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3134085925444652,
+      "grad_norm": 4.080793981103117,
+      "learning_rate": 4.018967194629261e-06,
+      "loss": 0.7868,
+      "step": 3859
+    },
+    {
+      "epoch": 0.31348980752050676,
+      "grad_norm": 3.847772634998914,
+      "learning_rate": 4.0184448235496685e-06,
+      "loss": 0.6058,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3135710224965484,
+      "grad_norm": 4.404444873697644,
+      "learning_rate": 4.017922347401731e-06,
+      "loss": 0.523,
+      "step": 3861
+    },
+    {
+      "epoch": 0.31365223747258997,
+      "grad_norm": 9.45105154996585,
+      "learning_rate": 4.017399766221599e-06,
+      "loss": 0.4319,
+      "step": 3862
+    },
+    {
+      "epoch": 0.31373345244863154,
+      "grad_norm": 4.265430944389896,
+      "learning_rate": 4.016877080045435e-06,
+      "loss": 0.4388,
+      "step": 3863
+    },
+    {
+      "epoch": 0.3138146674246731,
+      "grad_norm": 4.626292776222082,
+      "learning_rate": 4.016354288909405e-06,
+      "loss": 0.4834,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3138958824007147,
+      "grad_norm": 3.893386508353787,
+      "learning_rate": 4.0158313928496826e-06,
+      "loss": 0.5888,
+      "step": 3865
+    },
+    {
+      "epoch": 0.3139770973767563,
+      "grad_norm": 5.310652622885137,
+      "learning_rate": 4.015308391902452e-06,
+      "loss": 0.5323,
+      "step": 3866
+    },
+    {
+      "epoch": 0.31405831235279785,
+      "grad_norm": 6.885983808946358,
+      "learning_rate": 4.014785286103898e-06,
+      "loss": 0.5397,
+      "step": 3867
+    },
+    {
+      "epoch": 0.31413952732883943,
+      "grad_norm": 7.793035367661995,
+      "learning_rate": 4.014262075490221e-06,
+      "loss": 0.4684,
+      "step": 3868
+    },
+    {
+      "epoch": 0.314220742304881,
+      "grad_norm": 14.532888308919345,
+      "learning_rate": 4.013738760097622e-06,
+      "loss": 0.5751,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3143019572809226,
+      "grad_norm": 6.316239281714061,
+      "learning_rate": 4.0132153399623106e-06,
+      "loss": 0.5754,
+      "step": 3870
+    },
+    {
+      "epoch": 0.31438317225696416,
+      "grad_norm": 5.218130390396817,
+      "learning_rate": 4.012691815120508e-06,
+      "loss": 0.5718,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3144643872330058,
+      "grad_norm": 10.486211867182627,
+      "learning_rate": 4.012168185608437e-06,
+      "loss": 0.4341,
+      "step": 3872
+    },
+    {
+      "epoch": 0.31454560220904737,
+      "grad_norm": 13.04476716032136,
+      "learning_rate": 4.011644451462331e-06,
+      "loss": 0.6967,
+      "step": 3873
+    },
+    {
+      "epoch": 0.31462681718508895,
+      "grad_norm": 4.374789125818755,
+      "learning_rate": 4.011120612718429e-06,
+      "loss": 0.5778,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3147080321611305,
+      "grad_norm": 4.427126765436375,
+      "learning_rate": 4.010596669412978e-06,
+      "loss": 0.5147,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3147892471371721,
+      "grad_norm": 6.121898739230059,
+      "learning_rate": 4.010072621582233e-06,
+      "loss": 0.4541,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3148704621132137,
+      "grad_norm": 8.191495131008242,
+      "learning_rate": 4.009548469262453e-06,
+      "loss": 0.7081,
+      "step": 3877
+    },
+    {
+      "epoch": 0.31495167708925526,
+      "grad_norm": 5.939103893735796,
+      "learning_rate": 4.009024212489909e-06,
+      "loss": 0.588,
+      "step": 3878
+    },
+    {
+      "epoch": 0.31503289206529683,
+      "grad_norm": 5.468809556850152,
+      "learning_rate": 4.0084998513008765e-06,
+      "loss": 0.5442,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3151141070413384,
+      "grad_norm": 3.760788030695487,
+      "learning_rate": 4.007975385731637e-06,
+      "loss": 0.771,
+      "step": 3880
+    },
+    {
+      "epoch": 0.31519532201738,
+      "grad_norm": 4.150500328342237,
+      "learning_rate": 4.007450815818481e-06,
+      "loss": 0.5091,
+      "step": 3881
+    },
+    {
+      "epoch": 0.31527653699342156,
+      "grad_norm": 4.7154167246286285,
+      "learning_rate": 4.0069261415977075e-06,
+      "loss": 0.5819,
+      "step": 3882
+    },
+    {
+      "epoch": 0.3153577519694632,
+      "grad_norm": 4.032822924231925,
+      "learning_rate": 4.006401363105621e-06,
+      "loss": 0.5721,
+      "step": 3883
+    },
+    {
+      "epoch": 0.3154389669455048,
+      "grad_norm": 4.862866759718993,
+      "learning_rate": 4.0058764803785325e-06,
+      "loss": 0.6929,
+      "step": 3884
+    },
+    {
+      "epoch": 0.31552018192154635,
+      "grad_norm": 3.918915992085951,
+      "learning_rate": 4.00535149345276e-06,
+      "loss": 0.5521,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3156013968975879,
+      "grad_norm": 5.105255048171445,
+      "learning_rate": 4.0048264023646325e-06,
+      "loss": 0.5472,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3156826118736295,
+      "grad_norm": 8.077151245590299,
+      "learning_rate": 4.004301207150482e-06,
+      "loss": 0.4525,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3157638268496711,
+      "grad_norm": 8.364488519427972,
+      "learning_rate": 4.003775907846648e-06,
+      "loss": 0.566,
+      "step": 3888
+    },
+    {
+      "epoch": 0.31584504182571266,
+      "grad_norm": 4.65775942487372,
+      "learning_rate": 4.003250504489481e-06,
+      "loss": 0.4639,
+      "step": 3889
+    },
+    {
+      "epoch": 0.31592625680175423,
+      "grad_norm": 10.681842547013567,
+      "learning_rate": 4.002724997115335e-06,
+      "loss": 0.3836,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3160074717777958,
+      "grad_norm": 4.910721299022231,
+      "learning_rate": 4.002199385760571e-06,
+      "loss": 0.5612,
+      "step": 3891
+    },
+    {
+      "epoch": 0.3160886867538374,
+      "grad_norm": 5.963214780659344,
+      "learning_rate": 4.001673670461561e-06,
+      "loss": 0.6232,
+      "step": 3892
+    },
+    {
+      "epoch": 0.31616990172987897,
+      "grad_norm": 6.652157701535889,
+      "learning_rate": 4.0011478512546805e-06,
+      "loss": 0.5664,
+      "step": 3893
+    },
+    {
+      "epoch": 0.3162511167059206,
+      "grad_norm": 3.8712273283479233,
+      "learning_rate": 4.000621928176313e-06,
+      "loss": 0.5419,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3163323316819622,
+      "grad_norm": 5.721147211228416,
+      "learning_rate": 4.000095901262851e-06,
+      "loss": 0.5002,
+      "step": 3895
+    },
+    {
+      "epoch": 0.31641354665800375,
+      "grad_norm": 5.522015478399547,
+      "learning_rate": 3.99956977055069e-06,
+      "loss": 0.5785,
+      "step": 3896
+    },
+    {
+      "epoch": 0.31649476163404533,
+      "grad_norm": 12.710524189514066,
+      "learning_rate": 3.999043536076238e-06,
+      "loss": 0.5987,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3165759766100869,
+      "grad_norm": 15.432871133292979,
+      "learning_rate": 3.998517197875908e-06,
+      "loss": 0.5758,
+      "step": 3898
+    },
+    {
+      "epoch": 0.3166571915861285,
+      "grad_norm": 3.2538000933185622,
+      "learning_rate": 3.997990755986117e-06,
+      "loss": 0.3758,
+      "step": 3899
+    },
+    {
+      "epoch": 0.31673840656217006,
+      "grad_norm": 6.504572502780165,
+      "learning_rate": 3.9974642104432945e-06,
+      "loss": 0.6192,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31681962153821164,
+      "grad_norm": 5.380343912168226,
+      "learning_rate": 3.996937561283874e-06,
+      "loss": 0.5603,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3169008365142532,
+      "grad_norm": 8.424493108304915,
+      "learning_rate": 3.996410808544296e-06,
+      "loss": 0.5061,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3169820514902948,
+      "grad_norm": 5.3477146060500225,
+      "learning_rate": 3.99588395226101e-06,
+      "loss": 0.5234,
+      "step": 3903
+    },
+    {
+      "epoch": 0.31706326646633637,
+      "grad_norm": 12.609875677912806,
+      "learning_rate": 3.9953569924704715e-06,
+      "loss": 0.5047,
+      "step": 3904
+    },
+    {
+      "epoch": 0.317144481442378,
+      "grad_norm": 6.884048976659566,
+      "learning_rate": 3.994829929209143e-06,
+      "loss": 0.5393,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3172256964184196,
+      "grad_norm": 5.4396250935337465,
+      "learning_rate": 3.994302762513496e-06,
+      "loss": 0.5701,
+      "step": 3906
+    },
+    {
+      "epoch": 0.31730691139446116,
+      "grad_norm": 3.9789051349804905,
+      "learning_rate": 3.993775492420005e-06,
+      "loss": 0.8038,
+      "step": 3907
+    },
+    {
+      "epoch": 0.31738812637050273,
+      "grad_norm": 6.0730946208938335,
+      "learning_rate": 3.993248118965155e-06,
+      "loss": 0.5355,
+      "step": 3908
+    },
+    {
+      "epoch": 0.3174693413465443,
+      "grad_norm": 7.073767654590764,
+      "learning_rate": 3.992720642185439e-06,
+      "loss": 0.4949,
+      "step": 3909
+    },
+    {
+      "epoch": 0.3175505563225859,
+      "grad_norm": 6.028367836015118,
+      "learning_rate": 3.992193062117354e-06,
+      "loss": 0.5065,
+      "step": 3910
+    },
+    {
+      "epoch": 0.31763177129862746,
+      "grad_norm": 5.478458346134022,
+      "learning_rate": 3.991665378797408e-06,
+      "loss": 0.6328,
+      "step": 3911
+    },
+    {
+      "epoch": 0.31771298627466904,
+      "grad_norm": 5.80032080288049,
+      "learning_rate": 3.991137592262111e-06,
+      "loss": 0.3775,
+      "step": 3912
+    },
+    {
+      "epoch": 0.3177942012507106,
+      "grad_norm": 5.364477054479598,
+      "learning_rate": 3.990609702547985e-06,
+      "loss": 0.6169,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3178754162267522,
+      "grad_norm": 6.357814340494215,
+      "learning_rate": 3.990081709691556e-06,
+      "loss": 0.5666,
+      "step": 3914
+    },
+    {
+      "epoch": 0.31795663120279377,
+      "grad_norm": 3.0964122690765055,
+      "learning_rate": 3.989553613729359e-06,
+      "loss": 0.5291,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3180378461788354,
+      "grad_norm": 6.1633146191090615,
+      "learning_rate": 3.989025414697935e-06,
+      "loss": 0.4764,
+      "step": 3916
+    },
+    {
+      "epoch": 0.318119061154877,
+      "grad_norm": 5.380367903861603,
+      "learning_rate": 3.988497112633834e-06,
+      "loss": 0.4923,
+      "step": 3917
+    },
+    {
+      "epoch": 0.31820027613091856,
+      "grad_norm": 4.635335577517618,
+      "learning_rate": 3.98796870757361e-06,
+      "loss": 0.5211,
+      "step": 3918
+    },
+    {
+      "epoch": 0.31828149110696013,
+      "grad_norm": 4.837892560831567,
+      "learning_rate": 3.987440199553826e-06,
+      "loss": 0.5608,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3183627060830017,
+      "grad_norm": 5.412511747936819,
+      "learning_rate": 3.986911588611052e-06,
+      "loss": 0.5652,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3184439210590433,
+      "grad_norm": 6.754512301450377,
+      "learning_rate": 3.986382874781866e-06,
+      "loss": 0.6123,
+      "step": 3921
+    },
+    {
+      "epoch": 0.31852513603508487,
+      "grad_norm": 3.851560698751757,
+      "learning_rate": 3.985854058102851e-06,
+      "loss": 0.4627,
+      "step": 3922
+    },
+    {
+      "epoch": 0.31860635101112644,
+      "grad_norm": 9.180563910365933,
+      "learning_rate": 3.9853251386106e-06,
+      "loss": 0.5239,
+      "step": 3923
+    },
+    {
+      "epoch": 0.318687565987168,
+      "grad_norm": 6.398627354244881,
+      "learning_rate": 3.9847961163417094e-06,
+      "loss": 0.4989,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3187687809632096,
+      "grad_norm": 4.516597706332039,
+      "learning_rate": 3.984266991332787e-06,
+      "loss": 0.5573,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3188499959392512,
+      "grad_norm": 6.97958059945626,
+      "learning_rate": 3.9837377636204435e-06,
+      "loss": 0.3659,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3189312109152928,
+      "grad_norm": 9.380173505814613,
+      "learning_rate": 3.983208433241298e-06,
+      "loss": 0.5665,
+      "step": 3927
+    },
+    {
+      "epoch": 0.3190124258913344,
+      "grad_norm": 5.179839950219316,
+      "learning_rate": 3.98267900023198e-06,
+      "loss": 0.6098,
+      "step": 3928
+    },
+    {
+      "epoch": 0.31909364086737596,
+      "grad_norm": 11.243873982204496,
+      "learning_rate": 3.982149464629123e-06,
+      "loss": 0.5793,
+      "step": 3929
+    },
+    {
+      "epoch": 0.31917485584341754,
+      "grad_norm": 4.1291560709585164,
+      "learning_rate": 3.981619826469366e-06,
+      "loss": 0.5195,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3192560708194591,
+      "grad_norm": 4.373741253760804,
+      "learning_rate": 3.981090085789359e-06,
+      "loss": 0.5506,
+      "step": 3931
+    },
+    {
+      "epoch": 0.3193372857955007,
+      "grad_norm": 6.318431118742852,
+      "learning_rate": 3.980560242625756e-06,
+      "loss": 0.6606,
+      "step": 3932
+    },
+    {
+      "epoch": 0.31941850077154227,
+      "grad_norm": 4.240677812054353,
+      "learning_rate": 3.9800302970152205e-06,
+      "loss": 0.6272,
+      "step": 3933
+    },
+    {
+      "epoch": 0.31949971574758385,
+      "grad_norm": 5.672962140852729,
+      "learning_rate": 3.9795002489944216e-06,
+      "loss": 0.6,
+      "step": 3934
+    },
+    {
+      "epoch": 0.3195809307236254,
+      "grad_norm": 6.714289970142845,
+      "learning_rate": 3.978970098600035e-06,
+      "loss": 0.5626,
+      "step": 3935
+    },
+    {
+      "epoch": 0.319662145699667,
+      "grad_norm": 4.761339278563214,
+      "learning_rate": 3.978439845868745e-06,
+      "loss": 0.4723,
+      "step": 3936
+    },
+    {
+      "epoch": 0.3197433606757086,
+      "grad_norm": 13.677606899149971,
+      "learning_rate": 3.977909490837242e-06,
+      "loss": 0.5259,
+      "step": 3937
+    },
+    {
+      "epoch": 0.3198245756517502,
+      "grad_norm": 6.128118023978932,
+      "learning_rate": 3.977379033542225e-06,
+      "loss": 0.438,
+      "step": 3938
+    },
+    {
+      "epoch": 0.3199057906277918,
+      "grad_norm": 6.303253700795032,
+      "learning_rate": 3.976848474020397e-06,
+      "loss": 0.4292,
+      "step": 3939
+    },
+    {
+      "epoch": 0.31998700560383336,
+      "grad_norm": 10.396013099121385,
+      "learning_rate": 3.97631781230847e-06,
+      "loss": 0.635,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32006822057987494,
+      "grad_norm": 3.2724038031097575,
+      "learning_rate": 3.975787048443165e-06,
+      "loss": 0.7294,
+      "step": 3941
+    },
+    {
+      "epoch": 0.3201494355559165,
+      "grad_norm": 5.0292671115469645,
+      "learning_rate": 3.975256182461206e-06,
+      "loss": 0.5492,
+      "step": 3942
+    },
+    {
+      "epoch": 0.3202306505319581,
+      "grad_norm": 15.222409035535888,
+      "learning_rate": 3.9747252143993265e-06,
+      "loss": 0.5846,
+      "step": 3943
+    },
+    {
+      "epoch": 0.32031186550799967,
+      "grad_norm": 3.721523920711308,
+      "learning_rate": 3.9741941442942685e-06,
+      "loss": 0.7537,
+      "step": 3944
+    },
+    {
+      "epoch": 0.32039308048404125,
+      "grad_norm": 9.490205761589818,
+      "learning_rate": 3.973662972182777e-06,
+      "loss": 0.4823,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3204742954600828,
+      "grad_norm": 5.92729992222592,
+      "learning_rate": 3.973131698101606e-06,
+      "loss": 0.6342,
+      "step": 3946
+    },
+    {
+      "epoch": 0.3205555104361244,
+      "grad_norm": 3.889265293501103,
+      "learning_rate": 3.97260032208752e-06,
+      "loss": 0.6895,
+      "step": 3947
+    },
+    {
+      "epoch": 0.320636725412166,
+      "grad_norm": 5.718930345280851,
+      "learning_rate": 3.972068844177284e-06,
+      "loss": 0.5711,
+      "step": 3948
+    },
+    {
+      "epoch": 0.3207179403882076,
+      "grad_norm": 5.665338399507542,
+      "learning_rate": 3.971537264407674e-06,
+      "loss": 0.4434,
+      "step": 3949
+    },
+    {
+      "epoch": 0.3207991553642492,
+      "grad_norm": 5.195191497507738,
+      "learning_rate": 3.971005582815475e-06,
+      "loss": 0.5603,
+      "step": 3950
+    },
+    {
+      "epoch": 0.32088037034029077,
+      "grad_norm": 4.595813111552932,
+      "learning_rate": 3.970473799437475e-06,
+      "loss": 0.5229,
+      "step": 3951
+    },
+    {
+      "epoch": 0.32096158531633234,
+      "grad_norm": 7.011551829999244,
+      "learning_rate": 3.969941914310469e-06,
+      "loss": 0.5054,
+      "step": 3952
+    },
+    {
+      "epoch": 0.3210428002923739,
+      "grad_norm": 5.133064719413326,
+      "learning_rate": 3.969409927471263e-06,
+      "loss": 0.4704,
+      "step": 3953
+    },
+    {
+      "epoch": 0.3211240152684155,
+      "grad_norm": 4.606505759316552,
+      "learning_rate": 3.968877838956667e-06,
+      "loss": 0.525,
+      "step": 3954
+    },
+    {
+      "epoch": 0.3212052302444571,
+      "grad_norm": 5.298403703345239,
+      "learning_rate": 3.968345648803497e-06,
+      "loss": 0.4454,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32128644522049865,
+      "grad_norm": 3.561393835187397,
+      "learning_rate": 3.96781335704858e-06,
+      "loss": 0.5577,
+      "step": 3956
+    },
+    {
+      "epoch": 0.32136766019654023,
+      "grad_norm": 4.307683957878129,
+      "learning_rate": 3.967280963728748e-06,
+      "loss": 0.4494,
+      "step": 3957
+    },
+    {
+      "epoch": 0.3214488751725818,
+      "grad_norm": 6.747635355041429,
+      "learning_rate": 3.966748468880838e-06,
+      "loss": 0.6197,
+      "step": 3958
+    },
+    {
+      "epoch": 0.3215300901486234,
+      "grad_norm": 7.367076215445246,
+      "learning_rate": 3.9662158725416964e-06,
+      "loss": 0.5736,
+      "step": 3959
+    },
+    {
+      "epoch": 0.321611305124665,
+      "grad_norm": 4.986542323666267,
+      "learning_rate": 3.965683174748176e-06,
+      "loss": 0.6222,
+      "step": 3960
+    },
+    {
+      "epoch": 0.3216925201007066,
+      "grad_norm": 4.343391538341711,
+      "learning_rate": 3.965150375537137e-06,
+      "loss": 0.4032,
+      "step": 3961
+    },
+    {
+      "epoch": 0.32177373507674817,
+      "grad_norm": 10.317677958429377,
+      "learning_rate": 3.964617474945447e-06,
+      "loss": 0.5128,
+      "step": 3962
+    },
+    {
+      "epoch": 0.32185495005278975,
+      "grad_norm": 5.753574432660448,
+      "learning_rate": 3.9640844730099795e-06,
+      "loss": 0.5081,
+      "step": 3963
+    },
+    {
+      "epoch": 0.3219361650288313,
+      "grad_norm": 3.9575067307483045,
+      "learning_rate": 3.963551369767613e-06,
+      "loss": 0.5913,
+      "step": 3964
+    },
+    {
+      "epoch": 0.3220173800048729,
+      "grad_norm": 4.7333217124955365,
+      "learning_rate": 3.963018165255239e-06,
+      "loss": 0.5454,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3220985949809145,
+      "grad_norm": 4.107636292267094,
+      "learning_rate": 3.962484859509751e-06,
+      "loss": 0.4283,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32217980995695605,
+      "grad_norm": 3.710054258246495,
+      "learning_rate": 3.96195145256805e-06,
+      "loss": 0.5421,
+      "step": 3967
+    },
+    {
+      "epoch": 0.32226102493299763,
+      "grad_norm": 4.914093323605704,
+      "learning_rate": 3.961417944467046e-06,
+      "loss": 0.624,
+      "step": 3968
+    },
+    {
+      "epoch": 0.3223422399090392,
+      "grad_norm": 6.367334808510856,
+      "learning_rate": 3.960884335243655e-06,
+      "loss": 0.557,
+      "step": 3969
+    },
+    {
+      "epoch": 0.3224234548850808,
+      "grad_norm": 7.4083421441872455,
+      "learning_rate": 3.9603506249348e-06,
+      "loss": 0.7381,
+      "step": 3970
+    },
+    {
+      "epoch": 0.3225046698611224,
+      "grad_norm": 5.365590371657556,
+      "learning_rate": 3.959816813577409e-06,
+      "loss": 0.4419,
+      "step": 3971
+    },
+    {
+      "epoch": 0.322585884837164,
+      "grad_norm": 4.032690708340649,
+      "learning_rate": 3.959282901208422e-06,
+      "loss": 0.5859,
+      "step": 3972
+    },
+    {
+      "epoch": 0.32266709981320557,
+      "grad_norm": 5.285352909806383,
+      "learning_rate": 3.9587488878647816e-06,
+      "loss": 0.5464,
+      "step": 3973
+    },
+    {
+      "epoch": 0.32274831478924715,
+      "grad_norm": 6.70864234598179,
+      "learning_rate": 3.958214773583437e-06,
+      "loss": 0.5481,
+      "step": 3974
+    },
+    {
+      "epoch": 0.3228295297652887,
+      "grad_norm": 8.065418291262025,
+      "learning_rate": 3.957680558401348e-06,
+      "loss": 0.5129,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3229107447413303,
+      "grad_norm": 5.8858306697747755,
+      "learning_rate": 3.95714624235548e-06,
+      "loss": 0.5457,
+      "step": 3976
+    },
+    {
+      "epoch": 0.3229919597173719,
+      "grad_norm": 4.006111281727507,
+      "learning_rate": 3.956611825482803e-06,
+      "loss": 0.5355,
+      "step": 3977
+    },
+    {
+      "epoch": 0.32307317469341346,
+      "grad_norm": 5.150201779401238,
+      "learning_rate": 3.956077307820296e-06,
+      "loss": 0.4682,
+      "step": 3978
+    },
+    {
+      "epoch": 0.32315438966945503,
+      "grad_norm": 5.941131977925157,
+      "learning_rate": 3.955542689404948e-06,
+      "loss": 0.5036,
+      "step": 3979
+    },
+    {
+      "epoch": 0.3232356046454966,
+      "grad_norm": 5.8735315327446,
+      "learning_rate": 3.955007970273747e-06,
+      "loss": 0.7358,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3233168196215382,
+      "grad_norm": 6.004080782992699,
+      "learning_rate": 3.954473150463696e-06,
+      "loss": 0.4277,
+      "step": 3981
+    },
+    {
+      "epoch": 0.3233980345975798,
+      "grad_norm": 3.969562702014225,
+      "learning_rate": 3.9539382300117995e-06,
+      "loss": 0.6674,
+      "step": 3982
+    },
+    {
+      "epoch": 0.3234792495736214,
+      "grad_norm": 5.552544120013278,
+      "learning_rate": 3.953403208955074e-06,
+      "loss": 0.5466,
+      "step": 3983
+    },
+    {
+      "epoch": 0.323560464549663,
+      "grad_norm": 6.198037037595219,
+      "learning_rate": 3.952868087330537e-06,
+      "loss": 0.5557,
+      "step": 3984
+    },
+    {
+      "epoch": 0.32364167952570455,
+      "grad_norm": 3.7865591645101278,
+      "learning_rate": 3.952332865175218e-06,
+      "loss": 0.622,
+      "step": 3985
+    },
+    {
+      "epoch": 0.32372289450174613,
+      "grad_norm": 4.933123244556156,
+      "learning_rate": 3.951797542526151e-06,
+      "loss": 0.5681,
+      "step": 3986
+    },
+    {
+      "epoch": 0.3238041094777877,
+      "grad_norm": 5.397627483042807,
+      "learning_rate": 3.951262119420378e-06,
+      "loss": 0.4988,
+      "step": 3987
+    },
+    {
+      "epoch": 0.3238853244538293,
+      "grad_norm": 4.851399386478982,
+      "learning_rate": 3.950726595894947e-06,
+      "loss": 0.4111,
+      "step": 3988
+    },
+    {
+      "epoch": 0.32396653942987086,
+      "grad_norm": 3.1067739446054636,
+      "learning_rate": 3.950190971986913e-06,
+      "loss": 0.6122,
+      "step": 3989
+    },
+    {
+      "epoch": 0.32404775440591244,
+      "grad_norm": 5.367173504535532,
+      "learning_rate": 3.9496552477333396e-06,
+      "loss": 0.5503,
+      "step": 3990
+    },
+    {
+      "epoch": 0.324128969381954,
+      "grad_norm": 6.546774201268415,
+      "learning_rate": 3.9491194231712945e-06,
+      "loss": 0.4951,
+      "step": 3991
+    },
+    {
+      "epoch": 0.3242101843579956,
+      "grad_norm": 3.3008359196847876,
+      "learning_rate": 3.948583498337854e-06,
+      "loss": 0.4695,
+      "step": 3992
+    },
+    {
+      "epoch": 0.3242913993340372,
+      "grad_norm": 6.2106989027955315,
+      "learning_rate": 3.9480474732701034e-06,
+      "loss": 0.6426,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3243726143100788,
+      "grad_norm": 7.437097992706253,
+      "learning_rate": 3.9475113480051305e-06,
+      "loss": 0.5088,
+      "step": 3994
+    },
+    {
+      "epoch": 0.3244538292861204,
+      "grad_norm": 4.649897071407217,
+      "learning_rate": 3.9469751225800344e-06,
+      "loss": 0.3348,
+      "step": 3995
+    },
+    {
+      "epoch": 0.32453504426216195,
+      "grad_norm": 7.807412990752628,
+      "learning_rate": 3.946438797031916e-06,
+      "loss": 0.5809,
+      "step": 3996
+    },
+    {
+      "epoch": 0.32461625923820353,
+      "grad_norm": 10.891441494223873,
+      "learning_rate": 3.9459023713978895e-06,
+      "loss": 0.4846,
+      "step": 3997
+    },
+    {
+      "epoch": 0.3246974742142451,
+      "grad_norm": 8.904826186725101,
+      "learning_rate": 3.94536584571507e-06,
+      "loss": 0.6143,
+      "step": 3998
+    },
+    {
+      "epoch": 0.3247786891902867,
+      "grad_norm": 5.6954977350735785,
+      "learning_rate": 3.944829220020584e-06,
+      "loss": 0.5203,
+      "step": 3999
+    },
+    {
+      "epoch": 0.32485990416632826,
+      "grad_norm": 4.93125171789273,
+      "learning_rate": 3.944292494351563e-06,
+      "loss": 0.6325,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32494111914236984,
+      "grad_norm": 4.965862139362356,
+      "learning_rate": 3.943755668745145e-06,
+      "loss": 0.5805,
+      "step": 4001
+    },
+    {
+      "epoch": 0.3250223341184114,
+      "grad_norm": 5.16787159248045,
+      "learning_rate": 3.943218743238476e-06,
+      "loss": 0.5562,
+      "step": 4002
+    },
+    {
+      "epoch": 0.325103549094453,
+      "grad_norm": 7.5362717832466055,
+      "learning_rate": 3.942681717868707e-06,
+      "loss": 0.5688,
+      "step": 4003
+    },
+    {
+      "epoch": 0.3251847640704946,
+      "grad_norm": 4.981243847271942,
+      "learning_rate": 3.942144592673e-06,
+      "loss": 0.5032,
+      "step": 4004
+    },
+    {
+      "epoch": 0.3252659790465362,
+      "grad_norm": 5.040020101083127,
+      "learning_rate": 3.941607367688518e-06,
+      "loss": 0.6819,
+      "step": 4005
+    },
+    {
+      "epoch": 0.3253471940225778,
+      "grad_norm": 5.368844738609336,
+      "learning_rate": 3.941070042952437e-06,
+      "loss": 0.602,
+      "step": 4006
+    },
+    {
+      "epoch": 0.32542840899861936,
+      "grad_norm": 6.620173821128225,
+      "learning_rate": 3.940532618501935e-06,
+      "loss": 0.4266,
+      "step": 4007
+    },
+    {
+      "epoch": 0.32550962397466093,
+      "grad_norm": 3.763790781156802,
+      "learning_rate": 3.9399950943742e-06,
+      "loss": 0.6278,
+      "step": 4008
+    },
+    {
+      "epoch": 0.3255908389507025,
+      "grad_norm": 5.280317791344952,
+      "learning_rate": 3.939457470606426e-06,
+      "loss": 0.5235,
+      "step": 4009
+    },
+    {
+      "epoch": 0.3256720539267441,
+      "grad_norm": 6.207070016608509,
+      "learning_rate": 3.938919747235812e-06,
+      "loss": 0.498,
+      "step": 4010
+    },
+    {
+      "epoch": 0.32575326890278566,
+      "grad_norm": 3.9148661505895923,
+      "learning_rate": 3.938381924299568e-06,
+      "loss": 0.6105,
+      "step": 4011
+    },
+    {
+      "epoch": 0.32583448387882724,
+      "grad_norm": 11.982123261840117,
+      "learning_rate": 3.937844001834907e-06,
+      "loss": 0.4771,
+      "step": 4012
+    },
+    {
+      "epoch": 0.3259156988548688,
+      "grad_norm": 8.177076369022492,
+      "learning_rate": 3.93730597987905e-06,
+      "loss": 0.5762,
+      "step": 4013
+    },
+    {
+      "epoch": 0.3259969138309104,
+      "grad_norm": 4.802261707731817,
+      "learning_rate": 3.936767858469228e-06,
+      "loss": 0.4574,
+      "step": 4014
+    },
+    {
+      "epoch": 0.32607812880695203,
+      "grad_norm": 4.140558850278409,
+      "learning_rate": 3.936229637642672e-06,
+      "loss": 0.4855,
+      "step": 4015
+    },
+    {
+      "epoch": 0.3261593437829936,
+      "grad_norm": 3.5682197931038644,
+      "learning_rate": 3.935691317436628e-06,
+      "loss": 0.6508,
+      "step": 4016
+    },
+    {
+      "epoch": 0.3262405587590352,
+      "grad_norm": 4.124877825005949,
+      "learning_rate": 3.9351528978883425e-06,
+      "loss": 0.6438,
+      "step": 4017
+    },
+    {
+      "epoch": 0.32632177373507676,
+      "grad_norm": 5.830213074804944,
+      "learning_rate": 3.934614379035071e-06,
+      "loss": 0.5822,
+      "step": 4018
+    },
+    {
+      "epoch": 0.32640298871111834,
+      "grad_norm": 8.315171390345546,
+      "learning_rate": 3.9340757609140785e-06,
+      "loss": 0.5308,
+      "step": 4019
+    },
+    {
+      "epoch": 0.3264842036871599,
+      "grad_norm": 3.4244913994169313,
+      "learning_rate": 3.933537043562632e-06,
+      "loss": 0.4855,
+      "step": 4020
+    },
+    {
+      "epoch": 0.3265654186632015,
+      "grad_norm": 4.156835180938438,
+      "learning_rate": 3.932998227018009e-06,
+      "loss": 0.6688,
+      "step": 4021
+    },
+    {
+      "epoch": 0.32664663363924307,
+      "grad_norm": 3.4822522397229956,
+      "learning_rate": 3.932459311317494e-06,
+      "loss": 0.5877,
+      "step": 4022
+    },
+    {
+      "epoch": 0.32672784861528464,
+      "grad_norm": 4.707979946257839,
+      "learning_rate": 3.931920296498374e-06,
+      "loss": 0.5496,
+      "step": 4023
+    },
+    {
+      "epoch": 0.3268090635913262,
+      "grad_norm": 5.369501192774783,
+      "learning_rate": 3.931381182597949e-06,
+      "loss": 0.4053,
+      "step": 4024
+    },
+    {
+      "epoch": 0.3268902785673678,
+      "grad_norm": 7.0165291185775,
+      "learning_rate": 3.930841969653521e-06,
+      "loss": 0.5032,
+      "step": 4025
+    },
+    {
+      "epoch": 0.32697149354340943,
+      "grad_norm": 3.7285656291064844,
+      "learning_rate": 3.930302657702402e-06,
+      "loss": 0.5277,
+      "step": 4026
+    },
+    {
+      "epoch": 0.327052708519451,
+      "grad_norm": 4.479547869133663,
+      "learning_rate": 3.929763246781909e-06,
+      "loss": 0.5248,
+      "step": 4027
+    },
+    {
+      "epoch": 0.3271339234954926,
+      "grad_norm": 5.115409985178063,
+      "learning_rate": 3.929223736929366e-06,
+      "loss": 0.4248,
+      "step": 4028
+    },
+    {
+      "epoch": 0.32721513847153416,
+      "grad_norm": 6.753042464948224,
+      "learning_rate": 3.928684128182104e-06,
+      "loss": 0.5976,
+      "step": 4029
+    },
+    {
+      "epoch": 0.32729635344757574,
+      "grad_norm": 6.503506149392067,
+      "learning_rate": 3.9281444205774625e-06,
+      "loss": 0.5214,
+      "step": 4030
+    },
+    {
+      "epoch": 0.3273775684236173,
+      "grad_norm": 3.8958786789276787,
+      "learning_rate": 3.927604614152784e-06,
+      "loss": 0.6028,
+      "step": 4031
+    },
+    {
+      "epoch": 0.3274587833996589,
+      "grad_norm": 4.602684285068746,
+      "learning_rate": 3.927064708945423e-06,
+      "loss": 0.5836,
+      "step": 4032
+    },
+    {
+      "epoch": 0.32753999837570047,
+      "grad_norm": 5.4356792062574515,
+      "learning_rate": 3.926524704992736e-06,
+      "loss": 0.4976,
+      "step": 4033
+    },
+    {
+      "epoch": 0.32762121335174205,
+      "grad_norm": 5.1282337398644,
+      "learning_rate": 3.9259846023320895e-06,
+      "loss": 0.6328,
+      "step": 4034
+    },
+    {
+      "epoch": 0.3277024283277836,
+      "grad_norm": 5.574289088333828,
+      "learning_rate": 3.925444401000855e-06,
+      "loss": 0.5181,
+      "step": 4035
+    },
+    {
+      "epoch": 0.3277836433038252,
+      "grad_norm": 4.638010244640014,
+      "learning_rate": 3.924904101036413e-06,
+      "loss": 0.8373,
+      "step": 4036
+    },
+    {
+      "epoch": 0.32786485827986683,
+      "grad_norm": 3.1082588154104225,
+      "learning_rate": 3.924363702476147e-06,
+      "loss": 0.5468,
+      "step": 4037
+    },
+    {
+      "epoch": 0.3279460732559084,
+      "grad_norm": 3.8088490874389427,
+      "learning_rate": 3.923823205357453e-06,
+      "loss": 0.555,
+      "step": 4038
+    },
+    {
+      "epoch": 0.32802728823195,
+      "grad_norm": 3.64966498327372,
+      "learning_rate": 3.923282609717727e-06,
+      "loss": 0.4554,
+      "step": 4039
+    },
+    {
+      "epoch": 0.32810850320799156,
+      "grad_norm": 5.086174759269899,
+      "learning_rate": 3.922741915594378e-06,
+      "loss": 0.5311,
+      "step": 4040
+    },
+    {
+      "epoch": 0.32818971818403314,
+      "grad_norm": 3.596353017219979,
+      "learning_rate": 3.9222011230248175e-06,
+      "loss": 0.6889,
+      "step": 4041
+    },
+    {
+      "epoch": 0.3282709331600747,
+      "grad_norm": 4.391572461946081,
+      "learning_rate": 3.9216602320464655e-06,
+      "loss": 0.6185,
+      "step": 4042
+    },
+    {
+      "epoch": 0.3283521481361163,
+      "grad_norm": 3.615775625258503,
+      "learning_rate": 3.921119242696751e-06,
+      "loss": 0.5438,
+      "step": 4043
+    },
+    {
+      "epoch": 0.3284333631121579,
+      "grad_norm": 5.820794815401509,
+      "learning_rate": 3.920578155013106e-06,
+      "loss": 0.6468,
+      "step": 4044
+    },
+    {
+      "epoch": 0.32851457808819945,
+      "grad_norm": 5.467081465670503,
+      "learning_rate": 3.92003696903297e-06,
+      "loss": 0.4855,
+      "step": 4045
+    },
+    {
+      "epoch": 0.328595793064241,
+      "grad_norm": 5.561814863652521,
+      "learning_rate": 3.919495684793792e-06,
+      "loss": 0.6468,
+      "step": 4046
+    },
+    {
+      "epoch": 0.3286770080402826,
+      "grad_norm": 3.807058673477291,
+      "learning_rate": 3.918954302333025e-06,
+      "loss": 0.5918,
+      "step": 4047
+    },
+    {
+      "epoch": 0.32875822301632424,
+      "grad_norm": 13.31883734675245,
+      "learning_rate": 3.91841282168813e-06,
+      "loss": 0.6432,
+      "step": 4048
+    },
+    {
+      "epoch": 0.3288394379923658,
+      "grad_norm": 7.056921038692695,
+      "learning_rate": 3.917871242896575e-06,
+      "loss": 0.4209,
+      "step": 4049
+    },
+    {
+      "epoch": 0.3289206529684074,
+      "grad_norm": 3.9195936437606025,
+      "learning_rate": 3.917329565995833e-06,
+      "loss": 0.5187,
+      "step": 4050
+    },
+    {
+      "epoch": 0.32900186794444897,
+      "grad_norm": 4.550380167725966,
+      "learning_rate": 3.916787791023386e-06,
+      "loss": 0.5355,
+      "step": 4051
+    },
+    {
+      "epoch": 0.32908308292049054,
+      "grad_norm": 3.6800755148359148,
+      "learning_rate": 3.916245918016724e-06,
+      "loss": 0.5518,
+      "step": 4052
+    },
+    {
+      "epoch": 0.3291642978965321,
+      "grad_norm": 3.478748983074201,
+      "learning_rate": 3.915703947013338e-06,
+      "loss": 0.7117,
+      "step": 4053
+    },
+    {
+      "epoch": 0.3292455128725737,
+      "grad_norm": 3.7123489856745957,
+      "learning_rate": 3.9151618780507316e-06,
+      "loss": 0.5708,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3293267278486153,
+      "grad_norm": 5.643512507779105,
+      "learning_rate": 3.914619711166413e-06,
+      "loss": 0.5249,
+      "step": 4055
+    },
+    {
+      "epoch": 0.32940794282465685,
+      "grad_norm": 3.920365869559619,
+      "learning_rate": 3.914077446397897e-06,
+      "loss": 0.493,
+      "step": 4056
+    },
+    {
+      "epoch": 0.32948915780069843,
+      "grad_norm": 5.253760200255425,
+      "learning_rate": 3.913535083782707e-06,
+      "loss": 0.4857,
+      "step": 4057
+    },
+    {
+      "epoch": 0.32957037277674,
+      "grad_norm": 5.037318300624211,
+      "learning_rate": 3.912992623358368e-06,
+      "loss": 0.5694,
+      "step": 4058
+    },
+    {
+      "epoch": 0.32965158775278164,
+      "grad_norm": 16.068629096649826,
+      "learning_rate": 3.91245006516242e-06,
+      "loss": 0.407,
+      "step": 4059
+    },
+    {
+      "epoch": 0.3297328027288232,
+      "grad_norm": 8.438998004731252,
+      "learning_rate": 3.911907409232402e-06,
+      "loss": 0.5709,
+      "step": 4060
+    },
+    {
+      "epoch": 0.3298140177048648,
+      "grad_norm": 4.031413575021964,
+      "learning_rate": 3.911364655605863e-06,
+      "loss": 0.5698,
+      "step": 4061
+    },
+    {
+      "epoch": 0.32989523268090637,
+      "grad_norm": 6.129803113623354,
+      "learning_rate": 3.9108218043203595e-06,
+      "loss": 0.5103,
+      "step": 4062
+    },
+    {
+      "epoch": 0.32997644765694795,
+      "grad_norm": 5.08608452932562,
+      "learning_rate": 3.910278855413454e-06,
+      "loss": 0.6426,
+      "step": 4063
+    },
+    {
+      "epoch": 0.3300576626329895,
+      "grad_norm": 4.487322847861987,
+      "learning_rate": 3.909735808922716e-06,
+      "loss": 0.5326,
+      "step": 4064
+    },
+    {
+      "epoch": 0.3301388776090311,
+      "grad_norm": 5.951925940505768,
+      "learning_rate": 3.90919266488572e-06,
+      "loss": 0.481,
+      "step": 4065
+    },
+    {
+      "epoch": 0.3302200925850727,
+      "grad_norm": 5.479126876701789,
+      "learning_rate": 3.908649423340049e-06,
+      "loss": 0.5089,
+      "step": 4066
+    },
+    {
+      "epoch": 0.33030130756111425,
+      "grad_norm": 4.922261403979778,
+      "learning_rate": 3.908106084323295e-06,
+      "loss": 0.3857,
+      "step": 4067
+    },
+    {
+      "epoch": 0.33038252253715583,
+      "grad_norm": 5.574237760697828,
+      "learning_rate": 3.9075626478730515e-06,
+      "loss": 0.6416,
+      "step": 4068
+    },
+    {
+      "epoch": 0.3304637375131974,
+      "grad_norm": 3.790844587172487,
+      "learning_rate": 3.907019114026922e-06,
+      "loss": 0.5985,
+      "step": 4069
+    },
+    {
+      "epoch": 0.33054495248923904,
+      "grad_norm": 6.112434755772752,
+      "learning_rate": 3.906475482822517e-06,
+      "loss": 0.5749,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3306261674652806,
+      "grad_norm": 3.6083314382705836,
+      "learning_rate": 3.905931754297451e-06,
+      "loss": 0.5349,
+      "step": 4071
+    },
+    {
+      "epoch": 0.3307073824413222,
+      "grad_norm": 7.284570880294314,
+      "learning_rate": 3.905387928489349e-06,
+      "loss": 1.0363,
+      "step": 4072
+    },
+    {
+      "epoch": 0.33078859741736377,
+      "grad_norm": 4.232160296762864,
+      "learning_rate": 3.904844005435841e-06,
+      "loss": 0.5585,
+      "step": 4073
+    },
+    {
+      "epoch": 0.33086981239340535,
+      "grad_norm": 6.814310707158021,
+      "learning_rate": 3.904299985174562e-06,
+      "loss": 0.6862,
+      "step": 4074
+    },
+    {
+      "epoch": 0.3309510273694469,
+      "grad_norm": 4.171877579913468,
+      "learning_rate": 3.903755867743156e-06,
+      "loss": 0.5983,
+      "step": 4075
+    },
+    {
+      "epoch": 0.3310322423454885,
+      "grad_norm": 7.101609370383942,
+      "learning_rate": 3.9032116531792745e-06,
+      "loss": 0.5303,
+      "step": 4076
+    },
+    {
+      "epoch": 0.3311134573215301,
+      "grad_norm": 5.802670937749591,
+      "learning_rate": 3.902667341520572e-06,
+      "loss": 0.5949,
+      "step": 4077
+    },
+    {
+      "epoch": 0.33119467229757166,
+      "grad_norm": 4.414654939089757,
+      "learning_rate": 3.902122932804713e-06,
+      "loss": 0.5337,
+      "step": 4078
+    },
+    {
+      "epoch": 0.33127588727361323,
+      "grad_norm": 5.085650564698827,
+      "learning_rate": 3.901578427069368e-06,
+      "loss": 0.5897,
+      "step": 4079
+    },
+    {
+      "epoch": 0.3313571022496548,
+      "grad_norm": 3.7722391508616036,
+      "learning_rate": 3.901033824352213e-06,
+      "loss": 0.5486,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33143831722569644,
+      "grad_norm": 6.237526026700791,
+      "learning_rate": 3.9004891246909325e-06,
+      "loss": 0.5186,
+      "step": 4081
+    },
+    {
+      "epoch": 0.331519532201738,
+      "grad_norm": 4.2314785310516365,
+      "learning_rate": 3.8999443281232175e-06,
+      "loss": 0.727,
+      "step": 4082
+    },
+    {
+      "epoch": 0.3316007471777796,
+      "grad_norm": 4.575453746497402,
+      "learning_rate": 3.899399434686762e-06,
+      "loss": 0.4363,
+      "step": 4083
+    },
+    {
+      "epoch": 0.3316819621538212,
+      "grad_norm": 5.329498253898677,
+      "learning_rate": 3.898854444419274e-06,
+      "loss": 0.4231,
+      "step": 4084
+    },
+    {
+      "epoch": 0.33176317712986275,
+      "grad_norm": 7.175459151315008,
+      "learning_rate": 3.8983093573584605e-06,
+      "loss": 0.4597,
+      "step": 4085
+    },
+    {
+      "epoch": 0.33184439210590433,
+      "grad_norm": 4.3131878789803855,
+      "learning_rate": 3.89776417354204e-06,
+      "loss": 0.4702,
+      "step": 4086
+    },
+    {
+      "epoch": 0.3319256070819459,
+      "grad_norm": 4.921638293579504,
+      "learning_rate": 3.897218893007737e-06,
+      "loss": 0.5382,
+      "step": 4087
+    },
+    {
+      "epoch": 0.3320068220579875,
+      "grad_norm": 5.520197777335041,
+      "learning_rate": 3.896673515793281e-06,
+      "loss": 0.5162,
+      "step": 4088
+    },
+    {
+      "epoch": 0.33208803703402906,
+      "grad_norm": 4.2991891020242265,
+      "learning_rate": 3.89612804193641e-06,
+      "loss": 0.4928,
+      "step": 4089
+    },
+    {
+      "epoch": 0.33216925201007064,
+      "grad_norm": 5.160012794460195,
+      "learning_rate": 3.895582471474866e-06,
+      "loss": 0.5771,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3322504669861122,
+      "grad_norm": 4.051259036428905,
+      "learning_rate": 3.895036804446402e-06,
+      "loss": 0.4006,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33233168196215385,
+      "grad_norm": 8.065546594512844,
+      "learning_rate": 3.894491040888774e-06,
+      "loss": 0.7044,
+      "step": 4092
+    },
+    {
+      "epoch": 0.3324128969381954,
+      "grad_norm": 5.703512084951918,
+      "learning_rate": 3.893945180839747e-06,
+      "loss": 0.6109,
+      "step": 4093
+    },
+    {
+      "epoch": 0.332494111914237,
+      "grad_norm": 6.178693238141215,
+      "learning_rate": 3.893399224337089e-06,
+      "loss": 0.5347,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3325753268902786,
+      "grad_norm": 3.9476533225077293,
+      "learning_rate": 3.892853171418581e-06,
+      "loss": 0.7827,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33265654186632015,
+      "grad_norm": 6.5472391710424755,
+      "learning_rate": 3.8923070221220035e-06,
+      "loss": 0.5795,
+      "step": 4096
+    },
+    {
+      "epoch": 0.33273775684236173,
+      "grad_norm": 6.934804355699152,
+      "learning_rate": 3.891760776485151e-06,
+      "loss": 0.4096,
+      "step": 4097
+    },
+    {
+      "epoch": 0.3328189718184033,
+      "grad_norm": 4.704286066411406,
+      "learning_rate": 3.891214434545817e-06,
+      "loss": 0.5433,
+      "step": 4098
+    },
+    {
+      "epoch": 0.3329001867944449,
+      "grad_norm": 6.41984014774679,
+      "learning_rate": 3.890667996341806e-06,
+      "loss": 0.4911,
+      "step": 4099
+    },
+    {
+      "epoch": 0.33298140177048646,
+      "grad_norm": 5.021861875613623,
+      "learning_rate": 3.8901214619109315e-06,
+      "loss": 0.6617,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33306261674652804,
+      "grad_norm": 5.862846079718947,
+      "learning_rate": 3.889574831291008e-06,
+      "loss": 0.5158,
+      "step": 4101
+    },
+    {
+      "epoch": 0.3331438317225696,
+      "grad_norm": 3.942541898863967,
+      "learning_rate": 3.88902810451986e-06,
+      "loss": 0.6377,
+      "step": 4102
+    },
+    {
+      "epoch": 0.33322504669861125,
+      "grad_norm": 4.915257424154396,
+      "learning_rate": 3.88848128163532e-06,
+      "loss": 0.516,
+      "step": 4103
+    },
+    {
+      "epoch": 0.3333062616746528,
+      "grad_norm": 4.848991134824205,
+      "learning_rate": 3.887934362675223e-06,
+      "loss": 0.427,
+      "step": 4104
+    },
+    {
+      "epoch": 0.3333874766506944,
+      "grad_norm": 5.880446166851823,
+      "learning_rate": 3.887387347677413e-06,
+      "loss": 0.4525,
+      "step": 4105
+    },
+    {
+      "epoch": 0.333468691626736,
+      "grad_norm": 7.1524682868349405,
+      "learning_rate": 3.886840236679742e-06,
+      "loss": 0.4618,
+      "step": 4106
+    },
+    {
+      "epoch": 0.33354990660277756,
+      "grad_norm": 4.096340585575771,
+      "learning_rate": 3.8862930297200665e-06,
+      "loss": 0.4553,
+      "step": 4107
+    },
+    {
+      "epoch": 0.33363112157881913,
+      "grad_norm": 3.963306450481558,
+      "learning_rate": 3.885745726836249e-06,
+      "loss": 0.5802,
+      "step": 4108
+    },
+    {
+      "epoch": 0.3337123365548607,
+      "grad_norm": 7.460142742604042,
+      "learning_rate": 3.885198328066163e-06,
+      "loss": 0.4519,
+      "step": 4109
+    },
+    {
+      "epoch": 0.3337935515309023,
+      "grad_norm": 4.260196594904515,
+      "learning_rate": 3.8846508334476824e-06,
+      "loss": 0.5484,
+      "step": 4110
+    },
+    {
+      "epoch": 0.33387476650694387,
+      "grad_norm": 3.7783084323283074,
+      "learning_rate": 3.884103243018693e-06,
+      "loss": 0.5141,
+      "step": 4111
+    },
+    {
+      "epoch": 0.33395598148298544,
+      "grad_norm": 5.894477787381174,
+      "learning_rate": 3.883555556817083e-06,
+      "loss": 0.6465,
+      "step": 4112
+    },
+    {
+      "epoch": 0.334037196459027,
+      "grad_norm": 3.0063514235175917,
+      "learning_rate": 3.883007774880753e-06,
+      "loss": 0.4949,
+      "step": 4113
+    },
+    {
+      "epoch": 0.33411841143506865,
+      "grad_norm": 5.033096732224493,
+      "learning_rate": 3.882459897247603e-06,
+      "loss": 0.3705,
+      "step": 4114
+    },
+    {
+      "epoch": 0.33419962641111023,
+      "grad_norm": 5.991095754571292,
+      "learning_rate": 3.881911923955545e-06,
+      "loss": 0.4903,
+      "step": 4115
+    },
+    {
+      "epoch": 0.3342808413871518,
+      "grad_norm": 6.237218731747614,
+      "learning_rate": 3.881363855042496e-06,
+      "loss": 0.5439,
+      "step": 4116
+    },
+    {
+      "epoch": 0.3343620563631934,
+      "grad_norm": 4.522440392622378,
+      "learning_rate": 3.880815690546378e-06,
+      "loss": 0.4756,
+      "step": 4117
+    },
+    {
+      "epoch": 0.33444327133923496,
+      "grad_norm": 12.342772678402822,
+      "learning_rate": 3.880267430505123e-06,
+      "loss": 0.5839,
+      "step": 4118
+    },
+    {
+      "epoch": 0.33452448631527654,
+      "grad_norm": 4.148599127985545,
+      "learning_rate": 3.879719074956667e-06,
+      "loss": 0.5127,
+      "step": 4119
+    },
+    {
+      "epoch": 0.3346057012913181,
+      "grad_norm": 4.58831638518393,
+      "learning_rate": 3.879170623938951e-06,
+      "loss": 0.6514,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3346869162673597,
+      "grad_norm": 4.795703097401847,
+      "learning_rate": 3.878622077489929e-06,
+      "loss": 0.5412,
+      "step": 4121
+    },
+    {
+      "epoch": 0.33476813124340127,
+      "grad_norm": 2.8973056633483325,
+      "learning_rate": 3.8780734356475555e-06,
+      "loss": 0.5741,
+      "step": 4122
+    },
+    {
+      "epoch": 0.33484934621944284,
+      "grad_norm": 3.5866605917438474,
+      "learning_rate": 3.8775246984497924e-06,
+      "loss": 0.5285,
+      "step": 4123
+    },
+    {
+      "epoch": 0.3349305611954844,
+      "grad_norm": 4.6085635915185,
+      "learning_rate": 3.876975865934612e-06,
+      "loss": 0.6005,
+      "step": 4124
+    },
+    {
+      "epoch": 0.33501177617152605,
+      "grad_norm": 8.852381490643886,
+      "learning_rate": 3.876426938139988e-06,
+      "loss": 0.3847,
+      "step": 4125
+    },
+    {
+      "epoch": 0.33509299114756763,
+      "grad_norm": 4.677167179306538,
+      "learning_rate": 3.875877915103905e-06,
+      "loss": 0.5997,
+      "step": 4126
+    },
+    {
+      "epoch": 0.3351742061236092,
+      "grad_norm": 5.192252334603382,
+      "learning_rate": 3.875328796864353e-06,
+      "loss": 0.4351,
+      "step": 4127
+    },
+    {
+      "epoch": 0.3352554210996508,
+      "grad_norm": 5.52994136456474,
+      "learning_rate": 3.8747795834593255e-06,
+      "loss": 0.5296,
+      "step": 4128
+    },
+    {
+      "epoch": 0.33533663607569236,
+      "grad_norm": 9.716260977214207,
+      "learning_rate": 3.8742302749268264e-06,
+      "loss": 0.5424,
+      "step": 4129
+    },
+    {
+      "epoch": 0.33541785105173394,
+      "grad_norm": 5.297918507808388,
+      "learning_rate": 3.873680871304867e-06,
+      "loss": 0.6156,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3354990660277755,
+      "grad_norm": 6.873079350204181,
+      "learning_rate": 3.8731313726314615e-06,
+      "loss": 0.5344,
+      "step": 4131
+    },
+    {
+      "epoch": 0.3355802810038171,
+      "grad_norm": 7.36681510636934,
+      "learning_rate": 3.87258177894463e-06,
+      "loss": 0.4929,
+      "step": 4132
+    },
+    {
+      "epoch": 0.33566149597985867,
+      "grad_norm": 5.602356345828153,
+      "learning_rate": 3.872032090282406e-06,
+      "loss": 0.5442,
+      "step": 4133
+    },
+    {
+      "epoch": 0.33574271095590025,
+      "grad_norm": 7.3672095285990435,
+      "learning_rate": 3.871482306682821e-06,
+      "loss": 0.492,
+      "step": 4134
+    },
+    {
+      "epoch": 0.3358239259319418,
+      "grad_norm": 3.5366167997788427,
+      "learning_rate": 3.8709324281839205e-06,
+      "loss": 0.5198,
+      "step": 4135
+    },
+    {
+      "epoch": 0.33590514090798346,
+      "grad_norm": 6.858040921971297,
+      "learning_rate": 3.87038245482375e-06,
+      "loss": 0.5951,
+      "step": 4136
+    },
+    {
+      "epoch": 0.33598635588402503,
+      "grad_norm": 6.617980225631782,
+      "learning_rate": 3.869832386640367e-06,
+      "loss": 0.5853,
+      "step": 4137
+    },
+    {
+      "epoch": 0.3360675708600666,
+      "grad_norm": 4.219229609437151,
+      "learning_rate": 3.8692822236718334e-06,
+      "loss": 0.5662,
+      "step": 4138
+    },
+    {
+      "epoch": 0.3361487858361082,
+      "grad_norm": 7.439696385730913,
+      "learning_rate": 3.868731965956215e-06,
+      "loss": 0.4658,
+      "step": 4139
+    },
+    {
+      "epoch": 0.33623000081214977,
+      "grad_norm": 5.487826086727248,
+      "learning_rate": 3.86818161353159e-06,
+      "loss": 0.4537,
+      "step": 4140
+    },
+    {
+      "epoch": 0.33631121578819134,
+      "grad_norm": 4.200360102230179,
+      "learning_rate": 3.867631166436038e-06,
+      "loss": 0.5663,
+      "step": 4141
+    },
+    {
+      "epoch": 0.3363924307642329,
+      "grad_norm": 6.128604530592701,
+      "learning_rate": 3.867080624707647e-06,
+      "loss": 0.6134,
+      "step": 4142
+    },
+    {
+      "epoch": 0.3364736457402745,
+      "grad_norm": 8.81104997626726,
+      "learning_rate": 3.866529988384512e-06,
+      "loss": 0.5773,
+      "step": 4143
+    },
+    {
+      "epoch": 0.3365548607163161,
+      "grad_norm": 5.16122326363201,
+      "learning_rate": 3.865979257504734e-06,
+      "loss": 0.6132,
+      "step": 4144
+    },
+    {
+      "epoch": 0.33663607569235765,
+      "grad_norm": 5.870897322328888,
+      "learning_rate": 3.8654284321064205e-06,
+      "loss": 0.6016,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3367172906683992,
+      "grad_norm": 5.314926965529579,
+      "learning_rate": 3.864877512227686e-06,
+      "loss": 0.5678,
+      "step": 4146
+    },
+    {
+      "epoch": 0.33679850564444086,
+      "grad_norm": 4.818645386450623,
+      "learning_rate": 3.864326497906652e-06,
+      "loss": 0.5558,
+      "step": 4147
+    },
+    {
+      "epoch": 0.33687972062048244,
+      "grad_norm": 14.41001497081986,
+      "learning_rate": 3.8637753891814435e-06,
+      "loss": 0.5669,
+      "step": 4148
+    },
+    {
+      "epoch": 0.336960935596524,
+      "grad_norm": 10.608448522114792,
+      "learning_rate": 3.863224186090197e-06,
+      "loss": 0.5596,
+      "step": 4149
+    },
+    {
+      "epoch": 0.3370421505725656,
+      "grad_norm": 6.7059941990937775,
+      "learning_rate": 3.862672888671051e-06,
+      "loss": 0.5213,
+      "step": 4150
+    },
+    {
+      "epoch": 0.33712336554860717,
+      "grad_norm": 3.7804743609104237,
+      "learning_rate": 3.862121496962153e-06,
+      "loss": 0.7185,
+      "step": 4151
+    },
+    {
+      "epoch": 0.33720458052464874,
+      "grad_norm": 5.24213546954736,
+      "learning_rate": 3.861570011001658e-06,
+      "loss": 0.5231,
+      "step": 4152
+    },
+    {
+      "epoch": 0.3372857955006903,
+      "grad_norm": 7.547577006376037,
+      "learning_rate": 3.8610184308277216e-06,
+      "loss": 0.4785,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3373670104767319,
+      "grad_norm": 2.5300512242174387,
+      "learning_rate": 3.860466756478514e-06,
+      "loss": 0.6946,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3374482254527735,
+      "grad_norm": 4.653767821488039,
+      "learning_rate": 3.859914987992207e-06,
+      "loss": 0.7683,
+      "step": 4155
+    },
+    {
+      "epoch": 0.33752944042881505,
+      "grad_norm": 4.930191892040406,
+      "learning_rate": 3.85936312540698e-06,
+      "loss": 0.5301,
+      "step": 4156
+    },
+    {
+      "epoch": 0.33761065540485663,
+      "grad_norm": 5.429582397608964,
+      "learning_rate": 3.858811168761019e-06,
+      "loss": 0.4413,
+      "step": 4157
+    },
+    {
+      "epoch": 0.33769187038089826,
+      "grad_norm": 4.863739044933661,
+      "learning_rate": 3.8582591180925164e-06,
+      "loss": 0.4454,
+      "step": 4158
+    },
+    {
+      "epoch": 0.33777308535693984,
+      "grad_norm": 5.522590385270159,
+      "learning_rate": 3.857706973439672e-06,
+      "loss": 0.507,
+      "step": 4159
+    },
+    {
+      "epoch": 0.3378543003329814,
+      "grad_norm": 5.8411296020442,
+      "learning_rate": 3.85715473484069e-06,
+      "loss": 0.5373,
+      "step": 4160
+    },
+    {
+      "epoch": 0.337935515309023,
+      "grad_norm": 6.215435578421796,
+      "learning_rate": 3.856602402333783e-06,
+      "loss": 0.5099,
+      "step": 4161
+    },
+    {
+      "epoch": 0.33801673028506457,
+      "grad_norm": 3.6082614721689215,
+      "learning_rate": 3.85604997595717e-06,
+      "loss": 0.5364,
+      "step": 4162
+    },
+    {
+      "epoch": 0.33809794526110615,
+      "grad_norm": 6.122674136049005,
+      "learning_rate": 3.855497455749076e-06,
+      "loss": 0.5525,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3381791602371477,
+      "grad_norm": 7.470369967210209,
+      "learning_rate": 3.854944841747731e-06,
+      "loss": 0.4537,
+      "step": 4164
+    },
+    {
+      "epoch": 0.3382603752131893,
+      "grad_norm": 3.13970191041356,
+      "learning_rate": 3.854392133991373e-06,
+      "loss": 0.4119,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3383415901892309,
+      "grad_norm": 5.817426875860977,
+      "learning_rate": 3.853839332518249e-06,
+      "loss": 0.6385,
+      "step": 4166
+    },
+    {
+      "epoch": 0.33842280516527246,
+      "grad_norm": 6.211091908912435,
+      "learning_rate": 3.8532864373666076e-06,
+      "loss": 0.5794,
+      "step": 4167
+    },
+    {
+      "epoch": 0.33850402014131403,
+      "grad_norm": 5.831264510593337,
+      "learning_rate": 3.852733448574707e-06,
+      "loss": 0.53,
+      "step": 4168
+    },
+    {
+      "epoch": 0.33858523511735567,
+      "grad_norm": 6.324531527655222,
+      "learning_rate": 3.8521803661808105e-06,
+      "loss": 0.4364,
+      "step": 4169
+    },
+    {
+      "epoch": 0.33866645009339724,
+      "grad_norm": 4.057887664623083,
+      "learning_rate": 3.851627190223189e-06,
+      "loss": 0.6626,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3387476650694388,
+      "grad_norm": 6.855870724748997,
+      "learning_rate": 3.85107392074012e-06,
+      "loss": 0.5495,
+      "step": 4171
+    },
+    {
+      "epoch": 0.3388288800454804,
+      "grad_norm": 4.761555193030634,
+      "learning_rate": 3.850520557769886e-06,
+      "loss": 0.5127,
+      "step": 4172
+    },
+    {
+      "epoch": 0.338910095021522,
+      "grad_norm": 4.705148709464969,
+      "learning_rate": 3.849967101350777e-06,
+      "loss": 0.4324,
+      "step": 4173
+    },
+    {
+      "epoch": 0.33899130999756355,
+      "grad_norm": 6.188404278264953,
+      "learning_rate": 3.849413551521089e-06,
+      "loss": 0.4936,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3390725249736051,
+      "grad_norm": 4.011266087707078,
+      "learning_rate": 3.848859908319124e-06,
+      "loss": 0.4387,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3391537399496467,
+      "grad_norm": 3.339221489587599,
+      "learning_rate": 3.8483061717831935e-06,
+      "loss": 0.6163,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3392349549256883,
+      "grad_norm": 5.751859436326058,
+      "learning_rate": 3.8477523419516115e-06,
+      "loss": 0.5683,
+      "step": 4177
+    },
+    {
+      "epoch": 0.33931616990172986,
+      "grad_norm": 5.088627831398807,
+      "learning_rate": 3.8471984188627e-06,
+      "loss": 0.5431,
+      "step": 4178
+    },
+    {
+      "epoch": 0.33939738487777144,
+      "grad_norm": 5.37045736248948,
+      "learning_rate": 3.846644402554788e-06,
+      "loss": 0.7555,
+      "step": 4179
+    },
+    {
+      "epoch": 0.33947859985381307,
+      "grad_norm": 5.257725496690353,
+      "learning_rate": 3.84609029306621e-06,
+      "loss": 0.4728,
+      "step": 4180
+    },
+    {
+      "epoch": 0.33955981482985464,
+      "grad_norm": 4.575264421254666,
+      "learning_rate": 3.845536090435308e-06,
+      "loss": 0.5399,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3396410298058962,
+      "grad_norm": 8.153018797116616,
+      "learning_rate": 3.84498179470043e-06,
+      "loss": 0.6235,
+      "step": 4182
+    },
+    {
+      "epoch": 0.3397222447819378,
+      "grad_norm": 6.682613939663991,
+      "learning_rate": 3.8444274058999295e-06,
+      "loss": 0.4972,
+      "step": 4183
+    },
+    {
+      "epoch": 0.3398034597579794,
+      "grad_norm": 4.184180026129862,
+      "learning_rate": 3.843872924072168e-06,
+      "loss": 0.4771,
+      "step": 4184
+    },
+    {
+      "epoch": 0.33988467473402095,
+      "grad_norm": 5.132706408054481,
+      "learning_rate": 3.843318349255512e-06,
+      "loss": 0.4656,
+      "step": 4185
+    },
+    {
+      "epoch": 0.33996588971006253,
+      "grad_norm": 4.503889789040779,
+      "learning_rate": 3.842763681488337e-06,
+      "loss": 0.5627,
+      "step": 4186
+    },
+    {
+      "epoch": 0.3400471046861041,
+      "grad_norm": 8.418497213540109,
+      "learning_rate": 3.84220892080902e-06,
+      "loss": 0.4399,
+      "step": 4187
+    },
+    {
+      "epoch": 0.3401283196621457,
+      "grad_norm": 5.704270836674129,
+      "learning_rate": 3.841654067255951e-06,
+      "loss": 0.4365,
+      "step": 4188
+    },
+    {
+      "epoch": 0.34020953463818726,
+      "grad_norm": 3.7971463715319813,
+      "learning_rate": 3.84109912086752e-06,
+      "loss": 0.6448,
+      "step": 4189
+    },
+    {
+      "epoch": 0.34029074961422884,
+      "grad_norm": 5.144040812396076,
+      "learning_rate": 3.840544081682128e-06,
+      "loss": 0.7299,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34037196459027047,
+      "grad_norm": 4.053682645967981,
+      "learning_rate": 3.839988949738179e-06,
+      "loss": 0.4847,
+      "step": 4191
+    },
+    {
+      "epoch": 0.34045317956631205,
+      "grad_norm": 4.2685472291584245,
+      "learning_rate": 3.8394337250740886e-06,
+      "loss": 0.4542,
+      "step": 4192
+    },
+    {
+      "epoch": 0.3405343945423536,
+      "grad_norm": 5.165758993343546,
+      "learning_rate": 3.838878407728272e-06,
+      "loss": 0.5573,
+      "step": 4193
+    },
+    {
+      "epoch": 0.3406156095183952,
+      "grad_norm": 4.470607814383757,
+      "learning_rate": 3.838322997739155e-06,
+      "loss": 0.6386,
+      "step": 4194
+    },
+    {
+      "epoch": 0.3406968244944368,
+      "grad_norm": 6.678206627035974,
+      "learning_rate": 3.837767495145171e-06,
+      "loss": 0.6893,
+      "step": 4195
+    },
+    {
+      "epoch": 0.34077803947047836,
+      "grad_norm": 3.54047842198611,
+      "learning_rate": 3.837211899984756e-06,
+      "loss": 0.6608,
+      "step": 4196
+    },
+    {
+      "epoch": 0.34085925444651993,
+      "grad_norm": 2.4752964852325023,
+      "learning_rate": 3.836656212296353e-06,
+      "loss": 0.752,
+      "step": 4197
+    },
+    {
+      "epoch": 0.3409404694225615,
+      "grad_norm": 5.131230574367217,
+      "learning_rate": 3.836100432118416e-06,
+      "loss": 0.5224,
+      "step": 4198
+    },
+    {
+      "epoch": 0.3410216843986031,
+      "grad_norm": 7.654106670047192,
+      "learning_rate": 3.8355445594894e-06,
+      "loss": 0.5236,
+      "step": 4199
+    },
+    {
+      "epoch": 0.34110289937464466,
+      "grad_norm": 5.475141337183228,
+      "learning_rate": 3.834988594447768e-06,
+      "loss": 0.3241,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34118411435068624,
+      "grad_norm": 4.9511971467274964,
+      "learning_rate": 3.8344325370319914e-06,
+      "loss": 0.5689,
+      "step": 4201
+    },
+    {
+      "epoch": 0.3412653293267279,
+      "grad_norm": 4.8632486482468,
+      "learning_rate": 3.833876387280546e-06,
+      "loss": 0.6184,
+      "step": 4202
+    },
+    {
+      "epoch": 0.34134654430276945,
+      "grad_norm": 8.274806032942742,
+      "learning_rate": 3.833320145231913e-06,
+      "loss": 0.6777,
+      "step": 4203
+    },
+    {
+      "epoch": 0.341427759278811,
+      "grad_norm": 3.411012348367635,
+      "learning_rate": 3.832763810924583e-06,
+      "loss": 0.5455,
+      "step": 4204
+    },
+    {
+      "epoch": 0.3415089742548526,
+      "grad_norm": 4.294223710510589,
+      "learning_rate": 3.832207384397051e-06,
+      "loss": 0.5742,
+      "step": 4205
+    },
+    {
+      "epoch": 0.3415901892308942,
+      "grad_norm": 7.056992590176889,
+      "learning_rate": 3.831650865687818e-06,
+      "loss": 0.8163,
+      "step": 4206
+    },
+    {
+      "epoch": 0.34167140420693576,
+      "grad_norm": 4.513215530320878,
+      "learning_rate": 3.831094254835393e-06,
+      "loss": 0.4881,
+      "step": 4207
+    },
+    {
+      "epoch": 0.34175261918297734,
+      "grad_norm": 7.046767735752562,
+      "learning_rate": 3.8305375518782905e-06,
+      "loss": 0.5084,
+      "step": 4208
+    },
+    {
+      "epoch": 0.3418338341590189,
+      "grad_norm": 4.65613704507043,
+      "learning_rate": 3.829980756855032e-06,
+      "loss": 0.4564,
+      "step": 4209
+    },
+    {
+      "epoch": 0.3419150491350605,
+      "grad_norm": 5.501379878695886,
+      "learning_rate": 3.829423869804143e-06,
+      "loss": 0.5426,
+      "step": 4210
+    },
+    {
+      "epoch": 0.34199626411110207,
+      "grad_norm": 7.787760682762794,
+      "learning_rate": 3.828866890764157e-06,
+      "loss": 0.4953,
+      "step": 4211
+    },
+    {
+      "epoch": 0.34207747908714364,
+      "grad_norm": 4.761244297351993,
+      "learning_rate": 3.828309819773617e-06,
+      "loss": 0.7606,
+      "step": 4212
+    },
+    {
+      "epoch": 0.3421586940631853,
+      "grad_norm": 6.816779994137413,
+      "learning_rate": 3.827752656871067e-06,
+      "loss": 0.4022,
+      "step": 4213
+    },
+    {
+      "epoch": 0.34223990903922685,
+      "grad_norm": 5.030056000377152,
+      "learning_rate": 3.827195402095059e-06,
+      "loss": 0.5904,
+      "step": 4214
+    },
+    {
+      "epoch": 0.34232112401526843,
+      "grad_norm": 5.852140493007803,
+      "learning_rate": 3.826638055484154e-06,
+      "loss": 0.423,
+      "step": 4215
+    },
+    {
+      "epoch": 0.34240233899131,
+      "grad_norm": 4.727427200553123,
+      "learning_rate": 3.826080617076917e-06,
+      "loss": 0.6244,
+      "step": 4216
+    },
+    {
+      "epoch": 0.3424835539673516,
+      "grad_norm": 8.080683087721372,
+      "learning_rate": 3.825523086911919e-06,
+      "loss": 0.4194,
+      "step": 4217
+    },
+    {
+      "epoch": 0.34256476894339316,
+      "grad_norm": 3.4664753816902834,
+      "learning_rate": 3.824965465027739e-06,
+      "loss": 0.5493,
+      "step": 4218
+    },
+    {
+      "epoch": 0.34264598391943474,
+      "grad_norm": 8.5527804211556,
+      "learning_rate": 3.824407751462962e-06,
+      "loss": 0.5962,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3427271988954763,
+      "grad_norm": 9.91272395777568,
+      "learning_rate": 3.823849946256176e-06,
+      "loss": 0.5413,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3428084138715179,
+      "grad_norm": 5.49826382214047,
+      "learning_rate": 3.82329204944598e-06,
+      "loss": 0.4414,
+      "step": 4221
+    },
+    {
+      "epoch": 0.34288962884755947,
+      "grad_norm": 5.466358398271932,
+      "learning_rate": 3.822734061070979e-06,
+      "loss": 0.653,
+      "step": 4222
+    },
+    {
+      "epoch": 0.34297084382360105,
+      "grad_norm": 11.56485219391147,
+      "learning_rate": 3.8221759811697814e-06,
+      "loss": 0.6647,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3430520587996427,
+      "grad_norm": 17.302847263409163,
+      "learning_rate": 3.821617809781004e-06,
+      "loss": 0.5001,
+      "step": 4224
+    },
+    {
+      "epoch": 0.34313327377568426,
+      "grad_norm": 4.9393254820446,
+      "learning_rate": 3.821059546943268e-06,
+      "loss": 0.4429,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34321448875172583,
+      "grad_norm": 4.737433746889213,
+      "learning_rate": 3.820501192695202e-06,
+      "loss": 0.6441,
+      "step": 4226
+    },
+    {
+      "epoch": 0.3432957037277674,
+      "grad_norm": 5.321040703740516,
+      "learning_rate": 3.819942747075443e-06,
+      "loss": 0.4669,
+      "step": 4227
+    },
+    {
+      "epoch": 0.343376918703809,
+      "grad_norm": 5.28796113251718,
+      "learning_rate": 3.819384210122631e-06,
+      "loss": 0.682,
+      "step": 4228
+    },
+    {
+      "epoch": 0.34345813367985056,
+      "grad_norm": 4.339078198914844,
+      "learning_rate": 3.818825581875415e-06,
+      "loss": 0.5883,
+      "step": 4229
+    },
+    {
+      "epoch": 0.34353934865589214,
+      "grad_norm": 4.128567458915762,
+      "learning_rate": 3.818266862372449e-06,
+      "loss": 0.5184,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3436205636319337,
+      "grad_norm": 3.9206753012437146,
+      "learning_rate": 3.817708051652392e-06,
+      "loss": 0.6334,
+      "step": 4231
+    },
+    {
+      "epoch": 0.3437017786079753,
+      "grad_norm": 4.303153750803405,
+      "learning_rate": 3.817149149753912e-06,
+      "loss": 0.5891,
+      "step": 4232
+    },
+    {
+      "epoch": 0.34378299358401687,
+      "grad_norm": 4.474909831406013,
+      "learning_rate": 3.816590156715682e-06,
+      "loss": 0.5449,
+      "step": 4233
+    },
+    {
+      "epoch": 0.34386420856005845,
+      "grad_norm": 4.504556577904222,
+      "learning_rate": 3.81603107257638e-06,
+      "loss": 0.4061,
+      "step": 4234
+    },
+    {
+      "epoch": 0.3439454235361001,
+      "grad_norm": 2.939069265586592,
+      "learning_rate": 3.815471897374695e-06,
+      "loss": 0.5448,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34402663851214166,
+      "grad_norm": 4.384216019391756,
+      "learning_rate": 3.814912631149315e-06,
+      "loss": 0.4506,
+      "step": 4236
+    },
+    {
+      "epoch": 0.34410785348818324,
+      "grad_norm": 5.19610868271772,
+      "learning_rate": 3.8143532739389403e-06,
+      "loss": 0.5923,
+      "step": 4237
+    },
+    {
+      "epoch": 0.3441890684642248,
+      "grad_norm": 4.2187938416631185,
+      "learning_rate": 3.813793825782276e-06,
+      "loss": 0.6411,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3442702834402664,
+      "grad_norm": 5.412344609770045,
+      "learning_rate": 3.8132342867180318e-06,
+      "loss": 0.6972,
+      "step": 4239
+    },
+    {
+      "epoch": 0.34435149841630797,
+      "grad_norm": 3.8719949393667887,
+      "learning_rate": 3.812674656784924e-06,
+      "loss": 0.6241,
+      "step": 4240
+    },
+    {
+      "epoch": 0.34443271339234954,
+      "grad_norm": 6.272686879652281,
+      "learning_rate": 3.812114936021678e-06,
+      "loss": 0.4416,
+      "step": 4241
+    },
+    {
+      "epoch": 0.3445139283683911,
+      "grad_norm": 4.522533410945012,
+      "learning_rate": 3.811555124467023e-06,
+      "loss": 0.6361,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3445951433444327,
+      "grad_norm": 3.7359506733942554,
+      "learning_rate": 3.8109952221596948e-06,
+      "loss": 0.5521,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3446763583204743,
+      "grad_norm": 5.599008701288348,
+      "learning_rate": 3.810435229138435e-06,
+      "loss": 0.6105,
+      "step": 4244
+    },
+    {
+      "epoch": 0.34475757329651585,
+      "grad_norm": 4.77660925406526,
+      "learning_rate": 3.8098751454419925e-06,
+      "loss": 0.6491,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3448387882725575,
+      "grad_norm": 5.070740770359748,
+      "learning_rate": 3.8093149711091227e-06,
+      "loss": 0.4824,
+      "step": 4246
+    },
+    {
+      "epoch": 0.34492000324859906,
+      "grad_norm": 5.29618525895374,
+      "learning_rate": 3.8087547061785864e-06,
+      "loss": 0.591,
+      "step": 4247
+    },
+    {
+      "epoch": 0.34500121822464064,
+      "grad_norm": 4.011238439114341,
+      "learning_rate": 3.8081943506891505e-06,
+      "loss": 0.6667,
+      "step": 4248
+    },
+    {
+      "epoch": 0.3450824332006822,
+      "grad_norm": 4.640224694054099,
+      "learning_rate": 3.8076339046795897e-06,
+      "loss": 0.5231,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3451636481767238,
+      "grad_norm": 9.930057551543705,
+      "learning_rate": 3.807073368188683e-06,
+      "loss": 0.7279,
+      "step": 4250
+    },
+    {
+      "epoch": 0.34524486315276537,
+      "grad_norm": 10.280051980432813,
+      "learning_rate": 3.8065127412552172e-06,
+      "loss": 0.7463,
+      "step": 4251
+    },
+    {
+      "epoch": 0.34532607812880695,
+      "grad_norm": 4.260596124125055,
+      "learning_rate": 3.8059520239179836e-06,
+      "loss": 0.7628,
+      "step": 4252
+    },
+    {
+      "epoch": 0.3454072931048485,
+      "grad_norm": 6.511349904040633,
+      "learning_rate": 3.805391216215782e-06,
+      "loss": 0.4457,
+      "step": 4253
+    },
+    {
+      "epoch": 0.3454885080808901,
+      "grad_norm": 7.176157624874702,
+      "learning_rate": 3.8048303181874167e-06,
+      "loss": 0.4474,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3455697230569317,
+      "grad_norm": 5.9798967177829345,
+      "learning_rate": 3.8042693298717e-06,
+      "loss": 0.7248,
+      "step": 4255
+    },
+    {
+      "epoch": 0.34565093803297325,
+      "grad_norm": 5.3870467284981,
+      "learning_rate": 3.8037082513074468e-06,
+      "loss": 0.4971,
+      "step": 4256
+    },
+    {
+      "epoch": 0.3457321530090149,
+      "grad_norm": 6.0361440421296,
+      "learning_rate": 3.8031470825334838e-06,
+      "loss": 0.5185,
+      "step": 4257
+    },
+    {
+      "epoch": 0.34581336798505646,
+      "grad_norm": 3.7390306125476243,
+      "learning_rate": 3.8025858235886394e-06,
+      "loss": 0.7618,
+      "step": 4258
+    },
+    {
+      "epoch": 0.34589458296109804,
+      "grad_norm": 6.423197360909206,
+      "learning_rate": 3.802024474511749e-06,
+      "loss": 0.5033,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3459757979371396,
+      "grad_norm": 4.523265015480748,
+      "learning_rate": 3.801463035341656e-06,
+      "loss": 0.5696,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3460570129131812,
+      "grad_norm": 6.030313585209124,
+      "learning_rate": 3.8009015061172095e-06,
+      "loss": 0.5074,
+      "step": 4261
+    },
+    {
+      "epoch": 0.34613822788922277,
+      "grad_norm": 9.4302885564273,
+      "learning_rate": 3.8003398868772635e-06,
+      "loss": 0.4557,
+      "step": 4262
+    },
+    {
+      "epoch": 0.34621944286526435,
+      "grad_norm": 7.919711884223888,
+      "learning_rate": 3.799778177660679e-06,
+      "loss": 0.5377,
+      "step": 4263
+    },
+    {
+      "epoch": 0.3463006578413059,
+      "grad_norm": 4.176073302897445,
+      "learning_rate": 3.7992163785063236e-06,
+      "loss": 0.5704,
+      "step": 4264
+    },
+    {
+      "epoch": 0.3463818728173475,
+      "grad_norm": 5.09213708548835,
+      "learning_rate": 3.798654489453071e-06,
+      "loss": 0.554,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3464630877933891,
+      "grad_norm": 6.0436953163637295,
+      "learning_rate": 3.7980925105398004e-06,
+      "loss": 0.526,
+      "step": 4266
+    },
+    {
+      "epoch": 0.34654430276943066,
+      "grad_norm": 6.156030657781099,
+      "learning_rate": 3.7975304418053986e-06,
+      "loss": 0.4978,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3466255177454723,
+      "grad_norm": 8.589190270382112,
+      "learning_rate": 3.796968283288758e-06,
+      "loss": 0.6419,
+      "step": 4268
+    },
+    {
+      "epoch": 0.34670673272151387,
+      "grad_norm": 3.828448877259962,
+      "learning_rate": 3.7964060350287747e-06,
+      "loss": 0.5272,
+      "step": 4269
+    },
+    {
+      "epoch": 0.34678794769755544,
+      "grad_norm": 4.335800260568463,
+      "learning_rate": 3.795843697064355e-06,
+      "loss": 0.5072,
+      "step": 4270
+    },
+    {
+      "epoch": 0.346869162673597,
+      "grad_norm": 4.112493113260704,
+      "learning_rate": 3.795281269434411e-06,
+      "loss": 0.5839,
+      "step": 4271
+    },
+    {
+      "epoch": 0.3469503776496386,
+      "grad_norm": 7.2088319680381225,
+      "learning_rate": 3.794718752177857e-06,
+      "loss": 0.7074,
+      "step": 4272
+    },
+    {
+      "epoch": 0.3470315926256802,
+      "grad_norm": 3.766153162278056,
+      "learning_rate": 3.7941561453336184e-06,
+      "loss": 0.5309,
+      "step": 4273
+    },
+    {
+      "epoch": 0.34711280760172175,
+      "grad_norm": 5.845126706777332,
+      "learning_rate": 3.7935934489406232e-06,
+      "loss": 0.3967,
+      "step": 4274
+    },
+    {
+      "epoch": 0.34719402257776333,
+      "grad_norm": 6.9520337995936785,
+      "learning_rate": 3.7930306630378085e-06,
+      "loss": 0.7975,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3472752375538049,
+      "grad_norm": 4.890242996197673,
+      "learning_rate": 3.7924677876641147e-06,
+      "loss": 0.4203,
+      "step": 4276
+    },
+    {
+      "epoch": 0.3473564525298465,
+      "grad_norm": 3.8913871864084255,
+      "learning_rate": 3.79190482285849e-06,
+      "loss": 0.7376,
+      "step": 4277
+    },
+    {
+      "epoch": 0.34743766750588806,
+      "grad_norm": 3.544505934601897,
+      "learning_rate": 3.7913417686598886e-06,
+      "loss": 0.6519,
+      "step": 4278
+    },
+    {
+      "epoch": 0.3475188824819297,
+      "grad_norm": 9.771874636578227,
+      "learning_rate": 3.790778625107272e-06,
+      "loss": 0.5073,
+      "step": 4279
+    },
+    {
+      "epoch": 0.34760009745797127,
+      "grad_norm": 3.505229136041032,
+      "learning_rate": 3.790215392239606e-06,
+      "loss": 0.5759,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34768131243401285,
+      "grad_norm": 5.243230166877009,
+      "learning_rate": 3.7896520700958616e-06,
+      "loss": 0.5125,
+      "step": 4281
+    },
+    {
+      "epoch": 0.3477625274100544,
+      "grad_norm": 5.947817581890729,
+      "learning_rate": 3.789088658715021e-06,
+      "loss": 0.5668,
+      "step": 4282
+    },
+    {
+      "epoch": 0.347843742386096,
+      "grad_norm": 4.83590576883994,
+      "learning_rate": 3.788525158136067e-06,
+      "loss": 0.5129,
+      "step": 4283
+    },
+    {
+      "epoch": 0.3479249573621376,
+      "grad_norm": 4.089901831736869,
+      "learning_rate": 3.787961568397992e-06,
+      "loss": 0.45,
+      "step": 4284
+    },
+    {
+      "epoch": 0.34800617233817915,
+      "grad_norm": 4.292363577189182,
+      "learning_rate": 3.787397889539792e-06,
+      "loss": 0.699,
+      "step": 4285
+    },
+    {
+      "epoch": 0.34808738731422073,
+      "grad_norm": 3.731168656658335,
+      "learning_rate": 3.786834121600472e-06,
+      "loss": 0.4771,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3481686022902623,
+      "grad_norm": 6.663689172947077,
+      "learning_rate": 3.7862702646190415e-06,
+      "loss": 0.6777,
+      "step": 4287
+    },
+    {
+      "epoch": 0.3482498172663039,
+      "grad_norm": 7.125149701358761,
+      "learning_rate": 3.7857063186345156e-06,
+      "loss": 0.62,
+      "step": 4288
+    },
+    {
+      "epoch": 0.34833103224234546,
+      "grad_norm": 3.677264775633149,
+      "learning_rate": 3.7851422836859177e-06,
+      "loss": 0.541,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3484122472183871,
+      "grad_norm": 7.094843934887101,
+      "learning_rate": 3.7845781598122743e-06,
+      "loss": 0.4561,
+      "step": 4290
+    },
+    {
+      "epoch": 0.34849346219442867,
+      "grad_norm": 4.886642342030592,
+      "learning_rate": 3.7840139470526215e-06,
+      "loss": 0.4937,
+      "step": 4291
+    },
+    {
+      "epoch": 0.34857467717047025,
+      "grad_norm": 5.62838070355612,
+      "learning_rate": 3.783449645445999e-06,
+      "loss": 0.513,
+      "step": 4292
+    },
+    {
+      "epoch": 0.3486558921465118,
+      "grad_norm": 4.6812249317048025,
+      "learning_rate": 3.782885255031453e-06,
+      "loss": 0.5147,
+      "step": 4293
+    },
+    {
+      "epoch": 0.3487371071225534,
+      "grad_norm": 6.938556921862774,
+      "learning_rate": 3.782320775848038e-06,
+      "loss": 0.3674,
+      "step": 4294
+    },
+    {
+      "epoch": 0.348818322098595,
+      "grad_norm": 6.723176214680427,
+      "learning_rate": 3.7817562079348114e-06,
+      "loss": 0.4626,
+      "step": 4295
+    },
+    {
+      "epoch": 0.34889953707463656,
+      "grad_norm": 5.553775061400944,
+      "learning_rate": 3.7811915513308382e-06,
+      "loss": 0.4768,
+      "step": 4296
+    },
+    {
+      "epoch": 0.34898075205067813,
+      "grad_norm": 5.884939622386394,
+      "learning_rate": 3.7806268060751914e-06,
+      "loss": 0.4368,
+      "step": 4297
+    },
+    {
+      "epoch": 0.3490619670267197,
+      "grad_norm": 6.5667330861720234,
+      "learning_rate": 3.7800619722069464e-06,
+      "loss": 0.444,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3491431820027613,
+      "grad_norm": 6.393075158086913,
+      "learning_rate": 3.7794970497651877e-06,
+      "loss": 0.4569,
+      "step": 4299
+    },
+    {
+      "epoch": 0.34922439697880286,
+      "grad_norm": 7.173171177630671,
+      "learning_rate": 3.7789320387890056e-06,
+      "loss": 0.6791,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3493056119548445,
+      "grad_norm": 4.165736108586291,
+      "learning_rate": 3.778366939317494e-06,
+      "loss": 0.5986,
+      "step": 4301
+    },
+    {
+      "epoch": 0.3493868269308861,
+      "grad_norm": 3.349799899482817,
+      "learning_rate": 3.777801751389757e-06,
+      "loss": 0.5417,
+      "step": 4302
+    },
+    {
+      "epoch": 0.34946804190692765,
+      "grad_norm": 5.017370500577692,
+      "learning_rate": 3.7772364750449002e-06,
+      "loss": 0.7214,
+      "step": 4303
+    },
+    {
+      "epoch": 0.34954925688296923,
+      "grad_norm": 6.384615890109909,
+      "learning_rate": 3.77667111032204e-06,
+      "loss": 0.4845,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3496304718590108,
+      "grad_norm": 5.139067733395603,
+      "learning_rate": 3.776105657260295e-06,
+      "loss": 0.5029,
+      "step": 4305
+    },
+    {
+      "epoch": 0.3497116868350524,
+      "grad_norm": 30.243450603041254,
+      "learning_rate": 3.7755401158987926e-06,
+      "loss": 0.6012,
+      "step": 4306
+    },
+    {
+      "epoch": 0.34979290181109396,
+      "grad_norm": 8.484420160765751,
+      "learning_rate": 3.774974486276664e-06,
+      "loss": 0.4776,
+      "step": 4307
+    },
+    {
+      "epoch": 0.34987411678713554,
+      "grad_norm": 6.935408912143071,
+      "learning_rate": 3.77440876843305e-06,
+      "loss": 0.4172,
+      "step": 4308
+    },
+    {
+      "epoch": 0.3499553317631771,
+      "grad_norm": 4.024333397155922,
+      "learning_rate": 3.773842962407093e-06,
+      "loss": 0.7109,
+      "step": 4309
+    },
+    {
+      "epoch": 0.3500365467392187,
+      "grad_norm": 4.11100179002336,
+      "learning_rate": 3.773277068237945e-06,
+      "loss": 0.5926,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35011776171526027,
+      "grad_norm": 4.947318220787382,
+      "learning_rate": 3.7727110859647627e-06,
+      "loss": 0.6069,
+      "step": 4311
+    },
+    {
+      "epoch": 0.3501989766913019,
+      "grad_norm": 3.6871411628254696,
+      "learning_rate": 3.772145015626709e-06,
+      "loss": 0.5277,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3502801916673435,
+      "grad_norm": 10.305081134224809,
+      "learning_rate": 3.771578857262953e-06,
+      "loss": 0.4547,
+      "step": 4313
+    },
+    {
+      "epoch": 0.35036140664338505,
+      "grad_norm": 4.429330825416072,
+      "learning_rate": 3.771012610912669e-06,
+      "loss": 0.6503,
+      "step": 4314
+    },
+    {
+      "epoch": 0.35044262161942663,
+      "grad_norm": 8.456969721413778,
+      "learning_rate": 3.7704462766150396e-06,
+      "loss": 0.5715,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3505238365954682,
+      "grad_norm": 5.543349480716538,
+      "learning_rate": 3.7698798544092525e-06,
+      "loss": 0.3989,
+      "step": 4316
+    },
+    {
+      "epoch": 0.3506050515715098,
+      "grad_norm": 4.0226607503884315,
+      "learning_rate": 3.7693133443344986e-06,
+      "loss": 0.7712,
+      "step": 4317
+    },
+    {
+      "epoch": 0.35068626654755136,
+      "grad_norm": 3.3414221408894798,
+      "learning_rate": 3.7687467464299797e-06,
+      "loss": 0.5677,
+      "step": 4318
+    },
+    {
+      "epoch": 0.35076748152359294,
+      "grad_norm": 3.3587069368493263,
+      "learning_rate": 3.7681800607349017e-06,
+      "loss": 0.5779,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3508486964996345,
+      "grad_norm": 7.359687891817828,
+      "learning_rate": 3.767613287288474e-06,
+      "loss": 0.5286,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3509299114756761,
+      "grad_norm": 4.5580897772433095,
+      "learning_rate": 3.767046426129917e-06,
+      "loss": 0.6514,
+      "step": 4321
+    },
+    {
+      "epoch": 0.35101112645171767,
+      "grad_norm": 6.868236388509833,
+      "learning_rate": 3.7664794772984515e-06,
+      "loss": 0.4803,
+      "step": 4322
+    },
+    {
+      "epoch": 0.3510923414277593,
+      "grad_norm": 4.0864068693665985,
+      "learning_rate": 3.7659124408333094e-06,
+      "loss": 0.4627,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3511735564038009,
+      "grad_norm": 4.220792384857057,
+      "learning_rate": 3.7653453167737263e-06,
+      "loss": 0.6832,
+      "step": 4324
+    },
+    {
+      "epoch": 0.35125477137984246,
+      "grad_norm": 6.391965671992134,
+      "learning_rate": 3.7647781051589436e-06,
+      "loss": 0.7653,
+      "step": 4325
+    },
+    {
+      "epoch": 0.35133598635588403,
+      "grad_norm": 6.3932460863379905,
+      "learning_rate": 3.76421080602821e-06,
+      "loss": 0.4652,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3514172013319256,
+      "grad_norm": 5.223581522055953,
+      "learning_rate": 3.76364341942078e-06,
+      "loss": 0.6826,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3514984163079672,
+      "grad_norm": 6.233483510000385,
+      "learning_rate": 3.7630759453759123e-06,
+      "loss": 0.4624,
+      "step": 4328
+    },
+    {
+      "epoch": 0.35157963128400876,
+      "grad_norm": 4.989931764108342,
+      "learning_rate": 3.7625083839328747e-06,
+      "loss": 0.4736,
+      "step": 4329
+    },
+    {
+      "epoch": 0.35166084626005034,
+      "grad_norm": 3.2608228815714755,
+      "learning_rate": 3.7619407351309377e-06,
+      "loss": 0.4673,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3517420612360919,
+      "grad_norm": 4.925326222217201,
+      "learning_rate": 3.761372999009381e-06,
+      "loss": 0.5807,
+      "step": 4331
+    },
+    {
+      "epoch": 0.3518232762121335,
+      "grad_norm": 3.6058579203829844,
+      "learning_rate": 3.7608051756074894e-06,
+      "loss": 0.4816,
+      "step": 4332
+    },
+    {
+      "epoch": 0.3519044911881751,
+      "grad_norm": 5.412076104516127,
+      "learning_rate": 3.7602372649645512e-06,
+      "loss": 0.6296,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3519857061642167,
+      "grad_norm": 4.984953620322977,
+      "learning_rate": 3.759669267119864e-06,
+      "loss": 0.5238,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3520669211402583,
+      "grad_norm": 3.348233179207022,
+      "learning_rate": 3.759101182112731e-06,
+      "loss": 0.6843,
+      "step": 4335
+    },
+    {
+      "epoch": 0.35214813611629986,
+      "grad_norm": 3.9680570346308204,
+      "learning_rate": 3.758533009982459e-06,
+      "loss": 0.4943,
+      "step": 4336
+    },
+    {
+      "epoch": 0.35222935109234144,
+      "grad_norm": 7.538460902230132,
+      "learning_rate": 3.7579647507683636e-06,
+      "loss": 0.4964,
+      "step": 4337
+    },
+    {
+      "epoch": 0.352310566068383,
+      "grad_norm": 4.840273083850319,
+      "learning_rate": 3.7573964045097655e-06,
+      "loss": 0.6186,
+      "step": 4338
+    },
+    {
+      "epoch": 0.3523917810444246,
+      "grad_norm": 4.652079301299005,
+      "learning_rate": 3.7568279712459908e-06,
+      "loss": 0.4687,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35247299602046617,
+      "grad_norm": 4.542830944823784,
+      "learning_rate": 3.7562594510163718e-06,
+      "loss": 0.6544,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35255421099650774,
+      "grad_norm": 5.562323687848427,
+      "learning_rate": 3.755690843860248e-06,
+      "loss": 0.598,
+      "step": 4341
+    },
+    {
+      "epoch": 0.3526354259725493,
+      "grad_norm": 3.056102850388366,
+      "learning_rate": 3.7551221498169633e-06,
+      "loss": 0.46,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3527166409485909,
+      "grad_norm": 6.370494566318182,
+      "learning_rate": 3.7545533689258683e-06,
+      "loss": 0.5024,
+      "step": 4343
+    },
+    {
+      "epoch": 0.3527978559246325,
+      "grad_norm": 3.674196156661001,
+      "learning_rate": 3.75398450122632e-06,
+      "loss": 0.4952,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3528790709006741,
+      "grad_norm": 5.618292271167161,
+      "learning_rate": 3.7534155467576805e-06,
+      "loss": 0.545,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3529602858767157,
+      "grad_norm": 7.155057050102085,
+      "learning_rate": 3.7528465055593186e-06,
+      "loss": 0.7136,
+      "step": 4346
+    },
+    {
+      "epoch": 0.35304150085275726,
+      "grad_norm": 4.835019495981251,
+      "learning_rate": 3.75227737767061e-06,
+      "loss": 0.5152,
+      "step": 4347
+    },
+    {
+      "epoch": 0.35312271582879884,
+      "grad_norm": 6.80042426573597,
+      "learning_rate": 3.7517081631309336e-06,
+      "loss": 0.5799,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3532039308048404,
+      "grad_norm": 7.706095450683844,
+      "learning_rate": 3.751138861979678e-06,
+      "loss": 0.5641,
+      "step": 4349
+    },
+    {
+      "epoch": 0.353285145780882,
+      "grad_norm": 8.667289733551144,
+      "learning_rate": 3.750569474256233e-06,
+      "loss": 0.5249,
+      "step": 4350
+    },
+    {
+      "epoch": 0.35336636075692357,
+      "grad_norm": 4.311680279675531,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.4826,
+      "step": 4351
+    },
+    {
+      "epoch": 0.35344757573296515,
+      "grad_norm": 5.148268090014955,
+      "learning_rate": 3.7494304392503826e-06,
+      "loss": 0.5173,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3535287907090067,
+      "grad_norm": 5.2133873799522705,
+      "learning_rate": 3.7488607920467912e-06,
+      "loss": 0.7347,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3536100056850483,
+      "grad_norm": 5.251842412363714,
+      "learning_rate": 3.7482910584286424e-06,
+      "loss": 0.4798,
+      "step": 4354
+    },
+    {
+      "epoch": 0.3536912206610899,
+      "grad_norm": 4.4703093132843,
+      "learning_rate": 3.747721238435359e-06,
+      "loss": 0.6538,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3537724356371315,
+      "grad_norm": 4.814603055251707,
+      "learning_rate": 3.747151332106369e-06,
+      "loss": 0.487,
+      "step": 4356
+    },
+    {
+      "epoch": 0.3538536506131731,
+      "grad_norm": 5.542109813139826,
+      "learning_rate": 3.746581339481108e-06,
+      "loss": 0.5932,
+      "step": 4357
+    },
+    {
+      "epoch": 0.35393486558921466,
+      "grad_norm": 5.541416801300144,
+      "learning_rate": 3.746011260599015e-06,
+      "loss": 0.605,
+      "step": 4358
+    },
+    {
+      "epoch": 0.35401608056525624,
+      "grad_norm": 4.197242421152716,
+      "learning_rate": 3.7454410954995375e-06,
+      "loss": 0.4671,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3540972955412978,
+      "grad_norm": 6.220911708551648,
+      "learning_rate": 3.7448708442221277e-06,
+      "loss": 0.5889,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3541785105173394,
+      "grad_norm": 4.456428745711773,
+      "learning_rate": 3.744300506806243e-06,
+      "loss": 0.5731,
+      "step": 4361
+    },
+    {
+      "epoch": 0.354259725493381,
+      "grad_norm": 4.103784001652695,
+      "learning_rate": 3.7437300832913503e-06,
+      "loss": 0.4561,
+      "step": 4362
+    },
+    {
+      "epoch": 0.35434094046942255,
+      "grad_norm": 16.445100012740834,
+      "learning_rate": 3.743159573716917e-06,
+      "loss": 0.6277,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3544221554454641,
+      "grad_norm": 4.616837308526195,
+      "learning_rate": 3.7425889781224204e-06,
+      "loss": 0.5266,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3545033704215057,
+      "grad_norm": 8.670573299062688,
+      "learning_rate": 3.742018296547344e-06,
+      "loss": 0.5105,
+      "step": 4365
+    },
+    {
+      "epoch": 0.3545845853975473,
+      "grad_norm": 6.449318189960221,
+      "learning_rate": 3.741447529031173e-06,
+      "loss": 0.7134,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3546658003735889,
+      "grad_norm": 4.211623063096266,
+      "learning_rate": 3.7408766756134046e-06,
+      "loss": 0.5255,
+      "step": 4367
+    },
+    {
+      "epoch": 0.3547470153496305,
+      "grad_norm": 4.907991529043284,
+      "learning_rate": 3.740305736333537e-06,
+      "loss": 0.6893,
+      "step": 4368
+    },
+    {
+      "epoch": 0.35482823032567207,
+      "grad_norm": 4.845747717269982,
+      "learning_rate": 3.7397347112310767e-06,
+      "loss": 0.6383,
+      "step": 4369
+    },
+    {
+      "epoch": 0.35490944530171364,
+      "grad_norm": 4.148877926596722,
+      "learning_rate": 3.7391636003455355e-06,
+      "loss": 0.4795,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3549906602777552,
+      "grad_norm": 5.568518592335486,
+      "learning_rate": 3.7385924037164316e-06,
+      "loss": 0.5019,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3550718752537968,
+      "grad_norm": 11.380910354185877,
+      "learning_rate": 3.7380211213832882e-06,
+      "loss": 0.4622,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3551530902298384,
+      "grad_norm": 4.898875896209035,
+      "learning_rate": 3.737449753385636e-06,
+      "loss": 0.5913,
+      "step": 4373
+    },
+    {
+      "epoch": 0.35523430520587995,
+      "grad_norm": 3.7018730566328433,
+      "learning_rate": 3.7368782997630093e-06,
+      "loss": 0.7925,
+      "step": 4374
+    },
+    {
+      "epoch": 0.35531552018192153,
+      "grad_norm": 3.8066706472807565,
+      "learning_rate": 3.7363067605549515e-06,
+      "loss": 0.5056,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3553967351579631,
+      "grad_norm": 2.709504723389777,
+      "learning_rate": 3.7357351358010075e-06,
+      "loss": 0.5972,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3554779501340047,
+      "grad_norm": 4.824489245992463,
+      "learning_rate": 3.735163425540732e-06,
+      "loss": 0.6907,
+      "step": 4377
+    },
+    {
+      "epoch": 0.3555591651100463,
+      "grad_norm": 7.943391610774761,
+      "learning_rate": 3.734591629813686e-06,
+      "loss": 0.5539,
+      "step": 4378
+    },
+    {
+      "epoch": 0.3556403800860879,
+      "grad_norm": 4.381215594871206,
+      "learning_rate": 3.7340197486594315e-06,
+      "loss": 0.5431,
+      "step": 4379
+    },
+    {
+      "epoch": 0.35572159506212947,
+      "grad_norm": 4.946575395839248,
+      "learning_rate": 3.7334477821175424e-06,
+      "loss": 0.6081,
+      "step": 4380
+    },
+    {
+      "epoch": 0.35580281003817105,
+      "grad_norm": 4.52740218848989,
+      "learning_rate": 3.732875730227595e-06,
+      "loss": 0.4524,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3558840250142126,
+      "grad_norm": 6.069550060126085,
+      "learning_rate": 3.7323035930291706e-06,
+      "loss": 0.5258,
+      "step": 4382
+    },
+    {
+      "epoch": 0.3559652399902542,
+      "grad_norm": 5.631666918507212,
+      "learning_rate": 3.731731370561861e-06,
+      "loss": 0.8438,
+      "step": 4383
+    },
+    {
+      "epoch": 0.3560464549662958,
+      "grad_norm": 8.095060386123581,
+      "learning_rate": 3.7311590628652584e-06,
+      "loss": 0.6436,
+      "step": 4384
+    },
+    {
+      "epoch": 0.35612766994233735,
+      "grad_norm": 5.278732076516886,
+      "learning_rate": 3.730586669978965e-06,
+      "loss": 0.5972,
+      "step": 4385
+    },
+    {
+      "epoch": 0.35620888491837893,
+      "grad_norm": 6.566634009856689,
+      "learning_rate": 3.7300141919425865e-06,
+      "loss": 0.4352,
+      "step": 4386
+    },
+    {
+      "epoch": 0.3562900998944205,
+      "grad_norm": 4.301691821909096,
+      "learning_rate": 3.729441628795736e-06,
+      "loss": 0.4451,
+      "step": 4387
+    },
+    {
+      "epoch": 0.3563713148704621,
+      "grad_norm": 3.538764460520504,
+      "learning_rate": 3.728868980578031e-06,
+      "loss": 0.4408,
+      "step": 4388
+    },
+    {
+      "epoch": 0.3564525298465037,
+      "grad_norm": 8.009167633892602,
+      "learning_rate": 3.7282962473290964e-06,
+      "loss": 0.6824,
+      "step": 4389
+    },
+    {
+      "epoch": 0.3565337448225453,
+      "grad_norm": 3.6954739187392414,
+      "learning_rate": 3.727723429088562e-06,
+      "loss": 0.6465,
+      "step": 4390
+    },
+    {
+      "epoch": 0.3566149597985869,
+      "grad_norm": 5.835592705497657,
+      "learning_rate": 3.7271505258960644e-06,
+      "loss": 0.3879,
+      "step": 4391
+    },
+    {
+      "epoch": 0.35669617477462845,
+      "grad_norm": 4.621090146943028,
+      "learning_rate": 3.726577537791245e-06,
+      "loss": 0.6456,
+      "step": 4392
+    },
+    {
+      "epoch": 0.35677738975067,
+      "grad_norm": 7.345583344323817,
+      "learning_rate": 3.726004464813752e-06,
+      "loss": 0.5299,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3568586047267116,
+      "grad_norm": 5.887510819023688,
+      "learning_rate": 3.725431307003238e-06,
+      "loss": 0.4683,
+      "step": 4394
+    },
+    {
+      "epoch": 0.3569398197027532,
+      "grad_norm": 4.4086817353864785,
+      "learning_rate": 3.7248580643993625e-06,
+      "loss": 0.5324,
+      "step": 4395
+    },
+    {
+      "epoch": 0.35702103467879476,
+      "grad_norm": 5.875362001143908,
+      "learning_rate": 3.724284737041792e-06,
+      "loss": 0.5235,
+      "step": 4396
+    },
+    {
+      "epoch": 0.35710224965483633,
+      "grad_norm": 4.84990196629755,
+      "learning_rate": 3.723711324970197e-06,
+      "loss": 0.4776,
+      "step": 4397
+    },
+    {
+      "epoch": 0.3571834646308779,
+      "grad_norm": 4.446080183771416,
+      "learning_rate": 3.723137828224255e-06,
+      "loss": 0.5653,
+      "step": 4398
+    },
+    {
+      "epoch": 0.3572646796069195,
+      "grad_norm": 11.963154592877663,
+      "learning_rate": 3.722564246843648e-06,
+      "loss": 0.6452,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3573458945829611,
+      "grad_norm": 5.038523322391007,
+      "learning_rate": 3.7219905808680663e-06,
+      "loss": 0.5563,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3574271095590027,
+      "grad_norm": 5.775055041451304,
+      "learning_rate": 3.7214168303372033e-06,
+      "loss": 0.4416,
+      "step": 4401
+    },
+    {
+      "epoch": 0.3575083245350443,
+      "grad_norm": 13.159177334829804,
+      "learning_rate": 3.72084299529076e-06,
+      "loss": 0.6903,
+      "step": 4402
+    },
+    {
+      "epoch": 0.35758953951108585,
+      "grad_norm": 6.4873380102452325,
+      "learning_rate": 3.720269075768442e-06,
+      "loss": 0.5353,
+      "step": 4403
+    },
+    {
+      "epoch": 0.35767075448712743,
+      "grad_norm": 6.187953414055915,
+      "learning_rate": 3.7196950718099636e-06,
+      "loss": 0.5224,
+      "step": 4404
+    },
+    {
+      "epoch": 0.357751969463169,
+      "grad_norm": 6.813848598601859,
+      "learning_rate": 3.71912098345504e-06,
+      "loss": 0.4981,
+      "step": 4405
+    },
+    {
+      "epoch": 0.3578331844392106,
+      "grad_norm": 5.500592742588713,
+      "learning_rate": 3.7185468107433966e-06,
+      "loss": 0.6389,
+      "step": 4406
+    },
+    {
+      "epoch": 0.35791439941525216,
+      "grad_norm": 4.764003153750818,
+      "learning_rate": 3.7179725537147638e-06,
+      "loss": 0.6802,
+      "step": 4407
+    },
+    {
+      "epoch": 0.35799561439129374,
+      "grad_norm": 23.629219188415377,
+      "learning_rate": 3.717398212408875e-06,
+      "loss": 0.6526,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3580768293673353,
+      "grad_norm": 6.635582406761952,
+      "learning_rate": 3.716823786865474e-06,
+      "loss": 0.4644,
+      "step": 4409
+    },
+    {
+      "epoch": 0.3581580443433769,
+      "grad_norm": 5.158135221215257,
+      "learning_rate": 3.7162492771243068e-06,
+      "loss": 0.5585,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3582392593194185,
+      "grad_norm": 3.8038224782647925,
+      "learning_rate": 3.7156746832251266e-06,
+      "loss": 0.6006,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3583204742954601,
+      "grad_norm": 3.935552912541937,
+      "learning_rate": 3.7151000052076913e-06,
+      "loss": 0.5972,
+      "step": 4412
+    },
+    {
+      "epoch": 0.3584016892715017,
+      "grad_norm": 5.2873448761373885,
+      "learning_rate": 3.7145252431117672e-06,
+      "loss": 0.5808,
+      "step": 4413
+    },
+    {
+      "epoch": 0.35848290424754325,
+      "grad_norm": 7.369629869130491,
+      "learning_rate": 3.713950396977124e-06,
+      "loss": 0.7058,
+      "step": 4414
+    },
+    {
+      "epoch": 0.35856411922358483,
+      "grad_norm": 5.201267760709762,
+      "learning_rate": 3.7133754668435377e-06,
+      "loss": 0.8023,
+      "step": 4415
+    },
+    {
+      "epoch": 0.3586453341996264,
+      "grad_norm": 3.1998164797424598,
+      "learning_rate": 3.7128004527507916e-06,
+      "loss": 0.596,
+      "step": 4416
+    },
+    {
+      "epoch": 0.358726549175668,
+      "grad_norm": 5.052174245450653,
+      "learning_rate": 3.712225354738672e-06,
+      "loss": 0.6137,
+      "step": 4417
+    },
+    {
+      "epoch": 0.35880776415170956,
+      "grad_norm": 4.804582850754989,
+      "learning_rate": 3.7116501728469746e-06,
+      "loss": 0.5863,
+      "step": 4418
+    },
+    {
+      "epoch": 0.35888897912775114,
+      "grad_norm": 4.754942942838345,
+      "learning_rate": 3.711074907115497e-06,
+      "loss": 0.5453,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3589701941037927,
+      "grad_norm": 4.182528054308997,
+      "learning_rate": 3.710499557584045e-06,
+      "loss": 0.6604,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3590514090798343,
+      "grad_norm": 5.430420823976959,
+      "learning_rate": 3.7099241242924306e-06,
+      "loss": 0.6099,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3591326240558759,
+      "grad_norm": 6.298870190762428,
+      "learning_rate": 3.7093486072804696e-06,
+      "loss": 0.5275,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3592138390319175,
+      "grad_norm": 4.826705123382431,
+      "learning_rate": 3.7087730065879862e-06,
+      "loss": 0.7317,
+      "step": 4423
+    },
+    {
+      "epoch": 0.3592950540079591,
+      "grad_norm": 13.171633048916624,
+      "learning_rate": 3.708197322254807e-06,
+      "loss": 0.6044,
+      "step": 4424
+    },
+    {
+      "epoch": 0.35937626898400066,
+      "grad_norm": 4.774839704277342,
+      "learning_rate": 3.7076215543207688e-06,
+      "loss": 0.6298,
+      "step": 4425
+    },
+    {
+      "epoch": 0.35945748396004223,
+      "grad_norm": 7.708019365223568,
+      "learning_rate": 3.7070457028257095e-06,
+      "loss": 0.4906,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3595386989360838,
+      "grad_norm": 6.0983843136049805,
+      "learning_rate": 3.7064697678094765e-06,
+      "loss": 0.459,
+      "step": 4427
+    },
+    {
+      "epoch": 0.3596199139121254,
+      "grad_norm": 4.100219190551162,
+      "learning_rate": 3.7058937493119195e-06,
+      "loss": 0.553,
+      "step": 4428
+    },
+    {
+      "epoch": 0.35970112888816697,
+      "grad_norm": 5.689077880893839,
+      "learning_rate": 3.705317647372898e-06,
+      "loss": 0.4724,
+      "step": 4429
+    },
+    {
+      "epoch": 0.35978234386420854,
+      "grad_norm": 3.858145204321467,
+      "learning_rate": 3.704741462032274e-06,
+      "loss": 0.4586,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3598635588402501,
+      "grad_norm": 11.983671864011523,
+      "learning_rate": 3.7041651933299167e-06,
+      "loss": 0.5293,
+      "step": 4431
+    },
+    {
+      "epoch": 0.3599447738162917,
+      "grad_norm": 5.870016078637099,
+      "learning_rate": 3.703588841305702e-06,
+      "loss": 0.6165,
+      "step": 4432
+    },
+    {
+      "epoch": 0.36002598879233333,
+      "grad_norm": 11.720283579984393,
+      "learning_rate": 3.7030124059995086e-06,
+      "loss": 0.5869,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3601072037683749,
+      "grad_norm": 3.3466945140934072,
+      "learning_rate": 3.7024358874512235e-06,
+      "loss": 0.5978,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3601884187444165,
+      "grad_norm": 6.264649818329499,
+      "learning_rate": 3.7018592857007386e-06,
+      "loss": 0.5366,
+      "step": 4435
+    },
+    {
+      "epoch": 0.36026963372045806,
+      "grad_norm": 5.505223118383528,
+      "learning_rate": 3.701282600787952e-06,
+      "loss": 0.5043,
+      "step": 4436
+    },
+    {
+      "epoch": 0.36035084869649964,
+      "grad_norm": 3.9050434398042793,
+      "learning_rate": 3.700705832752768e-06,
+      "loss": 0.4023,
+      "step": 4437
+    },
+    {
+      "epoch": 0.3604320636725412,
+      "grad_norm": 3.8486979131156818,
+      "learning_rate": 3.700128981635094e-06,
+      "loss": 0.7087,
+      "step": 4438
+    },
+    {
+      "epoch": 0.3605132786485828,
+      "grad_norm": 4.97236792052597,
+      "learning_rate": 3.6995520474748457e-06,
+      "loss": 0.676,
+      "step": 4439
+    },
+    {
+      "epoch": 0.36059449362462437,
+      "grad_norm": 5.54885373013609,
+      "learning_rate": 3.698975030311946e-06,
+      "loss": 0.4853,
+      "step": 4440
+    },
+    {
+      "epoch": 0.36067570860066595,
+      "grad_norm": 4.00154383333082,
+      "learning_rate": 3.6983979301863184e-06,
+      "loss": 0.6239,
+      "step": 4441
+    },
+    {
+      "epoch": 0.3607569235767075,
+      "grad_norm": 8.644737774833692,
+      "learning_rate": 3.6978207471378965e-06,
+      "loss": 0.648,
+      "step": 4442
+    },
+    {
+      "epoch": 0.3608381385527491,
+      "grad_norm": 4.192845183053263,
+      "learning_rate": 3.697243481206619e-06,
+      "loss": 0.5129,
+      "step": 4443
+    },
+    {
+      "epoch": 0.36091935352879073,
+      "grad_norm": 4.71407305163305,
+      "learning_rate": 3.6966661324324278e-06,
+      "loss": 0.6095,
+      "step": 4444
+    },
+    {
+      "epoch": 0.3610005685048323,
+      "grad_norm": 7.220930209286327,
+      "learning_rate": 3.6960887008552743e-06,
+      "loss": 0.4677,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3610817834808739,
+      "grad_norm": 4.962625757719452,
+      "learning_rate": 3.6955111865151127e-06,
+      "loss": 0.5154,
+      "step": 4446
+    },
+    {
+      "epoch": 0.36116299845691546,
+      "grad_norm": 10.29073366584436,
+      "learning_rate": 3.6949335894519033e-06,
+      "loss": 0.5977,
+      "step": 4447
+    },
+    {
+      "epoch": 0.36124421343295704,
+      "grad_norm": 4.688756562689801,
+      "learning_rate": 3.6943559097056155e-06,
+      "loss": 0.4716,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3613254284089986,
+      "grad_norm": 5.833182973807545,
+      "learning_rate": 3.6937781473162183e-06,
+      "loss": 0.6092,
+      "step": 4449
+    },
+    {
+      "epoch": 0.3614066433850402,
+      "grad_norm": 7.160318638424944,
+      "learning_rate": 3.6932003023236916e-06,
+      "loss": 0.6076,
+      "step": 4450
+    },
+    {
+      "epoch": 0.36148785836108177,
+      "grad_norm": 9.311074485235293,
+      "learning_rate": 3.692622374768019e-06,
+      "loss": 0.5193,
+      "step": 4451
+    },
+    {
+      "epoch": 0.36156907333712335,
+      "grad_norm": 12.351430012812038,
+      "learning_rate": 3.69204436468919e-06,
+      "loss": 0.4899,
+      "step": 4452
+    },
+    {
+      "epoch": 0.3616502883131649,
+      "grad_norm": 5.254888121313631,
+      "learning_rate": 3.6914662721272e-06,
+      "loss": 0.5991,
+      "step": 4453
+    },
+    {
+      "epoch": 0.36173150328920656,
+      "grad_norm": 4.86844411981734,
+      "learning_rate": 3.6908880971220494e-06,
+      "loss": 0.5256,
+      "step": 4454
+    },
+    {
+      "epoch": 0.36181271826524813,
+      "grad_norm": 11.036443384726775,
+      "learning_rate": 3.690309839713745e-06,
+      "loss": 0.504,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3618939332412897,
+      "grad_norm": 3.81687688264698,
+      "learning_rate": 3.6897314999423e-06,
+      "loss": 0.4975,
+      "step": 4456
+    },
+    {
+      "epoch": 0.3619751482173313,
+      "grad_norm": 3.12299941091288,
+      "learning_rate": 3.6891530778477306e-06,
+      "loss": 0.5774,
+      "step": 4457
+    },
+    {
+      "epoch": 0.36205636319337287,
+      "grad_norm": 5.036876057872486,
+      "learning_rate": 3.6885745734700628e-06,
+      "loss": 0.4885,
+      "step": 4458
+    },
+    {
+      "epoch": 0.36213757816941444,
+      "grad_norm": 8.62327137178916,
+      "learning_rate": 3.687995986849325e-06,
+      "loss": 0.6869,
+      "step": 4459
+    },
+    {
+      "epoch": 0.362218793145456,
+      "grad_norm": 3.4708246203276443,
+      "learning_rate": 3.687417318025551e-06,
+      "loss": 0.5733,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3623000081214976,
+      "grad_norm": 4.973444884464531,
+      "learning_rate": 3.686838567038784e-06,
+      "loss": 0.6222,
+      "step": 4461
+    },
+    {
+      "epoch": 0.3623812230975392,
+      "grad_norm": 5.667375423236751,
+      "learning_rate": 3.68625973392907e-06,
+      "loss": 0.5412,
+      "step": 4462
+    },
+    {
+      "epoch": 0.36246243807358075,
+      "grad_norm": 6.103470863477208,
+      "learning_rate": 3.6856808187364594e-06,
+      "loss": 0.5223,
+      "step": 4463
+    },
+    {
+      "epoch": 0.3625436530496223,
+      "grad_norm": 7.865435634910888,
+      "learning_rate": 3.685101821501012e-06,
+      "loss": 0.4664,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36262486802566396,
+      "grad_norm": 4.7253648152592715,
+      "learning_rate": 3.6845227422627904e-06,
+      "loss": 0.3839,
+      "step": 4465
+    },
+    {
+      "epoch": 0.36270608300170554,
+      "grad_norm": 5.338042758405891,
+      "learning_rate": 3.683943581061864e-06,
+      "loss": 0.587,
+      "step": 4466
+    },
+    {
+      "epoch": 0.3627872979777471,
+      "grad_norm": 5.179836245143301,
+      "learning_rate": 3.683364337938308e-06,
+      "loss": 0.5633,
+      "step": 4467
+    },
+    {
+      "epoch": 0.3628685129537887,
+      "grad_norm": 6.353705665638407,
+      "learning_rate": 3.6827850129322017e-06,
+      "loss": 0.6154,
+      "step": 4468
+    },
+    {
+      "epoch": 0.36294972792983027,
+      "grad_norm": 4.584366349679948,
+      "learning_rate": 3.682205606083633e-06,
+      "loss": 0.4864,
+      "step": 4469
+    },
+    {
+      "epoch": 0.36303094290587185,
+      "grad_norm": 6.208825994136881,
+      "learning_rate": 3.681626117432693e-06,
+      "loss": 0.4577,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3631121578819134,
+      "grad_norm": 6.366010507681265,
+      "learning_rate": 3.6810465470194796e-06,
+      "loss": 0.4515,
+      "step": 4471
+    },
+    {
+      "epoch": 0.363193372857955,
+      "grad_norm": 4.554722007028088,
+      "learning_rate": 3.680466894884096e-06,
+      "loss": 0.5824,
+      "step": 4472
+    },
+    {
+      "epoch": 0.3632745878339966,
+      "grad_norm": 14.16578555542569,
+      "learning_rate": 3.6798871610666497e-06,
+      "loss": 0.5687,
+      "step": 4473
+    },
+    {
+      "epoch": 0.36335580281003815,
+      "grad_norm": 6.940126197441687,
+      "learning_rate": 3.679307345607257e-06,
+      "loss": 0.5116,
+      "step": 4474
+    },
+    {
+      "epoch": 0.36343701778607973,
+      "grad_norm": 11.501609688473003,
+      "learning_rate": 3.6787274485460377e-06,
+      "loss": 0.5931,
+      "step": 4475
+    },
+    {
+      "epoch": 0.36351823276212136,
+      "grad_norm": 9.174827684495176,
+      "learning_rate": 3.678147469923117e-06,
+      "loss": 0.5515,
+      "step": 4476
+    },
+    {
+      "epoch": 0.36359944773816294,
+      "grad_norm": 4.650571061924677,
+      "learning_rate": 3.677567409778626e-06,
+      "loss": 0.4882,
+      "step": 4477
+    },
+    {
+      "epoch": 0.3636806627142045,
+      "grad_norm": 4.716688957959236,
+      "learning_rate": 3.6769872681527036e-06,
+      "loss": 0.5081,
+      "step": 4478
+    },
+    {
+      "epoch": 0.3637618776902461,
+      "grad_norm": 6.347553656163701,
+      "learning_rate": 3.6764070450854907e-06,
+      "loss": 0.4636,
+      "step": 4479
+    },
+    {
+      "epoch": 0.36384309266628767,
+      "grad_norm": 10.156227178827374,
+      "learning_rate": 3.675826740617136e-06,
+      "loss": 0.4511,
+      "step": 4480
+    },
+    {
+      "epoch": 0.36392430764232925,
+      "grad_norm": 3.559954335727195,
+      "learning_rate": 3.6752463547877946e-06,
+      "loss": 0.6027,
+      "step": 4481
+    },
+    {
+      "epoch": 0.3640055226183708,
+      "grad_norm": 4.72021021192598,
+      "learning_rate": 3.674665887637625e-06,
+      "loss": 0.5531,
+      "step": 4482
+    },
+    {
+      "epoch": 0.3640867375944124,
+      "grad_norm": 11.263712709683006,
+      "learning_rate": 3.6740853392067925e-06,
+      "loss": 0.609,
+      "step": 4483
+    },
+    {
+      "epoch": 0.364167952570454,
+      "grad_norm": 5.123832459796128,
+      "learning_rate": 3.6735047095354693e-06,
+      "loss": 0.458,
+      "step": 4484
+    },
+    {
+      "epoch": 0.36424916754649556,
+      "grad_norm": 4.265471041437096,
+      "learning_rate": 3.67292399866383e-06,
+      "loss": 0.8243,
+      "step": 4485
+    },
+    {
+      "epoch": 0.36433038252253713,
+      "grad_norm": 7.3136697189021405,
+      "learning_rate": 3.6723432066320575e-06,
+      "loss": 0.5279,
+      "step": 4486
+    },
+    {
+      "epoch": 0.36441159749857877,
+      "grad_norm": 3.4708908613272618,
+      "learning_rate": 3.67176233348034e-06,
+      "loss": 0.5274,
+      "step": 4487
+    },
+    {
+      "epoch": 0.36449281247462034,
+      "grad_norm": 3.2379003246205205,
+      "learning_rate": 3.6711813792488706e-06,
+      "loss": 0.4907,
+      "step": 4488
+    },
+    {
+      "epoch": 0.3645740274506619,
+      "grad_norm": 4.61818923987606,
+      "learning_rate": 3.6706003439778476e-06,
+      "loss": 0.5109,
+      "step": 4489
+    },
+    {
+      "epoch": 0.3646552424267035,
+      "grad_norm": 6.1238028357576555,
+      "learning_rate": 3.6700192277074766e-06,
+      "loss": 0.4814,
+      "step": 4490
+    },
+    {
+      "epoch": 0.3647364574027451,
+      "grad_norm": 5.108669074454857,
+      "learning_rate": 3.6694380304779676e-06,
+      "loss": 0.4514,
+      "step": 4491
+    },
+    {
+      "epoch": 0.36481767237878665,
+      "grad_norm": 4.555539206226715,
+      "learning_rate": 3.6688567523295356e-06,
+      "loss": 0.6225,
+      "step": 4492
+    },
+    {
+      "epoch": 0.3648988873548282,
+      "grad_norm": 4.947835075771073,
+      "learning_rate": 3.668275393302402e-06,
+      "loss": 0.526,
+      "step": 4493
+    },
+    {
+      "epoch": 0.3649801023308698,
+      "grad_norm": 4.987280419667079,
+      "learning_rate": 3.667693953436795e-06,
+      "loss": 0.5885,
+      "step": 4494
+    },
+    {
+      "epoch": 0.3650613173069114,
+      "grad_norm": 5.33019974406346,
+      "learning_rate": 3.6671124327729457e-06,
+      "loss": 0.5906,
+      "step": 4495
+    },
+    {
+      "epoch": 0.36514253228295296,
+      "grad_norm": 3.911550454532029,
+      "learning_rate": 3.6665308313510927e-06,
+      "loss": 0.4641,
+      "step": 4496
+    },
+    {
+      "epoch": 0.36522374725899454,
+      "grad_norm": 7.327451479070638,
+      "learning_rate": 3.665949149211481e-06,
+      "loss": 0.5346,
+      "step": 4497
+    },
+    {
+      "epoch": 0.36530496223503617,
+      "grad_norm": 6.679619506230374,
+      "learning_rate": 3.6653673863943584e-06,
+      "loss": 0.564,
+      "step": 4498
+    },
+    {
+      "epoch": 0.36538617721107775,
+      "grad_norm": 7.979525255569825,
+      "learning_rate": 3.6647855429399803e-06,
+      "loss": 0.5412,
+      "step": 4499
+    },
+    {
+      "epoch": 0.3654673921871193,
+      "grad_norm": 5.930675497170029,
+      "learning_rate": 3.6642036188886072e-06,
+      "loss": 0.5653,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3655486071631609,
+      "grad_norm": 4.872227933745589,
+      "learning_rate": 3.663621614280505e-06,
+      "loss": 0.6143,
+      "step": 4501
+    },
+    {
+      "epoch": 0.3656298221392025,
+      "grad_norm": 4.942347208045127,
+      "learning_rate": 3.663039529155945e-06,
+      "loss": 0.5996,
+      "step": 4502
+    },
+    {
+      "epoch": 0.36571103711524405,
+      "grad_norm": 3.9093013103451084,
+      "learning_rate": 3.6624573635552056e-06,
+      "loss": 0.6432,
+      "step": 4503
+    },
+    {
+      "epoch": 0.36579225209128563,
+      "grad_norm": 6.258360673482065,
+      "learning_rate": 3.6618751175185687e-06,
+      "loss": 0.5461,
+      "step": 4504
+    },
+    {
+      "epoch": 0.3658734670673272,
+      "grad_norm": 6.512518797476682,
+      "learning_rate": 3.6612927910863235e-06,
+      "loss": 0.4864,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3659546820433688,
+      "grad_norm": 4.622906130526112,
+      "learning_rate": 3.660710384298762e-06,
+      "loss": 0.6799,
+      "step": 4506
+    },
+    {
+      "epoch": 0.36603589701941036,
+      "grad_norm": 3.157312017939539,
+      "learning_rate": 3.6601278971961853e-06,
+      "loss": 0.6479,
+      "step": 4507
+    },
+    {
+      "epoch": 0.36611711199545194,
+      "grad_norm": 5.831757782243094,
+      "learning_rate": 3.659545329818898e-06,
+      "loss": 0.3765,
+      "step": 4508
+    },
+    {
+      "epoch": 0.36619832697149357,
+      "grad_norm": 4.871188343403577,
+      "learning_rate": 3.6589626822072105e-06,
+      "loss": 0.5067,
+      "step": 4509
+    },
+    {
+      "epoch": 0.36627954194753515,
+      "grad_norm": 3.3143048064685825,
+      "learning_rate": 3.6583799544014397e-06,
+      "loss": 0.7395,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3663607569235767,
+      "grad_norm": 6.5927342002006535,
+      "learning_rate": 3.6577971464419064e-06,
+      "loss": 0.4955,
+      "step": 4511
+    },
+    {
+      "epoch": 0.3664419718996183,
+      "grad_norm": 5.678543552265716,
+      "learning_rate": 3.6572142583689372e-06,
+      "loss": 0.5946,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3665231868756599,
+      "grad_norm": 9.978179156767617,
+      "learning_rate": 3.656631290222867e-06,
+      "loss": 0.6531,
+      "step": 4513
+    },
+    {
+      "epoch": 0.36660440185170146,
+      "grad_norm": 4.4736754108351215,
+      "learning_rate": 3.656048242044033e-06,
+      "loss": 0.545,
+      "step": 4514
+    },
+    {
+      "epoch": 0.36668561682774303,
+      "grad_norm": 3.069124245403996,
+      "learning_rate": 3.655465113872779e-06,
+      "loss": 0.4614,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3667668318037846,
+      "grad_norm": 5.78003187791633,
+      "learning_rate": 3.6548819057494533e-06,
+      "loss": 0.5642,
+      "step": 4516
+    },
+    {
+      "epoch": 0.3668480467798262,
+      "grad_norm": 6.651983645474769,
+      "learning_rate": 3.6542986177144124e-06,
+      "loss": 0.726,
+      "step": 4517
+    },
+    {
+      "epoch": 0.36692926175586776,
+      "grad_norm": 5.61902839950427,
+      "learning_rate": 3.6537152498080165e-06,
+      "loss": 0.5679,
+      "step": 4518
+    },
+    {
+      "epoch": 0.36701047673190934,
+      "grad_norm": 3.6610484670477597,
+      "learning_rate": 3.653131802070631e-06,
+      "loss": 0.6392,
+      "step": 4519
+    },
+    {
+      "epoch": 0.367091691707951,
+      "grad_norm": 4.744064286312601,
+      "learning_rate": 3.6525482745426277e-06,
+      "loss": 0.5744,
+      "step": 4520
+    },
+    {
+      "epoch": 0.36717290668399255,
+      "grad_norm": 5.713260181456036,
+      "learning_rate": 3.6519646672643837e-06,
+      "loss": 0.647,
+      "step": 4521
+    },
+    {
+      "epoch": 0.3672541216600341,
+      "grad_norm": 4.789967162852778,
+      "learning_rate": 3.6513809802762805e-06,
+      "loss": 0.5129,
+      "step": 4522
+    },
+    {
+      "epoch": 0.3673353366360757,
+      "grad_norm": 5.3194037700176215,
+      "learning_rate": 3.6507972136187082e-06,
+      "loss": 0.53,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3674165516121173,
+      "grad_norm": 4.077739067704118,
+      "learning_rate": 3.650213367332059e-06,
+      "loss": 0.4182,
+      "step": 4524
+    },
+    {
+      "epoch": 0.36749776658815886,
+      "grad_norm": 5.963767765697606,
+      "learning_rate": 3.6496294414567313e-06,
+      "loss": 0.5525,
+      "step": 4525
+    },
+    {
+      "epoch": 0.36757898156420044,
+      "grad_norm": 16.366327402411947,
+      "learning_rate": 3.649045436033132e-06,
+      "loss": 0.6207,
+      "step": 4526
+    },
+    {
+      "epoch": 0.367660196540242,
+      "grad_norm": 4.600677095064222,
+      "learning_rate": 3.6484613511016693e-06,
+      "loss": 0.5597,
+      "step": 4527
+    },
+    {
+      "epoch": 0.3677414115162836,
+      "grad_norm": 4.635569455138606,
+      "learning_rate": 3.6478771867027585e-06,
+      "loss": 0.5274,
+      "step": 4528
+    },
+    {
+      "epoch": 0.36782262649232517,
+      "grad_norm": 13.8186155722183,
+      "learning_rate": 3.647292942876822e-06,
+      "loss": 0.3283,
+      "step": 4529
+    },
+    {
+      "epoch": 0.36790384146836674,
+      "grad_norm": 4.921988696473463,
+      "learning_rate": 3.646708619664286e-06,
+      "loss": 0.666,
+      "step": 4530
+    },
+    {
+      "epoch": 0.3679850564444084,
+      "grad_norm": 5.125112781896423,
+      "learning_rate": 3.646124217105582e-06,
+      "loss": 0.504,
+      "step": 4531
+    },
+    {
+      "epoch": 0.36806627142044995,
+      "grad_norm": 4.160106735681602,
+      "learning_rate": 3.645539735241148e-06,
+      "loss": 0.4889,
+      "step": 4532
+    },
+    {
+      "epoch": 0.36814748639649153,
+      "grad_norm": 11.018187691856639,
+      "learning_rate": 3.6449551741114277e-06,
+      "loss": 0.5853,
+      "step": 4533
+    },
+    {
+      "epoch": 0.3682287013725331,
+      "grad_norm": 6.301041751261872,
+      "learning_rate": 3.6443705337568683e-06,
+      "loss": 0.5506,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3683099163485747,
+      "grad_norm": 2.9536104737688635,
+      "learning_rate": 3.643785814217924e-06,
+      "loss": 0.4832,
+      "step": 4535
+    },
+    {
+      "epoch": 0.36839113132461626,
+      "grad_norm": 6.237830789303934,
+      "learning_rate": 3.6432010155350556e-06,
+      "loss": 0.5408,
+      "step": 4536
+    },
+    {
+      "epoch": 0.36847234630065784,
+      "grad_norm": 5.326352917032812,
+      "learning_rate": 3.642616137748727e-06,
+      "loss": 0.5208,
+      "step": 4537
+    },
+    {
+      "epoch": 0.3685535612766994,
+      "grad_norm": 6.248306452774462,
+      "learning_rate": 3.6420311808994084e-06,
+      "loss": 0.5739,
+      "step": 4538
+    },
+    {
+      "epoch": 0.368634776252741,
+      "grad_norm": 4.147693647179564,
+      "learning_rate": 3.641446145027577e-06,
+      "loss": 0.6192,
+      "step": 4539
+    },
+    {
+      "epoch": 0.36871599122878257,
+      "grad_norm": 6.469141494304793,
+      "learning_rate": 3.640861030173713e-06,
+      "loss": 0.5873,
+      "step": 4540
+    },
+    {
+      "epoch": 0.36879720620482415,
+      "grad_norm": 7.33657703163369,
+      "learning_rate": 3.6402758363783037e-06,
+      "loss": 0.6911,
+      "step": 4541
+    },
+    {
+      "epoch": 0.3688784211808658,
+      "grad_norm": 5.919960884683257,
+      "learning_rate": 3.639690563681841e-06,
+      "loss": 0.5873,
+      "step": 4542
+    },
+    {
+      "epoch": 0.36895963615690736,
+      "grad_norm": 18.615597531360145,
+      "learning_rate": 3.6391052121248233e-06,
+      "loss": 0.5995,
+      "step": 4543
+    },
+    {
+      "epoch": 0.36904085113294893,
+      "grad_norm": 7.146041859144082,
+      "learning_rate": 3.6385197817477535e-06,
+      "loss": 0.4588,
+      "step": 4544
+    },
+    {
+      "epoch": 0.3691220661089905,
+      "grad_norm": 4.196856311260542,
+      "learning_rate": 3.6379342725911402e-06,
+      "loss": 0.4738,
+      "step": 4545
+    },
+    {
+      "epoch": 0.3692032810850321,
+      "grad_norm": 4.917160222161063,
+      "learning_rate": 3.637348684695498e-06,
+      "loss": 0.6132,
+      "step": 4546
+    },
+    {
+      "epoch": 0.36928449606107366,
+      "grad_norm": 3.5260428355619218,
+      "learning_rate": 3.6367630181013457e-06,
+      "loss": 0.3245,
+      "step": 4547
+    },
+    {
+      "epoch": 0.36936571103711524,
+      "grad_norm": 5.486577536708969,
+      "learning_rate": 3.6361772728492096e-06,
+      "loss": 0.6407,
+      "step": 4548
+    },
+    {
+      "epoch": 0.3694469260131568,
+      "grad_norm": 5.527101563539024,
+      "learning_rate": 3.6355914489796185e-06,
+      "loss": 0.645,
+      "step": 4549
+    },
+    {
+      "epoch": 0.3695281409891984,
+      "grad_norm": 5.8535848869679405,
+      "learning_rate": 3.6350055465331098e-06,
+      "loss": 0.5124,
+      "step": 4550
+    },
+    {
+      "epoch": 0.36960935596523997,
+      "grad_norm": 3.0247220854088592,
+      "learning_rate": 3.6344195655502233e-06,
+      "loss": 0.5123,
+      "step": 4551
+    },
+    {
+      "epoch": 0.36969057094128155,
+      "grad_norm": 5.325222924482676,
+      "learning_rate": 3.633833506071508e-06,
+      "loss": 0.3802,
+      "step": 4552
+    },
+    {
+      "epoch": 0.3697717859173232,
+      "grad_norm": 6.707191666111479,
+      "learning_rate": 3.6332473681375146e-06,
+      "loss": 0.4215,
+      "step": 4553
+    },
+    {
+      "epoch": 0.36985300089336476,
+      "grad_norm": 6.3473511974694885,
+      "learning_rate": 3.6326611517888e-06,
+      "loss": 0.5247,
+      "step": 4554
+    },
+    {
+      "epoch": 0.36993421586940634,
+      "grad_norm": 4.62624452294825,
+      "learning_rate": 3.632074857065928e-06,
+      "loss": 0.4702,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3700154308454479,
+      "grad_norm": 3.421641642355826,
+      "learning_rate": 3.631488484009469e-06,
+      "loss": 0.5228,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3700966458214895,
+      "grad_norm": 6.538600958475513,
+      "learning_rate": 3.630902032659994e-06,
+      "loss": 0.5604,
+      "step": 4557
+    },
+    {
+      "epoch": 0.37017786079753107,
+      "grad_norm": 6.779583853613752,
+      "learning_rate": 3.6303155030580834e-06,
+      "loss": 0.5324,
+      "step": 4558
+    },
+    {
+      "epoch": 0.37025907577357264,
+      "grad_norm": 4.786962125388907,
+      "learning_rate": 3.629728895244323e-06,
+      "loss": 0.4619,
+      "step": 4559
+    },
+    {
+      "epoch": 0.3703402907496142,
+      "grad_norm": 5.970025695408968,
+      "learning_rate": 3.6291422092593016e-06,
+      "loss": 0.5009,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3704215057256558,
+      "grad_norm": 5.054283857837389,
+      "learning_rate": 3.628555445143615e-06,
+      "loss": 0.5614,
+      "step": 4561
+    },
+    {
+      "epoch": 0.3705027207016974,
+      "grad_norm": 4.038161358633893,
+      "learning_rate": 3.6279686029378646e-06,
+      "loss": 0.5086,
+      "step": 4562
+    },
+    {
+      "epoch": 0.37058393567773895,
+      "grad_norm": 4.169489051463636,
+      "learning_rate": 3.6273816826826565e-06,
+      "loss": 0.4921,
+      "step": 4563
+    },
+    {
+      "epoch": 0.3706651506537806,
+      "grad_norm": 6.938462515442908,
+      "learning_rate": 3.6267946844186023e-06,
+      "loss": 0.5351,
+      "step": 4564
+    },
+    {
+      "epoch": 0.37074636562982216,
+      "grad_norm": 4.177658193519644,
+      "learning_rate": 3.6262076081863195e-06,
+      "loss": 0.4798,
+      "step": 4565
+    },
+    {
+      "epoch": 0.37082758060586374,
+      "grad_norm": 3.350413459828565,
+      "learning_rate": 3.625620454026431e-06,
+      "loss": 0.468,
+      "step": 4566
+    },
+    {
+      "epoch": 0.3709087955819053,
+      "grad_norm": 5.601772436876186,
+      "learning_rate": 3.625033221979564e-06,
+      "loss": 0.5411,
+      "step": 4567
+    },
+    {
+      "epoch": 0.3709900105579469,
+      "grad_norm": 5.46442194527407,
+      "learning_rate": 3.624445912086352e-06,
+      "loss": 0.6332,
+      "step": 4568
+    },
+    {
+      "epoch": 0.37107122553398847,
+      "grad_norm": 5.8554252343208,
+      "learning_rate": 3.6238585243874346e-06,
+      "loss": 0.6209,
+      "step": 4569
+    },
+    {
+      "epoch": 0.37115244051003005,
+      "grad_norm": 6.38111885179741,
+      "learning_rate": 3.6232710589234556e-06,
+      "loss": 0.5771,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3712336554860716,
+      "grad_norm": 4.123032307940846,
+      "learning_rate": 3.6226835157350625e-06,
+      "loss": 0.6513,
+      "step": 4571
+    },
+    {
+      "epoch": 0.3713148704621132,
+      "grad_norm": 5.925335677806806,
+      "learning_rate": 3.6220958948629137e-06,
+      "loss": 0.6454,
+      "step": 4572
+    },
+    {
+      "epoch": 0.3713960854381548,
+      "grad_norm": 4.123001819171552,
+      "learning_rate": 3.621508196347667e-06,
+      "loss": 0.5576,
+      "step": 4573
+    },
+    {
+      "epoch": 0.37147730041419635,
+      "grad_norm": 6.967825569297834,
+      "learning_rate": 3.6209204202299875e-06,
+      "loss": 0.6256,
+      "step": 4574
+    },
+    {
+      "epoch": 0.371558515390238,
+      "grad_norm": 6.347779650753475,
+      "learning_rate": 3.6203325665505486e-06,
+      "loss": 0.632,
+      "step": 4575
+    },
+    {
+      "epoch": 0.37163973036627956,
+      "grad_norm": 3.8118384188679943,
+      "learning_rate": 3.619744635350025e-06,
+      "loss": 0.5919,
+      "step": 4576
+    },
+    {
+      "epoch": 0.37172094534232114,
+      "grad_norm": 7.993997703219614,
+      "learning_rate": 3.619156626669098e-06,
+      "loss": 0.6338,
+      "step": 4577
+    },
+    {
+      "epoch": 0.3718021603183627,
+      "grad_norm": 4.85057042384623,
+      "learning_rate": 3.6185685405484566e-06,
+      "loss": 0.5769,
+      "step": 4578
+    },
+    {
+      "epoch": 0.3718833752944043,
+      "grad_norm": 4.65594156616089,
+      "learning_rate": 3.6179803770287913e-06,
+      "loss": 0.4109,
+      "step": 4579
+    },
+    {
+      "epoch": 0.37196459027044587,
+      "grad_norm": 4.986423129791739,
+      "learning_rate": 3.6173921361508012e-06,
+      "loss": 0.3943,
+      "step": 4580
+    },
+    {
+      "epoch": 0.37204580524648745,
+      "grad_norm": 6.284947322407295,
+      "learning_rate": 3.616803817955189e-06,
+      "loss": 0.5428,
+      "step": 4581
+    },
+    {
+      "epoch": 0.372127020222529,
+      "grad_norm": 5.691836227913848,
+      "learning_rate": 3.6162154224826627e-06,
+      "loss": 0.5465,
+      "step": 4582
+    },
+    {
+      "epoch": 0.3722082351985706,
+      "grad_norm": 8.95784788971493,
+      "learning_rate": 3.615626949773937e-06,
+      "loss": 0.4766,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3722894501746122,
+      "grad_norm": 4.7473375463729495,
+      "learning_rate": 3.6150383998697315e-06,
+      "loss": 0.5862,
+      "step": 4584
+    },
+    {
+      "epoch": 0.37237066515065376,
+      "grad_norm": 5.4505240290945585,
+      "learning_rate": 3.614449772810769e-06,
+      "loss": 0.4044,
+      "step": 4585
+    },
+    {
+      "epoch": 0.3724518801266954,
+      "grad_norm": 6.977930137629381,
+      "learning_rate": 3.613861068637781e-06,
+      "loss": 0.4798,
+      "step": 4586
+    },
+    {
+      "epoch": 0.37253309510273697,
+      "grad_norm": 6.5403690266388335,
+      "learning_rate": 3.6132722873915017e-06,
+      "loss": 0.6262,
+      "step": 4587
+    },
+    {
+      "epoch": 0.37261431007877854,
+      "grad_norm": 6.834395602891912,
+      "learning_rate": 3.6126834291126724e-06,
+      "loss": 0.4946,
+      "step": 4588
+    },
+    {
+      "epoch": 0.3726955250548201,
+      "grad_norm": 3.6009773174215054,
+      "learning_rate": 3.6120944938420384e-06,
+      "loss": 0.552,
+      "step": 4589
+    },
+    {
+      "epoch": 0.3727767400308617,
+      "grad_norm": 6.184936256815004,
+      "learning_rate": 3.6115054816203504e-06,
+      "loss": 0.4826,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3728579550069033,
+      "grad_norm": 4.521800781478745,
+      "learning_rate": 3.6109163924883668e-06,
+      "loss": 0.595,
+      "step": 4591
+    },
+    {
+      "epoch": 0.37293916998294485,
+      "grad_norm": 5.129658876888231,
+      "learning_rate": 3.6103272264868473e-06,
+      "loss": 0.5459,
+      "step": 4592
+    },
+    {
+      "epoch": 0.37302038495898643,
+      "grad_norm": 3.955425189400771,
+      "learning_rate": 3.6097379836565604e-06,
+      "loss": 0.7445,
+      "step": 4593
+    },
+    {
+      "epoch": 0.373101599935028,
+      "grad_norm": 4.651648963857242,
+      "learning_rate": 3.6091486640382785e-06,
+      "loss": 0.5907,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3731828149110696,
+      "grad_norm": 3.3474693180043045,
+      "learning_rate": 3.6085592676727786e-06,
+      "loss": 0.332,
+      "step": 4595
+    },
+    {
+      "epoch": 0.37326402988711116,
+      "grad_norm": 6.634033909104456,
+      "learning_rate": 3.6079697946008453e-06,
+      "loss": 0.4753,
+      "step": 4596
+    },
+    {
+      "epoch": 0.3733452448631528,
+      "grad_norm": 5.83533093603698,
+      "learning_rate": 3.607380244863265e-06,
+      "loss": 0.5514,
+      "step": 4597
+    },
+    {
+      "epoch": 0.37342645983919437,
+      "grad_norm": 4.575836973417688,
+      "learning_rate": 3.6067906185008328e-06,
+      "loss": 0.3765,
+      "step": 4598
+    },
+    {
+      "epoch": 0.37350767481523595,
+      "grad_norm": 4.848076544705409,
+      "learning_rate": 3.6062009155543483e-06,
+      "loss": 0.5876,
+      "step": 4599
+    },
+    {
+      "epoch": 0.3735888897912775,
+      "grad_norm": 5.836658897942606,
+      "learning_rate": 3.6056111360646134e-06,
+      "loss": 0.4459,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3736701047673191,
+      "grad_norm": 3.9899151337710306,
+      "learning_rate": 3.6050212800724403e-06,
+      "loss": 0.4275,
+      "step": 4601
+    },
+    {
+      "epoch": 0.3737513197433607,
+      "grad_norm": 4.289228152113736,
+      "learning_rate": 3.6044313476186433e-06,
+      "loss": 0.5786,
+      "step": 4602
+    },
+    {
+      "epoch": 0.37383253471940225,
+      "grad_norm": 5.434462312775084,
+      "learning_rate": 3.603841338744041e-06,
+      "loss": 0.4954,
+      "step": 4603
+    },
+    {
+      "epoch": 0.37391374969544383,
+      "grad_norm": 3.5351327522854423,
+      "learning_rate": 3.6032512534894597e-06,
+      "loss": 0.6879,
+      "step": 4604
+    },
+    {
+      "epoch": 0.3739949646714854,
+      "grad_norm": 5.549918211724535,
+      "learning_rate": 3.602661091895732e-06,
+      "loss": 0.43,
+      "step": 4605
+    },
+    {
+      "epoch": 0.374076179647527,
+      "grad_norm": 4.083602872009912,
+      "learning_rate": 3.602070854003692e-06,
+      "loss": 0.5157,
+      "step": 4606
+    },
+    {
+      "epoch": 0.37415739462356856,
+      "grad_norm": 7.5939379592765786,
+      "learning_rate": 3.6014805398541815e-06,
+      "loss": 0.5669,
+      "step": 4607
+    },
+    {
+      "epoch": 0.3742386095996102,
+      "grad_norm": 3.8544939676204355,
+      "learning_rate": 3.6008901494880467e-06,
+      "loss": 0.4815,
+      "step": 4608
+    },
+    {
+      "epoch": 0.37431982457565177,
+      "grad_norm": 6.619794981662382,
+      "learning_rate": 3.60029968294614e-06,
+      "loss": 0.8072,
+      "step": 4609
+    },
+    {
+      "epoch": 0.37440103955169335,
+      "grad_norm": 9.067352524497199,
+      "learning_rate": 3.599709140269319e-06,
+      "loss": 0.409,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3744822545277349,
+      "grad_norm": 4.980681642925656,
+      "learning_rate": 3.599118521498445e-06,
+      "loss": 0.6207,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3745634695037765,
+      "grad_norm": 5.8795928027879585,
+      "learning_rate": 3.598527826674387e-06,
+      "loss": 0.4077,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3746446844798181,
+      "grad_norm": 6.664130906786667,
+      "learning_rate": 3.597937055838017e-06,
+      "loss": 0.4889,
+      "step": 4613
+    },
+    {
+      "epoch": 0.37472589945585966,
+      "grad_norm": 3.512866216429364,
+      "learning_rate": 3.5973462090302137e-06,
+      "loss": 0.556,
+      "step": 4614
+    },
+    {
+      "epoch": 0.37480711443190123,
+      "grad_norm": 7.915343003804626,
+      "learning_rate": 3.5967552862918603e-06,
+      "loss": 0.5955,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3748883294079428,
+      "grad_norm": 5.142753619848735,
+      "learning_rate": 3.596164287663845e-06,
+      "loss": 0.4933,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3749695443839844,
+      "grad_norm": 5.185440275795067,
+      "learning_rate": 3.5955732131870626e-06,
+      "loss": 0.9477,
+      "step": 4617
+    },
+    {
+      "epoch": 0.37505075936002596,
+      "grad_norm": 12.990297623448031,
+      "learning_rate": 3.594982062902412e-06,
+      "loss": 0.6607,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3751319743360676,
+      "grad_norm": 4.182193438396625,
+      "learning_rate": 3.5943908368507985e-06,
+      "loss": 0.6018,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3752131893121092,
+      "grad_norm": 3.8071948876468733,
+      "learning_rate": 3.59379953507313e-06,
+      "loss": 0.5165,
+      "step": 4620
+    },
+    {
+      "epoch": 0.37529440428815075,
+      "grad_norm": 4.47056756713465,
+      "learning_rate": 3.593208157610324e-06,
+      "loss": 0.4829,
+      "step": 4621
+    },
+    {
+      "epoch": 0.37537561926419233,
+      "grad_norm": 4.861658029251169,
+      "learning_rate": 3.592616704503298e-06,
+      "loss": 0.5083,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3754568342402339,
+      "grad_norm": 4.194610134145657,
+      "learning_rate": 3.5920251757929787e-06,
+      "loss": 0.5095,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3755380492162755,
+      "grad_norm": 6.791531269079618,
+      "learning_rate": 3.5914335715202976e-06,
+      "loss": 0.4922,
+      "step": 4624
+    },
+    {
+      "epoch": 0.37561926419231706,
+      "grad_norm": 5.2361522290402185,
+      "learning_rate": 3.590841891726189e-06,
+      "loss": 0.6261,
+      "step": 4625
+    },
+    {
+      "epoch": 0.37570047916835864,
+      "grad_norm": 10.082840154908705,
+      "learning_rate": 3.5902501364515945e-06,
+      "loss": 0.5765,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3757816941444002,
+      "grad_norm": 6.544537456866241,
+      "learning_rate": 3.5896583057374607e-06,
+      "loss": 0.3993,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3758629091204418,
+      "grad_norm": 4.601861451010068,
+      "learning_rate": 3.589066399624739e-06,
+      "loss": 0.6202,
+      "step": 4628
+    },
+    {
+      "epoch": 0.37594412409648337,
+      "grad_norm": 4.584192839631751,
+      "learning_rate": 3.5884744181543868e-06,
+      "loss": 0.5738,
+      "step": 4629
+    },
+    {
+      "epoch": 0.376025339072525,
+      "grad_norm": 3.856688627858152,
+      "learning_rate": 3.5878823613673652e-06,
+      "loss": 0.4293,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3761065540485666,
+      "grad_norm": 3.3977632115418106,
+      "learning_rate": 3.5872902293046417e-06,
+      "loss": 0.4848,
+      "step": 4631
+    },
+    {
+      "epoch": 0.37618776902460815,
+      "grad_norm": 5.9959616576129395,
+      "learning_rate": 3.586698022007189e-06,
+      "loss": 0.554,
+      "step": 4632
+    },
+    {
+      "epoch": 0.37626898400064973,
+      "grad_norm": 4.593599550308795,
+      "learning_rate": 3.5861057395159837e-06,
+      "loss": 0.494,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3763501989766913,
+      "grad_norm": 3.9971449166055644,
+      "learning_rate": 3.5855133818720106e-06,
+      "loss": 0.3877,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3764314139527329,
+      "grad_norm": 10.899762207965088,
+      "learning_rate": 3.5849209491162555e-06,
+      "loss": 0.4489,
+      "step": 4635
+    },
+    {
+      "epoch": 0.37651262892877446,
+      "grad_norm": 3.0450268530430598,
+      "learning_rate": 3.5843284412897127e-06,
+      "loss": 0.6891,
+      "step": 4636
+    },
+    {
+      "epoch": 0.37659384390481604,
+      "grad_norm": 5.435757415343299,
+      "learning_rate": 3.5837358584333814e-06,
+      "loss": 0.6111,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3766750588808576,
+      "grad_norm": 4.283255422002926,
+      "learning_rate": 3.583143200588263e-06,
+      "loss": 0.4667,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3767562738568992,
+      "grad_norm": 2.6058528884262553,
+      "learning_rate": 3.5825504677953684e-06,
+      "loss": 0.5796,
+      "step": 4639
+    },
+    {
+      "epoch": 0.37683748883294077,
+      "grad_norm": 5.535677617104462,
+      "learning_rate": 3.581957660095711e-06,
+      "loss": 0.579,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3769187038089824,
+      "grad_norm": 7.993394293530673,
+      "learning_rate": 3.5813647775303084e-06,
+      "loss": 0.531,
+      "step": 4641
+    },
+    {
+      "epoch": 0.376999918785024,
+      "grad_norm": 6.421520761195438,
+      "learning_rate": 3.580771820140187e-06,
+      "loss": 0.5836,
+      "step": 4642
+    },
+    {
+      "epoch": 0.37708113376106556,
+      "grad_norm": 4.4706924859775174,
+      "learning_rate": 3.580178787966376e-06,
+      "loss": 0.6717,
+      "step": 4643
+    },
+    {
+      "epoch": 0.37716234873710713,
+      "grad_norm": 4.900856116172455,
+      "learning_rate": 3.5795856810499085e-06,
+      "loss": 0.5729,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3772435637131487,
+      "grad_norm": 3.182583890174541,
+      "learning_rate": 3.5789924994318267e-06,
+      "loss": 0.6078,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3773247786891903,
+      "grad_norm": 3.865947434731064,
+      "learning_rate": 3.578399243153174e-06,
+      "loss": 0.6606,
+      "step": 4646
+    },
+    {
+      "epoch": 0.37740599366523186,
+      "grad_norm": 3.349782633519335,
+      "learning_rate": 3.5778059122550007e-06,
+      "loss": 0.544,
+      "step": 4647
+    },
+    {
+      "epoch": 0.37748720864127344,
+      "grad_norm": 4.885217069177799,
+      "learning_rate": 3.5772125067783624e-06,
+      "loss": 0.5011,
+      "step": 4648
+    },
+    {
+      "epoch": 0.377568423617315,
+      "grad_norm": 4.590378597766298,
+      "learning_rate": 3.57661902676432e-06,
+      "loss": 0.5884,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3776496385933566,
+      "grad_norm": 4.527590654884785,
+      "learning_rate": 3.576025472253939e-06,
+      "loss": 0.4238,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3777308535693982,
+      "grad_norm": 4.451704643586669,
+      "learning_rate": 3.5754318432882907e-06,
+      "loss": 0.4485,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3778120685454398,
+      "grad_norm": 5.927309052944885,
+      "learning_rate": 3.5748381399084492e-06,
+      "loss": 0.5697,
+      "step": 4652
+    },
+    {
+      "epoch": 0.3778932835214814,
+      "grad_norm": 5.139872795558716,
+      "learning_rate": 3.5742443621554977e-06,
+      "loss": 0.5761,
+      "step": 4653
+    },
+    {
+      "epoch": 0.37797449849752296,
+      "grad_norm": 6.387127931906058,
+      "learning_rate": 3.5736505100705223e-06,
+      "loss": 0.6974,
+      "step": 4654
+    },
+    {
+      "epoch": 0.37805571347356454,
+      "grad_norm": 5.810815572808849,
+      "learning_rate": 3.573056583694612e-06,
+      "loss": 0.6572,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3781369284496061,
+      "grad_norm": 4.719558057872907,
+      "learning_rate": 3.5724625830688667e-06,
+      "loss": 0.5495,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3782181434256477,
+      "grad_norm": 6.208002944055565,
+      "learning_rate": 3.571868508234386e-06,
+      "loss": 0.4823,
+      "step": 4657
+    },
+    {
+      "epoch": 0.37829935840168927,
+      "grad_norm": 3.68160438978963,
+      "learning_rate": 3.5712743592322775e-06,
+      "loss": 0.504,
+      "step": 4658
+    },
+    {
+      "epoch": 0.37838057337773084,
+      "grad_norm": 4.064268236543746,
+      "learning_rate": 3.570680136103653e-06,
+      "loss": 0.571,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3784617883537724,
+      "grad_norm": 7.721913069819413,
+      "learning_rate": 3.57008583888963e-06,
+      "loss": 0.4338,
+      "step": 4660
+    },
+    {
+      "epoch": 0.378543003329814,
+      "grad_norm": 8.025619823408007,
+      "learning_rate": 3.569491467631329e-06,
+      "loss": 0.4907,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3786242183058556,
+      "grad_norm": 4.4108267729746204,
+      "learning_rate": 3.568897022369879e-06,
+      "loss": 0.6222,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3787054332818972,
+      "grad_norm": 15.163395858330826,
+      "learning_rate": 3.568302503146413e-06,
+      "loss": 0.5193,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3787866482579388,
+      "grad_norm": 3.2649054015739525,
+      "learning_rate": 3.567707910002068e-06,
+      "loss": 0.5473,
+      "step": 4664
+    },
+    {
+      "epoch": 0.37886786323398036,
+      "grad_norm": 5.082393091909739,
+      "learning_rate": 3.5671132429779847e-06,
+      "loss": 0.4679,
+      "step": 4665
+    },
+    {
+      "epoch": 0.37894907821002194,
+      "grad_norm": 7.079599784559318,
+      "learning_rate": 3.566518502115314e-06,
+      "loss": 0.501,
+      "step": 4666
+    },
+    {
+      "epoch": 0.3790302931860635,
+      "grad_norm": 6.472514959171711,
+      "learning_rate": 3.565923687455207e-06,
+      "loss": 0.6414,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3791115081621051,
+      "grad_norm": 6.442050896136885,
+      "learning_rate": 3.565328799038822e-06,
+      "loss": 0.4772,
+      "step": 4668
+    },
+    {
+      "epoch": 0.37919272313814667,
+      "grad_norm": 6.208659634243696,
+      "learning_rate": 3.5647338369073225e-06,
+      "loss": 0.6315,
+      "step": 4669
+    },
+    {
+      "epoch": 0.37927393811418825,
+      "grad_norm": 4.322414723790994,
+      "learning_rate": 3.5641388011018764e-06,
+      "loss": 0.4861,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3793551530902298,
+      "grad_norm": 6.68918224709947,
+      "learning_rate": 3.563543691663657e-06,
+      "loss": 0.6495,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3794363680662714,
+      "grad_norm": 6.412987095133581,
+      "learning_rate": 3.5629485086338432e-06,
+      "loss": 0.7778,
+      "step": 4672
+    },
+    {
+      "epoch": 0.379517583042313,
+      "grad_norm": 4.2253670535048204,
+      "learning_rate": 3.562353252053618e-06,
+      "loss": 0.4884,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3795987980183546,
+      "grad_norm": 5.325608575778679,
+      "learning_rate": 3.56175792196417e-06,
+      "loss": 0.4152,
+      "step": 4674
+    },
+    {
+      "epoch": 0.3796800129943962,
+      "grad_norm": 5.859119696052518,
+      "learning_rate": 3.561162518406693e-06,
+      "loss": 0.5208,
+      "step": 4675
+    },
+    {
+      "epoch": 0.37976122797043776,
+      "grad_norm": 3.929599806942015,
+      "learning_rate": 3.5605670414223866e-06,
+      "loss": 0.6021,
+      "step": 4676
+    },
+    {
+      "epoch": 0.37984244294647934,
+      "grad_norm": 10.494504954752514,
+      "learning_rate": 3.559971491052453e-06,
+      "loss": 0.6292,
+      "step": 4677
+    },
+    {
+      "epoch": 0.3799236579225209,
+      "grad_norm": 4.64718386046636,
+      "learning_rate": 3.559375867338103e-06,
+      "loss": 0.4315,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3800048728985625,
+      "grad_norm": 4.74521085436876,
+      "learning_rate": 3.5587801703205486e-06,
+      "loss": 0.4433,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3800860878746041,
+      "grad_norm": 9.15577004594742,
+      "learning_rate": 3.558184400041011e-06,
+      "loss": 0.5353,
+      "step": 4680
+    },
+    {
+      "epoch": 0.38016730285064565,
+      "grad_norm": 4.324557986539743,
+      "learning_rate": 3.557588556540712e-06,
+      "loss": 0.6834,
+      "step": 4681
+    },
+    {
+      "epoch": 0.3802485178266872,
+      "grad_norm": 4.785865260255061,
+      "learning_rate": 3.556992639860883e-06,
+      "loss": 0.487,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3803297328027288,
+      "grad_norm": 7.647172576935408,
+      "learning_rate": 3.5563966500427577e-06,
+      "loss": 0.4949,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3804109477787704,
+      "grad_norm": 6.176240715797557,
+      "learning_rate": 3.555800587127574e-06,
+      "loss": 0.6732,
+      "step": 4684
+    },
+    {
+      "epoch": 0.380492162754812,
+      "grad_norm": 2.9442581188680306,
+      "learning_rate": 3.5552044511565783e-06,
+      "loss": 0.6017,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3805733777308536,
+      "grad_norm": 3.6958273548877916,
+      "learning_rate": 3.554608242171019e-06,
+      "loss": 0.5588,
+      "step": 4686
+    },
+    {
+      "epoch": 0.38065459270689517,
+      "grad_norm": 4.174711443180175,
+      "learning_rate": 3.554011960212151e-06,
+      "loss": 0.4675,
+      "step": 4687
+    },
+    {
+      "epoch": 0.38073580768293674,
+      "grad_norm": 4.393276038211761,
+      "learning_rate": 3.5534156053212333e-06,
+      "loss": 0.5403,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3808170226589783,
+      "grad_norm": 4.053882314908021,
+      "learning_rate": 3.5528191775395304e-06,
+      "loss": 0.5421,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3808982376350199,
+      "grad_norm": 2.5408049268953437,
+      "learning_rate": 3.552222676908313e-06,
+      "loss": 0.4942,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3809794526110615,
+      "grad_norm": 10.13235114058035,
+      "learning_rate": 3.5516261034688547e-06,
+      "loss": 0.5421,
+      "step": 4691
+    },
+    {
+      "epoch": 0.38106066758710305,
+      "grad_norm": 7.384145049649841,
+      "learning_rate": 3.5510294572624358e-06,
+      "loss": 0.4957,
+      "step": 4692
+    },
+    {
+      "epoch": 0.38114188256314463,
+      "grad_norm": 11.958189835069302,
+      "learning_rate": 3.5504327383303415e-06,
+      "loss": 0.5927,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3812230975391862,
+      "grad_norm": 3.403027900822607,
+      "learning_rate": 3.549835946713861e-06,
+      "loss": 0.5608,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3813043125152278,
+      "grad_norm": 3.057095877975725,
+      "learning_rate": 3.5492390824542887e-06,
+      "loss": 0.5047,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3813855274912694,
+      "grad_norm": 6.33274087310348,
+      "learning_rate": 3.5486421455929253e-06,
+      "loss": 0.4971,
+      "step": 4696
+    },
+    {
+      "epoch": 0.381466742467311,
+      "grad_norm": 4.1556831093555555,
+      "learning_rate": 3.5480451361710744e-06,
+      "loss": 0.6343,
+      "step": 4697
+    },
+    {
+      "epoch": 0.38154795744335257,
+      "grad_norm": 3.1197068090047786,
+      "learning_rate": 3.5474480542300475e-06,
+      "loss": 0.6561,
+      "step": 4698
+    },
+    {
+      "epoch": 0.38162917241939415,
+      "grad_norm": 4.160059988171047,
+      "learning_rate": 3.5468508998111596e-06,
+      "loss": 0.6627,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3817103873954357,
+      "grad_norm": 5.478519363239558,
+      "learning_rate": 3.5462536729557284e-06,
+      "loss": 0.6216,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3817916023714773,
+      "grad_norm": 6.8557884315946795,
+      "learning_rate": 3.545656373705081e-06,
+      "loss": 0.6758,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3818728173475189,
+      "grad_norm": 4.172187135110137,
+      "learning_rate": 3.5450590021005465e-06,
+      "loss": 0.4609,
+      "step": 4702
+    },
+    {
+      "epoch": 0.38195403232356046,
+      "grad_norm": 7.593654425254116,
+      "learning_rate": 3.5444615581834595e-06,
+      "loss": 0.5303,
+      "step": 4703
+    },
+    {
+      "epoch": 0.38203524729960203,
+      "grad_norm": 2.9965381248103466,
+      "learning_rate": 3.5438640419951608e-06,
+      "loss": 0.6348,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3821164622756436,
+      "grad_norm": 3.965887732610272,
+      "learning_rate": 3.5432664535769952e-06,
+      "loss": 0.7041,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3821976772516852,
+      "grad_norm": 10.865736295673367,
+      "learning_rate": 3.5426687929703117e-06,
+      "loss": 0.5335,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3822788922277268,
+      "grad_norm": 6.090855516494487,
+      "learning_rate": 3.5420710602164665e-06,
+      "loss": 0.6845,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3823601072037684,
+      "grad_norm": 3.356332999735362,
+      "learning_rate": 3.5414732553568194e-06,
+      "loss": 0.5373,
+      "step": 4708
+    },
+    {
+      "epoch": 0.38244132217981,
+      "grad_norm": 6.469401787137639,
+      "learning_rate": 3.5408753784327344e-06,
+      "loss": 0.5448,
+      "step": 4709
+    },
+    {
+      "epoch": 0.38252253715585155,
+      "grad_norm": 5.575865042285411,
+      "learning_rate": 3.540277429485582e-06,
+      "loss": 0.474,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3826037521318931,
+      "grad_norm": 5.15216848497321,
+      "learning_rate": 3.539679408556737e-06,
+      "loss": 0.4949,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3826849671079347,
+      "grad_norm": 4.447503975145806,
+      "learning_rate": 3.5390813156875792e-06,
+      "loss": 0.4684,
+      "step": 4712
+    },
+    {
+      "epoch": 0.3827661820839763,
+      "grad_norm": 6.602421808698497,
+      "learning_rate": 3.538483150919494e-06,
+      "loss": 0.4992,
+      "step": 4713
+    },
+    {
+      "epoch": 0.38284739706001786,
+      "grad_norm": 9.959981744936794,
+      "learning_rate": 3.537884914293871e-06,
+      "loss": 0.5644,
+      "step": 4714
+    },
+    {
+      "epoch": 0.38292861203605943,
+      "grad_norm": 4.773698731849422,
+      "learning_rate": 3.537286605852105e-06,
+      "loss": 0.5316,
+      "step": 4715
+    },
+    {
+      "epoch": 0.383009827012101,
+      "grad_norm": 5.852883004242007,
+      "learning_rate": 3.536688225635595e-06,
+      "loss": 0.4393,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3830910419881426,
+      "grad_norm": 3.5878863623629442,
+      "learning_rate": 3.5360897736857464e-06,
+      "loss": 0.7132,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3831722569641842,
+      "grad_norm": 5.620156517899439,
+      "learning_rate": 3.5354912500439696e-06,
+      "loss": 0.5248,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3832534719402258,
+      "grad_norm": 7.422807956148856,
+      "learning_rate": 3.5348926547516783e-06,
+      "loss": 0.5087,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3833346869162674,
+      "grad_norm": 5.38906725850609,
+      "learning_rate": 3.534293987850291e-06,
+      "loss": 0.7483,
+      "step": 4720
+    },
+    {
+      "epoch": 0.38341590189230895,
+      "grad_norm": 4.767084586004944,
+      "learning_rate": 3.5336952493812353e-06,
+      "loss": 0.6177,
+      "step": 4721
+    },
+    {
+      "epoch": 0.38349711686835053,
+      "grad_norm": 4.383751010173588,
+      "learning_rate": 3.533096439385939e-06,
+      "loss": 0.5982,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3835783318443921,
+      "grad_norm": 18.60580065511833,
+      "learning_rate": 3.532497557905836e-06,
+      "loss": 0.6254,
+      "step": 4723
+    },
+    {
+      "epoch": 0.3836595468204337,
+      "grad_norm": 4.724647861890121,
+      "learning_rate": 3.531898604982367e-06,
+      "loss": 0.4813,
+      "step": 4724
+    },
+    {
+      "epoch": 0.38374076179647526,
+      "grad_norm": 5.525218151959553,
+      "learning_rate": 3.5312995806569754e-06,
+      "loss": 0.604,
+      "step": 4725
+    },
+    {
+      "epoch": 0.38382197677251684,
+      "grad_norm": 4.419767437379395,
+      "learning_rate": 3.5307004849711114e-06,
+      "loss": 0.6971,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3839031917485584,
+      "grad_norm": 4.847805495392348,
+      "learning_rate": 3.530101317966228e-06,
+      "loss": 0.4638,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3839844067246,
+      "grad_norm": 4.89939462212519,
+      "learning_rate": 3.5295020796837854e-06,
+      "loss": 0.503,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3840656217006416,
+      "grad_norm": 4.244557296775247,
+      "learning_rate": 3.528902770165248e-06,
+      "loss": 0.5719,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3841468366766832,
+      "grad_norm": 8.538711454427421,
+      "learning_rate": 3.5283033894520836e-06,
+      "loss": 0.4718,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3842280516527248,
+      "grad_norm": 3.6208666339316995,
+      "learning_rate": 3.5277039375857677e-06,
+      "loss": 0.5757,
+      "step": 4731
+    },
+    {
+      "epoch": 0.38430926662876636,
+      "grad_norm": 6.751106985384788,
+      "learning_rate": 3.5271044146077773e-06,
+      "loss": 0.5931,
+      "step": 4732
+    },
+    {
+      "epoch": 0.38439048160480793,
+      "grad_norm": 7.411379001602781,
+      "learning_rate": 3.5265048205595976e-06,
+      "loss": 0.4918,
+      "step": 4733
+    },
+    {
+      "epoch": 0.3844716965808495,
+      "grad_norm": 4.0826091607694455,
+      "learning_rate": 3.5259051554827175e-06,
+      "loss": 0.4503,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3845529115568911,
+      "grad_norm": 4.479671138778023,
+      "learning_rate": 3.5253054194186297e-06,
+      "loss": 0.5551,
+      "step": 4735
+    },
+    {
+      "epoch": 0.38463412653293266,
+      "grad_norm": 6.005050130154772,
+      "learning_rate": 3.524705612408833e-06,
+      "loss": 0.6683,
+      "step": 4736
+    },
+    {
+      "epoch": 0.38471534150897424,
+      "grad_norm": 5.171375305373519,
+      "learning_rate": 3.5241057344948317e-06,
+      "loss": 0.5399,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3847965564850158,
+      "grad_norm": 10.084203556597553,
+      "learning_rate": 3.523505785718133e-06,
+      "loss": 0.6731,
+      "step": 4738
+    },
+    {
+      "epoch": 0.3848777714610574,
+      "grad_norm": 6.878409019634097,
+      "learning_rate": 3.5229057661202513e-06,
+      "loss": 0.4689,
+      "step": 4739
+    },
+    {
+      "epoch": 0.384958986437099,
+      "grad_norm": 5.680734133112059,
+      "learning_rate": 3.5223056757427044e-06,
+      "loss": 0.5154,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3850402014131406,
+      "grad_norm": 2.861081172022858,
+      "learning_rate": 3.5217055146270144e-06,
+      "loss": 0.5012,
+      "step": 4741
+    },
+    {
+      "epoch": 0.3851214163891822,
+      "grad_norm": 7.512509192657636,
+      "learning_rate": 3.5211052828147114e-06,
+      "loss": 0.5743,
+      "step": 4742
+    },
+    {
+      "epoch": 0.38520263136522376,
+      "grad_norm": 7.512493844062562,
+      "learning_rate": 3.5205049803473257e-06,
+      "loss": 0.5381,
+      "step": 4743
+    },
+    {
+      "epoch": 0.38528384634126533,
+      "grad_norm": 4.916065054864969,
+      "learning_rate": 3.5199046072663968e-06,
+      "loss": 0.5969,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3853650613173069,
+      "grad_norm": 5.767505645700775,
+      "learning_rate": 3.5193041636134673e-06,
+      "loss": 0.896,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3854462762933485,
+      "grad_norm": 3.7412254239286566,
+      "learning_rate": 3.518703649430083e-06,
+      "loss": 0.6188,
+      "step": 4746
+    },
+    {
+      "epoch": 0.38552749126939007,
+      "grad_norm": 5.626427073356312,
+      "learning_rate": 3.518103064757798e-06,
+      "loss": 0.4346,
+      "step": 4747
+    },
+    {
+      "epoch": 0.38560870624543164,
+      "grad_norm": 5.8362139577169,
+      "learning_rate": 3.51750240963817e-06,
+      "loss": 0.5342,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3856899212214732,
+      "grad_norm": 4.982453547043835,
+      "learning_rate": 3.516901684112759e-06,
+      "loss": 0.6567,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3857711361975148,
+      "grad_norm": 4.093696003385388,
+      "learning_rate": 3.5163008882231347e-06,
+      "loss": 0.5791,
+      "step": 4750
+    },
+    {
+      "epoch": 0.38585235117355643,
+      "grad_norm": 5.642536053815455,
+      "learning_rate": 3.5157000220108674e-06,
+      "loss": 0.4339,
+      "step": 4751
+    },
+    {
+      "epoch": 0.385933566149598,
+      "grad_norm": 4.2134737972824,
+      "learning_rate": 3.5150990855175337e-06,
+      "loss": 0.6173,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3860147811256396,
+      "grad_norm": 7.7835339423428405,
+      "learning_rate": 3.5144980787847155e-06,
+      "loss": 0.7697,
+      "step": 4753
+    },
+    {
+      "epoch": 0.38609599610168116,
+      "grad_norm": 4.445884072963403,
+      "learning_rate": 3.5138970018539998e-06,
+      "loss": 0.4681,
+      "step": 4754
+    },
+    {
+      "epoch": 0.38617721107772274,
+      "grad_norm": 7.397822019738059,
+      "learning_rate": 3.513295854766977e-06,
+      "loss": 0.4439,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3862584260537643,
+      "grad_norm": 5.83757001159602,
+      "learning_rate": 3.5126946375652443e-06,
+      "loss": 0.5881,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3863396410298059,
+      "grad_norm": 5.649456450527899,
+      "learning_rate": 3.512093350290402e-06,
+      "loss": 0.4453,
+      "step": 4757
+    },
+    {
+      "epoch": 0.38642085600584747,
+      "grad_norm": 4.735284890427706,
+      "learning_rate": 3.511491992984057e-06,
+      "loss": 0.6149,
+      "step": 4758
+    },
+    {
+      "epoch": 0.38650207098188905,
+      "grad_norm": 9.984154759576063,
+      "learning_rate": 3.510890565687818e-06,
+      "loss": 0.4296,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3865832859579306,
+      "grad_norm": 5.043149999950322,
+      "learning_rate": 3.5102890684433026e-06,
+      "loss": 0.5894,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3866645009339722,
+      "grad_norm": 4.065716569234592,
+      "learning_rate": 3.509687501292132e-06,
+      "loss": 0.3679,
+      "step": 4761
+    },
+    {
+      "epoch": 0.38674571591001383,
+      "grad_norm": 5.553031924098756,
+      "learning_rate": 3.5090858642759273e-06,
+      "loss": 0.4891,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3868269308860554,
+      "grad_norm": 9.139353955383072,
+      "learning_rate": 3.5084841574363227e-06,
+      "loss": 0.5457,
+      "step": 4763
+    },
+    {
+      "epoch": 0.386908145862097,
+      "grad_norm": 5.211077508553389,
+      "learning_rate": 3.507882380814952e-06,
+      "loss": 0.7012,
+      "step": 4764
+    },
+    {
+      "epoch": 0.38698936083813856,
+      "grad_norm": 3.9781818293321027,
+      "learning_rate": 3.507280534453454e-06,
+      "loss": 0.5159,
+      "step": 4765
+    },
+    {
+      "epoch": 0.38707057581418014,
+      "grad_norm": 3.7855417874921455,
+      "learning_rate": 3.5066786183934743e-06,
+      "loss": 0.409,
+      "step": 4766
+    },
+    {
+      "epoch": 0.3871517907902217,
+      "grad_norm": 5.51441031251894,
+      "learning_rate": 3.5060766326766626e-06,
+      "loss": 0.5835,
+      "step": 4767
+    },
+    {
+      "epoch": 0.3872330057662633,
+      "grad_norm": 5.729949936654051,
+      "learning_rate": 3.505474577344672e-06,
+      "loss": 0.4933,
+      "step": 4768
+    },
+    {
+      "epoch": 0.38731422074230487,
+      "grad_norm": 4.617510877977197,
+      "learning_rate": 3.504872452439162e-06,
+      "loss": 0.6594,
+      "step": 4769
+    },
+    {
+      "epoch": 0.38739543571834645,
+      "grad_norm": 4.618436696346881,
+      "learning_rate": 3.504270258001796e-06,
+      "loss": 0.6615,
+      "step": 4770
+    },
+    {
+      "epoch": 0.387476650694388,
+      "grad_norm": 3.771460379690749,
+      "learning_rate": 3.503667994074244e-06,
+      "loss": 0.2883,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3875578656704296,
+      "grad_norm": 6.252805853476455,
+      "learning_rate": 3.5030656606981783e-06,
+      "loss": 0.4846,
+      "step": 4772
+    },
+    {
+      "epoch": 0.38763908064647123,
+      "grad_norm": 6.604659609809073,
+      "learning_rate": 3.5024632579152775e-06,
+      "loss": 0.4682,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3877202956225128,
+      "grad_norm": 8.26513746625276,
+      "learning_rate": 3.501860785767225e-06,
+      "loss": 0.5166,
+      "step": 4774
+    },
+    {
+      "epoch": 0.3878015105985544,
+      "grad_norm": 4.4406417232924085,
+      "learning_rate": 3.5012582442957077e-06,
+      "loss": 0.7179,
+      "step": 4775
+    },
+    {
+      "epoch": 0.38788272557459597,
+      "grad_norm": 13.64277952343109,
+      "learning_rate": 3.5006556335424197e-06,
+      "loss": 0.3686,
+      "step": 4776
+    },
+    {
+      "epoch": 0.38796394055063754,
+      "grad_norm": 4.226503480824964,
+      "learning_rate": 3.500052953549058e-06,
+      "loss": 0.6139,
+      "step": 4777
+    },
+    {
+      "epoch": 0.3880451555266791,
+      "grad_norm": 4.974191546797816,
+      "learning_rate": 3.4994502043573237e-06,
+      "loss": 0.7821,
+      "step": 4778
+    },
+    {
+      "epoch": 0.3881263705027207,
+      "grad_norm": 6.7453632874753815,
+      "learning_rate": 3.498847386008925e-06,
+      "loss": 0.5492,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3882075854787623,
+      "grad_norm": 4.381360711445991,
+      "learning_rate": 3.4982444985455744e-06,
+      "loss": 0.5192,
+      "step": 4780
+    },
+    {
+      "epoch": 0.38828880045480385,
+      "grad_norm": 4.91701938364453,
+      "learning_rate": 3.4976415420089865e-06,
+      "loss": 0.5407,
+      "step": 4781
+    },
+    {
+      "epoch": 0.38837001543084543,
+      "grad_norm": 5.048580518545317,
+      "learning_rate": 3.4970385164408837e-06,
+      "loss": 0.466,
+      "step": 4782
+    },
+    {
+      "epoch": 0.388451230406887,
+      "grad_norm": 4.973082626749845,
+      "learning_rate": 3.496435421882994e-06,
+      "loss": 0.7516,
+      "step": 4783
+    },
+    {
+      "epoch": 0.38853244538292864,
+      "grad_norm": 5.743564392198817,
+      "learning_rate": 3.4958322583770453e-06,
+      "loss": 0.593,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3886136603589702,
+      "grad_norm": 4.979299976934781,
+      "learning_rate": 3.495229025964775e-06,
+      "loss": 0.4979,
+      "step": 4785
+    },
+    {
+      "epoch": 0.3886948753350118,
+      "grad_norm": 12.613122878358682,
+      "learning_rate": 3.494625724687923e-06,
+      "loss": 0.6043,
+      "step": 4786
+    },
+    {
+      "epoch": 0.38877609031105337,
+      "grad_norm": 3.85614729618497,
+      "learning_rate": 3.494022354588235e-06,
+      "loss": 0.7142,
+      "step": 4787
+    },
+    {
+      "epoch": 0.38885730528709495,
+      "grad_norm": 6.046549551329485,
+      "learning_rate": 3.493418915707461e-06,
+      "loss": 0.4985,
+      "step": 4788
+    },
+    {
+      "epoch": 0.3889385202631365,
+      "grad_norm": 5.004952219579293,
+      "learning_rate": 3.4928154080873556e-06,
+      "loss": 0.7017,
+      "step": 4789
+    },
+    {
+      "epoch": 0.3890197352391781,
+      "grad_norm": 6.147470697065457,
+      "learning_rate": 3.4922118317696785e-06,
+      "loss": 0.5423,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3891009502152197,
+      "grad_norm": 6.876508550025277,
+      "learning_rate": 3.491608186796193e-06,
+      "loss": 0.6039,
+      "step": 4791
+    },
+    {
+      "epoch": 0.38918216519126125,
+      "grad_norm": 10.41135603101845,
+      "learning_rate": 3.49100447320867e-06,
+      "loss": 0.5495,
+      "step": 4792
+    },
+    {
+      "epoch": 0.38926338016730283,
+      "grad_norm": 3.362847757169651,
+      "learning_rate": 3.4904006910488824e-06,
+      "loss": 0.5267,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3893445951433444,
+      "grad_norm": 8.868539359623025,
+      "learning_rate": 3.489796840358608e-06,
+      "loss": 0.5027,
+      "step": 4794
+    },
+    {
+      "epoch": 0.38942581011938604,
+      "grad_norm": 3.869570039313511,
+      "learning_rate": 3.4891929211796303e-06,
+      "loss": 0.6388,
+      "step": 4795
+    },
+    {
+      "epoch": 0.3895070250954276,
+      "grad_norm": 18.46976613986865,
+      "learning_rate": 3.488588933553739e-06,
+      "loss": 0.7654,
+      "step": 4796
+    },
+    {
+      "epoch": 0.3895882400714692,
+      "grad_norm": 4.462040282578044,
+      "learning_rate": 3.4879848775227243e-06,
+      "loss": 0.5233,
+      "step": 4797
+    },
+    {
+      "epoch": 0.38966945504751077,
+      "grad_norm": 16.476898959715022,
+      "learning_rate": 3.487380753128385e-06,
+      "loss": 0.463,
+      "step": 4798
+    },
+    {
+      "epoch": 0.38975067002355235,
+      "grad_norm": 4.641132700498576,
+      "learning_rate": 3.4867765604125236e-06,
+      "loss": 0.3472,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3898318849995939,
+      "grad_norm": 4.809948864229084,
+      "learning_rate": 3.4861722994169466e-06,
+      "loss": 0.4521,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3899130999756355,
+      "grad_norm": 4.6897556369620625,
+      "learning_rate": 3.485567970183466e-06,
+      "loss": 0.5308,
+      "step": 4801
+    },
+    {
+      "epoch": 0.3899943149516771,
+      "grad_norm": 4.712889214741922,
+      "learning_rate": 3.484963572753898e-06,
+      "loss": 0.4595,
+      "step": 4802
+    },
+    {
+      "epoch": 0.39007552992771866,
+      "grad_norm": 3.914385123227235,
+      "learning_rate": 3.4843591071700627e-06,
+      "loss": 0.6773,
+      "step": 4803
+    },
+    {
+      "epoch": 0.39015674490376023,
+      "grad_norm": 3.7046943355774147,
+      "learning_rate": 3.4837545734737877e-06,
+      "loss": 0.4701,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3902379598798018,
+      "grad_norm": 5.273455209488689,
+      "learning_rate": 3.483149971706902e-06,
+      "loss": 0.5245,
+      "step": 4805
+    },
+    {
+      "epoch": 0.39031917485584344,
+      "grad_norm": 5.259747071201792,
+      "learning_rate": 3.482545301911242e-06,
+      "loss": 0.58,
+      "step": 4806
+    },
+    {
+      "epoch": 0.390400389831885,
+      "grad_norm": 3.865781563186537,
+      "learning_rate": 3.4819405641286476e-06,
+      "loss": 0.5728,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3904816048079266,
+      "grad_norm": 4.430734119933856,
+      "learning_rate": 3.481335758400962e-06,
+      "loss": 0.406,
+      "step": 4808
+    },
+    {
+      "epoch": 0.3905628197839682,
+      "grad_norm": 3.1468197398544295,
+      "learning_rate": 3.480730884770036e-06,
+      "loss": 0.5959,
+      "step": 4809
+    },
+    {
+      "epoch": 0.39064403476000975,
+      "grad_norm": 4.928724212153909,
+      "learning_rate": 3.4801259432777236e-06,
+      "loss": 0.534,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3907252497360513,
+      "grad_norm": 5.338771481365462,
+      "learning_rate": 3.479520933965882e-06,
+      "loss": 0.6451,
+      "step": 4811
+    },
+    {
+      "epoch": 0.3908064647120929,
+      "grad_norm": 4.865233384961484,
+      "learning_rate": 3.4789158568763777e-06,
+      "loss": 0.6479,
+      "step": 4812
+    },
+    {
+      "epoch": 0.3908876796881345,
+      "grad_norm": 6.708639239866096,
+      "learning_rate": 3.4783107120510758e-06,
+      "loss": 0.4542,
+      "step": 4813
+    },
+    {
+      "epoch": 0.39096889466417606,
+      "grad_norm": 5.797570016494007,
+      "learning_rate": 3.4777054995318493e-06,
+      "loss": 0.6198,
+      "step": 4814
+    },
+    {
+      "epoch": 0.39105010964021764,
+      "grad_norm": 3.723719927731895,
+      "learning_rate": 3.4771002193605783e-06,
+      "loss": 0.5544,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3911313246162592,
+      "grad_norm": 4.734496419671341,
+      "learning_rate": 3.4764948715791425e-06,
+      "loss": 0.5319,
+      "step": 4816
+    },
+    {
+      "epoch": 0.39121253959230085,
+      "grad_norm": 3.483732371245711,
+      "learning_rate": 3.47588945622943e-06,
+      "loss": 0.6067,
+      "step": 4817
+    },
+    {
+      "epoch": 0.3912937545683424,
+      "grad_norm": 3.7596387481376143,
+      "learning_rate": 3.4752839733533315e-06,
+      "loss": 0.7024,
+      "step": 4818
+    },
+    {
+      "epoch": 0.391374969544384,
+      "grad_norm": 6.235595740178511,
+      "learning_rate": 3.4746784229927445e-06,
+      "loss": 0.5705,
+      "step": 4819
+    },
+    {
+      "epoch": 0.3914561845204256,
+      "grad_norm": 4.914796686357341,
+      "learning_rate": 3.4740728051895683e-06,
+      "loss": 0.4127,
+      "step": 4820
+    },
+    {
+      "epoch": 0.39153739949646715,
+      "grad_norm": 6.037035106034122,
+      "learning_rate": 3.4734671199857093e-06,
+      "loss": 0.461,
+      "step": 4821
+    },
+    {
+      "epoch": 0.39161861447250873,
+      "grad_norm": 4.084676397060007,
+      "learning_rate": 3.4728613674230777e-06,
+      "loss": 0.4533,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3916998294485503,
+      "grad_norm": 5.820986362019269,
+      "learning_rate": 3.472255547543589e-06,
+      "loss": 0.4948,
+      "step": 4823
+    },
+    {
+      "epoch": 0.3917810444245919,
+      "grad_norm": 6.625284347900165,
+      "learning_rate": 3.4716496603891605e-06,
+      "loss": 0.6856,
+      "step": 4824
+    },
+    {
+      "epoch": 0.39186225940063346,
+      "grad_norm": 4.751431998722198,
+      "learning_rate": 3.471043706001719e-06,
+      "loss": 0.5442,
+      "step": 4825
+    },
+    {
+      "epoch": 0.39194347437667504,
+      "grad_norm": 3.789894887233057,
+      "learning_rate": 3.4704376844231922e-06,
+      "loss": 0.5568,
+      "step": 4826
+    },
+    {
+      "epoch": 0.3920246893527166,
+      "grad_norm": 3.666796443871811,
+      "learning_rate": 3.4698315956955125e-06,
+      "loss": 0.6599,
+      "step": 4827
+    },
+    {
+      "epoch": 0.39210590432875825,
+      "grad_norm": 3.6434242436779973,
+      "learning_rate": 3.46922543986062e-06,
+      "loss": 0.4903,
+      "step": 4828
+    },
+    {
+      "epoch": 0.3921871193047998,
+      "grad_norm": 7.08366419413291,
+      "learning_rate": 3.468619216960457e-06,
+      "loss": 0.6005,
+      "step": 4829
+    },
+    {
+      "epoch": 0.3922683342808414,
+      "grad_norm": 4.495677796038025,
+      "learning_rate": 3.46801292703697e-06,
+      "loss": 0.5686,
+      "step": 4830
+    },
+    {
+      "epoch": 0.392349549256883,
+      "grad_norm": 6.22783462766733,
+      "learning_rate": 3.467406570132112e-06,
+      "loss": 0.402,
+      "step": 4831
+    },
+    {
+      "epoch": 0.39243076423292456,
+      "grad_norm": 6.8479413586138165,
+      "learning_rate": 3.4668001462878386e-06,
+      "loss": 0.4031,
+      "step": 4832
+    },
+    {
+      "epoch": 0.39251197920896613,
+      "grad_norm": 11.848480365877366,
+      "learning_rate": 3.466193655546112e-06,
+      "loss": 0.4203,
+      "step": 4833
+    },
+    {
+      "epoch": 0.3925931941850077,
+      "grad_norm": 4.1344833900535205,
+      "learning_rate": 3.465587097948898e-06,
+      "loss": 0.4155,
+      "step": 4834
+    },
+    {
+      "epoch": 0.3926744091610493,
+      "grad_norm": 4.430449430971486,
+      "learning_rate": 3.4649804735381675e-06,
+      "loss": 0.6314,
+      "step": 4835
+    },
+    {
+      "epoch": 0.39275562413709086,
+      "grad_norm": 5.347837303662017,
+      "learning_rate": 3.4643737823558947e-06,
+      "loss": 0.5447,
+      "step": 4836
+    },
+    {
+      "epoch": 0.39283683911313244,
+      "grad_norm": 6.546391191124815,
+      "learning_rate": 3.463767024444061e-06,
+      "loss": 0.5371,
+      "step": 4837
+    },
+    {
+      "epoch": 0.392918054089174,
+      "grad_norm": 4.80456942652959,
+      "learning_rate": 3.4631601998446484e-06,
+      "loss": 0.638,
+      "step": 4838
+    },
+    {
+      "epoch": 0.39299926906521565,
+      "grad_norm": 5.280861827964533,
+      "learning_rate": 3.4625533085996495e-06,
+      "loss": 0.4545,
+      "step": 4839
+    },
+    {
+      "epoch": 0.3930804840412572,
+      "grad_norm": 6.5672307967606995,
+      "learning_rate": 3.4619463507510536e-06,
+      "loss": 0.4733,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3931616990172988,
+      "grad_norm": 4.308991992577792,
+      "learning_rate": 3.4613393263408625e-06,
+      "loss": 0.5039,
+      "step": 4841
+    },
+    {
+      "epoch": 0.3932429139933404,
+      "grad_norm": 5.784233850167831,
+      "learning_rate": 3.4607322354110785e-06,
+      "loss": 0.5651,
+      "step": 4842
+    },
+    {
+      "epoch": 0.39332412896938196,
+      "grad_norm": 5.463637822033822,
+      "learning_rate": 3.4601250780037064e-06,
+      "loss": 0.6328,
+      "step": 4843
+    },
+    {
+      "epoch": 0.39340534394542354,
+      "grad_norm": 6.116722144553743,
+      "learning_rate": 3.4595178541607616e-06,
+      "loss": 0.63,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3934865589214651,
+      "grad_norm": 5.568944318775315,
+      "learning_rate": 3.45891056392426e-06,
+      "loss": 0.4404,
+      "step": 4845
+    },
+    {
+      "epoch": 0.3935677738975067,
+      "grad_norm": 5.00032757592686,
+      "learning_rate": 3.4583032073362216e-06,
+      "loss": 0.4921,
+      "step": 4846
+    },
+    {
+      "epoch": 0.39364898887354827,
+      "grad_norm": 6.833348088070499,
+      "learning_rate": 3.4576957844386728e-06,
+      "loss": 0.5459,
+      "step": 4847
+    },
+    {
+      "epoch": 0.39373020384958984,
+      "grad_norm": 3.622191745550694,
+      "learning_rate": 3.4570882952736445e-06,
+      "loss": 0.5338,
+      "step": 4848
+    },
+    {
+      "epoch": 0.3938114188256314,
+      "grad_norm": 3.7561292429825617,
+      "learning_rate": 3.4564807398831716e-06,
+      "loss": 0.5432,
+      "step": 4849
+    },
+    {
+      "epoch": 0.39389263380167305,
+      "grad_norm": 6.7697920687749145,
+      "learning_rate": 3.4558731183092936e-06,
+      "loss": 0.4724,
+      "step": 4850
+    },
+    {
+      "epoch": 0.39397384877771463,
+      "grad_norm": 17.901118220763788,
+      "learning_rate": 3.4552654305940546e-06,
+      "loss": 0.4965,
+      "step": 4851
+    },
+    {
+      "epoch": 0.3940550637537562,
+      "grad_norm": 5.591762564408774,
+      "learning_rate": 3.4546576767795036e-06,
+      "loss": 0.4596,
+      "step": 4852
+    },
+    {
+      "epoch": 0.3941362787297978,
+      "grad_norm": 4.5991945379468975,
+      "learning_rate": 3.4540498569076935e-06,
+      "loss": 0.6449,
+      "step": 4853
+    },
+    {
+      "epoch": 0.39421749370583936,
+      "grad_norm": 4.012140667999568,
+      "learning_rate": 3.453441971020682e-06,
+      "loss": 0.5522,
+      "step": 4854
+    },
+    {
+      "epoch": 0.39429870868188094,
+      "grad_norm": 4.026685263806628,
+      "learning_rate": 3.4528340191605336e-06,
+      "loss": 0.5232,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3943799236579225,
+      "grad_norm": 3.9644370555022226,
+      "learning_rate": 3.452226001369313e-06,
+      "loss": 0.4068,
+      "step": 4856
+    },
+    {
+      "epoch": 0.3944611386339641,
+      "grad_norm": 4.955057245622184,
+      "learning_rate": 3.451617917689093e-06,
+      "loss": 0.5034,
+      "step": 4857
+    },
+    {
+      "epoch": 0.39454235361000567,
+      "grad_norm": 6.121308963517397,
+      "learning_rate": 3.4510097681619497e-06,
+      "loss": 0.5604,
+      "step": 4858
+    },
+    {
+      "epoch": 0.39462356858604725,
+      "grad_norm": 3.955911011778945,
+      "learning_rate": 3.4504015528299633e-06,
+      "loss": 0.742,
+      "step": 4859
+    },
+    {
+      "epoch": 0.3947047835620888,
+      "grad_norm": 3.416220952136702,
+      "learning_rate": 3.449793271735219e-06,
+      "loss": 0.4727,
+      "step": 4860
+    },
+    {
+      "epoch": 0.39478599853813046,
+      "grad_norm": 4.42715063532882,
+      "learning_rate": 3.4491849249198074e-06,
+      "loss": 0.6083,
+      "step": 4861
+    },
+    {
+      "epoch": 0.39486721351417203,
+      "grad_norm": 5.7756208575555466,
+      "learning_rate": 3.4485765124258223e-06,
+      "loss": 0.537,
+      "step": 4862
+    },
+    {
+      "epoch": 0.3949484284902136,
+      "grad_norm": 5.659261574392475,
+      "learning_rate": 3.4479680342953627e-06,
+      "loss": 0.4879,
+      "step": 4863
+    },
+    {
+      "epoch": 0.3950296434662552,
+      "grad_norm": 3.9294665757315625,
+      "learning_rate": 3.4473594905705326e-06,
+      "loss": 0.5727,
+      "step": 4864
+    },
+    {
+      "epoch": 0.39511085844229676,
+      "grad_norm": 4.426724017012553,
+      "learning_rate": 3.446750881293439e-06,
+      "loss": 0.5823,
+      "step": 4865
+    },
+    {
+      "epoch": 0.39519207341833834,
+      "grad_norm": 4.23549319138764,
+      "learning_rate": 3.4461422065061957e-06,
+      "loss": 0.594,
+      "step": 4866
+    },
+    {
+      "epoch": 0.3952732883943799,
+      "grad_norm": 4.069048549480606,
+      "learning_rate": 3.4455334662509186e-06,
+      "loss": 0.5666,
+      "step": 4867
+    },
+    {
+      "epoch": 0.3953545033704215,
+      "grad_norm": 9.920612867004065,
+      "learning_rate": 3.44492466056973e-06,
+      "loss": 0.5656,
+      "step": 4868
+    },
+    {
+      "epoch": 0.39543571834646307,
+      "grad_norm": 7.365805382978001,
+      "learning_rate": 3.4443157895047556e-06,
+      "loss": 0.5565,
+      "step": 4869
+    },
+    {
+      "epoch": 0.39551693332250465,
+      "grad_norm": 4.079639656340487,
+      "learning_rate": 3.4437068530981266e-06,
+      "loss": 0.5486,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3955981482985462,
+      "grad_norm": 5.1014095863826805,
+      "learning_rate": 3.4430978513919777e-06,
+      "loss": 0.4873,
+      "step": 4871
+    },
+    {
+      "epoch": 0.39567936327458786,
+      "grad_norm": 6.57666055283231,
+      "learning_rate": 3.4424887844284492e-06,
+      "loss": 0.5425,
+      "step": 4872
+    },
+    {
+      "epoch": 0.39576057825062944,
+      "grad_norm": 4.768382603290578,
+      "learning_rate": 3.4418796522496845e-06,
+      "loss": 0.3941,
+      "step": 4873
+    },
+    {
+      "epoch": 0.395841793226671,
+      "grad_norm": 6.114180863563064,
+      "learning_rate": 3.4412704548978326e-06,
+      "loss": 0.7732,
+      "step": 4874
+    },
+    {
+      "epoch": 0.3959230082027126,
+      "grad_norm": 3.93227698865023,
+      "learning_rate": 3.4406611924150468e-06,
+      "loss": 0.5043,
+      "step": 4875
+    },
+    {
+      "epoch": 0.39600422317875417,
+      "grad_norm": 5.0863186506106794,
+      "learning_rate": 3.440051864843485e-06,
+      "loss": 0.4199,
+      "step": 4876
+    },
+    {
+      "epoch": 0.39608543815479574,
+      "grad_norm": 8.186606675125352,
+      "learning_rate": 3.4394424722253095e-06,
+      "loss": 0.5594,
+      "step": 4877
+    },
+    {
+      "epoch": 0.3961666531308373,
+      "grad_norm": 3.1309990632230185,
+      "learning_rate": 3.4388330146026865e-06,
+      "loss": 0.515,
+      "step": 4878
+    },
+    {
+      "epoch": 0.3962478681068789,
+      "grad_norm": 3.0856049239044676,
+      "learning_rate": 3.438223492017787e-06,
+      "loss": 0.65,
+      "step": 4879
+    },
+    {
+      "epoch": 0.3963290830829205,
+      "grad_norm": 5.758345655080891,
+      "learning_rate": 3.4376139045127886e-06,
+      "loss": 0.5401,
+      "step": 4880
+    },
+    {
+      "epoch": 0.39641029805896205,
+      "grad_norm": 6.386676934781451,
+      "learning_rate": 3.4370042521298697e-06,
+      "loss": 0.63,
+      "step": 4881
+    },
+    {
+      "epoch": 0.39649151303500363,
+      "grad_norm": 4.289237032932299,
+      "learning_rate": 3.436394534911216e-06,
+      "loss": 0.4628,
+      "step": 4882
+    },
+    {
+      "epoch": 0.39657272801104526,
+      "grad_norm": 4.493221527603865,
+      "learning_rate": 3.4357847528990157e-06,
+      "loss": 0.4507,
+      "step": 4883
+    },
+    {
+      "epoch": 0.39665394298708684,
+      "grad_norm": 5.931364030127042,
+      "learning_rate": 3.4351749061354634e-06,
+      "loss": 0.4214,
+      "step": 4884
+    },
+    {
+      "epoch": 0.3967351579631284,
+      "grad_norm": 5.201076643919345,
+      "learning_rate": 3.4345649946627567e-06,
+      "loss": 0.4338,
+      "step": 4885
+    },
+    {
+      "epoch": 0.39681637293917,
+      "grad_norm": 6.755107310201019,
+      "learning_rate": 3.4339550185230985e-06,
+      "loss": 0.633,
+      "step": 4886
+    },
+    {
+      "epoch": 0.39689758791521157,
+      "grad_norm": 6.012302549365938,
+      "learning_rate": 3.4333449777586957e-06,
+      "loss": 0.4477,
+      "step": 4887
+    },
+    {
+      "epoch": 0.39697880289125315,
+      "grad_norm": 5.404553280600934,
+      "learning_rate": 3.432734872411761e-06,
+      "loss": 0.555,
+      "step": 4888
+    },
+    {
+      "epoch": 0.3970600178672947,
+      "grad_norm": 6.802478190345189,
+      "learning_rate": 3.4321247025245084e-06,
+      "loss": 0.56,
+      "step": 4889
+    },
+    {
+      "epoch": 0.3971412328433363,
+      "grad_norm": 4.629306992412694,
+      "learning_rate": 3.4315144681391604e-06,
+      "loss": 0.5331,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3972224478193779,
+      "grad_norm": 4.785707649092083,
+      "learning_rate": 3.430904169297941e-06,
+      "loss": 0.9165,
+      "step": 4891
+    },
+    {
+      "epoch": 0.39730366279541945,
+      "grad_norm": 7.873306295295412,
+      "learning_rate": 3.4302938060430794e-06,
+      "loss": 0.6152,
+      "step": 4892
+    },
+    {
+      "epoch": 0.39738487777146103,
+      "grad_norm": 3.4652132108801195,
+      "learning_rate": 3.429683378416811e-06,
+      "loss": 0.4936,
+      "step": 4893
+    },
+    {
+      "epoch": 0.39746609274750266,
+      "grad_norm": 4.301920992179896,
+      "learning_rate": 3.429072886461372e-06,
+      "loss": 0.5562,
+      "step": 4894
+    },
+    {
+      "epoch": 0.39754730772354424,
+      "grad_norm": 5.010865178574669,
+      "learning_rate": 3.428462330219007e-06,
+      "loss": 0.5153,
+      "step": 4895
+    },
+    {
+      "epoch": 0.3976285226995858,
+      "grad_norm": 6.077200433598698,
+      "learning_rate": 3.4278517097319617e-06,
+      "loss": 0.4894,
+      "step": 4896
+    },
+    {
+      "epoch": 0.3977097376756274,
+      "grad_norm": 4.201260945415446,
+      "learning_rate": 3.4272410250424893e-06,
+      "loss": 0.5466,
+      "step": 4897
+    },
+    {
+      "epoch": 0.39779095265166897,
+      "grad_norm": 4.064316394961935,
+      "learning_rate": 3.4266302761928453e-06,
+      "loss": 0.499,
+      "step": 4898
+    },
+    {
+      "epoch": 0.39787216762771055,
+      "grad_norm": 10.729804958408987,
+      "learning_rate": 3.4260194632252903e-06,
+      "loss": 0.6487,
+      "step": 4899
+    },
+    {
+      "epoch": 0.3979533826037521,
+      "grad_norm": 5.44190759303068,
+      "learning_rate": 3.4254085861820895e-06,
+      "loss": 0.4588,
+      "step": 4900
+    },
+    {
+      "epoch": 0.3980345975797937,
+      "grad_norm": 6.394017389790841,
+      "learning_rate": 3.424797645105512e-06,
+      "loss": 0.5658,
+      "step": 4901
+    },
+    {
+      "epoch": 0.3981158125558353,
+      "grad_norm": 5.684350734119721,
+      "learning_rate": 3.4241866400378315e-06,
+      "loss": 0.5158,
+      "step": 4902
+    },
+    {
+      "epoch": 0.39819702753187686,
+      "grad_norm": 3.9864422041154275,
+      "learning_rate": 3.423575571021327e-06,
+      "loss": 0.4298,
+      "step": 4903
+    },
+    {
+      "epoch": 0.39827824250791843,
+      "grad_norm": 5.696110565863003,
+      "learning_rate": 3.4229644380982817e-06,
+      "loss": 0.6485,
+      "step": 4904
+    },
+    {
+      "epoch": 0.39835945748396007,
+      "grad_norm": 8.085273365710513,
+      "learning_rate": 3.4223532413109807e-06,
+      "loss": 0.5311,
+      "step": 4905
+    },
+    {
+      "epoch": 0.39844067246000164,
+      "grad_norm": 21.38684299111223,
+      "learning_rate": 3.4217419807017177e-06,
+      "loss": 0.3467,
+      "step": 4906
+    },
+    {
+      "epoch": 0.3985218874360432,
+      "grad_norm": 6.879258693117489,
+      "learning_rate": 3.4211306563127876e-06,
+      "loss": 0.3502,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3986031024120848,
+      "grad_norm": 5.751298879545744,
+      "learning_rate": 3.4205192681864905e-06,
+      "loss": 0.5344,
+      "step": 4908
+    },
+    {
+      "epoch": 0.3986843173881264,
+      "grad_norm": 9.337940570841411,
+      "learning_rate": 3.4199078163651335e-06,
+      "loss": 0.5033,
+      "step": 4909
+    },
+    {
+      "epoch": 0.39876553236416795,
+      "grad_norm": 4.8306108732275925,
+      "learning_rate": 3.419296300891023e-06,
+      "loss": 0.518,
+      "step": 4910
+    },
+    {
+      "epoch": 0.39884674734020953,
+      "grad_norm": 7.933167629628174,
+      "learning_rate": 3.418684721806474e-06,
+      "loss": 0.4499,
+      "step": 4911
+    },
+    {
+      "epoch": 0.3989279623162511,
+      "grad_norm": 6.472677681508684,
+      "learning_rate": 3.418073079153804e-06,
+      "loss": 0.534,
+      "step": 4912
+    },
+    {
+      "epoch": 0.3990091772922927,
+      "grad_norm": 6.504622980241599,
+      "learning_rate": 3.4174613729753364e-06,
+      "loss": 0.569,
+      "step": 4913
+    },
+    {
+      "epoch": 0.39909039226833426,
+      "grad_norm": 7.727056330599826,
+      "learning_rate": 3.4168496033133968e-06,
+      "loss": 0.5061,
+      "step": 4914
+    },
+    {
+      "epoch": 0.39917160724437584,
+      "grad_norm": 4.697545157125827,
+      "learning_rate": 3.416237770210317e-06,
+      "loss": 0.4645,
+      "step": 4915
+    },
+    {
+      "epoch": 0.39925282222041747,
+      "grad_norm": 4.809263186617258,
+      "learning_rate": 3.415625873708433e-06,
+      "loss": 0.4595,
+      "step": 4916
+    },
+    {
+      "epoch": 0.39933403719645905,
+      "grad_norm": 10.22371082626964,
+      "learning_rate": 3.4150139138500843e-06,
+      "loss": 0.4436,
+      "step": 4917
+    },
+    {
+      "epoch": 0.3994152521725006,
+      "grad_norm": 5.9377177585980245,
+      "learning_rate": 3.4144018906776155e-06,
+      "loss": 0.5012,
+      "step": 4918
+    },
+    {
+      "epoch": 0.3994964671485422,
+      "grad_norm": 5.523225188806799,
+      "learning_rate": 3.413789804233375e-06,
+      "loss": 0.6356,
+      "step": 4919
+    },
+    {
+      "epoch": 0.3995776821245838,
+      "grad_norm": 4.922432155658762,
+      "learning_rate": 3.413177654559717e-06,
+      "loss": 0.4785,
+      "step": 4920
+    },
+    {
+      "epoch": 0.39965889710062535,
+      "grad_norm": 8.522209977514363,
+      "learning_rate": 3.4125654416989975e-06,
+      "loss": 0.4049,
+      "step": 4921
+    },
+    {
+      "epoch": 0.39974011207666693,
+      "grad_norm": 4.166534006928631,
+      "learning_rate": 3.411953165693579e-06,
+      "loss": 0.5051,
+      "step": 4922
+    },
+    {
+      "epoch": 0.3998213270527085,
+      "grad_norm": 7.556966122206325,
+      "learning_rate": 3.4113408265858282e-06,
+      "loss": 0.5134,
+      "step": 4923
+    },
+    {
+      "epoch": 0.3999025420287501,
+      "grad_norm": 4.46432321290014,
+      "learning_rate": 3.4107284244181154e-06,
+      "loss": 0.5773,
+      "step": 4924
+    },
+    {
+      "epoch": 0.39998375700479166,
+      "grad_norm": 3.3296185748575837,
+      "learning_rate": 3.4101159592328148e-06,
+      "loss": 0.492,
+      "step": 4925
+    },
+    {
+      "epoch": 0.40006497198083324,
+      "grad_norm": 4.808058470791979,
+      "learning_rate": 3.409503431072308e-06,
+      "loss": 0.4053,
+      "step": 4926
+    },
+    {
+      "epoch": 0.40014618695687487,
+      "grad_norm": 5.658421289602297,
+      "learning_rate": 3.408890839978976e-06,
+      "loss": 0.515,
+      "step": 4927
+    },
+    {
+      "epoch": 0.40022740193291645,
+      "grad_norm": 6.896229712559321,
+      "learning_rate": 3.4082781859952087e-06,
+      "loss": 0.5547,
+      "step": 4928
+    },
+    {
+      "epoch": 0.400308616908958,
+      "grad_norm": 5.824806238919424,
+      "learning_rate": 3.407665469163398e-06,
+      "loss": 0.6033,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4003898318849996,
+      "grad_norm": 4.135425915903601,
+      "learning_rate": 3.4070526895259403e-06,
+      "loss": 0.5083,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4004710468610412,
+      "grad_norm": 5.133129422170843,
+      "learning_rate": 3.4064398471252367e-06,
+      "loss": 0.5962,
+      "step": 4931
+    },
+    {
+      "epoch": 0.40055226183708276,
+      "grad_norm": 5.46292327134982,
+      "learning_rate": 3.4058269420036937e-06,
+      "loss": 0.4848,
+      "step": 4932
+    },
+    {
+      "epoch": 0.40063347681312433,
+      "grad_norm": 5.055147156646343,
+      "learning_rate": 3.40521397420372e-06,
+      "loss": 0.5223,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4007146917891659,
+      "grad_norm": 6.056907838015249,
+      "learning_rate": 3.4046009437677296e-06,
+      "loss": 0.658,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4007959067652075,
+      "grad_norm": 5.370500579188346,
+      "learning_rate": 3.403987850738142e-06,
+      "loss": 0.4064,
+      "step": 4935
+    },
+    {
+      "epoch": 0.40087712174124907,
+      "grad_norm": 5.583756706162014,
+      "learning_rate": 3.4033746951573797e-06,
+      "loss": 0.4991,
+      "step": 4936
+    },
+    {
+      "epoch": 0.40095833671729064,
+      "grad_norm": 7.016728588738892,
+      "learning_rate": 3.4027614770678695e-06,
+      "loss": 0.6236,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4010395516933323,
+      "grad_norm": 4.595088770001303,
+      "learning_rate": 3.402148196512042e-06,
+      "loss": 0.4602,
+      "step": 4938
+    },
+    {
+      "epoch": 0.40112076666937385,
+      "grad_norm": 8.586997426840448,
+      "learning_rate": 3.4015348535323344e-06,
+      "loss": 0.582,
+      "step": 4939
+    },
+    {
+      "epoch": 0.40120198164541543,
+      "grad_norm": 6.420810576653918,
+      "learning_rate": 3.400921448171187e-06,
+      "loss": 0.456,
+      "step": 4940
+    },
+    {
+      "epoch": 0.401283196621457,
+      "grad_norm": 3.598889910541628,
+      "learning_rate": 3.4003079804710414e-06,
+      "loss": 0.5049,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4013644115974986,
+      "grad_norm": 4.85229866007985,
+      "learning_rate": 3.39969445047435e-06,
+      "loss": 0.5795,
+      "step": 4942
+    },
+    {
+      "epoch": 0.40144562657354016,
+      "grad_norm": 5.046704044838436,
+      "learning_rate": 3.399080858223564e-06,
+      "loss": 0.4898,
+      "step": 4943
+    },
+    {
+      "epoch": 0.40152684154958174,
+      "grad_norm": 4.201283039118606,
+      "learning_rate": 3.3984672037611403e-06,
+      "loss": 0.4583,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4016080565256233,
+      "grad_norm": 3.6245756239233855,
+      "learning_rate": 3.3978534871295423e-06,
+      "loss": 0.569,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4016892715016649,
+      "grad_norm": 5.943049720731405,
+      "learning_rate": 3.3972397083712337e-06,
+      "loss": 0.4635,
+      "step": 4946
+    },
+    {
+      "epoch": 0.40177048647770647,
+      "grad_norm": 4.4825133089550615,
+      "learning_rate": 3.3966258675286868e-06,
+      "loss": 0.5545,
+      "step": 4947
+    },
+    {
+      "epoch": 0.40185170145374804,
+      "grad_norm": 7.159533755424081,
+      "learning_rate": 3.3960119646443743e-06,
+      "loss": 0.5132,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4019329164297897,
+      "grad_norm": 5.668432680762583,
+      "learning_rate": 3.395397999760777e-06,
+      "loss": 0.4824,
+      "step": 4949
+    },
+    {
+      "epoch": 0.40201413140583125,
+      "grad_norm": 7.74454321710248,
+      "learning_rate": 3.394783972920376e-06,
+      "loss": 0.5292,
+      "step": 4950
+    },
+    {
+      "epoch": 0.40209534638187283,
+      "grad_norm": 3.9688138836277864,
+      "learning_rate": 3.3941698841656594e-06,
+      "loss": 0.5123,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4021765613579144,
+      "grad_norm": 3.6092591329939196,
+      "learning_rate": 3.3935557335391194e-06,
+      "loss": 0.5618,
+      "step": 4952
+    },
+    {
+      "epoch": 0.402257776333956,
+      "grad_norm": 5.268354030573778,
+      "learning_rate": 3.3929415210832526e-06,
+      "loss": 0.4638,
+      "step": 4953
+    },
+    {
+      "epoch": 0.40233899130999756,
+      "grad_norm": 4.725073559638671,
+      "learning_rate": 3.392327246840558e-06,
+      "loss": 0.5442,
+      "step": 4954
+    },
+    {
+      "epoch": 0.40242020628603914,
+      "grad_norm": 5.0239266752483225,
+      "learning_rate": 3.39171291085354e-06,
+      "loss": 0.5132,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4025014212620807,
+      "grad_norm": 4.688205783968622,
+      "learning_rate": 3.3910985131647077e-06,
+      "loss": 0.6074,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4025826362381223,
+      "grad_norm": 4.306614565733134,
+      "learning_rate": 3.3904840538165745e-06,
+      "loss": 0.4564,
+      "step": 4957
+    },
+    {
+      "epoch": 0.40266385121416387,
+      "grad_norm": 13.862109865488437,
+      "learning_rate": 3.3898695328516585e-06,
+      "loss": 0.536,
+      "step": 4958
+    },
+    {
+      "epoch": 0.40274506619020545,
+      "grad_norm": 4.512610902634363,
+      "learning_rate": 3.38925495031248e-06,
+      "loss": 0.6559,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4028262811662471,
+      "grad_norm": 4.1778178120621225,
+      "learning_rate": 3.3886403062415653e-06,
+      "loss": 0.5741,
+      "step": 4960
+    },
+    {
+      "epoch": 0.40290749614228866,
+      "grad_norm": 5.5923461249681194,
+      "learning_rate": 3.3880256006814436e-06,
+      "loss": 0.7565,
+      "step": 4961
+    },
+    {
+      "epoch": 0.40298871111833023,
+      "grad_norm": 10.781234356503536,
+      "learning_rate": 3.387410833674651e-06,
+      "loss": 0.817,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4030699260943718,
+      "grad_norm": 7.255219164441723,
+      "learning_rate": 3.386796005263725e-06,
+      "loss": 0.4918,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4031511410704134,
+      "grad_norm": 3.7214218539504094,
+      "learning_rate": 3.3861811154912085e-06,
+      "loss": 0.5502,
+      "step": 4964
+    },
+    {
+      "epoch": 0.40323235604645496,
+      "grad_norm": 3.7233458707878246,
+      "learning_rate": 3.385566164399649e-06,
+      "loss": 0.6731,
+      "step": 4965
+    },
+    {
+      "epoch": 0.40331357102249654,
+      "grad_norm": 6.244845838241695,
+      "learning_rate": 3.3849511520315986e-06,
+      "loss": 0.426,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4033947859985381,
+      "grad_norm": 5.384785695599885,
+      "learning_rate": 3.384336078429611e-06,
+      "loss": 0.7065,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4034760009745797,
+      "grad_norm": 4.49252865805685,
+      "learning_rate": 3.3837209436362473e-06,
+      "loss": 0.477,
+      "step": 4968
+    },
+    {
+      "epoch": 0.4035572159506213,
+      "grad_norm": 3.4933203948485643,
+      "learning_rate": 3.3831057476940716e-06,
+      "loss": 0.5459,
+      "step": 4969
+    },
+    {
+      "epoch": 0.40363843092666285,
+      "grad_norm": 7.33852450348436,
+      "learning_rate": 3.382490490645651e-06,
+      "loss": 0.559,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4037196459027045,
+      "grad_norm": 5.565835641557815,
+      "learning_rate": 3.3818751725335595e-06,
+      "loss": 0.4383,
+      "step": 4971
+    },
+    {
+      "epoch": 0.40380086087874606,
+      "grad_norm": 6.069976549417349,
+      "learning_rate": 3.3812597934003746e-06,
+      "loss": 0.6346,
+      "step": 4972
+    },
+    {
+      "epoch": 0.40388207585478764,
+      "grad_norm": 5.41297727878491,
+      "learning_rate": 3.3806443532886736e-06,
+      "loss": 0.4902,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4039632908308292,
+      "grad_norm": 3.474757683944031,
+      "learning_rate": 3.3800288522410464e-06,
+      "loss": 0.6833,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4040445058068708,
+      "grad_norm": 18.494129306251725,
+      "learning_rate": 3.3794132903000787e-06,
+      "loss": 0.4717,
+      "step": 4975
+    },
+    {
+      "epoch": 0.40412572078291237,
+      "grad_norm": 5.201629259993079,
+      "learning_rate": 3.3787976675083657e-06,
+      "loss": 0.4586,
+      "step": 4976
+    },
+    {
+      "epoch": 0.40420693575895394,
+      "grad_norm": 3.067579639195292,
+      "learning_rate": 3.3781819839085056e-06,
+      "loss": 0.6093,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4042881507349955,
+      "grad_norm": 6.51098650482296,
+      "learning_rate": 3.3775662395431e-06,
+      "loss": 0.4642,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4043693657110371,
+      "grad_norm": 3.2528809931282057,
+      "learning_rate": 3.376950434454754e-06,
+      "loss": 0.6634,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4044505806870787,
+      "grad_norm": 6.255509398772904,
+      "learning_rate": 3.37633456868608e-06,
+      "loss": 0.5264,
+      "step": 4980
+    },
+    {
+      "epoch": 0.40453179566312025,
+      "grad_norm": 6.628156482525939,
+      "learning_rate": 3.3757186422796918e-06,
+      "loss": 0.4249,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4046130106391619,
+      "grad_norm": 5.667448673111919,
+      "learning_rate": 3.3751026552782085e-06,
+      "loss": 0.5736,
+      "step": 4982
+    },
+    {
+      "epoch": 0.40469422561520346,
+      "grad_norm": 4.987055006193434,
+      "learning_rate": 3.3744866077242516e-06,
+      "loss": 0.585,
+      "step": 4983
+    },
+    {
+      "epoch": 0.40477544059124504,
+      "grad_norm": 5.82552976044451,
+      "learning_rate": 3.3738704996604505e-06,
+      "loss": 0.6341,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4048566555672866,
+      "grad_norm": 4.093717103910503,
+      "learning_rate": 3.373254331129436e-06,
+      "loss": 0.4046,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4049378705433282,
+      "grad_norm": 7.070638776876593,
+      "learning_rate": 3.3726381021738426e-06,
+      "loss": 0.4569,
+      "step": 4986
+    },
+    {
+      "epoch": 0.40501908551936977,
+      "grad_norm": 4.327438597298289,
+      "learning_rate": 3.372021812836311e-06,
+      "loss": 0.522,
+      "step": 4987
+    },
+    {
+      "epoch": 0.40510030049541135,
+      "grad_norm": 7.802424985063589,
+      "learning_rate": 3.371405463159486e-06,
+      "loss": 0.5055,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4051815154714529,
+      "grad_norm": 5.2484985718748405,
+      "learning_rate": 3.3707890531860143e-06,
+      "loss": 0.5691,
+      "step": 4989
+    },
+    {
+      "epoch": 0.4052627304474945,
+      "grad_norm": 3.961595770162228,
+      "learning_rate": 3.3701725829585484e-06,
+      "loss": 0.5087,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4053439454235361,
+      "grad_norm": 5.115368545445615,
+      "learning_rate": 3.369556052519746e-06,
+      "loss": 0.5014,
+      "step": 4991
+    },
+    {
+      "epoch": 0.40542516039957766,
+      "grad_norm": 9.264859588158524,
+      "learning_rate": 3.3689394619122654e-06,
+      "loss": 0.5217,
+      "step": 4992
+    },
+    {
+      "epoch": 0.4055063753756193,
+      "grad_norm": 5.809302524376221,
+      "learning_rate": 3.3683228111787738e-06,
+      "loss": 0.5209,
+      "step": 4993
+    },
+    {
+      "epoch": 0.40558759035166086,
+      "grad_norm": 6.347688148270995,
+      "learning_rate": 3.367706100361939e-06,
+      "loss": 0.5284,
+      "step": 4994
+    },
+    {
+      "epoch": 0.40566880532770244,
+      "grad_norm": 5.2075723248409185,
+      "learning_rate": 3.3670893295044344e-06,
+      "loss": 0.6141,
+      "step": 4995
+    },
+    {
+      "epoch": 0.405750020303744,
+      "grad_norm": 6.241751717955846,
+      "learning_rate": 3.3664724986489368e-06,
+      "loss": 0.539,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4058312352797856,
+      "grad_norm": 5.913411596718994,
+      "learning_rate": 3.3658556078381283e-06,
+      "loss": 0.4779,
+      "step": 4997
+    },
+    {
+      "epoch": 0.4059124502558272,
+      "grad_norm": 5.987603839595508,
+      "learning_rate": 3.3652386571146945e-06,
+      "loss": 0.4415,
+      "step": 4998
+    },
+    {
+      "epoch": 0.40599366523186875,
+      "grad_norm": 5.342255084971644,
+      "learning_rate": 3.3646216465213245e-06,
+      "loss": 0.5426,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4060748802079103,
+      "grad_norm": 5.881978781645484,
+      "learning_rate": 3.364004576100712e-06,
+      "loss": 0.5235,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4061560951839519,
+      "grad_norm": 5.907061959419017,
+      "learning_rate": 3.3633874458955573e-06,
+      "loss": 0.5061,
+      "step": 5001
+    },
+    {
+      "epoch": 0.4062373101599935,
+      "grad_norm": 4.640005094027489,
+      "learning_rate": 3.362770255948559e-06,
+      "loss": 0.5892,
+      "step": 5002
+    },
+    {
+      "epoch": 0.40631852513603506,
+      "grad_norm": 5.450649196042516,
+      "learning_rate": 3.3621530063024257e-06,
+      "loss": 0.6451,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4063997401120767,
+      "grad_norm": 12.30013823840977,
+      "learning_rate": 3.3615356969998676e-06,
+      "loss": 0.5278,
+      "step": 5004
+    },
+    {
+      "epoch": 0.40648095508811827,
+      "grad_norm": 3.2285712876853156,
+      "learning_rate": 3.360918328083598e-06,
+      "loss": 0.6247,
+      "step": 5005
+    },
+    {
+      "epoch": 0.40656217006415984,
+      "grad_norm": 4.8576215977033135,
+      "learning_rate": 3.3603008995963373e-06,
+      "loss": 0.527,
+      "step": 5006
+    },
+    {
+      "epoch": 0.4066433850402014,
+      "grad_norm": 6.983405839371138,
+      "learning_rate": 3.3596834115808074e-06,
+      "loss": 0.5264,
+      "step": 5007
+    },
+    {
+      "epoch": 0.406724600016243,
+      "grad_norm": 4.393414313571411,
+      "learning_rate": 3.3590658640797346e-06,
+      "loss": 0.6397,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4068058149922846,
+      "grad_norm": 5.416974817120511,
+      "learning_rate": 3.3584482571358513e-06,
+      "loss": 0.4797,
+      "step": 5009
+    },
+    {
+      "epoch": 0.40688702996832615,
+      "grad_norm": 3.780442832500498,
+      "learning_rate": 3.357830590791891e-06,
+      "loss": 0.452,
+      "step": 5010
+    },
+    {
+      "epoch": 0.40696824494436773,
+      "grad_norm": 4.203958087565441,
+      "learning_rate": 3.3572128650905946e-06,
+      "loss": 0.4962,
+      "step": 5011
+    },
+    {
+      "epoch": 0.4070494599204093,
+      "grad_norm": 7.016929878457646,
+      "learning_rate": 3.3565950800747038e-06,
+      "loss": 0.5625,
+      "step": 5012
+    },
+    {
+      "epoch": 0.4071306748964509,
+      "grad_norm": 5.779316536373502,
+      "learning_rate": 3.355977235786968e-06,
+      "loss": 0.544,
+      "step": 5013
+    },
+    {
+      "epoch": 0.40721188987249246,
+      "grad_norm": 4.419872849655686,
+      "learning_rate": 3.3553593322701374e-06,
+      "loss": 0.5445,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4072931048485341,
+      "grad_norm": 19.583565435812158,
+      "learning_rate": 3.3547413695669673e-06,
+      "loss": 0.4479,
+      "step": 5015
+    },
+    {
+      "epoch": 0.40737431982457567,
+      "grad_norm": 6.737000577753693,
+      "learning_rate": 3.3541233477202184e-06,
+      "loss": 0.6231,
+      "step": 5016
+    },
+    {
+      "epoch": 0.40745553480061725,
+      "grad_norm": 5.651619729539659,
+      "learning_rate": 3.3535052667726546e-06,
+      "loss": 0.4791,
+      "step": 5017
+    },
+    {
+      "epoch": 0.4075367497766588,
+      "grad_norm": 13.01736060108638,
+      "learning_rate": 3.352887126767043e-06,
+      "loss": 0.618,
+      "step": 5018
+    },
+    {
+      "epoch": 0.4076179647527004,
+      "grad_norm": 7.913220552107575,
+      "learning_rate": 3.352268927746156e-06,
+      "loss": 0.5014,
+      "step": 5019
+    },
+    {
+      "epoch": 0.407699179728742,
+      "grad_norm": 5.00487843260526,
+      "learning_rate": 3.3516506697527706e-06,
+      "loss": 0.7005,
+      "step": 5020
+    },
+    {
+      "epoch": 0.40778039470478356,
+      "grad_norm": 4.0699036925411045,
+      "learning_rate": 3.3510323528296656e-06,
+      "loss": 0.5042,
+      "step": 5021
+    },
+    {
+      "epoch": 0.40786160968082513,
+      "grad_norm": 3.9550635463784944,
+      "learning_rate": 3.3504139770196252e-06,
+      "loss": 0.5318,
+      "step": 5022
+    },
+    {
+      "epoch": 0.4079428246568667,
+      "grad_norm": 3.9690809902318605,
+      "learning_rate": 3.3497955423654395e-06,
+      "loss": 0.5501,
+      "step": 5023
+    },
+    {
+      "epoch": 0.4080240396329083,
+      "grad_norm": 5.238253417658213,
+      "learning_rate": 3.349177048909899e-06,
+      "loss": 0.4221,
+      "step": 5024
+    },
+    {
+      "epoch": 0.40810525460894986,
+      "grad_norm": 6.31641058902131,
+      "learning_rate": 3.3485584966958005e-06,
+      "loss": 0.5599,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4081864695849915,
+      "grad_norm": 4.773613373392485,
+      "learning_rate": 3.3479398857659464e-06,
+      "loss": 0.672,
+      "step": 5026
+    },
+    {
+      "epoch": 0.4082676845610331,
+      "grad_norm": 6.145123089078632,
+      "learning_rate": 3.3473212161631385e-06,
+      "loss": 0.4964,
+      "step": 5027
+    },
+    {
+      "epoch": 0.40834889953707465,
+      "grad_norm": 4.502728069285821,
+      "learning_rate": 3.3467024879301873e-06,
+      "loss": 0.633,
+      "step": 5028
+    },
+    {
+      "epoch": 0.4084301145131162,
+      "grad_norm": 4.232356275831273,
+      "learning_rate": 3.346083701109905e-06,
+      "loss": 0.7541,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4085113294891578,
+      "grad_norm": 6.5669331099437045,
+      "learning_rate": 3.3454648557451087e-06,
+      "loss": 0.6154,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4085925444651994,
+      "grad_norm": 3.585238962477474,
+      "learning_rate": 3.3448459518786193e-06,
+      "loss": 0.563,
+      "step": 5031
+    },
+    {
+      "epoch": 0.40867375944124096,
+      "grad_norm": 7.113560531413456,
+      "learning_rate": 3.3442269895532604e-06,
+      "loss": 0.4665,
+      "step": 5032
+    },
+    {
+      "epoch": 0.40875497441728254,
+      "grad_norm": 4.184427215416845,
+      "learning_rate": 3.3436079688118618e-06,
+      "loss": 0.4255,
+      "step": 5033
+    },
+    {
+      "epoch": 0.4088361893933241,
+      "grad_norm": 3.690634979166768,
+      "learning_rate": 3.3429888896972575e-06,
+      "loss": 0.4262,
+      "step": 5034
+    },
+    {
+      "epoch": 0.4089174043693657,
+      "grad_norm": 3.7301420071620424,
+      "learning_rate": 3.3423697522522823e-06,
+      "loss": 0.5172,
+      "step": 5035
+    },
+    {
+      "epoch": 0.40899861934540727,
+      "grad_norm": 4.87720646565865,
+      "learning_rate": 3.3417505565197794e-06,
+      "loss": 0.6716,
+      "step": 5036
+    },
+    {
+      "epoch": 0.4090798343214489,
+      "grad_norm": 3.6748482560950615,
+      "learning_rate": 3.3411313025425927e-06,
+      "loss": 0.5472,
+      "step": 5037
+    },
+    {
+      "epoch": 0.4091610492974905,
+      "grad_norm": 5.7048262277305275,
+      "learning_rate": 3.340511990363571e-06,
+      "loss": 0.3746,
+      "step": 5038
+    },
+    {
+      "epoch": 0.40924226427353205,
+      "grad_norm": 5.363584407358412,
+      "learning_rate": 3.3398926200255684e-06,
+      "loss": 0.3669,
+      "step": 5039
+    },
+    {
+      "epoch": 0.40932347924957363,
+      "grad_norm": 4.185826585370702,
+      "learning_rate": 3.3392731915714417e-06,
+      "loss": 0.5765,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4094046942256152,
+      "grad_norm": 5.9673921212077286,
+      "learning_rate": 3.338653705044051e-06,
+      "loss": 0.574,
+      "step": 5041
+    },
+    {
+      "epoch": 0.4094859092016568,
+      "grad_norm": 13.697669906869592,
+      "learning_rate": 3.3380341604862633e-06,
+      "loss": 0.4787,
+      "step": 5042
+    },
+    {
+      "epoch": 0.40956712417769836,
+      "grad_norm": 18.101167067110552,
+      "learning_rate": 3.3374145579409467e-06,
+      "loss": 0.5793,
+      "step": 5043
+    },
+    {
+      "epoch": 0.40964833915373994,
+      "grad_norm": 33.302750260887564,
+      "learning_rate": 3.3367948974509743e-06,
+      "loss": 0.4709,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4097295541297815,
+      "grad_norm": 5.565821515126126,
+      "learning_rate": 3.336175179059224e-06,
+      "loss": 0.4976,
+      "step": 5045
+    },
+    {
+      "epoch": 0.4098107691058231,
+      "grad_norm": 4.92473685068305,
+      "learning_rate": 3.335555402808577e-06,
+      "loss": 0.5237,
+      "step": 5046
+    },
+    {
+      "epoch": 0.40989198408186467,
+      "grad_norm": 3.6230723915327343,
+      "learning_rate": 3.334935568741918e-06,
+      "loss": 0.505,
+      "step": 5047
+    },
+    {
+      "epoch": 0.4099731990579063,
+      "grad_norm": 4.073098074480497,
+      "learning_rate": 3.3343156769021355e-06,
+      "loss": 0.5275,
+      "step": 5048
+    },
+    {
+      "epoch": 0.4100544140339479,
+      "grad_norm": 3.908868759278294,
+      "learning_rate": 3.333695727332125e-06,
+      "loss": 0.635,
+      "step": 5049
+    },
+    {
+      "epoch": 0.41013562900998946,
+      "grad_norm": 4.898135196202766,
+      "learning_rate": 3.3330757200747828e-06,
+      "loss": 0.4958,
+      "step": 5050
+    },
+    {
+      "epoch": 0.41021684398603103,
+      "grad_norm": 11.463379753224626,
+      "learning_rate": 3.332455655173008e-06,
+      "loss": 0.4647,
+      "step": 5051
+    },
+    {
+      "epoch": 0.4102980589620726,
+      "grad_norm": 4.243407156064371,
+      "learning_rate": 3.3318355326697093e-06,
+      "loss": 0.6904,
+      "step": 5052
+    },
+    {
+      "epoch": 0.4103792739381142,
+      "grad_norm": 3.5903815519776585,
+      "learning_rate": 3.3312153526077933e-06,
+      "loss": 0.5248,
+      "step": 5053
+    },
+    {
+      "epoch": 0.41046048891415576,
+      "grad_norm": 5.5607811919006345,
+      "learning_rate": 3.330595115030174e-06,
+      "loss": 0.4837,
+      "step": 5054
+    },
+    {
+      "epoch": 0.41054170389019734,
+      "grad_norm": 4.396444235590735,
+      "learning_rate": 3.3299748199797686e-06,
+      "loss": 0.4331,
+      "step": 5055
+    },
+    {
+      "epoch": 0.4106229188662389,
+      "grad_norm": 5.6326755354518845,
+      "learning_rate": 3.3293544674994987e-06,
+      "loss": 0.5109,
+      "step": 5056
+    },
+    {
+      "epoch": 0.4107041338422805,
+      "grad_norm": 6.458559629689074,
+      "learning_rate": 3.328734057632289e-06,
+      "loss": 0.5488,
+      "step": 5057
+    },
+    {
+      "epoch": 0.41078534881832207,
+      "grad_norm": 6.259903972340533,
+      "learning_rate": 3.328113590421068e-06,
+      "loss": 0.787,
+      "step": 5058
+    },
+    {
+      "epoch": 0.4108665637943637,
+      "grad_norm": 5.031847327165066,
+      "learning_rate": 3.3274930659087694e-06,
+      "loss": 0.6045,
+      "step": 5059
+    },
+    {
+      "epoch": 0.4109477787704053,
+      "grad_norm": 12.757281627069613,
+      "learning_rate": 3.3268724841383302e-06,
+      "loss": 0.5007,
+      "step": 5060
+    },
+    {
+      "epoch": 0.41102899374644686,
+      "grad_norm": 8.972711433026783,
+      "learning_rate": 3.3262518451526916e-06,
+      "loss": 0.6009,
+      "step": 5061
+    },
+    {
+      "epoch": 0.41111020872248843,
+      "grad_norm": 3.924314034573167,
+      "learning_rate": 3.3256311489947973e-06,
+      "loss": 0.4718,
+      "step": 5062
+    },
+    {
+      "epoch": 0.41119142369853,
+      "grad_norm": 3.8326582602133796,
+      "learning_rate": 3.3250103957075987e-06,
+      "loss": 0.6721,
+      "step": 5063
+    },
+    {
+      "epoch": 0.4112726386745716,
+      "grad_norm": 5.858057317848575,
+      "learning_rate": 3.3243895853340445e-06,
+      "loss": 0.3982,
+      "step": 5064
+    },
+    {
+      "epoch": 0.41135385365061317,
+      "grad_norm": 4.79530936974041,
+      "learning_rate": 3.323768717917096e-06,
+      "loss": 0.6906,
+      "step": 5065
+    },
+    {
+      "epoch": 0.41143506862665474,
+      "grad_norm": 3.761887070675124,
+      "learning_rate": 3.323147793499712e-06,
+      "loss": 0.5835,
+      "step": 5066
+    },
+    {
+      "epoch": 0.4115162836026963,
+      "grad_norm": 4.8103574968475735,
+      "learning_rate": 3.3225268121248567e-06,
+      "loss": 0.5361,
+      "step": 5067
+    },
+    {
+      "epoch": 0.4115974985787379,
+      "grad_norm": 3.859720909374451,
+      "learning_rate": 3.321905773835498e-06,
+      "loss": 0.6076,
+      "step": 5068
+    },
+    {
+      "epoch": 0.4116787135547795,
+      "grad_norm": 3.758573255947487,
+      "learning_rate": 3.3212846786746113e-06,
+      "loss": 0.4216,
+      "step": 5069
+    },
+    {
+      "epoch": 0.4117599285308211,
+      "grad_norm": 7.220042667589109,
+      "learning_rate": 3.3206635266851707e-06,
+      "loss": 0.5195,
+      "step": 5070
+    },
+    {
+      "epoch": 0.4118411435068627,
+      "grad_norm": 5.446095850430871,
+      "learning_rate": 3.320042317910157e-06,
+      "loss": 0.5241,
+      "step": 5071
+    },
+    {
+      "epoch": 0.41192235848290426,
+      "grad_norm": 4.589008245862078,
+      "learning_rate": 3.319421052392556e-06,
+      "loss": 0.6736,
+      "step": 5072
+    },
+    {
+      "epoch": 0.41200357345894584,
+      "grad_norm": 9.586999090184612,
+      "learning_rate": 3.318799730175354e-06,
+      "loss": 0.5927,
+      "step": 5073
+    },
+    {
+      "epoch": 0.4120847884349874,
+      "grad_norm": 3.790231648461632,
+      "learning_rate": 3.3181783513015443e-06,
+      "loss": 0.5262,
+      "step": 5074
+    },
+    {
+      "epoch": 0.412166003411029,
+      "grad_norm": 7.8339239381427435,
+      "learning_rate": 3.317556915814123e-06,
+      "loss": 0.5578,
+      "step": 5075
+    },
+    {
+      "epoch": 0.41224721838707057,
+      "grad_norm": 4.189621669708876,
+      "learning_rate": 3.31693542375609e-06,
+      "loss": 0.4758,
+      "step": 5076
+    },
+    {
+      "epoch": 0.41232843336311215,
+      "grad_norm": 3.23569345024193,
+      "learning_rate": 3.316313875170449e-06,
+      "loss": 0.5989,
+      "step": 5077
+    },
+    {
+      "epoch": 0.4124096483391537,
+      "grad_norm": 5.092573629488183,
+      "learning_rate": 3.3156922701002082e-06,
+      "loss": 0.4451,
+      "step": 5078
+    },
+    {
+      "epoch": 0.4124908633151953,
+      "grad_norm": 3.7634789724489774,
+      "learning_rate": 3.3150706085883795e-06,
+      "loss": 0.6206,
+      "step": 5079
+    },
+    {
+      "epoch": 0.4125720782912369,
+      "grad_norm": 9.494128622068938,
+      "learning_rate": 3.3144488906779775e-06,
+      "loss": 0.5531,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4126532932672785,
+      "grad_norm": 5.01807369882204,
+      "learning_rate": 3.3138271164120235e-06,
+      "loss": 0.4668,
+      "step": 5081
+    },
+    {
+      "epoch": 0.4127345082433201,
+      "grad_norm": 4.224562759189403,
+      "learning_rate": 3.3132052858335405e-06,
+      "loss": 0.471,
+      "step": 5082
+    },
+    {
+      "epoch": 0.41281572321936166,
+      "grad_norm": 4.452581770491441,
+      "learning_rate": 3.312583398985555e-06,
+      "loss": 0.478,
+      "step": 5083
+    },
+    {
+      "epoch": 0.41289693819540324,
+      "grad_norm": 6.547729750220411,
+      "learning_rate": 3.3119614559110986e-06,
+      "loss": 0.7315,
+      "step": 5084
+    },
+    {
+      "epoch": 0.4129781531714448,
+      "grad_norm": 3.7189334139526307,
+      "learning_rate": 3.3113394566532076e-06,
+      "loss": 0.4778,
+      "step": 5085
+    },
+    {
+      "epoch": 0.4130593681474864,
+      "grad_norm": 5.682620271914815,
+      "learning_rate": 3.310717401254919e-06,
+      "loss": 0.6956,
+      "step": 5086
+    },
+    {
+      "epoch": 0.41314058312352797,
+      "grad_norm": 5.539361693130007,
+      "learning_rate": 3.3100952897592774e-06,
+      "loss": 0.4507,
+      "step": 5087
+    },
+    {
+      "epoch": 0.41322179809956955,
+      "grad_norm": 5.822291658506241,
+      "learning_rate": 3.3094731222093297e-06,
+      "loss": 0.5914,
+      "step": 5088
+    },
+    {
+      "epoch": 0.4133030130756111,
+      "grad_norm": 4.73471914454109,
+      "learning_rate": 3.3088508986481256e-06,
+      "loss": 0.4881,
+      "step": 5089
+    },
+    {
+      "epoch": 0.4133842280516527,
+      "grad_norm": 6.22074010943859,
+      "learning_rate": 3.30822861911872e-06,
+      "loss": 0.603,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4134654430276943,
+      "grad_norm": 6.1051785076416385,
+      "learning_rate": 3.3076062836641716e-06,
+      "loss": 0.43,
+      "step": 5091
+    },
+    {
+      "epoch": 0.4135466580037359,
+      "grad_norm": 4.391606365906377,
+      "learning_rate": 3.306983892327542e-06,
+      "loss": 0.4868,
+      "step": 5092
+    },
+    {
+      "epoch": 0.4136278729797775,
+      "grad_norm": 3.796872062788742,
+      "learning_rate": 3.306361445151899e-06,
+      "loss": 0.5556,
+      "step": 5093
+    },
+    {
+      "epoch": 0.41370908795581907,
+      "grad_norm": 5.3352051929174165,
+      "learning_rate": 3.3057389421803104e-06,
+      "loss": 0.5305,
+      "step": 5094
+    },
+    {
+      "epoch": 0.41379030293186064,
+      "grad_norm": 5.507108563864587,
+      "learning_rate": 3.305116383455852e-06,
+      "loss": 0.8252,
+      "step": 5095
+    },
+    {
+      "epoch": 0.4138715179079022,
+      "grad_norm": 3.239020837952215,
+      "learning_rate": 3.304493769021601e-06,
+      "loss": 0.5663,
+      "step": 5096
+    },
+    {
+      "epoch": 0.4139527328839438,
+      "grad_norm": 5.771163705049402,
+      "learning_rate": 3.3038710989206386e-06,
+      "loss": 0.5291,
+      "step": 5097
+    },
+    {
+      "epoch": 0.4140339478599854,
+      "grad_norm": 7.098970415020295,
+      "learning_rate": 3.303248373196051e-06,
+      "loss": 0.5858,
+      "step": 5098
+    },
+    {
+      "epoch": 0.41411516283602695,
+      "grad_norm": 6.2763287339806535,
+      "learning_rate": 3.3026255918909267e-06,
+      "loss": 0.4381,
+      "step": 5099
+    },
+    {
+      "epoch": 0.41419637781206853,
+      "grad_norm": 3.6575950544234708,
+      "learning_rate": 3.302002755048359e-06,
+      "loss": 0.7504,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4142775927881101,
+      "grad_norm": 4.90381745333394,
+      "learning_rate": 3.3013798627114457e-06,
+      "loss": 0.3803,
+      "step": 5101
+    },
+    {
+      "epoch": 0.4143588077641517,
+      "grad_norm": 4.885461667100054,
+      "learning_rate": 3.300756914923287e-06,
+      "loss": 0.4982,
+      "step": 5102
+    },
+    {
+      "epoch": 0.4144400227401933,
+      "grad_norm": 4.517776876693077,
+      "learning_rate": 3.3001339117269883e-06,
+      "loss": 0.6321,
+      "step": 5103
+    },
+    {
+      "epoch": 0.4145212377162349,
+      "grad_norm": 4.594992279285769,
+      "learning_rate": 3.2995108531656566e-06,
+      "loss": 0.4652,
+      "step": 5104
+    },
+    {
+      "epoch": 0.41460245269227647,
+      "grad_norm": 4.353302582897617,
+      "learning_rate": 3.298887739282406e-06,
+      "loss": 0.5871,
+      "step": 5105
+    },
+    {
+      "epoch": 0.41468366766831805,
+      "grad_norm": 8.458802751943896,
+      "learning_rate": 3.298264570120351e-06,
+      "loss": 0.7833,
+      "step": 5106
+    },
+    {
+      "epoch": 0.4147648826443596,
+      "grad_norm": 3.170384706895752,
+      "learning_rate": 3.297641345722613e-06,
+      "loss": 0.5037,
+      "step": 5107
+    },
+    {
+      "epoch": 0.4148460976204012,
+      "grad_norm": 2.9676348563178094,
+      "learning_rate": 3.2970180661323155e-06,
+      "loss": 0.6279,
+      "step": 5108
+    },
+    {
+      "epoch": 0.4149273125964428,
+      "grad_norm": 4.616218599976382,
+      "learning_rate": 3.2963947313925857e-06,
+      "loss": 0.559,
+      "step": 5109
+    },
+    {
+      "epoch": 0.41500852757248435,
+      "grad_norm": 6.559911673941787,
+      "learning_rate": 3.295771341546555e-06,
+      "loss": 0.5066,
+      "step": 5110
+    },
+    {
+      "epoch": 0.41508974254852593,
+      "grad_norm": 4.702456652104927,
+      "learning_rate": 3.2951478966373602e-06,
+      "loss": 0.4724,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4151709575245675,
+      "grad_norm": 6.625773908660462,
+      "learning_rate": 3.2945243967081386e-06,
+      "loss": 0.5304,
+      "step": 5112
+    },
+    {
+      "epoch": 0.4152521725006091,
+      "grad_norm": 4.794677100016037,
+      "learning_rate": 3.2939008418020334e-06,
+      "loss": 0.5455,
+      "step": 5113
+    },
+    {
+      "epoch": 0.4153333874766507,
+      "grad_norm": 4.228847169637584,
+      "learning_rate": 3.293277231962192e-06,
+      "loss": 0.4838,
+      "step": 5114
+    },
+    {
+      "epoch": 0.4154146024526923,
+      "grad_norm": 5.780889261663269,
+      "learning_rate": 3.292653567231765e-06,
+      "loss": 0.5851,
+      "step": 5115
+    },
+    {
+      "epoch": 0.41549581742873387,
+      "grad_norm": 3.5767019478443514,
+      "learning_rate": 3.2920298476539047e-06,
+      "loss": 0.4035,
+      "step": 5116
+    },
+    {
+      "epoch": 0.41557703240477545,
+      "grad_norm": 3.173589475313961,
+      "learning_rate": 3.2914060732717725e-06,
+      "loss": 0.4834,
+      "step": 5117
+    },
+    {
+      "epoch": 0.415658247380817,
+      "grad_norm": 5.702832664803554,
+      "learning_rate": 3.290782244128527e-06,
+      "loss": 0.4122,
+      "step": 5118
+    },
+    {
+      "epoch": 0.4157394623568586,
+      "grad_norm": 4.350245123381758,
+      "learning_rate": 3.290158360267336e-06,
+      "loss": 0.7049,
+      "step": 5119
+    },
+    {
+      "epoch": 0.4158206773329002,
+      "grad_norm": 3.6272992204161887,
+      "learning_rate": 3.2895344217313683e-06,
+      "loss": 0.5299,
+      "step": 5120
+    },
+    {
+      "epoch": 0.41590189230894176,
+      "grad_norm": 3.9213077791760966,
+      "learning_rate": 3.2889104285637967e-06,
+      "loss": 0.5795,
+      "step": 5121
+    },
+    {
+      "epoch": 0.41598310728498333,
+      "grad_norm": 4.304699561537021,
+      "learning_rate": 3.2882863808077993e-06,
+      "loss": 0.5066,
+      "step": 5122
+    },
+    {
+      "epoch": 0.4160643222610249,
+      "grad_norm": 13.4372391747002,
+      "learning_rate": 3.287662278506556e-06,
+      "loss": 0.4601,
+      "step": 5123
+    },
+    {
+      "epoch": 0.4161455372370665,
+      "grad_norm": 8.076930536538704,
+      "learning_rate": 3.2870381217032522e-06,
+      "loss": 0.4771,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4162267522131081,
+      "grad_norm": 5.712865061327913,
+      "learning_rate": 3.2864139104410753e-06,
+      "loss": 0.3819,
+      "step": 5125
+    },
+    {
+      "epoch": 0.4163079671891497,
+      "grad_norm": 8.244135425876427,
+      "learning_rate": 3.2857896447632174e-06,
+      "loss": 0.5816,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4163891821651913,
+      "grad_norm": 3.9304755927131785,
+      "learning_rate": 3.2851653247128755e-06,
+      "loss": 0.5813,
+      "step": 5127
+    },
+    {
+      "epoch": 0.41647039714123285,
+      "grad_norm": 3.476794981739987,
+      "learning_rate": 3.2845409503332488e-06,
+      "loss": 0.7026,
+      "step": 5128
+    },
+    {
+      "epoch": 0.41655161211727443,
+      "grad_norm": 6.988919037400744,
+      "learning_rate": 3.2839165216675396e-06,
+      "loss": 0.3883,
+      "step": 5129
+    },
+    {
+      "epoch": 0.416632827093316,
+      "grad_norm": 9.043393482730245,
+      "learning_rate": 3.283292038758956e-06,
+      "loss": 0.6782,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4167140420693576,
+      "grad_norm": 9.047180698274353,
+      "learning_rate": 3.2826675016507094e-06,
+      "loss": 0.3717,
+      "step": 5131
+    },
+    {
+      "epoch": 0.41679525704539916,
+      "grad_norm": 4.352324808412342,
+      "learning_rate": 3.2820429103860133e-06,
+      "loss": 0.5285,
+      "step": 5132
+    },
+    {
+      "epoch": 0.41687647202144074,
+      "grad_norm": 4.3428722591294155,
+      "learning_rate": 3.281418265008087e-06,
+      "loss": 0.571,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4169576869974823,
+      "grad_norm": 4.529875570698177,
+      "learning_rate": 3.280793565560153e-06,
+      "loss": 0.6332,
+      "step": 5134
+    },
+    {
+      "epoch": 0.41703890197352395,
+      "grad_norm": 10.119978089195198,
+      "learning_rate": 3.280168812085436e-06,
+      "loss": 0.4528,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4171201169495655,
+      "grad_norm": 3.973556997891654,
+      "learning_rate": 3.279544004627166e-06,
+      "loss": 0.6027,
+      "step": 5136
+    },
+    {
+      "epoch": 0.4172013319256071,
+      "grad_norm": 3.7856125466570814,
+      "learning_rate": 3.2789191432285767e-06,
+      "loss": 0.5312,
+      "step": 5137
+    },
+    {
+      "epoch": 0.4172825469016487,
+      "grad_norm": 3.897046711276253,
+      "learning_rate": 3.278294227932905e-06,
+      "loss": 0.4928,
+      "step": 5138
+    },
+    {
+      "epoch": 0.41736376187769025,
+      "grad_norm": 3.9807540904239715,
+      "learning_rate": 3.277669258783391e-06,
+      "loss": 0.4981,
+      "step": 5139
+    },
+    {
+      "epoch": 0.41744497685373183,
+      "grad_norm": 4.193176888500685,
+      "learning_rate": 3.277044235823281e-06,
+      "loss": 0.5247,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4175261918297734,
+      "grad_norm": 4.102295880640859,
+      "learning_rate": 3.2764191590958234e-06,
+      "loss": 0.5928,
+      "step": 5141
+    },
+    {
+      "epoch": 0.417607406805815,
+      "grad_norm": 4.4166185861935,
+      "learning_rate": 3.2757940286442676e-06,
+      "loss": 0.4061,
+      "step": 5142
+    },
+    {
+      "epoch": 0.41768862178185656,
+      "grad_norm": 6.205963915441601,
+      "learning_rate": 3.2751688445118705e-06,
+      "loss": 0.5939,
+      "step": 5143
+    },
+    {
+      "epoch": 0.41776983675789814,
+      "grad_norm": 9.375588558168431,
+      "learning_rate": 3.2745436067418934e-06,
+      "loss": 0.5445,
+      "step": 5144
+    },
+    {
+      "epoch": 0.4178510517339397,
+      "grad_norm": 7.231124759113749,
+      "learning_rate": 3.2739183153775964e-06,
+      "loss": 0.3809,
+      "step": 5145
+    },
+    {
+      "epoch": 0.41793226670998135,
+      "grad_norm": 5.36109001138212,
+      "learning_rate": 3.2732929704622485e-06,
+      "loss": 0.4421,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4180134816860229,
+      "grad_norm": 5.1614659053586545,
+      "learning_rate": 3.2726675720391203e-06,
+      "loss": 0.4925,
+      "step": 5147
+    },
+    {
+      "epoch": 0.4180946966620645,
+      "grad_norm": 5.095668465921769,
+      "learning_rate": 3.272042120151485e-06,
+      "loss": 0.7233,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4181759116381061,
+      "grad_norm": 6.655912818912801,
+      "learning_rate": 3.2714166148426204e-06,
+      "loss": 0.4635,
+      "step": 5149
+    },
+    {
+      "epoch": 0.41825712661414766,
+      "grad_norm": 5.621140760979749,
+      "learning_rate": 3.27079105615581e-06,
+      "loss": 0.5281,
+      "step": 5150
+    },
+    {
+      "epoch": 0.41833834159018923,
+      "grad_norm": 5.0469957881501255,
+      "learning_rate": 3.2701654441343365e-06,
+      "loss": 0.5175,
+      "step": 5151
+    },
+    {
+      "epoch": 0.4184195565662308,
+      "grad_norm": 4.710792152616251,
+      "learning_rate": 3.269539778821491e-06,
+      "loss": 0.4824,
+      "step": 5152
+    },
+    {
+      "epoch": 0.4185007715422724,
+      "grad_norm": 5.248614874618813,
+      "learning_rate": 3.268914060260565e-06,
+      "loss": 0.6031,
+      "step": 5153
+    },
+    {
+      "epoch": 0.41858198651831396,
+      "grad_norm": 3.7091517078378096,
+      "learning_rate": 3.2682882884948557e-06,
+      "loss": 0.445,
+      "step": 5154
+    },
+    {
+      "epoch": 0.41866320149435554,
+      "grad_norm": 7.473085780555477,
+      "learning_rate": 3.2676624635676637e-06,
+      "loss": 0.4225,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4187444164703971,
+      "grad_norm": 5.984937157684743,
+      "learning_rate": 3.267036585522291e-06,
+      "loss": 0.3815,
+      "step": 5156
+    },
+    {
+      "epoch": 0.41882563144643875,
+      "grad_norm": 7.910654251052096,
+      "learning_rate": 3.2664106544020464e-06,
+      "loss": 0.4595,
+      "step": 5157
+    },
+    {
+      "epoch": 0.41890684642248033,
+      "grad_norm": 5.447690662144098,
+      "learning_rate": 3.2657846702502404e-06,
+      "loss": 0.5105,
+      "step": 5158
+    },
+    {
+      "epoch": 0.4189880613985219,
+      "grad_norm": 6.61183045497076,
+      "learning_rate": 3.2651586331101887e-06,
+      "loss": 0.4271,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4190692763745635,
+      "grad_norm": 6.227375988300761,
+      "learning_rate": 3.2645325430252096e-06,
+      "loss": 0.5066,
+      "step": 5160
+    },
+    {
+      "epoch": 0.41915049135060506,
+      "grad_norm": 3.5873779321139714,
+      "learning_rate": 3.2639064000386236e-06,
+      "loss": 0.5637,
+      "step": 5161
+    },
+    {
+      "epoch": 0.41923170632664664,
+      "grad_norm": 3.2262674778585376,
+      "learning_rate": 3.2632802041937574e-06,
+      "loss": 0.4754,
+      "step": 5162
+    },
+    {
+      "epoch": 0.4193129213026882,
+      "grad_norm": 4.976278018847007,
+      "learning_rate": 3.262653955533942e-06,
+      "loss": 0.4554,
+      "step": 5163
+    },
+    {
+      "epoch": 0.4193941362787298,
+      "grad_norm": 7.7409127238618245,
+      "learning_rate": 3.262027654102508e-06,
+      "loss": 0.5582,
+      "step": 5164
+    },
+    {
+      "epoch": 0.41947535125477137,
+      "grad_norm": 5.01210383268172,
+      "learning_rate": 3.2614012999427934e-06,
+      "loss": 0.4712,
+      "step": 5165
+    },
+    {
+      "epoch": 0.41955656623081294,
+      "grad_norm": 4.527219081383319,
+      "learning_rate": 3.26077489309814e-06,
+      "loss": 0.7056,
+      "step": 5166
+    },
+    {
+      "epoch": 0.4196377812068545,
+      "grad_norm": 4.790861705461241,
+      "learning_rate": 3.2601484336118887e-06,
+      "loss": 0.6324,
+      "step": 5167
+    },
+    {
+      "epoch": 0.41971899618289615,
+      "grad_norm": 5.6189045494809875,
+      "learning_rate": 3.2595219215273895e-06,
+      "loss": 0.4605,
+      "step": 5168
+    },
+    {
+      "epoch": 0.41980021115893773,
+      "grad_norm": 3.3926840995979206,
+      "learning_rate": 3.258895356887993e-06,
+      "loss": 0.4653,
+      "step": 5169
+    },
+    {
+      "epoch": 0.4198814261349793,
+      "grad_norm": 4.7133012677803405,
+      "learning_rate": 3.2582687397370538e-06,
+      "loss": 0.3422,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4199626411110209,
+      "grad_norm": 6.7508413558662355,
+      "learning_rate": 3.257642070117931e-06,
+      "loss": 0.4689,
+      "step": 5171
+    },
+    {
+      "epoch": 0.42004385608706246,
+      "grad_norm": 7.457776695287912,
+      "learning_rate": 3.2570153480739867e-06,
+      "loss": 0.5847,
+      "step": 5172
+    },
+    {
+      "epoch": 0.42012507106310404,
+      "grad_norm": 5.26972034775471,
+      "learning_rate": 3.2563885736485873e-06,
+      "loss": 0.5858,
+      "step": 5173
+    },
+    {
+      "epoch": 0.4202062860391456,
+      "grad_norm": 2.8327813340648325,
+      "learning_rate": 3.255761746885101e-06,
+      "loss": 0.437,
+      "step": 5174
+    },
+    {
+      "epoch": 0.4202875010151872,
+      "grad_norm": 4.515606181218866,
+      "learning_rate": 3.2551348678269023e-06,
+      "loss": 0.6866,
+      "step": 5175
+    },
+    {
+      "epoch": 0.42036871599122877,
+      "grad_norm": 4.52820690426378,
+      "learning_rate": 3.2545079365173672e-06,
+      "loss": 0.4457,
+      "step": 5176
+    },
+    {
+      "epoch": 0.42044993096727035,
+      "grad_norm": 5.8094460720614505,
+      "learning_rate": 3.253880952999876e-06,
+      "loss": 0.501,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4205311459433119,
+      "grad_norm": 9.872169395018048,
+      "learning_rate": 3.2532539173178125e-06,
+      "loss": 0.6308,
+      "step": 5178
+    },
+    {
+      "epoch": 0.42061236091935356,
+      "grad_norm": 8.660617050371519,
+      "learning_rate": 3.2526268295145647e-06,
+      "loss": 0.5709,
+      "step": 5179
+    },
+    {
+      "epoch": 0.42069357589539513,
+      "grad_norm": 4.736443287957975,
+      "learning_rate": 3.251999689633523e-06,
+      "loss": 0.4353,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4207747908714367,
+      "grad_norm": 10.700216416621961,
+      "learning_rate": 3.2513724977180828e-06,
+      "loss": 0.623,
+      "step": 5181
+    },
+    {
+      "epoch": 0.4208560058474783,
+      "grad_norm": 11.331063237198684,
+      "learning_rate": 3.250745253811643e-06,
+      "loss": 0.4766,
+      "step": 5182
+    },
+    {
+      "epoch": 0.42093722082351986,
+      "grad_norm": 4.634568286716608,
+      "learning_rate": 3.250117957957604e-06,
+      "loss": 0.5879,
+      "step": 5183
+    },
+    {
+      "epoch": 0.42101843579956144,
+      "grad_norm": 4.031580348842282,
+      "learning_rate": 3.249490610199373e-06,
+      "loss": 0.4788,
+      "step": 5184
+    },
+    {
+      "epoch": 0.421099650775603,
+      "grad_norm": 5.071428970099418,
+      "learning_rate": 3.248863210580358e-06,
+      "loss": 0.4319,
+      "step": 5185
+    },
+    {
+      "epoch": 0.4211808657516446,
+      "grad_norm": 4.271403990802303,
+      "learning_rate": 3.248235759143972e-06,
+      "loss": 0.4285,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4212620807276862,
+      "grad_norm": 5.620600198973997,
+      "learning_rate": 3.247608255933632e-06,
+      "loss": 0.5508,
+      "step": 5187
+    },
+    {
+      "epoch": 0.42134329570372775,
+      "grad_norm": 3.5656379095963366,
+      "learning_rate": 3.2469807009927568e-06,
+      "loss": 0.5465,
+      "step": 5188
+    },
+    {
+      "epoch": 0.4214245106797693,
+      "grad_norm": 4.267563683154685,
+      "learning_rate": 3.2463530943647708e-06,
+      "loss": 0.6901,
+      "step": 5189
+    },
+    {
+      "epoch": 0.42150572565581096,
+      "grad_norm": 6.968445424375631,
+      "learning_rate": 3.2457254360931013e-06,
+      "loss": 0.6524,
+      "step": 5190
+    },
+    {
+      "epoch": 0.42158694063185254,
+      "grad_norm": 6.404940362152012,
+      "learning_rate": 3.245097726221177e-06,
+      "loss": 0.4577,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4216681556078941,
+      "grad_norm": 10.422891227755217,
+      "learning_rate": 3.244469964792434e-06,
+      "loss": 0.487,
+      "step": 5192
+    },
+    {
+      "epoch": 0.4217493705839357,
+      "grad_norm": 4.235778167208473,
+      "learning_rate": 3.24384215185031e-06,
+      "loss": 0.5221,
+      "step": 5193
+    },
+    {
+      "epoch": 0.42183058555997727,
+      "grad_norm": 5.314122103787476,
+      "learning_rate": 3.2432142874382442e-06,
+      "loss": 0.5772,
+      "step": 5194
+    },
+    {
+      "epoch": 0.42191180053601884,
+      "grad_norm": 34.115592182605965,
+      "learning_rate": 3.2425863715996852e-06,
+      "loss": 0.6579,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4219930155120604,
+      "grad_norm": 5.189198561105788,
+      "learning_rate": 3.241958404378078e-06,
+      "loss": 0.5213,
+      "step": 5196
+    },
+    {
+      "epoch": 0.422074230488102,
+      "grad_norm": 4.933122374718706,
+      "learning_rate": 3.2413303858168767e-06,
+      "loss": 0.6707,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4221554454641436,
+      "grad_norm": 4.33962971377441,
+      "learning_rate": 3.2407023159595356e-06,
+      "loss": 0.4882,
+      "step": 5198
+    },
+    {
+      "epoch": 0.42223666044018515,
+      "grad_norm": 6.399904760111607,
+      "learning_rate": 3.2400741948495146e-06,
+      "loss": 0.383,
+      "step": 5199
+    },
+    {
+      "epoch": 0.42231787541622673,
+      "grad_norm": 11.60358668566535,
+      "learning_rate": 3.239446022530276e-06,
+      "loss": 0.6851,
+      "step": 5200
+    },
+    {
+      "epoch": 0.42239909039226836,
+      "grad_norm": 10.376518506368765,
+      "learning_rate": 3.2388177990452863e-06,
+      "loss": 0.5153,
+      "step": 5201
+    },
+    {
+      "epoch": 0.42248030536830994,
+      "grad_norm": 5.087706195580212,
+      "learning_rate": 3.2381895244380146e-06,
+      "loss": 0.4698,
+      "step": 5202
+    },
+    {
+      "epoch": 0.4225615203443515,
+      "grad_norm": 3.447914836981935,
+      "learning_rate": 3.237561198751935e-06,
+      "loss": 0.5198,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4226427353203931,
+      "grad_norm": 7.767514821519397,
+      "learning_rate": 3.2369328220305242e-06,
+      "loss": 0.5854,
+      "step": 5204
+    },
+    {
+      "epoch": 0.42272395029643467,
+      "grad_norm": 7.5527718662260295,
+      "learning_rate": 3.2363043943172616e-06,
+      "loss": 0.4495,
+      "step": 5205
+    },
+    {
+      "epoch": 0.42280516527247625,
+      "grad_norm": 6.408787873539771,
+      "learning_rate": 3.235675915655633e-06,
+      "loss": 0.4103,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4228863802485178,
+      "grad_norm": 8.10134326274207,
+      "learning_rate": 3.235047386089123e-06,
+      "loss": 0.4829,
+      "step": 5207
+    },
+    {
+      "epoch": 0.4229675952245594,
+      "grad_norm": 4.1969050205146425,
+      "learning_rate": 3.2344188056612247e-06,
+      "loss": 0.5541,
+      "step": 5208
+    },
+    {
+      "epoch": 0.423048810200601,
+      "grad_norm": 4.954025585754433,
+      "learning_rate": 3.233790174415432e-06,
+      "loss": 0.4104,
+      "step": 5209
+    },
+    {
+      "epoch": 0.42313002517664255,
+      "grad_norm": 5.191981721617316,
+      "learning_rate": 3.2331614923952424e-06,
+      "loss": 0.6211,
+      "step": 5210
+    },
+    {
+      "epoch": 0.42321124015268413,
+      "grad_norm": 10.46456809201404,
+      "learning_rate": 3.232532759644158e-06,
+      "loss": 0.5638,
+      "step": 5211
+    },
+    {
+      "epoch": 0.42329245512872576,
+      "grad_norm": 4.916424280031669,
+      "learning_rate": 3.231903976205684e-06,
+      "loss": 0.4139,
+      "step": 5212
+    },
+    {
+      "epoch": 0.42337367010476734,
+      "grad_norm": 4.247004575276243,
+      "learning_rate": 3.231275142123328e-06,
+      "loss": 0.3899,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4234548850808089,
+      "grad_norm": 4.899674792955789,
+      "learning_rate": 3.2306462574406024e-06,
+      "loss": 0.697,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4235361000568505,
+      "grad_norm": 4.934275124900704,
+      "learning_rate": 3.2300173222010225e-06,
+      "loss": 0.6492,
+      "step": 5215
+    },
+    {
+      "epoch": 0.42361731503289207,
+      "grad_norm": 4.326952640941383,
+      "learning_rate": 3.229388336448107e-06,
+      "loss": 0.5192,
+      "step": 5216
+    },
+    {
+      "epoch": 0.42369853000893365,
+      "grad_norm": 6.478214535298061,
+      "learning_rate": 3.22875930022538e-06,
+      "loss": 0.5203,
+      "step": 5217
+    },
+    {
+      "epoch": 0.4237797449849752,
+      "grad_norm": 3.7343533496893393,
+      "learning_rate": 3.2281302135763655e-06,
+      "loss": 0.5004,
+      "step": 5218
+    },
+    {
+      "epoch": 0.4238609599610168,
+      "grad_norm": 4.189623891419984,
+      "learning_rate": 3.227501076544594e-06,
+      "loss": 0.5008,
+      "step": 5219
+    },
+    {
+      "epoch": 0.4239421749370584,
+      "grad_norm": 9.421806026431668,
+      "learning_rate": 3.2268718891735985e-06,
+      "loss": 0.4428,
+      "step": 5220
+    },
+    {
+      "epoch": 0.42402338991309996,
+      "grad_norm": 8.411484589237242,
+      "learning_rate": 3.2262426515069144e-06,
+      "loss": 0.5658,
+      "step": 5221
+    },
+    {
+      "epoch": 0.42410460488914153,
+      "grad_norm": 5.809866223556941,
+      "learning_rate": 3.225613363588084e-06,
+      "loss": 0.5688,
+      "step": 5222
+    },
+    {
+      "epoch": 0.42418581986518317,
+      "grad_norm": 4.3592044879636465,
+      "learning_rate": 3.2249840254606474e-06,
+      "loss": 0.5421,
+      "step": 5223
+    },
+    {
+      "epoch": 0.42426703484122474,
+      "grad_norm": 2.8417643428931147,
+      "learning_rate": 3.2243546371681535e-06,
+      "loss": 0.5438,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4243482498172663,
+      "grad_norm": 4.321132502798912,
+      "learning_rate": 3.2237251987541535e-06,
+      "loss": 0.669,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4244294647933079,
+      "grad_norm": 17.616950980213982,
+      "learning_rate": 3.223095710262199e-06,
+      "loss": 0.5083,
+      "step": 5226
+    },
+    {
+      "epoch": 0.4245106797693495,
+      "grad_norm": 6.679913112655923,
+      "learning_rate": 3.2224661717358474e-06,
+      "loss": 0.5919,
+      "step": 5227
+    },
+    {
+      "epoch": 0.42459189474539105,
+      "grad_norm": 4.47571261524808,
+      "learning_rate": 3.221836583218662e-06,
+      "loss": 0.376,
+      "step": 5228
+    },
+    {
+      "epoch": 0.42467310972143263,
+      "grad_norm": 8.501783709365114,
+      "learning_rate": 3.221206944754205e-06,
+      "loss": 0.4275,
+      "step": 5229
+    },
+    {
+      "epoch": 0.4247543246974742,
+      "grad_norm": 4.892245082709042,
+      "learning_rate": 3.220577256386043e-06,
+      "loss": 0.6864,
+      "step": 5230
+    },
+    {
+      "epoch": 0.4248355396735158,
+      "grad_norm": 5.054086864480611,
+      "learning_rate": 3.21994751815775e-06,
+      "loss": 0.5241,
+      "step": 5231
+    },
+    {
+      "epoch": 0.42491675464955736,
+      "grad_norm": 3.2768063978983095,
+      "learning_rate": 3.2193177301128985e-06,
+      "loss": 0.5298,
+      "step": 5232
+    },
+    {
+      "epoch": 0.42499796962559894,
+      "grad_norm": 4.649050909544704,
+      "learning_rate": 3.2186878922950672e-06,
+      "loss": 0.6465,
+      "step": 5233
+    },
+    {
+      "epoch": 0.42507918460164057,
+      "grad_norm": 4.194611280947634,
+      "learning_rate": 3.218058004747837e-06,
+      "loss": 0.5813,
+      "step": 5234
+    },
+    {
+      "epoch": 0.42516039957768215,
+      "grad_norm": 12.035010001320948,
+      "learning_rate": 3.2174280675147933e-06,
+      "loss": 0.444,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4252416145537237,
+      "grad_norm": 3.9279173401772285,
+      "learning_rate": 3.2167980806395244e-06,
+      "loss": 0.4574,
+      "step": 5236
+    },
+    {
+      "epoch": 0.4253228295297653,
+      "grad_norm": 3.760359692561372,
+      "learning_rate": 3.216168044165622e-06,
+      "loss": 0.4997,
+      "step": 5237
+    },
+    {
+      "epoch": 0.4254040445058069,
+      "grad_norm": 6.915480718748514,
+      "learning_rate": 3.215537958136681e-06,
+      "loss": 0.5207,
+      "step": 5238
+    },
+    {
+      "epoch": 0.42548525948184845,
+      "grad_norm": 9.393926456415134,
+      "learning_rate": 3.2149078225963e-06,
+      "loss": 0.4996,
+      "step": 5239
+    },
+    {
+      "epoch": 0.42556647445789003,
+      "grad_norm": 5.607193609498181,
+      "learning_rate": 3.2142776375880814e-06,
+      "loss": 0.6179,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4256476894339316,
+      "grad_norm": 7.777688950913407,
+      "learning_rate": 3.213647403155631e-06,
+      "loss": 0.4407,
+      "step": 5241
+    },
+    {
+      "epoch": 0.4257289044099732,
+      "grad_norm": 5.406198429944911,
+      "learning_rate": 3.213017119342557e-06,
+      "loss": 0.4342,
+      "step": 5242
+    },
+    {
+      "epoch": 0.42581011938601476,
+      "grad_norm": 4.332500644815715,
+      "learning_rate": 3.2123867861924705e-06,
+      "loss": 0.547,
+      "step": 5243
+    },
+    {
+      "epoch": 0.42589133436205634,
+      "grad_norm": 4.158064015085327,
+      "learning_rate": 3.211756403748991e-06,
+      "loss": 0.6803,
+      "step": 5244
+    },
+    {
+      "epoch": 0.42597254933809797,
+      "grad_norm": 4.670204213436716,
+      "learning_rate": 3.211125972055734e-06,
+      "loss": 0.3712,
+      "step": 5245
+    },
+    {
+      "epoch": 0.42605376431413955,
+      "grad_norm": 3.1072025340370835,
+      "learning_rate": 3.210495491156323e-06,
+      "loss": 0.6592,
+      "step": 5246
+    },
+    {
+      "epoch": 0.4261349792901811,
+      "grad_norm": 3.845276286103103,
+      "learning_rate": 3.2098649610943855e-06,
+      "loss": 0.3973,
+      "step": 5247
+    },
+    {
+      "epoch": 0.4262161942662227,
+      "grad_norm": 5.263686826495678,
+      "learning_rate": 3.2092343819135485e-06,
+      "loss": 0.4361,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4262974092422643,
+      "grad_norm": 4.706614514686621,
+      "learning_rate": 3.2086037536574467e-06,
+      "loss": 0.5642,
+      "step": 5249
+    },
+    {
+      "epoch": 0.42637862421830586,
+      "grad_norm": 7.9955278197924455,
+      "learning_rate": 3.207973076369715e-06,
+      "loss": 0.4473,
+      "step": 5250
+    },
+    {
+      "epoch": 0.42645983919434743,
+      "grad_norm": 5.6591134035265505,
+      "learning_rate": 3.2073423500939926e-06,
+      "loss": 0.4853,
+      "step": 5251
+    },
+    {
+      "epoch": 0.426541054170389,
+      "grad_norm": 8.612516192884062,
+      "learning_rate": 3.206711574873924e-06,
+      "loss": 0.462,
+      "step": 5252
+    },
+    {
+      "epoch": 0.4266222691464306,
+      "grad_norm": 5.102810963488392,
+      "learning_rate": 3.2060807507531545e-06,
+      "loss": 0.5559,
+      "step": 5253
+    },
+    {
+      "epoch": 0.42670348412247217,
+      "grad_norm": 7.341129074246249,
+      "learning_rate": 3.2054498777753335e-06,
+      "loss": 0.5183,
+      "step": 5254
+    },
+    {
+      "epoch": 0.42678469909851374,
+      "grad_norm": 5.00922975998508,
+      "learning_rate": 3.204818955984115e-06,
+      "loss": 0.4557,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4268659140745554,
+      "grad_norm": 3.0225137055666087,
+      "learning_rate": 3.2041879854231545e-06,
+      "loss": 0.6202,
+      "step": 5256
+    },
+    {
+      "epoch": 0.42694712905059695,
+      "grad_norm": 5.862930940720101,
+      "learning_rate": 3.203556966136113e-06,
+      "loss": 0.5674,
+      "step": 5257
+    },
+    {
+      "epoch": 0.42702834402663853,
+      "grad_norm": 8.760401643981206,
+      "learning_rate": 3.202925898166652e-06,
+      "loss": 0.4603,
+      "step": 5258
+    },
+    {
+      "epoch": 0.4271095590026801,
+      "grad_norm": 7.198009661677629,
+      "learning_rate": 3.2022947815584393e-06,
+      "loss": 0.4481,
+      "step": 5259
+    },
+    {
+      "epoch": 0.4271907739787217,
+      "grad_norm": 4.513288458419195,
+      "learning_rate": 3.2016636163551456e-06,
+      "loss": 0.4721,
+      "step": 5260
+    },
+    {
+      "epoch": 0.42727198895476326,
+      "grad_norm": 4.556711076123282,
+      "learning_rate": 3.2010324026004425e-06,
+      "loss": 0.5153,
+      "step": 5261
+    },
+    {
+      "epoch": 0.42735320393080484,
+      "grad_norm": 4.403874321167366,
+      "learning_rate": 3.200401140338007e-06,
+      "loss": 0.5405,
+      "step": 5262
+    },
+    {
+      "epoch": 0.4274344189068464,
+      "grad_norm": 15.464524857508628,
+      "learning_rate": 3.1997698296115192e-06,
+      "loss": 0.5812,
+      "step": 5263
+    },
+    {
+      "epoch": 0.427515633882888,
+      "grad_norm": 6.763078104210325,
+      "learning_rate": 3.1991384704646632e-06,
+      "loss": 0.4366,
+      "step": 5264
+    },
+    {
+      "epoch": 0.42759684885892957,
+      "grad_norm": 3.8083736623019377,
+      "learning_rate": 3.198507062941125e-06,
+      "loss": 0.7049,
+      "step": 5265
+    },
+    {
+      "epoch": 0.42767806383497114,
+      "grad_norm": 4.411432581791518,
+      "learning_rate": 3.197875607084595e-06,
+      "loss": 0.5536,
+      "step": 5266
+    },
+    {
+      "epoch": 0.4277592788110128,
+      "grad_norm": 5.209356808212503,
+      "learning_rate": 3.1972441029387664e-06,
+      "loss": 0.7173,
+      "step": 5267
+    },
+    {
+      "epoch": 0.42784049378705435,
+      "grad_norm": 4.422812878298527,
+      "learning_rate": 3.196612550547336e-06,
+      "loss": 0.5126,
+      "step": 5268
+    },
+    {
+      "epoch": 0.42792170876309593,
+      "grad_norm": 4.063375030795528,
+      "learning_rate": 3.1959809499540033e-06,
+      "loss": 0.4163,
+      "step": 5269
+    },
+    {
+      "epoch": 0.4280029237391375,
+      "grad_norm": 6.439580078200473,
+      "learning_rate": 3.1953493012024728e-06,
+      "loss": 0.46,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4280841387151791,
+      "grad_norm": 6.644696648989759,
+      "learning_rate": 3.1947176043364512e-06,
+      "loss": 0.5559,
+      "step": 5271
+    },
+    {
+      "epoch": 0.42816535369122066,
+      "grad_norm": 9.625730715608832,
+      "learning_rate": 3.194085859399647e-06,
+      "loss": 0.4542,
+      "step": 5272
+    },
+    {
+      "epoch": 0.42824656866726224,
+      "grad_norm": 6.80723990351529,
+      "learning_rate": 3.1934540664357756e-06,
+      "loss": 0.3875,
+      "step": 5273
+    },
+    {
+      "epoch": 0.4283277836433038,
+      "grad_norm": 4.063505294236836,
+      "learning_rate": 3.1928222254885527e-06,
+      "loss": 0.4263,
+      "step": 5274
+    },
+    {
+      "epoch": 0.4284089986193454,
+      "grad_norm": 5.254426921833707,
+      "learning_rate": 3.192190336601698e-06,
+      "loss": 0.601,
+      "step": 5275
+    },
+    {
+      "epoch": 0.42849021359538697,
+      "grad_norm": 3.3439818153250274,
+      "learning_rate": 3.1915583998189365e-06,
+      "loss": 0.8393,
+      "step": 5276
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 4.423002542739268,
+      "learning_rate": 3.190926415183993e-06,
+      "loss": 0.4905,
+      "step": 5277
+    },
+    {
+      "epoch": 0.4286526435474702,
+      "grad_norm": 4.969141698478439,
+      "learning_rate": 3.190294382740598e-06,
+      "loss": 0.4598,
+      "step": 5278
+    },
+    {
+      "epoch": 0.42873385852351176,
+      "grad_norm": 5.329263080541811,
+      "learning_rate": 3.189662302532486e-06,
+      "loss": 0.5027,
+      "step": 5279
+    },
+    {
+      "epoch": 0.42881507349955333,
+      "grad_norm": 8.465008482689347,
+      "learning_rate": 3.1890301746033914e-06,
+      "loss": 0.4451,
+      "step": 5280
+    },
+    {
+      "epoch": 0.4288962884755949,
+      "grad_norm": 4.288959841432594,
+      "learning_rate": 3.188397998997056e-06,
+      "loss": 0.5379,
+      "step": 5281
+    },
+    {
+      "epoch": 0.4289775034516365,
+      "grad_norm": 4.307272162236529,
+      "learning_rate": 3.1877657757572223e-06,
+      "loss": 0.3864,
+      "step": 5282
+    },
+    {
+      "epoch": 0.42905871842767807,
+      "grad_norm": 4.853723061541425,
+      "learning_rate": 3.187133504927637e-06,
+      "loss": 0.5744,
+      "step": 5283
+    },
+    {
+      "epoch": 0.42913993340371964,
+      "grad_norm": 9.888573881930817,
+      "learning_rate": 3.18650118655205e-06,
+      "loss": 0.4017,
+      "step": 5284
+    },
+    {
+      "epoch": 0.4292211483797612,
+      "grad_norm": 6.195263958719191,
+      "learning_rate": 3.1858688206742135e-06,
+      "loss": 0.4613,
+      "step": 5285
+    },
+    {
+      "epoch": 0.4293023633558028,
+      "grad_norm": 5.993567524504076,
+      "learning_rate": 3.1852364073378845e-06,
+      "loss": 0.5304,
+      "step": 5286
+    },
+    {
+      "epoch": 0.4293835783318444,
+      "grad_norm": 6.246122494535498,
+      "learning_rate": 3.1846039465868233e-06,
+      "loss": 0.4809,
+      "step": 5287
+    },
+    {
+      "epoch": 0.42946479330788595,
+      "grad_norm": 11.344295485158227,
+      "learning_rate": 3.1839714384647914e-06,
+      "loss": 0.416,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4295460082839276,
+      "grad_norm": 6.789926972632632,
+      "learning_rate": 3.1833388830155564e-06,
+      "loss": 0.5286,
+      "step": 5289
+    },
+    {
+      "epoch": 0.42962722325996916,
+      "grad_norm": 4.731313984743315,
+      "learning_rate": 3.1827062802828878e-06,
+      "loss": 0.6368,
+      "step": 5290
+    },
+    {
+      "epoch": 0.42970843823601074,
+      "grad_norm": 8.422362176653303,
+      "learning_rate": 3.182073630310557e-06,
+      "loss": 0.4566,
+      "step": 5291
+    },
+    {
+      "epoch": 0.4297896532120523,
+      "grad_norm": 5.811200383485644,
+      "learning_rate": 3.18144093314234e-06,
+      "loss": 0.7081,
+      "step": 5292
+    },
+    {
+      "epoch": 0.4298708681880939,
+      "grad_norm": 3.464910361407899,
+      "learning_rate": 3.180808188822019e-06,
+      "loss": 0.6889,
+      "step": 5293
+    },
+    {
+      "epoch": 0.42995208316413547,
+      "grad_norm": 3.8106177663295115,
+      "learning_rate": 3.180175397393373e-06,
+      "loss": 0.6863,
+      "step": 5294
+    },
+    {
+      "epoch": 0.43003329814017704,
+      "grad_norm": 8.071444985141785,
+      "learning_rate": 3.1795425589001896e-06,
+      "loss": 0.5488,
+      "step": 5295
+    },
+    {
+      "epoch": 0.4301145131162186,
+      "grad_norm": 4.430290323012749,
+      "learning_rate": 3.178909673386257e-06,
+      "loss": 0.5592,
+      "step": 5296
+    },
+    {
+      "epoch": 0.4301957280922602,
+      "grad_norm": 4.5330502482145505,
+      "learning_rate": 3.178276740895369e-06,
+      "loss": 0.5002,
+      "step": 5297
+    },
+    {
+      "epoch": 0.4302769430683018,
+      "grad_norm": 7.279026055668376,
+      "learning_rate": 3.1776437614713197e-06,
+      "loss": 0.547,
+      "step": 5298
+    },
+    {
+      "epoch": 0.43035815804434335,
+      "grad_norm": 5.6210852302334855,
+      "learning_rate": 3.177010735157909e-06,
+      "loss": 0.6157,
+      "step": 5299
+    },
+    {
+      "epoch": 0.430439373020385,
+      "grad_norm": 6.112959552581171,
+      "learning_rate": 3.1763776619989377e-06,
+      "loss": 0.5323,
+      "step": 5300
+    },
+    {
+      "epoch": 0.43052058799642656,
+      "grad_norm": 4.4197983016582745,
+      "learning_rate": 3.175744542038212e-06,
+      "loss": 0.525,
+      "step": 5301
+    },
+    {
+      "epoch": 0.43060180297246814,
+      "grad_norm": 5.4564596091826,
+      "learning_rate": 3.175111375319541e-06,
+      "loss": 0.5436,
+      "step": 5302
+    },
+    {
+      "epoch": 0.4306830179485097,
+      "grad_norm": 4.9689391465007615,
+      "learning_rate": 3.174478161886736e-06,
+      "loss": 0.4957,
+      "step": 5303
+    },
+    {
+      "epoch": 0.4307642329245513,
+      "grad_norm": 9.04638437397121,
+      "learning_rate": 3.1738449017836102e-06,
+      "loss": 0.4726,
+      "step": 5304
+    },
+    {
+      "epoch": 0.43084544790059287,
+      "grad_norm": 6.518271460980143,
+      "learning_rate": 3.173211595053985e-06,
+      "loss": 0.4154,
+      "step": 5305
+    },
+    {
+      "epoch": 0.43092666287663445,
+      "grad_norm": 9.139018032286032,
+      "learning_rate": 3.17257824174168e-06,
+      "loss": 0.6605,
+      "step": 5306
+    },
+    {
+      "epoch": 0.431007877852676,
+      "grad_norm": 5.893703080064373,
+      "learning_rate": 3.17194484189052e-06,
+      "loss": 0.5479,
+      "step": 5307
+    },
+    {
+      "epoch": 0.4310890928287176,
+      "grad_norm": 6.561881133428007,
+      "learning_rate": 3.171311395544333e-06,
+      "loss": 0.5071,
+      "step": 5308
+    },
+    {
+      "epoch": 0.4311703078047592,
+      "grad_norm": 5.394616296616023,
+      "learning_rate": 3.170677902746951e-06,
+      "loss": 0.4559,
+      "step": 5309
+    },
+    {
+      "epoch": 0.43125152278080076,
+      "grad_norm": 6.001462406111083,
+      "learning_rate": 3.170044363542207e-06,
+      "loss": 0.4429,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4313327377568424,
+      "grad_norm": 5.400528463179634,
+      "learning_rate": 3.1694107779739394e-06,
+      "loss": 0.4939,
+      "step": 5311
+    },
+    {
+      "epoch": 0.43141395273288397,
+      "grad_norm": 3.4971169628242347,
+      "learning_rate": 3.1687771460859886e-06,
+      "loss": 0.5207,
+      "step": 5312
+    },
+    {
+      "epoch": 0.43149516770892554,
+      "grad_norm": 5.801443009456409,
+      "learning_rate": 3.168143467922199e-06,
+      "loss": 0.537,
+      "step": 5313
+    },
+    {
+      "epoch": 0.4315763826849671,
+      "grad_norm": 3.877634512463603,
+      "learning_rate": 3.1675097435264175e-06,
+      "loss": 0.6002,
+      "step": 5314
+    },
+    {
+      "epoch": 0.4316575976610087,
+      "grad_norm": 4.436235293442774,
+      "learning_rate": 3.166875972942494e-06,
+      "loss": 0.5651,
+      "step": 5315
+    },
+    {
+      "epoch": 0.4317388126370503,
+      "grad_norm": 3.8720642376323857,
+      "learning_rate": 3.166242156214283e-06,
+      "loss": 0.5255,
+      "step": 5316
+    },
+    {
+      "epoch": 0.43182002761309185,
+      "grad_norm": 5.096619815050037,
+      "learning_rate": 3.1656082933856415e-06,
+      "loss": 0.4563,
+      "step": 5317
+    },
+    {
+      "epoch": 0.4319012425891334,
+      "grad_norm": 4.129271114889403,
+      "learning_rate": 3.1649743845004275e-06,
+      "loss": 0.4742,
+      "step": 5318
+    },
+    {
+      "epoch": 0.431982457565175,
+      "grad_norm": 5.369554351025192,
+      "learning_rate": 3.164340429602506e-06,
+      "loss": 0.5313,
+      "step": 5319
+    },
+    {
+      "epoch": 0.4320636725412166,
+      "grad_norm": 5.6537665065078615,
+      "learning_rate": 3.1637064287357433e-06,
+      "loss": 0.427,
+      "step": 5320
+    },
+    {
+      "epoch": 0.43214488751725816,
+      "grad_norm": 5.909754479859293,
+      "learning_rate": 3.1630723819440075e-06,
+      "loss": 0.4343,
+      "step": 5321
+    },
+    {
+      "epoch": 0.4322261024932998,
+      "grad_norm": 9.806257811882773,
+      "learning_rate": 3.1624382892711724e-06,
+      "loss": 0.4722,
+      "step": 5322
+    },
+    {
+      "epoch": 0.43230731746934137,
+      "grad_norm": 9.73741516646952,
+      "learning_rate": 3.161804150761114e-06,
+      "loss": 0.5959,
+      "step": 5323
+    },
+    {
+      "epoch": 0.43238853244538294,
+      "grad_norm": 4.940223216087869,
+      "learning_rate": 3.16116996645771e-06,
+      "loss": 0.5979,
+      "step": 5324
+    },
+    {
+      "epoch": 0.4324697474214245,
+      "grad_norm": 16.649964044548284,
+      "learning_rate": 3.1605357364048446e-06,
+      "loss": 0.4802,
+      "step": 5325
+    },
+    {
+      "epoch": 0.4325509623974661,
+      "grad_norm": 7.719840433898037,
+      "learning_rate": 3.159901460646401e-06,
+      "loss": 0.5291,
+      "step": 5326
+    },
+    {
+      "epoch": 0.4326321773735077,
+      "grad_norm": 5.612098859952399,
+      "learning_rate": 3.15926713922627e-06,
+      "loss": 0.4709,
+      "step": 5327
+    },
+    {
+      "epoch": 0.43271339234954925,
+      "grad_norm": 15.165120351544296,
+      "learning_rate": 3.1586327721883416e-06,
+      "loss": 0.4699,
+      "step": 5328
+    },
+    {
+      "epoch": 0.43279460732559083,
+      "grad_norm": 7.373093805345169,
+      "learning_rate": 3.1579983595765107e-06,
+      "loss": 0.5246,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4328758223016324,
+      "grad_norm": 4.232276270918911,
+      "learning_rate": 3.1573639014346756e-06,
+      "loss": 0.413,
+      "step": 5330
+    },
+    {
+      "epoch": 0.432957037277674,
+      "grad_norm": 5.475928131532231,
+      "learning_rate": 3.1567293978067383e-06,
+      "loss": 0.4798,
+      "step": 5331
+    },
+    {
+      "epoch": 0.43303825225371556,
+      "grad_norm": 5.319170320864578,
+      "learning_rate": 3.1560948487366016e-06,
+      "loss": 0.5221,
+      "step": 5332
+    },
+    {
+      "epoch": 0.4331194672297572,
+      "grad_norm": 4.7595654317237654,
+      "learning_rate": 3.1554602542681746e-06,
+      "loss": 0.5395,
+      "step": 5333
+    },
+    {
+      "epoch": 0.43320068220579877,
+      "grad_norm": 5.163449251799242,
+      "learning_rate": 3.154825614445366e-06,
+      "loss": 0.5755,
+      "step": 5334
+    },
+    {
+      "epoch": 0.43328189718184035,
+      "grad_norm": 4.2774589507058085,
+      "learning_rate": 3.154190929312091e-06,
+      "loss": 0.4156,
+      "step": 5335
+    },
+    {
+      "epoch": 0.4333631121578819,
+      "grad_norm": 6.636509210502926,
+      "learning_rate": 3.1535561989122667e-06,
+      "loss": 0.3532,
+      "step": 5336
+    },
+    {
+      "epoch": 0.4334443271339235,
+      "grad_norm": 4.679244393970052,
+      "learning_rate": 3.152921423289811e-06,
+      "loss": 0.5302,
+      "step": 5337
+    },
+    {
+      "epoch": 0.4335255421099651,
+      "grad_norm": 4.600315390689705,
+      "learning_rate": 3.1522866024886497e-06,
+      "loss": 0.5156,
+      "step": 5338
+    },
+    {
+      "epoch": 0.43360675708600666,
+      "grad_norm": 7.387419218755155,
+      "learning_rate": 3.1516517365527064e-06,
+      "loss": 0.6254,
+      "step": 5339
+    },
+    {
+      "epoch": 0.43368797206204823,
+      "grad_norm": 8.038125663875935,
+      "learning_rate": 3.151016825525912e-06,
+      "loss": 0.4804,
+      "step": 5340
+    },
+    {
+      "epoch": 0.4337691870380898,
+      "grad_norm": 5.664820632074068,
+      "learning_rate": 3.1503818694521993e-06,
+      "loss": 0.5997,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4338504020141314,
+      "grad_norm": 5.420884590250101,
+      "learning_rate": 3.1497468683755027e-06,
+      "loss": 0.5103,
+      "step": 5342
+    },
+    {
+      "epoch": 0.43393161699017296,
+      "grad_norm": 4.344293142047334,
+      "learning_rate": 3.1491118223397622e-06,
+      "loss": 0.6567,
+      "step": 5343
+    },
+    {
+      "epoch": 0.4340128319662146,
+      "grad_norm": 3.3509931709068184,
+      "learning_rate": 3.1484767313889186e-06,
+      "loss": 0.5217,
+      "step": 5344
+    },
+    {
+      "epoch": 0.4340940469422562,
+      "grad_norm": 6.46488127239872,
+      "learning_rate": 3.1478415955669174e-06,
+      "loss": 0.5403,
+      "step": 5345
+    },
+    {
+      "epoch": 0.43417526191829775,
+      "grad_norm": 28.243274098376233,
+      "learning_rate": 3.1472064149177063e-06,
+      "loss": 0.613,
+      "step": 5346
+    },
+    {
+      "epoch": 0.4342564768943393,
+      "grad_norm": 6.961851378871304,
+      "learning_rate": 3.1465711894852364e-06,
+      "loss": 0.5199,
+      "step": 5347
+    },
+    {
+      "epoch": 0.4343376918703809,
+      "grad_norm": 4.438081889262291,
+      "learning_rate": 3.145935919313462e-06,
+      "loss": 0.6373,
+      "step": 5348
+    },
+    {
+      "epoch": 0.4344189068464225,
+      "grad_norm": 4.323565064610984,
+      "learning_rate": 3.1453006044463417e-06,
+      "loss": 0.5597,
+      "step": 5349
+    },
+    {
+      "epoch": 0.43450012182246406,
+      "grad_norm": 3.0858340228520826,
+      "learning_rate": 3.144665244927833e-06,
+      "loss": 0.5194,
+      "step": 5350
+    },
+    {
+      "epoch": 0.43458133679850564,
+      "grad_norm": 4.877755892925608,
+      "learning_rate": 3.144029840801902e-06,
+      "loss": 0.7601,
+      "step": 5351
+    },
+    {
+      "epoch": 0.4346625517745472,
+      "grad_norm": 4.213210244527702,
+      "learning_rate": 3.1433943921125154e-06,
+      "loss": 0.5054,
+      "step": 5352
+    },
+    {
+      "epoch": 0.4347437667505888,
+      "grad_norm": 4.009304731887304,
+      "learning_rate": 3.1427588989036406e-06,
+      "loss": 0.8653,
+      "step": 5353
+    },
+    {
+      "epoch": 0.43482498172663037,
+      "grad_norm": 6.671797591935914,
+      "learning_rate": 3.1421233612192527e-06,
+      "loss": 0.5943,
+      "step": 5354
+    },
+    {
+      "epoch": 0.434906196702672,
+      "grad_norm": 6.798919847330098,
+      "learning_rate": 3.1414877791033267e-06,
+      "loss": 0.5151,
+      "step": 5355
+    },
+    {
+      "epoch": 0.4349874116787136,
+      "grad_norm": 5.3539572195583975,
+      "learning_rate": 3.1408521525998403e-06,
+      "loss": 0.4387,
+      "step": 5356
+    },
+    {
+      "epoch": 0.43506862665475515,
+      "grad_norm": 4.194017130183013,
+      "learning_rate": 3.1402164817527776e-06,
+      "loss": 0.3866,
+      "step": 5357
+    },
+    {
+      "epoch": 0.43514984163079673,
+      "grad_norm": 3.6695150698980354,
+      "learning_rate": 3.1395807666061223e-06,
+      "loss": 0.6504,
+      "step": 5358
+    },
+    {
+      "epoch": 0.4352310566068383,
+      "grad_norm": 4.641952406086023,
+      "learning_rate": 3.138945007203863e-06,
+      "loss": 0.4588,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4353122715828799,
+      "grad_norm": 10.931096401774482,
+      "learning_rate": 3.1383092035899903e-06,
+      "loss": 0.5878,
+      "step": 5360
+    },
+    {
+      "epoch": 0.43539348655892146,
+      "grad_norm": 5.762955444790604,
+      "learning_rate": 3.1376733558084994e-06,
+      "loss": 0.3661,
+      "step": 5361
+    },
+    {
+      "epoch": 0.43547470153496304,
+      "grad_norm": 4.461787974097396,
+      "learning_rate": 3.1370374639033876e-06,
+      "loss": 0.5696,
+      "step": 5362
+    },
+    {
+      "epoch": 0.4355559165110046,
+      "grad_norm": 3.0791414260530683,
+      "learning_rate": 3.1364015279186537e-06,
+      "loss": 0.6584,
+      "step": 5363
+    },
+    {
+      "epoch": 0.4356371314870462,
+      "grad_norm": 5.145988507189497,
+      "learning_rate": 3.1357655478983028e-06,
+      "loss": 0.4157,
+      "step": 5364
+    },
+    {
+      "epoch": 0.43571834646308777,
+      "grad_norm": 4.936564393357792,
+      "learning_rate": 3.135129523886341e-06,
+      "loss": 0.6098,
+      "step": 5365
+    },
+    {
+      "epoch": 0.4357995614391294,
+      "grad_norm": 5.83447257075552,
+      "learning_rate": 3.1344934559267763e-06,
+      "loss": 0.4316,
+      "step": 5366
+    },
+    {
+      "epoch": 0.435880776415171,
+      "grad_norm": 4.5475608942051196,
+      "learning_rate": 3.1338573440636232e-06,
+      "loss": 0.591,
+      "step": 5367
+    },
+    {
+      "epoch": 0.43596199139121256,
+      "grad_norm": 6.710109555046167,
+      "learning_rate": 3.133221188340897e-06,
+      "loss": 0.5388,
+      "step": 5368
+    },
+    {
+      "epoch": 0.43604320636725413,
+      "grad_norm": 6.706321579158362,
+      "learning_rate": 3.132584988802615e-06,
+      "loss": 0.5046,
+      "step": 5369
+    },
+    {
+      "epoch": 0.4361244213432957,
+      "grad_norm": 5.443757710169265,
+      "learning_rate": 3.1319487454928005e-06,
+      "loss": 0.5206,
+      "step": 5370
+    },
+    {
+      "epoch": 0.4362056363193373,
+      "grad_norm": 43.84951492925826,
+      "learning_rate": 3.1313124584554772e-06,
+      "loss": 0.601,
+      "step": 5371
+    },
+    {
+      "epoch": 0.43628685129537886,
+      "grad_norm": 2.965181185591869,
+      "learning_rate": 3.130676127734673e-06,
+      "loss": 0.5649,
+      "step": 5372
+    },
+    {
+      "epoch": 0.43636806627142044,
+      "grad_norm": 4.0688538065516955,
+      "learning_rate": 3.1300397533744176e-06,
+      "loss": 0.5637,
+      "step": 5373
+    },
+    {
+      "epoch": 0.436449281247462,
+      "grad_norm": 4.285947029576996,
+      "learning_rate": 3.129403335418747e-06,
+      "loss": 0.6039,
+      "step": 5374
+    },
+    {
+      "epoch": 0.4365304962235036,
+      "grad_norm": 9.05418272699839,
+      "learning_rate": 3.128766873911696e-06,
+      "loss": 0.504,
+      "step": 5375
+    },
+    {
+      "epoch": 0.43661171119954517,
+      "grad_norm": 7.150334825744675,
+      "learning_rate": 3.1281303688973054e-06,
+      "loss": 0.437,
+      "step": 5376
+    },
+    {
+      "epoch": 0.4366929261755868,
+      "grad_norm": 4.999179068937834,
+      "learning_rate": 3.127493820419617e-06,
+      "loss": 0.4809,
+      "step": 5377
+    },
+    {
+      "epoch": 0.4367741411516284,
+      "grad_norm": 4.8473879562154565,
+      "learning_rate": 3.1268572285226773e-06,
+      "loss": 0.4894,
+      "step": 5378
+    },
+    {
+      "epoch": 0.43685535612766996,
+      "grad_norm": 6.929506260339618,
+      "learning_rate": 3.1262205932505353e-06,
+      "loss": 0.4363,
+      "step": 5379
+    },
+    {
+      "epoch": 0.43693657110371154,
+      "grad_norm": 5.5615771450834846,
+      "learning_rate": 3.125583914647242e-06,
+      "loss": 0.5205,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4370177860797531,
+      "grad_norm": 3.3462484960680072,
+      "learning_rate": 3.124947192756853e-06,
+      "loss": 0.6615,
+      "step": 5381
+    },
+    {
+      "epoch": 0.4370990010557947,
+      "grad_norm": 6.634956155249173,
+      "learning_rate": 3.124310427623426e-06,
+      "loss": 0.4979,
+      "step": 5382
+    },
+    {
+      "epoch": 0.43718021603183627,
+      "grad_norm": 5.3106857326366095,
+      "learning_rate": 3.123673619291021e-06,
+      "loss": 0.5231,
+      "step": 5383
+    },
+    {
+      "epoch": 0.43726143100787784,
+      "grad_norm": 4.925146872849586,
+      "learning_rate": 3.123036767803703e-06,
+      "loss": 0.6155,
+      "step": 5384
+    },
+    {
+      "epoch": 0.4373426459839194,
+      "grad_norm": 6.351273194254498,
+      "learning_rate": 3.122399873205538e-06,
+      "loss": 0.4055,
+      "step": 5385
+    },
+    {
+      "epoch": 0.437423860959961,
+      "grad_norm": 7.439635575652295,
+      "learning_rate": 3.121762935540595e-06,
+      "loss": 0.4508,
+      "step": 5386
+    },
+    {
+      "epoch": 0.4375050759360026,
+      "grad_norm": 11.221500508022611,
+      "learning_rate": 3.121125954852948e-06,
+      "loss": 0.4207,
+      "step": 5387
+    },
+    {
+      "epoch": 0.4375862909120442,
+      "grad_norm": 15.28900713021598,
+      "learning_rate": 3.120488931186672e-06,
+      "loss": 0.4755,
+      "step": 5388
+    },
+    {
+      "epoch": 0.4376675058880858,
+      "grad_norm": 3.903889410795801,
+      "learning_rate": 3.1198518645858455e-06,
+      "loss": 0.4798,
+      "step": 5389
+    },
+    {
+      "epoch": 0.43774872086412736,
+      "grad_norm": 6.674485525301077,
+      "learning_rate": 3.1192147550945517e-06,
+      "loss": 0.6514,
+      "step": 5390
+    },
+    {
+      "epoch": 0.43782993584016894,
+      "grad_norm": 14.258332144198288,
+      "learning_rate": 3.118577602756873e-06,
+      "loss": 0.4016,
+      "step": 5391
+    },
+    {
+      "epoch": 0.4379111508162105,
+      "grad_norm": 4.922640216869197,
+      "learning_rate": 3.1179404076168983e-06,
+      "loss": 0.5567,
+      "step": 5392
+    },
+    {
+      "epoch": 0.4379923657922521,
+      "grad_norm": 5.597787362812228,
+      "learning_rate": 3.1173031697187178e-06,
+      "loss": 0.6444,
+      "step": 5393
+    },
+    {
+      "epoch": 0.43807358076829367,
+      "grad_norm": 3.8553056656955143,
+      "learning_rate": 3.116665889106425e-06,
+      "loss": 0.3967,
+      "step": 5394
+    },
+    {
+      "epoch": 0.43815479574433525,
+      "grad_norm": 3.517771835764705,
+      "learning_rate": 3.1160285658241157e-06,
+      "loss": 0.4896,
+      "step": 5395
+    },
+    {
+      "epoch": 0.4382360107203768,
+      "grad_norm": 3.6077248593647715,
+      "learning_rate": 3.11539119991589e-06,
+      "loss": 0.6903,
+      "step": 5396
+    },
+    {
+      "epoch": 0.4383172256964184,
+      "grad_norm": 3.6023017655336056,
+      "learning_rate": 3.1147537914258513e-06,
+      "loss": 0.602,
+      "step": 5397
+    },
+    {
+      "epoch": 0.43839844067246,
+      "grad_norm": 5.836048987707051,
+      "learning_rate": 3.1141163403981033e-06,
+      "loss": 0.4444,
+      "step": 5398
+    },
+    {
+      "epoch": 0.4384796556485016,
+      "grad_norm": 4.215159805640962,
+      "learning_rate": 3.113478846876754e-06,
+      "loss": 0.6145,
+      "step": 5399
+    },
+    {
+      "epoch": 0.4385608706245432,
+      "grad_norm": 7.095922534739042,
+      "learning_rate": 3.1128413109059164e-06,
+      "loss": 0.5145,
+      "step": 5400
+    },
+    {
+      "epoch": 0.43864208560058476,
+      "grad_norm": 3.5538230920359077,
+      "learning_rate": 3.1122037325297027e-06,
+      "loss": 0.6229,
+      "step": 5401
+    },
+    {
+      "epoch": 0.43872330057662634,
+      "grad_norm": 5.0267826629365455,
+      "learning_rate": 3.1115661117922307e-06,
+      "loss": 0.5481,
+      "step": 5402
+    },
+    {
+      "epoch": 0.4388045155526679,
+      "grad_norm": 3.179251364264237,
+      "learning_rate": 3.1109284487376213e-06,
+      "loss": 0.5039,
+      "step": 5403
+    },
+    {
+      "epoch": 0.4388857305287095,
+      "grad_norm": 3.2478475873654626,
+      "learning_rate": 3.1102907434099962e-06,
+      "loss": 0.5207,
+      "step": 5404
+    },
+    {
+      "epoch": 0.43896694550475107,
+      "grad_norm": 7.568604033338787,
+      "learning_rate": 3.1096529958534805e-06,
+      "loss": 0.5009,
+      "step": 5405
+    },
+    {
+      "epoch": 0.43904816048079265,
+      "grad_norm": 5.352814742876498,
+      "learning_rate": 3.1090152061122053e-06,
+      "loss": 0.4465,
+      "step": 5406
+    },
+    {
+      "epoch": 0.4391293754568342,
+      "grad_norm": 6.7191876007979525,
+      "learning_rate": 3.1083773742303003e-06,
+      "loss": 0.5651,
+      "step": 5407
+    },
+    {
+      "epoch": 0.4392105904328758,
+      "grad_norm": 4.940449250528321,
+      "learning_rate": 3.1077395002519013e-06,
+      "loss": 0.5716,
+      "step": 5408
+    },
+    {
+      "epoch": 0.4392918054089174,
+      "grad_norm": 3.648123205660983,
+      "learning_rate": 3.1071015842211447e-06,
+      "loss": 0.9121,
+      "step": 5409
+    },
+    {
+      "epoch": 0.439373020384959,
+      "grad_norm": 4.953572403071827,
+      "learning_rate": 3.1064636261821716e-06,
+      "loss": 0.6146,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4394542353610006,
+      "grad_norm": 6.9246508482882145,
+      "learning_rate": 3.105825626179126e-06,
+      "loss": 0.3964,
+      "step": 5411
+    },
+    {
+      "epoch": 0.43953545033704217,
+      "grad_norm": 4.790951487900791,
+      "learning_rate": 3.1051875842561523e-06,
+      "loss": 0.6518,
+      "step": 5412
+    },
+    {
+      "epoch": 0.43961666531308374,
+      "grad_norm": 4.736239149088806,
+      "learning_rate": 3.1045495004574017e-06,
+      "loss": 0.5212,
+      "step": 5413
+    },
+    {
+      "epoch": 0.4396978802891253,
+      "grad_norm": 7.540621113454675,
+      "learning_rate": 3.1039113748270248e-06,
+      "loss": 0.5596,
+      "step": 5414
+    },
+    {
+      "epoch": 0.4397790952651669,
+      "grad_norm": 3.720380110888026,
+      "learning_rate": 3.1032732074091765e-06,
+      "loss": 0.4385,
+      "step": 5415
+    },
+    {
+      "epoch": 0.4398603102412085,
+      "grad_norm": 4.7658804425871235,
+      "learning_rate": 3.1026349982480153e-06,
+      "loss": 0.6445,
+      "step": 5416
+    },
+    {
+      "epoch": 0.43994152521725005,
+      "grad_norm": 3.8751042373300963,
+      "learning_rate": 3.101996747387702e-06,
+      "loss": 0.5791,
+      "step": 5417
+    },
+    {
+      "epoch": 0.44002274019329163,
+      "grad_norm": 4.852177339352803,
+      "learning_rate": 3.101358454872399e-06,
+      "loss": 0.5171,
+      "step": 5418
+    },
+    {
+      "epoch": 0.4401039551693332,
+      "grad_norm": 4.223969934330754,
+      "learning_rate": 3.1007201207462745e-06,
+      "loss": 0.491,
+      "step": 5419
+    },
+    {
+      "epoch": 0.4401851701453748,
+      "grad_norm": 6.43559751438046,
+      "learning_rate": 3.1000817450534964e-06,
+      "loss": 0.4736,
+      "step": 5420
+    },
+    {
+      "epoch": 0.4402663851214164,
+      "grad_norm": 4.415584725720336,
+      "learning_rate": 3.0994433278382374e-06,
+      "loss": 0.5335,
+      "step": 5421
+    },
+    {
+      "epoch": 0.440347600097458,
+      "grad_norm": 7.044600702022488,
+      "learning_rate": 3.0988048691446733e-06,
+      "loss": 0.4042,
+      "step": 5422
+    },
+    {
+      "epoch": 0.44042881507349957,
+      "grad_norm": 5.958216205084174,
+      "learning_rate": 3.0981663690169806e-06,
+      "loss": 0.5884,
+      "step": 5423
+    },
+    {
+      "epoch": 0.44051003004954115,
+      "grad_norm": 7.972188724833584,
+      "learning_rate": 3.097527827499341e-06,
+      "loss": 0.4748,
+      "step": 5424
+    },
+    {
+      "epoch": 0.4405912450255827,
+      "grad_norm": 5.3560614787912195,
+      "learning_rate": 3.0968892446359383e-06,
+      "loss": 0.4807,
+      "step": 5425
+    },
+    {
+      "epoch": 0.4406724600016243,
+      "grad_norm": 4.339200995867066,
+      "learning_rate": 3.0962506204709587e-06,
+      "loss": 0.5928,
+      "step": 5426
+    },
+    {
+      "epoch": 0.4407536749776659,
+      "grad_norm": 5.4395379188651605,
+      "learning_rate": 3.0956119550485925e-06,
+      "loss": 0.4506,
+      "step": 5427
+    },
+    {
+      "epoch": 0.44083488995370745,
+      "grad_norm": 6.180414632693948,
+      "learning_rate": 3.09497324841303e-06,
+      "loss": 0.4891,
+      "step": 5428
+    },
+    {
+      "epoch": 0.44091610492974903,
+      "grad_norm": 4.088720798041899,
+      "learning_rate": 3.0943345006084678e-06,
+      "loss": 0.4157,
+      "step": 5429
+    },
+    {
+      "epoch": 0.4409973199057906,
+      "grad_norm": 4.834350877095814,
+      "learning_rate": 3.0936957116791048e-06,
+      "loss": 0.4818,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4410785348818322,
+      "grad_norm": 6.729193046534626,
+      "learning_rate": 3.0930568816691394e-06,
+      "loss": 0.4463,
+      "step": 5431
+    },
+    {
+      "epoch": 0.4411597498578738,
+      "grad_norm": 4.907909407785353,
+      "learning_rate": 3.092418010622777e-06,
+      "loss": 0.5464,
+      "step": 5432
+    },
+    {
+      "epoch": 0.4412409648339154,
+      "grad_norm": 3.5985370572492146,
+      "learning_rate": 3.091779098584224e-06,
+      "loss": 0.5931,
+      "step": 5433
+    },
+    {
+      "epoch": 0.44132217980995697,
+      "grad_norm": 3.530928584827468,
+      "learning_rate": 3.0911401455976882e-06,
+      "loss": 0.483,
+      "step": 5434
+    },
+    {
+      "epoch": 0.44140339478599855,
+      "grad_norm": 4.3804647274752035,
+      "learning_rate": 3.0905011517073834e-06,
+      "loss": 0.4682,
+      "step": 5435
+    },
+    {
+      "epoch": 0.4414846097620401,
+      "grad_norm": 8.630610987447541,
+      "learning_rate": 3.089862116957525e-06,
+      "loss": 0.5017,
+      "step": 5436
+    },
+    {
+      "epoch": 0.4415658247380817,
+      "grad_norm": 4.490184934510654,
+      "learning_rate": 3.089223041392329e-06,
+      "loss": 0.4949,
+      "step": 5437
+    },
+    {
+      "epoch": 0.4416470397141233,
+      "grad_norm": 4.853522689003282,
+      "learning_rate": 3.0885839250560172e-06,
+      "loss": 0.6344,
+      "step": 5438
+    },
+    {
+      "epoch": 0.44172825469016486,
+      "grad_norm": 7.016095469900614,
+      "learning_rate": 3.087944767992813e-06,
+      "loss": 0.5162,
+      "step": 5439
+    },
+    {
+      "epoch": 0.44180946966620643,
+      "grad_norm": 5.994420464445846,
+      "learning_rate": 3.0873055702469416e-06,
+      "loss": 0.6458,
+      "step": 5440
+    },
+    {
+      "epoch": 0.441890684642248,
+      "grad_norm": 4.376884747350049,
+      "learning_rate": 3.086666331862634e-06,
+      "loss": 0.4059,
+      "step": 5441
+    },
+    {
+      "epoch": 0.4419718996182896,
+      "grad_norm": 5.692559176862132,
+      "learning_rate": 3.0860270528841208e-06,
+      "loss": 0.4641,
+      "step": 5442
+    },
+    {
+      "epoch": 0.4420531145943312,
+      "grad_norm": 4.8077643907663665,
+      "learning_rate": 3.085387733355637e-06,
+      "loss": 0.5179,
+      "step": 5443
+    },
+    {
+      "epoch": 0.4421343295703728,
+      "grad_norm": 5.021716165337083,
+      "learning_rate": 3.08474837332142e-06,
+      "loss": 0.6234,
+      "step": 5444
+    },
+    {
+      "epoch": 0.4422155445464144,
+      "grad_norm": 5.826683036908708,
+      "learning_rate": 3.0841089728257108e-06,
+      "loss": 0.5355,
+      "step": 5445
+    },
+    {
+      "epoch": 0.44229675952245595,
+      "grad_norm": 4.024099529077247,
+      "learning_rate": 3.0834695319127516e-06,
+      "loss": 0.5641,
+      "step": 5446
+    },
+    {
+      "epoch": 0.44237797449849753,
+      "grad_norm": 6.876907227052113,
+      "learning_rate": 3.082830050626789e-06,
+      "loss": 0.4901,
+      "step": 5447
+    },
+    {
+      "epoch": 0.4424591894745391,
+      "grad_norm": 5.144105316785434,
+      "learning_rate": 3.0821905290120712e-06,
+      "loss": 0.3913,
+      "step": 5448
+    },
+    {
+      "epoch": 0.4425404044505807,
+      "grad_norm": 3.975940531276462,
+      "learning_rate": 3.0815509671128506e-06,
+      "loss": 0.4229,
+      "step": 5449
+    },
+    {
+      "epoch": 0.44262161942662226,
+      "grad_norm": 5.375633378244454,
+      "learning_rate": 3.0809113649733803e-06,
+      "loss": 0.5113,
+      "step": 5450
+    },
+    {
+      "epoch": 0.44270283440266384,
+      "grad_norm": 12.651478454129343,
+      "learning_rate": 3.0802717226379175e-06,
+      "loss": 0.5075,
+      "step": 5451
+    },
+    {
+      "epoch": 0.4427840493787054,
+      "grad_norm": 4.831174765479994,
+      "learning_rate": 3.079632040150724e-06,
+      "loss": 0.5238,
+      "step": 5452
+    },
+    {
+      "epoch": 0.442865264354747,
+      "grad_norm": 5.064193661858532,
+      "learning_rate": 3.07899231755606e-06,
+      "loss": 0.5808,
+      "step": 5453
+    },
+    {
+      "epoch": 0.4429464793307886,
+      "grad_norm": 5.352945242031246,
+      "learning_rate": 3.0783525548981917e-06,
+      "loss": 0.4835,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4430276943068302,
+      "grad_norm": 4.965261328987609,
+      "learning_rate": 3.077712752221388e-06,
+      "loss": 0.4251,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4431089092828718,
+      "grad_norm": 4.9254182519569465,
+      "learning_rate": 3.0770729095699194e-06,
+      "loss": 0.6029,
+      "step": 5456
+    },
+    {
+      "epoch": 0.44319012425891335,
+      "grad_norm": 3.5617209008138992,
+      "learning_rate": 3.0764330269880593e-06,
+      "loss": 0.5718,
+      "step": 5457
+    },
+    {
+      "epoch": 0.44327133923495493,
+      "grad_norm": 4.355025485119006,
+      "learning_rate": 3.0757931045200844e-06,
+      "loss": 0.6286,
+      "step": 5458
+    },
+    {
+      "epoch": 0.4433525542109965,
+      "grad_norm": 6.239139168135333,
+      "learning_rate": 3.075153142210274e-06,
+      "loss": 0.5216,
+      "step": 5459
+    },
+    {
+      "epoch": 0.4434337691870381,
+      "grad_norm": 4.784319073015132,
+      "learning_rate": 3.0745131401029105e-06,
+      "loss": 0.5841,
+      "step": 5460
+    },
+    {
+      "epoch": 0.44351498416307966,
+      "grad_norm": 4.463298659345465,
+      "learning_rate": 3.073873098242278e-06,
+      "loss": 0.6327,
+      "step": 5461
+    },
+    {
+      "epoch": 0.44359619913912124,
+      "grad_norm": 6.8742335301264585,
+      "learning_rate": 3.0732330166726644e-06,
+      "loss": 0.5043,
+      "step": 5462
+    },
+    {
+      "epoch": 0.4436774141151628,
+      "grad_norm": 3.366707977466441,
+      "learning_rate": 3.07259289543836e-06,
+      "loss": 0.5506,
+      "step": 5463
+    },
+    {
+      "epoch": 0.4437586290912044,
+      "grad_norm": 4.220402613518687,
+      "learning_rate": 3.0719527345836568e-06,
+      "loss": 0.4197,
+      "step": 5464
+    },
+    {
+      "epoch": 0.443839844067246,
+      "grad_norm": 3.555095408073302,
+      "learning_rate": 3.0713125341528527e-06,
+      "loss": 0.4006,
+      "step": 5465
+    },
+    {
+      "epoch": 0.4439210590432876,
+      "grad_norm": 5.959143884494173,
+      "learning_rate": 3.0706722941902438e-06,
+      "loss": 0.5243,
+      "step": 5466
+    },
+    {
+      "epoch": 0.4440022740193292,
+      "grad_norm": 3.7731524034106707,
+      "learning_rate": 3.0700320147401324e-06,
+      "loss": 0.5004,
+      "step": 5467
+    },
+    {
+      "epoch": 0.44408348899537076,
+      "grad_norm": 4.782660700197082,
+      "learning_rate": 3.0693916958468236e-06,
+      "loss": 0.5797,
+      "step": 5468
+    },
+    {
+      "epoch": 0.44416470397141233,
+      "grad_norm": 4.37229758317183,
+      "learning_rate": 3.0687513375546216e-06,
+      "loss": 0.614,
+      "step": 5469
+    },
+    {
+      "epoch": 0.4442459189474539,
+      "grad_norm": 2.828225215491486,
+      "learning_rate": 3.0681109399078375e-06,
+      "loss": 0.7196,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4443271339234955,
+      "grad_norm": 6.309177247959491,
+      "learning_rate": 3.0674705029507833e-06,
+      "loss": 0.476,
+      "step": 5471
+    },
+    {
+      "epoch": 0.44440834889953706,
+      "grad_norm": 4.103834477568316,
+      "learning_rate": 3.0668300267277735e-06,
+      "loss": 0.6375,
+      "step": 5472
+    },
+    {
+      "epoch": 0.44448956387557864,
+      "grad_norm": 6.093953734178574,
+      "learning_rate": 3.066189511283126e-06,
+      "loss": 0.4653,
+      "step": 5473
+    },
+    {
+      "epoch": 0.4445707788516202,
+      "grad_norm": 5.225914425304106,
+      "learning_rate": 3.0655489566611603e-06,
+      "loss": 0.6781,
+      "step": 5474
+    },
+    {
+      "epoch": 0.4446519938276618,
+      "grad_norm": 5.8819081687868335,
+      "learning_rate": 3.0649083629062e-06,
+      "loss": 0.5596,
+      "step": 5475
+    },
+    {
+      "epoch": 0.44473320880370343,
+      "grad_norm": 7.739981641220584,
+      "learning_rate": 3.0642677300625704e-06,
+      "loss": 0.3823,
+      "step": 5476
+    },
+    {
+      "epoch": 0.444814423779745,
+      "grad_norm": 5.131722265818073,
+      "learning_rate": 3.063627058174601e-06,
+      "loss": 0.4458,
+      "step": 5477
+    },
+    {
+      "epoch": 0.4448956387557866,
+      "grad_norm": 3.2411651551006364,
+      "learning_rate": 3.062986347286622e-06,
+      "loss": 0.7122,
+      "step": 5478
+    },
+    {
+      "epoch": 0.44497685373182816,
+      "grad_norm": 4.225618329584794,
+      "learning_rate": 3.0623455974429677e-06,
+      "loss": 0.4025,
+      "step": 5479
+    },
+    {
+      "epoch": 0.44505806870786974,
+      "grad_norm": 7.665112766366524,
+      "learning_rate": 3.061704808687973e-06,
+      "loss": 0.6129,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4451392836839113,
+      "grad_norm": 9.07304807101099,
+      "learning_rate": 3.061063981065979e-06,
+      "loss": 0.5051,
+      "step": 5481
+    },
+    {
+      "epoch": 0.4452204986599529,
+      "grad_norm": 4.554173085391561,
+      "learning_rate": 3.0604231146213276e-06,
+      "loss": 0.6264,
+      "step": 5482
+    },
+    {
+      "epoch": 0.44530171363599447,
+      "grad_norm": 3.5414231796303515,
+      "learning_rate": 3.0597822093983614e-06,
+      "loss": 0.5022,
+      "step": 5483
+    },
+    {
+      "epoch": 0.44538292861203604,
+      "grad_norm": 8.230686535869143,
+      "learning_rate": 3.0591412654414297e-06,
+      "loss": 0.6486,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4454641435880776,
+      "grad_norm": 5.579450710250304,
+      "learning_rate": 3.058500282794882e-06,
+      "loss": 0.5926,
+      "step": 5485
+    },
+    {
+      "epoch": 0.4455453585641192,
+      "grad_norm": 6.0408658738191425,
+      "learning_rate": 3.0578592615030693e-06,
+      "loss": 0.3853,
+      "step": 5486
+    },
+    {
+      "epoch": 0.44562657354016083,
+      "grad_norm": 4.852372309573421,
+      "learning_rate": 3.057218201610349e-06,
+      "loss": 0.4482,
+      "step": 5487
+    },
+    {
+      "epoch": 0.4457077885162024,
+      "grad_norm": 5.300082736246172,
+      "learning_rate": 3.056577103161078e-06,
+      "loss": 0.5611,
+      "step": 5488
+    },
+    {
+      "epoch": 0.445789003492244,
+      "grad_norm": 5.314374108815902,
+      "learning_rate": 3.055935966199617e-06,
+      "loss": 0.505,
+      "step": 5489
+    },
+    {
+      "epoch": 0.44587021846828556,
+      "grad_norm": 3.462874288033993,
+      "learning_rate": 3.0552947907703296e-06,
+      "loss": 0.5165,
+      "step": 5490
+    },
+    {
+      "epoch": 0.44595143344432714,
+      "grad_norm": 5.895017788703498,
+      "learning_rate": 3.054653576917581e-06,
+      "loss": 0.3758,
+      "step": 5491
+    },
+    {
+      "epoch": 0.4460326484203687,
+      "grad_norm": 3.720905933238603,
+      "learning_rate": 3.054012324685742e-06,
+      "loss": 0.574,
+      "step": 5492
+    },
+    {
+      "epoch": 0.4461138633964103,
+      "grad_norm": 6.446011438422685,
+      "learning_rate": 3.05337103411918e-06,
+      "loss": 0.6723,
+      "step": 5493
+    },
+    {
+      "epoch": 0.44619507837245187,
+      "grad_norm": 4.27621399758535,
+      "learning_rate": 3.0527297052622724e-06,
+      "loss": 0.5515,
+      "step": 5494
+    },
+    {
+      "epoch": 0.44627629334849345,
+      "grad_norm": 14.21365870200448,
+      "learning_rate": 3.0520883381593945e-06,
+      "loss": 0.4223,
+      "step": 5495
+    },
+    {
+      "epoch": 0.446357508324535,
+      "grad_norm": 3.727959140755334,
+      "learning_rate": 3.0514469328549244e-06,
+      "loss": 0.7102,
+      "step": 5496
+    },
+    {
+      "epoch": 0.4464387233005766,
+      "grad_norm": 4.155096776178402,
+      "learning_rate": 3.050805489393246e-06,
+      "loss": 0.6996,
+      "step": 5497
+    },
+    {
+      "epoch": 0.44651993827661823,
+      "grad_norm": 3.376388149465253,
+      "learning_rate": 3.0501640078187433e-06,
+      "loss": 0.6374,
+      "step": 5498
+    },
+    {
+      "epoch": 0.4466011532526598,
+      "grad_norm": 6.291374977431427,
+      "learning_rate": 3.049522488175802e-06,
+      "loss": 0.5527,
+      "step": 5499
+    },
+    {
+      "epoch": 0.4466823682287014,
+      "grad_norm": 6.926029198647835,
+      "learning_rate": 3.048880930508813e-06,
+      "loss": 0.4501,
+      "step": 5500
+    },
+    {
+      "epoch": 0.44676358320474296,
+      "grad_norm": 8.251812163196625,
+      "learning_rate": 3.0482393348621686e-06,
+      "loss": 0.4361,
+      "step": 5501
+    },
+    {
+      "epoch": 0.44684479818078454,
+      "grad_norm": 4.312548651784504,
+      "learning_rate": 3.0475977012802636e-06,
+      "loss": 0.4884,
+      "step": 5502
+    },
+    {
+      "epoch": 0.4469260131568261,
+      "grad_norm": 8.316308066336365,
+      "learning_rate": 3.0469560298074963e-06,
+      "loss": 0.4457,
+      "step": 5503
+    },
+    {
+      "epoch": 0.4470072281328677,
+      "grad_norm": 5.919440004233395,
+      "learning_rate": 3.046314320488266e-06,
+      "loss": 0.5825,
+      "step": 5504
+    },
+    {
+      "epoch": 0.4470884431089093,
+      "grad_norm": 3.520589949942169,
+      "learning_rate": 3.045672573366976e-06,
+      "loss": 0.7033,
+      "step": 5505
+    },
+    {
+      "epoch": 0.44716965808495085,
+      "grad_norm": 5.061834295396976,
+      "learning_rate": 3.045030788488032e-06,
+      "loss": 0.5824,
+      "step": 5506
+    },
+    {
+      "epoch": 0.4472508730609924,
+      "grad_norm": 5.285749057766182,
+      "learning_rate": 3.0443889658958425e-06,
+      "loss": 0.5358,
+      "step": 5507
+    },
+    {
+      "epoch": 0.447332088037034,
+      "grad_norm": 6.668648467416671,
+      "learning_rate": 3.043747105634817e-06,
+      "loss": 0.4972,
+      "step": 5508
+    },
+    {
+      "epoch": 0.44741330301307564,
+      "grad_norm": 5.315769012630028,
+      "learning_rate": 3.0431052077493693e-06,
+      "loss": 0.7238,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4474945179891172,
+      "grad_norm": 4.114487698288283,
+      "learning_rate": 3.0424632722839164e-06,
+      "loss": 0.7628,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4475757329651588,
+      "grad_norm": 5.636091284907231,
+      "learning_rate": 3.041821299282876e-06,
+      "loss": 0.6014,
+      "step": 5511
+    },
+    {
+      "epoch": 0.44765694794120037,
+      "grad_norm": 5.436008606726467,
+      "learning_rate": 3.0411792887906684e-06,
+      "loss": 0.4414,
+      "step": 5512
+    },
+    {
+      "epoch": 0.44773816291724194,
+      "grad_norm": 5.179891754796302,
+      "learning_rate": 3.0405372408517187e-06,
+      "loss": 0.477,
+      "step": 5513
+    },
+    {
+      "epoch": 0.4478193778932835,
+      "grad_norm": 2.443055617371502,
+      "learning_rate": 3.0398951555104528e-06,
+      "loss": 0.6141,
+      "step": 5514
+    },
+    {
+      "epoch": 0.4479005928693251,
+      "grad_norm": 3.1073744907433745,
+      "learning_rate": 3.0392530328112997e-06,
+      "loss": 0.4829,
+      "step": 5515
+    },
+    {
+      "epoch": 0.4479818078453667,
+      "grad_norm": 10.2090255345531,
+      "learning_rate": 3.0386108727986903e-06,
+      "loss": 0.4541,
+      "step": 5516
+    },
+    {
+      "epoch": 0.44806302282140825,
+      "grad_norm": 5.848825993848949,
+      "learning_rate": 3.037968675517059e-06,
+      "loss": 0.4629,
+      "step": 5517
+    },
+    {
+      "epoch": 0.44814423779744983,
+      "grad_norm": 5.418590209098287,
+      "learning_rate": 3.0373264410108422e-06,
+      "loss": 0.512,
+      "step": 5518
+    },
+    {
+      "epoch": 0.4482254527734914,
+      "grad_norm": 4.498788774198856,
+      "learning_rate": 3.03668416932448e-06,
+      "loss": 0.4333,
+      "step": 5519
+    },
+    {
+      "epoch": 0.44830666774953304,
+      "grad_norm": 7.415061980682369,
+      "learning_rate": 3.0360418605024134e-06,
+      "loss": 0.4415,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4483878827255746,
+      "grad_norm": 4.0616647064171465,
+      "learning_rate": 3.0353995145890868e-06,
+      "loss": 0.6796,
+      "step": 5521
+    },
+    {
+      "epoch": 0.4484690977016162,
+      "grad_norm": 5.786539246564089,
+      "learning_rate": 3.0347571316289476e-06,
+      "loss": 0.4574,
+      "step": 5522
+    },
+    {
+      "epoch": 0.44855031267765777,
+      "grad_norm": 4.2099399908062445,
+      "learning_rate": 3.0341147116664455e-06,
+      "loss": 0.5367,
+      "step": 5523
+    },
+    {
+      "epoch": 0.44863152765369935,
+      "grad_norm": 4.538423749900379,
+      "learning_rate": 3.0334722547460317e-06,
+      "loss": 0.374,
+      "step": 5524
+    },
+    {
+      "epoch": 0.4487127426297409,
+      "grad_norm": 3.5387099780140137,
+      "learning_rate": 3.032829760912161e-06,
+      "loss": 0.4963,
+      "step": 5525
+    },
+    {
+      "epoch": 0.4487939576057825,
+      "grad_norm": 5.798358992841778,
+      "learning_rate": 3.032187230209291e-06,
+      "loss": 0.5552,
+      "step": 5526
+    },
+    {
+      "epoch": 0.4488751725818241,
+      "grad_norm": 4.773816159502634,
+      "learning_rate": 3.0315446626818816e-06,
+      "loss": 0.5549,
+      "step": 5527
+    },
+    {
+      "epoch": 0.44895638755786565,
+      "grad_norm": 4.150570620974484,
+      "learning_rate": 3.030902058374394e-06,
+      "loss": 0.5485,
+      "step": 5528
+    },
+    {
+      "epoch": 0.44903760253390723,
+      "grad_norm": 5.126013922801945,
+      "learning_rate": 3.0302594173312937e-06,
+      "loss": 0.4232,
+      "step": 5529
+    },
+    {
+      "epoch": 0.4491188175099488,
+      "grad_norm": 3.268316271468154,
+      "learning_rate": 3.0296167395970494e-06,
+      "loss": 0.5323,
+      "step": 5530
+    },
+    {
+      "epoch": 0.44920003248599044,
+      "grad_norm": 5.730548961152452,
+      "learning_rate": 3.0289740252161288e-06,
+      "loss": 0.6075,
+      "step": 5531
+    },
+    {
+      "epoch": 0.449281247462032,
+      "grad_norm": 7.380827782342893,
+      "learning_rate": 3.0283312742330044e-06,
+      "loss": 0.5684,
+      "step": 5532
+    },
+    {
+      "epoch": 0.4493624624380736,
+      "grad_norm": 3.0968839249512983,
+      "learning_rate": 3.027688486692153e-06,
+      "loss": 0.6726,
+      "step": 5533
+    },
+    {
+      "epoch": 0.4494436774141152,
+      "grad_norm": 3.139400489756557,
+      "learning_rate": 3.027045662638051e-06,
+      "loss": 0.5246,
+      "step": 5534
+    },
+    {
+      "epoch": 0.44952489239015675,
+      "grad_norm": 5.977877338701136,
+      "learning_rate": 3.026402802115178e-06,
+      "loss": 0.4677,
+      "step": 5535
+    },
+    {
+      "epoch": 0.4496061073661983,
+      "grad_norm": 4.58739972834906,
+      "learning_rate": 3.0257599051680175e-06,
+      "loss": 0.4622,
+      "step": 5536
+    },
+    {
+      "epoch": 0.4496873223422399,
+      "grad_norm": 6.184504437794764,
+      "learning_rate": 3.025116971841054e-06,
+      "loss": 0.6765,
+      "step": 5537
+    },
+    {
+      "epoch": 0.4497685373182815,
+      "grad_norm": 8.256944628820975,
+      "learning_rate": 3.0244740021787756e-06,
+      "loss": 0.5273,
+      "step": 5538
+    },
+    {
+      "epoch": 0.44984975229432306,
+      "grad_norm": 5.127185634205651,
+      "learning_rate": 3.023830996225671e-06,
+      "loss": 0.5202,
+      "step": 5539
+    },
+    {
+      "epoch": 0.44993096727036463,
+      "grad_norm": 5.783243249616182,
+      "learning_rate": 3.023187954026234e-06,
+      "loss": 0.588,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4500121822464062,
+      "grad_norm": 11.669014789188214,
+      "learning_rate": 3.0225448756249605e-06,
+      "loss": 0.5103,
+      "step": 5541
+    },
+    {
+      "epoch": 0.45009339722244784,
+      "grad_norm": 12.31667513372858,
+      "learning_rate": 3.0219017610663466e-06,
+      "loss": 0.6313,
+      "step": 5542
+    },
+    {
+      "epoch": 0.4501746121984894,
+      "grad_norm": 4.705587679569699,
+      "learning_rate": 3.0212586103948933e-06,
+      "loss": 0.5984,
+      "step": 5543
+    },
+    {
+      "epoch": 0.450255827174531,
+      "grad_norm": 7.657296950948716,
+      "learning_rate": 3.020615423655102e-06,
+      "loss": 0.4608,
+      "step": 5544
+    },
+    {
+      "epoch": 0.4503370421505726,
+      "grad_norm": 4.260385063187866,
+      "learning_rate": 3.0199722008914787e-06,
+      "loss": 0.6098,
+      "step": 5545
+    },
+    {
+      "epoch": 0.45041825712661415,
+      "grad_norm": 5.405032331240298,
+      "learning_rate": 3.0193289421485317e-06,
+      "loss": 0.3907,
+      "step": 5546
+    },
+    {
+      "epoch": 0.45049947210265573,
+      "grad_norm": 5.755862088555348,
+      "learning_rate": 3.0186856474707705e-06,
+      "loss": 0.4768,
+      "step": 5547
+    },
+    {
+      "epoch": 0.4505806870786973,
+      "grad_norm": 8.389991068275073,
+      "learning_rate": 3.0180423169027067e-06,
+      "loss": 0.4632,
+      "step": 5548
+    },
+    {
+      "epoch": 0.4506619020547389,
+      "grad_norm": 7.221227795632766,
+      "learning_rate": 3.0173989504888573e-06,
+      "loss": 0.4662,
+      "step": 5549
+    },
+    {
+      "epoch": 0.45074311703078046,
+      "grad_norm": 3.6157966531181467,
+      "learning_rate": 3.0167555482737384e-06,
+      "loss": 0.6383,
+      "step": 5550
+    },
+    {
+      "epoch": 0.45082433200682204,
+      "grad_norm": 5.5152149011394185,
+      "learning_rate": 3.01611211030187e-06,
+      "loss": 0.3908,
+      "step": 5551
+    },
+    {
+      "epoch": 0.4509055469828636,
+      "grad_norm": 4.18620086470904,
+      "learning_rate": 3.0154686366177753e-06,
+      "loss": 0.4462,
+      "step": 5552
+    },
+    {
+      "epoch": 0.45098676195890525,
+      "grad_norm": 5.069886109587496,
+      "learning_rate": 3.0148251272659795e-06,
+      "loss": 0.4242,
+      "step": 5553
+    },
+    {
+      "epoch": 0.4510679769349468,
+      "grad_norm": 4.107158289610819,
+      "learning_rate": 3.0141815822910094e-06,
+      "loss": 0.5341,
+      "step": 5554
+    },
+    {
+      "epoch": 0.4511491919109884,
+      "grad_norm": 4.955579663758407,
+      "learning_rate": 3.013538001737395e-06,
+      "loss": 0.561,
+      "step": 5555
+    },
+    {
+      "epoch": 0.45123040688703,
+      "grad_norm": 4.9599952397922635,
+      "learning_rate": 3.0128943856496686e-06,
+      "loss": 0.6114,
+      "step": 5556
+    },
+    {
+      "epoch": 0.45131162186307155,
+      "grad_norm": 8.882975709525924,
+      "learning_rate": 3.0122507340723656e-06,
+      "loss": 0.6247,
+      "step": 5557
+    },
+    {
+      "epoch": 0.45139283683911313,
+      "grad_norm": 5.7173513452837135,
+      "learning_rate": 3.011607047050022e-06,
+      "loss": 0.4825,
+      "step": 5558
+    },
+    {
+      "epoch": 0.4514740518151547,
+      "grad_norm": 6.378003204165053,
+      "learning_rate": 3.0109633246271783e-06,
+      "loss": 0.4608,
+      "step": 5559
+    },
+    {
+      "epoch": 0.4515552667911963,
+      "grad_norm": 8.215078154867541,
+      "learning_rate": 3.0103195668483787e-06,
+      "loss": 0.4853,
+      "step": 5560
+    },
+    {
+      "epoch": 0.45163648176723786,
+      "grad_norm": 5.6937027402469305,
+      "learning_rate": 3.009675773758164e-06,
+      "loss": 0.5338,
+      "step": 5561
+    },
+    {
+      "epoch": 0.45171769674327944,
+      "grad_norm": 9.181599736624747,
+      "learning_rate": 3.009031945401084e-06,
+      "loss": 0.4927,
+      "step": 5562
+    },
+    {
+      "epoch": 0.451798911719321,
+      "grad_norm": 11.154194461173983,
+      "learning_rate": 3.008388081821687e-06,
+      "loss": 0.5629,
+      "step": 5563
+    },
+    {
+      "epoch": 0.45188012669536265,
+      "grad_norm": 8.09877927664861,
+      "learning_rate": 3.0077441830645256e-06,
+      "loss": 0.6261,
+      "step": 5564
+    },
+    {
+      "epoch": 0.4519613416714042,
+      "grad_norm": 6.075046325765384,
+      "learning_rate": 3.0071002491741537e-06,
+      "loss": 0.5261,
+      "step": 5565
+    },
+    {
+      "epoch": 0.4520425566474458,
+      "grad_norm": 7.918146195582449,
+      "learning_rate": 3.0064562801951286e-06,
+      "loss": 0.4545,
+      "step": 5566
+    },
+    {
+      "epoch": 0.4521237716234874,
+      "grad_norm": 4.644487714982916,
+      "learning_rate": 3.005812276172009e-06,
+      "loss": 0.5743,
+      "step": 5567
+    },
+    {
+      "epoch": 0.45220498659952896,
+      "grad_norm": 3.5759922555457924,
+      "learning_rate": 3.005168237149357e-06,
+      "loss": 0.572,
+      "step": 5568
+    },
+    {
+      "epoch": 0.45228620157557053,
+      "grad_norm": 6.775625726215614,
+      "learning_rate": 3.0045241631717366e-06,
+      "loss": 0.5287,
+      "step": 5569
+    },
+    {
+      "epoch": 0.4523674165516121,
+      "grad_norm": 3.7500419034479457,
+      "learning_rate": 3.0038800542837137e-06,
+      "loss": 0.5277,
+      "step": 5570
+    },
+    {
+      "epoch": 0.4524486315276537,
+      "grad_norm": 4.670720788070081,
+      "learning_rate": 3.003235910529859e-06,
+      "loss": 0.4724,
+      "step": 5571
+    },
+    {
+      "epoch": 0.45252984650369527,
+      "grad_norm": 7.410732606803197,
+      "learning_rate": 3.0025917319547417e-06,
+      "loss": 0.6593,
+      "step": 5572
+    },
+    {
+      "epoch": 0.45261106147973684,
+      "grad_norm": 7.469496813543684,
+      "learning_rate": 3.001947518602937e-06,
+      "loss": 0.4562,
+      "step": 5573
+    },
+    {
+      "epoch": 0.4526922764557784,
+      "grad_norm": 4.482461581886912,
+      "learning_rate": 3.0013032705190196e-06,
+      "loss": 0.5644,
+      "step": 5574
+    },
+    {
+      "epoch": 0.45277349143182005,
+      "grad_norm": 4.5986439128006,
+      "learning_rate": 3.00065898774757e-06,
+      "loss": 0.5426,
+      "step": 5575
+    },
+    {
+      "epoch": 0.45285470640786163,
+      "grad_norm": 4.16197803163674,
+      "learning_rate": 3.000014670333168e-06,
+      "loss": 0.5278,
+      "step": 5576
+    },
+    {
+      "epoch": 0.4529359213839032,
+      "grad_norm": 3.5560084963087646,
+      "learning_rate": 2.9993703183203963e-06,
+      "loss": 0.5385,
+      "step": 5577
+    },
+    {
+      "epoch": 0.4530171363599448,
+      "grad_norm": 4.260872328467781,
+      "learning_rate": 2.998725931753842e-06,
+      "loss": 0.449,
+      "step": 5578
+    },
+    {
+      "epoch": 0.45309835133598636,
+      "grad_norm": 4.8192798552612635,
+      "learning_rate": 2.9980815106780937e-06,
+      "loss": 0.605,
+      "step": 5579
+    },
+    {
+      "epoch": 0.45317956631202794,
+      "grad_norm": 3.949873091272697,
+      "learning_rate": 2.9974370551377396e-06,
+      "loss": 0.5863,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4532607812880695,
+      "grad_norm": 8.728645349811258,
+      "learning_rate": 2.9967925651773745e-06,
+      "loss": 0.4891,
+      "step": 5581
+    },
+    {
+      "epoch": 0.4533419962641111,
+      "grad_norm": 3.623472869472515,
+      "learning_rate": 2.9961480408415926e-06,
+      "loss": 0.6374,
+      "step": 5582
+    },
+    {
+      "epoch": 0.45342321124015267,
+      "grad_norm": 4.651906826052307,
+      "learning_rate": 2.995503482174993e-06,
+      "loss": 0.4461,
+      "step": 5583
+    },
+    {
+      "epoch": 0.45350442621619425,
+      "grad_norm": 4.256628817272692,
+      "learning_rate": 2.9948588892221744e-06,
+      "loss": 0.4862,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4535856411922358,
+      "grad_norm": 3.3916680603373797,
+      "learning_rate": 2.9942142620277394e-06,
+      "loss": 0.5934,
+      "step": 5585
+    },
+    {
+      "epoch": 0.45366685616827745,
+      "grad_norm": 6.635234044978131,
+      "learning_rate": 2.993569600636293e-06,
+      "loss": 0.5009,
+      "step": 5586
+    },
+    {
+      "epoch": 0.45374807114431903,
+      "grad_norm": 5.742703156297073,
+      "learning_rate": 2.9929249050924424e-06,
+      "loss": 0.4523,
+      "step": 5587
+    },
+    {
+      "epoch": 0.4538292861203606,
+      "grad_norm": 4.30344798988703,
+      "learning_rate": 2.992280175440797e-06,
+      "loss": 0.5456,
+      "step": 5588
+    },
+    {
+      "epoch": 0.4539105010964022,
+      "grad_norm": 4.517980897014619,
+      "learning_rate": 2.99163541172597e-06,
+      "loss": 0.4736,
+      "step": 5589
+    },
+    {
+      "epoch": 0.45399171607244376,
+      "grad_norm": 5.224517211850113,
+      "learning_rate": 2.990990613992573e-06,
+      "loss": 0.6306,
+      "step": 5590
+    },
+    {
+      "epoch": 0.45407293104848534,
+      "grad_norm": 4.522124261787193,
+      "learning_rate": 2.990345782285225e-06,
+      "loss": 0.5577,
+      "step": 5591
+    },
+    {
+      "epoch": 0.4541541460245269,
+      "grad_norm": 4.777313229511017,
+      "learning_rate": 2.989700916648544e-06,
+      "loss": 0.4167,
+      "step": 5592
+    },
+    {
+      "epoch": 0.4542353610005685,
+      "grad_norm": 6.094818732275617,
+      "learning_rate": 2.989056017127151e-06,
+      "loss": 0.4435,
+      "step": 5593
+    },
+    {
+      "epoch": 0.45431657597661007,
+      "grad_norm": 7.062295268512343,
+      "learning_rate": 2.988411083765669e-06,
+      "loss": 0.5327,
+      "step": 5594
+    },
+    {
+      "epoch": 0.45439779095265165,
+      "grad_norm": 5.878383502169189,
+      "learning_rate": 2.9877661166087265e-06,
+      "loss": 0.7565,
+      "step": 5595
+    },
+    {
+      "epoch": 0.4544790059286932,
+      "grad_norm": 5.252077496243823,
+      "learning_rate": 2.9871211157009496e-06,
+      "loss": 0.3842,
+      "step": 5596
+    },
+    {
+      "epoch": 0.45456022090473486,
+      "grad_norm": 4.8053252445722014,
+      "learning_rate": 2.986476081086969e-06,
+      "loss": 0.5382,
+      "step": 5597
+    },
+    {
+      "epoch": 0.45464143588077643,
+      "grad_norm": 5.29723639288913,
+      "learning_rate": 2.9858310128114187e-06,
+      "loss": 0.5417,
+      "step": 5598
+    },
+    {
+      "epoch": 0.454722650856818,
+      "grad_norm": 4.021237021928457,
+      "learning_rate": 2.9851859109189335e-06,
+      "loss": 0.54,
+      "step": 5599
+    },
+    {
+      "epoch": 0.4548038658328596,
+      "grad_norm": 26.465640352460827,
+      "learning_rate": 2.9845407754541513e-06,
+      "loss": 0.4593,
+      "step": 5600
+    },
+    {
+      "epoch": 0.45488508080890117,
+      "grad_norm": 4.300973423387741,
+      "learning_rate": 2.9838956064617108e-06,
+      "loss": 0.547,
+      "step": 5601
+    },
+    {
+      "epoch": 0.45496629578494274,
+      "grad_norm": 4.972732875615428,
+      "learning_rate": 2.9832504039862564e-06,
+      "loss": 0.5216,
+      "step": 5602
+    },
+    {
+      "epoch": 0.4550475107609843,
+      "grad_norm": 3.8151254239492074,
+      "learning_rate": 2.982605168072431e-06,
+      "loss": 0.5508,
+      "step": 5603
+    },
+    {
+      "epoch": 0.4551287257370259,
+      "grad_norm": 4.251587340007891,
+      "learning_rate": 2.981959898764882e-06,
+      "loss": 0.4318,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4552099407130675,
+      "grad_norm": 4.436180317192108,
+      "learning_rate": 2.9813145961082594e-06,
+      "loss": 0.5097,
+      "step": 5605
+    },
+    {
+      "epoch": 0.45529115568910905,
+      "grad_norm": 4.5049604397224785,
+      "learning_rate": 2.9806692601472143e-06,
+      "loss": 0.6497,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4553723706651506,
+      "grad_norm": 4.817837180058034,
+      "learning_rate": 2.9800238909263994e-06,
+      "loss": 0.6143,
+      "step": 5607
+    },
+    {
+      "epoch": 0.45545358564119226,
+      "grad_norm": 3.4243930089969785,
+      "learning_rate": 2.9793784884904733e-06,
+      "loss": 0.5063,
+      "step": 5608
+    },
+    {
+      "epoch": 0.45553480061723384,
+      "grad_norm": 5.925953178663676,
+      "learning_rate": 2.9787330528840915e-06,
+      "loss": 0.4992,
+      "step": 5609
+    },
+    {
+      "epoch": 0.4556160155932754,
+      "grad_norm": 5.497205043778295,
+      "learning_rate": 2.978087584151915e-06,
+      "loss": 0.5372,
+      "step": 5610
+    },
+    {
+      "epoch": 0.455697230569317,
+      "grad_norm": 8.33939404348421,
+      "learning_rate": 2.9774420823386104e-06,
+      "loss": 0.5619,
+      "step": 5611
+    },
+    {
+      "epoch": 0.45577844554535857,
+      "grad_norm": 6.195975570900406,
+      "learning_rate": 2.9767965474888395e-06,
+      "loss": 0.5683,
+      "step": 5612
+    },
+    {
+      "epoch": 0.45585966052140015,
+      "grad_norm": 14.703831195583096,
+      "learning_rate": 2.9761509796472697e-06,
+      "loss": 0.3429,
+      "step": 5613
+    },
+    {
+      "epoch": 0.4559408754974417,
+      "grad_norm": 4.631555776984385,
+      "learning_rate": 2.975505378858574e-06,
+      "loss": 0.6409,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4560220904734833,
+      "grad_norm": 3.4809174977075803,
+      "learning_rate": 2.974859745167422e-06,
+      "loss": 0.769,
+      "step": 5615
+    },
+    {
+      "epoch": 0.4561033054495249,
+      "grad_norm": 5.098517264596913,
+      "learning_rate": 2.9742140786184885e-06,
+      "loss": 0.5612,
+      "step": 5616
+    },
+    {
+      "epoch": 0.45618452042556645,
+      "grad_norm": 6.00392072759688,
+      "learning_rate": 2.9735683792564506e-06,
+      "loss": 0.5576,
+      "step": 5617
+    },
+    {
+      "epoch": 0.45626573540160803,
+      "grad_norm": 7.263834884088008,
+      "learning_rate": 2.9729226471259877e-06,
+      "loss": 0.6709,
+      "step": 5618
+    },
+    {
+      "epoch": 0.45634695037764966,
+      "grad_norm": 3.554328673137164,
+      "learning_rate": 2.9722768822717795e-06,
+      "loss": 0.6453,
+      "step": 5619
+    },
+    {
+      "epoch": 0.45642816535369124,
+      "grad_norm": 5.778288795426547,
+      "learning_rate": 2.971631084738511e-06,
+      "loss": 0.4381,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4565093803297328,
+      "grad_norm": 3.465143813973219,
+      "learning_rate": 2.9709852545708677e-06,
+      "loss": 0.5124,
+      "step": 5621
+    },
+    {
+      "epoch": 0.4565905953057744,
+      "grad_norm": 4.670081929190185,
+      "learning_rate": 2.9703393918135383e-06,
+      "loss": 0.4459,
+      "step": 5622
+    },
+    {
+      "epoch": 0.45667181028181597,
+      "grad_norm": 4.713194770845547,
+      "learning_rate": 2.96969349651121e-06,
+      "loss": 0.4847,
+      "step": 5623
+    },
+    {
+      "epoch": 0.45675302525785755,
+      "grad_norm": 3.630426739892531,
+      "learning_rate": 2.9690475687085795e-06,
+      "loss": 0.4093,
+      "step": 5624
+    },
+    {
+      "epoch": 0.4568342402338991,
+      "grad_norm": 5.486371575732508,
+      "learning_rate": 2.968401608450339e-06,
+      "loss": 0.6108,
+      "step": 5625
+    },
+    {
+      "epoch": 0.4569154552099407,
+      "grad_norm": 8.575804166016505,
+      "learning_rate": 2.967755615781186e-06,
+      "loss": 0.45,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4569966701859823,
+      "grad_norm": 8.699853217920039,
+      "learning_rate": 2.9671095907458203e-06,
+      "loss": 0.5099,
+      "step": 5627
+    },
+    {
+      "epoch": 0.45707788516202386,
+      "grad_norm": 3.801057038330998,
+      "learning_rate": 2.966463533388943e-06,
+      "loss": 0.5301,
+      "step": 5628
+    },
+    {
+      "epoch": 0.45715910013806543,
+      "grad_norm": 6.044511448542259,
+      "learning_rate": 2.9658174437552577e-06,
+      "loss": 0.5236,
+      "step": 5629
+    },
+    {
+      "epoch": 0.45724031511410707,
+      "grad_norm": 5.875396782939016,
+      "learning_rate": 2.9651713218894706e-06,
+      "loss": 0.6417,
+      "step": 5630
+    },
+    {
+      "epoch": 0.45732153009014864,
+      "grad_norm": 4.000366715370684,
+      "learning_rate": 2.96452516783629e-06,
+      "loss": 0.5186,
+      "step": 5631
+    },
+    {
+      "epoch": 0.4574027450661902,
+      "grad_norm": 4.69919331836541,
+      "learning_rate": 2.9638789816404264e-06,
+      "loss": 0.5162,
+      "step": 5632
+    },
+    {
+      "epoch": 0.4574839600422318,
+      "grad_norm": 7.3603896088494185,
+      "learning_rate": 2.9632327633465917e-06,
+      "loss": 0.4416,
+      "step": 5633
+    },
+    {
+      "epoch": 0.4575651750182734,
+      "grad_norm": 7.25776019771969,
+      "learning_rate": 2.9625865129995023e-06,
+      "loss": 0.5364,
+      "step": 5634
+    },
+    {
+      "epoch": 0.45764638999431495,
+      "grad_norm": 5.174876153569623,
+      "learning_rate": 2.9619402306438738e-06,
+      "loss": 0.412,
+      "step": 5635
+    },
+    {
+      "epoch": 0.4577276049703565,
+      "grad_norm": 5.4010502742692355,
+      "learning_rate": 2.9612939163244266e-06,
+      "loss": 0.4053,
+      "step": 5636
+    },
+    {
+      "epoch": 0.4578088199463981,
+      "grad_norm": 3.835214497106897,
+      "learning_rate": 2.960647570085881e-06,
+      "loss": 0.4996,
+      "step": 5637
+    },
+    {
+      "epoch": 0.4578900349224397,
+      "grad_norm": 6.055794052072197,
+      "learning_rate": 2.960001191972963e-06,
+      "loss": 0.7144,
+      "step": 5638
+    },
+    {
+      "epoch": 0.45797124989848126,
+      "grad_norm": 4.158855522188884,
+      "learning_rate": 2.9593547820303954e-06,
+      "loss": 0.6862,
+      "step": 5639
+    },
+    {
+      "epoch": 0.45805246487452284,
+      "grad_norm": 3.708348870716642,
+      "learning_rate": 2.958708340302908e-06,
+      "loss": 0.5555,
+      "step": 5640
+    },
+    {
+      "epoch": 0.45813367985056447,
+      "grad_norm": 5.983420776955866,
+      "learning_rate": 2.958061866835232e-06,
+      "loss": 0.4188,
+      "step": 5641
+    },
+    {
+      "epoch": 0.45821489482660605,
+      "grad_norm": 8.29550387074976,
+      "learning_rate": 2.9574153616720986e-06,
+      "loss": 0.5177,
+      "step": 5642
+    },
+    {
+      "epoch": 0.4582961098026476,
+      "grad_norm": 5.444764698582431,
+      "learning_rate": 2.9567688248582436e-06,
+      "loss": 0.5237,
+      "step": 5643
+    },
+    {
+      "epoch": 0.4583773247786892,
+      "grad_norm": 10.29419590630684,
+      "learning_rate": 2.956122256438403e-06,
+      "loss": 0.4966,
+      "step": 5644
+    },
+    {
+      "epoch": 0.4584585397547308,
+      "grad_norm": 6.484747843668005,
+      "learning_rate": 2.955475656457316e-06,
+      "loss": 0.5255,
+      "step": 5645
+    },
+    {
+      "epoch": 0.45853975473077235,
+      "grad_norm": 5.119025976889396,
+      "learning_rate": 2.9548290249597246e-06,
+      "loss": 0.5482,
+      "step": 5646
+    },
+    {
+      "epoch": 0.45862096970681393,
+      "grad_norm": 6.402280776009877,
+      "learning_rate": 2.9541823619903716e-06,
+      "loss": 0.6823,
+      "step": 5647
+    },
+    {
+      "epoch": 0.4587021846828555,
+      "grad_norm": 7.6633301689603375,
+      "learning_rate": 2.9535356675940023e-06,
+      "loss": 0.4238,
+      "step": 5648
+    },
+    {
+      "epoch": 0.4587833996588971,
+      "grad_norm": 4.171651081928115,
+      "learning_rate": 2.952888941815366e-06,
+      "loss": 0.6524,
+      "step": 5649
+    },
+    {
+      "epoch": 0.45886461463493866,
+      "grad_norm": 5.133593022664716,
+      "learning_rate": 2.952242184699211e-06,
+      "loss": 0.5472,
+      "step": 5650
+    },
+    {
+      "epoch": 0.45894582961098024,
+      "grad_norm": 5.9594049521898205,
+      "learning_rate": 2.9515953962902914e-06,
+      "loss": 0.5447,
+      "step": 5651
+    },
+    {
+      "epoch": 0.45902704458702187,
+      "grad_norm": 6.3931523146679785,
+      "learning_rate": 2.950948576633359e-06,
+      "loss": 0.5729,
+      "step": 5652
+    },
+    {
+      "epoch": 0.45910825956306345,
+      "grad_norm": 5.8713619633879,
+      "learning_rate": 2.9503017257731727e-06,
+      "loss": 0.4821,
+      "step": 5653
+    },
+    {
+      "epoch": 0.459189474539105,
+      "grad_norm": 3.2931039167346743,
+      "learning_rate": 2.9496548437544905e-06,
+      "loss": 0.4733,
+      "step": 5654
+    },
+    {
+      "epoch": 0.4592706895151466,
+      "grad_norm": 5.1941445369327335,
+      "learning_rate": 2.9490079306220714e-06,
+      "loss": 0.3829,
+      "step": 5655
+    },
+    {
+      "epoch": 0.4593519044911882,
+      "grad_norm": 4.360689715599596,
+      "learning_rate": 2.9483609864206808e-06,
+      "loss": 0.5581,
+      "step": 5656
+    },
+    {
+      "epoch": 0.45943311946722976,
+      "grad_norm": 7.1909194878489195,
+      "learning_rate": 2.9477140111950834e-06,
+      "loss": 0.4832,
+      "step": 5657
+    },
+    {
+      "epoch": 0.45951433444327133,
+      "grad_norm": 5.716565178813941,
+      "learning_rate": 2.947067004990045e-06,
+      "loss": 0.458,
+      "step": 5658
+    },
+    {
+      "epoch": 0.4595955494193129,
+      "grad_norm": 5.12640741752529,
+      "learning_rate": 2.9464199678503364e-06,
+      "loss": 0.5096,
+      "step": 5659
+    },
+    {
+      "epoch": 0.4596767643953545,
+      "grad_norm": 4.73475880030873,
+      "learning_rate": 2.9457728998207286e-06,
+      "loss": 0.46,
+      "step": 5660
+    },
+    {
+      "epoch": 0.45975797937139606,
+      "grad_norm": 5.497100257560182,
+      "learning_rate": 2.9451258009459947e-06,
+      "loss": 0.4947,
+      "step": 5661
+    },
+    {
+      "epoch": 0.45983919434743764,
+      "grad_norm": 3.188631639464191,
+      "learning_rate": 2.9444786712709122e-06,
+      "loss": 0.4633,
+      "step": 5662
+    },
+    {
+      "epoch": 0.4599204093234793,
+      "grad_norm": 6.625940309443984,
+      "learning_rate": 2.943831510840257e-06,
+      "loss": 0.5155,
+      "step": 5663
+    },
+    {
+      "epoch": 0.46000162429952085,
+      "grad_norm": 6.021653211612759,
+      "learning_rate": 2.9431843196988107e-06,
+      "loss": 0.4657,
+      "step": 5664
+    },
+    {
+      "epoch": 0.4600828392755624,
+      "grad_norm": 4.027902151297978,
+      "learning_rate": 2.942537097891355e-06,
+      "loss": 0.5745,
+      "step": 5665
+    },
+    {
+      "epoch": 0.460164054251604,
+      "grad_norm": 4.556130080376428,
+      "learning_rate": 2.9418898454626744e-06,
+      "loss": 0.7196,
+      "step": 5666
+    },
+    {
+      "epoch": 0.4602452692276456,
+      "grad_norm": 7.887493386006087,
+      "learning_rate": 2.9412425624575553e-06,
+      "loss": 0.5629,
+      "step": 5667
+    },
+    {
+      "epoch": 0.46032648420368716,
+      "grad_norm": 11.085374599142877,
+      "learning_rate": 2.9405952489207858e-06,
+      "loss": 0.4118,
+      "step": 5668
+    },
+    {
+      "epoch": 0.46040769917972874,
+      "grad_norm": 5.492625782917754,
+      "learning_rate": 2.9399479048971567e-06,
+      "loss": 0.4832,
+      "step": 5669
+    },
+    {
+      "epoch": 0.4604889141557703,
+      "grad_norm": 3.5056589438634216,
+      "learning_rate": 2.939300530431462e-06,
+      "loss": 0.5736,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4605701291318119,
+      "grad_norm": 4.465558544811287,
+      "learning_rate": 2.9386531255684942e-06,
+      "loss": 0.7238,
+      "step": 5671
+    },
+    {
+      "epoch": 0.46065134410785347,
+      "grad_norm": 48.57316377462239,
+      "learning_rate": 2.938005690353052e-06,
+      "loss": 0.5249,
+      "step": 5672
+    },
+    {
+      "epoch": 0.46073255908389504,
+      "grad_norm": 5.242926405538982,
+      "learning_rate": 2.937358224829935e-06,
+      "loss": 0.5339,
+      "step": 5673
+    },
+    {
+      "epoch": 0.4608137740599367,
+      "grad_norm": 5.8937374649076695,
+      "learning_rate": 2.936710729043943e-06,
+      "loss": 0.6889,
+      "step": 5674
+    },
+    {
+      "epoch": 0.46089498903597825,
+      "grad_norm": 5.462741459326538,
+      "learning_rate": 2.936063203039879e-06,
+      "loss": 0.5676,
+      "step": 5675
+    },
+    {
+      "epoch": 0.46097620401201983,
+      "grad_norm": 6.268297125768787,
+      "learning_rate": 2.93541564686255e-06,
+      "loss": 0.526,
+      "step": 5676
+    },
+    {
+      "epoch": 0.4610574189880614,
+      "grad_norm": 3.614454345003604,
+      "learning_rate": 2.9347680605567624e-06,
+      "loss": 0.4487,
+      "step": 5677
+    },
+    {
+      "epoch": 0.461138633964103,
+      "grad_norm": 8.297993020626132,
+      "learning_rate": 2.9341204441673267e-06,
+      "loss": 0.4531,
+      "step": 5678
+    },
+    {
+      "epoch": 0.46121984894014456,
+      "grad_norm": 4.364678727568973,
+      "learning_rate": 2.9334727977390526e-06,
+      "loss": 0.6027,
+      "step": 5679
+    },
+    {
+      "epoch": 0.46130106391618614,
+      "grad_norm": 8.596844992148577,
+      "learning_rate": 2.9328251213167557e-06,
+      "loss": 0.5818,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4613822788922277,
+      "grad_norm": 4.6174846839314005,
+      "learning_rate": 2.9321774149452507e-06,
+      "loss": 0.4268,
+      "step": 5681
+    },
+    {
+      "epoch": 0.4614634938682693,
+      "grad_norm": 10.308065553433666,
+      "learning_rate": 2.9315296786693564e-06,
+      "loss": 0.5332,
+      "step": 5682
+    },
+    {
+      "epoch": 0.46154470884431087,
+      "grad_norm": 7.029864284315326,
+      "learning_rate": 2.9308819125338923e-06,
+      "loss": 0.6391,
+      "step": 5683
+    },
+    {
+      "epoch": 0.46162592382035245,
+      "grad_norm": 6.141294742654424,
+      "learning_rate": 2.9302341165836794e-06,
+      "loss": 0.7152,
+      "step": 5684
+    },
+    {
+      "epoch": 0.4617071387963941,
+      "grad_norm": 5.578226095205441,
+      "learning_rate": 2.9295862908635436e-06,
+      "loss": 0.5101,
+      "step": 5685
+    },
+    {
+      "epoch": 0.46178835377243566,
+      "grad_norm": 5.083560208470007,
+      "learning_rate": 2.92893843541831e-06,
+      "loss": 0.5236,
+      "step": 5686
+    },
+    {
+      "epoch": 0.46186956874847723,
+      "grad_norm": 3.7370396960027654,
+      "learning_rate": 2.928290550292806e-06,
+      "loss": 0.6088,
+      "step": 5687
+    },
+    {
+      "epoch": 0.4619507837245188,
+      "grad_norm": 8.921482639240043,
+      "learning_rate": 2.9276426355318625e-06,
+      "loss": 0.4895,
+      "step": 5688
+    },
+    {
+      "epoch": 0.4620319987005604,
+      "grad_norm": 3.2133566714283055,
+      "learning_rate": 2.9269946911803134e-06,
+      "loss": 0.5798,
+      "step": 5689
+    },
+    {
+      "epoch": 0.46211321367660196,
+      "grad_norm": 6.204381262030037,
+      "learning_rate": 2.92634671728299e-06,
+      "loss": 0.4417,
+      "step": 5690
+    },
+    {
+      "epoch": 0.46219442865264354,
+      "grad_norm": 5.691308583371654,
+      "learning_rate": 2.9256987138847302e-06,
+      "loss": 0.5512,
+      "step": 5691
+    },
+    {
+      "epoch": 0.4622756436286851,
+      "grad_norm": 4.319078154150966,
+      "learning_rate": 2.925050681030373e-06,
+      "loss": 0.5221,
+      "step": 5692
+    },
+    {
+      "epoch": 0.4623568586047267,
+      "grad_norm": 5.04058444742641,
+      "learning_rate": 2.9244026187647584e-06,
+      "loss": 0.647,
+      "step": 5693
+    },
+    {
+      "epoch": 0.46243807358076827,
+      "grad_norm": 5.620845349439251,
+      "learning_rate": 2.923754527132728e-06,
+      "loss": 0.4942,
+      "step": 5694
+    },
+    {
+      "epoch": 0.46251928855680985,
+      "grad_norm": 5.022513745576825,
+      "learning_rate": 2.9231064061791277e-06,
+      "loss": 0.3824,
+      "step": 5695
+    },
+    {
+      "epoch": 0.4626005035328515,
+      "grad_norm": 4.5284930571538835,
+      "learning_rate": 2.922458255948803e-06,
+      "loss": 0.4568,
+      "step": 5696
+    },
+    {
+      "epoch": 0.46268171850889306,
+      "grad_norm": 4.600857423481818,
+      "learning_rate": 2.9218100764866025e-06,
+      "loss": 0.6157,
+      "step": 5697
+    },
+    {
+      "epoch": 0.46276293348493464,
+      "grad_norm": 3.265626698109701,
+      "learning_rate": 2.9211618678373775e-06,
+      "loss": 0.4392,
+      "step": 5698
+    },
+    {
+      "epoch": 0.4628441484609762,
+      "grad_norm": 4.624889936593594,
+      "learning_rate": 2.9205136300459803e-06,
+      "loss": 0.463,
+      "step": 5699
+    },
+    {
+      "epoch": 0.4629253634370178,
+      "grad_norm": 5.8554370873391655,
+      "learning_rate": 2.919865363157265e-06,
+      "loss": 0.6943,
+      "step": 5700
+    },
+    {
+      "epoch": 0.46300657841305937,
+      "grad_norm": 10.564403475889055,
+      "learning_rate": 2.9192170672160892e-06,
+      "loss": 0.5849,
+      "step": 5701
+    },
+    {
+      "epoch": 0.46308779338910094,
+      "grad_norm": 4.570004688540801,
+      "learning_rate": 2.9185687422673103e-06,
+      "loss": 0.5091,
+      "step": 5702
+    },
+    {
+      "epoch": 0.4631690083651425,
+      "grad_norm": 7.370321832435026,
+      "learning_rate": 2.917920388355791e-06,
+      "loss": 0.4578,
+      "step": 5703
+    },
+    {
+      "epoch": 0.4632502233411841,
+      "grad_norm": 4.488088245228176,
+      "learning_rate": 2.9172720055263916e-06,
+      "loss": 0.3637,
+      "step": 5704
+    },
+    {
+      "epoch": 0.4633314383172257,
+      "grad_norm": 3.3278944590303086,
+      "learning_rate": 2.9166235938239785e-06,
+      "loss": 0.4415,
+      "step": 5705
+    },
+    {
+      "epoch": 0.46341265329326725,
+      "grad_norm": 25.965744402127452,
+      "learning_rate": 2.9159751532934165e-06,
+      "loss": 0.4605,
+      "step": 5706
+    },
+    {
+      "epoch": 0.4634938682693089,
+      "grad_norm": 5.292291969576549,
+      "learning_rate": 2.9153266839795756e-06,
+      "loss": 0.5707,
+      "step": 5707
+    },
+    {
+      "epoch": 0.46357508324535046,
+      "grad_norm": 7.540226421886441,
+      "learning_rate": 2.9146781859273276e-06,
+      "loss": 0.4891,
+      "step": 5708
+    },
+    {
+      "epoch": 0.46365629822139204,
+      "grad_norm": 3.896646180404308,
+      "learning_rate": 2.9140296591815425e-06,
+      "loss": 0.5333,
+      "step": 5709
+    },
+    {
+      "epoch": 0.4637375131974336,
+      "grad_norm": 5.700465799524542,
+      "learning_rate": 2.913381103787097e-06,
+      "loss": 0.391,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4638187281734752,
+      "grad_norm": 4.798705492265267,
+      "learning_rate": 2.9127325197888663e-06,
+      "loss": 0.404,
+      "step": 5711
+    },
+    {
+      "epoch": 0.46389994314951677,
+      "grad_norm": 3.5471558345306877,
+      "learning_rate": 2.91208390723173e-06,
+      "loss": 0.7544,
+      "step": 5712
+    },
+    {
+      "epoch": 0.46398115812555835,
+      "grad_norm": 6.891850689361664,
+      "learning_rate": 2.911435266160568e-06,
+      "loss": 0.4801,
+      "step": 5713
+    },
+    {
+      "epoch": 0.4640623731015999,
+      "grad_norm": 4.753158541053688,
+      "learning_rate": 2.910786596620263e-06,
+      "loss": 0.4848,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4641435880776415,
+      "grad_norm": 5.463566847148066,
+      "learning_rate": 2.9101378986556996e-06,
+      "loss": 0.4833,
+      "step": 5715
+    },
+    {
+      "epoch": 0.4642248030536831,
+      "grad_norm": 3.542984250968721,
+      "learning_rate": 2.909489172311765e-06,
+      "loss": 0.7555,
+      "step": 5716
+    },
+    {
+      "epoch": 0.46430601802972465,
+      "grad_norm": 4.081018198125554,
+      "learning_rate": 2.9088404176333456e-06,
+      "loss": 0.5655,
+      "step": 5717
+    },
+    {
+      "epoch": 0.4643872330057663,
+      "grad_norm": 5.0004125127548,
+      "learning_rate": 2.9081916346653333e-06,
+      "loss": 0.6157,
+      "step": 5718
+    },
+    {
+      "epoch": 0.46446844798180786,
+      "grad_norm": 6.309500154267012,
+      "learning_rate": 2.9075428234526215e-06,
+      "loss": 0.404,
+      "step": 5719
+    },
+    {
+      "epoch": 0.46454966295784944,
+      "grad_norm": 5.199891325095606,
+      "learning_rate": 2.9068939840401018e-06,
+      "loss": 0.4213,
+      "step": 5720
+    },
+    {
+      "epoch": 0.464630877933891,
+      "grad_norm": 3.3927116646573685,
+      "learning_rate": 2.906245116472672e-06,
+      "loss": 0.5366,
+      "step": 5721
+    },
+    {
+      "epoch": 0.4647120929099326,
+      "grad_norm": 5.355679239024366,
+      "learning_rate": 2.905596220795231e-06,
+      "loss": 0.5221,
+      "step": 5722
+    },
+    {
+      "epoch": 0.46479330788597417,
+      "grad_norm": 4.733947923123343,
+      "learning_rate": 2.9049472970526777e-06,
+      "loss": 0.49,
+      "step": 5723
+    },
+    {
+      "epoch": 0.46487452286201575,
+      "grad_norm": 6.624090170841801,
+      "learning_rate": 2.904298345289914e-06,
+      "loss": 0.4846,
+      "step": 5724
+    },
+    {
+      "epoch": 0.4649557378380573,
+      "grad_norm": 5.72885175850521,
+      "learning_rate": 2.9036493655518456e-06,
+      "loss": 0.4319,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4650369528140989,
+      "grad_norm": 4.332865010791888,
+      "learning_rate": 2.9030003578833765e-06,
+      "loss": 0.5975,
+      "step": 5726
+    },
+    {
+      "epoch": 0.4651181677901405,
+      "grad_norm": 6.313800965150897,
+      "learning_rate": 2.902351322329416e-06,
+      "loss": 0.5159,
+      "step": 5727
+    },
+    {
+      "epoch": 0.46519938276618206,
+      "grad_norm": 5.227382270145744,
+      "learning_rate": 2.9017022589348733e-06,
+      "loss": 0.393,
+      "step": 5728
+    },
+    {
+      "epoch": 0.4652805977422237,
+      "grad_norm": 2.840558709344795,
+      "learning_rate": 2.9010531677446602e-06,
+      "loss": 0.52,
+      "step": 5729
+    },
+    {
+      "epoch": 0.46536181271826527,
+      "grad_norm": 5.366773140432362,
+      "learning_rate": 2.90040404880369e-06,
+      "loss": 0.506,
+      "step": 5730
+    },
+    {
+      "epoch": 0.46544302769430684,
+      "grad_norm": 3.2313831572145193,
+      "learning_rate": 2.8997549021568792e-06,
+      "loss": 0.6582,
+      "step": 5731
+    },
+    {
+      "epoch": 0.4655242426703484,
+      "grad_norm": 5.926467612756465,
+      "learning_rate": 2.899105727849145e-06,
+      "loss": 0.4414,
+      "step": 5732
+    },
+    {
+      "epoch": 0.46560545764639,
+      "grad_norm": 4.447511140460557,
+      "learning_rate": 2.898456525925406e-06,
+      "loss": 0.54,
+      "step": 5733
+    },
+    {
+      "epoch": 0.4656866726224316,
+      "grad_norm": 5.303134033820951,
+      "learning_rate": 2.8978072964305848e-06,
+      "loss": 0.4579,
+      "step": 5734
+    },
+    {
+      "epoch": 0.46576788759847315,
+      "grad_norm": 5.195062102266287,
+      "learning_rate": 2.8971580394096043e-06,
+      "loss": 0.4434,
+      "step": 5735
+    },
+    {
+      "epoch": 0.46584910257451473,
+      "grad_norm": 5.8469679642289165,
+      "learning_rate": 2.896508754907389e-06,
+      "loss": 0.5036,
+      "step": 5736
+    },
+    {
+      "epoch": 0.4659303175505563,
+      "grad_norm": 3.484649611013429,
+      "learning_rate": 2.8958594429688656e-06,
+      "loss": 0.483,
+      "step": 5737
+    },
+    {
+      "epoch": 0.4660115325265979,
+      "grad_norm": 7.276763839964852,
+      "learning_rate": 2.895210103638966e-06,
+      "loss": 0.5142,
+      "step": 5738
+    },
+    {
+      "epoch": 0.46609274750263946,
+      "grad_norm": 4.316468383561267,
+      "learning_rate": 2.894560736962617e-06,
+      "loss": 0.6067,
+      "step": 5739
+    },
+    {
+      "epoch": 0.4661739624786811,
+      "grad_norm": 3.8962363928693184,
+      "learning_rate": 2.893911342984754e-06,
+      "loss": 0.7335,
+      "step": 5740
+    },
+    {
+      "epoch": 0.46625517745472267,
+      "grad_norm": 4.124297053875989,
+      "learning_rate": 2.89326192175031e-06,
+      "loss": 0.4862,
+      "step": 5741
+    },
+    {
+      "epoch": 0.46633639243076425,
+      "grad_norm": 5.37302630861064,
+      "learning_rate": 2.8926124733042228e-06,
+      "loss": 0.3562,
+      "step": 5742
+    },
+    {
+      "epoch": 0.4664176074068058,
+      "grad_norm": 5.046203124192214,
+      "learning_rate": 2.89196299769143e-06,
+      "loss": 0.7901,
+      "step": 5743
+    },
+    {
+      "epoch": 0.4664988223828474,
+      "grad_norm": 8.053971161102101,
+      "learning_rate": 2.8913134949568726e-06,
+      "loss": 0.5199,
+      "step": 5744
+    },
+    {
+      "epoch": 0.466580037358889,
+      "grad_norm": 3.327749003210284,
+      "learning_rate": 2.890663965145492e-06,
+      "loss": 0.5486,
+      "step": 5745
+    },
+    {
+      "epoch": 0.46666125233493055,
+      "grad_norm": 5.0353868986052674,
+      "learning_rate": 2.890014408302233e-06,
+      "loss": 0.6335,
+      "step": 5746
+    },
+    {
+      "epoch": 0.46674246731097213,
+      "grad_norm": 6.139502965392141,
+      "learning_rate": 2.8893648244720406e-06,
+      "loss": 0.4263,
+      "step": 5747
+    },
+    {
+      "epoch": 0.4668236822870137,
+      "grad_norm": 5.019570855865613,
+      "learning_rate": 2.8887152136998644e-06,
+      "loss": 0.5657,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4669048972630553,
+      "grad_norm": 16.504398142409077,
+      "learning_rate": 2.8880655760306507e-06,
+      "loss": 0.3836,
+      "step": 5749
+    },
+    {
+      "epoch": 0.46698611223909686,
+      "grad_norm": 5.525848295027518,
+      "learning_rate": 2.887415911509354e-06,
+      "loss": 0.6306,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4670673272151385,
+      "grad_norm": 4.288458640459229,
+      "learning_rate": 2.8867662201809266e-06,
+      "loss": 0.6479,
+      "step": 5751
+    },
+    {
+      "epoch": 0.46714854219118007,
+      "grad_norm": 8.844918615051897,
+      "learning_rate": 2.8861165020903235e-06,
+      "loss": 0.5476,
+      "step": 5752
+    },
+    {
+      "epoch": 0.46722975716722165,
+      "grad_norm": 8.708089067131917,
+      "learning_rate": 2.8854667572825013e-06,
+      "loss": 0.4022,
+      "step": 5753
+    },
+    {
+      "epoch": 0.4673109721432632,
+      "grad_norm": 3.9421995828081897,
+      "learning_rate": 2.8848169858024206e-06,
+      "loss": 0.4588,
+      "step": 5754
+    },
+    {
+      "epoch": 0.4673921871193048,
+      "grad_norm": 3.8977256522289374,
+      "learning_rate": 2.8841671876950404e-06,
+      "loss": 0.4142,
+      "step": 5755
+    },
+    {
+      "epoch": 0.4674734020953464,
+      "grad_norm": 8.084593014494807,
+      "learning_rate": 2.8835173630053244e-06,
+      "loss": 0.4468,
+      "step": 5756
+    },
+    {
+      "epoch": 0.46755461707138796,
+      "grad_norm": 3.533484576984449,
+      "learning_rate": 2.882867511778237e-06,
+      "loss": 0.5784,
+      "step": 5757
+    },
+    {
+      "epoch": 0.46763583204742953,
+      "grad_norm": 4.263459871167963,
+      "learning_rate": 2.8822176340587434e-06,
+      "loss": 0.6078,
+      "step": 5758
+    },
+    {
+      "epoch": 0.4677170470234711,
+      "grad_norm": 3.9845148250837865,
+      "learning_rate": 2.881567729891812e-06,
+      "loss": 0.4972,
+      "step": 5759
+    },
+    {
+      "epoch": 0.4677982619995127,
+      "grad_norm": 7.062828717835998,
+      "learning_rate": 2.8809177993224143e-06,
+      "loss": 0.5564,
+      "step": 5760
+    },
+    {
+      "epoch": 0.46787947697555426,
+      "grad_norm": 4.435936639131757,
+      "learning_rate": 2.88026784239552e-06,
+      "loss": 0.573,
+      "step": 5761
+    },
+    {
+      "epoch": 0.4679606919515959,
+      "grad_norm": 4.512414174366236,
+      "learning_rate": 2.8796178591561035e-06,
+      "loss": 0.4828,
+      "step": 5762
+    },
+    {
+      "epoch": 0.4680419069276375,
+      "grad_norm": 5.094918568491417,
+      "learning_rate": 2.8789678496491407e-06,
+      "loss": 0.5475,
+      "step": 5763
+    },
+    {
+      "epoch": 0.46812312190367905,
+      "grad_norm": 2.6728500983692514,
+      "learning_rate": 2.878317813919608e-06,
+      "loss": 0.5026,
+      "step": 5764
+    },
+    {
+      "epoch": 0.46820433687972063,
+      "grad_norm": 4.7597544395019495,
+      "learning_rate": 2.877667752012485e-06,
+      "loss": 0.7266,
+      "step": 5765
+    },
+    {
+      "epoch": 0.4682855518557622,
+      "grad_norm": 6.8641010734258,
+      "learning_rate": 2.877017663972752e-06,
+      "loss": 0.4941,
+      "step": 5766
+    },
+    {
+      "epoch": 0.4683667668318038,
+      "grad_norm": 5.927018173381962,
+      "learning_rate": 2.876367549845393e-06,
+      "loss": 0.5093,
+      "step": 5767
+    },
+    {
+      "epoch": 0.46844798180784536,
+      "grad_norm": 3.4289610085754383,
+      "learning_rate": 2.875717409675391e-06,
+      "loss": 0.5454,
+      "step": 5768
+    },
+    {
+      "epoch": 0.46852919678388694,
+      "grad_norm": 7.9437461607693,
+      "learning_rate": 2.875067243507732e-06,
+      "loss": 0.4647,
+      "step": 5769
+    },
+    {
+      "epoch": 0.4686104117599285,
+      "grad_norm": 3.7116341834744078,
+      "learning_rate": 2.8744170513874054e-06,
+      "loss": 0.4881,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4686916267359701,
+      "grad_norm": 4.2367362028156,
+      "learning_rate": 2.8737668333594005e-06,
+      "loss": 0.4672,
+      "step": 5771
+    },
+    {
+      "epoch": 0.46877284171201167,
+      "grad_norm": 4.273237686231666,
+      "learning_rate": 2.873116589468708e-06,
+      "loss": 0.4611,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4688540566880533,
+      "grad_norm": 5.096615588091853,
+      "learning_rate": 2.872466319760323e-06,
+      "loss": 0.5954,
+      "step": 5773
+    },
+    {
+      "epoch": 0.4689352716640949,
+      "grad_norm": 4.825210645748332,
+      "learning_rate": 2.87181602427924e-06,
+      "loss": 0.5718,
+      "step": 5774
+    },
+    {
+      "epoch": 0.46901648664013645,
+      "grad_norm": 5.991461149755983,
+      "learning_rate": 2.8711657030704553e-06,
+      "loss": 0.4037,
+      "step": 5775
+    },
+    {
+      "epoch": 0.46909770161617803,
+      "grad_norm": 3.4800097173343776,
+      "learning_rate": 2.870515356178969e-06,
+      "loss": 0.513,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4691789165922196,
+      "grad_norm": 6.256528064000865,
+      "learning_rate": 2.8698649836497805e-06,
+      "loss": 0.6655,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4692601315682612,
+      "grad_norm": 5.04158800857517,
+      "learning_rate": 2.869214585527893e-06,
+      "loss": 0.4901,
+      "step": 5778
+    },
+    {
+      "epoch": 0.46934134654430276,
+      "grad_norm": 4.617317169571152,
+      "learning_rate": 2.8685641618583098e-06,
+      "loss": 0.4887,
+      "step": 5779
+    },
+    {
+      "epoch": 0.46942256152034434,
+      "grad_norm": 3.9855464372411147,
+      "learning_rate": 2.8679137126860373e-06,
+      "loss": 0.647,
+      "step": 5780
+    },
+    {
+      "epoch": 0.4695037764963859,
+      "grad_norm": 4.5232125393874965,
+      "learning_rate": 2.867263238056084e-06,
+      "loss": 0.5514,
+      "step": 5781
+    },
+    {
+      "epoch": 0.4695849914724275,
+      "grad_norm": 6.069972056894211,
+      "learning_rate": 2.866612738013457e-06,
+      "loss": 0.6096,
+      "step": 5782
+    },
+    {
+      "epoch": 0.46966620644846907,
+      "grad_norm": 4.326041253791683,
+      "learning_rate": 2.8659622126031687e-06,
+      "loss": 0.5519,
+      "step": 5783
+    },
+    {
+      "epoch": 0.4697474214245107,
+      "grad_norm": 6.02872157760923,
+      "learning_rate": 2.8653116618702338e-06,
+      "loss": 0.5394,
+      "step": 5784
+    },
+    {
+      "epoch": 0.4698286364005523,
+      "grad_norm": 12.764955573129805,
+      "learning_rate": 2.8646610858596635e-06,
+      "loss": 0.5991,
+      "step": 5785
+    },
+    {
+      "epoch": 0.46990985137659386,
+      "grad_norm": 5.345811501835503,
+      "learning_rate": 2.864010484616477e-06,
+      "loss": 0.54,
+      "step": 5786
+    },
+    {
+      "epoch": 0.46999106635263543,
+      "grad_norm": 4.1825362641291415,
+      "learning_rate": 2.8633598581856915e-06,
+      "loss": 0.515,
+      "step": 5787
+    },
+    {
+      "epoch": 0.470072281328677,
+      "grad_norm": 4.283001560216305,
+      "learning_rate": 2.8627092066123263e-06,
+      "loss": 0.4081,
+      "step": 5788
+    },
+    {
+      "epoch": 0.4701534963047186,
+      "grad_norm": 4.308589294597862,
+      "learning_rate": 2.8620585299414038e-06,
+      "loss": 0.5021,
+      "step": 5789
+    },
+    {
+      "epoch": 0.47023471128076016,
+      "grad_norm": 8.01361742978466,
+      "learning_rate": 2.861407828217947e-06,
+      "loss": 0.4754,
+      "step": 5790
+    },
+    {
+      "epoch": 0.47031592625680174,
+      "grad_norm": 5.074919316163974,
+      "learning_rate": 2.8607571014869816e-06,
+      "loss": 0.5081,
+      "step": 5791
+    },
+    {
+      "epoch": 0.4703971412328433,
+      "grad_norm": 4.9471885543147325,
+      "learning_rate": 2.860106349793534e-06,
+      "loss": 0.6083,
+      "step": 5792
+    },
+    {
+      "epoch": 0.4704783562088849,
+      "grad_norm": 5.520036465193203,
+      "learning_rate": 2.859455573182632e-06,
+      "loss": 0.5181,
+      "step": 5793
+    },
+    {
+      "epoch": 0.4705595711849265,
+      "grad_norm": 8.857086777181047,
+      "learning_rate": 2.8588047716993084e-06,
+      "loss": 0.402,
+      "step": 5794
+    },
+    {
+      "epoch": 0.4706407861609681,
+      "grad_norm": 5.525749548650858,
+      "learning_rate": 2.858153945388592e-06,
+      "loss": 0.4228,
+      "step": 5795
+    },
+    {
+      "epoch": 0.4707220011370097,
+      "grad_norm": 4.365954557012976,
+      "learning_rate": 2.8575030942955185e-06,
+      "loss": 0.3335,
+      "step": 5796
+    },
+    {
+      "epoch": 0.47080321611305126,
+      "grad_norm": 5.357132843700224,
+      "learning_rate": 2.856852218465124e-06,
+      "loss": 0.5517,
+      "step": 5797
+    },
+    {
+      "epoch": 0.47088443108909284,
+      "grad_norm": 5.2190595114312694,
+      "learning_rate": 2.856201317942443e-06,
+      "loss": 0.4705,
+      "step": 5798
+    },
+    {
+      "epoch": 0.4709656460651344,
+      "grad_norm": 3.424227721385115,
+      "learning_rate": 2.8555503927725164e-06,
+      "loss": 0.4894,
+      "step": 5799
+    },
+    {
+      "epoch": 0.471046861041176,
+      "grad_norm": 5.0569787088343325,
+      "learning_rate": 2.854899443000385e-06,
+      "loss": 0.4604,
+      "step": 5800
+    },
+    {
+      "epoch": 0.47112807601721757,
+      "grad_norm": 8.187150772848494,
+      "learning_rate": 2.8542484686710896e-06,
+      "loss": 0.5102,
+      "step": 5801
+    },
+    {
+      "epoch": 0.47120929099325914,
+      "grad_norm": 6.207010087185806,
+      "learning_rate": 2.8535974698296765e-06,
+      "loss": 0.3528,
+      "step": 5802
+    },
+    {
+      "epoch": 0.4712905059693007,
+      "grad_norm": 4.227888017544308,
+      "learning_rate": 2.8529464465211886e-06,
+      "loss": 0.6394,
+      "step": 5803
+    },
+    {
+      "epoch": 0.4713717209453423,
+      "grad_norm": 4.665116991045132,
+      "learning_rate": 2.852295398790675e-06,
+      "loss": 0.6159,
+      "step": 5804
+    },
+    {
+      "epoch": 0.4714529359213839,
+      "grad_norm": 3.8703424664055857,
+      "learning_rate": 2.8516443266831837e-06,
+      "loss": 0.673,
+      "step": 5805
+    },
+    {
+      "epoch": 0.4715341508974255,
+      "grad_norm": 3.9676189612736388,
+      "learning_rate": 2.8509932302437665e-06,
+      "loss": 0.5117,
+      "step": 5806
+    },
+    {
+      "epoch": 0.4716153658734671,
+      "grad_norm": 4.329273160183401,
+      "learning_rate": 2.850342109517475e-06,
+      "loss": 0.3692,
+      "step": 5807
+    },
+    {
+      "epoch": 0.47169658084950866,
+      "grad_norm": 11.340360099111553,
+      "learning_rate": 2.8496909645493642e-06,
+      "loss": 0.4246,
+      "step": 5808
+    },
+    {
+      "epoch": 0.47177779582555024,
+      "grad_norm": 3.8942256837186626,
+      "learning_rate": 2.849039795384489e-06,
+      "loss": 0.5581,
+      "step": 5809
+    },
+    {
+      "epoch": 0.4718590108015918,
+      "grad_norm": 9.155900656113834,
+      "learning_rate": 2.8483886020679075e-06,
+      "loss": 0.4519,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4719402257776334,
+      "grad_norm": 6.0033962365746465,
+      "learning_rate": 2.847737384644678e-06,
+      "loss": 0.5612,
+      "step": 5811
+    },
+    {
+      "epoch": 0.47202144075367497,
+      "grad_norm": 5.889887605386142,
+      "learning_rate": 2.8470861431598623e-06,
+      "loss": 0.6559,
+      "step": 5812
+    },
+    {
+      "epoch": 0.47210265572971655,
+      "grad_norm": 4.599377492190549,
+      "learning_rate": 2.8464348776585234e-06,
+      "loss": 0.635,
+      "step": 5813
+    },
+    {
+      "epoch": 0.4721838707057581,
+      "grad_norm": 10.76890029919621,
+      "learning_rate": 2.8457835881857227e-06,
+      "loss": 0.4829,
+      "step": 5814
+    },
+    {
+      "epoch": 0.4722650856817997,
+      "grad_norm": 7.220464203607489,
+      "learning_rate": 2.8451322747865286e-06,
+      "loss": 0.5789,
+      "step": 5815
+    },
+    {
+      "epoch": 0.47234630065784133,
+      "grad_norm": 4.1108394338985725,
+      "learning_rate": 2.844480937506008e-06,
+      "loss": 0.609,
+      "step": 5816
+    },
+    {
+      "epoch": 0.4724275156338829,
+      "grad_norm": 9.04222984053817,
+      "learning_rate": 2.843829576389229e-06,
+      "loss": 0.4926,
+      "step": 5817
+    },
+    {
+      "epoch": 0.4725087306099245,
+      "grad_norm": 4.528427577254921,
+      "learning_rate": 2.843178191481263e-06,
+      "loss": 0.6624,
+      "step": 5818
+    },
+    {
+      "epoch": 0.47258994558596606,
+      "grad_norm": 3.3028903657064825,
+      "learning_rate": 2.842526782827183e-06,
+      "loss": 0.5929,
+      "step": 5819
+    },
+    {
+      "epoch": 0.47267116056200764,
+      "grad_norm": 5.026903257750425,
+      "learning_rate": 2.841875350472062e-06,
+      "loss": 0.5268,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4727523755380492,
+      "grad_norm": 6.5104574690823975,
+      "learning_rate": 2.841223894460976e-06,
+      "loss": 0.6567,
+      "step": 5821
+    },
+    {
+      "epoch": 0.4728335905140908,
+      "grad_norm": 4.314065979269395,
+      "learning_rate": 2.8405724148390023e-06,
+      "loss": 0.449,
+      "step": 5822
+    },
+    {
+      "epoch": 0.4729148054901324,
+      "grad_norm": 3.312866780207277,
+      "learning_rate": 2.8399209116512204e-06,
+      "loss": 0.5964,
+      "step": 5823
+    },
+    {
+      "epoch": 0.47299602046617395,
+      "grad_norm": 5.37935898304279,
+      "learning_rate": 2.83926938494271e-06,
+      "loss": 0.5702,
+      "step": 5824
+    },
+    {
+      "epoch": 0.4730772354422155,
+      "grad_norm": 4.881465697522925,
+      "learning_rate": 2.838617834758554e-06,
+      "loss": 0.4697,
+      "step": 5825
+    },
+    {
+      "epoch": 0.4731584504182571,
+      "grad_norm": 5.744126603133613,
+      "learning_rate": 2.8379662611438356e-06,
+      "loss": 0.4512,
+      "step": 5826
+    },
+    {
+      "epoch": 0.47323966539429874,
+      "grad_norm": 5.291309755595148,
+      "learning_rate": 2.8373146641436413e-06,
+      "loss": 0.5276,
+      "step": 5827
+    },
+    {
+      "epoch": 0.4733208803703403,
+      "grad_norm": 3.9537639542161407,
+      "learning_rate": 2.836663043803057e-06,
+      "loss": 0.5143,
+      "step": 5828
+    },
+    {
+      "epoch": 0.4734020953463819,
+      "grad_norm": 7.745428674706201,
+      "learning_rate": 2.8360114001671724e-06,
+      "loss": 0.4687,
+      "step": 5829
+    },
+    {
+      "epoch": 0.47348331032242347,
+      "grad_norm": 4.510709091006437,
+      "learning_rate": 2.835359733281077e-06,
+      "loss": 0.4782,
+      "step": 5830
+    },
+    {
+      "epoch": 0.47356452529846504,
+      "grad_norm": 7.03651123334896,
+      "learning_rate": 2.834708043189862e-06,
+      "loss": 0.4816,
+      "step": 5831
+    },
+    {
+      "epoch": 0.4736457402745066,
+      "grad_norm": 4.581264549358578,
+      "learning_rate": 2.8340563299386226e-06,
+      "loss": 0.3454,
+      "step": 5832
+    },
+    {
+      "epoch": 0.4737269552505482,
+      "grad_norm": 3.39063784376629,
+      "learning_rate": 2.833404593572453e-06,
+      "loss": 0.3696,
+      "step": 5833
+    },
+    {
+      "epoch": 0.4738081702265898,
+      "grad_norm": 5.734013115764664,
+      "learning_rate": 2.832752834136449e-06,
+      "loss": 0.4435,
+      "step": 5834
+    },
+    {
+      "epoch": 0.47388938520263135,
+      "grad_norm": 4.343417166849715,
+      "learning_rate": 2.832101051675712e-06,
+      "loss": 0.6581,
+      "step": 5835
+    },
+    {
+      "epoch": 0.47397060017867293,
+      "grad_norm": 3.661535617789465,
+      "learning_rate": 2.8314492462353386e-06,
+      "loss": 0.5384,
+      "step": 5836
+    },
+    {
+      "epoch": 0.4740518151547145,
+      "grad_norm": 4.785958660433438,
+      "learning_rate": 2.8307974178604312e-06,
+      "loss": 0.5199,
+      "step": 5837
+    },
+    {
+      "epoch": 0.47413303013075614,
+      "grad_norm": 7.777693529299737,
+      "learning_rate": 2.830145566596094e-06,
+      "loss": 0.5311,
+      "step": 5838
+    },
+    {
+      "epoch": 0.4742142451067977,
+      "grad_norm": 6.275994532015221,
+      "learning_rate": 2.8294936924874304e-06,
+      "loss": 0.6261,
+      "step": 5839
+    },
+    {
+      "epoch": 0.4742954600828393,
+      "grad_norm": 5.198840400244198,
+      "learning_rate": 2.8288417955795476e-06,
+      "loss": 0.5628,
+      "step": 5840
+    },
+    {
+      "epoch": 0.47437667505888087,
+      "grad_norm": 3.7789894977053615,
+      "learning_rate": 2.828189875917553e-06,
+      "loss": 0.6039,
+      "step": 5841
+    },
+    {
+      "epoch": 0.47445789003492245,
+      "grad_norm": 4.356995519309679,
+      "learning_rate": 2.827537933546555e-06,
+      "loss": 0.6307,
+      "step": 5842
+    },
+    {
+      "epoch": 0.474539105010964,
+      "grad_norm": 4.052004440671419,
+      "learning_rate": 2.8268859685116663e-06,
+      "loss": 0.5372,
+      "step": 5843
+    },
+    {
+      "epoch": 0.4746203199870056,
+      "grad_norm": 3.622092711482535,
+      "learning_rate": 2.826233980857998e-06,
+      "loss": 0.534,
+      "step": 5844
+    },
+    {
+      "epoch": 0.4747015349630472,
+      "grad_norm": 6.265338779708121,
+      "learning_rate": 2.8255819706306653e-06,
+      "loss": 0.4377,
+      "step": 5845
+    },
+    {
+      "epoch": 0.47478274993908876,
+      "grad_norm": 5.902804627865366,
+      "learning_rate": 2.8249299378747833e-06,
+      "loss": 0.3812,
+      "step": 5846
+    },
+    {
+      "epoch": 0.47486396491513033,
+      "grad_norm": 7.988753927468825,
+      "learning_rate": 2.824277882635469e-06,
+      "loss": 0.5673,
+      "step": 5847
+    },
+    {
+      "epoch": 0.4749451798911719,
+      "grad_norm": 10.87808274886899,
+      "learning_rate": 2.8236258049578418e-06,
+      "loss": 0.5389,
+      "step": 5848
+    },
+    {
+      "epoch": 0.47502639486721354,
+      "grad_norm": 5.350053974435865,
+      "learning_rate": 2.8229737048870216e-06,
+      "loss": 0.4301,
+      "step": 5849
+    },
+    {
+      "epoch": 0.4751076098432551,
+      "grad_norm": 5.918757588156862,
+      "learning_rate": 2.8223215824681295e-06,
+      "loss": 0.5695,
+      "step": 5850
+    },
+    {
+      "epoch": 0.4751888248192967,
+      "grad_norm": 12.49565829789524,
+      "learning_rate": 2.821669437746291e-06,
+      "loss": 0.6008,
+      "step": 5851
+    },
+    {
+      "epoch": 0.4752700397953383,
+      "grad_norm": 7.165842116453641,
+      "learning_rate": 2.8210172707666296e-06,
+      "loss": 0.5504,
+      "step": 5852
+    },
+    {
+      "epoch": 0.47535125477137985,
+      "grad_norm": 5.786887937045634,
+      "learning_rate": 2.820365081574271e-06,
+      "loss": 0.5192,
+      "step": 5853
+    },
+    {
+      "epoch": 0.4754324697474214,
+      "grad_norm": 6.157943347201467,
+      "learning_rate": 2.819712870214345e-06,
+      "loss": 0.5769,
+      "step": 5854
+    },
+    {
+      "epoch": 0.475513684723463,
+      "grad_norm": 15.746368000393133,
+      "learning_rate": 2.8190606367319806e-06,
+      "loss": 0.4668,
+      "step": 5855
+    },
+    {
+      "epoch": 0.4755948996995046,
+      "grad_norm": 7.868624147224612,
+      "learning_rate": 2.8184083811723083e-06,
+      "loss": 0.4084,
+      "step": 5856
+    },
+    {
+      "epoch": 0.47567611467554616,
+      "grad_norm": 16.215616004236644,
+      "learning_rate": 2.817756103580461e-06,
+      "loss": 0.588,
+      "step": 5857
+    },
+    {
+      "epoch": 0.47575732965158773,
+      "grad_norm": 4.747406090572309,
+      "learning_rate": 2.8171038040015737e-06,
+      "loss": 0.5907,
+      "step": 5858
+    },
+    {
+      "epoch": 0.4758385446276293,
+      "grad_norm": 4.9449474953603305,
+      "learning_rate": 2.8164514824807814e-06,
+      "loss": 0.5015,
+      "step": 5859
+    },
+    {
+      "epoch": 0.47591975960367094,
+      "grad_norm": 13.074384603626209,
+      "learning_rate": 2.8157991390632206e-06,
+      "loss": 0.6872,
+      "step": 5860
+    },
+    {
+      "epoch": 0.4760009745797125,
+      "grad_norm": 6.893964042250645,
+      "learning_rate": 2.8151467737940312e-06,
+      "loss": 0.4646,
+      "step": 5861
+    },
+    {
+      "epoch": 0.4760821895557541,
+      "grad_norm": 11.048526461851749,
+      "learning_rate": 2.8144943867183535e-06,
+      "loss": 0.6181,
+      "step": 5862
+    },
+    {
+      "epoch": 0.4761634045317957,
+      "grad_norm": 5.148676616822644,
+      "learning_rate": 2.8138419778813274e-06,
+      "loss": 0.5295,
+      "step": 5863
+    },
+    {
+      "epoch": 0.47624461950783725,
+      "grad_norm": 7.31376261027769,
+      "learning_rate": 2.8131895473280985e-06,
+      "loss": 0.6112,
+      "step": 5864
+    },
+    {
+      "epoch": 0.47632583448387883,
+      "grad_norm": 8.847836919115894,
+      "learning_rate": 2.81253709510381e-06,
+      "loss": 0.5427,
+      "step": 5865
+    },
+    {
+      "epoch": 0.4764070494599204,
+      "grad_norm": 3.549876129014136,
+      "learning_rate": 2.811884621253608e-06,
+      "loss": 0.5135,
+      "step": 5866
+    },
+    {
+      "epoch": 0.476488264435962,
+      "grad_norm": 4.676010284588651,
+      "learning_rate": 2.811232125822642e-06,
+      "loss": 0.6499,
+      "step": 5867
+    },
+    {
+      "epoch": 0.47656947941200356,
+      "grad_norm": 4.475545565711852,
+      "learning_rate": 2.81057960885606e-06,
+      "loss": 0.4845,
+      "step": 5868
+    },
+    {
+      "epoch": 0.47665069438804514,
+      "grad_norm": 5.150829015821627,
+      "learning_rate": 2.8099270703990124e-06,
+      "loss": 0.4768,
+      "step": 5869
+    },
+    {
+      "epoch": 0.4767319093640867,
+      "grad_norm": 6.981268549767182,
+      "learning_rate": 2.8092745104966514e-06,
+      "loss": 0.5269,
+      "step": 5870
+    },
+    {
+      "epoch": 0.47681312434012835,
+      "grad_norm": 4.317381690472616,
+      "learning_rate": 2.8086219291941314e-06,
+      "loss": 0.5675,
+      "step": 5871
+    },
+    {
+      "epoch": 0.4768943393161699,
+      "grad_norm": 6.190644029515262,
+      "learning_rate": 2.807969326536607e-06,
+      "loss": 0.4985,
+      "step": 5872
+    },
+    {
+      "epoch": 0.4769755542922115,
+      "grad_norm": 3.4066085952901073,
+      "learning_rate": 2.8073167025692354e-06,
+      "loss": 0.5183,
+      "step": 5873
+    },
+    {
+      "epoch": 0.4770567692682531,
+      "grad_norm": 5.692886429945919,
+      "learning_rate": 2.8066640573371747e-06,
+      "loss": 0.5322,
+      "step": 5874
+    },
+    {
+      "epoch": 0.47713798424429466,
+      "grad_norm": 7.642876548549243,
+      "learning_rate": 2.8060113908855847e-06,
+      "loss": 0.6323,
+      "step": 5875
+    },
+    {
+      "epoch": 0.47721919922033623,
+      "grad_norm": 17.72297787277452,
+      "learning_rate": 2.805358703259624e-06,
+      "loss": 0.5077,
+      "step": 5876
+    },
+    {
+      "epoch": 0.4773004141963778,
+      "grad_norm": 5.581364142761338,
+      "learning_rate": 2.8047059945044585e-06,
+      "loss": 0.4381,
+      "step": 5877
+    },
+    {
+      "epoch": 0.4773816291724194,
+      "grad_norm": 9.67091630066501,
+      "learning_rate": 2.8040532646652515e-06,
+      "loss": 0.4763,
+      "step": 5878
+    },
+    {
+      "epoch": 0.47746284414846096,
+      "grad_norm": 4.1891971217435575,
+      "learning_rate": 2.803400513787166e-06,
+      "loss": 0.5316,
+      "step": 5879
+    },
+    {
+      "epoch": 0.47754405912450254,
+      "grad_norm": 6.17817844727735,
+      "learning_rate": 2.802747741915372e-06,
+      "loss": 0.5111,
+      "step": 5880
+    },
+    {
+      "epoch": 0.4776252741005441,
+      "grad_norm": 4.9847568946301015,
+      "learning_rate": 2.8020949490950367e-06,
+      "loss": 0.5109,
+      "step": 5881
+    },
+    {
+      "epoch": 0.47770648907658575,
+      "grad_norm": 4.473462329646816,
+      "learning_rate": 2.801442135371329e-06,
+      "loss": 0.6185,
+      "step": 5882
+    },
+    {
+      "epoch": 0.4777877040526273,
+      "grad_norm": 6.199201466381722,
+      "learning_rate": 2.800789300789421e-06,
+      "loss": 0.4709,
+      "step": 5883
+    },
+    {
+      "epoch": 0.4778689190286689,
+      "grad_norm": 3.4927971670876157,
+      "learning_rate": 2.8001364453944853e-06,
+      "loss": 0.658,
+      "step": 5884
+    },
+    {
+      "epoch": 0.4779501340047105,
+      "grad_norm": 9.252290801408948,
+      "learning_rate": 2.799483569231696e-06,
+      "loss": 0.3258,
+      "step": 5885
+    },
+    {
+      "epoch": 0.47803134898075206,
+      "grad_norm": 5.159111959600434,
+      "learning_rate": 2.798830672346229e-06,
+      "loss": 0.7327,
+      "step": 5886
+    },
+    {
+      "epoch": 0.47811256395679363,
+      "grad_norm": 4.997715592914455,
+      "learning_rate": 2.7981777547832604e-06,
+      "loss": 0.5373,
+      "step": 5887
+    },
+    {
+      "epoch": 0.4781937789328352,
+      "grad_norm": 7.611885925164203,
+      "learning_rate": 2.7975248165879697e-06,
+      "loss": 0.4249,
+      "step": 5888
+    },
+    {
+      "epoch": 0.4782749939088768,
+      "grad_norm": 8.402898696861591,
+      "learning_rate": 2.7968718578055365e-06,
+      "loss": 0.5686,
+      "step": 5889
+    },
+    {
+      "epoch": 0.47835620888491837,
+      "grad_norm": 4.514667421391678,
+      "learning_rate": 2.796218878481142e-06,
+      "loss": 0.5125,
+      "step": 5890
+    },
+    {
+      "epoch": 0.47843742386095994,
+      "grad_norm": 6.200592103587213,
+      "learning_rate": 2.7955658786599688e-06,
+      "loss": 0.4591,
+      "step": 5891
+    },
+    {
+      "epoch": 0.4785186388370015,
+      "grad_norm": 3.9141827656240262,
+      "learning_rate": 2.7949128583872e-06,
+      "loss": 0.6035,
+      "step": 5892
+    },
+    {
+      "epoch": 0.47859985381304315,
+      "grad_norm": 2.881200756727726,
+      "learning_rate": 2.7942598177080233e-06,
+      "loss": 0.5105,
+      "step": 5893
+    },
+    {
+      "epoch": 0.47868106878908473,
+      "grad_norm": 4.876865531343913,
+      "learning_rate": 2.7936067566676244e-06,
+      "loss": 0.4788,
+      "step": 5894
+    },
+    {
+      "epoch": 0.4787622837651263,
+      "grad_norm": 4.129321830180821,
+      "learning_rate": 2.792953675311192e-06,
+      "loss": 0.4933,
+      "step": 5895
+    },
+    {
+      "epoch": 0.4788434987411679,
+      "grad_norm": 4.975168884090526,
+      "learning_rate": 2.792300573683915e-06,
+      "loss": 0.6777,
+      "step": 5896
+    },
+    {
+      "epoch": 0.47892471371720946,
+      "grad_norm": 4.373706953851931,
+      "learning_rate": 2.7916474518309854e-06,
+      "loss": 0.6101,
+      "step": 5897
+    },
+    {
+      "epoch": 0.47900592869325104,
+      "grad_norm": 6.787308432476194,
+      "learning_rate": 2.790994309797596e-06,
+      "loss": 0.5524,
+      "step": 5898
+    },
+    {
+      "epoch": 0.4790871436692926,
+      "grad_norm": 3.130380915868563,
+      "learning_rate": 2.79034114762894e-06,
+      "loss": 0.5455,
+      "step": 5899
+    },
+    {
+      "epoch": 0.4791683586453342,
+      "grad_norm": 5.155056558227334,
+      "learning_rate": 2.789687965370214e-06,
+      "loss": 0.405,
+      "step": 5900
+    },
+    {
+      "epoch": 0.47924957362137577,
+      "grad_norm": 5.772687941455542,
+      "learning_rate": 2.7890347630666135e-06,
+      "loss": 0.498,
+      "step": 5901
+    },
+    {
+      "epoch": 0.47933078859741735,
+      "grad_norm": 4.464910286196713,
+      "learning_rate": 2.788381540763337e-06,
+      "loss": 0.4895,
+      "step": 5902
+    },
+    {
+      "epoch": 0.4794120035734589,
+      "grad_norm": 6.222904764201984,
+      "learning_rate": 2.787728298505584e-06,
+      "loss": 0.456,
+      "step": 5903
+    },
+    {
+      "epoch": 0.47949321854950055,
+      "grad_norm": 30.493510617091204,
+      "learning_rate": 2.787075036338556e-06,
+      "loss": 0.4607,
+      "step": 5904
+    },
+    {
+      "epoch": 0.47957443352554213,
+      "grad_norm": 12.321251377438697,
+      "learning_rate": 2.7864217543074544e-06,
+      "loss": 0.545,
+      "step": 5905
+    },
+    {
+      "epoch": 0.4796556485015837,
+      "grad_norm": 5.80807114653115,
+      "learning_rate": 2.7857684524574833e-06,
+      "loss": 0.6029,
+      "step": 5906
+    },
+    {
+      "epoch": 0.4797368634776253,
+      "grad_norm": 7.259661847105047,
+      "learning_rate": 2.7851151308338483e-06,
+      "loss": 0.5074,
+      "step": 5907
+    },
+    {
+      "epoch": 0.47981807845366686,
+      "grad_norm": 5.490831475001269,
+      "learning_rate": 2.784461789481754e-06,
+      "loss": 0.4931,
+      "step": 5908
+    },
+    {
+      "epoch": 0.47989929342970844,
+      "grad_norm": 10.451899945538207,
+      "learning_rate": 2.7838084284464105e-06,
+      "loss": 0.3872,
+      "step": 5909
+    },
+    {
+      "epoch": 0.47998050840575,
+      "grad_norm": 7.293146517834257,
+      "learning_rate": 2.7831550477730255e-06,
+      "loss": 0.6436,
+      "step": 5910
+    },
+    {
+      "epoch": 0.4800617233817916,
+      "grad_norm": 6.817978018689368,
+      "learning_rate": 2.78250164750681e-06,
+      "loss": 0.4786,
+      "step": 5911
+    },
+    {
+      "epoch": 0.48014293835783317,
+      "grad_norm": 5.17758792887419,
+      "learning_rate": 2.781848227692974e-06,
+      "loss": 0.316,
+      "step": 5912
+    },
+    {
+      "epoch": 0.48022415333387475,
+      "grad_norm": 9.128124210915745,
+      "learning_rate": 2.7811947883767343e-06,
+      "loss": 0.4606,
+      "step": 5913
+    },
+    {
+      "epoch": 0.4803053683099163,
+      "grad_norm": 6.296126142033211,
+      "learning_rate": 2.780541329603303e-06,
+      "loss": 0.4745,
+      "step": 5914
+    },
+    {
+      "epoch": 0.48038658328595796,
+      "grad_norm": 6.296563355774902,
+      "learning_rate": 2.7798878514178955e-06,
+      "loss": 0.5816,
+      "step": 5915
+    },
+    {
+      "epoch": 0.48046779826199953,
+      "grad_norm": 4.08831948588252,
+      "learning_rate": 2.779234353865731e-06,
+      "loss": 0.4533,
+      "step": 5916
+    },
+    {
+      "epoch": 0.4805490132380411,
+      "grad_norm": 6.198220168009877,
+      "learning_rate": 2.7785808369920263e-06,
+      "loss": 0.52,
+      "step": 5917
+    },
+    {
+      "epoch": 0.4806302282140827,
+      "grad_norm": 7.348946572016875,
+      "learning_rate": 2.777927300842003e-06,
+      "loss": 0.4988,
+      "step": 5918
+    },
+    {
+      "epoch": 0.48071144319012427,
+      "grad_norm": 7.4243657712578,
+      "learning_rate": 2.7772737454608804e-06,
+      "loss": 0.5427,
+      "step": 5919
+    },
+    {
+      "epoch": 0.48079265816616584,
+      "grad_norm": 6.240520113282074,
+      "learning_rate": 2.7766201708938823e-06,
+      "loss": 0.6156,
+      "step": 5920
+    },
+    {
+      "epoch": 0.4808738731422074,
+      "grad_norm": 4.092737052613775,
+      "learning_rate": 2.7759665771862324e-06,
+      "loss": 0.4755,
+      "step": 5921
+    },
+    {
+      "epoch": 0.480955088118249,
+      "grad_norm": 4.543429446499594,
+      "learning_rate": 2.775312964383156e-06,
+      "loss": 0.4982,
+      "step": 5922
+    },
+    {
+      "epoch": 0.4810363030942906,
+      "grad_norm": 4.139943817706591,
+      "learning_rate": 2.77465933252988e-06,
+      "loss": 0.6698,
+      "step": 5923
+    },
+    {
+      "epoch": 0.48111751807033215,
+      "grad_norm": 4.420260702078683,
+      "learning_rate": 2.7740056816716317e-06,
+      "loss": 0.4612,
+      "step": 5924
+    },
+    {
+      "epoch": 0.48119873304637373,
+      "grad_norm": 8.399576878392056,
+      "learning_rate": 2.7733520118536395e-06,
+      "loss": 0.4372,
+      "step": 5925
+    },
+    {
+      "epoch": 0.48127994802241536,
+      "grad_norm": 8.529789641771623,
+      "learning_rate": 2.772698323121135e-06,
+      "loss": 0.4788,
+      "step": 5926
+    },
+    {
+      "epoch": 0.48136116299845694,
+      "grad_norm": 6.033494247535142,
+      "learning_rate": 2.7720446155193503e-06,
+      "loss": 0.8274,
+      "step": 5927
+    },
+    {
+      "epoch": 0.4814423779744985,
+      "grad_norm": 3.680362638542244,
+      "learning_rate": 2.7713908890935177e-06,
+      "loss": 0.5214,
+      "step": 5928
+    },
+    {
+      "epoch": 0.4815235929505401,
+      "grad_norm": 7.379140973686685,
+      "learning_rate": 2.770737143888872e-06,
+      "loss": 0.7151,
+      "step": 5929
+    },
+    {
+      "epoch": 0.48160480792658167,
+      "grad_norm": 3.6816364647280584,
+      "learning_rate": 2.7700833799506487e-06,
+      "loss": 0.553,
+      "step": 5930
+    },
+    {
+      "epoch": 0.48168602290262325,
+      "grad_norm": 5.987951257678209,
+      "learning_rate": 2.7694295973240848e-06,
+      "loss": 0.4937,
+      "step": 5931
+    },
+    {
+      "epoch": 0.4817672378786648,
+      "grad_norm": 4.956448565368726,
+      "learning_rate": 2.7687757960544193e-06,
+      "loss": 0.4982,
+      "step": 5932
+    },
+    {
+      "epoch": 0.4818484528547064,
+      "grad_norm": 4.49940857003615,
+      "learning_rate": 2.7681219761868905e-06,
+      "loss": 0.6454,
+      "step": 5933
+    },
+    {
+      "epoch": 0.481929667830748,
+      "grad_norm": 4.009481443453415,
+      "learning_rate": 2.7674681377667403e-06,
+      "loss": 0.5949,
+      "step": 5934
+    },
+    {
+      "epoch": 0.48201088280678955,
+      "grad_norm": 6.276103431385119,
+      "learning_rate": 2.7668142808392102e-06,
+      "loss": 0.6751,
+      "step": 5935
+    },
+    {
+      "epoch": 0.48209209778283113,
+      "grad_norm": 5.264109990138893,
+      "learning_rate": 2.7661604054495447e-06,
+      "loss": 0.5605,
+      "step": 5936
+    },
+    {
+      "epoch": 0.48217331275887276,
+      "grad_norm": 6.623257155307936,
+      "learning_rate": 2.765506511642987e-06,
+      "loss": 0.5536,
+      "step": 5937
+    },
+    {
+      "epoch": 0.48225452773491434,
+      "grad_norm": 8.180552462619701,
+      "learning_rate": 2.764852599464784e-06,
+      "loss": 0.5155,
+      "step": 5938
+    },
+    {
+      "epoch": 0.4823357427109559,
+      "grad_norm": 3.6982416953033996,
+      "learning_rate": 2.764198668960183e-06,
+      "loss": 0.4975,
+      "step": 5939
+    },
+    {
+      "epoch": 0.4824169576869975,
+      "grad_norm": 4.58146111767092,
+      "learning_rate": 2.7635447201744324e-06,
+      "loss": 0.6719,
+      "step": 5940
+    },
+    {
+      "epoch": 0.48249817266303907,
+      "grad_norm": 4.792428868707111,
+      "learning_rate": 2.7628907531527815e-06,
+      "loss": 0.5106,
+      "step": 5941
+    },
+    {
+      "epoch": 0.48257938763908065,
+      "grad_norm": 5.140027653654479,
+      "learning_rate": 2.762236767940482e-06,
+      "loss": 0.3998,
+      "step": 5942
+    },
+    {
+      "epoch": 0.4826606026151222,
+      "grad_norm": 4.42487569925798,
+      "learning_rate": 2.761582764582787e-06,
+      "loss": 0.4726,
+      "step": 5943
+    },
+    {
+      "epoch": 0.4827418175911638,
+      "grad_norm": 4.997518890904461,
+      "learning_rate": 2.760928743124948e-06,
+      "loss": 0.5601,
+      "step": 5944
+    },
+    {
+      "epoch": 0.4828230325672054,
+      "grad_norm": 3.799079838934848,
+      "learning_rate": 2.7602747036122213e-06,
+      "loss": 0.6182,
+      "step": 5945
+    },
+    {
+      "epoch": 0.48290424754324696,
+      "grad_norm": 10.353271040341017,
+      "learning_rate": 2.759620646089863e-06,
+      "loss": 0.476,
+      "step": 5946
+    },
+    {
+      "epoch": 0.48298546251928853,
+      "grad_norm": 5.666756315530278,
+      "learning_rate": 2.758966570603129e-06,
+      "loss": 0.5052,
+      "step": 5947
+    },
+    {
+      "epoch": 0.48306667749533017,
+      "grad_norm": 6.165341654780337,
+      "learning_rate": 2.7583124771972797e-06,
+      "loss": 0.4994,
+      "step": 5948
+    },
+    {
+      "epoch": 0.48314789247137174,
+      "grad_norm": 4.569577065572749,
+      "learning_rate": 2.7576583659175738e-06,
+      "loss": 0.439,
+      "step": 5949
+    },
+    {
+      "epoch": 0.4832291074474133,
+      "grad_norm": 3.098941821564416,
+      "learning_rate": 2.7570042368092724e-06,
+      "loss": 0.5527,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4833103224234549,
+      "grad_norm": 5.497608597420854,
+      "learning_rate": 2.7563500899176383e-06,
+      "loss": 0.4795,
+      "step": 5951
+    },
+    {
+      "epoch": 0.4833915373994965,
+      "grad_norm": 5.7165228874758505,
+      "learning_rate": 2.7556959252879345e-06,
+      "loss": 0.4909,
+      "step": 5952
+    },
+    {
+      "epoch": 0.48347275237553805,
+      "grad_norm": 5.874962862560617,
+      "learning_rate": 2.755041742965426e-06,
+      "loss": 0.4677,
+      "step": 5953
+    },
+    {
+      "epoch": 0.4835539673515796,
+      "grad_norm": 4.095451834207825,
+      "learning_rate": 2.7543875429953787e-06,
+      "loss": 0.6461,
+      "step": 5954
+    },
+    {
+      "epoch": 0.4836351823276212,
+      "grad_norm": 4.473317225148218,
+      "learning_rate": 2.7537333254230596e-06,
+      "loss": 0.4963,
+      "step": 5955
+    },
+    {
+      "epoch": 0.4837163973036628,
+      "grad_norm": 5.773342053411397,
+      "learning_rate": 2.7530790902937376e-06,
+      "loss": 0.4534,
+      "step": 5956
+    },
+    {
+      "epoch": 0.48379761227970436,
+      "grad_norm": 3.510346924849262,
+      "learning_rate": 2.752424837652681e-06,
+      "loss": 0.6875,
+      "step": 5957
+    },
+    {
+      "epoch": 0.48387882725574594,
+      "grad_norm": 7.132250653049774,
+      "learning_rate": 2.751770567545163e-06,
+      "loss": 0.5412,
+      "step": 5958
+    },
+    {
+      "epoch": 0.48396004223178757,
+      "grad_norm": 5.085327115862414,
+      "learning_rate": 2.7511162800164536e-06,
+      "loss": 0.7837,
+      "step": 5959
+    },
+    {
+      "epoch": 0.48404125720782915,
+      "grad_norm": 7.899170324549878,
+      "learning_rate": 2.7504619751118266e-06,
+      "loss": 0.5815,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4841224721838707,
+      "grad_norm": 12.534688873820935,
+      "learning_rate": 2.749807652876556e-06,
+      "loss": 0.5867,
+      "step": 5961
+    },
+    {
+      "epoch": 0.4842036871599123,
+      "grad_norm": 5.888585867450327,
+      "learning_rate": 2.749153313355919e-06,
+      "loss": 0.4844,
+      "step": 5962
+    },
+    {
+      "epoch": 0.4842849021359539,
+      "grad_norm": 4.658146495411607,
+      "learning_rate": 2.74849895659519e-06,
+      "loss": 0.4221,
+      "step": 5963
+    },
+    {
+      "epoch": 0.48436611711199545,
+      "grad_norm": 8.00901759996569,
+      "learning_rate": 2.7478445826396495e-06,
+      "loss": 0.3995,
+      "step": 5964
+    },
+    {
+      "epoch": 0.48444733208803703,
+      "grad_norm": 4.722557156452973,
+      "learning_rate": 2.747190191534575e-06,
+      "loss": 0.5922,
+      "step": 5965
+    },
+    {
+      "epoch": 0.4845285470640786,
+      "grad_norm": 4.907315907025934,
+      "learning_rate": 2.7465357833252477e-06,
+      "loss": 0.483,
+      "step": 5966
+    },
+    {
+      "epoch": 0.4846097620401202,
+      "grad_norm": 3.780776893956639,
+      "learning_rate": 2.7458813580569487e-06,
+      "loss": 0.5136,
+      "step": 5967
+    },
+    {
+      "epoch": 0.48469097701616176,
+      "grad_norm": 4.759267167903013,
+      "learning_rate": 2.7452269157749614e-06,
+      "loss": 0.6048,
+      "step": 5968
+    },
+    {
+      "epoch": 0.48477219199220334,
+      "grad_norm": 5.977936036351052,
+      "learning_rate": 2.744572456524569e-06,
+      "loss": 0.5071,
+      "step": 5969
+    },
+    {
+      "epoch": 0.48485340696824497,
+      "grad_norm": 5.801052897330955,
+      "learning_rate": 2.7439179803510567e-06,
+      "loss": 0.3259,
+      "step": 5970
+    },
+    {
+      "epoch": 0.48493462194428655,
+      "grad_norm": 5.504537267230917,
+      "learning_rate": 2.7432634872997123e-06,
+      "loss": 0.454,
+      "step": 5971
+    },
+    {
+      "epoch": 0.4850158369203281,
+      "grad_norm": 5.980973589761268,
+      "learning_rate": 2.7426089774158217e-06,
+      "loss": 0.5151,
+      "step": 5972
+    },
+    {
+      "epoch": 0.4850970518963697,
+      "grad_norm": 6.090787296944255,
+      "learning_rate": 2.7419544507446727e-06,
+      "loss": 0.5434,
+      "step": 5973
+    },
+    {
+      "epoch": 0.4851782668724113,
+      "grad_norm": 5.184785779288021,
+      "learning_rate": 2.7412999073315567e-06,
+      "loss": 0.6548,
+      "step": 5974
+    },
+    {
+      "epoch": 0.48525948184845286,
+      "grad_norm": 3.168315464569746,
+      "learning_rate": 2.7406453472217654e-06,
+      "loss": 0.5838,
+      "step": 5975
+    },
+    {
+      "epoch": 0.48534069682449443,
+      "grad_norm": 5.7401424667695204,
+      "learning_rate": 2.7399907704605884e-06,
+      "loss": 0.4268,
+      "step": 5976
+    },
+    {
+      "epoch": 0.485421911800536,
+      "grad_norm": 8.886061000991853,
+      "learning_rate": 2.7393361770933198e-06,
+      "loss": 0.4986,
+      "step": 5977
+    },
+    {
+      "epoch": 0.4855031267765776,
+      "grad_norm": 4.804673355461681,
+      "learning_rate": 2.7386815671652556e-06,
+      "loss": 0.6466,
+      "step": 5978
+    },
+    {
+      "epoch": 0.48558434175261916,
+      "grad_norm": 5.321396681856843,
+      "learning_rate": 2.7380269407216896e-06,
+      "loss": 0.4684,
+      "step": 5979
+    },
+    {
+      "epoch": 0.48566555672866074,
+      "grad_norm": 5.686891915170021,
+      "learning_rate": 2.737372297807919e-06,
+      "loss": 0.6947,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4857467717047024,
+      "grad_norm": 6.348245493300976,
+      "learning_rate": 2.7367176384692425e-06,
+      "loss": 0.4924,
+      "step": 5981
+    },
+    {
+      "epoch": 0.48582798668074395,
+      "grad_norm": 6.0903339297311545,
+      "learning_rate": 2.736062962750957e-06,
+      "loss": 0.5165,
+      "step": 5982
+    },
+    {
+      "epoch": 0.4859092016567855,
+      "grad_norm": 11.757378358765155,
+      "learning_rate": 2.735408270698364e-06,
+      "loss": 0.4462,
+      "step": 5983
+    },
+    {
+      "epoch": 0.4859904166328271,
+      "grad_norm": 16.683533350645916,
+      "learning_rate": 2.7347535623567656e-06,
+      "loss": 0.6067,
+      "step": 5984
+    },
+    {
+      "epoch": 0.4860716316088687,
+      "grad_norm": 7.944296108437706,
+      "learning_rate": 2.734098837771462e-06,
+      "loss": 0.5032,
+      "step": 5985
+    },
+    {
+      "epoch": 0.48615284658491026,
+      "grad_norm": 5.943425981392983,
+      "learning_rate": 2.7334440969877584e-06,
+      "loss": 0.5087,
+      "step": 5986
+    },
+    {
+      "epoch": 0.48623406156095184,
+      "grad_norm": 3.3987635620526686,
+      "learning_rate": 2.7327893400509586e-06,
+      "loss": 0.5818,
+      "step": 5987
+    },
+    {
+      "epoch": 0.4863152765369934,
+      "grad_norm": 6.620696249187532,
+      "learning_rate": 2.732134567006368e-06,
+      "loss": 0.4364,
+      "step": 5988
+    },
+    {
+      "epoch": 0.486396491513035,
+      "grad_norm": 5.572056490122121,
+      "learning_rate": 2.731479777899295e-06,
+      "loss": 0.4405,
+      "step": 5989
+    },
+    {
+      "epoch": 0.48647770648907657,
+      "grad_norm": 6.354284008710806,
+      "learning_rate": 2.730824972775045e-06,
+      "loss": 0.6067,
+      "step": 5990
+    },
+    {
+      "epoch": 0.48655892146511814,
+      "grad_norm": 5.802616735842974,
+      "learning_rate": 2.7301701516789303e-06,
+      "loss": 0.3966,
+      "step": 5991
+    },
+    {
+      "epoch": 0.4866401364411598,
+      "grad_norm": 4.459624839521087,
+      "learning_rate": 2.729515314656258e-06,
+      "loss": 0.5996,
+      "step": 5992
+    },
+    {
+      "epoch": 0.48672135141720135,
+      "grad_norm": 7.157939161361248,
+      "learning_rate": 2.7288604617523405e-06,
+      "loss": 0.5517,
+      "step": 5993
+    },
+    {
+      "epoch": 0.48680256639324293,
+      "grad_norm": 5.292457922278461,
+      "learning_rate": 2.728205593012491e-06,
+      "loss": 0.4596,
+      "step": 5994
+    },
+    {
+      "epoch": 0.4868837813692845,
+      "grad_norm": 5.048505829129327,
+      "learning_rate": 2.7275507084820226e-06,
+      "loss": 0.6614,
+      "step": 5995
+    },
+    {
+      "epoch": 0.4869649963453261,
+      "grad_norm": 5.999484690389678,
+      "learning_rate": 2.726895808206248e-06,
+      "loss": 0.4645,
+      "step": 5996
+    },
+    {
+      "epoch": 0.48704621132136766,
+      "grad_norm": 4.707719844104531,
+      "learning_rate": 2.7262408922304857e-06,
+      "loss": 0.5658,
+      "step": 5997
+    },
+    {
+      "epoch": 0.48712742629740924,
+      "grad_norm": 9.302118644783453,
+      "learning_rate": 2.72558596060005e-06,
+      "loss": 0.4844,
+      "step": 5998
+    },
+    {
+      "epoch": 0.4872086412734508,
+      "grad_norm": 6.41310409364449,
+      "learning_rate": 2.72493101336026e-06,
+      "loss": 0.4755,
+      "step": 5999
+    },
+    {
+      "epoch": 0.4872898562494924,
+      "grad_norm": 7.372592645487124,
+      "learning_rate": 2.7242760505564346e-06,
+      "loss": 0.4443,
+      "step": 6000
+    },
+    {
+      "epoch": 0.48737107122553397,
+      "grad_norm": 5.478164117803782,
+      "learning_rate": 2.7236210722338936e-06,
+      "loss": 0.6266,
+      "step": 6001
+    },
+    {
+      "epoch": 0.48745228620157555,
+      "grad_norm": 4.451621100299445,
+      "learning_rate": 2.7229660784379575e-06,
+      "loss": 0.6028,
+      "step": 6002
+    },
+    {
+      "epoch": 0.4875335011776172,
+      "grad_norm": 4.829301062695959,
+      "learning_rate": 2.7223110692139487e-06,
+      "loss": 0.3843,
+      "step": 6003
+    },
+    {
+      "epoch": 0.48761471615365876,
+      "grad_norm": 8.161182349000478,
+      "learning_rate": 2.7216560446071904e-06,
+      "loss": 0.5373,
+      "step": 6004
+    },
+    {
+      "epoch": 0.48769593112970033,
+      "grad_norm": 5.075301630204499,
+      "learning_rate": 2.721001004663008e-06,
+      "loss": 0.5209,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4877771461057419,
+      "grad_norm": 3.5392282378757094,
+      "learning_rate": 2.7203459494267243e-06,
+      "loss": 0.4714,
+      "step": 6006
+    },
+    {
+      "epoch": 0.4878583610817835,
+      "grad_norm": 3.3493451655565147,
+      "learning_rate": 2.719690878943668e-06,
+      "loss": 0.6381,
+      "step": 6007
+    },
+    {
+      "epoch": 0.48793957605782506,
+      "grad_norm": 11.279877740153177,
+      "learning_rate": 2.7190357932591653e-06,
+      "loss": 0.5869,
+      "step": 6008
+    },
+    {
+      "epoch": 0.48802079103386664,
+      "grad_norm": 15.655656745532893,
+      "learning_rate": 2.7183806924185447e-06,
+      "loss": 0.4589,
+      "step": 6009
+    },
+    {
+      "epoch": 0.4881020060099082,
+      "grad_norm": 7.534260268163045,
+      "learning_rate": 2.717725576467136e-06,
+      "loss": 0.5564,
+      "step": 6010
+    },
+    {
+      "epoch": 0.4881832209859498,
+      "grad_norm": 8.16602240519616,
+      "learning_rate": 2.71707044545027e-06,
+      "loss": 0.4806,
+      "step": 6011
+    },
+    {
+      "epoch": 0.48826443596199137,
+      "grad_norm": 5.065942666535822,
+      "learning_rate": 2.716415299413278e-06,
+      "loss": 0.4948,
+      "step": 6012
+    },
+    {
+      "epoch": 0.48834565093803295,
+      "grad_norm": 6.1717583953297455,
+      "learning_rate": 2.7157601384014927e-06,
+      "loss": 0.5663,
+      "step": 6013
+    },
+    {
+      "epoch": 0.4884268659140746,
+      "grad_norm": 5.9450854471553125,
+      "learning_rate": 2.7151049624602473e-06,
+      "loss": 0.6468,
+      "step": 6014
+    },
+    {
+      "epoch": 0.48850808089011616,
+      "grad_norm": 6.376587879616875,
+      "learning_rate": 2.714449771634877e-06,
+      "loss": 0.4685,
+      "step": 6015
+    },
+    {
+      "epoch": 0.48858929586615774,
+      "grad_norm": 4.341090257406891,
+      "learning_rate": 2.713794565970718e-06,
+      "loss": 0.438,
+      "step": 6016
+    },
+    {
+      "epoch": 0.4886705108421993,
+      "grad_norm": 4.016157690853844,
+      "learning_rate": 2.7131393455131057e-06,
+      "loss": 0.6089,
+      "step": 6017
+    },
+    {
+      "epoch": 0.4887517258182409,
+      "grad_norm": 2.638895965006909,
+      "learning_rate": 2.7124841103073794e-06,
+      "loss": 0.3652,
+      "step": 6018
+    },
+    {
+      "epoch": 0.48883294079428247,
+      "grad_norm": 9.501098441788494,
+      "learning_rate": 2.711828860398877e-06,
+      "loss": 0.5685,
+      "step": 6019
+    },
+    {
+      "epoch": 0.48891415577032404,
+      "grad_norm": 6.054200225549085,
+      "learning_rate": 2.7111735958329383e-06,
+      "loss": 0.3793,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4889953707463656,
+      "grad_norm": 5.616508580837288,
+      "learning_rate": 2.7105183166549048e-06,
+      "loss": 0.547,
+      "step": 6021
+    },
+    {
+      "epoch": 0.4890765857224072,
+      "grad_norm": 6.361440708041633,
+      "learning_rate": 2.7098630229101174e-06,
+      "loss": 0.5794,
+      "step": 6022
+    },
+    {
+      "epoch": 0.4891578006984488,
+      "grad_norm": 9.543869316221988,
+      "learning_rate": 2.70920771464392e-06,
+      "loss": 0.6857,
+      "step": 6023
+    },
+    {
+      "epoch": 0.48923901567449035,
+      "grad_norm": 4.931811725830596,
+      "learning_rate": 2.708552391901656e-06,
+      "loss": 0.6798,
+      "step": 6024
+    },
+    {
+      "epoch": 0.489320230650532,
+      "grad_norm": 7.691562606616547,
+      "learning_rate": 2.70789705472867e-06,
+      "loss": 0.6456,
+      "step": 6025
+    },
+    {
+      "epoch": 0.48940144562657356,
+      "grad_norm": 3.9609368324412286,
+      "learning_rate": 2.707241703170308e-06,
+      "loss": 0.4363,
+      "step": 6026
+    },
+    {
+      "epoch": 0.48948266060261514,
+      "grad_norm": 5.514465718417427,
+      "learning_rate": 2.706586337271917e-06,
+      "loss": 0.4752,
+      "step": 6027
+    },
+    {
+      "epoch": 0.4895638755786567,
+      "grad_norm": 4.986568690370059,
+      "learning_rate": 2.705930957078845e-06,
+      "loss": 0.5089,
+      "step": 6028
+    },
+    {
+      "epoch": 0.4896450905546983,
+      "grad_norm": 4.59026441450655,
+      "learning_rate": 2.705275562636441e-06,
+      "loss": 0.5348,
+      "step": 6029
+    },
+    {
+      "epoch": 0.48972630553073987,
+      "grad_norm": 5.882838457335099,
+      "learning_rate": 2.7046201539900537e-06,
+      "loss": 0.5789,
+      "step": 6030
+    },
+    {
+      "epoch": 0.48980752050678145,
+      "grad_norm": 6.273818536620701,
+      "learning_rate": 2.7039647311850347e-06,
+      "loss": 0.5857,
+      "step": 6031
+    },
+    {
+      "epoch": 0.489888735482823,
+      "grad_norm": 4.353727277852687,
+      "learning_rate": 2.7033092942667362e-06,
+      "loss": 0.4136,
+      "step": 6032
+    },
+    {
+      "epoch": 0.4899699504588646,
+      "grad_norm": 11.49874908610111,
+      "learning_rate": 2.70265384328051e-06,
+      "loss": 0.4196,
+      "step": 6033
+    },
+    {
+      "epoch": 0.4900511654349062,
+      "grad_norm": 8.579866589240709,
+      "learning_rate": 2.701998378271711e-06,
+      "loss": 0.4894,
+      "step": 6034
+    },
+    {
+      "epoch": 0.49013238041094775,
+      "grad_norm": 25.037657431765965,
+      "learning_rate": 2.7013428992856925e-06,
+      "loss": 0.5139,
+      "step": 6035
+    },
+    {
+      "epoch": 0.4902135953869894,
+      "grad_norm": 11.228244377773777,
+      "learning_rate": 2.700687406367812e-06,
+      "loss": 0.6387,
+      "step": 6036
+    },
+    {
+      "epoch": 0.49029481036303096,
+      "grad_norm": 4.944749953225672,
+      "learning_rate": 2.700031899563425e-06,
+      "loss": 0.5617,
+      "step": 6037
+    },
+    {
+      "epoch": 0.49037602533907254,
+      "grad_norm": 11.318432666073637,
+      "learning_rate": 2.6993763789178885e-06,
+      "loss": 0.6037,
+      "step": 6038
+    },
+    {
+      "epoch": 0.4904572403151141,
+      "grad_norm": 4.831316548854263,
+      "learning_rate": 2.698720844476562e-06,
+      "loss": 0.6919,
+      "step": 6039
+    },
+    {
+      "epoch": 0.4905384552911557,
+      "grad_norm": 6.464718545938012,
+      "learning_rate": 2.6980652962848055e-06,
+      "loss": 0.5024,
+      "step": 6040
+    },
+    {
+      "epoch": 0.49061967026719727,
+      "grad_norm": 8.179746403330494,
+      "learning_rate": 2.697409734387978e-06,
+      "loss": 0.4774,
+      "step": 6041
+    },
+    {
+      "epoch": 0.49070088524323885,
+      "grad_norm": 6.753372733627089,
+      "learning_rate": 2.6967541588314413e-06,
+      "loss": 0.4216,
+      "step": 6042
+    },
+    {
+      "epoch": 0.4907821002192804,
+      "grad_norm": 3.5059703839980707,
+      "learning_rate": 2.6960985696605583e-06,
+      "loss": 0.5735,
+      "step": 6043
+    },
+    {
+      "epoch": 0.490863315195322,
+      "grad_norm": 6.5528726281723335,
+      "learning_rate": 2.695442966920693e-06,
+      "loss": 0.4922,
+      "step": 6044
+    },
+    {
+      "epoch": 0.4909445301713636,
+      "grad_norm": 6.2801153498685744,
+      "learning_rate": 2.6947873506572083e-06,
+      "loss": 0.5194,
+      "step": 6045
+    },
+    {
+      "epoch": 0.49102574514740516,
+      "grad_norm": 2.729788705777957,
+      "learning_rate": 2.6941317209154694e-06,
+      "loss": 0.7296,
+      "step": 6046
+    },
+    {
+      "epoch": 0.4911069601234468,
+      "grad_norm": 5.110271057688403,
+      "learning_rate": 2.693476077740843e-06,
+      "loss": 0.4946,
+      "step": 6047
+    },
+    {
+      "epoch": 0.49118817509948837,
+      "grad_norm": 5.133921081919526,
+      "learning_rate": 2.6928204211786957e-06,
+      "loss": 0.5102,
+      "step": 6048
+    },
+    {
+      "epoch": 0.49126939007552994,
+      "grad_norm": 4.1327143511126,
+      "learning_rate": 2.6921647512743963e-06,
+      "loss": 0.4642,
+      "step": 6049
+    },
+    {
+      "epoch": 0.4913506050515715,
+      "grad_norm": 13.617171548443578,
+      "learning_rate": 2.691509068073313e-06,
+      "loss": 0.356,
+      "step": 6050
+    },
+    {
+      "epoch": 0.4914318200276131,
+      "grad_norm": 5.719645297157029,
+      "learning_rate": 2.6908533716208157e-06,
+      "loss": 0.5327,
+      "step": 6051
+    },
+    {
+      "epoch": 0.4915130350036547,
+      "grad_norm": 3.9479135140894224,
+      "learning_rate": 2.690197661962275e-06,
+      "loss": 0.6186,
+      "step": 6052
+    },
+    {
+      "epoch": 0.49159424997969625,
+      "grad_norm": 7.503761117502973,
+      "learning_rate": 2.6895419391430635e-06,
+      "loss": 0.3574,
+      "step": 6053
+    },
+    {
+      "epoch": 0.49167546495573783,
+      "grad_norm": 4.597046243404988,
+      "learning_rate": 2.688886203208552e-06,
+      "loss": 0.4214,
+      "step": 6054
+    },
+    {
+      "epoch": 0.4917566799317794,
+      "grad_norm": 19.152422694361487,
+      "learning_rate": 2.6882304542041147e-06,
+      "loss": 0.4696,
+      "step": 6055
+    },
+    {
+      "epoch": 0.491837894907821,
+      "grad_norm": 6.515260679020386,
+      "learning_rate": 2.687574692175127e-06,
+      "loss": 0.4959,
+      "step": 6056
+    },
+    {
+      "epoch": 0.49191910988386256,
+      "grad_norm": 4.2821690669001224,
+      "learning_rate": 2.6869189171669637e-06,
+      "loss": 0.3626,
+      "step": 6057
+    },
+    {
+      "epoch": 0.4920003248599042,
+      "grad_norm": 7.192130700433451,
+      "learning_rate": 2.686263129224999e-06,
+      "loss": 0.5963,
+      "step": 6058
+    },
+    {
+      "epoch": 0.49208153983594577,
+      "grad_norm": 7.6817361440233585,
+      "learning_rate": 2.685607328394613e-06,
+      "loss": 0.4168,
+      "step": 6059
+    },
+    {
+      "epoch": 0.49216275481198735,
+      "grad_norm": 11.249409008964918,
+      "learning_rate": 2.6849515147211814e-06,
+      "loss": 0.4758,
+      "step": 6060
+    },
+    {
+      "epoch": 0.4922439697880289,
+      "grad_norm": 3.9392524421599346,
+      "learning_rate": 2.6842956882500843e-06,
+      "loss": 0.445,
+      "step": 6061
+    },
+    {
+      "epoch": 0.4923251847640705,
+      "grad_norm": 9.559526227459175,
+      "learning_rate": 2.6836398490267006e-06,
+      "loss": 0.5492,
+      "step": 6062
+    },
+    {
+      "epoch": 0.4924063997401121,
+      "grad_norm": 3.917087406483406,
+      "learning_rate": 2.6829839970964112e-06,
+      "loss": 0.4452,
+      "step": 6063
+    },
+    {
+      "epoch": 0.49248761471615365,
+      "grad_norm": 3.7273772258833566,
+      "learning_rate": 2.682328132504598e-06,
+      "loss": 0.4458,
+      "step": 6064
+    },
+    {
+      "epoch": 0.49256882969219523,
+      "grad_norm": 4.9038911903534474,
+      "learning_rate": 2.6816722552966423e-06,
+      "loss": 0.5387,
+      "step": 6065
+    },
+    {
+      "epoch": 0.4926500446682368,
+      "grad_norm": 5.697167255856347,
+      "learning_rate": 2.6810163655179287e-06,
+      "loss": 0.4933,
+      "step": 6066
+    },
+    {
+      "epoch": 0.4927312596442784,
+      "grad_norm": 4.553361726145492,
+      "learning_rate": 2.6803604632138403e-06,
+      "loss": 0.381,
+      "step": 6067
+    },
+    {
+      "epoch": 0.49281247462031996,
+      "grad_norm": 4.7949903203120305,
+      "learning_rate": 2.6797045484297624e-06,
+      "loss": 0.5172,
+      "step": 6068
+    },
+    {
+      "epoch": 0.4928936895963616,
+      "grad_norm": 20.378872084713798,
+      "learning_rate": 2.6790486212110812e-06,
+      "loss": 0.4133,
+      "step": 6069
+    },
+    {
+      "epoch": 0.49297490457240317,
+      "grad_norm": 6.870346383155372,
+      "learning_rate": 2.678392681603183e-06,
+      "loss": 0.4685,
+      "step": 6070
+    },
+    {
+      "epoch": 0.49305611954844475,
+      "grad_norm": 4.53706289191984,
+      "learning_rate": 2.6777367296514547e-06,
+      "loss": 0.4881,
+      "step": 6071
+    },
+    {
+      "epoch": 0.4931373345244863,
+      "grad_norm": 5.005808865354581,
+      "learning_rate": 2.677080765401286e-06,
+      "loss": 0.5477,
+      "step": 6072
+    },
+    {
+      "epoch": 0.4932185495005279,
+      "grad_norm": 8.68217397091714,
+      "learning_rate": 2.6764247888980654e-06,
+      "loss": 0.4074,
+      "step": 6073
+    },
+    {
+      "epoch": 0.4932997644765695,
+      "grad_norm": 6.058970447210371,
+      "learning_rate": 2.675768800187182e-06,
+      "loss": 0.5763,
+      "step": 6074
+    },
+    {
+      "epoch": 0.49338097945261106,
+      "grad_norm": 2.9831711571626522,
+      "learning_rate": 2.67511279931403e-06,
+      "loss": 0.6046,
+      "step": 6075
+    },
+    {
+      "epoch": 0.49346219442865263,
+      "grad_norm": 3.841362882001304,
+      "learning_rate": 2.674456786323998e-06,
+      "loss": 0.505,
+      "step": 6076
+    },
+    {
+      "epoch": 0.4935434094046942,
+      "grad_norm": 5.296826365325529,
+      "learning_rate": 2.6738007612624792e-06,
+      "loss": 0.5466,
+      "step": 6077
+    },
+    {
+      "epoch": 0.4936246243807358,
+      "grad_norm": 4.678775568920299,
+      "learning_rate": 2.673144724174868e-06,
+      "loss": 0.5204,
+      "step": 6078
+    },
+    {
+      "epoch": 0.49370583935677737,
+      "grad_norm": 4.857667826957571,
+      "learning_rate": 2.6724886751065584e-06,
+      "loss": 0.5329,
+      "step": 6079
+    },
+    {
+      "epoch": 0.493787054332819,
+      "grad_norm": 5.592492702134483,
+      "learning_rate": 2.671832614102945e-06,
+      "loss": 0.529,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4938682693088606,
+      "grad_norm": 61.655374457513666,
+      "learning_rate": 2.671176541209424e-06,
+      "loss": 0.5004,
+      "step": 6081
+    },
+    {
+      "epoch": 0.49394948428490215,
+      "grad_norm": 6.490521162736462,
+      "learning_rate": 2.6705204564713927e-06,
+      "loss": 0.4508,
+      "step": 6082
+    },
+    {
+      "epoch": 0.49403069926094373,
+      "grad_norm": 6.042977369262763,
+      "learning_rate": 2.669864359934249e-06,
+      "loss": 0.4754,
+      "step": 6083
+    },
+    {
+      "epoch": 0.4941119142369853,
+      "grad_norm": 7.550162317594159,
+      "learning_rate": 2.6692082516433886e-06,
+      "loss": 0.4356,
+      "step": 6084
+    },
+    {
+      "epoch": 0.4941931292130269,
+      "grad_norm": 3.954161538122901,
+      "learning_rate": 2.668552131644214e-06,
+      "loss": 0.6147,
+      "step": 6085
+    },
+    {
+      "epoch": 0.49427434418906846,
+      "grad_norm": 8.837611857542369,
+      "learning_rate": 2.667895999982124e-06,
+      "loss": 0.4406,
+      "step": 6086
+    },
+    {
+      "epoch": 0.49435555916511004,
+      "grad_norm": 5.060653912888538,
+      "learning_rate": 2.6672398567025188e-06,
+      "loss": 0.4472,
+      "step": 6087
+    },
+    {
+      "epoch": 0.4944367741411516,
+      "grad_norm": 9.229957236777977,
+      "learning_rate": 2.666583701850802e-06,
+      "loss": 0.507,
+      "step": 6088
+    },
+    {
+      "epoch": 0.4945179891171932,
+      "grad_norm": 4.035963295981821,
+      "learning_rate": 2.6659275354723735e-06,
+      "loss": 0.4867,
+      "step": 6089
+    },
+    {
+      "epoch": 0.49459920409323477,
+      "grad_norm": 8.351371921914652,
+      "learning_rate": 2.6652713576126376e-06,
+      "loss": 0.6755,
+      "step": 6090
+    },
+    {
+      "epoch": 0.4946804190692764,
+      "grad_norm": 5.78678933022576,
+      "learning_rate": 2.6646151683169985e-06,
+      "loss": 0.6439,
+      "step": 6091
+    },
+    {
+      "epoch": 0.494761634045318,
+      "grad_norm": 4.566506684904742,
+      "learning_rate": 2.6639589676308614e-06,
+      "loss": 0.7211,
+      "step": 6092
+    },
+    {
+      "epoch": 0.49484284902135955,
+      "grad_norm": 7.609049634242752,
+      "learning_rate": 2.663302755599631e-06,
+      "loss": 0.6008,
+      "step": 6093
+    },
+    {
+      "epoch": 0.49492406399740113,
+      "grad_norm": 3.3025767017322893,
+      "learning_rate": 2.6626465322687144e-06,
+      "loss": 0.5058,
+      "step": 6094
+    },
+    {
+      "epoch": 0.4950052789734427,
+      "grad_norm": 5.250107319379549,
+      "learning_rate": 2.6619902976835187e-06,
+      "loss": 0.3995,
+      "step": 6095
+    },
+    {
+      "epoch": 0.4950864939494843,
+      "grad_norm": 4.264694509387339,
+      "learning_rate": 2.6613340518894513e-06,
+      "loss": 0.4428,
+      "step": 6096
+    },
+    {
+      "epoch": 0.49516770892552586,
+      "grad_norm": 9.523848099068603,
+      "learning_rate": 2.6606777949319217e-06,
+      "loss": 0.4568,
+      "step": 6097
+    },
+    {
+      "epoch": 0.49524892390156744,
+      "grad_norm": 4.645197872989053,
+      "learning_rate": 2.6600215268563396e-06,
+      "loss": 0.7066,
+      "step": 6098
+    },
+    {
+      "epoch": 0.495330138877609,
+      "grad_norm": 6.723227949020278,
+      "learning_rate": 2.6593652477081146e-06,
+      "loss": 0.5934,
+      "step": 6099
+    },
+    {
+      "epoch": 0.4954113538536506,
+      "grad_norm": 6.836596280031364,
+      "learning_rate": 2.658708957532657e-06,
+      "loss": 0.4197,
+      "step": 6100
+    },
+    {
+      "epoch": 0.49549256882969217,
+      "grad_norm": 5.530454100104951,
+      "learning_rate": 2.6580526563753794e-06,
+      "loss": 0.5521,
+      "step": 6101
+    },
+    {
+      "epoch": 0.4955737838057338,
+      "grad_norm": 8.014803649260598,
+      "learning_rate": 2.6573963442816957e-06,
+      "loss": 0.4716,
+      "step": 6102
+    },
+    {
+      "epoch": 0.4956549987817754,
+      "grad_norm": 5.651578109021286,
+      "learning_rate": 2.656740021297017e-06,
+      "loss": 0.6337,
+      "step": 6103
+    },
+    {
+      "epoch": 0.49573621375781696,
+      "grad_norm": 4.497707754410902,
+      "learning_rate": 2.6560836874667584e-06,
+      "loss": 0.4835,
+      "step": 6104
+    },
+    {
+      "epoch": 0.49581742873385853,
+      "grad_norm": 6.748930438264841,
+      "learning_rate": 2.6554273428363352e-06,
+      "loss": 0.4341,
+      "step": 6105
+    },
+    {
+      "epoch": 0.4958986437099001,
+      "grad_norm": 4.813991480179138,
+      "learning_rate": 2.6547709874511622e-06,
+      "loss": 0.4681,
+      "step": 6106
+    },
+    {
+      "epoch": 0.4959798586859417,
+      "grad_norm": 4.661897396362085,
+      "learning_rate": 2.654114621356656e-06,
+      "loss": 0.4306,
+      "step": 6107
+    },
+    {
+      "epoch": 0.49606107366198327,
+      "grad_norm": 6.397578426337775,
+      "learning_rate": 2.6534582445982338e-06,
+      "loss": 0.4603,
+      "step": 6108
+    },
+    {
+      "epoch": 0.49614228863802484,
+      "grad_norm": 4.315053588094933,
+      "learning_rate": 2.6528018572213133e-06,
+      "loss": 0.6069,
+      "step": 6109
+    },
+    {
+      "epoch": 0.4962235036140664,
+      "grad_norm": 3.5982681139239876,
+      "learning_rate": 2.6521454592713125e-06,
+      "loss": 0.4975,
+      "step": 6110
+    },
+    {
+      "epoch": 0.496304718590108,
+      "grad_norm": 4.778399998810193,
+      "learning_rate": 2.6514890507936515e-06,
+      "loss": 0.4753,
+      "step": 6111
+    },
+    {
+      "epoch": 0.4963859335661496,
+      "grad_norm": 5.875526293696564,
+      "learning_rate": 2.6508326318337498e-06,
+      "loss": 0.6111,
+      "step": 6112
+    },
+    {
+      "epoch": 0.4964671485421912,
+      "grad_norm": 4.292823551355768,
+      "learning_rate": 2.6501762024370283e-06,
+      "loss": 0.3784,
+      "step": 6113
+    },
+    {
+      "epoch": 0.4965483635182328,
+      "grad_norm": 8.401173587655329,
+      "learning_rate": 2.6495197626489082e-06,
+      "loss": 0.582,
+      "step": 6114
+    },
+    {
+      "epoch": 0.49662957849427436,
+      "grad_norm": 4.321012772875746,
+      "learning_rate": 2.6488633125148127e-06,
+      "loss": 0.4502,
+      "step": 6115
+    },
+    {
+      "epoch": 0.49671079347031594,
+      "grad_norm": 5.379767516829725,
+      "learning_rate": 2.6482068520801625e-06,
+      "loss": 0.5668,
+      "step": 6116
+    },
+    {
+      "epoch": 0.4967920084463575,
+      "grad_norm": 5.218798786070938,
+      "learning_rate": 2.647550381390383e-06,
+      "loss": 0.4885,
+      "step": 6117
+    },
+    {
+      "epoch": 0.4968732234223991,
+      "grad_norm": 6.280161135864912,
+      "learning_rate": 2.6468939004908987e-06,
+      "loss": 0.5571,
+      "step": 6118
+    },
+    {
+      "epoch": 0.49695443839844067,
+      "grad_norm": 6.147745243094484,
+      "learning_rate": 2.646237409427133e-06,
+      "loss": 0.6213,
+      "step": 6119
+    },
+    {
+      "epoch": 0.49703565337448224,
+      "grad_norm": 4.237997643901119,
+      "learning_rate": 2.645580908244513e-06,
+      "loss": 0.5256,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4971168683505238,
+      "grad_norm": 5.794983422357201,
+      "learning_rate": 2.644924396988465e-06,
+      "loss": 0.4717,
+      "step": 6121
+    },
+    {
+      "epoch": 0.4971980833265654,
+      "grad_norm": 3.028956483672569,
+      "learning_rate": 2.644267875704415e-06,
+      "loss": 0.605,
+      "step": 6122
+    },
+    {
+      "epoch": 0.497279298302607,
+      "grad_norm": 4.783447111346257,
+      "learning_rate": 2.6436113444377916e-06,
+      "loss": 0.3828,
+      "step": 6123
+    },
+    {
+      "epoch": 0.4973605132786486,
+      "grad_norm": 4.634896048686429,
+      "learning_rate": 2.6429548032340233e-06,
+      "loss": 0.4624,
+      "step": 6124
+    },
+    {
+      "epoch": 0.4974417282546902,
+      "grad_norm": 4.4180065168228575,
+      "learning_rate": 2.642298252138539e-06,
+      "loss": 0.471,
+      "step": 6125
+    },
+    {
+      "epoch": 0.49752294323073176,
+      "grad_norm": 7.029597857740861,
+      "learning_rate": 2.641641691196769e-06,
+      "loss": 0.3827,
+      "step": 6126
+    },
+    {
+      "epoch": 0.49760415820677334,
+      "grad_norm": 3.5588276162139922,
+      "learning_rate": 2.6409851204541435e-06,
+      "loss": 0.634,
+      "step": 6127
+    },
+    {
+      "epoch": 0.4976853731828149,
+      "grad_norm": 5.229212123221044,
+      "learning_rate": 2.640328539956094e-06,
+      "loss": 0.4562,
+      "step": 6128
+    },
+    {
+      "epoch": 0.4977665881588565,
+      "grad_norm": 4.266180122295215,
+      "learning_rate": 2.639671949748052e-06,
+      "loss": 0.5074,
+      "step": 6129
+    },
+    {
+      "epoch": 0.49784780313489807,
+      "grad_norm": 5.993522414034609,
+      "learning_rate": 2.6390153498754506e-06,
+      "loss": 0.5987,
+      "step": 6130
+    },
+    {
+      "epoch": 0.49792901811093965,
+      "grad_norm": 5.432626385109642,
+      "learning_rate": 2.638358740383723e-06,
+      "loss": 0.5943,
+      "step": 6131
+    },
+    {
+      "epoch": 0.4980102330869812,
+      "grad_norm": 7.744059010319982,
+      "learning_rate": 2.637702121318302e-06,
+      "loss": 0.4533,
+      "step": 6132
+    },
+    {
+      "epoch": 0.4980914480630228,
+      "grad_norm": 6.471183582152214,
+      "learning_rate": 2.6370454927246237e-06,
+      "loss": 0.4274,
+      "step": 6133
+    },
+    {
+      "epoch": 0.4981726630390644,
+      "grad_norm": 3.457874851254734,
+      "learning_rate": 2.6363888546481224e-06,
+      "loss": 0.4676,
+      "step": 6134
+    },
+    {
+      "epoch": 0.498253878015106,
+      "grad_norm": 8.481196069942994,
+      "learning_rate": 2.635732207134234e-06,
+      "loss": 0.4627,
+      "step": 6135
+    },
+    {
+      "epoch": 0.4983350929911476,
+      "grad_norm": 5.660075404955752,
+      "learning_rate": 2.635075550228395e-06,
+      "loss": 0.4784,
+      "step": 6136
+    },
+    {
+      "epoch": 0.49841630796718916,
+      "grad_norm": 4.313279838626725,
+      "learning_rate": 2.634418883976043e-06,
+      "loss": 0.4569,
+      "step": 6137
+    },
+    {
+      "epoch": 0.49849752294323074,
+      "grad_norm": 3.362477204863664,
+      "learning_rate": 2.6337622084226163e-06,
+      "loss": 0.5727,
+      "step": 6138
+    },
+    {
+      "epoch": 0.4985787379192723,
+      "grad_norm": 4.485740579944916,
+      "learning_rate": 2.633105523613551e-06,
+      "loss": 0.7045,
+      "step": 6139
+    },
+    {
+      "epoch": 0.4986599528953139,
+      "grad_norm": 3.989314569645018,
+      "learning_rate": 2.6324488295942897e-06,
+      "loss": 0.693,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4987411678713555,
+      "grad_norm": 6.808212058806507,
+      "learning_rate": 2.6317921264102697e-06,
+      "loss": 0.4904,
+      "step": 6141
+    },
+    {
+      "epoch": 0.49882238284739705,
+      "grad_norm": 4.01346195857619,
+      "learning_rate": 2.6311354141069324e-06,
+      "loss": 0.6021,
+      "step": 6142
+    },
+    {
+      "epoch": 0.4989035978234386,
+      "grad_norm": 5.471950249286309,
+      "learning_rate": 2.630478692729718e-06,
+      "loss": 0.5009,
+      "step": 6143
+    },
+    {
+      "epoch": 0.4989848127994802,
+      "grad_norm": 4.820722399207098,
+      "learning_rate": 2.6298219623240685e-06,
+      "loss": 0.3804,
+      "step": 6144
+    },
+    {
+      "epoch": 0.4990660277755218,
+      "grad_norm": 3.164263910898727,
+      "learning_rate": 2.6291652229354264e-06,
+      "loss": 0.4665,
+      "step": 6145
+    },
+    {
+      "epoch": 0.4991472427515634,
+      "grad_norm": 5.584269076148355,
+      "learning_rate": 2.6285084746092347e-06,
+      "loss": 0.5254,
+      "step": 6146
+    },
+    {
+      "epoch": 0.499228457727605,
+      "grad_norm": 10.9678056140828,
+      "learning_rate": 2.627851717390936e-06,
+      "loss": 0.4869,
+      "step": 6147
+    },
+    {
+      "epoch": 0.49930967270364657,
+      "grad_norm": 6.984083124060342,
+      "learning_rate": 2.6271949513259764e-06,
+      "loss": 0.5502,
+      "step": 6148
+    },
+    {
+      "epoch": 0.49939088767968814,
+      "grad_norm": 4.28910430937024,
+      "learning_rate": 2.626538176459798e-06,
+      "loss": 0.5626,
+      "step": 6149
+    },
+    {
+      "epoch": 0.4994721026557297,
+      "grad_norm": 5.809158832217803,
+      "learning_rate": 2.625881392837849e-06,
+      "loss": 0.5614,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4995533176317713,
+      "grad_norm": 4.895799235690263,
+      "learning_rate": 2.6252246005055725e-06,
+      "loss": 0.5047,
+      "step": 6151
+    },
+    {
+      "epoch": 0.4996345326078129,
+      "grad_norm": 3.584830115023326,
+      "learning_rate": 2.6245677995084163e-06,
+      "loss": 0.555,
+      "step": 6152
+    },
+    {
+      "epoch": 0.49971574758385445,
+      "grad_norm": 10.564765809267739,
+      "learning_rate": 2.6239109898918286e-06,
+      "loss": 0.4141,
+      "step": 6153
+    },
+    {
+      "epoch": 0.49979696255989603,
+      "grad_norm": 3.816580959787303,
+      "learning_rate": 2.6232541717012563e-06,
+      "loss": 0.5562,
+      "step": 6154
+    },
+    {
+      "epoch": 0.4998781775359376,
+      "grad_norm": 7.68222211219942,
+      "learning_rate": 2.6225973449821468e-06,
+      "loss": 0.4434,
+      "step": 6155
+    },
+    {
+      "epoch": 0.4999593925119792,
+      "grad_norm": 4.677405643402403,
+      "learning_rate": 2.6219405097799498e-06,
+      "loss": 0.67,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5000406074880208,
+      "grad_norm": 6.735859677395967,
+      "learning_rate": 2.6212836661401154e-06,
+      "loss": 0.5044,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5001218224640623,
+      "grad_norm": 5.559869392207337,
+      "learning_rate": 2.6206268141080924e-06,
+      "loss": 0.4942,
+      "step": 6158
+    },
+    {
+      "epoch": 0.500203037440104,
+      "grad_norm": 5.91589064626228,
+      "learning_rate": 2.619969953729333e-06,
+      "loss": 0.5946,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5002842524161455,
+      "grad_norm": 9.884104436935436,
+      "learning_rate": 2.6193130850492876e-06,
+      "loss": 0.6033,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5003654673921871,
+      "grad_norm": 5.191171578581443,
+      "learning_rate": 2.618656208113408e-06,
+      "loss": 0.4999,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5004466823682288,
+      "grad_norm": 4.869819153010546,
+      "learning_rate": 2.6179993229671473e-06,
+      "loss": 0.5341,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5005278973442703,
+      "grad_norm": 4.257762076221053,
+      "learning_rate": 2.6173424296559575e-06,
+      "loss": 0.5408,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5006091123203119,
+      "grad_norm": 4.96043740363745,
+      "learning_rate": 2.6166855282252933e-06,
+      "loss": 0.5863,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5006903272963534,
+      "grad_norm": 4.664873730144854,
+      "learning_rate": 2.616028618720607e-06,
+      "loss": 0.4936,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5007715422723951,
+      "grad_norm": 5.757723223582411,
+      "learning_rate": 2.615371701187355e-06,
+      "loss": 0.5604,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5008527572484366,
+      "grad_norm": 5.55920229153415,
+      "learning_rate": 2.6147147756709925e-06,
+      "loss": 0.5215,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5009339722244782,
+      "grad_norm": 3.916022531924741,
+      "learning_rate": 2.614057842216973e-06,
+      "loss": 0.6314,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5010151872005197,
+      "grad_norm": 3.748268469101728,
+      "learning_rate": 2.6134009008707555e-06,
+      "loss": 0.5058,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5010964021765614,
+      "grad_norm": 3.866469109499023,
+      "learning_rate": 2.6127439516777956e-06,
+      "loss": 0.6175,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5011776171526029,
+      "grad_norm": 9.784917821882507,
+      "learning_rate": 2.6120869946835513e-06,
+      "loss": 0.4369,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5012588321286445,
+      "grad_norm": 3.918198159806199,
+      "learning_rate": 2.61143002993348e-06,
+      "loss": 0.5966,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5013400471046862,
+      "grad_norm": 9.939958648361882,
+      "learning_rate": 2.61077305747304e-06,
+      "loss": 0.5353,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5014212620807277,
+      "grad_norm": 6.322410360361211,
+      "learning_rate": 2.610116077347691e-06,
+      "loss": 0.4513,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5015024770567693,
+      "grad_norm": 6.622717663567328,
+      "learning_rate": 2.609459089602892e-06,
+      "loss": 0.3957,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5015836920328108,
+      "grad_norm": 8.064816332776429,
+      "learning_rate": 2.6088020942841034e-06,
+      "loss": 0.3698,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5016649070088525,
+      "grad_norm": 7.086132865620177,
+      "learning_rate": 2.6081450914367864e-06,
+      "loss": 0.5031,
+      "step": 6177
+    },
+    {
+      "epoch": 0.501746121984894,
+      "grad_norm": 3.7954526083234077,
+      "learning_rate": 2.6074880811064003e-06,
+      "loss": 0.5348,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5018273369609356,
+      "grad_norm": 5.020997847466201,
+      "learning_rate": 2.606831063338408e-06,
+      "loss": 0.4605,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5019085519369771,
+      "grad_norm": 4.560830832766969,
+      "learning_rate": 2.6061740381782723e-06,
+      "loss": 0.5455,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5019897669130188,
+      "grad_norm": 4.229120124206621,
+      "learning_rate": 2.605517005671454e-06,
+      "loss": 0.5157,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5020709818890603,
+      "grad_norm": 4.581886089334441,
+      "learning_rate": 2.604859965863418e-06,
+      "loss": 0.4667,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5021521968651019,
+      "grad_norm": 4.4186999593426055,
+      "learning_rate": 2.6042029187996277e-06,
+      "loss": 0.5935,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5022334118411436,
+      "grad_norm": 4.736522021215294,
+      "learning_rate": 2.6035458645255467e-06,
+      "loss": 0.5252,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5023146268171851,
+      "grad_norm": 4.668585962104124,
+      "learning_rate": 2.602888803086639e-06,
+      "loss": 0.5201,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5023958417932267,
+      "grad_norm": 9.66768445142635,
+      "learning_rate": 2.602231734528372e-06,
+      "loss": 0.4524,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5024770567692682,
+      "grad_norm": 4.15432156915695,
+      "learning_rate": 2.601574658896209e-06,
+      "loss": 0.6911,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5025582717453099,
+      "grad_norm": 5.036750129329255,
+      "learning_rate": 2.6009175762356176e-06,
+      "loss": 0.587,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5026394867213514,
+      "grad_norm": 9.244386485882309,
+      "learning_rate": 2.6002604865920645e-06,
+      "loss": 0.5369,
+      "step": 6189
+    },
+    {
+      "epoch": 0.502720701697393,
+      "grad_norm": 4.712476893869534,
+      "learning_rate": 2.5996033900110155e-06,
+      "loss": 0.6125,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5028019166734345,
+      "grad_norm": 7.034161631409917,
+      "learning_rate": 2.5989462865379394e-06,
+      "loss": 0.6711,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5028831316494762,
+      "grad_norm": 2.537027536386944,
+      "learning_rate": 2.598289176218304e-06,
+      "loss": 0.5937,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5029643466255177,
+      "grad_norm": 3.394185226747455,
+      "learning_rate": 2.597632059097577e-06,
+      "loss": 0.6653,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5030455616015593,
+      "grad_norm": 4.376129626575339,
+      "learning_rate": 2.5969749352212294e-06,
+      "loss": 0.6337,
+      "step": 6194
+    },
+    {
+      "epoch": 0.503126776577601,
+      "grad_norm": 4.748486742998477,
+      "learning_rate": 2.5963178046347286e-06,
+      "loss": 0.4952,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5032079915536425,
+      "grad_norm": 7.369302442783826,
+      "learning_rate": 2.595660667383547e-06,
+      "loss": 0.3893,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5032892065296841,
+      "grad_norm": 5.509513918719035,
+      "learning_rate": 2.5950035235131515e-06,
+      "loss": 0.5277,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5033704215057256,
+      "grad_norm": 5.829645133957113,
+      "learning_rate": 2.594346373069016e-06,
+      "loss": 0.4237,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5034516364817673,
+      "grad_norm": 5.626826411794362,
+      "learning_rate": 2.593689216096611e-06,
+      "loss": 0.4906,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5035328514578088,
+      "grad_norm": 6.977550708147046,
+      "learning_rate": 2.5930320526414083e-06,
+      "loss": 0.548,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5036140664338504,
+      "grad_norm": 7.209561596520351,
+      "learning_rate": 2.592374882748879e-06,
+      "loss": 0.4672,
+      "step": 6201
+    },
+    {
+      "epoch": 0.503695281409892,
+      "grad_norm": 5.543565192935603,
+      "learning_rate": 2.5917177064644974e-06,
+      "loss": 0.5535,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5037764963859336,
+      "grad_norm": 5.366571236178294,
+      "learning_rate": 2.5910605238337355e-06,
+      "loss": 0.6972,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5038577113619751,
+      "grad_norm": 6.261566607697583,
+      "learning_rate": 2.5904033349020675e-06,
+      "loss": 0.4503,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5039389263380167,
+      "grad_norm": 5.805760768553691,
+      "learning_rate": 2.589746139714967e-06,
+      "loss": 0.4825,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5040201413140584,
+      "grad_norm": 8.328312111239873,
+      "learning_rate": 2.5890889383179086e-06,
+      "loss": 0.4764,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5041013562900999,
+      "grad_norm": 4.810122399520198,
+      "learning_rate": 2.588431730756367e-06,
+      "loss": 0.5084,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5041825712661415,
+      "grad_norm": 3.7065509135197545,
+      "learning_rate": 2.5877745170758177e-06,
+      "loss": 0.524,
+      "step": 6208
+    },
+    {
+      "epoch": 0.504263786242183,
+      "grad_norm": 5.121308484330212,
+      "learning_rate": 2.5871172973217367e-06,
+      "loss": 0.5104,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5043450012182247,
+      "grad_norm": 5.928958710988731,
+      "learning_rate": 2.5864600715396e-06,
+      "loss": 0.5344,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5044262161942662,
+      "grad_norm": 5.887834972208363,
+      "learning_rate": 2.585802839774883e-06,
+      "loss": 0.4809,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5045074311703078,
+      "grad_norm": 4.367368393282637,
+      "learning_rate": 2.5851456020730643e-06,
+      "loss": 0.6071,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5045886461463494,
+      "grad_norm": 3.4461457206000374,
+      "learning_rate": 2.584488358479621e-06,
+      "loss": 0.6266,
+      "step": 6213
+    },
+    {
+      "epoch": 0.504669861122391,
+      "grad_norm": 3.1291003154252004,
+      "learning_rate": 2.5838311090400293e-06,
+      "loss": 0.5782,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5047510760984325,
+      "grad_norm": 9.218149230525775,
+      "learning_rate": 2.58317385379977e-06,
+      "loss": 0.395,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5048322910744741,
+      "grad_norm": 7.375352110089412,
+      "learning_rate": 2.582516592804319e-06,
+      "loss": 0.4645,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5049135060505158,
+      "grad_norm": 6.15235163146789,
+      "learning_rate": 2.5818593260991565e-06,
+      "loss": 0.514,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5049947210265573,
+      "grad_norm": 16.29462482704463,
+      "learning_rate": 2.581202053729762e-06,
+      "loss": 0.5103,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5050759360025989,
+      "grad_norm": 4.3058949775554884,
+      "learning_rate": 2.580544775741616e-06,
+      "loss": 0.5023,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5051571509786404,
+      "grad_norm": 3.2866867580977606,
+      "learning_rate": 2.579887492180197e-06,
+      "loss": 0.6187,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5052383659546821,
+      "grad_norm": 5.219156744499874,
+      "learning_rate": 2.579230203090986e-06,
+      "loss": 0.3933,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5053195809307236,
+      "grad_norm": 4.147926897905404,
+      "learning_rate": 2.578572908519465e-06,
+      "loss": 0.506,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5054007959067652,
+      "grad_norm": 4.123405600563359,
+      "learning_rate": 2.577915608511114e-06,
+      "loss": 0.5677,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5054820108828068,
+      "grad_norm": 7.522333120828346,
+      "learning_rate": 2.5772583031114157e-06,
+      "loss": 0.4562,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5055632258588484,
+      "grad_norm": 6.046806353687886,
+      "learning_rate": 2.5766009923658516e-06,
+      "loss": 0.3849,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5056444408348899,
+      "grad_norm": 4.707865925528092,
+      "learning_rate": 2.5759436763199047e-06,
+      "loss": 0.4868,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5057256558109315,
+      "grad_norm": 4.715913190140127,
+      "learning_rate": 2.575286355019056e-06,
+      "loss": 0.587,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5058068707869732,
+      "grad_norm": 3.2497965234461264,
+      "learning_rate": 2.5746290285087912e-06,
+      "loss": 0.7003,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5058880857630147,
+      "grad_norm": 3.280084946712466,
+      "learning_rate": 2.5739716968345922e-06,
+      "loss": 0.6808,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5059693007390563,
+      "grad_norm": 5.724019303106518,
+      "learning_rate": 2.573314360041943e-06,
+      "loss": 0.4032,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5060505157150978,
+      "grad_norm": 4.723214741303482,
+      "learning_rate": 2.5726570181763286e-06,
+      "loss": 0.4832,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5061317306911395,
+      "grad_norm": 3.7300991619481363,
+      "learning_rate": 2.571999671283233e-06,
+      "loss": 0.4762,
+      "step": 6232
+    },
+    {
+      "epoch": 0.506212945667181,
+      "grad_norm": 4.889315053935509,
+      "learning_rate": 2.5713423194081404e-06,
+      "loss": 0.5565,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5062941606432226,
+      "grad_norm": 7.67062134201034,
+      "learning_rate": 2.570684962596538e-06,
+      "loss": 0.5569,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5063753756192642,
+      "grad_norm": 4.100525163986654,
+      "learning_rate": 2.5700276008939096e-06,
+      "loss": 0.4171,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5064565905953058,
+      "grad_norm": 4.883509578878357,
+      "learning_rate": 2.569370234345742e-06,
+      "loss": 0.4534,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5065378055713473,
+      "grad_norm": 17.999174929712392,
+      "learning_rate": 2.568712862997522e-06,
+      "loss": 0.4791,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5066190205473889,
+      "grad_norm": 3.501500494032202,
+      "learning_rate": 2.5680554868947346e-06,
+      "loss": 0.3923,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5067002355234306,
+      "grad_norm": 5.490863237584197,
+      "learning_rate": 2.5673981060828672e-06,
+      "loss": 0.557,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5067814504994721,
+      "grad_norm": 4.017635335289608,
+      "learning_rate": 2.5667407206074084e-06,
+      "loss": 0.4779,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5068626654755137,
+      "grad_norm": 5.193790698559395,
+      "learning_rate": 2.566083330513845e-06,
+      "loss": 0.5404,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5069438804515553,
+      "grad_norm": 3.781621395266743,
+      "learning_rate": 2.565425935847665e-06,
+      "loss": 0.5054,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5070250954275969,
+      "grad_norm": 6.125227421255686,
+      "learning_rate": 2.564768536654356e-06,
+      "loss": 0.565,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5071063104036384,
+      "grad_norm": 5.159395575524112,
+      "learning_rate": 2.564111132979407e-06,
+      "loss": 0.5639,
+      "step": 6244
+    },
+    {
+      "epoch": 0.50718752537968,
+      "grad_norm": 5.800011134638956,
+      "learning_rate": 2.563453724868308e-06,
+      "loss": 0.4247,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5072687403557216,
+      "grad_norm": 3.2153805898476606,
+      "learning_rate": 2.5627963123665455e-06,
+      "loss": 0.4201,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5073499553317632,
+      "grad_norm": 3.5092901295905405,
+      "learning_rate": 2.5621388955196113e-06,
+      "loss": 0.472,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5074311703078047,
+      "grad_norm": 8.695984681785939,
+      "learning_rate": 2.561481474372995e-06,
+      "loss": 0.4915,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5075123852838463,
+      "grad_norm": 4.710656984668509,
+      "learning_rate": 2.560824048972185e-06,
+      "loss": 0.4936,
+      "step": 6249
+    },
+    {
+      "epoch": 0.507593600259888,
+      "grad_norm": 7.293645049422129,
+      "learning_rate": 2.5601666193626735e-06,
+      "loss": 0.6042,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5076748152359295,
+      "grad_norm": 5.2621850945387285,
+      "learning_rate": 2.55950918558995e-06,
+      "loss": 0.443,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5077560302119711,
+      "grad_norm": 3.4152519821072578,
+      "learning_rate": 2.558851747699506e-06,
+      "loss": 0.697,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5078372451880127,
+      "grad_norm": 5.202402703658923,
+      "learning_rate": 2.5581943057368317e-06,
+      "loss": 0.4773,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5079184601640543,
+      "grad_norm": 7.606949152034,
+      "learning_rate": 2.5575368597474202e-06,
+      "loss": 0.3156,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5079996751400958,
+      "grad_norm": 3.929776691209423,
+      "learning_rate": 2.5568794097767624e-06,
+      "loss": 0.4507,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5080808901161374,
+      "grad_norm": 6.335298963383283,
+      "learning_rate": 2.5562219558703504e-06,
+      "loss": 0.6335,
+      "step": 6256
+    },
+    {
+      "epoch": 0.508162105092179,
+      "grad_norm": 7.8562095782475865,
+      "learning_rate": 2.555564498073677e-06,
+      "loss": 0.5108,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5082433200682206,
+      "grad_norm": 4.8650466475894145,
+      "learning_rate": 2.554907036432235e-06,
+      "loss": 0.5093,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5083245350442621,
+      "grad_norm": 5.229859161415607,
+      "learning_rate": 2.554249570991515e-06,
+      "loss": 0.5999,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5084057500203037,
+      "grad_norm": 4.359502656532959,
+      "learning_rate": 2.5535921017970123e-06,
+      "loss": 0.435,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5084869649963454,
+      "grad_norm": 15.539251038374521,
+      "learning_rate": 2.5529346288942203e-06,
+      "loss": 0.4546,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5085681799723869,
+      "grad_norm": 4.359760496582075,
+      "learning_rate": 2.5522771523286317e-06,
+      "loss": 0.4273,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5086493949484285,
+      "grad_norm": 5.6985722122365114,
+      "learning_rate": 2.551619672145741e-06,
+      "loss": 0.5175,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5087306099244701,
+      "grad_norm": 4.451602003719397,
+      "learning_rate": 2.5509621883910424e-06,
+      "loss": 0.5247,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5088118249005117,
+      "grad_norm": 6.12534073894256,
+      "learning_rate": 2.55030470111003e-06,
+      "loss": 0.4593,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5088930398765532,
+      "grad_norm": 5.010881145167957,
+      "learning_rate": 2.5496472103481984e-06,
+      "loss": 0.523,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5089742548525948,
+      "grad_norm": 6.63962229868484,
+      "learning_rate": 2.5489897161510425e-06,
+      "loss": 0.4728,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5090554698286364,
+      "grad_norm": 4.6745969311887485,
+      "learning_rate": 2.5483322185640575e-06,
+      "loss": 0.5094,
+      "step": 6268
+    },
+    {
+      "epoch": 0.509136684804678,
+      "grad_norm": 7.671233857618601,
+      "learning_rate": 2.547674717632739e-06,
+      "loss": 0.5129,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5092178997807195,
+      "grad_norm": 4.60865861862504,
+      "learning_rate": 2.547017213402582e-06,
+      "loss": 0.4483,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5092991147567612,
+      "grad_norm": 6.2621430489481265,
+      "learning_rate": 2.546359705919083e-06,
+      "loss": 0.4052,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5093803297328028,
+      "grad_norm": 5.501701524912082,
+      "learning_rate": 2.545702195227737e-06,
+      "loss": 0.504,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5094615447088443,
+      "grad_norm": 9.43685395208808,
+      "learning_rate": 2.545044681374042e-06,
+      "loss": 0.7419,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5095427596848859,
+      "grad_norm": 4.655667182094343,
+      "learning_rate": 2.544387164403493e-06,
+      "loss": 0.6057,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5096239746609275,
+      "grad_norm": 6.207757916402813,
+      "learning_rate": 2.543729644361587e-06,
+      "loss": 0.3979,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5097051896369691,
+      "grad_norm": 6.082343128329472,
+      "learning_rate": 2.5430721212938216e-06,
+      "loss": 0.4088,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5097864046130106,
+      "grad_norm": 4.720976085216124,
+      "learning_rate": 2.542414595245693e-06,
+      "loss": 0.4561,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5098676195890522,
+      "grad_norm": 4.73426813490108,
+      "learning_rate": 2.541757066262699e-06,
+      "loss": 0.6201,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5099488345650938,
+      "grad_norm": 3.383613942551711,
+      "learning_rate": 2.541099534390336e-06,
+      "loss": 0.4821,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5100300495411354,
+      "grad_norm": 8.433408631314988,
+      "learning_rate": 2.5404419996741042e-06,
+      "loss": 0.4494,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5101112645171769,
+      "grad_norm": 5.815113828274007,
+      "learning_rate": 2.5397844621594997e-06,
+      "loss": 0.5385,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5101924794932186,
+      "grad_norm": 4.687103952333867,
+      "learning_rate": 2.5391269218920202e-06,
+      "loss": 0.4344,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5102736944692602,
+      "grad_norm": 5.089752144632809,
+      "learning_rate": 2.5384693789171656e-06,
+      "loss": 0.4559,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5103549094453017,
+      "grad_norm": 5.854876777674836,
+      "learning_rate": 2.537811833280433e-06,
+      "loss": 0.5045,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5104361244213433,
+      "grad_norm": 4.799615858367655,
+      "learning_rate": 2.5371542850273224e-06,
+      "loss": 0.4898,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5105173393973849,
+      "grad_norm": 5.42400764526346,
+      "learning_rate": 2.5364967342033307e-06,
+      "loss": 0.6302,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5105985543734265,
+      "grad_norm": 4.933143417479578,
+      "learning_rate": 2.5358391808539597e-06,
+      "loss": 0.4977,
+      "step": 6287
+    },
+    {
+      "epoch": 0.510679769349468,
+      "grad_norm": 3.4223345719076113,
+      "learning_rate": 2.535181625024706e-06,
+      "loss": 0.4809,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5107609843255096,
+      "grad_norm": 9.061912369576826,
+      "learning_rate": 2.53452406676107e-06,
+      "loss": 0.3358,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5108421993015512,
+      "grad_norm": 10.525328047917656,
+      "learning_rate": 2.5338665061085518e-06,
+      "loss": 0.6054,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5109234142775928,
+      "grad_norm": 5.383474823413491,
+      "learning_rate": 2.5332089431126504e-06,
+      "loss": 0.6881,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5110046292536343,
+      "grad_norm": 11.602351673058207,
+      "learning_rate": 2.532551377818866e-06,
+      "loss": 0.6387,
+      "step": 6292
+    },
+    {
+      "epoch": 0.511085844229676,
+      "grad_norm": 4.317046810663442,
+      "learning_rate": 2.5318938102726985e-06,
+      "loss": 0.6418,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5111670592057176,
+      "grad_norm": 4.588601931278866,
+      "learning_rate": 2.5312362405196485e-06,
+      "loss": 0.4955,
+      "step": 6294
+    },
+    {
+      "epoch": 0.5112482741817591,
+      "grad_norm": 5.9618029800437204,
+      "learning_rate": 2.530578668605215e-06,
+      "loss": 0.4686,
+      "step": 6295
+    },
+    {
+      "epoch": 0.5113294891578007,
+      "grad_norm": 4.309047195757413,
+      "learning_rate": 2.5299210945749005e-06,
+      "loss": 0.624,
+      "step": 6296
+    },
+    {
+      "epoch": 0.5114107041338423,
+      "grad_norm": 6.062649085654284,
+      "learning_rate": 2.529263518474204e-06,
+      "loss": 0.5655,
+      "step": 6297
+    },
+    {
+      "epoch": 0.5114919191098839,
+      "grad_norm": 8.101899866072504,
+      "learning_rate": 2.5286059403486262e-06,
+      "loss": 0.6048,
+      "step": 6298
+    },
+    {
+      "epoch": 0.5115731340859254,
+      "grad_norm": 4.855939734927873,
+      "learning_rate": 2.52794836024367e-06,
+      "loss": 0.5202,
+      "step": 6299
+    },
+    {
+      "epoch": 0.511654349061967,
+      "grad_norm": 4.1533242601287315,
+      "learning_rate": 2.5272907782048343e-06,
+      "loss": 0.6758,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5117355640380086,
+      "grad_norm": 44.854170715654014,
+      "learning_rate": 2.526633194277622e-06,
+      "loss": 0.4042,
+      "step": 6301
+    },
+    {
+      "epoch": 0.5118167790140502,
+      "grad_norm": 4.477675057177277,
+      "learning_rate": 2.5259756085075333e-06,
+      "loss": 0.641,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5118979939900917,
+      "grad_norm": 5.00794940429514,
+      "learning_rate": 2.5253180209400697e-06,
+      "loss": 0.4153,
+      "step": 6303
+    },
+    {
+      "epoch": 0.5119792089661334,
+      "grad_norm": 35.355057930038036,
+      "learning_rate": 2.5246604316207327e-06,
+      "loss": 0.3842,
+      "step": 6304
+    },
+    {
+      "epoch": 0.512060423942175,
+      "grad_norm": 5.069001886917277,
+      "learning_rate": 2.524002840595025e-06,
+      "loss": 0.5766,
+      "step": 6305
+    },
+    {
+      "epoch": 0.5121416389182165,
+      "grad_norm": 4.7801204519502845,
+      "learning_rate": 2.523345247908448e-06,
+      "loss": 0.5046,
+      "step": 6306
+    },
+    {
+      "epoch": 0.5122228538942581,
+      "grad_norm": 13.094992788363395,
+      "learning_rate": 2.522687653606503e-06,
+      "loss": 0.4069,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5123040688702997,
+      "grad_norm": 3.4115277705733273,
+      "learning_rate": 2.5220300577346925e-06,
+      "loss": 0.6193,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5123852838463413,
+      "grad_norm": 4.934143576608908,
+      "learning_rate": 2.521372460338518e-06,
+      "loss": 0.4763,
+      "step": 6309
+    },
+    {
+      "epoch": 0.5124664988223828,
+      "grad_norm": 4.955171281673901,
+      "learning_rate": 2.5207148614634836e-06,
+      "loss": 0.5961,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5125477137984245,
+      "grad_norm": 5.804504802010579,
+      "learning_rate": 2.5200572611550893e-06,
+      "loss": 0.4608,
+      "step": 6311
+    },
+    {
+      "epoch": 0.512628928774466,
+      "grad_norm": 3.3522351499110346,
+      "learning_rate": 2.5193996594588395e-06,
+      "loss": 0.5245,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5127101437505076,
+      "grad_norm": 4.503464521908895,
+      "learning_rate": 2.5187420564202357e-06,
+      "loss": 0.6784,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5127913587265491,
+      "grad_norm": 4.619599612350149,
+      "learning_rate": 2.518084452084781e-06,
+      "loss": 0.5239,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5128725737025908,
+      "grad_norm": 6.2490361812009745,
+      "learning_rate": 2.5174268464979775e-06,
+      "loss": 0.4502,
+      "step": 6315
+    },
+    {
+      "epoch": 0.5129537886786324,
+      "grad_norm": 6.493719400660229,
+      "learning_rate": 2.516769239705328e-06,
+      "loss": 0.45,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5130350036546739,
+      "grad_norm": 15.674054593332599,
+      "learning_rate": 2.5161116317523367e-06,
+      "loss": 0.5184,
+      "step": 6317
+    },
+    {
+      "epoch": 0.5131162186307155,
+      "grad_norm": 5.072386439299523,
+      "learning_rate": 2.5154540226845053e-06,
+      "loss": 0.4148,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5131974336067571,
+      "grad_norm": 6.330463701552802,
+      "learning_rate": 2.514796412547337e-06,
+      "loss": 0.6593,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5132786485827987,
+      "grad_norm": 6.329569453939536,
+      "learning_rate": 2.5141388013863366e-06,
+      "loss": 0.6006,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5133598635588402,
+      "grad_norm": 4.995421040496941,
+      "learning_rate": 2.5134811892470046e-06,
+      "loss": 0.4651,
+      "step": 6321
+    },
+    {
+      "epoch": 0.5134410785348819,
+      "grad_norm": 6.312161136312402,
+      "learning_rate": 2.512823576174846e-06,
+      "loss": 0.34,
+      "step": 6322
+    },
+    {
+      "epoch": 0.5135222935109234,
+      "grad_norm": 4.377238702155564,
+      "learning_rate": 2.5121659622153643e-06,
+      "loss": 0.4297,
+      "step": 6323
+    },
+    {
+      "epoch": 0.513603508486965,
+      "grad_norm": 4.928516077514282,
+      "learning_rate": 2.511508347414062e-06,
+      "loss": 0.5385,
+      "step": 6324
+    },
+    {
+      "epoch": 0.5136847234630065,
+      "grad_norm": 7.762372366845343,
+      "learning_rate": 2.510850731816443e-06,
+      "loss": 0.3269,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5137659384390482,
+      "grad_norm": 5.504193583882282,
+      "learning_rate": 2.510193115468011e-06,
+      "loss": 0.3859,
+      "step": 6326
+    },
+    {
+      "epoch": 0.5138471534150898,
+      "grad_norm": 6.152137635079484,
+      "learning_rate": 2.5095354984142682e-06,
+      "loss": 0.4028,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5139283683911313,
+      "grad_norm": 7.085948581559598,
+      "learning_rate": 2.5088778807007203e-06,
+      "loss": 0.3954,
+      "step": 6328
+    },
+    {
+      "epoch": 0.514009583367173,
+      "grad_norm": 5.649925226999557,
+      "learning_rate": 2.5082202623728707e-06,
+      "loss": 0.4362,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5140907983432145,
+      "grad_norm": 7.520999779577865,
+      "learning_rate": 2.507562643476222e-06,
+      "loss": 0.7279,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5141720133192561,
+      "grad_norm": 3.4023713944158143,
+      "learning_rate": 2.5069050240562782e-06,
+      "loss": 0.745,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5142532282952976,
+      "grad_norm": 5.4405959398028125,
+      "learning_rate": 2.5062474041585432e-06,
+      "loss": 0.5726,
+      "step": 6332
+    },
+    {
+      "epoch": 0.5143344432713393,
+      "grad_norm": 5.517222494769225,
+      "learning_rate": 2.5055897838285207e-06,
+      "loss": 0.5267,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5144156582473808,
+      "grad_norm": 4.726623734798838,
+      "learning_rate": 2.504932163111715e-06,
+      "loss": 0.5198,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5144968732234224,
+      "grad_norm": 4.440896263973417,
+      "learning_rate": 2.5042745420536295e-06,
+      "loss": 0.5967,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5145780881994639,
+      "grad_norm": 3.37064042601589,
+      "learning_rate": 2.503616920699769e-06,
+      "loss": 0.5971,
+      "step": 6336
+    },
+    {
+      "epoch": 0.5146593031755056,
+      "grad_norm": 4.690221273319353,
+      "learning_rate": 2.502959299095636e-06,
+      "loss": 0.5007,
+      "step": 6337
+    },
+    {
+      "epoch": 0.5147405181515472,
+      "grad_norm": 4.908136131112301,
+      "learning_rate": 2.5023016772867353e-06,
+      "loss": 0.5678,
+      "step": 6338
+    },
+    {
+      "epoch": 0.5148217331275887,
+      "grad_norm": 5.443552173567539,
+      "learning_rate": 2.5016440553185718e-06,
+      "loss": 0.4118,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5149029481036304,
+      "grad_norm": 6.1838875143833665,
+      "learning_rate": 2.5009864332366467e-06,
+      "loss": 0.6241,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5149841630796719,
+      "grad_norm": 3.986184988106473,
+      "learning_rate": 2.5003288110864664e-06,
+      "loss": 0.5065,
+      "step": 6341
+    },
+    {
+      "epoch": 0.5150653780557135,
+      "grad_norm": 4.231672066174447,
+      "learning_rate": 2.4996711889135344e-06,
+      "loss": 0.6497,
+      "step": 6342
+    },
+    {
+      "epoch": 0.515146593031755,
+      "grad_norm": 5.815390035755676,
+      "learning_rate": 2.499013566763354e-06,
+      "loss": 0.4079,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5152278080077967,
+      "grad_norm": 7.9133098452043455,
+      "learning_rate": 2.4983559446814295e-06,
+      "loss": 0.438,
+      "step": 6344
+    },
+    {
+      "epoch": 0.5153090229838382,
+      "grad_norm": 3.207812612582073,
+      "learning_rate": 2.497698322713265e-06,
+      "loss": 0.5876,
+      "step": 6345
+    },
+    {
+      "epoch": 0.5153902379598798,
+      "grad_norm": 4.353200703794427,
+      "learning_rate": 2.4970407009043646e-06,
+      "loss": 0.5272,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5154714529359213,
+      "grad_norm": 5.892469802972017,
+      "learning_rate": 2.4963830793002313e-06,
+      "loss": 0.4392,
+      "step": 6347
+    },
+    {
+      "epoch": 0.515552667911963,
+      "grad_norm": 7.87771121157062,
+      "learning_rate": 2.495725457946371e-06,
+      "loss": 0.326,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5156338828880046,
+      "grad_norm": 4.905801737169398,
+      "learning_rate": 2.4950678368882863e-06,
+      "loss": 0.5542,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5157150978640461,
+      "grad_norm": 3.2384008143696117,
+      "learning_rate": 2.49441021617148e-06,
+      "loss": 0.5257,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5157963128400878,
+      "grad_norm": 8.399266140485144,
+      "learning_rate": 2.4937525958414576e-06,
+      "loss": 0.6298,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5158775278161293,
+      "grad_norm": 6.628609352541173,
+      "learning_rate": 2.4930949759437234e-06,
+      "loss": 0.3812,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5159587427921709,
+      "grad_norm": 7.882154633725088,
+      "learning_rate": 2.492437356523779e-06,
+      "loss": 0.5552,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5160399577682124,
+      "grad_norm": 9.337887778597489,
+      "learning_rate": 2.4917797376271297e-06,
+      "loss": 0.5093,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5161211727442541,
+      "grad_norm": 7.376950318246357,
+      "learning_rate": 2.49112211929928e-06,
+      "loss": 0.5847,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5162023877202956,
+      "grad_norm": 6.482321048724295,
+      "learning_rate": 2.4904645015857318e-06,
+      "loss": 0.5784,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5162836026963372,
+      "grad_norm": 5.797801300962893,
+      "learning_rate": 2.48980688453199e-06,
+      "loss": 0.5229,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5163648176723787,
+      "grad_norm": 5.360257972509032,
+      "learning_rate": 2.4891492681835584e-06,
+      "loss": 0.4893,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5164460326484204,
+      "grad_norm": 5.202599723312153,
+      "learning_rate": 2.4884916525859386e-06,
+      "loss": 0.4775,
+      "step": 6359
+    },
+    {
+      "epoch": 0.516527247624462,
+      "grad_norm": 6.783080832818119,
+      "learning_rate": 2.4878340377846365e-06,
+      "loss": 0.4474,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5166084626005035,
+      "grad_norm": 12.845276601042336,
+      "learning_rate": 2.4871764238251547e-06,
+      "loss": 0.4225,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5166896775765452,
+      "grad_norm": 7.494098434686925,
+      "learning_rate": 2.4865188107529963e-06,
+      "loss": 0.5043,
+      "step": 6362
+    },
+    {
+      "epoch": 0.5167708925525867,
+      "grad_norm": 5.004627980302835,
+      "learning_rate": 2.485861198613664e-06,
+      "loss": 0.4393,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5168521075286283,
+      "grad_norm": 6.726551242092707,
+      "learning_rate": 2.4852035874526632e-06,
+      "loss": 0.5499,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5169333225046698,
+      "grad_norm": 4.666439480933289,
+      "learning_rate": 2.4845459773154964e-06,
+      "loss": 0.5416,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5170145374807115,
+      "grad_norm": 5.180749129277349,
+      "learning_rate": 2.483888368247664e-06,
+      "loss": 0.5082,
+      "step": 6366
+    },
+    {
+      "epoch": 0.517095752456753,
+      "grad_norm": 4.494845006036604,
+      "learning_rate": 2.4832307602946726e-06,
+      "loss": 0.4328,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5171769674327946,
+      "grad_norm": 7.722256382211164,
+      "learning_rate": 2.4825731535020242e-06,
+      "loss": 0.4709,
+      "step": 6368
+    },
+    {
+      "epoch": 0.5172581824088361,
+      "grad_norm": 5.336177066714621,
+      "learning_rate": 2.48191554791522e-06,
+      "loss": 0.5804,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5173393973848778,
+      "grad_norm": 6.222339347839942,
+      "learning_rate": 2.481257943579765e-06,
+      "loss": 0.4719,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5174206123609194,
+      "grad_norm": 7.613124790360803,
+      "learning_rate": 2.4806003405411617e-06,
+      "loss": 0.6593,
+      "step": 6371
+    },
+    {
+      "epoch": 0.5175018273369609,
+      "grad_norm": 6.354106819523852,
+      "learning_rate": 2.479942738844911e-06,
+      "loss": 0.4365,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5175830423130026,
+      "grad_norm": 5.688366482514378,
+      "learning_rate": 2.479285138536517e-06,
+      "loss": 0.5308,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5176642572890441,
+      "grad_norm": 5.415154693434397,
+      "learning_rate": 2.4786275396614823e-06,
+      "loss": 0.4804,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5177454722650857,
+      "grad_norm": 28.82102125102892,
+      "learning_rate": 2.477969942265308e-06,
+      "loss": 0.3669,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5178266872411272,
+      "grad_norm": 6.810481547484077,
+      "learning_rate": 2.4773123463934973e-06,
+      "loss": 0.6299,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5179079022171689,
+      "grad_norm": 4.392949418938218,
+      "learning_rate": 2.476654752091553e-06,
+      "loss": 0.5736,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5179891171932104,
+      "grad_norm": 4.2704517486876234,
+      "learning_rate": 2.4759971594049763e-06,
+      "loss": 0.5811,
+      "step": 6378
+    },
+    {
+      "epoch": 0.518070332169252,
+      "grad_norm": 6.692688448203209,
+      "learning_rate": 2.4753395683792677e-06,
+      "loss": 0.4761,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5181515471452935,
+      "grad_norm": 7.334413784638161,
+      "learning_rate": 2.474681979059931e-06,
+      "loss": 0.5694,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5182327621213352,
+      "grad_norm": 4.544276408270827,
+      "learning_rate": 2.474024391492468e-06,
+      "loss": 0.4231,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5183139770973768,
+      "grad_norm": 4.593897927457953,
+      "learning_rate": 2.473366805722379e-06,
+      "loss": 0.6074,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5183951920734183,
+      "grad_norm": 4.897073039683832,
+      "learning_rate": 2.472709221795166e-06,
+      "loss": 0.4831,
+      "step": 6383
+    },
+    {
+      "epoch": 0.51847640704946,
+      "grad_norm": 4.507107061670829,
+      "learning_rate": 2.4720516397563314e-06,
+      "loss": 0.5052,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5185576220255015,
+      "grad_norm": 4.582461812162122,
+      "learning_rate": 2.471394059651374e-06,
+      "loss": 0.6067,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5186388370015431,
+      "grad_norm": 5.866771836678322,
+      "learning_rate": 2.470736481525797e-06,
+      "loss": 0.6531,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5187200519775846,
+      "grad_norm": 5.813106703432218,
+      "learning_rate": 2.470078905425101e-06,
+      "loss": 0.4097,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5188012669536263,
+      "grad_norm": 7.253453752131115,
+      "learning_rate": 2.4694213313947855e-06,
+      "loss": 0.4998,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5188824819296678,
+      "grad_norm": 3.081751300748125,
+      "learning_rate": 2.4687637594803527e-06,
+      "loss": 0.5751,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5189636969057094,
+      "grad_norm": 5.45405710599491,
+      "learning_rate": 2.4681061897273028e-06,
+      "loss": 0.4083,
+      "step": 6390
+    },
+    {
+      "epoch": 0.519044911881751,
+      "grad_norm": 4.843125937608803,
+      "learning_rate": 2.4674486221811345e-06,
+      "loss": 0.4051,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5191261268577926,
+      "grad_norm": 6.00791556685988,
+      "learning_rate": 2.46679105688735e-06,
+      "loss": 0.5527,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5192073418338342,
+      "grad_norm": 3.547137678101548,
+      "learning_rate": 2.466133493891449e-06,
+      "loss": 0.4875,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5192885568098757,
+      "grad_norm": 8.66243792163446,
+      "learning_rate": 2.46547593323893e-06,
+      "loss": 0.496,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5193697717859174,
+      "grad_norm": 4.430513402849331,
+      "learning_rate": 2.464818374975295e-06,
+      "loss": 0.4364,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5194509867619589,
+      "grad_norm": 3.7975317298710625,
+      "learning_rate": 2.4641608191460415e-06,
+      "loss": 0.6302,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5195322017380005,
+      "grad_norm": 3.888846410063618,
+      "learning_rate": 2.46350326579667e-06,
+      "loss": 0.5352,
+      "step": 6397
+    },
+    {
+      "epoch": 0.519613416714042,
+      "grad_norm": 4.412943996478983,
+      "learning_rate": 2.462845714972679e-06,
+      "loss": 0.4854,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5196946316900837,
+      "grad_norm": 3.706708426188841,
+      "learning_rate": 2.4621881667195676e-06,
+      "loss": 0.4375,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5197758466661252,
+      "grad_norm": 3.8449883186160787,
+      "learning_rate": 2.4615306210828357e-06,
+      "loss": 0.5516,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5198570616421668,
+      "grad_norm": 5.909942632055917,
+      "learning_rate": 2.46087307810798e-06,
+      "loss": 0.5237,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5199382766182084,
+      "grad_norm": 3.547279382037591,
+      "learning_rate": 2.460215537840501e-06,
+      "loss": 0.4175,
+      "step": 6402
+    },
+    {
+      "epoch": 0.52001949159425,
+      "grad_norm": 4.882379724186628,
+      "learning_rate": 2.459558000325897e-06,
+      "loss": 0.6123,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5201007065702916,
+      "grad_norm": 4.050921637738143,
+      "learning_rate": 2.458900465609664e-06,
+      "loss": 0.5215,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5201819215463331,
+      "grad_norm": 3.382346078488962,
+      "learning_rate": 2.4582429337373018e-06,
+      "loss": 0.4617,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5202631365223748,
+      "grad_norm": 7.257748442971193,
+      "learning_rate": 2.4575854047543082e-06,
+      "loss": 0.4909,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5203443514984163,
+      "grad_norm": 6.459780012681047,
+      "learning_rate": 2.456927878706179e-06,
+      "loss": 0.4125,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5204255664744579,
+      "grad_norm": 3.8936939687249863,
+      "learning_rate": 2.4562703556384136e-06,
+      "loss": 0.5746,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5205067814504994,
+      "grad_norm": 5.650603779541892,
+      "learning_rate": 2.4556128355965076e-06,
+      "loss": 0.4194,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5205879964265411,
+      "grad_norm": 6.233819656387275,
+      "learning_rate": 2.454955318625958e-06,
+      "loss": 0.5296,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5206692114025826,
+      "grad_norm": 5.175873926721092,
+      "learning_rate": 2.4542978047722633e-06,
+      "loss": 0.5687,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5207504263786242,
+      "grad_norm": 7.441015494915213,
+      "learning_rate": 2.453640294080918e-06,
+      "loss": 0.7367,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5208316413546658,
+      "grad_norm": 7.282413948152037,
+      "learning_rate": 2.452982786597419e-06,
+      "loss": 0.5059,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5209128563307074,
+      "grad_norm": 5.063377741570186,
+      "learning_rate": 2.452325282367262e-06,
+      "loss": 0.4086,
+      "step": 6414
+    },
+    {
+      "epoch": 0.520994071306749,
+      "grad_norm": 4.377231675145512,
+      "learning_rate": 2.4516677814359434e-06,
+      "loss": 0.5554,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5210752862827905,
+      "grad_norm": 5.111222932344912,
+      "learning_rate": 2.4510102838489587e-06,
+      "loss": 0.5666,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5211565012588322,
+      "grad_norm": 2.6572969128420363,
+      "learning_rate": 2.4503527896518025e-06,
+      "loss": 0.5856,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5212377162348737,
+      "grad_norm": 3.467977660333967,
+      "learning_rate": 2.449695298889971e-06,
+      "loss": 0.5549,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5213189312109153,
+      "grad_norm": 3.9586980008450547,
+      "learning_rate": 2.449037811608959e-06,
+      "loss": 0.5114,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5214001461869568,
+      "grad_norm": 6.6024044152724795,
+      "learning_rate": 2.4483803278542594e-06,
+      "loss": 0.4881,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5214813611629985,
+      "grad_norm": 3.872306998345536,
+      "learning_rate": 2.447722847671369e-06,
+      "loss": 0.4957,
+      "step": 6421
+    },
+    {
+      "epoch": 0.52156257613904,
+      "grad_norm": 4.928256043695277,
+      "learning_rate": 2.4470653711057805e-06,
+      "loss": 0.5453,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5216437911150816,
+      "grad_norm": 5.00049623611133,
+      "learning_rate": 2.446407898202988e-06,
+      "loss": 0.4302,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5217250060911232,
+      "grad_norm": 6.210683915021455,
+      "learning_rate": 2.445750429008486e-06,
+      "loss": 0.4867,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5218062210671648,
+      "grad_norm": 4.733137071274775,
+      "learning_rate": 2.4450929635677667e-06,
+      "loss": 0.5342,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5218874360432064,
+      "grad_norm": 3.573213441447143,
+      "learning_rate": 2.4444355019263235e-06,
+      "loss": 0.5391,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5219686510192479,
+      "grad_norm": 3.9218556675383685,
+      "learning_rate": 2.44377804412965e-06,
+      "loss": 0.5641,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5220498659952896,
+      "grad_norm": 3.72842203066307,
+      "learning_rate": 2.443120590223238e-06,
+      "loss": 0.5727,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5221310809713311,
+      "grad_norm": 4.5179452250246355,
+      "learning_rate": 2.4424631402525797e-06,
+      "loss": 0.5027,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5222122959473727,
+      "grad_norm": 7.296740350963874,
+      "learning_rate": 2.4418056942631687e-06,
+      "loss": 0.4627,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5222935109234143,
+      "grad_norm": 4.733022729564796,
+      "learning_rate": 2.4411482523004946e-06,
+      "loss": 0.4634,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5223747258994559,
+      "grad_norm": 5.089595564180344,
+      "learning_rate": 2.4404908144100513e-06,
+      "loss": 0.3586,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5224559408754974,
+      "grad_norm": 5.358613043214558,
+      "learning_rate": 2.4398333806373274e-06,
+      "loss": 0.465,
+      "step": 6433
+    },
+    {
+      "epoch": 0.522537155851539,
+      "grad_norm": 5.524190371189003,
+      "learning_rate": 2.4391759510278153e-06,
+      "loss": 0.5472,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5226183708275806,
+      "grad_norm": 4.103822972000649,
+      "learning_rate": 2.438518525627006e-06,
+      "loss": 0.5572,
+      "step": 6435
+    },
+    {
+      "epoch": 0.5226995858036222,
+      "grad_norm": 5.959127088930798,
+      "learning_rate": 2.4378611044803887e-06,
+      "loss": 0.4209,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5227808007796638,
+      "grad_norm": 5.31529709254031,
+      "learning_rate": 2.437203687633455e-06,
+      "loss": 0.3897,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5228620157557053,
+      "grad_norm": 3.7128349527804656,
+      "learning_rate": 2.436546275131693e-06,
+      "loss": 0.4428,
+      "step": 6438
+    },
+    {
+      "epoch": 0.522943230731747,
+      "grad_norm": 3.784355733006889,
+      "learning_rate": 2.435888867020593e-06,
+      "loss": 0.575,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5230244457077885,
+      "grad_norm": 4.907099811863354,
+      "learning_rate": 2.435231463345645e-06,
+      "loss": 0.5123,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5231056606838301,
+      "grad_norm": 3.094941647382156,
+      "learning_rate": 2.4345740641523362e-06,
+      "loss": 0.6211,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5231868756598717,
+      "grad_norm": 5.697788871531617,
+      "learning_rate": 2.4339166694861553e-06,
+      "loss": 0.4124,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5232680906359133,
+      "grad_norm": 5.825537615998602,
+      "learning_rate": 2.433259279392592e-06,
+      "loss": 0.4934,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5233493056119548,
+      "grad_norm": 4.894893357625769,
+      "learning_rate": 2.432601893917133e-06,
+      "loss": 0.5467,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5234305205879964,
+      "grad_norm": 7.012722371867207,
+      "learning_rate": 2.431944513105266e-06,
+      "loss": 0.5173,
+      "step": 6445
+    },
+    {
+      "epoch": 0.523511735564038,
+      "grad_norm": 3.8914614811391144,
+      "learning_rate": 2.4312871370024794e-06,
+      "loss": 0.5101,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5235929505400796,
+      "grad_norm": 4.15063424291914,
+      "learning_rate": 2.4306297656542584e-06,
+      "loss": 0.4239,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5236741655161212,
+      "grad_norm": 26.114252256627633,
+      "learning_rate": 2.4299723991060904e-06,
+      "loss": 0.5355,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5237553804921627,
+      "grad_norm": 5.921632572418434,
+      "learning_rate": 2.4293150374034625e-06,
+      "loss": 0.3273,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5238365954682044,
+      "grad_norm": 5.16163949915224,
+      "learning_rate": 2.4286576805918604e-06,
+      "loss": 0.4013,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5239178104442459,
+      "grad_norm": 3.8705275409101643,
+      "learning_rate": 2.4280003287167684e-06,
+      "loss": 0.4692,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5239990254202875,
+      "grad_norm": 4.713434827452213,
+      "learning_rate": 2.427342981823672e-06,
+      "loss": 0.4827,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5240802403963291,
+      "grad_norm": 9.847453330969838,
+      "learning_rate": 2.426685639958058e-06,
+      "loss": 0.597,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5241614553723707,
+      "grad_norm": 14.084536911041138,
+      "learning_rate": 2.426028303165409e-06,
+      "loss": 0.7119,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5242426703484122,
+      "grad_norm": 4.162924811617758,
+      "learning_rate": 2.425370971491209e-06,
+      "loss": 0.5353,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5243238853244538,
+      "grad_norm": 5.310608964267923,
+      "learning_rate": 2.424713644980945e-06,
+      "loss": 0.4668,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5244051003004954,
+      "grad_norm": 3.485581874385866,
+      "learning_rate": 2.424056323680097e-06,
+      "loss": 0.5376,
+      "step": 6457
+    },
+    {
+      "epoch": 0.524486315276537,
+      "grad_norm": 4.466249110985916,
+      "learning_rate": 2.423399007634149e-06,
+      "loss": 0.5362,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5245675302525786,
+      "grad_norm": 15.507085752685619,
+      "learning_rate": 2.422741696888585e-06,
+      "loss": 0.4823,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5246487452286202,
+      "grad_norm": 3.42283227124507,
+      "learning_rate": 2.4220843914888865e-06,
+      "loss": 0.5139,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5247299602046618,
+      "grad_norm": 4.471153348445109,
+      "learning_rate": 2.4214270914805353e-06,
+      "loss": 0.6101,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5248111751807033,
+      "grad_norm": 4.937962838447406,
+      "learning_rate": 2.4207697969090145e-06,
+      "loss": 0.5725,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5248923901567449,
+      "grad_norm": 3.4301025114527355,
+      "learning_rate": 2.420112507819804e-06,
+      "loss": 0.6678,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5249736051327865,
+      "grad_norm": 9.810088340742176,
+      "learning_rate": 2.4194552242583845e-06,
+      "loss": 0.4788,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5250548201088281,
+      "grad_norm": 5.247524236071172,
+      "learning_rate": 2.4187979462702382e-06,
+      "loss": 0.588,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5251360350848696,
+      "grad_norm": 3.8501987558213133,
+      "learning_rate": 2.4181406739008443e-06,
+      "loss": 0.5239,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5252172500609112,
+      "grad_norm": 4.286565115822655,
+      "learning_rate": 2.417483407195682e-06,
+      "loss": 0.4764,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5252984650369528,
+      "grad_norm": 2.9196698091729347,
+      "learning_rate": 2.416826146200231e-06,
+      "loss": 0.5093,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5253796800129944,
+      "grad_norm": 4.6923054622476,
+      "learning_rate": 2.4161688909599715e-06,
+      "loss": 0.6884,
+      "step": 6469
+    },
+    {
+      "epoch": 0.525460894989036,
+      "grad_norm": 3.966646784959062,
+      "learning_rate": 2.4155116415203804e-06,
+      "loss": 0.5132,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5255421099650776,
+      "grad_norm": 3.6916733502842956,
+      "learning_rate": 2.4148543979269357e-06,
+      "loss": 0.5949,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5256233249411192,
+      "grad_norm": 4.986568652074796,
+      "learning_rate": 2.4141971602251176e-06,
+      "loss": 0.5262,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5257045399171607,
+      "grad_norm": 7.761776773353264,
+      "learning_rate": 2.4135399284604012e-06,
+      "loss": 0.3962,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5257857548932023,
+      "grad_norm": 28.513128146924068,
+      "learning_rate": 2.4128827026782633e-06,
+      "loss": 0.655,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5258669698692439,
+      "grad_norm": 5.030638393005198,
+      "learning_rate": 2.4122254829241827e-06,
+      "loss": 0.6157,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5259481848452855,
+      "grad_norm": 3.759316711837708,
+      "learning_rate": 2.4115682692436337e-06,
+      "loss": 0.4513,
+      "step": 6476
+    },
+    {
+      "epoch": 0.526029399821327,
+      "grad_norm": 5.1999339488745395,
+      "learning_rate": 2.4109110616820918e-06,
+      "loss": 0.427,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5261106147973686,
+      "grad_norm": 9.081659712327916,
+      "learning_rate": 2.4102538602850337e-06,
+      "loss": 0.4293,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5261918297734102,
+      "grad_norm": 3.9469459629155574,
+      "learning_rate": 2.4095966650979342e-06,
+      "loss": 0.5247,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5262730447494518,
+      "grad_norm": 4.830126815254179,
+      "learning_rate": 2.4089394761662653e-06,
+      "loss": 0.5248,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5263542597254934,
+      "grad_norm": 5.427859891508367,
+      "learning_rate": 2.4082822935355035e-06,
+      "loss": 0.4429,
+      "step": 6481
+    },
+    {
+      "epoch": 0.526435474701535,
+      "grad_norm": 5.032571069469282,
+      "learning_rate": 2.4076251172511224e-06,
+      "loss": 0.4578,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5265166896775766,
+      "grad_norm": 3.8795340568732364,
+      "learning_rate": 2.4069679473585925e-06,
+      "loss": 0.5519,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5265979046536181,
+      "grad_norm": 3.9342593469573877,
+      "learning_rate": 2.4063107839033894e-06,
+      "loss": 0.5532,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5266791196296597,
+      "grad_norm": 6.08364360680254,
+      "learning_rate": 2.4056536269309847e-06,
+      "loss": 0.5749,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5267603346057013,
+      "grad_norm": 6.140920876184235,
+      "learning_rate": 2.4049964764868493e-06,
+      "loss": 0.3915,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5268415495817429,
+      "grad_norm": 4.532054993744361,
+      "learning_rate": 2.4043393326164536e-06,
+      "loss": 0.4485,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5269227645577844,
+      "grad_norm": 4.418973334061362,
+      "learning_rate": 2.403682195365272e-06,
+      "loss": 0.4327,
+      "step": 6488
+    },
+    {
+      "epoch": 0.527003979533826,
+      "grad_norm": 4.167489894614658,
+      "learning_rate": 2.4030250647787714e-06,
+      "loss": 0.545,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5270851945098676,
+      "grad_norm": 5.917681962851331,
+      "learning_rate": 2.402367940902423e-06,
+      "loss": 0.5363,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5271664094859092,
+      "grad_norm": 10.913996246318028,
+      "learning_rate": 2.401710823781697e-06,
+      "loss": 0.3902,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5272476244619508,
+      "grad_norm": 25.430239522056482,
+      "learning_rate": 2.4010537134620614e-06,
+      "loss": 0.4407,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5273288394379924,
+      "grad_norm": 9.226096785828776,
+      "learning_rate": 2.400396609988985e-06,
+      "loss": 0.5414,
+      "step": 6493
+    },
+    {
+      "epoch": 0.527410054414034,
+      "grad_norm": 4.553011519720875,
+      "learning_rate": 2.3997395134079367e-06,
+      "loss": 0.5125,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5274912693900755,
+      "grad_norm": 3.85392516623142,
+      "learning_rate": 2.399082423764383e-06,
+      "loss": 0.5804,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5275724843661171,
+      "grad_norm": 6.43430352311751,
+      "learning_rate": 2.3984253411037913e-06,
+      "loss": 0.4332,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5276536993421587,
+      "grad_norm": 5.348598603512892,
+      "learning_rate": 2.397768265471629e-06,
+      "loss": 0.4464,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5277349143182003,
+      "grad_norm": 5.503883803363463,
+      "learning_rate": 2.397111196913362e-06,
+      "loss": 0.5333,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5278161292942418,
+      "grad_norm": 5.084027665154841,
+      "learning_rate": 2.396454135474454e-06,
+      "loss": 0.5963,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5278973442702835,
+      "grad_norm": 4.104843883398135,
+      "learning_rate": 2.3957970812003727e-06,
+      "loss": 0.6582,
+      "step": 6500
+    },
+    {
+      "epoch": 0.527978559246325,
+      "grad_norm": 6.334197801454462,
+      "learning_rate": 2.3951400341365827e-06,
+      "loss": 0.5699,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5280597742223666,
+      "grad_norm": 6.9437969081416435,
+      "learning_rate": 2.394482994328546e-06,
+      "loss": 0.4565,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5281409891984082,
+      "grad_norm": 3.941363956260876,
+      "learning_rate": 2.393825961821728e-06,
+      "loss": 0.705,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5282222041744498,
+      "grad_norm": 3.7273474058501974,
+      "learning_rate": 2.3931689366615926e-06,
+      "loss": 0.488,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5283034191504914,
+      "grad_norm": 5.084976289181063,
+      "learning_rate": 2.392511918893601e-06,
+      "loss": 0.5142,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5283846341265329,
+      "grad_norm": 6.4658935525713295,
+      "learning_rate": 2.3918549085632145e-06,
+      "loss": 0.4072,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5284658491025745,
+      "grad_norm": 3.651970957929354,
+      "learning_rate": 2.3911979057158974e-06,
+      "loss": 0.4404,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5285470640786161,
+      "grad_norm": 3.5446244109539045,
+      "learning_rate": 2.3905409103971096e-06,
+      "loss": 0.5651,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5286282790546577,
+      "grad_norm": 3.9057802685762217,
+      "learning_rate": 2.38988392265231e-06,
+      "loss": 0.4663,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5287094940306992,
+      "grad_norm": 2.871351243716771,
+      "learning_rate": 2.389226942526961e-06,
+      "loss": 0.5126,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5287907090067409,
+      "grad_norm": 6.264655385586473,
+      "learning_rate": 2.3885699700665217e-06,
+      "loss": 0.463,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5288719239827824,
+      "grad_norm": 6.176994725304146,
+      "learning_rate": 2.3879130053164495e-06,
+      "loss": 0.3438,
+      "step": 6512
+    },
+    {
+      "epoch": 0.528953138958824,
+      "grad_norm": 3.841956377687345,
+      "learning_rate": 2.3872560483222048e-06,
+      "loss": 0.6078,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5290343539348656,
+      "grad_norm": 3.946093406576217,
+      "learning_rate": 2.3865990991292458e-06,
+      "loss": 0.49,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5291155689109072,
+      "grad_norm": 5.815463605481333,
+      "learning_rate": 2.3859421577830276e-06,
+      "loss": 0.5979,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5291967838869488,
+      "grad_norm": 7.2849496983935875,
+      "learning_rate": 2.385285224329009e-06,
+      "loss": 0.4522,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5292779988629903,
+      "grad_norm": 6.703027017567743,
+      "learning_rate": 2.384628298812646e-06,
+      "loss": 0.4208,
+      "step": 6517
+    },
+    {
+      "epoch": 0.529359213839032,
+      "grad_norm": 4.068730154097392,
+      "learning_rate": 2.383971381279393e-06,
+      "loss": 0.649,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5294404288150735,
+      "grad_norm": 3.9699040583000667,
+      "learning_rate": 2.383314471774707e-06,
+      "loss": 0.7164,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5295216437911151,
+      "grad_norm": 11.734590811153446,
+      "learning_rate": 2.382657570344043e-06,
+      "loss": 0.5432,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5296028587671566,
+      "grad_norm": 5.270660282246951,
+      "learning_rate": 2.382000677032854e-06,
+      "loss": 0.5122,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5296840737431983,
+      "grad_norm": 4.145643825845797,
+      "learning_rate": 2.3813437918865925e-06,
+      "loss": 0.5513,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5297652887192398,
+      "grad_norm": 5.695234292809091,
+      "learning_rate": 2.380686914950713e-06,
+      "loss": 0.5945,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5298465036952814,
+      "grad_norm": 5.187560295102548,
+      "learning_rate": 2.380030046270668e-06,
+      "loss": 0.5533,
+      "step": 6524
+    },
+    {
+      "epoch": 0.529927718671323,
+      "grad_norm": 5.2980757209337215,
+      "learning_rate": 2.379373185891908e-06,
+      "loss": 0.5801,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5300089336473646,
+      "grad_norm": 4.379144750425471,
+      "learning_rate": 2.3787163338598854e-06,
+      "loss": 0.6238,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5300901486234062,
+      "grad_norm": 5.906641818249719,
+      "learning_rate": 2.3780594902200515e-06,
+      "loss": 0.5211,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5301713635994477,
+      "grad_norm": 5.654605677438833,
+      "learning_rate": 2.377402655017854e-06,
+      "loss": 0.4426,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5302525785754894,
+      "grad_norm": 9.600316549753119,
+      "learning_rate": 2.376745828298745e-06,
+      "loss": 0.4189,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5303337935515309,
+      "grad_norm": 4.652507037908199,
+      "learning_rate": 2.376089010108172e-06,
+      "loss": 0.5635,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5304150085275725,
+      "grad_norm": 6.872480563300781,
+      "learning_rate": 2.3754322004915837e-06,
+      "loss": 0.452,
+      "step": 6531
+    },
+    {
+      "epoch": 0.530496223503614,
+      "grad_norm": 6.133795123377434,
+      "learning_rate": 2.3747753994944283e-06,
+      "loss": 0.5212,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5305774384796557,
+      "grad_norm": 5.688402297871484,
+      "learning_rate": 2.3741186071621523e-06,
+      "loss": 0.5803,
+      "step": 6533
+    },
+    {
+      "epoch": 0.5306586534556972,
+      "grad_norm": 4.147307899835006,
+      "learning_rate": 2.373461823540202e-06,
+      "loss": 0.4938,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5307398684317388,
+      "grad_norm": 4.16411315568013,
+      "learning_rate": 2.3728050486740244e-06,
+      "loss": 0.4921,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5308210834077804,
+      "grad_norm": 4.86479630808414,
+      "learning_rate": 2.3721482826090643e-06,
+      "loss": 0.5561,
+      "step": 6536
+    },
+    {
+      "epoch": 0.530902298383822,
+      "grad_norm": 4.549881463792877,
+      "learning_rate": 2.3714915253907657e-06,
+      "loss": 0.4727,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5309835133598636,
+      "grad_norm": 3.1960496743376745,
+      "learning_rate": 2.370834777064574e-06,
+      "loss": 0.5856,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5310647283359051,
+      "grad_norm": 6.183831960146142,
+      "learning_rate": 2.3701780376759323e-06,
+      "loss": 0.4186,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5311459433119468,
+      "grad_norm": 6.1626582028588075,
+      "learning_rate": 2.3695213072702834e-06,
+      "loss": 0.4421,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5312271582879883,
+      "grad_norm": 4.399409989452963,
+      "learning_rate": 2.368864585893069e-06,
+      "loss": 0.5147,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5313083732640299,
+      "grad_norm": 4.73410977261382,
+      "learning_rate": 2.368207873589731e-06,
+      "loss": 0.56,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5313895882400714,
+      "grad_norm": 4.575239688507527,
+      "learning_rate": 2.3675511704057115e-06,
+      "loss": 0.5333,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5314708032161131,
+      "grad_norm": 7.035858437872166,
+      "learning_rate": 2.3668944763864486e-06,
+      "loss": 0.5508,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5315520181921546,
+      "grad_norm": 5.241353250145363,
+      "learning_rate": 2.3662377915773845e-06,
+      "loss": 0.4836,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5316332331681962,
+      "grad_norm": 7.382490118024742,
+      "learning_rate": 2.365581116023958e-06,
+      "loss": 0.4395,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5317144481442379,
+      "grad_norm": 5.3401428892737455,
+      "learning_rate": 2.364924449771605e-06,
+      "loss": 0.427,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5317956631202794,
+      "grad_norm": 3.7061230832301066,
+      "learning_rate": 2.364267792865767e-06,
+      "loss": 0.8927,
+      "step": 6548
+    },
+    {
+      "epoch": 0.531876878096321,
+      "grad_norm": 4.420036708733726,
+      "learning_rate": 2.363611145351879e-06,
+      "loss": 0.5167,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5319580930723625,
+      "grad_norm": 4.783438045653566,
+      "learning_rate": 2.3629545072753767e-06,
+      "loss": 0.5279,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5320393080484042,
+      "grad_norm": 4.749522841051883,
+      "learning_rate": 2.3622978786816984e-06,
+      "loss": 0.4024,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5321205230244457,
+      "grad_norm": 7.7640179721652,
+      "learning_rate": 2.361641259616278e-06,
+      "loss": 0.5393,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5322017380004873,
+      "grad_norm": 3.6547210717885865,
+      "learning_rate": 2.3609846501245494e-06,
+      "loss": 0.4325,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5322829529765288,
+      "grad_norm": 7.3797843232620455,
+      "learning_rate": 2.3603280502519482e-06,
+      "loss": 0.5021,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5323641679525705,
+      "grad_norm": 3.694146153232864,
+      "learning_rate": 2.3596714600439062e-06,
+      "loss": 0.5116,
+      "step": 6555
+    },
+    {
+      "epoch": 0.532445382928612,
+      "grad_norm": 5.0985701151843354,
+      "learning_rate": 2.3590148795458577e-06,
+      "loss": 0.4945,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5325265979046536,
+      "grad_norm": 5.131669229194393,
+      "learning_rate": 2.3583583088032313e-06,
+      "loss": 0.5813,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5326078128806953,
+      "grad_norm": 4.005833213164181,
+      "learning_rate": 2.3577017478614613e-06,
+      "loss": 0.5053,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5326890278567368,
+      "grad_norm": 14.883116766398912,
+      "learning_rate": 2.357045196765978e-06,
+      "loss": 0.6758,
+      "step": 6559
+    },
+    {
+      "epoch": 0.5327702428327784,
+      "grad_norm": 4.9450151525987955,
+      "learning_rate": 2.3563886555622093e-06,
+      "loss": 0.561,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5328514578088199,
+      "grad_norm": 9.194980096203347,
+      "learning_rate": 2.355732124295586e-06,
+      "loss": 0.5195,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5329326727848616,
+      "grad_norm": 5.200787441007623,
+      "learning_rate": 2.3550756030115364e-06,
+      "loss": 0.6133,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5330138877609031,
+      "grad_norm": 4.874445965778859,
+      "learning_rate": 2.3544190917554875e-06,
+      "loss": 0.456,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5330951027369447,
+      "grad_norm": 4.734097340063761,
+      "learning_rate": 2.3537625905728677e-06,
+      "loss": 0.5303,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5331763177129862,
+      "grad_norm": 5.317506693611895,
+      "learning_rate": 2.3531060995091026e-06,
+      "loss": 0.6195,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5332575326890279,
+      "grad_norm": 5.085875518247843,
+      "learning_rate": 2.352449618609617e-06,
+      "loss": 0.6773,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5333387476650694,
+      "grad_norm": 5.216473307627287,
+      "learning_rate": 2.3517931479198383e-06,
+      "loss": 0.4075,
+      "step": 6567
+    },
+    {
+      "epoch": 0.533419962641111,
+      "grad_norm": 4.73626486723172,
+      "learning_rate": 2.3511366874851885e-06,
+      "loss": 0.5928,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5335011776171527,
+      "grad_norm": 4.068709628111098,
+      "learning_rate": 2.350480237351092e-06,
+      "loss": 0.5746,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5335823925931942,
+      "grad_norm": 7.162771272344813,
+      "learning_rate": 2.3498237975629726e-06,
+      "loss": 0.6726,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5336636075692358,
+      "grad_norm": 8.523429545622452,
+      "learning_rate": 2.349167368166251e-06,
+      "loss": 0.4304,
+      "step": 6571
+    },
+    {
+      "epoch": 0.5337448225452773,
+      "grad_norm": 5.274081883984625,
+      "learning_rate": 2.348510949206349e-06,
+      "loss": 0.5222,
+      "step": 6572
+    },
+    {
+      "epoch": 0.533826037521319,
+      "grad_norm": 3.8799868155561406,
+      "learning_rate": 2.3478545407286883e-06,
+      "loss": 0.6353,
+      "step": 6573
+    },
+    {
+      "epoch": 0.5339072524973605,
+      "grad_norm": 3.6165853886009844,
+      "learning_rate": 2.3471981427786875e-06,
+      "loss": 0.5822,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5339884674734021,
+      "grad_norm": 6.353757597164967,
+      "learning_rate": 2.3465417554017675e-06,
+      "loss": 0.4707,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5340696824494436,
+      "grad_norm": 5.10414117668738,
+      "learning_rate": 2.3458853786433444e-06,
+      "loss": 0.5357,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5341508974254853,
+      "grad_norm": 3.3925342392484317,
+      "learning_rate": 2.345229012548838e-06,
+      "loss": 0.6855,
+      "step": 6577
+    },
+    {
+      "epoch": 0.5342321124015268,
+      "grad_norm": 3.3042399836502776,
+      "learning_rate": 2.3445726571636656e-06,
+      "loss": 0.4581,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5343133273775684,
+      "grad_norm": 4.792803765102656,
+      "learning_rate": 2.3439163125332415e-06,
+      "loss": 0.4442,
+      "step": 6579
+    },
+    {
+      "epoch": 0.5343945423536101,
+      "grad_norm": 53.09136619996998,
+      "learning_rate": 2.343259978702984e-06,
+      "loss": 0.5999,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5344757573296516,
+      "grad_norm": 4.366575308474977,
+      "learning_rate": 2.3426036557183056e-06,
+      "loss": 0.55,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5345569723056932,
+      "grad_norm": 4.219747570498485,
+      "learning_rate": 2.3419473436246206e-06,
+      "loss": 0.6209,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5346381872817347,
+      "grad_norm": 5.505458655811916,
+      "learning_rate": 2.341291042467344e-06,
+      "loss": 0.5682,
+      "step": 6583
+    },
+    {
+      "epoch": 0.5347194022577764,
+      "grad_norm": 5.438215231835009,
+      "learning_rate": 2.3406347522918866e-06,
+      "loss": 0.4587,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5348006172338179,
+      "grad_norm": 9.221582577454688,
+      "learning_rate": 2.339978473143661e-06,
+      "loss": 0.6507,
+      "step": 6585
+    },
+    {
+      "epoch": 0.5348818322098595,
+      "grad_norm": 6.264732038581897,
+      "learning_rate": 2.3393222050680788e-06,
+      "loss": 0.5201,
+      "step": 6586
+    },
+    {
+      "epoch": 0.534963047185901,
+      "grad_norm": 5.801333658841605,
+      "learning_rate": 2.338665948110549e-06,
+      "loss": 0.5507,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5350442621619427,
+      "grad_norm": 6.810336490229261,
+      "learning_rate": 2.3380097023164813e-06,
+      "loss": 0.67,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5351254771379842,
+      "grad_norm": 5.781581076596003,
+      "learning_rate": 2.337353467731286e-06,
+      "loss": 0.391,
+      "step": 6589
+    },
+    {
+      "epoch": 0.5352066921140258,
+      "grad_norm": 4.370970902980655,
+      "learning_rate": 2.3366972444003698e-06,
+      "loss": 0.5351,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5352879070900675,
+      "grad_norm": 12.469579377782821,
+      "learning_rate": 2.3360410323691386e-06,
+      "loss": 0.4929,
+      "step": 6591
+    },
+    {
+      "epoch": 0.535369122066109,
+      "grad_norm": 10.008332152388622,
+      "learning_rate": 2.335384831683002e-06,
+      "loss": 0.4752,
+      "step": 6592
+    },
+    {
+      "epoch": 0.5354503370421506,
+      "grad_norm": 9.797876571328075,
+      "learning_rate": 2.334728642387363e-06,
+      "loss": 0.4668,
+      "step": 6593
+    },
+    {
+      "epoch": 0.5355315520181921,
+      "grad_norm": 5.754796284833849,
+      "learning_rate": 2.334072464527628e-06,
+      "loss": 0.3572,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5356127669942338,
+      "grad_norm": 4.316813249696164,
+      "learning_rate": 2.333416298149199e-06,
+      "loss": 0.4839,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5356939819702753,
+      "grad_norm": 4.974375179617703,
+      "learning_rate": 2.3327601432974817e-06,
+      "loss": 0.3869,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5357751969463169,
+      "grad_norm": 3.985538587500043,
+      "learning_rate": 2.332104000017877e-06,
+      "loss": 0.4175,
+      "step": 6597
+    },
+    {
+      "epoch": 0.5358564119223584,
+      "grad_norm": 3.2055836403047064,
+      "learning_rate": 2.3314478683557863e-06,
+      "loss": 0.5612,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5359376268984001,
+      "grad_norm": 5.214849251275779,
+      "learning_rate": 2.330791748356612e-06,
+      "loss": 0.6605,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5360188418744416,
+      "grad_norm": 5.888955286356365,
+      "learning_rate": 2.3301356400657527e-06,
+      "loss": 0.4432,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5361000568504832,
+      "grad_norm": 4.505062527177114,
+      "learning_rate": 2.3294795435286073e-06,
+      "loss": 0.6324,
+      "step": 6601
+    },
+    {
+      "epoch": 0.5361812718265249,
+      "grad_norm": 79.62251681114921,
+      "learning_rate": 2.3288234587905767e-06,
+      "loss": 0.4921,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5362624868025664,
+      "grad_norm": 7.158021669325846,
+      "learning_rate": 2.328167385897056e-06,
+      "loss": 0.4798,
+      "step": 6603
+    },
+    {
+      "epoch": 0.536343701778608,
+      "grad_norm": 5.336534392507409,
+      "learning_rate": 2.327511324893442e-06,
+      "loss": 0.6895,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5364249167546495,
+      "grad_norm": 5.969891236348079,
+      "learning_rate": 2.3268552758251327e-06,
+      "loss": 0.5348,
+      "step": 6605
+    },
+    {
+      "epoch": 0.5365061317306912,
+      "grad_norm": 5.339767902188447,
+      "learning_rate": 2.3261992387375216e-06,
+      "loss": 0.4408,
+      "step": 6606
+    },
+    {
+      "epoch": 0.5365873467067327,
+      "grad_norm": 6.745067120318976,
+      "learning_rate": 2.3255432136760026e-06,
+      "loss": 0.5484,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5366685616827743,
+      "grad_norm": 6.071978645783486,
+      "learning_rate": 2.324887200685971e-06,
+      "loss": 0.5867,
+      "step": 6608
+    },
+    {
+      "epoch": 0.5367497766588158,
+      "grad_norm": 6.2394195972445665,
+      "learning_rate": 2.3242311998128182e-06,
+      "loss": 0.5295,
+      "step": 6609
+    },
+    {
+      "epoch": 0.5368309916348575,
+      "grad_norm": 5.667495576556501,
+      "learning_rate": 2.3235752111019362e-06,
+      "loss": 0.4169,
+      "step": 6610
+    },
+    {
+      "epoch": 0.536912206610899,
+      "grad_norm": 7.823798452635792,
+      "learning_rate": 2.3229192345987146e-06,
+      "loss": 0.6219,
+      "step": 6611
+    },
+    {
+      "epoch": 0.5369934215869406,
+      "grad_norm": 6.983866314486278,
+      "learning_rate": 2.322263270348546e-06,
+      "loss": 0.5665,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5370746365629823,
+      "grad_norm": 6.526144378674829,
+      "learning_rate": 2.3216073183968184e-06,
+      "loss": 0.3971,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5371558515390238,
+      "grad_norm": 10.243690491824887,
+      "learning_rate": 2.320951378788919e-06,
+      "loss": 0.6378,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5372370665150654,
+      "grad_norm": 4.954484713664851,
+      "learning_rate": 2.3202954515702384e-06,
+      "loss": 0.548,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5373182814911069,
+      "grad_norm": 5.184321630416869,
+      "learning_rate": 2.3196395367861605e-06,
+      "loss": 0.5172,
+      "step": 6616
+    },
+    {
+      "epoch": 0.5373994964671486,
+      "grad_norm": 5.640278703192089,
+      "learning_rate": 2.3189836344820717e-06,
+      "loss": 0.4252,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5374807114431901,
+      "grad_norm": 6.756103044876508,
+      "learning_rate": 2.318327744703358e-06,
+      "loss": 0.5939,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5375619264192317,
+      "grad_norm": 4.587869437489592,
+      "learning_rate": 2.317671867495403e-06,
+      "loss": 0.5293,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5376431413952732,
+      "grad_norm": 5.744181498371175,
+      "learning_rate": 2.317016002903589e-06,
+      "loss": 0.4909,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5377243563713149,
+      "grad_norm": 5.53015069217161,
+      "learning_rate": 2.3163601509733e-06,
+      "loss": 0.52,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5378055713473564,
+      "grad_norm": 6.7221739959300555,
+      "learning_rate": 2.3157043117499174e-06,
+      "loss": 0.4972,
+      "step": 6622
+    },
+    {
+      "epoch": 0.537886786323398,
+      "grad_norm": 4.038968186465272,
+      "learning_rate": 2.3150484852788186e-06,
+      "loss": 0.4686,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5379680012994397,
+      "grad_norm": 5.334775479299188,
+      "learning_rate": 2.3143926716053876e-06,
+      "loss": 0.4335,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5380492162754812,
+      "grad_norm": 8.59183543983034,
+      "learning_rate": 2.3137368707750018e-06,
+      "loss": 0.6309,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5381304312515228,
+      "grad_norm": 4.578718128339019,
+      "learning_rate": 2.3130810828330375e-06,
+      "loss": 0.6889,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5382116462275643,
+      "grad_norm": 7.819423595541969,
+      "learning_rate": 2.3124253078248734e-06,
+      "loss": 0.43,
+      "step": 6627
+    },
+    {
+      "epoch": 0.538292861203606,
+      "grad_norm": 3.229990826836947,
+      "learning_rate": 2.3117695457958857e-06,
+      "loss": 0.7549,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5383740761796475,
+      "grad_norm": 4.562845996906024,
+      "learning_rate": 2.3111137967914492e-06,
+      "loss": 0.434,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5384552911556891,
+      "grad_norm": 5.915615501905179,
+      "learning_rate": 2.310458060856937e-06,
+      "loss": 0.3945,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5385365061317307,
+      "grad_norm": 10.277625736537166,
+      "learning_rate": 2.3098023380377257e-06,
+      "loss": 0.542,
+      "step": 6631
+    },
+    {
+      "epoch": 0.5386177211077723,
+      "grad_norm": 3.9982941253903843,
+      "learning_rate": 2.309146628379185e-06,
+      "loss": 0.4956,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5386989360838138,
+      "grad_norm": 5.587736755912178,
+      "learning_rate": 2.308490931926687e-06,
+      "loss": 0.3932,
+      "step": 6633
+    },
+    {
+      "epoch": 0.5387801510598554,
+      "grad_norm": 6.249822940854045,
+      "learning_rate": 2.3078352487256045e-06,
+      "loss": 0.551,
+      "step": 6634
+    },
+    {
+      "epoch": 0.5388613660358971,
+      "grad_norm": 4.330260149960894,
+      "learning_rate": 2.3071795788213047e-06,
+      "loss": 0.419,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5389425810119386,
+      "grad_norm": 3.9213426854190554,
+      "learning_rate": 2.3065239222591574e-06,
+      "loss": 0.912,
+      "step": 6636
+    },
+    {
+      "epoch": 0.5390237959879802,
+      "grad_norm": 5.204419720367973,
+      "learning_rate": 2.3058682790845314e-06,
+      "loss": 0.582,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5391050109640217,
+      "grad_norm": 6.739029726775385,
+      "learning_rate": 2.3052126493427934e-06,
+      "loss": 0.6209,
+      "step": 6638
+    },
+    {
+      "epoch": 0.5391862259400634,
+      "grad_norm": 4.96555228670431,
+      "learning_rate": 2.304557033079308e-06,
+      "loss": 0.431,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5392674409161049,
+      "grad_norm": 6.401695121326637,
+      "learning_rate": 2.303901430339442e-06,
+      "loss": 0.3933,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5393486558921465,
+      "grad_norm": 6.539736744319487,
+      "learning_rate": 2.30324584116856e-06,
+      "loss": 0.5304,
+      "step": 6641
+    },
+    {
+      "epoch": 0.539429870868188,
+      "grad_norm": 4.494895959996267,
+      "learning_rate": 2.302590265612023e-06,
+      "loss": 0.4851,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5395110858442297,
+      "grad_norm": 5.068900846135241,
+      "learning_rate": 2.301934703715196e-06,
+      "loss": 0.542,
+      "step": 6643
+    },
+    {
+      "epoch": 0.5395923008202712,
+      "grad_norm": 3.670185354744509,
+      "learning_rate": 2.301279155523439e-06,
+      "loss": 0.4716,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5396735157963128,
+      "grad_norm": 3.9777847287273436,
+      "learning_rate": 2.3006236210821127e-06,
+      "loss": 0.4095,
+      "step": 6645
+    },
+    {
+      "epoch": 0.5397547307723545,
+      "grad_norm": 8.621397414880517,
+      "learning_rate": 2.2999681004365755e-06,
+      "loss": 0.4291,
+      "step": 6646
+    },
+    {
+      "epoch": 0.539835945748396,
+      "grad_norm": 4.348209634682363,
+      "learning_rate": 2.299312593632189e-06,
+      "loss": 0.5603,
+      "step": 6647
+    },
+    {
+      "epoch": 0.5399171607244376,
+      "grad_norm": 5.453003219785878,
+      "learning_rate": 2.298657100714308e-06,
+      "loss": 0.4495,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5399983757004791,
+      "grad_norm": 6.222935801462418,
+      "learning_rate": 2.2980016217282892e-06,
+      "loss": 0.5602,
+      "step": 6649
+    },
+    {
+      "epoch": 0.5400795906765208,
+      "grad_norm": 6.456779266203264,
+      "learning_rate": 2.2973461567194903e-06,
+      "loss": 0.6055,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5401608056525623,
+      "grad_norm": 2.812177928800366,
+      "learning_rate": 2.296690705733265e-06,
+      "loss": 0.5931,
+      "step": 6651
+    },
+    {
+      "epoch": 0.5402420206286039,
+      "grad_norm": 5.628644054902416,
+      "learning_rate": 2.2960352688149657e-06,
+      "loss": 0.5323,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5403232356046455,
+      "grad_norm": 6.158902153206404,
+      "learning_rate": 2.295379846009947e-06,
+      "loss": 0.6064,
+      "step": 6653
+    },
+    {
+      "epoch": 0.5404044505806871,
+      "grad_norm": 4.596838238746014,
+      "learning_rate": 2.2947244373635608e-06,
+      "loss": 0.5383,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5404856655567286,
+      "grad_norm": 10.141289999607887,
+      "learning_rate": 2.294069042921156e-06,
+      "loss": 0.4802,
+      "step": 6655
+    },
+    {
+      "epoch": 0.5405668805327702,
+      "grad_norm": 3.56367135718309,
+      "learning_rate": 2.2934136627280834e-06,
+      "loss": 0.5922,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5406480955088119,
+      "grad_norm": 4.957818406371747,
+      "learning_rate": 2.292758296829693e-06,
+      "loss": 0.7018,
+      "step": 6657
+    },
+    {
+      "epoch": 0.5407293104848534,
+      "grad_norm": 4.552129764687744,
+      "learning_rate": 2.2921029452713305e-06,
+      "loss": 0.5328,
+      "step": 6658
+    },
+    {
+      "epoch": 0.540810525460895,
+      "grad_norm": 5.962108877745079,
+      "learning_rate": 2.291447608098345e-06,
+      "loss": 0.5459,
+      "step": 6659
+    },
+    {
+      "epoch": 0.5408917404369366,
+      "grad_norm": 3.831361014930282,
+      "learning_rate": 2.290792285356081e-06,
+      "loss": 0.445,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5409729554129782,
+      "grad_norm": 4.039508905590606,
+      "learning_rate": 2.290136977089883e-06,
+      "loss": 0.6447,
+      "step": 6661
+    },
+    {
+      "epoch": 0.5410541703890197,
+      "grad_norm": 4.701839781470267,
+      "learning_rate": 2.289481683345096e-06,
+      "loss": 0.4856,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5411353853650613,
+      "grad_norm": 5.28893125359557,
+      "learning_rate": 2.2888264041670625e-06,
+      "loss": 0.3886,
+      "step": 6663
+    },
+    {
+      "epoch": 0.5412166003411029,
+      "grad_norm": 4.226127643770201,
+      "learning_rate": 2.288171139601124e-06,
+      "loss": 0.451,
+      "step": 6664
+    },
+    {
+      "epoch": 0.5412978153171445,
+      "grad_norm": 4.150106558822198,
+      "learning_rate": 2.287515889692621e-06,
+      "loss": 0.585,
+      "step": 6665
+    },
+    {
+      "epoch": 0.541379030293186,
+      "grad_norm": 7.149187546807345,
+      "learning_rate": 2.2868606544868947e-06,
+      "loss": 0.5057,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5414602452692276,
+      "grad_norm": 4.941704759283249,
+      "learning_rate": 2.2862054340292835e-06,
+      "loss": 0.5628,
+      "step": 6667
+    },
+    {
+      "epoch": 0.5415414602452693,
+      "grad_norm": 6.843536039362226,
+      "learning_rate": 2.2855502283651238e-06,
+      "loss": 0.5538,
+      "step": 6668
+    },
+    {
+      "epoch": 0.5416226752213108,
+      "grad_norm": 6.747892070688352,
+      "learning_rate": 2.284895037539753e-06,
+      "loss": 0.5538,
+      "step": 6669
+    },
+    {
+      "epoch": 0.5417038901973524,
+      "grad_norm": 4.985721031141658,
+      "learning_rate": 2.2842398615985086e-06,
+      "loss": 0.5889,
+      "step": 6670
+    },
+    {
+      "epoch": 0.541785105173394,
+      "grad_norm": 3.9696526712453686,
+      "learning_rate": 2.283584700586723e-06,
+      "loss": 0.4235,
+      "step": 6671
+    },
+    {
+      "epoch": 0.5418663201494356,
+      "grad_norm": 4.370310826153604,
+      "learning_rate": 2.2829295545497304e-06,
+      "loss": 0.565,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5419475351254771,
+      "grad_norm": 7.303076408805493,
+      "learning_rate": 2.282274423532865e-06,
+      "loss": 0.4592,
+      "step": 6673
+    },
+    {
+      "epoch": 0.5420287501015187,
+      "grad_norm": 5.82847333716156,
+      "learning_rate": 2.2816193075814557e-06,
+      "loss": 0.519,
+      "step": 6674
+    },
+    {
+      "epoch": 0.5421099650775603,
+      "grad_norm": 3.1507733594386704,
+      "learning_rate": 2.280964206740835e-06,
+      "loss": 0.5885,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5421911800536019,
+      "grad_norm": 5.983802017759874,
+      "learning_rate": 2.280309121056333e-06,
+      "loss": 0.4888,
+      "step": 6676
+    },
+    {
+      "epoch": 0.5422723950296434,
+      "grad_norm": 3.19378590139564,
+      "learning_rate": 2.279654050573276e-06,
+      "loss": 0.4794,
+      "step": 6677
+    },
+    {
+      "epoch": 0.542353610005685,
+      "grad_norm": 4.270799718455597,
+      "learning_rate": 2.2789989953369924e-06,
+      "loss": 0.5408,
+      "step": 6678
+    },
+    {
+      "epoch": 0.5424348249817267,
+      "grad_norm": 5.277701316589272,
+      "learning_rate": 2.27834395539281e-06,
+      "loss": 0.4466,
+      "step": 6679
+    },
+    {
+      "epoch": 0.5425160399577682,
+      "grad_norm": 5.962388380985894,
+      "learning_rate": 2.2776889307860513e-06,
+      "loss": 0.4871,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5425972549338098,
+      "grad_norm": 4.937531617716085,
+      "learning_rate": 2.2770339215620433e-06,
+      "loss": 0.4022,
+      "step": 6681
+    },
+    {
+      "epoch": 0.5426784699098514,
+      "grad_norm": 4.989377028856358,
+      "learning_rate": 2.2763789277661077e-06,
+      "loss": 0.4676,
+      "step": 6682
+    },
+    {
+      "epoch": 0.542759684885893,
+      "grad_norm": 3.814875085144808,
+      "learning_rate": 2.2757239494435666e-06,
+      "loss": 0.4234,
+      "step": 6683
+    },
+    {
+      "epoch": 0.5428408998619345,
+      "grad_norm": 6.832068792784659,
+      "learning_rate": 2.2750689866397407e-06,
+      "loss": 0.6319,
+      "step": 6684
+    },
+    {
+      "epoch": 0.5429221148379761,
+      "grad_norm": 26.304115965543602,
+      "learning_rate": 2.2744140393999507e-06,
+      "loss": 0.3713,
+      "step": 6685
+    },
+    {
+      "epoch": 0.5430033298140177,
+      "grad_norm": 7.2769279119321135,
+      "learning_rate": 2.273759107769516e-06,
+      "loss": 0.6063,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5430845447900593,
+      "grad_norm": 4.819897180382021,
+      "learning_rate": 2.2731041917937524e-06,
+      "loss": 0.5373,
+      "step": 6687
+    },
+    {
+      "epoch": 0.5431657597661008,
+      "grad_norm": 5.898798059857416,
+      "learning_rate": 2.2724492915179787e-06,
+      "loss": 0.5572,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5432469747421425,
+      "grad_norm": 4.441345383933851,
+      "learning_rate": 2.27179440698751e-06,
+      "loss": 0.6711,
+      "step": 6689
+    },
+    {
+      "epoch": 0.5433281897181841,
+      "grad_norm": 3.8157128193761185,
+      "learning_rate": 2.2711395382476595e-06,
+      "loss": 0.6712,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5434094046942256,
+      "grad_norm": 4.388902512240195,
+      "learning_rate": 2.2704846853437424e-06,
+      "loss": 0.5042,
+      "step": 6691
+    },
+    {
+      "epoch": 0.5434906196702672,
+      "grad_norm": 7.873275069114776,
+      "learning_rate": 2.269829848321071e-06,
+      "loss": 0.4374,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5435718346463088,
+      "grad_norm": 5.506546269665942,
+      "learning_rate": 2.2691750272249545e-06,
+      "loss": 0.3936,
+      "step": 6693
+    },
+    {
+      "epoch": 0.5436530496223504,
+      "grad_norm": 4.315922170242298,
+      "learning_rate": 2.2685202221007057e-06,
+      "loss": 0.66,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5437342645983919,
+      "grad_norm": 41.128866529656484,
+      "learning_rate": 2.2678654329936322e-06,
+      "loss": 0.4566,
+      "step": 6695
+    },
+    {
+      "epoch": 0.5438154795744335,
+      "grad_norm": 4.547047318098447,
+      "learning_rate": 2.267210659949042e-06,
+      "loss": 0.5562,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5438966945504751,
+      "grad_norm": 5.996426610029105,
+      "learning_rate": 2.2665559030122424e-06,
+      "loss": 0.4198,
+      "step": 6697
+    },
+    {
+      "epoch": 0.5439779095265167,
+      "grad_norm": 4.366420114566318,
+      "learning_rate": 2.2659011622285383e-06,
+      "loss": 0.468,
+      "step": 6698
+    },
+    {
+      "epoch": 0.5440591245025582,
+      "grad_norm": 4.742561220461673,
+      "learning_rate": 2.265246437643236e-06,
+      "loss": 0.4206,
+      "step": 6699
+    },
+    {
+      "epoch": 0.5441403394785999,
+      "grad_norm": 5.981908620139316,
+      "learning_rate": 2.2645917293016363e-06,
+      "loss": 0.6577,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5442215544546415,
+      "grad_norm": 3.9854122807772723,
+      "learning_rate": 2.2639370372490434e-06,
+      "loss": 0.442,
+      "step": 6701
+    },
+    {
+      "epoch": 0.544302769430683,
+      "grad_norm": 4.317869994988899,
+      "learning_rate": 2.263282361530759e-06,
+      "loss": 0.6317,
+      "step": 6702
+    },
+    {
+      "epoch": 0.5443839844067246,
+      "grad_norm": 3.2344627144715874,
+      "learning_rate": 2.2626277021920813e-06,
+      "loss": 0.4615,
+      "step": 6703
+    },
+    {
+      "epoch": 0.5444651993827662,
+      "grad_norm": 6.42718836422913,
+      "learning_rate": 2.2619730592783108e-06,
+      "loss": 0.5471,
+      "step": 6704
+    },
+    {
+      "epoch": 0.5445464143588078,
+      "grad_norm": 4.488283000655936,
+      "learning_rate": 2.2613184328347453e-06,
+      "loss": 0.4092,
+      "step": 6705
+    },
+    {
+      "epoch": 0.5446276293348493,
+      "grad_norm": 6.305967356057076,
+      "learning_rate": 2.2606638229066802e-06,
+      "loss": 0.6374,
+      "step": 6706
+    },
+    {
+      "epoch": 0.544708844310891,
+      "grad_norm": 4.2258432165081565,
+      "learning_rate": 2.2600092295394125e-06,
+      "loss": 0.5745,
+      "step": 6707
+    },
+    {
+      "epoch": 0.5447900592869325,
+      "grad_norm": 4.384272018008662,
+      "learning_rate": 2.2593546527782362e-06,
+      "loss": 0.433,
+      "step": 6708
+    },
+    {
+      "epoch": 0.5448712742629741,
+      "grad_norm": 6.576525934865906,
+      "learning_rate": 2.2587000926684432e-06,
+      "loss": 0.5326,
+      "step": 6709
+    },
+    {
+      "epoch": 0.5449524892390156,
+      "grad_norm": 5.555291917382588,
+      "learning_rate": 2.258045549255328e-06,
+      "loss": 0.4881,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5450337042150573,
+      "grad_norm": 5.546197578764206,
+      "learning_rate": 2.25739102258418e-06,
+      "loss": 0.479,
+      "step": 6711
+    },
+    {
+      "epoch": 0.5451149191910989,
+      "grad_norm": 6.1439652570444325,
+      "learning_rate": 2.256736512700288e-06,
+      "loss": 0.4543,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5451961341671404,
+      "grad_norm": 5.653213757078279,
+      "learning_rate": 2.2560820196489437e-06,
+      "loss": 0.5768,
+      "step": 6713
+    },
+    {
+      "epoch": 0.545277349143182,
+      "grad_norm": 4.21691379783433,
+      "learning_rate": 2.255427543475432e-06,
+      "loss": 0.4927,
+      "step": 6714
+    },
+    {
+      "epoch": 0.5453585641192236,
+      "grad_norm": 6.798777933865642,
+      "learning_rate": 2.254773084225039e-06,
+      "loss": 0.4237,
+      "step": 6715
+    },
+    {
+      "epoch": 0.5454397790952652,
+      "grad_norm": 4.819909494343674,
+      "learning_rate": 2.254118641943052e-06,
+      "loss": 0.4288,
+      "step": 6716
+    },
+    {
+      "epoch": 0.5455209940713067,
+      "grad_norm": 7.439931633022933,
+      "learning_rate": 2.253464216674753e-06,
+      "loss": 0.482,
+      "step": 6717
+    },
+    {
+      "epoch": 0.5456022090473484,
+      "grad_norm": 7.792402778020558,
+      "learning_rate": 2.2528098084654262e-06,
+      "loss": 0.614,
+      "step": 6718
+    },
+    {
+      "epoch": 0.5456834240233899,
+      "grad_norm": 4.495435217507245,
+      "learning_rate": 2.2521554173603513e-06,
+      "loss": 0.4562,
+      "step": 6719
+    },
+    {
+      "epoch": 0.5457646389994315,
+      "grad_norm": 4.419124752755961,
+      "learning_rate": 2.25150104340481e-06,
+      "loss": 0.5417,
+      "step": 6720
+    },
+    {
+      "epoch": 0.545845853975473,
+      "grad_norm": 6.332131149232858,
+      "learning_rate": 2.2508466866440824e-06,
+      "loss": 0.4154,
+      "step": 6721
+    },
+    {
+      "epoch": 0.5459270689515147,
+      "grad_norm": 6.610876686409083,
+      "learning_rate": 2.2501923471234444e-06,
+      "loss": 0.49,
+      "step": 6722
+    },
+    {
+      "epoch": 0.5460082839275563,
+      "grad_norm": 17.017885967526706,
+      "learning_rate": 2.249538024888174e-06,
+      "loss": 0.4733,
+      "step": 6723
+    },
+    {
+      "epoch": 0.5460894989035978,
+      "grad_norm": 7.28235162312456,
+      "learning_rate": 2.2488837199835477e-06,
+      "loss": 0.6985,
+      "step": 6724
+    },
+    {
+      "epoch": 0.5461707138796394,
+      "grad_norm": 6.165256861279543,
+      "learning_rate": 2.2482294324548376e-06,
+      "loss": 0.5424,
+      "step": 6725
+    },
+    {
+      "epoch": 0.546251928855681,
+      "grad_norm": 7.302972322656833,
+      "learning_rate": 2.2475751623473193e-06,
+      "loss": 0.6049,
+      "step": 6726
+    },
+    {
+      "epoch": 0.5463331438317226,
+      "grad_norm": 4.71990096240607,
+      "learning_rate": 2.2469209097062637e-06,
+      "loss": 0.4342,
+      "step": 6727
+    },
+    {
+      "epoch": 0.5464143588077641,
+      "grad_norm": 6.900667478651783,
+      "learning_rate": 2.246266674576941e-06,
+      "loss": 0.4572,
+      "step": 6728
+    },
+    {
+      "epoch": 0.5464955737838058,
+      "grad_norm": 2.982802060269296,
+      "learning_rate": 2.245612457004622e-06,
+      "loss": 0.4723,
+      "step": 6729
+    },
+    {
+      "epoch": 0.5465767887598473,
+      "grad_norm": 4.176653744332102,
+      "learning_rate": 2.244958257034575e-06,
+      "loss": 0.4191,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5466580037358889,
+      "grad_norm": 4.511011865460297,
+      "learning_rate": 2.244304074712066e-06,
+      "loss": 0.4973,
+      "step": 6731
+    },
+    {
+      "epoch": 0.5467392187119304,
+      "grad_norm": 4.202369107086116,
+      "learning_rate": 2.243649910082363e-06,
+      "loss": 0.6125,
+      "step": 6732
+    },
+    {
+      "epoch": 0.5468204336879721,
+      "grad_norm": 7.339054515036435,
+      "learning_rate": 2.2429957631907285e-06,
+      "loss": 0.444,
+      "step": 6733
+    },
+    {
+      "epoch": 0.5469016486640137,
+      "grad_norm": 4.854339006065371,
+      "learning_rate": 2.2423416340824266e-06,
+      "loss": 0.4205,
+      "step": 6734
+    },
+    {
+      "epoch": 0.5469828636400552,
+      "grad_norm": 5.693035824826316,
+      "learning_rate": 2.241687522802721e-06,
+      "loss": 0.5089,
+      "step": 6735
+    },
+    {
+      "epoch": 0.5470640786160968,
+      "grad_norm": 4.855926081415383,
+      "learning_rate": 2.2410334293968716e-06,
+      "loss": 0.43,
+      "step": 6736
+    },
+    {
+      "epoch": 0.5471452935921384,
+      "grad_norm": 4.266917082773724,
+      "learning_rate": 2.2403793539101387e-06,
+      "loss": 0.4145,
+      "step": 6737
+    },
+    {
+      "epoch": 0.54722650856818,
+      "grad_norm": 5.411690627496338,
+      "learning_rate": 2.2397252963877795e-06,
+      "loss": 0.5073,
+      "step": 6738
+    },
+    {
+      "epoch": 0.5473077235442215,
+      "grad_norm": 2.6098814011197096,
+      "learning_rate": 2.239071256875053e-06,
+      "loss": 0.5423,
+      "step": 6739
+    },
+    {
+      "epoch": 0.5473889385202632,
+      "grad_norm": 5.3581580919616885,
+      "learning_rate": 2.238417235417214e-06,
+      "loss": 0.4069,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5474701534963047,
+      "grad_norm": 4.544117591382317,
+      "learning_rate": 2.237763232059518e-06,
+      "loss": 0.4756,
+      "step": 6741
+    },
+    {
+      "epoch": 0.5475513684723463,
+      "grad_norm": 18.0511878541294,
+      "learning_rate": 2.2371092468472193e-06,
+      "loss": 0.5058,
+      "step": 6742
+    },
+    {
+      "epoch": 0.5476325834483878,
+      "grad_norm": 6.268300319177437,
+      "learning_rate": 2.236455279825569e-06,
+      "loss": 0.4712,
+      "step": 6743
+    },
+    {
+      "epoch": 0.5477137984244295,
+      "grad_norm": 3.663452778757879,
+      "learning_rate": 2.2358013310398174e-06,
+      "loss": 0.5031,
+      "step": 6744
+    },
+    {
+      "epoch": 0.5477950134004711,
+      "grad_norm": 4.936451070405641,
+      "learning_rate": 2.235147400535217e-06,
+      "loss": 0.4914,
+      "step": 6745
+    },
+    {
+      "epoch": 0.5478762283765126,
+      "grad_norm": 5.976409510815185,
+      "learning_rate": 2.2344934883570143e-06,
+      "loss": 0.5417,
+      "step": 6746
+    },
+    {
+      "epoch": 0.5479574433525543,
+      "grad_norm": 7.332839588724573,
+      "learning_rate": 2.2338395945504557e-06,
+      "loss": 0.525,
+      "step": 6747
+    },
+    {
+      "epoch": 0.5480386583285958,
+      "grad_norm": 5.026062475739817,
+      "learning_rate": 2.23318571916079e-06,
+      "loss": 0.6906,
+      "step": 6748
+    },
+    {
+      "epoch": 0.5481198733046374,
+      "grad_norm": 5.070833605866744,
+      "learning_rate": 2.2325318622332606e-06,
+      "loss": 0.4563,
+      "step": 6749
+    },
+    {
+      "epoch": 0.5482010882806789,
+      "grad_norm": 4.9673547801455955,
+      "learning_rate": 2.2318780238131095e-06,
+      "loss": 0.5338,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5482823032567206,
+      "grad_norm": 3.9729119791581744,
+      "learning_rate": 2.2312242039455816e-06,
+      "loss": 0.503,
+      "step": 6751
+    },
+    {
+      "epoch": 0.5483635182327621,
+      "grad_norm": 3.9917626185437878,
+      "learning_rate": 2.230570402675916e-06,
+      "loss": 0.3857,
+      "step": 6752
+    },
+    {
+      "epoch": 0.5484447332088037,
+      "grad_norm": 4.014300920446785,
+      "learning_rate": 2.2299166200493526e-06,
+      "loss": 0.4657,
+      "step": 6753
+    },
+    {
+      "epoch": 0.5485259481848452,
+      "grad_norm": 5.148556943279063,
+      "learning_rate": 2.2292628561111285e-06,
+      "loss": 0.4921,
+      "step": 6754
+    },
+    {
+      "epoch": 0.5486071631608869,
+      "grad_norm": 3.2886345422623715,
+      "learning_rate": 2.228609110906483e-06,
+      "loss": 0.4815,
+      "step": 6755
+    },
+    {
+      "epoch": 0.5486883781369285,
+      "grad_norm": 9.805600602297103,
+      "learning_rate": 2.2279553844806506e-06,
+      "loss": 0.5324,
+      "step": 6756
+    },
+    {
+      "epoch": 0.54876959311297,
+      "grad_norm": 7.9338566902011385,
+      "learning_rate": 2.2273016768788653e-06,
+      "loss": 0.4963,
+      "step": 6757
+    },
+    {
+      "epoch": 0.5488508080890117,
+      "grad_norm": 4.352389013230028,
+      "learning_rate": 2.2266479881463614e-06,
+      "loss": 0.4975,
+      "step": 6758
+    },
+    {
+      "epoch": 0.5489320230650532,
+      "grad_norm": 5.035317373817535,
+      "learning_rate": 2.2259943183283696e-06,
+      "loss": 0.5316,
+      "step": 6759
+    },
+    {
+      "epoch": 0.5490132380410948,
+      "grad_norm": 6.329671082512638,
+      "learning_rate": 2.2253406674701206e-06,
+      "loss": 0.4758,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5490944530171363,
+      "grad_norm": 4.350812102992751,
+      "learning_rate": 2.2246870356168447e-06,
+      "loss": 0.3731,
+      "step": 6761
+    },
+    {
+      "epoch": 0.549175667993178,
+      "grad_norm": 4.411297111261324,
+      "learning_rate": 2.224033422813768e-06,
+      "loss": 0.4601,
+      "step": 6762
+    },
+    {
+      "epoch": 0.5492568829692195,
+      "grad_norm": 3.3546196364460172,
+      "learning_rate": 2.2233798291061177e-06,
+      "loss": 0.5048,
+      "step": 6763
+    },
+    {
+      "epoch": 0.5493380979452611,
+      "grad_norm": 5.75969034058185,
+      "learning_rate": 2.2227262545391204e-06,
+      "loss": 0.5135,
+      "step": 6764
+    },
+    {
+      "epoch": 0.5494193129213026,
+      "grad_norm": 4.676293581342606,
+      "learning_rate": 2.222072699157998e-06,
+      "loss": 0.382,
+      "step": 6765
+    },
+    {
+      "epoch": 0.5495005278973443,
+      "grad_norm": 4.301771743349808,
+      "learning_rate": 2.2214191630079733e-06,
+      "loss": 0.4019,
+      "step": 6766
+    },
+    {
+      "epoch": 0.5495817428733859,
+      "grad_norm": 8.356786267103574,
+      "learning_rate": 2.2207656461342696e-06,
+      "loss": 0.486,
+      "step": 6767
+    },
+    {
+      "epoch": 0.5496629578494274,
+      "grad_norm": 5.187735340758213,
+      "learning_rate": 2.2201121485821053e-06,
+      "loss": 0.5106,
+      "step": 6768
+    },
+    {
+      "epoch": 0.5497441728254691,
+      "grad_norm": 14.002790250484441,
+      "learning_rate": 2.2194586703966976e-06,
+      "loss": 0.5786,
+      "step": 6769
+    },
+    {
+      "epoch": 0.5498253878015106,
+      "grad_norm": 4.598645970086092,
+      "learning_rate": 2.218805211623266e-06,
+      "loss": 0.6202,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5499066027775522,
+      "grad_norm": 7.035714823525913,
+      "learning_rate": 2.2181517723070263e-06,
+      "loss": 0.3767,
+      "step": 6771
+    },
+    {
+      "epoch": 0.5499878177535937,
+      "grad_norm": 3.789629383739064,
+      "learning_rate": 2.2174983524931916e-06,
+      "loss": 0.4665,
+      "step": 6772
+    },
+    {
+      "epoch": 0.5500690327296354,
+      "grad_norm": 5.153278702932309,
+      "learning_rate": 2.216844952226975e-06,
+      "loss": 0.5535,
+      "step": 6773
+    },
+    {
+      "epoch": 0.5501502477056769,
+      "grad_norm": 5.090968399667655,
+      "learning_rate": 2.2161915715535903e-06,
+      "loss": 0.5416,
+      "step": 6774
+    },
+    {
+      "epoch": 0.5502314626817185,
+      "grad_norm": 4.880285899644429,
+      "learning_rate": 2.2155382105182462e-06,
+      "loss": 0.528,
+      "step": 6775
+    },
+    {
+      "epoch": 0.55031267765776,
+      "grad_norm": 7.133831222495926,
+      "learning_rate": 2.214884869166152e-06,
+      "loss": 0.4645,
+      "step": 6776
+    },
+    {
+      "epoch": 0.5503938926338017,
+      "grad_norm": 5.111553296411102,
+      "learning_rate": 2.214231547542517e-06,
+      "loss": 0.5628,
+      "step": 6777
+    },
+    {
+      "epoch": 0.5504751076098433,
+      "grad_norm": 4.076454306491572,
+      "learning_rate": 2.213578245692546e-06,
+      "loss": 0.4838,
+      "step": 6778
+    },
+    {
+      "epoch": 0.5505563225858848,
+      "grad_norm": 5.211719936927261,
+      "learning_rate": 2.2129249636614443e-06,
+      "loss": 0.563,
+      "step": 6779
+    },
+    {
+      "epoch": 0.5506375375619265,
+      "grad_norm": 12.302787876796701,
+      "learning_rate": 2.2122717014944167e-06,
+      "loss": 0.4786,
+      "step": 6780
+    },
+    {
+      "epoch": 0.550718752537968,
+      "grad_norm": 4.330340043687264,
+      "learning_rate": 2.2116184592366643e-06,
+      "loss": 0.5747,
+      "step": 6781
+    },
+    {
+      "epoch": 0.5507999675140096,
+      "grad_norm": 5.034937065269623,
+      "learning_rate": 2.2109652369333873e-06,
+      "loss": 0.4252,
+      "step": 6782
+    },
+    {
+      "epoch": 0.5508811824900511,
+      "grad_norm": 7.614248716990825,
+      "learning_rate": 2.2103120346297864e-06,
+      "loss": 0.4404,
+      "step": 6783
+    },
+    {
+      "epoch": 0.5509623974660928,
+      "grad_norm": 7.1625705175917425,
+      "learning_rate": 2.2096588523710606e-06,
+      "loss": 0.5373,
+      "step": 6784
+    },
+    {
+      "epoch": 0.5510436124421343,
+      "grad_norm": 4.064726317144616,
+      "learning_rate": 2.2090056902024045e-06,
+      "loss": 0.4773,
+      "step": 6785
+    },
+    {
+      "epoch": 0.5511248274181759,
+      "grad_norm": 4.1799707253524065,
+      "learning_rate": 2.208352548169015e-06,
+      "loss": 0.5596,
+      "step": 6786
+    },
+    {
+      "epoch": 0.5512060423942174,
+      "grad_norm": 5.2430508668568425,
+      "learning_rate": 2.2076994263160863e-06,
+      "loss": 0.5455,
+      "step": 6787
+    },
+    {
+      "epoch": 0.5512872573702591,
+      "grad_norm": 5.514510034389823,
+      "learning_rate": 2.2070463246888094e-06,
+      "loss": 0.4733,
+      "step": 6788
+    },
+    {
+      "epoch": 0.5513684723463007,
+      "grad_norm": 4.50738218564209,
+      "learning_rate": 2.206393243332376e-06,
+      "loss": 0.5448,
+      "step": 6789
+    },
+    {
+      "epoch": 0.5514496873223422,
+      "grad_norm": 4.34724451667795,
+      "learning_rate": 2.2057401822919775e-06,
+      "loss": 0.5013,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5515309022983839,
+      "grad_norm": 5.578815799717539,
+      "learning_rate": 2.2050871416128005e-06,
+      "loss": 0.5814,
+      "step": 6791
+    },
+    {
+      "epoch": 0.5516121172744254,
+      "grad_norm": 6.701219984709083,
+      "learning_rate": 2.204434121340032e-06,
+      "loss": 0.3959,
+      "step": 6792
+    },
+    {
+      "epoch": 0.551693332250467,
+      "grad_norm": 3.6134666388120404,
+      "learning_rate": 2.203781121518859e-06,
+      "loss": 0.536,
+      "step": 6793
+    },
+    {
+      "epoch": 0.5517745472265085,
+      "grad_norm": 6.117283840596765,
+      "learning_rate": 2.2031281421944643e-06,
+      "loss": 0.4515,
+      "step": 6794
+    },
+    {
+      "epoch": 0.5518557622025502,
+      "grad_norm": 3.9112049782621003,
+      "learning_rate": 2.2024751834120302e-06,
+      "loss": 0.4622,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5519369771785917,
+      "grad_norm": 3.8079103420465437,
+      "learning_rate": 2.20182224521674e-06,
+      "loss": 0.5821,
+      "step": 6796
+    },
+    {
+      "epoch": 0.5520181921546333,
+      "grad_norm": 5.019201884297034,
+      "learning_rate": 2.2011693276537722e-06,
+      "loss": 0.6276,
+      "step": 6797
+    },
+    {
+      "epoch": 0.5520994071306748,
+      "grad_norm": 5.4407005403339825,
+      "learning_rate": 2.2005164307683047e-06,
+      "loss": 0.4406,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5521806221067165,
+      "grad_norm": 4.301332089064263,
+      "learning_rate": 2.199863554605515e-06,
+      "loss": 0.5167,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5522618370827581,
+      "grad_norm": 7.493043226544131,
+      "learning_rate": 2.19921069921058e-06,
+      "loss": 0.5069,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5523430520587996,
+      "grad_norm": 3.8437035341706967,
+      "learning_rate": 2.1985578646286717e-06,
+      "loss": 0.6306,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5524242670348413,
+      "grad_norm": 3.805644122254566,
+      "learning_rate": 2.197905050904964e-06,
+      "loss": 0.4094,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5525054820108828,
+      "grad_norm": 6.899378255822541,
+      "learning_rate": 2.197252258084629e-06,
+      "loss": 0.4034,
+      "step": 6803
+    },
+    {
+      "epoch": 0.5525866969869244,
+      "grad_norm": 7.710934181517784,
+      "learning_rate": 2.196599486212834e-06,
+      "loss": 0.5314,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5526679119629659,
+      "grad_norm": 3.3575643225535856,
+      "learning_rate": 2.1959467353347494e-06,
+      "loss": 0.4281,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5527491269390076,
+      "grad_norm": 3.714838991436428,
+      "learning_rate": 2.195294005495542e-06,
+      "loss": 0.5778,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5528303419150491,
+      "grad_norm": 6.171375505356923,
+      "learning_rate": 2.1946412967403763e-06,
+      "loss": 0.5032,
+      "step": 6807
+    },
+    {
+      "epoch": 0.5529115568910907,
+      "grad_norm": 5.843750427655172,
+      "learning_rate": 2.1939886091144165e-06,
+      "loss": 0.4943,
+      "step": 6808
+    },
+    {
+      "epoch": 0.5529927718671322,
+      "grad_norm": 3.222741223173242,
+      "learning_rate": 2.193335942662826e-06,
+      "loss": 0.4707,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5530739868431739,
+      "grad_norm": 3.916188730239071,
+      "learning_rate": 2.192683297430766e-06,
+      "loss": 0.4572,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5531552018192155,
+      "grad_norm": 4.698273133823479,
+      "learning_rate": 2.1920306734633932e-06,
+      "loss": 0.4173,
+      "step": 6811
+    },
+    {
+      "epoch": 0.553236416795257,
+      "grad_norm": 5.974971528842556,
+      "learning_rate": 2.1913780708058694e-06,
+      "loss": 0.5129,
+      "step": 6812
+    },
+    {
+      "epoch": 0.5533176317712987,
+      "grad_norm": 4.899970058534168,
+      "learning_rate": 2.19072548950335e-06,
+      "loss": 0.4309,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5533988467473402,
+      "grad_norm": 4.224538543746903,
+      "learning_rate": 2.190072929600989e-06,
+      "loss": 0.6603,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5534800617233818,
+      "grad_norm": 9.421465257051333,
+      "learning_rate": 2.189420391143941e-06,
+      "loss": 0.4734,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5535612766994233,
+      "grad_norm": 6.665428434932173,
+      "learning_rate": 2.1887678741773592e-06,
+      "loss": 0.5195,
+      "step": 6816
+    },
+    {
+      "epoch": 0.553642491675465,
+      "grad_norm": 2.89387484029611,
+      "learning_rate": 2.188115378746392e-06,
+      "loss": 0.4763,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5537237066515065,
+      "grad_norm": 6.823773735416532,
+      "learning_rate": 2.1874629048961904e-06,
+      "loss": 0.5321,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5538049216275481,
+      "grad_norm": 4.521778809799006,
+      "learning_rate": 2.1868104526719023e-06,
+      "loss": 0.5816,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5538861366035897,
+      "grad_norm": 3.3396433922902906,
+      "learning_rate": 2.1861580221186726e-06,
+      "loss": 0.4168,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5539673515796313,
+      "grad_norm": 3.3374583977198036,
+      "learning_rate": 2.185505613281647e-06,
+      "loss": 0.4614,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5540485665556729,
+      "grad_norm": 6.068612637720992,
+      "learning_rate": 2.1848532262059696e-06,
+      "loss": 0.5797,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5541297815317144,
+      "grad_norm": 5.0334311376181144,
+      "learning_rate": 2.1842008609367794e-06,
+      "loss": 0.6781,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5542109965077561,
+      "grad_norm": 4.81150612829465,
+      "learning_rate": 2.183548517519219e-06,
+      "loss": 0.4113,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5542922114837976,
+      "grad_norm": 7.137103277777565,
+      "learning_rate": 2.1828961959984267e-06,
+      "loss": 0.502,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5543734264598392,
+      "grad_norm": 4.93103788302294,
+      "learning_rate": 2.18224389641954e-06,
+      "loss": 0.3781,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5544546414358807,
+      "grad_norm": 4.17706299919582,
+      "learning_rate": 2.1815916188276925e-06,
+      "loss": 0.5585,
+      "step": 6827
+    },
+    {
+      "epoch": 0.5545358564119224,
+      "grad_norm": 6.194948711383427,
+      "learning_rate": 2.18093936326802e-06,
+      "loss": 0.5219,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5546170713879639,
+      "grad_norm": 7.761708704297234,
+      "learning_rate": 2.180287129785656e-06,
+      "loss": 0.4665,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5546982863640055,
+      "grad_norm": 5.591231620591249,
+      "learning_rate": 2.1796349184257294e-06,
+      "loss": 0.4973,
+      "step": 6830
+    },
+    {
+      "epoch": 0.554779501340047,
+      "grad_norm": 3.9216643074294595,
+      "learning_rate": 2.1789827292333717e-06,
+      "loss": 0.3763,
+      "step": 6831
+    },
+    {
+      "epoch": 0.5548607163160887,
+      "grad_norm": 4.444201370705435,
+      "learning_rate": 2.1783305622537106e-06,
+      "loss": 0.522,
+      "step": 6832
+    },
+    {
+      "epoch": 0.5549419312921303,
+      "grad_norm": 3.5234577542993972,
+      "learning_rate": 2.1776784175318705e-06,
+      "loss": 0.6173,
+      "step": 6833
+    },
+    {
+      "epoch": 0.5550231462681718,
+      "grad_norm": 5.522357961927791,
+      "learning_rate": 2.1770262951129792e-06,
+      "loss": 0.4304,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5551043612442135,
+      "grad_norm": 2.753675801015695,
+      "learning_rate": 2.1763741950421595e-06,
+      "loss": 0.5176,
+      "step": 6835
+    },
+    {
+      "epoch": 0.555185576220255,
+      "grad_norm": 5.5643466943030715,
+      "learning_rate": 2.175722117364531e-06,
+      "loss": 0.6027,
+      "step": 6836
+    },
+    {
+      "epoch": 0.5552667911962966,
+      "grad_norm": 2.6499029128692957,
+      "learning_rate": 2.175070062125217e-06,
+      "loss": 0.472,
+      "step": 6837
+    },
+    {
+      "epoch": 0.5553480061723381,
+      "grad_norm": 11.241112599357884,
+      "learning_rate": 2.1744180293693355e-06,
+      "loss": 0.5573,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5554292211483798,
+      "grad_norm": 5.094638009416633,
+      "learning_rate": 2.173766019142002e-06,
+      "loss": 0.4747,
+      "step": 6839
+    },
+    {
+      "epoch": 0.5555104361244213,
+      "grad_norm": 6.191994268871611,
+      "learning_rate": 2.1731140314883346e-06,
+      "loss": 0.5245,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5555916511004629,
+      "grad_norm": 4.2743419408023495,
+      "learning_rate": 2.1724620664534453e-06,
+      "loss": 0.508,
+      "step": 6841
+    },
+    {
+      "epoch": 0.5556728660765045,
+      "grad_norm": 12.369287470212933,
+      "learning_rate": 2.1718101240824485e-06,
+      "loss": 0.6018,
+      "step": 6842
+    },
+    {
+      "epoch": 0.5557540810525461,
+      "grad_norm": 6.34277452834017,
+      "learning_rate": 2.171158204420453e-06,
+      "loss": 0.5897,
+      "step": 6843
+    },
+    {
+      "epoch": 0.5558352960285877,
+      "grad_norm": 3.6842271055521763,
+      "learning_rate": 2.17050630751257e-06,
+      "loss": 0.3811,
+      "step": 6844
+    },
+    {
+      "epoch": 0.5559165110046292,
+      "grad_norm": 4.1774312445120545,
+      "learning_rate": 2.169854433403907e-06,
+      "loss": 0.7328,
+      "step": 6845
+    },
+    {
+      "epoch": 0.5559977259806709,
+      "grad_norm": 4.702814140349739,
+      "learning_rate": 2.169202582139569e-06,
+      "loss": 0.5899,
+      "step": 6846
+    },
+    {
+      "epoch": 0.5560789409567124,
+      "grad_norm": 5.095589215614225,
+      "learning_rate": 2.1685507537646622e-06,
+      "loss": 0.4751,
+      "step": 6847
+    },
+    {
+      "epoch": 0.556160155932754,
+      "grad_norm": 5.703171345481269,
+      "learning_rate": 2.1678989483242896e-06,
+      "loss": 0.4403,
+      "step": 6848
+    },
+    {
+      "epoch": 0.5562413709087956,
+      "grad_norm": 4.75054750690896,
+      "learning_rate": 2.1672471658635506e-06,
+      "loss": 0.5135,
+      "step": 6849
+    },
+    {
+      "epoch": 0.5563225858848372,
+      "grad_norm": 3.3902339048047305,
+      "learning_rate": 2.166595406427548e-06,
+      "loss": 0.6971,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5564038008608787,
+      "grad_norm": 5.852734969187135,
+      "learning_rate": 2.1659436700613787e-06,
+      "loss": 0.4691,
+      "step": 6851
+    },
+    {
+      "epoch": 0.5564850158369203,
+      "grad_norm": 3.727932064595546,
+      "learning_rate": 2.1652919568101386e-06,
+      "loss": 0.5747,
+      "step": 6852
+    },
+    {
+      "epoch": 0.5565662308129619,
+      "grad_norm": 5.153139343207781,
+      "learning_rate": 2.1646402667189245e-06,
+      "loss": 0.4892,
+      "step": 6853
+    },
+    {
+      "epoch": 0.5566474457890035,
+      "grad_norm": 3.8440190223031125,
+      "learning_rate": 2.1639885998328293e-06,
+      "loss": 0.4627,
+      "step": 6854
+    },
+    {
+      "epoch": 0.5567286607650451,
+      "grad_norm": 8.606115530142445,
+      "learning_rate": 2.1633369561969435e-06,
+      "loss": 0.4342,
+      "step": 6855
+    },
+    {
+      "epoch": 0.5568098757410866,
+      "grad_norm": 3.6671273924850407,
+      "learning_rate": 2.1626853358563595e-06,
+      "loss": 0.5365,
+      "step": 6856
+    },
+    {
+      "epoch": 0.5568910907171283,
+      "grad_norm": 3.835808189282639,
+      "learning_rate": 2.162033738856165e-06,
+      "loss": 0.491,
+      "step": 6857
+    },
+    {
+      "epoch": 0.5569723056931698,
+      "grad_norm": 6.132790993906602,
+      "learning_rate": 2.161382165241446e-06,
+      "loss": 0.4935,
+      "step": 6858
+    },
+    {
+      "epoch": 0.5570535206692114,
+      "grad_norm": 4.4718395237709485,
+      "learning_rate": 2.1607306150572905e-06,
+      "loss": 0.5225,
+      "step": 6859
+    },
+    {
+      "epoch": 0.557134735645253,
+      "grad_norm": 5.4291132456188445,
+      "learning_rate": 2.1600790883487805e-06,
+      "loss": 0.4914,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5572159506212946,
+      "grad_norm": 3.396009857568952,
+      "learning_rate": 2.159427585160999e-06,
+      "loss": 0.6867,
+      "step": 6861
+    },
+    {
+      "epoch": 0.5572971655973361,
+      "grad_norm": 6.2945750984242945,
+      "learning_rate": 2.1587761055390247e-06,
+      "loss": 0.3675,
+      "step": 6862
+    },
+    {
+      "epoch": 0.5573783805733777,
+      "grad_norm": 10.344490630116207,
+      "learning_rate": 2.1581246495279388e-06,
+      "loss": 0.4676,
+      "step": 6863
+    },
+    {
+      "epoch": 0.5574595955494193,
+      "grad_norm": 4.03264581660929,
+      "learning_rate": 2.1574732171728187e-06,
+      "loss": 0.5014,
+      "step": 6864
+    },
+    {
+      "epoch": 0.5575408105254609,
+      "grad_norm": 5.595299416089842,
+      "learning_rate": 2.1568218085187375e-06,
+      "loss": 0.5567,
+      "step": 6865
+    },
+    {
+      "epoch": 0.5576220255015025,
+      "grad_norm": 4.615982346308546,
+      "learning_rate": 2.1561704236107715e-06,
+      "loss": 0.4441,
+      "step": 6866
+    },
+    {
+      "epoch": 0.557703240477544,
+      "grad_norm": 3.371649845113746,
+      "learning_rate": 2.1555190624939933e-06,
+      "loss": 0.4734,
+      "step": 6867
+    },
+    {
+      "epoch": 0.5577844554535857,
+      "grad_norm": 4.993213905650965,
+      "learning_rate": 2.154867725213472e-06,
+      "loss": 0.5791,
+      "step": 6868
+    },
+    {
+      "epoch": 0.5578656704296272,
+      "grad_norm": 5.433984223606404,
+      "learning_rate": 2.154216411814278e-06,
+      "loss": 0.6413,
+      "step": 6869
+    },
+    {
+      "epoch": 0.5579468854056688,
+      "grad_norm": 4.1089167750307505,
+      "learning_rate": 2.1535651223414783e-06,
+      "loss": 0.4065,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5580281003817104,
+      "grad_norm": 6.410732794666789,
+      "learning_rate": 2.1529138568401377e-06,
+      "loss": 0.4132,
+      "step": 6871
+    },
+    {
+      "epoch": 0.558109315357752,
+      "grad_norm": 2.952240951187637,
+      "learning_rate": 2.1522626153553224e-06,
+      "loss": 0.5569,
+      "step": 6872
+    },
+    {
+      "epoch": 0.5581905303337935,
+      "grad_norm": 5.778680431168145,
+      "learning_rate": 2.1516113979320937e-06,
+      "loss": 0.615,
+      "step": 6873
+    },
+    {
+      "epoch": 0.5582717453098351,
+      "grad_norm": 5.40543144487348,
+      "learning_rate": 2.150960204615511e-06,
+      "loss": 0.4005,
+      "step": 6874
+    },
+    {
+      "epoch": 0.5583529602858767,
+      "grad_norm": 5.360050799209055,
+      "learning_rate": 2.1503090354506366e-06,
+      "loss": 0.4878,
+      "step": 6875
+    },
+    {
+      "epoch": 0.5584341752619183,
+      "grad_norm": 4.768908334280203,
+      "learning_rate": 2.1496578904825253e-06,
+      "loss": 0.606,
+      "step": 6876
+    },
+    {
+      "epoch": 0.5585153902379599,
+      "grad_norm": 4.8420372746526255,
+      "learning_rate": 2.149006769756234e-06,
+      "loss": 0.3823,
+      "step": 6877
+    },
+    {
+      "epoch": 0.5585966052140015,
+      "grad_norm": 7.2191793614780835,
+      "learning_rate": 2.148355673316817e-06,
+      "loss": 0.4933,
+      "step": 6878
+    },
+    {
+      "epoch": 0.5586778201900431,
+      "grad_norm": 5.635609934888895,
+      "learning_rate": 2.1477046012093263e-06,
+      "loss": 0.4284,
+      "step": 6879
+    },
+    {
+      "epoch": 0.5587590351660846,
+      "grad_norm": 4.860255565023494,
+      "learning_rate": 2.147053553478813e-06,
+      "loss": 0.4215,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5588402501421262,
+      "grad_norm": 6.326189480768318,
+      "learning_rate": 2.1464025301703243e-06,
+      "loss": 0.3612,
+      "step": 6881
+    },
+    {
+      "epoch": 0.5589214651181678,
+      "grad_norm": 10.398795327314401,
+      "learning_rate": 2.145751531328911e-06,
+      "loss": 0.638,
+      "step": 6882
+    },
+    {
+      "epoch": 0.5590026800942094,
+      "grad_norm": 5.9840259751672455,
+      "learning_rate": 2.1451005569996157e-06,
+      "loss": 0.4752,
+      "step": 6883
+    },
+    {
+      "epoch": 0.5590838950702509,
+      "grad_norm": 6.729449032368889,
+      "learning_rate": 2.144449607227483e-06,
+      "loss": 0.4235,
+      "step": 6884
+    },
+    {
+      "epoch": 0.5591651100462925,
+      "grad_norm": 4.99524830446243,
+      "learning_rate": 2.143798682057558e-06,
+      "loss": 0.565,
+      "step": 6885
+    },
+    {
+      "epoch": 0.5592463250223341,
+      "grad_norm": 11.826585338715425,
+      "learning_rate": 2.1431477815348775e-06,
+      "loss": 0.644,
+      "step": 6886
+    },
+    {
+      "epoch": 0.5593275399983757,
+      "grad_norm": 4.772181141872521,
+      "learning_rate": 2.1424969057044815e-06,
+      "loss": 0.6231,
+      "step": 6887
+    },
+    {
+      "epoch": 0.5594087549744173,
+      "grad_norm": 5.569287892562355,
+      "learning_rate": 2.1418460546114087e-06,
+      "loss": 0.4135,
+      "step": 6888
+    },
+    {
+      "epoch": 0.5594899699504589,
+      "grad_norm": 3.637243154196817,
+      "learning_rate": 2.141195228300693e-06,
+      "loss": 0.4882,
+      "step": 6889
+    },
+    {
+      "epoch": 0.5595711849265005,
+      "grad_norm": 4.970081667441746,
+      "learning_rate": 2.140544426817368e-06,
+      "loss": 0.4451,
+      "step": 6890
+    },
+    {
+      "epoch": 0.559652399902542,
+      "grad_norm": 6.0593249594190715,
+      "learning_rate": 2.139893650206467e-06,
+      "loss": 0.4589,
+      "step": 6891
+    },
+    {
+      "epoch": 0.5597336148785836,
+      "grad_norm": 4.272966835485669,
+      "learning_rate": 2.1392428985130192e-06,
+      "loss": 0.4929,
+      "step": 6892
+    },
+    {
+      "epoch": 0.5598148298546252,
+      "grad_norm": 4.935701907357544,
+      "learning_rate": 2.138592171782053e-06,
+      "loss": 0.6183,
+      "step": 6893
+    },
+    {
+      "epoch": 0.5598960448306668,
+      "grad_norm": 4.965836504024703,
+      "learning_rate": 2.137941470058597e-06,
+      "loss": 0.5274,
+      "step": 6894
+    },
+    {
+      "epoch": 0.5599772598067083,
+      "grad_norm": 3.8763705524472245,
+      "learning_rate": 2.1372907933876745e-06,
+      "loss": 0.6235,
+      "step": 6895
+    },
+    {
+      "epoch": 0.56005847478275,
+      "grad_norm": 3.7922687516844387,
+      "learning_rate": 2.13664014181431e-06,
+      "loss": 0.4554,
+      "step": 6896
+    },
+    {
+      "epoch": 0.5601396897587915,
+      "grad_norm": 5.143643713764753,
+      "learning_rate": 2.1359895153835235e-06,
+      "loss": 0.7668,
+      "step": 6897
+    },
+    {
+      "epoch": 0.5602209047348331,
+      "grad_norm": 3.9929186611253686,
+      "learning_rate": 2.1353389141403373e-06,
+      "loss": 0.5105,
+      "step": 6898
+    },
+    {
+      "epoch": 0.5603021197108747,
+      "grad_norm": 7.4139111080640605,
+      "learning_rate": 2.134688338129768e-06,
+      "loss": 0.3908,
+      "step": 6899
+    },
+    {
+      "epoch": 0.5603833346869163,
+      "grad_norm": 4.298081345320913,
+      "learning_rate": 2.1340377873968313e-06,
+      "loss": 0.3907,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5604645496629579,
+      "grad_norm": 3.6665218877774404,
+      "learning_rate": 2.133387261986544e-06,
+      "loss": 0.4315,
+      "step": 6901
+    },
+    {
+      "epoch": 0.5605457646389994,
+      "grad_norm": 3.345078936648102,
+      "learning_rate": 2.132736761943917e-06,
+      "loss": 0.4787,
+      "step": 6902
+    },
+    {
+      "epoch": 0.560626979615041,
+      "grad_norm": 5.603195213119475,
+      "learning_rate": 2.1320862873139627e-06,
+      "loss": 0.517,
+      "step": 6903
+    },
+    {
+      "epoch": 0.5607081945910826,
+      "grad_norm": 4.582210135280795,
+      "learning_rate": 2.1314358381416906e-06,
+      "loss": 0.4446,
+      "step": 6904
+    },
+    {
+      "epoch": 0.5607894095671242,
+      "grad_norm": 5.520444524716127,
+      "learning_rate": 2.130785414472108e-06,
+      "loss": 0.5447,
+      "step": 6905
+    },
+    {
+      "epoch": 0.5608706245431657,
+      "grad_norm": 3.594858666968153,
+      "learning_rate": 2.1301350163502194e-06,
+      "loss": 0.4709,
+      "step": 6906
+    },
+    {
+      "epoch": 0.5609518395192074,
+      "grad_norm": 5.775944364817623,
+      "learning_rate": 2.1294846438210316e-06,
+      "loss": 0.5108,
+      "step": 6907
+    },
+    {
+      "epoch": 0.5610330544952489,
+      "grad_norm": 4.911293017998235,
+      "learning_rate": 2.128834296929545e-06,
+      "loss": 0.458,
+      "step": 6908
+    },
+    {
+      "epoch": 0.5611142694712905,
+      "grad_norm": 5.313549191131658,
+      "learning_rate": 2.12818397572076e-06,
+      "loss": 0.5297,
+      "step": 6909
+    },
+    {
+      "epoch": 0.5611954844473321,
+      "grad_norm": 5.235033497622972,
+      "learning_rate": 2.1275336802396775e-06,
+      "loss": 0.6024,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5612766994233737,
+      "grad_norm": 8.31540179839902,
+      "learning_rate": 2.1268834105312926e-06,
+      "loss": 0.4589,
+      "step": 6911
+    },
+    {
+      "epoch": 0.5613579143994153,
+      "grad_norm": 5.844285401664914,
+      "learning_rate": 2.1262331666406003e-06,
+      "loss": 0.687,
+      "step": 6912
+    },
+    {
+      "epoch": 0.5614391293754568,
+      "grad_norm": 5.506426123033813,
+      "learning_rate": 2.125582948612595e-06,
+      "loss": 0.3788,
+      "step": 6913
+    },
+    {
+      "epoch": 0.5615203443514984,
+      "grad_norm": 4.938418738241158,
+      "learning_rate": 2.124932756492269e-06,
+      "loss": 0.4786,
+      "step": 6914
+    },
+    {
+      "epoch": 0.56160155932754,
+      "grad_norm": 5.011490627219829,
+      "learning_rate": 2.1242825903246104e-06,
+      "loss": 0.5104,
+      "step": 6915
+    },
+    {
+      "epoch": 0.5616827743035816,
+      "grad_norm": 4.130351403264876,
+      "learning_rate": 2.1236324501546073e-06,
+      "loss": 0.5,
+      "step": 6916
+    },
+    {
+      "epoch": 0.5617639892796231,
+      "grad_norm": 5.131299996818871,
+      "learning_rate": 2.1229823360272483e-06,
+      "loss": 0.5647,
+      "step": 6917
+    },
+    {
+      "epoch": 0.5618452042556648,
+      "grad_norm": 4.930401157312547,
+      "learning_rate": 2.1223322479875157e-06,
+      "loss": 0.6934,
+      "step": 6918
+    },
+    {
+      "epoch": 0.5619264192317063,
+      "grad_norm": 4.218515382201296,
+      "learning_rate": 2.1216821860803922e-06,
+      "loss": 0.5194,
+      "step": 6919
+    },
+    {
+      "epoch": 0.5620076342077479,
+      "grad_norm": 4.955917407462902,
+      "learning_rate": 2.12103215035086e-06,
+      "loss": 0.548,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5620888491837895,
+      "grad_norm": 7.097785428246493,
+      "learning_rate": 2.1203821408438973e-06,
+      "loss": 0.3326,
+      "step": 6921
+    },
+    {
+      "epoch": 0.5621700641598311,
+      "grad_norm": 5.512144520285594,
+      "learning_rate": 2.1197321576044803e-06,
+      "loss": 0.5434,
+      "step": 6922
+    },
+    {
+      "epoch": 0.5622512791358727,
+      "grad_norm": 4.559962704037825,
+      "learning_rate": 2.119082200677587e-06,
+      "loss": 0.706,
+      "step": 6923
+    },
+    {
+      "epoch": 0.5623324941119142,
+      "grad_norm": 3.716157164787584,
+      "learning_rate": 2.1184322701081884e-06,
+      "loss": 0.4711,
+      "step": 6924
+    },
+    {
+      "epoch": 0.5624137090879558,
+      "grad_norm": 5.005561099415801,
+      "learning_rate": 2.117782365941257e-06,
+      "loss": 0.6163,
+      "step": 6925
+    },
+    {
+      "epoch": 0.5624949240639974,
+      "grad_norm": 3.3652327554641808,
+      "learning_rate": 2.1171324882217644e-06,
+      "loss": 0.5811,
+      "step": 6926
+    },
+    {
+      "epoch": 0.562576139040039,
+      "grad_norm": 3.4936659274070485,
+      "learning_rate": 2.116482636994677e-06,
+      "loss": 0.4976,
+      "step": 6927
+    },
+    {
+      "epoch": 0.5626573540160805,
+      "grad_norm": 15.606156521919054,
+      "learning_rate": 2.11583281230496e-06,
+      "loss": 0.3923,
+      "step": 6928
+    },
+    {
+      "epoch": 0.5627385689921222,
+      "grad_norm": 8.796650533681337,
+      "learning_rate": 2.11518301419758e-06,
+      "loss": 0.5009,
+      "step": 6929
+    },
+    {
+      "epoch": 0.5628197839681637,
+      "grad_norm": 5.043513216497029,
+      "learning_rate": 2.1145332427174995e-06,
+      "loss": 0.3586,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5629009989442053,
+      "grad_norm": 5.302763968985061,
+      "learning_rate": 2.1138834979096778e-06,
+      "loss": 0.4416,
+      "step": 6931
+    },
+    {
+      "epoch": 0.5629822139202469,
+      "grad_norm": 4.618030423336223,
+      "learning_rate": 2.1132337798190743e-06,
+      "loss": 0.5591,
+      "step": 6932
+    },
+    {
+      "epoch": 0.5630634288962885,
+      "grad_norm": 5.305923379739884,
+      "learning_rate": 2.112584088490647e-06,
+      "loss": 0.6132,
+      "step": 6933
+    },
+    {
+      "epoch": 0.5631446438723301,
+      "grad_norm": 8.514031471933345,
+      "learning_rate": 2.11193442396935e-06,
+      "loss": 0.4011,
+      "step": 6934
+    },
+    {
+      "epoch": 0.5632258588483716,
+      "grad_norm": 4.100042747142503,
+      "learning_rate": 2.111284786300137e-06,
+      "loss": 0.6595,
+      "step": 6935
+    },
+    {
+      "epoch": 0.5633070738244133,
+      "grad_norm": 6.623608750925487,
+      "learning_rate": 2.11063517552796e-06,
+      "loss": 0.4761,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5633882888004548,
+      "grad_norm": 4.043561752629549,
+      "learning_rate": 2.1099855916977676e-06,
+      "loss": 0.4361,
+      "step": 6937
+    },
+    {
+      "epoch": 0.5634695037764964,
+      "grad_norm": 3.492107617173054,
+      "learning_rate": 2.109336034854508e-06,
+      "loss": 0.6809,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5635507187525379,
+      "grad_norm": 5.163084090773521,
+      "learning_rate": 2.1086865050431283e-06,
+      "loss": 0.4603,
+      "step": 6939
+    },
+    {
+      "epoch": 0.5636319337285796,
+      "grad_norm": 4.691574509899781,
+      "learning_rate": 2.1080370023085713e-06,
+      "loss": 0.6142,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5637131487046211,
+      "grad_norm": 4.128014703538365,
+      "learning_rate": 2.107387526695778e-06,
+      "loss": 0.4824,
+      "step": 6941
+    },
+    {
+      "epoch": 0.5637943636806627,
+      "grad_norm": 11.05696154296604,
+      "learning_rate": 2.106738078249691e-06,
+      "loss": 0.5197,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5638755786567043,
+      "grad_norm": 5.222620839074155,
+      "learning_rate": 2.1060886570152477e-06,
+      "loss": 0.4981,
+      "step": 6943
+    },
+    {
+      "epoch": 0.5639567936327459,
+      "grad_norm": 4.697414322539678,
+      "learning_rate": 2.105439263037384e-06,
+      "loss": 0.4387,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5640380086087875,
+      "grad_norm": 3.8406258019033004,
+      "learning_rate": 2.1047898963610354e-06,
+      "loss": 0.4677,
+      "step": 6945
+    },
+    {
+      "epoch": 0.564119223584829,
+      "grad_norm": 5.2714849848498115,
+      "learning_rate": 2.1041405570311348e-06,
+      "loss": 0.3368,
+      "step": 6946
+    },
+    {
+      "epoch": 0.5642004385608707,
+      "grad_norm": 7.429954267821541,
+      "learning_rate": 2.1034912450926114e-06,
+      "loss": 0.4286,
+      "step": 6947
+    },
+    {
+      "epoch": 0.5642816535369122,
+      "grad_norm": 4.461747197120303,
+      "learning_rate": 2.102841960590396e-06,
+      "loss": 0.4246,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5643628685129538,
+      "grad_norm": 6.852938483133885,
+      "learning_rate": 2.102192703569416e-06,
+      "loss": 0.4819,
+      "step": 6949
+    },
+    {
+      "epoch": 0.5644440834889953,
+      "grad_norm": 4.945485942357694,
+      "learning_rate": 2.1015434740745944e-06,
+      "loss": 0.5224,
+      "step": 6950
+    },
+    {
+      "epoch": 0.564525298465037,
+      "grad_norm": 5.1220985098046095,
+      "learning_rate": 2.1008942721508553e-06,
+      "loss": 0.4534,
+      "step": 6951
+    },
+    {
+      "epoch": 0.5646065134410785,
+      "grad_norm": 4.645911138798781,
+      "learning_rate": 2.1002450978431216e-06,
+      "loss": 0.4874,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5646877284171201,
+      "grad_norm": 4.138063372246911,
+      "learning_rate": 2.099595951196311e-06,
+      "loss": 0.4335,
+      "step": 6953
+    },
+    {
+      "epoch": 0.5647689433931617,
+      "grad_norm": 5.760613534796062,
+      "learning_rate": 2.09894683225534e-06,
+      "loss": 0.5575,
+      "step": 6954
+    },
+    {
+      "epoch": 0.5648501583692033,
+      "grad_norm": 5.719720985157515,
+      "learning_rate": 2.0982977410651276e-06,
+      "loss": 0.5573,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5649313733452449,
+      "grad_norm": 5.21317528533535,
+      "learning_rate": 2.0976486776705853e-06,
+      "loss": 0.4523,
+      "step": 6956
+    },
+    {
+      "epoch": 0.5650125883212864,
+      "grad_norm": 6.455893264664072,
+      "learning_rate": 2.0969996421166243e-06,
+      "loss": 0.5827,
+      "step": 6957
+    },
+    {
+      "epoch": 0.5650938032973281,
+      "grad_norm": 3.152544873395565,
+      "learning_rate": 2.0963506344481556e-06,
+      "loss": 0.4638,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5651750182733696,
+      "grad_norm": 3.711902575933437,
+      "learning_rate": 2.0957016547100867e-06,
+      "loss": 0.5174,
+      "step": 6959
+    },
+    {
+      "epoch": 0.5652562332494112,
+      "grad_norm": 3.1937905928416774,
+      "learning_rate": 2.095052702947323e-06,
+      "loss": 0.5568,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5653374482254527,
+      "grad_norm": 3.8506487142937336,
+      "learning_rate": 2.09440377920477e-06,
+      "loss": 0.6765,
+      "step": 6961
+    },
+    {
+      "epoch": 0.5654186632014944,
+      "grad_norm": 4.161307544084725,
+      "learning_rate": 2.0937548835273285e-06,
+      "loss": 0.4279,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5654998781775359,
+      "grad_norm": 5.783786158584384,
+      "learning_rate": 2.0931060159598986e-06,
+      "loss": 0.483,
+      "step": 6963
+    },
+    {
+      "epoch": 0.5655810931535775,
+      "grad_norm": 4.391669807993531,
+      "learning_rate": 2.0924571765473793e-06,
+      "loss": 0.5543,
+      "step": 6964
+    },
+    {
+      "epoch": 0.5656623081296192,
+      "grad_norm": 11.934622866843036,
+      "learning_rate": 2.091808365334667e-06,
+      "loss": 0.5861,
+      "step": 6965
+    },
+    {
+      "epoch": 0.5657435231056607,
+      "grad_norm": 6.272947824575495,
+      "learning_rate": 2.091159582366655e-06,
+      "loss": 0.4361,
+      "step": 6966
+    },
+    {
+      "epoch": 0.5658247380817023,
+      "grad_norm": 2.896511468026234,
+      "learning_rate": 2.0905108276882356e-06,
+      "loss": 0.4904,
+      "step": 6967
+    },
+    {
+      "epoch": 0.5659059530577438,
+      "grad_norm": 4.7888166758801205,
+      "learning_rate": 2.089862101344301e-06,
+      "loss": 0.4592,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5659871680337855,
+      "grad_norm": 10.0912728353721,
+      "learning_rate": 2.0892134033797383e-06,
+      "loss": 0.4861,
+      "step": 6969
+    },
+    {
+      "epoch": 0.566068383009827,
+      "grad_norm": 2.573138862813564,
+      "learning_rate": 2.088564733839433e-06,
+      "loss": 0.3952,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5661495979858686,
+      "grad_norm": 2.924408214469329,
+      "learning_rate": 2.087916092768271e-06,
+      "loss": 0.4989,
+      "step": 6971
+    },
+    {
+      "epoch": 0.5662308129619101,
+      "grad_norm": 5.175327347055258,
+      "learning_rate": 2.087267480211135e-06,
+      "loss": 0.593,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5663120279379518,
+      "grad_norm": 5.327737397653494,
+      "learning_rate": 2.086618896212904e-06,
+      "loss": 0.4824,
+      "step": 6973
+    },
+    {
+      "epoch": 0.5663932429139933,
+      "grad_norm": 3.5981451480047344,
+      "learning_rate": 2.0859703408184583e-06,
+      "loss": 0.569,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5664744578900349,
+      "grad_norm": 3.98952188483392,
+      "learning_rate": 2.085321814072674e-06,
+      "loss": 0.5055,
+      "step": 6975
+    },
+    {
+      "epoch": 0.5665556728660766,
+      "grad_norm": 4.760263855647576,
+      "learning_rate": 2.0846733160204244e-06,
+      "loss": 0.5048,
+      "step": 6976
+    },
+    {
+      "epoch": 0.5666368878421181,
+      "grad_norm": 4.040158060702142,
+      "learning_rate": 2.084024846706584e-06,
+      "loss": 0.474,
+      "step": 6977
+    },
+    {
+      "epoch": 0.5667181028181597,
+      "grad_norm": 3.3443500978774394,
+      "learning_rate": 2.083376406176023e-06,
+      "loss": 0.6342,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5667993177942012,
+      "grad_norm": 6.809697122943693,
+      "learning_rate": 2.082727994473609e-06,
+      "loss": 0.4007,
+      "step": 6979
+    },
+    {
+      "epoch": 0.5668805327702429,
+      "grad_norm": 4.007751918991223,
+      "learning_rate": 2.08207961164421e-06,
+      "loss": 0.3879,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5669617477462844,
+      "grad_norm": 4.107776861520255,
+      "learning_rate": 2.08143125773269e-06,
+      "loss": 0.4471,
+      "step": 6981
+    },
+    {
+      "epoch": 0.567042962722326,
+      "grad_norm": 3.806832164663562,
+      "learning_rate": 2.080782932783911e-06,
+      "loss": 0.4513,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5671241776983675,
+      "grad_norm": 6.942546340030462,
+      "learning_rate": 2.0801346368427356e-06,
+      "loss": 0.6031,
+      "step": 6983
+    },
+    {
+      "epoch": 0.5672053926744092,
+      "grad_norm": 3.225972794340873,
+      "learning_rate": 2.0794863699540206e-06,
+      "loss": 0.6745,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5672866076504507,
+      "grad_norm": 4.7018910207675955,
+      "learning_rate": 2.0788381321626237e-06,
+      "loss": 0.531,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5673678226264923,
+      "grad_norm": 3.88377602984253,
+      "learning_rate": 2.0781899235133984e-06,
+      "loss": 0.6038,
+      "step": 6986
+    },
+    {
+      "epoch": 0.567449037602534,
+      "grad_norm": 4.743493354886736,
+      "learning_rate": 2.077541744051198e-06,
+      "loss": 0.5044,
+      "step": 6987
+    },
+    {
+      "epoch": 0.5675302525785755,
+      "grad_norm": 3.725682821827385,
+      "learning_rate": 2.0768935938208735e-06,
+      "loss": 0.545,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5676114675546171,
+      "grad_norm": 4.23366272227734,
+      "learning_rate": 2.0762454728672727e-06,
+      "loss": 0.6513,
+      "step": 6989
+    },
+    {
+      "epoch": 0.5676926825306586,
+      "grad_norm": 3.219731697269269,
+      "learning_rate": 2.0755973812352424e-06,
+      "loss": 0.5489,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5677738975067003,
+      "grad_norm": 5.426598112663618,
+      "learning_rate": 2.074949318969628e-06,
+      "loss": 0.5292,
+      "step": 6991
+    },
+    {
+      "epoch": 0.5678551124827418,
+      "grad_norm": 5.501771039935805,
+      "learning_rate": 2.07430128611527e-06,
+      "loss": 0.4586,
+      "step": 6992
+    },
+    {
+      "epoch": 0.5679363274587834,
+      "grad_norm": 4.038443635513911,
+      "learning_rate": 2.0736532827170107e-06,
+      "loss": 0.414,
+      "step": 6993
+    },
+    {
+      "epoch": 0.5680175424348249,
+      "grad_norm": 5.939660501414922,
+      "learning_rate": 2.0730053088196883e-06,
+      "loss": 0.4727,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5680987574108666,
+      "grad_norm": 4.71954105859692,
+      "learning_rate": 2.072357364468138e-06,
+      "loss": 0.4108,
+      "step": 6995
+    },
+    {
+      "epoch": 0.5681799723869081,
+      "grad_norm": 12.024015711126426,
+      "learning_rate": 2.0717094497071945e-06,
+      "loss": 0.4805,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5682611873629497,
+      "grad_norm": 4.1133512422429455,
+      "learning_rate": 2.0710615645816913e-06,
+      "loss": 0.4311,
+      "step": 6997
+    },
+    {
+      "epoch": 0.5683424023389914,
+      "grad_norm": 3.993669100784519,
+      "learning_rate": 2.0704137091364568e-06,
+      "loss": 0.5012,
+      "step": 6998
+    },
+    {
+      "epoch": 0.5684236173150329,
+      "grad_norm": 5.302686271074888,
+      "learning_rate": 2.069765883416321e-06,
+      "loss": 0.43,
+      "step": 6999
+    },
+    {
+      "epoch": 0.5685048322910745,
+      "grad_norm": 4.755390559997945,
+      "learning_rate": 2.0691180874661086e-06,
+      "loss": 0.5289,
+      "step": 7000
+    },
+    {
+      "epoch": 0.568586047267116,
+      "grad_norm": 2.9673265351316704,
+      "learning_rate": 2.0684703213306435e-06,
+      "loss": 0.4691,
+      "step": 7001
+    },
+    {
+      "epoch": 0.5686672622431577,
+      "grad_norm": 3.9211031513326486,
+      "learning_rate": 2.0678225850547497e-06,
+      "loss": 0.5234,
+      "step": 7002
+    },
+    {
+      "epoch": 0.5687484772191992,
+      "grad_norm": 4.581102578387739,
+      "learning_rate": 2.0671748786832447e-06,
+      "loss": 0.5117,
+      "step": 7003
+    },
+    {
+      "epoch": 0.5688296921952408,
+      "grad_norm": 6.219007379440202,
+      "learning_rate": 2.0665272022609482e-06,
+      "loss": 0.5086,
+      "step": 7004
+    },
+    {
+      "epoch": 0.5689109071712823,
+      "grad_norm": 6.0686864713771485,
+      "learning_rate": 2.0658795558326745e-06,
+      "loss": 0.4348,
+      "step": 7005
+    },
+    {
+      "epoch": 0.568992122147324,
+      "grad_norm": 3.6037676191639885,
+      "learning_rate": 2.065231939443238e-06,
+      "loss": 0.4322,
+      "step": 7006
+    },
+    {
+      "epoch": 0.5690733371233655,
+      "grad_norm": 3.160547290178154,
+      "learning_rate": 2.064584353137451e-06,
+      "loss": 0.4967,
+      "step": 7007
+    },
+    {
+      "epoch": 0.5691545520994071,
+      "grad_norm": 4.373900459536156,
+      "learning_rate": 2.0639367969601215e-06,
+      "loss": 0.4513,
+      "step": 7008
+    },
+    {
+      "epoch": 0.5692357670754488,
+      "grad_norm": 10.5295783218965,
+      "learning_rate": 2.063289270956058e-06,
+      "loss": 0.4657,
+      "step": 7009
+    },
+    {
+      "epoch": 0.5693169820514903,
+      "grad_norm": 6.441791257822959,
+      "learning_rate": 2.0626417751700664e-06,
+      "loss": 0.3973,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5693981970275319,
+      "grad_norm": 7.65548642747038,
+      "learning_rate": 2.0619943096469484e-06,
+      "loss": 0.4463,
+      "step": 7011
+    },
+    {
+      "epoch": 0.5694794120035734,
+      "grad_norm": 2.7981688902513864,
+      "learning_rate": 2.061346874431507e-06,
+      "loss": 0.5198,
+      "step": 7012
+    },
+    {
+      "epoch": 0.5695606269796151,
+      "grad_norm": 3.8831757918226586,
+      "learning_rate": 2.0606994695685396e-06,
+      "loss": 0.3686,
+      "step": 7013
+    },
+    {
+      "epoch": 0.5696418419556566,
+      "grad_norm": 3.8701501681776156,
+      "learning_rate": 2.0600520951028437e-06,
+      "loss": 0.5688,
+      "step": 7014
+    },
+    {
+      "epoch": 0.5697230569316982,
+      "grad_norm": 3.537194498397044,
+      "learning_rate": 2.059404751079215e-06,
+      "loss": 0.4693,
+      "step": 7015
+    },
+    {
+      "epoch": 0.5698042719077397,
+      "grad_norm": 4.5238431030094794,
+      "learning_rate": 2.0587574375424456e-06,
+      "loss": 0.4932,
+      "step": 7016
+    },
+    {
+      "epoch": 0.5698854868837814,
+      "grad_norm": 6.314721658579622,
+      "learning_rate": 2.0581101545373255e-06,
+      "loss": 0.5723,
+      "step": 7017
+    },
+    {
+      "epoch": 0.5699667018598229,
+      "grad_norm": 4.543665245055623,
+      "learning_rate": 2.057462902108645e-06,
+      "loss": 0.6108,
+      "step": 7018
+    },
+    {
+      "epoch": 0.5700479168358645,
+      "grad_norm": 4.326213404795697,
+      "learning_rate": 2.0568156803011897e-06,
+      "loss": 0.5064,
+      "step": 7019
+    },
+    {
+      "epoch": 0.5701291318119062,
+      "grad_norm": 6.344731487919164,
+      "learning_rate": 2.056168489159744e-06,
+      "loss": 0.4286,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5702103467879477,
+      "grad_norm": 6.407501842611541,
+      "learning_rate": 2.0555213287290886e-06,
+      "loss": 0.4281,
+      "step": 7021
+    },
+    {
+      "epoch": 0.5702915617639893,
+      "grad_norm": 3.9016889868541873,
+      "learning_rate": 2.0548741990540057e-06,
+      "loss": 0.5229,
+      "step": 7022
+    },
+    {
+      "epoch": 0.5703727767400308,
+      "grad_norm": 4.485329708138932,
+      "learning_rate": 2.0542271001792726e-06,
+      "loss": 0.4988,
+      "step": 7023
+    },
+    {
+      "epoch": 0.5704539917160725,
+      "grad_norm": 7.011271727948778,
+      "learning_rate": 2.0535800321496645e-06,
+      "loss": 0.4174,
+      "step": 7024
+    },
+    {
+      "epoch": 0.570535206692114,
+      "grad_norm": 21.104012028716472,
+      "learning_rate": 2.0529329950099554e-06,
+      "loss": 0.566,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5706164216681556,
+      "grad_norm": 5.5743586700224,
+      "learning_rate": 2.052285988804918e-06,
+      "loss": 0.4475,
+      "step": 7026
+    },
+    {
+      "epoch": 0.5706976366441971,
+      "grad_norm": 4.416468163895723,
+      "learning_rate": 2.0516390135793192e-06,
+      "loss": 0.3382,
+      "step": 7027
+    },
+    {
+      "epoch": 0.5707788516202388,
+      "grad_norm": 4.986468294717369,
+      "learning_rate": 2.050992069377929e-06,
+      "loss": 0.3999,
+      "step": 7028
+    },
+    {
+      "epoch": 0.5708600665962803,
+      "grad_norm": 9.4114543685177,
+      "learning_rate": 2.050345156245511e-06,
+      "loss": 0.4653,
+      "step": 7029
+    },
+    {
+      "epoch": 0.5709412815723219,
+      "grad_norm": 4.613503552061298,
+      "learning_rate": 2.0496982742268273e-06,
+      "loss": 0.3624,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5710224965483636,
+      "grad_norm": 6.062105080879825,
+      "learning_rate": 2.0490514233666413e-06,
+      "loss": 0.5138,
+      "step": 7031
+    },
+    {
+      "epoch": 0.5711037115244051,
+      "grad_norm": 3.9841008788202132,
+      "learning_rate": 2.04840460370971e-06,
+      "loss": 0.4635,
+      "step": 7032
+    },
+    {
+      "epoch": 0.5711849265004467,
+      "grad_norm": 5.48691904012669,
+      "learning_rate": 2.0477578153007887e-06,
+      "loss": 0.4516,
+      "step": 7033
+    },
+    {
+      "epoch": 0.5712661414764882,
+      "grad_norm": 3.3836486599021636,
+      "learning_rate": 2.047111058184635e-06,
+      "loss": 0.5242,
+      "step": 7034
+    },
+    {
+      "epoch": 0.5713473564525299,
+      "grad_norm": 3.8672222024944856,
+      "learning_rate": 2.046464332405998e-06,
+      "loss": 0.5653,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 4.314382703074949,
+      "learning_rate": 2.045817638009629e-06,
+      "loss": 0.5402,
+      "step": 7036
+    },
+    {
+      "epoch": 0.571509786404613,
+      "grad_norm": 3.567072195815995,
+      "learning_rate": 2.045170975040276e-06,
+      "loss": 0.4213,
+      "step": 7037
+    },
+    {
+      "epoch": 0.5715910013806546,
+      "grad_norm": 4.48375869912088,
+      "learning_rate": 2.0445243435426847e-06,
+      "loss": 0.6564,
+      "step": 7038
+    },
+    {
+      "epoch": 0.5716722163566962,
+      "grad_norm": 3.552656399851541,
+      "learning_rate": 2.043877743561598e-06,
+      "loss": 0.5135,
+      "step": 7039
+    },
+    {
+      "epoch": 0.5717534313327377,
+      "grad_norm": 3.8377915597339873,
+      "learning_rate": 2.0432311751417568e-06,
+      "loss": 0.6527,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5718346463087793,
+      "grad_norm": 6.426418479233554,
+      "learning_rate": 2.042584638327902e-06,
+      "loss": 0.3723,
+      "step": 7041
+    },
+    {
+      "epoch": 0.571915861284821,
+      "grad_norm": 4.241032012671777,
+      "learning_rate": 2.0419381331647687e-06,
+      "loss": 0.4957,
+      "step": 7042
+    },
+    {
+      "epoch": 0.5719970762608625,
+      "grad_norm": 3.735539966685467,
+      "learning_rate": 2.0412916596970918e-06,
+      "loss": 0.566,
+      "step": 7043
+    },
+    {
+      "epoch": 0.5720782912369041,
+      "grad_norm": 3.8956911217710917,
+      "learning_rate": 2.040645217969606e-06,
+      "loss": 0.4935,
+      "step": 7044
+    },
+    {
+      "epoch": 0.5721595062129456,
+      "grad_norm": 20.493625970905985,
+      "learning_rate": 2.0399988080270384e-06,
+      "loss": 0.4629,
+      "step": 7045
+    },
+    {
+      "epoch": 0.5722407211889873,
+      "grad_norm": 29.937743604580216,
+      "learning_rate": 2.039352429914119e-06,
+      "loss": 0.4411,
+      "step": 7046
+    },
+    {
+      "epoch": 0.5723219361650288,
+      "grad_norm": 6.109372993647007,
+      "learning_rate": 2.038706083675574e-06,
+      "loss": 0.4187,
+      "step": 7047
+    },
+    {
+      "epoch": 0.5724031511410704,
+      "grad_norm": 4.3015816031863086,
+      "learning_rate": 2.038059769356127e-06,
+      "loss": 0.5831,
+      "step": 7048
+    },
+    {
+      "epoch": 0.572484366117112,
+      "grad_norm": 4.347753553845655,
+      "learning_rate": 2.037413487000498e-06,
+      "loss": 0.4969,
+      "step": 7049
+    },
+    {
+      "epoch": 0.5725655810931536,
+      "grad_norm": 5.01369388436399,
+      "learning_rate": 2.0367672366534087e-06,
+      "loss": 0.4838,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5726467960691951,
+      "grad_norm": 6.956995413047726,
+      "learning_rate": 2.036121018359574e-06,
+      "loss": 0.4995,
+      "step": 7051
+    },
+    {
+      "epoch": 0.5727280110452367,
+      "grad_norm": 7.014322634335082,
+      "learning_rate": 2.03547483216371e-06,
+      "loss": 0.5013,
+      "step": 7052
+    },
+    {
+      "epoch": 0.5728092260212784,
+      "grad_norm": 6.622483650279607,
+      "learning_rate": 2.0348286781105302e-06,
+      "loss": 0.4124,
+      "step": 7053
+    },
+    {
+      "epoch": 0.5728904409973199,
+      "grad_norm": 4.159914158121862,
+      "learning_rate": 2.0341825562447427e-06,
+      "loss": 0.6059,
+      "step": 7054
+    },
+    {
+      "epoch": 0.5729716559733615,
+      "grad_norm": 5.011801450523911,
+      "learning_rate": 2.0335364666110572e-06,
+      "loss": 0.4856,
+      "step": 7055
+    },
+    {
+      "epoch": 0.573052870949403,
+      "grad_norm": 4.952363770440026,
+      "learning_rate": 2.03289040925418e-06,
+      "loss": 0.4473,
+      "step": 7056
+    },
+    {
+      "epoch": 0.5731340859254447,
+      "grad_norm": 7.83403745870332,
+      "learning_rate": 2.032244384218815e-06,
+      "loss": 0.6729,
+      "step": 7057
+    },
+    {
+      "epoch": 0.5732153009014862,
+      "grad_norm": 6.993413026032686,
+      "learning_rate": 2.031598391549662e-06,
+      "loss": 0.5622,
+      "step": 7058
+    },
+    {
+      "epoch": 0.5732965158775278,
+      "grad_norm": 6.260135249773573,
+      "learning_rate": 2.030952431291421e-06,
+      "loss": 0.5174,
+      "step": 7059
+    },
+    {
+      "epoch": 0.5733777308535694,
+      "grad_norm": 4.075720359030366,
+      "learning_rate": 2.0303065034887904e-06,
+      "loss": 0.5199,
+      "step": 7060
+    },
+    {
+      "epoch": 0.573458945829611,
+      "grad_norm": 3.9202159123881475,
+      "learning_rate": 2.0296606081864634e-06,
+      "loss": 0.5269,
+      "step": 7061
+    },
+    {
+      "epoch": 0.5735401608056525,
+      "grad_norm": 4.8268240982250274,
+      "learning_rate": 2.0290147454291323e-06,
+      "loss": 0.4606,
+      "step": 7062
+    },
+    {
+      "epoch": 0.5736213757816941,
+      "grad_norm": 5.603064677109073,
+      "learning_rate": 2.0283689152614896e-06,
+      "loss": 0.6205,
+      "step": 7063
+    },
+    {
+      "epoch": 0.5737025907577358,
+      "grad_norm": 4.633814523131622,
+      "learning_rate": 2.0277231177282213e-06,
+      "loss": 0.5806,
+      "step": 7064
+    },
+    {
+      "epoch": 0.5737838057337773,
+      "grad_norm": 11.093950775524117,
+      "learning_rate": 2.0270773528740127e-06,
+      "loss": 0.5424,
+      "step": 7065
+    },
+    {
+      "epoch": 0.5738650207098189,
+      "grad_norm": 4.963416503789594,
+      "learning_rate": 2.02643162074355e-06,
+      "loss": 0.5403,
+      "step": 7066
+    },
+    {
+      "epoch": 0.5739462356858605,
+      "grad_norm": 7.269597973886369,
+      "learning_rate": 2.0257859213815123e-06,
+      "loss": 0.649,
+      "step": 7067
+    },
+    {
+      "epoch": 0.5740274506619021,
+      "grad_norm": 4.685273966619034,
+      "learning_rate": 2.0251402548325783e-06,
+      "loss": 0.6087,
+      "step": 7068
+    },
+    {
+      "epoch": 0.5741086656379436,
+      "grad_norm": 4.846673111295444,
+      "learning_rate": 2.0244946211414267e-06,
+      "loss": 0.5899,
+      "step": 7069
+    },
+    {
+      "epoch": 0.5741898806139852,
+      "grad_norm": 3.9278515943561625,
+      "learning_rate": 2.0238490203527307e-06,
+      "loss": 0.5397,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5742710955900268,
+      "grad_norm": 4.034403858762683,
+      "learning_rate": 2.0232034525111617e-06,
+      "loss": 0.5535,
+      "step": 7071
+    },
+    {
+      "epoch": 0.5743523105660684,
+      "grad_norm": 6.517927780743827,
+      "learning_rate": 2.0225579176613905e-06,
+      "loss": 0.4494,
+      "step": 7072
+    },
+    {
+      "epoch": 0.5744335255421099,
+      "grad_norm": 3.068515611615937,
+      "learning_rate": 2.0219124158480853e-06,
+      "loss": 0.3643,
+      "step": 7073
+    },
+    {
+      "epoch": 0.5745147405181515,
+      "grad_norm": 3.6556023426111084,
+      "learning_rate": 2.0212669471159098e-06,
+      "loss": 0.4402,
+      "step": 7074
+    },
+    {
+      "epoch": 0.5745959554941932,
+      "grad_norm": 3.202736868542401,
+      "learning_rate": 2.020621511509528e-06,
+      "loss": 0.5183,
+      "step": 7075
+    },
+    {
+      "epoch": 0.5746771704702347,
+      "grad_norm": 5.842839385752603,
+      "learning_rate": 2.019976109073601e-06,
+      "loss": 0.477,
+      "step": 7076
+    },
+    {
+      "epoch": 0.5747583854462763,
+      "grad_norm": 12.847781405058441,
+      "learning_rate": 2.0193307398527865e-06,
+      "loss": 0.4842,
+      "step": 7077
+    },
+    {
+      "epoch": 0.5748396004223179,
+      "grad_norm": 21.13344792621275,
+      "learning_rate": 2.0186854038917405e-06,
+      "loss": 0.4538,
+      "step": 7078
+    },
+    {
+      "epoch": 0.5749208153983595,
+      "grad_norm": 10.50726009998217,
+      "learning_rate": 2.0180401012351182e-06,
+      "loss": 0.4578,
+      "step": 7079
+    },
+    {
+      "epoch": 0.575002030374401,
+      "grad_norm": 5.987355841136969,
+      "learning_rate": 2.0173948319275696e-06,
+      "loss": 0.5223,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5750832453504426,
+      "grad_norm": 5.722234729191682,
+      "learning_rate": 2.016749596013744e-06,
+      "loss": 0.4965,
+      "step": 7081
+    },
+    {
+      "epoch": 0.5751644603264842,
+      "grad_norm": 4.664393900667678,
+      "learning_rate": 2.0161043935382897e-06,
+      "loss": 0.4677,
+      "step": 7082
+    },
+    {
+      "epoch": 0.5752456753025258,
+      "grad_norm": 7.498350619883631,
+      "learning_rate": 2.0154592245458504e-06,
+      "loss": 0.5619,
+      "step": 7083
+    },
+    {
+      "epoch": 0.5753268902785673,
+      "grad_norm": 5.775102597864225,
+      "learning_rate": 2.014814089081067e-06,
+      "loss": 0.5719,
+      "step": 7084
+    },
+    {
+      "epoch": 0.575408105254609,
+      "grad_norm": 5.1155695309255025,
+      "learning_rate": 2.014168987188582e-06,
+      "loss": 0.7247,
+      "step": 7085
+    },
+    {
+      "epoch": 0.5754893202306506,
+      "grad_norm": 4.2278332646738175,
+      "learning_rate": 2.0135239189130325e-06,
+      "loss": 0.6624,
+      "step": 7086
+    },
+    {
+      "epoch": 0.5755705352066921,
+      "grad_norm": 3.7678111182822334,
+      "learning_rate": 2.0128788842990516e-06,
+      "loss": 0.7632,
+      "step": 7087
+    },
+    {
+      "epoch": 0.5756517501827337,
+      "grad_norm": 11.009075470970467,
+      "learning_rate": 2.0122338833912743e-06,
+      "loss": 0.4833,
+      "step": 7088
+    },
+    {
+      "epoch": 0.5757329651587753,
+      "grad_norm": 4.5002280122486535,
+      "learning_rate": 2.0115889162343316e-06,
+      "loss": 0.4961,
+      "step": 7089
+    },
+    {
+      "epoch": 0.5758141801348169,
+      "grad_norm": 4.760191730901538,
+      "learning_rate": 2.01094398287285e-06,
+      "loss": 0.475,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5758953951108584,
+      "grad_norm": 3.118348188458208,
+      "learning_rate": 2.010299083351457e-06,
+      "loss": 0.353,
+      "step": 7091
+    },
+    {
+      "epoch": 0.5759766100869,
+      "grad_norm": 6.734823400251969,
+      "learning_rate": 2.009654217714776e-06,
+      "loss": 0.5105,
+      "step": 7092
+    },
+    {
+      "epoch": 0.5760578250629416,
+      "grad_norm": 5.003848469090975,
+      "learning_rate": 2.0090093860074273e-06,
+      "loss": 0.5659,
+      "step": 7093
+    },
+    {
+      "epoch": 0.5761390400389832,
+      "grad_norm": 3.690892599491099,
+      "learning_rate": 2.008364588274031e-06,
+      "loss": 0.5335,
+      "step": 7094
+    },
+    {
+      "epoch": 0.5762202550150247,
+      "grad_norm": 5.097697074216427,
+      "learning_rate": 2.0077198245592033e-06,
+      "loss": 0.5175,
+      "step": 7095
+    },
+    {
+      "epoch": 0.5763014699910664,
+      "grad_norm": 3.9972301907638133,
+      "learning_rate": 2.0070750949075584e-06,
+      "loss": 0.5674,
+      "step": 7096
+    },
+    {
+      "epoch": 0.576382684967108,
+      "grad_norm": 3.765506245804216,
+      "learning_rate": 2.0064303993637073e-06,
+      "loss": 0.4085,
+      "step": 7097
+    },
+    {
+      "epoch": 0.5764638999431495,
+      "grad_norm": 5.010605186017302,
+      "learning_rate": 2.005785737972262e-06,
+      "loss": 0.5775,
+      "step": 7098
+    },
+    {
+      "epoch": 0.5765451149191911,
+      "grad_norm": 3.786228217501147,
+      "learning_rate": 2.0051411107778273e-06,
+      "loss": 0.5185,
+      "step": 7099
+    },
+    {
+      "epoch": 0.5766263298952327,
+      "grad_norm": 9.0650925529505,
+      "learning_rate": 2.004496517825008e-06,
+      "loss": 0.4735,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5767075448712743,
+      "grad_norm": 5.189793367646856,
+      "learning_rate": 2.0038519591584078e-06,
+      "loss": 0.5995,
+      "step": 7101
+    },
+    {
+      "epoch": 0.5767887598473158,
+      "grad_norm": 4.700410500709092,
+      "learning_rate": 2.0032074348226268e-06,
+      "loss": 0.5761,
+      "step": 7102
+    },
+    {
+      "epoch": 0.5768699748233574,
+      "grad_norm": 4.234749790705928,
+      "learning_rate": 2.002562944862261e-06,
+      "loss": 0.5744,
+      "step": 7103
+    },
+    {
+      "epoch": 0.576951189799399,
+      "grad_norm": 3.629545684362975,
+      "learning_rate": 2.0019184893219076e-06,
+      "loss": 0.5361,
+      "step": 7104
+    },
+    {
+      "epoch": 0.5770324047754406,
+      "grad_norm": 5.227066105381711,
+      "learning_rate": 2.0012740682461585e-06,
+      "loss": 0.424,
+      "step": 7105
+    },
+    {
+      "epoch": 0.5771136197514821,
+      "grad_norm": 8.03973120733085,
+      "learning_rate": 2.0006296816796037e-06,
+      "loss": 0.4179,
+      "step": 7106
+    },
+    {
+      "epoch": 0.5771948347275238,
+      "grad_norm": 3.3013104736748575,
+      "learning_rate": 1.9999853296668326e-06,
+      "loss": 0.657,
+      "step": 7107
+    },
+    {
+      "epoch": 0.5772760497035654,
+      "grad_norm": 4.319311268255759,
+      "learning_rate": 1.999341012252431e-06,
+      "loss": 0.4993,
+      "step": 7108
+    },
+    {
+      "epoch": 0.5773572646796069,
+      "grad_norm": 3.6319991209865945,
+      "learning_rate": 1.9986967294809804e-06,
+      "loss": 0.474,
+      "step": 7109
+    },
+    {
+      "epoch": 0.5774384796556485,
+      "grad_norm": 4.960374042694389,
+      "learning_rate": 1.9980524813970635e-06,
+      "loss": 0.5666,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5775196946316901,
+      "grad_norm": 2.9962262052676314,
+      "learning_rate": 1.997408268045259e-06,
+      "loss": 0.6228,
+      "step": 7111
+    },
+    {
+      "epoch": 0.5776009096077317,
+      "grad_norm": 5.668495286840983,
+      "learning_rate": 1.9967640894701424e-06,
+      "loss": 0.5147,
+      "step": 7112
+    },
+    {
+      "epoch": 0.5776821245837732,
+      "grad_norm": 4.163547326343956,
+      "learning_rate": 1.9961199457162867e-06,
+      "loss": 0.4194,
+      "step": 7113
+    },
+    {
+      "epoch": 0.5777633395598148,
+      "grad_norm": 3.4932620061985555,
+      "learning_rate": 1.995475836828264e-06,
+      "loss": 0.5275,
+      "step": 7114
+    },
+    {
+      "epoch": 0.5778445545358564,
+      "grad_norm": 4.368032739037839,
+      "learning_rate": 1.9948317628506444e-06,
+      "loss": 0.4277,
+      "step": 7115
+    },
+    {
+      "epoch": 0.577925769511898,
+      "grad_norm": 3.876730464227753,
+      "learning_rate": 1.994187723827992e-06,
+      "loss": 0.4987,
+      "step": 7116
+    },
+    {
+      "epoch": 0.5780069844879395,
+      "grad_norm": 3.2062830745638506,
+      "learning_rate": 1.9935437198048722e-06,
+      "loss": 0.5308,
+      "step": 7117
+    },
+    {
+      "epoch": 0.5780881994639812,
+      "grad_norm": 7.356711618544573,
+      "learning_rate": 1.9928997508258475e-06,
+      "loss": 0.464,
+      "step": 7118
+    },
+    {
+      "epoch": 0.5781694144400228,
+      "grad_norm": 5.484693237704757,
+      "learning_rate": 1.9922558169354752e-06,
+      "loss": 0.4484,
+      "step": 7119
+    },
+    {
+      "epoch": 0.5782506294160643,
+      "grad_norm": 6.681361631875744,
+      "learning_rate": 1.9916119181783135e-06,
+      "loss": 0.4322,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5783318443921059,
+      "grad_norm": 4.438508741670482,
+      "learning_rate": 1.9909680545989175e-06,
+      "loss": 0.567,
+      "step": 7121
+    },
+    {
+      "epoch": 0.5784130593681475,
+      "grad_norm": 5.678949037298267,
+      "learning_rate": 1.9903242262418366e-06,
+      "loss": 0.5764,
+      "step": 7122
+    },
+    {
+      "epoch": 0.5784942743441891,
+      "grad_norm": 4.194676378904089,
+      "learning_rate": 1.989680433151622e-06,
+      "loss": 0.7083,
+      "step": 7123
+    },
+    {
+      "epoch": 0.5785754893202306,
+      "grad_norm": 8.002991142166685,
+      "learning_rate": 1.989036675372822e-06,
+      "loss": 0.5588,
+      "step": 7124
+    },
+    {
+      "epoch": 0.5786567042962723,
+      "grad_norm": 3.764738047010275,
+      "learning_rate": 1.988392952949978e-06,
+      "loss": 0.5561,
+      "step": 7125
+    },
+    {
+      "epoch": 0.5787379192723138,
+      "grad_norm": 7.746991321514468,
+      "learning_rate": 1.9877492659276353e-06,
+      "loss": 0.473,
+      "step": 7126
+    },
+    {
+      "epoch": 0.5788191342483554,
+      "grad_norm": 5.541041864204196,
+      "learning_rate": 1.9871056143503322e-06,
+      "loss": 0.417,
+      "step": 7127
+    },
+    {
+      "epoch": 0.5789003492243969,
+      "grad_norm": 3.424180901190808,
+      "learning_rate": 1.9864619982626064e-06,
+      "loss": 0.743,
+      "step": 7128
+    },
+    {
+      "epoch": 0.5789815642004386,
+      "grad_norm": 16.31331645341726,
+      "learning_rate": 1.9858184177089915e-06,
+      "loss": 0.6825,
+      "step": 7129
+    },
+    {
+      "epoch": 0.5790627791764802,
+      "grad_norm": 3.593693975763408,
+      "learning_rate": 1.9851748727340214e-06,
+      "loss": 0.5404,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5791439941525217,
+      "grad_norm": 6.678160792934489,
+      "learning_rate": 1.9845313633822255e-06,
+      "loss": 0.5591,
+      "step": 7131
+    },
+    {
+      "epoch": 0.5792252091285633,
+      "grad_norm": 4.456993833787158,
+      "learning_rate": 1.9838878896981303e-06,
+      "loss": 0.5188,
+      "step": 7132
+    },
+    {
+      "epoch": 0.5793064241046049,
+      "grad_norm": 4.230178294928199,
+      "learning_rate": 1.9832444517262625e-06,
+      "loss": 0.3748,
+      "step": 7133
+    },
+    {
+      "epoch": 0.5793876390806465,
+      "grad_norm": 4.218687976196768,
+      "learning_rate": 1.982601049511144e-06,
+      "loss": 0.4318,
+      "step": 7134
+    },
+    {
+      "epoch": 0.579468854056688,
+      "grad_norm": 3.9051525132078386,
+      "learning_rate": 1.9819576830972938e-06,
+      "loss": 0.5105,
+      "step": 7135
+    },
+    {
+      "epoch": 0.5795500690327297,
+      "grad_norm": 5.953375128637914,
+      "learning_rate": 1.9813143525292304e-06,
+      "loss": 0.5147,
+      "step": 7136
+    },
+    {
+      "epoch": 0.5796312840087712,
+      "grad_norm": 5.299949856256123,
+      "learning_rate": 1.980671057851469e-06,
+      "loss": 0.5801,
+      "step": 7137
+    },
+    {
+      "epoch": 0.5797124989848128,
+      "grad_norm": 5.575356667344601,
+      "learning_rate": 1.9800277991085217e-06,
+      "loss": 0.484,
+      "step": 7138
+    },
+    {
+      "epoch": 0.5797937139608543,
+      "grad_norm": 4.75965558776345,
+      "learning_rate": 1.9793845763448987e-06,
+      "loss": 0.3816,
+      "step": 7139
+    },
+    {
+      "epoch": 0.579874928936896,
+      "grad_norm": 3.335625778162441,
+      "learning_rate": 1.9787413896051084e-06,
+      "loss": 0.4796,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5799561439129376,
+      "grad_norm": 4.108846663125449,
+      "learning_rate": 1.978098238933654e-06,
+      "loss": 0.4803,
+      "step": 7141
+    },
+    {
+      "epoch": 0.5800373588889791,
+      "grad_norm": 4.357715753076788,
+      "learning_rate": 1.9774551243750403e-06,
+      "loss": 0.3848,
+      "step": 7142
+    },
+    {
+      "epoch": 0.5801185738650207,
+      "grad_norm": 4.9420482058322435,
+      "learning_rate": 1.9768120459737663e-06,
+      "loss": 0.4834,
+      "step": 7143
+    },
+    {
+      "epoch": 0.5801997888410623,
+      "grad_norm": 3.610962949080674,
+      "learning_rate": 1.9761690037743293e-06,
+      "loss": 0.4072,
+      "step": 7144
+    },
+    {
+      "epoch": 0.5802810038171039,
+      "grad_norm": 4.466463892907367,
+      "learning_rate": 1.9755259978212253e-06,
+      "loss": 0.5826,
+      "step": 7145
+    },
+    {
+      "epoch": 0.5803622187931454,
+      "grad_norm": 5.2380939327824345,
+      "learning_rate": 1.9748830281589464e-06,
+      "loss": 0.6036,
+      "step": 7146
+    },
+    {
+      "epoch": 0.5804434337691871,
+      "grad_norm": 3.3553525016104704,
+      "learning_rate": 1.9742400948319838e-06,
+      "loss": 0.6842,
+      "step": 7147
+    },
+    {
+      "epoch": 0.5805246487452286,
+      "grad_norm": 5.736174896340501,
+      "learning_rate": 1.9735971978848224e-06,
+      "loss": 0.5493,
+      "step": 7148
+    },
+    {
+      "epoch": 0.5806058637212702,
+      "grad_norm": 3.703679606497901,
+      "learning_rate": 1.9729543373619497e-06,
+      "loss": 0.4979,
+      "step": 7149
+    },
+    {
+      "epoch": 0.5806870786973117,
+      "grad_norm": 9.728903722310404,
+      "learning_rate": 1.972311513307848e-06,
+      "loss": 0.45,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5807682936733534,
+      "grad_norm": 5.187888574328153,
+      "learning_rate": 1.971668725766996e-06,
+      "loss": 0.7753,
+      "step": 7151
+    },
+    {
+      "epoch": 0.580849508649395,
+      "grad_norm": 3.3139229056956396,
+      "learning_rate": 1.971025974783872e-06,
+      "loss": 0.675,
+      "step": 7152
+    },
+    {
+      "epoch": 0.5809307236254365,
+      "grad_norm": 9.118516211139692,
+      "learning_rate": 1.9703832604029523e-06,
+      "loss": 0.5264,
+      "step": 7153
+    },
+    {
+      "epoch": 0.5810119386014782,
+      "grad_norm": 4.415066688147641,
+      "learning_rate": 1.9697405826687063e-06,
+      "loss": 0.5034,
+      "step": 7154
+    },
+    {
+      "epoch": 0.5810931535775197,
+      "grad_norm": 5.533151171983695,
+      "learning_rate": 1.9690979416256062e-06,
+      "loss": 0.5316,
+      "step": 7155
+    },
+    {
+      "epoch": 0.5811743685535613,
+      "grad_norm": 3.4203284332256274,
+      "learning_rate": 1.9684553373181197e-06,
+      "loss": 0.4886,
+      "step": 7156
+    },
+    {
+      "epoch": 0.5812555835296028,
+      "grad_norm": 7.5906819086747435,
+      "learning_rate": 1.967812769790709e-06,
+      "loss": 0.5435,
+      "step": 7157
+    },
+    {
+      "epoch": 0.5813367985056445,
+      "grad_norm": 5.6092591367647815,
+      "learning_rate": 1.9671702390878396e-06,
+      "loss": 0.4718,
+      "step": 7158
+    },
+    {
+      "epoch": 0.581418013481686,
+      "grad_norm": 4.25266168413679,
+      "learning_rate": 1.9665277452539696e-06,
+      "loss": 0.5233,
+      "step": 7159
+    },
+    {
+      "epoch": 0.5814992284577276,
+      "grad_norm": 3.843541994359124,
+      "learning_rate": 1.965885288333555e-06,
+      "loss": 0.4787,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5815804434337691,
+      "grad_norm": 4.549147478757539,
+      "learning_rate": 1.965242868371053e-06,
+      "loss": 0.6732,
+      "step": 7161
+    },
+    {
+      "epoch": 0.5816616584098108,
+      "grad_norm": 7.3370587404827345,
+      "learning_rate": 1.9646004854109136e-06,
+      "loss": 0.3952,
+      "step": 7162
+    },
+    {
+      "epoch": 0.5817428733858524,
+      "grad_norm": 4.832489166222658,
+      "learning_rate": 1.963958139497588e-06,
+      "loss": 0.5612,
+      "step": 7163
+    },
+    {
+      "epoch": 0.5818240883618939,
+      "grad_norm": 4.612979610999288,
+      "learning_rate": 1.9633158306755206e-06,
+      "loss": 0.4501,
+      "step": 7164
+    },
+    {
+      "epoch": 0.5819053033379356,
+      "grad_norm": 3.842889295187542,
+      "learning_rate": 1.962673558989158e-06,
+      "loss": 0.479,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5819865183139771,
+      "grad_norm": 3.615074737738207,
+      "learning_rate": 1.9620313244829423e-06,
+      "loss": 0.5962,
+      "step": 7166
+    },
+    {
+      "epoch": 0.5820677332900187,
+      "grad_norm": 8.385548343766278,
+      "learning_rate": 1.961389127201311e-06,
+      "loss": 0.5366,
+      "step": 7167
+    },
+    {
+      "epoch": 0.5821489482660602,
+      "grad_norm": 4.0857765219914945,
+      "learning_rate": 1.9607469671887015e-06,
+      "loss": 0.3576,
+      "step": 7168
+    },
+    {
+      "epoch": 0.5822301632421019,
+      "grad_norm": 4.7701105433833995,
+      "learning_rate": 1.960104844489548e-06,
+      "loss": 0.5,
+      "step": 7169
+    },
+    {
+      "epoch": 0.5823113782181434,
+      "grad_norm": 4.079111419609443,
+      "learning_rate": 1.9594627591482817e-06,
+      "loss": 0.6516,
+      "step": 7170
+    },
+    {
+      "epoch": 0.582392593194185,
+      "grad_norm": 5.075085483225634,
+      "learning_rate": 1.9588207112093324e-06,
+      "loss": 0.5357,
+      "step": 7171
+    },
+    {
+      "epoch": 0.5824738081702265,
+      "grad_norm": 3.9201086563815624,
+      "learning_rate": 1.958178700717125e-06,
+      "loss": 0.4631,
+      "step": 7172
+    },
+    {
+      "epoch": 0.5825550231462682,
+      "grad_norm": 4.27616202832829,
+      "learning_rate": 1.957536727716084e-06,
+      "loss": 0.4648,
+      "step": 7173
+    },
+    {
+      "epoch": 0.5826362381223098,
+      "grad_norm": 4.882524062302524,
+      "learning_rate": 1.956894792250631e-06,
+      "loss": 0.5446,
+      "step": 7174
+    },
+    {
+      "epoch": 0.5827174530983513,
+      "grad_norm": 3.8190132907051435,
+      "learning_rate": 1.9562528943651837e-06,
+      "loss": 0.54,
+      "step": 7175
+    },
+    {
+      "epoch": 0.582798668074393,
+      "grad_norm": 6.140657105349779,
+      "learning_rate": 1.955611034104158e-06,
+      "loss": 0.4492,
+      "step": 7176
+    },
+    {
+      "epoch": 0.5828798830504345,
+      "grad_norm": 4.341900527053686,
+      "learning_rate": 1.9549692115119685e-06,
+      "loss": 0.5207,
+      "step": 7177
+    },
+    {
+      "epoch": 0.5829610980264761,
+      "grad_norm": 3.7483344786953685,
+      "learning_rate": 1.9543274266330244e-06,
+      "loss": 0.4006,
+      "step": 7178
+    },
+    {
+      "epoch": 0.5830423130025176,
+      "grad_norm": 3.3504281004368854,
+      "learning_rate": 1.9536856795117344e-06,
+      "loss": 0.5293,
+      "step": 7179
+    },
+    {
+      "epoch": 0.5831235279785593,
+      "grad_norm": 6.428753807471014,
+      "learning_rate": 1.9530439701925046e-06,
+      "loss": 0.6898,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5832047429546008,
+      "grad_norm": 9.364262509671228,
+      "learning_rate": 1.952402298719737e-06,
+      "loss": 0.3292,
+      "step": 7181
+    },
+    {
+      "epoch": 0.5832859579306424,
+      "grad_norm": 3.942276858330493,
+      "learning_rate": 1.951760665137832e-06,
+      "loss": 0.4671,
+      "step": 7182
+    },
+    {
+      "epoch": 0.5833671729066839,
+      "grad_norm": 6.118096625229616,
+      "learning_rate": 1.9511190694911875e-06,
+      "loss": 0.4731,
+      "step": 7183
+    },
+    {
+      "epoch": 0.5834483878827256,
+      "grad_norm": 4.555831582312394,
+      "learning_rate": 1.9504775118241987e-06,
+      "loss": 0.4638,
+      "step": 7184
+    },
+    {
+      "epoch": 0.5835296028587672,
+      "grad_norm": 3.4499740673047214,
+      "learning_rate": 1.9498359921812583e-06,
+      "loss": 0.6663,
+      "step": 7185
+    },
+    {
+      "epoch": 0.5836108178348087,
+      "grad_norm": 6.8178554531301705,
+      "learning_rate": 1.9491945106067544e-06,
+      "loss": 0.4266,
+      "step": 7186
+    },
+    {
+      "epoch": 0.5836920328108504,
+      "grad_norm": 4.578346378430701,
+      "learning_rate": 1.948553067145076e-06,
+      "loss": 0.5886,
+      "step": 7187
+    },
+    {
+      "epoch": 0.5837732477868919,
+      "grad_norm": 3.0536821725267718,
+      "learning_rate": 1.947911661840607e-06,
+      "loss": 0.3962,
+      "step": 7188
+    },
+    {
+      "epoch": 0.5838544627629335,
+      "grad_norm": 10.782933310564555,
+      "learning_rate": 1.947270294737728e-06,
+      "loss": 0.3495,
+      "step": 7189
+    },
+    {
+      "epoch": 0.583935677738975,
+      "grad_norm": 4.129900561498041,
+      "learning_rate": 1.9466289658808207e-06,
+      "loss": 0.4919,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5840168927150167,
+      "grad_norm": 6.249402586479176,
+      "learning_rate": 1.9459876753142593e-06,
+      "loss": 0.541,
+      "step": 7191
+    },
+    {
+      "epoch": 0.5840981076910582,
+      "grad_norm": 2.926784134616971,
+      "learning_rate": 1.9453464230824186e-06,
+      "loss": 0.4936,
+      "step": 7192
+    },
+    {
+      "epoch": 0.5841793226670998,
+      "grad_norm": 3.6484600396561433,
+      "learning_rate": 1.9447052092296712e-06,
+      "loss": 0.5192,
+      "step": 7193
+    },
+    {
+      "epoch": 0.5842605376431413,
+      "grad_norm": 3.898125434059147,
+      "learning_rate": 1.9440640338003835e-06,
+      "loss": 0.4092,
+      "step": 7194
+    },
+    {
+      "epoch": 0.584341752619183,
+      "grad_norm": 3.683502965273959,
+      "learning_rate": 1.943422896838922e-06,
+      "loss": 0.5507,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5844229675952246,
+      "grad_norm": 5.288308403487764,
+      "learning_rate": 1.9427817983896518e-06,
+      "loss": 0.4068,
+      "step": 7196
+    },
+    {
+      "epoch": 0.5845041825712661,
+      "grad_norm": 5.445064431182992,
+      "learning_rate": 1.942140738496931e-06,
+      "loss": 0.7837,
+      "step": 7197
+    },
+    {
+      "epoch": 0.5845853975473078,
+      "grad_norm": 6.058709142095907,
+      "learning_rate": 1.9414997172051184e-06,
+      "loss": 0.3739,
+      "step": 7198
+    },
+    {
+      "epoch": 0.5846666125233493,
+      "grad_norm": 3.970322164486651,
+      "learning_rate": 1.9408587345585707e-06,
+      "loss": 0.5194,
+      "step": 7199
+    },
+    {
+      "epoch": 0.5847478274993909,
+      "grad_norm": 5.3016872496159495,
+      "learning_rate": 1.9402177906016395e-06,
+      "loss": 0.4401,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5848290424754324,
+      "grad_norm": 4.775951466036485,
+      "learning_rate": 1.939576885378674e-06,
+      "loss": 0.398,
+      "step": 7201
+    },
+    {
+      "epoch": 0.5849102574514741,
+      "grad_norm": 6.899462009323301,
+      "learning_rate": 1.9389360189340213e-06,
+      "loss": 0.5067,
+      "step": 7202
+    },
+    {
+      "epoch": 0.5849914724275156,
+      "grad_norm": 3.2063799922509464,
+      "learning_rate": 1.9382951913120276e-06,
+      "loss": 0.5829,
+      "step": 7203
+    },
+    {
+      "epoch": 0.5850726874035572,
+      "grad_norm": 9.321053496169446,
+      "learning_rate": 1.937654402557034e-06,
+      "loss": 0.4999,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5851539023795987,
+      "grad_norm": 5.677431258954087,
+      "learning_rate": 1.937013652713378e-06,
+      "loss": 0.2643,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5852351173556404,
+      "grad_norm": 3.3598751265364446,
+      "learning_rate": 1.9363729418253995e-06,
+      "loss": 0.7468,
+      "step": 7206
+    },
+    {
+      "epoch": 0.585316332331682,
+      "grad_norm": 4.156902235894881,
+      "learning_rate": 1.93573226993743e-06,
+      "loss": 0.5862,
+      "step": 7207
+    },
+    {
+      "epoch": 0.5853975473077235,
+      "grad_norm": 4.5056414028942084,
+      "learning_rate": 1.9350916370938004e-06,
+      "loss": 0.6588,
+      "step": 7208
+    },
+    {
+      "epoch": 0.5854787622837652,
+      "grad_norm": 5.759552128771596,
+      "learning_rate": 1.9344510433388405e-06,
+      "loss": 0.8105,
+      "step": 7209
+    },
+    {
+      "epoch": 0.5855599772598067,
+      "grad_norm": 4.3224511149017255,
+      "learning_rate": 1.9338104887168753e-06,
+      "loss": 0.5087,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5856411922358483,
+      "grad_norm": 4.7674728901987455,
+      "learning_rate": 1.933169973272227e-06,
+      "loss": 0.4611,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5857224072118898,
+      "grad_norm": 5.687185959055942,
+      "learning_rate": 1.932529497049217e-06,
+      "loss": 0.5603,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5858036221879315,
+      "grad_norm": 6.09715437191108,
+      "learning_rate": 1.9318890600921638e-06,
+      "loss": 0.6074,
+      "step": 7213
+    },
+    {
+      "epoch": 0.585884837163973,
+      "grad_norm": 5.020060984695505,
+      "learning_rate": 1.9312486624453783e-06,
+      "loss": 0.6328,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5859660521400146,
+      "grad_norm": 4.477953926588951,
+      "learning_rate": 1.9306083041531773e-06,
+      "loss": 0.5313,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5860472671160561,
+      "grad_norm": 3.519455685248619,
+      "learning_rate": 1.9299679852598684e-06,
+      "loss": 0.6649,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5861284820920978,
+      "grad_norm": 3.6704238824484325,
+      "learning_rate": 1.929327705809757e-06,
+      "loss": 0.4758,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5862096970681394,
+      "grad_norm": 6.283905565847572,
+      "learning_rate": 1.928687465847148e-06,
+      "loss": 0.4217,
+      "step": 7218
+    },
+    {
+      "epoch": 0.5862909120441809,
+      "grad_norm": 6.945906876806998,
+      "learning_rate": 1.9280472654163436e-06,
+      "loss": 0.4945,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5863721270202226,
+      "grad_norm": 8.370884932771967,
+      "learning_rate": 1.927407104561641e-06,
+      "loss": 0.4238,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5864533419962641,
+      "grad_norm": 4.383870501450878,
+      "learning_rate": 1.926766983327336e-06,
+      "loss": 0.5634,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5865345569723057,
+      "grad_norm": 9.241632141950406,
+      "learning_rate": 1.9261269017577228e-06,
+      "loss": 0.634,
+      "step": 7222
+    },
+    {
+      "epoch": 0.5866157719483472,
+      "grad_norm": 3.7586621345897853,
+      "learning_rate": 1.9254868598970904e-06,
+      "loss": 0.5618,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5866969869243889,
+      "grad_norm": 4.209074416276335,
+      "learning_rate": 1.924846857789726e-06,
+      "loss": 0.6119,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5867782019004304,
+      "grad_norm": 4.546889702288765,
+      "learning_rate": 1.924206895479916e-06,
+      "loss": 0.4853,
+      "step": 7225
+    },
+    {
+      "epoch": 0.586859416876472,
+      "grad_norm": 3.7823061514718423,
+      "learning_rate": 1.9235669730119415e-06,
+      "loss": 0.3448,
+      "step": 7226
+    },
+    {
+      "epoch": 0.5869406318525136,
+      "grad_norm": 4.256499430588995,
+      "learning_rate": 1.922927090430081e-06,
+      "loss": 0.5403,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5870218468285552,
+      "grad_norm": 44.81613726842311,
+      "learning_rate": 1.9222872477786124e-06,
+      "loss": 0.4421,
+      "step": 7228
+    },
+    {
+      "epoch": 0.5871030618045968,
+      "grad_norm": 4.765715400723174,
+      "learning_rate": 1.921647445101809e-06,
+      "loss": 0.4627,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5871842767806383,
+      "grad_norm": 20.740588315180986,
+      "learning_rate": 1.921007682443941e-06,
+      "loss": 0.4778,
+      "step": 7230
+    },
+    {
+      "epoch": 0.58726549175668,
+      "grad_norm": 5.271002543096211,
+      "learning_rate": 1.920367959849277e-06,
+      "loss": 0.4699,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5873467067327215,
+      "grad_norm": 4.633497702253883,
+      "learning_rate": 1.919728277362083e-06,
+      "loss": 0.6411,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5874279217087631,
+      "grad_norm": 2.739394598878589,
+      "learning_rate": 1.91908863502662e-06,
+      "loss": 0.3667,
+      "step": 7233
+    },
+    {
+      "epoch": 0.5875091366848046,
+      "grad_norm": 5.224714666393266,
+      "learning_rate": 1.9184490328871502e-06,
+      "loss": 0.5021,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5875903516608463,
+      "grad_norm": 7.019642199719357,
+      "learning_rate": 1.9178094709879296e-06,
+      "loss": 0.4818,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5876715666368878,
+      "grad_norm": 5.128939865748304,
+      "learning_rate": 1.9171699493732122e-06,
+      "loss": 0.455,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5877527816129294,
+      "grad_norm": 7.227653257937603,
+      "learning_rate": 1.916530468087249e-06,
+      "loss": 0.4862,
+      "step": 7237
+    },
+    {
+      "epoch": 0.587833996588971,
+      "grad_norm": 5.527188153857498,
+      "learning_rate": 1.9158910271742905e-06,
+      "loss": 0.5265,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5879152115650126,
+      "grad_norm": 2.6827965014779593,
+      "learning_rate": 1.9152516266785807e-06,
+      "loss": 0.5964,
+      "step": 7239
+    },
+    {
+      "epoch": 0.5879964265410542,
+      "grad_norm": 5.211112335596169,
+      "learning_rate": 1.9146122666443635e-06,
+      "loss": 0.5542,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5880776415170957,
+      "grad_norm": 4.03076289953267,
+      "learning_rate": 1.91397294711588e-06,
+      "loss": 0.4605,
+      "step": 7241
+    },
+    {
+      "epoch": 0.5881588564931374,
+      "grad_norm": 5.996591886378709,
+      "learning_rate": 1.9133336681373673e-06,
+      "loss": 0.5431,
+      "step": 7242
+    },
+    {
+      "epoch": 0.5882400714691789,
+      "grad_norm": 4.7750369105364845,
+      "learning_rate": 1.912694429753059e-06,
+      "loss": 0.4664,
+      "step": 7243
+    },
+    {
+      "epoch": 0.5883212864452205,
+      "grad_norm": 4.118454697868038,
+      "learning_rate": 1.912055232007188e-06,
+      "loss": 0.4708,
+      "step": 7244
+    },
+    {
+      "epoch": 0.588402501421262,
+      "grad_norm": 5.287731670886094,
+      "learning_rate": 1.911416074943984e-06,
+      "loss": 0.4099,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5884837163973037,
+      "grad_norm": 3.463709326164506,
+      "learning_rate": 1.9107769586076716e-06,
+      "loss": 0.4555,
+      "step": 7246
+    },
+    {
+      "epoch": 0.5885649313733452,
+      "grad_norm": 3.9158367444108855,
+      "learning_rate": 1.9101378830424758e-06,
+      "loss": 0.5739,
+      "step": 7247
+    },
+    {
+      "epoch": 0.5886461463493868,
+      "grad_norm": 4.558330070983507,
+      "learning_rate": 1.909498848292617e-06,
+      "loss": 0.561,
+      "step": 7248
+    },
+    {
+      "epoch": 0.5887273613254284,
+      "grad_norm": 5.495600846510651,
+      "learning_rate": 1.9088598544023118e-06,
+      "loss": 0.3776,
+      "step": 7249
+    },
+    {
+      "epoch": 0.58880857630147,
+      "grad_norm": 4.632815620841959,
+      "learning_rate": 1.908220901415777e-06,
+      "loss": 0.4483,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5888897912775116,
+      "grad_norm": 4.978619608664077,
+      "learning_rate": 1.907581989377224e-06,
+      "loss": 0.5724,
+      "step": 7251
+    },
+    {
+      "epoch": 0.5889710062535531,
+      "grad_norm": 5.346540069308749,
+      "learning_rate": 1.9069431183308615e-06,
+      "loss": 0.5294,
+      "step": 7252
+    },
+    {
+      "epoch": 0.5890522212295948,
+      "grad_norm": 4.59400323772315,
+      "learning_rate": 1.906304288320896e-06,
+      "loss": 0.6616,
+      "step": 7253
+    },
+    {
+      "epoch": 0.5891334362056363,
+      "grad_norm": 6.3719239881505665,
+      "learning_rate": 1.9056654993915326e-06,
+      "loss": 0.6185,
+      "step": 7254
+    },
+    {
+      "epoch": 0.5892146511816779,
+      "grad_norm": 6.682687810031172,
+      "learning_rate": 1.9050267515869709e-06,
+      "loss": 0.4781,
+      "step": 7255
+    },
+    {
+      "epoch": 0.5892958661577195,
+      "grad_norm": 3.7501427973660397,
+      "learning_rate": 1.9043880449514085e-06,
+      "loss": 0.8261,
+      "step": 7256
+    },
+    {
+      "epoch": 0.5893770811337611,
+      "grad_norm": 5.558288691226038,
+      "learning_rate": 1.9037493795290421e-06,
+      "loss": 0.4972,
+      "step": 7257
+    },
+    {
+      "epoch": 0.5894582961098026,
+      "grad_norm": 5.421642176266866,
+      "learning_rate": 1.9031107553640632e-06,
+      "loss": 0.4625,
+      "step": 7258
+    },
+    {
+      "epoch": 0.5895395110858442,
+      "grad_norm": 5.221069697025148,
+      "learning_rate": 1.9024721725006598e-06,
+      "loss": 0.3844,
+      "step": 7259
+    },
+    {
+      "epoch": 0.5896207260618858,
+      "grad_norm": 7.976877536523762,
+      "learning_rate": 1.9018336309830202e-06,
+      "loss": 0.3608,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5897019410379274,
+      "grad_norm": 9.70876511343524,
+      "learning_rate": 1.9011951308553284e-06,
+      "loss": 0.5597,
+      "step": 7261
+    },
+    {
+      "epoch": 0.589783156013969,
+      "grad_norm": 4.2405529559362405,
+      "learning_rate": 1.900556672161763e-06,
+      "loss": 0.5158,
+      "step": 7262
+    },
+    {
+      "epoch": 0.5898643709900105,
+      "grad_norm": 6.1774490371345845,
+      "learning_rate": 1.899918254946504e-06,
+      "loss": 0.5368,
+      "step": 7263
+    },
+    {
+      "epoch": 0.5899455859660522,
+      "grad_norm": 5.486539007392727,
+      "learning_rate": 1.8992798792537265e-06,
+      "loss": 0.6,
+      "step": 7264
+    },
+    {
+      "epoch": 0.5900268009420937,
+      "grad_norm": 4.8768276862477995,
+      "learning_rate": 1.898641545127601e-06,
+      "loss": 0.3974,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5901080159181353,
+      "grad_norm": 5.691046376635863,
+      "learning_rate": 1.8980032526122985e-06,
+      "loss": 0.5479,
+      "step": 7266
+    },
+    {
+      "epoch": 0.5901892308941769,
+      "grad_norm": 5.4416215817890325,
+      "learning_rate": 1.8973650017519855e-06,
+      "loss": 0.4471,
+      "step": 7267
+    },
+    {
+      "epoch": 0.5902704458702185,
+      "grad_norm": 3.7432268661179324,
+      "learning_rate": 1.8967267925908237e-06,
+      "loss": 0.5173,
+      "step": 7268
+    },
+    {
+      "epoch": 0.59035166084626,
+      "grad_norm": 8.040626967946537,
+      "learning_rate": 1.8960886251729756e-06,
+      "loss": 0.4184,
+      "step": 7269
+    },
+    {
+      "epoch": 0.5904328758223016,
+      "grad_norm": 6.080428319987492,
+      "learning_rate": 1.8954504995425994e-06,
+      "loss": 0.4146,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5905140907983432,
+      "grad_norm": 4.115370064107381,
+      "learning_rate": 1.8948124157438485e-06,
+      "loss": 0.4314,
+      "step": 7271
+    },
+    {
+      "epoch": 0.5905953057743848,
+      "grad_norm": 4.654941886199049,
+      "learning_rate": 1.8941743738208752e-06,
+      "loss": 0.4831,
+      "step": 7272
+    },
+    {
+      "epoch": 0.5906765207504264,
+      "grad_norm": 6.074230738214155,
+      "learning_rate": 1.8935363738178288e-06,
+      "loss": 0.5219,
+      "step": 7273
+    },
+    {
+      "epoch": 0.590757735726468,
+      "grad_norm": 4.209033114032256,
+      "learning_rate": 1.8928984157788565e-06,
+      "loss": 0.6185,
+      "step": 7274
+    },
+    {
+      "epoch": 0.5908389507025096,
+      "grad_norm": 6.854702739452237,
+      "learning_rate": 1.8922604997480998e-06,
+      "loss": 0.5069,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5909201656785511,
+      "grad_norm": 5.0281604588869975,
+      "learning_rate": 1.8916226257697004e-06,
+      "loss": 0.4144,
+      "step": 7276
+    },
+    {
+      "epoch": 0.5910013806545927,
+      "grad_norm": 3.904693540082544,
+      "learning_rate": 1.8909847938877962e-06,
+      "loss": 0.697,
+      "step": 7277
+    },
+    {
+      "epoch": 0.5910825956306343,
+      "grad_norm": 4.638587360552655,
+      "learning_rate": 1.89034700414652e-06,
+      "loss": 0.3997,
+      "step": 7278
+    },
+    {
+      "epoch": 0.5911638106066759,
+      "grad_norm": 4.955343175267275,
+      "learning_rate": 1.8897092565900048e-06,
+      "loss": 0.5474,
+      "step": 7279
+    },
+    {
+      "epoch": 0.5912450255827174,
+      "grad_norm": 4.586670514862452,
+      "learning_rate": 1.8890715512623802e-06,
+      "loss": 0.5885,
+      "step": 7280
+    },
+    {
+      "epoch": 0.591326240558759,
+      "grad_norm": 6.135861467307459,
+      "learning_rate": 1.8884338882077697e-06,
+      "loss": 0.4193,
+      "step": 7281
+    },
+    {
+      "epoch": 0.5914074555348006,
+      "grad_norm": 8.91363405049214,
+      "learning_rate": 1.8877962674702977e-06,
+      "loss": 0.4749,
+      "step": 7282
+    },
+    {
+      "epoch": 0.5914886705108422,
+      "grad_norm": 3.5396738186174317,
+      "learning_rate": 1.8871586890940847e-06,
+      "loss": 0.4059,
+      "step": 7283
+    },
+    {
+      "epoch": 0.5915698854868838,
+      "grad_norm": 4.623629413772576,
+      "learning_rate": 1.886521153123246e-06,
+      "loss": 0.4456,
+      "step": 7284
+    },
+    {
+      "epoch": 0.5916511004629254,
+      "grad_norm": 4.952672810545965,
+      "learning_rate": 1.8858836596018973e-06,
+      "loss": 0.4957,
+      "step": 7285
+    },
+    {
+      "epoch": 0.591732315438967,
+      "grad_norm": 4.741145039545014,
+      "learning_rate": 1.8852462085741497e-06,
+      "loss": 0.5374,
+      "step": 7286
+    },
+    {
+      "epoch": 0.5918135304150085,
+      "grad_norm": 8.418571766778543,
+      "learning_rate": 1.8846088000841096e-06,
+      "loss": 0.558,
+      "step": 7287
+    },
+    {
+      "epoch": 0.5918947453910501,
+      "grad_norm": 3.602706463619327,
+      "learning_rate": 1.8839714341758847e-06,
+      "loss": 0.5848,
+      "step": 7288
+    },
+    {
+      "epoch": 0.5919759603670917,
+      "grad_norm": 10.651397817035003,
+      "learning_rate": 1.883334110893576e-06,
+      "loss": 0.5071,
+      "step": 7289
+    },
+    {
+      "epoch": 0.5920571753431333,
+      "grad_norm": 4.361973439215471,
+      "learning_rate": 1.8826968302812837e-06,
+      "loss": 0.4555,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5921383903191748,
+      "grad_norm": 6.18398302417084,
+      "learning_rate": 1.8820595923831025e-06,
+      "loss": 0.6062,
+      "step": 7291
+    },
+    {
+      "epoch": 0.5922196052952164,
+      "grad_norm": 8.248248546443998,
+      "learning_rate": 1.8814223972431276e-06,
+      "loss": 0.5944,
+      "step": 7292
+    },
+    {
+      "epoch": 0.592300820271258,
+      "grad_norm": 7.741858864310114,
+      "learning_rate": 1.8807852449054497e-06,
+      "loss": 0.5043,
+      "step": 7293
+    },
+    {
+      "epoch": 0.5923820352472996,
+      "grad_norm": 3.9704756953268565,
+      "learning_rate": 1.8801481354141547e-06,
+      "loss": 0.5163,
+      "step": 7294
+    },
+    {
+      "epoch": 0.5924632502233412,
+      "grad_norm": 4.432871358786943,
+      "learning_rate": 1.8795110688133283e-06,
+      "loss": 0.4582,
+      "step": 7295
+    },
+    {
+      "epoch": 0.5925444651993828,
+      "grad_norm": 4.035829611958314,
+      "learning_rate": 1.878874045147053e-06,
+      "loss": 0.4393,
+      "step": 7296
+    },
+    {
+      "epoch": 0.5926256801754244,
+      "grad_norm": 3.2780411870082284,
+      "learning_rate": 1.8782370644594055e-06,
+      "loss": 0.62,
+      "step": 7297
+    },
+    {
+      "epoch": 0.5927068951514659,
+      "grad_norm": 5.278700539156712,
+      "learning_rate": 1.8776001267944628e-06,
+      "loss": 0.5785,
+      "step": 7298
+    },
+    {
+      "epoch": 0.5927881101275075,
+      "grad_norm": 3.827532458925898,
+      "learning_rate": 1.876963232196298e-06,
+      "loss": 0.57,
+      "step": 7299
+    },
+    {
+      "epoch": 0.5928693251035491,
+      "grad_norm": 4.0256389154851595,
+      "learning_rate": 1.876326380708979e-06,
+      "loss": 0.609,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5929505400795907,
+      "grad_norm": 8.391342343121048,
+      "learning_rate": 1.8756895723765747e-06,
+      "loss": 0.5095,
+      "step": 7301
+    },
+    {
+      "epoch": 0.5930317550556322,
+      "grad_norm": 3.698568278989539,
+      "learning_rate": 1.8750528072431477e-06,
+      "loss": 0.5275,
+      "step": 7302
+    },
+    {
+      "epoch": 0.5931129700316738,
+      "grad_norm": 7.336981984750687,
+      "learning_rate": 1.8744160853527579e-06,
+      "loss": 0.6133,
+      "step": 7303
+    },
+    {
+      "epoch": 0.5931941850077154,
+      "grad_norm": 4.166581863576458,
+      "learning_rate": 1.8737794067494656e-06,
+      "loss": 0.4626,
+      "step": 7304
+    },
+    {
+      "epoch": 0.593275399983757,
+      "grad_norm": 4.274646132254657,
+      "learning_rate": 1.8731427714773233e-06,
+      "loss": 0.5315,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5933566149597986,
+      "grad_norm": 3.911879665279761,
+      "learning_rate": 1.8725061795803846e-06,
+      "loss": 0.4508,
+      "step": 7306
+    },
+    {
+      "epoch": 0.5934378299358402,
+      "grad_norm": 2.642462457610074,
+      "learning_rate": 1.8718696311026956e-06,
+      "loss": 0.5147,
+      "step": 7307
+    },
+    {
+      "epoch": 0.5935190449118818,
+      "grad_norm": 4.675230745750343,
+      "learning_rate": 1.871233126088305e-06,
+      "loss": 0.3801,
+      "step": 7308
+    },
+    {
+      "epoch": 0.5936002598879233,
+      "grad_norm": 10.255898660318655,
+      "learning_rate": 1.8705966645812544e-06,
+      "loss": 0.5329,
+      "step": 7309
+    },
+    {
+      "epoch": 0.5936814748639649,
+      "grad_norm": 4.954517136307957,
+      "learning_rate": 1.8699602466255828e-06,
+      "loss": 0.4131,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5937626898400065,
+      "grad_norm": 4.751776132607896,
+      "learning_rate": 1.8693238722653278e-06,
+      "loss": 0.4266,
+      "step": 7311
+    },
+    {
+      "epoch": 0.5938439048160481,
+      "grad_norm": 3.047587348683506,
+      "learning_rate": 1.8686875415445238e-06,
+      "loss": 0.4889,
+      "step": 7312
+    },
+    {
+      "epoch": 0.5939251197920896,
+      "grad_norm": 12.990954504278774,
+      "learning_rate": 1.8680512545071999e-06,
+      "loss": 0.4731,
+      "step": 7313
+    },
+    {
+      "epoch": 0.5940063347681313,
+      "grad_norm": 4.070689985920183,
+      "learning_rate": 1.8674150111973854e-06,
+      "loss": 0.6631,
+      "step": 7314
+    },
+    {
+      "epoch": 0.5940875497441728,
+      "grad_norm": 2.864000413306198,
+      "learning_rate": 1.866778811659104e-06,
+      "loss": 0.8279,
+      "step": 7315
+    },
+    {
+      "epoch": 0.5941687647202144,
+      "grad_norm": 3.763015690605599,
+      "learning_rate": 1.8661426559363768e-06,
+      "loss": 0.6624,
+      "step": 7316
+    },
+    {
+      "epoch": 0.594249979696256,
+      "grad_norm": 6.349440373441788,
+      "learning_rate": 1.8655065440732243e-06,
+      "loss": 0.5768,
+      "step": 7317
+    },
+    {
+      "epoch": 0.5943311946722976,
+      "grad_norm": 3.739887750298035,
+      "learning_rate": 1.8648704761136604e-06,
+      "loss": 0.54,
+      "step": 7318
+    },
+    {
+      "epoch": 0.5944124096483392,
+      "grad_norm": 3.8465470522434977,
+      "learning_rate": 1.8642344521016974e-06,
+      "loss": 0.4265,
+      "step": 7319
+    },
+    {
+      "epoch": 0.5944936246243807,
+      "grad_norm": 5.442222857316663,
+      "learning_rate": 1.8635984720813471e-06,
+      "loss": 0.5427,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5945748396004223,
+      "grad_norm": 5.036275208595901,
+      "learning_rate": 1.8629625360966137e-06,
+      "loss": 0.4966,
+      "step": 7321
+    },
+    {
+      "epoch": 0.5946560545764639,
+      "grad_norm": 4.325139913021066,
+      "learning_rate": 1.8623266441915006e-06,
+      "loss": 0.4192,
+      "step": 7322
+    },
+    {
+      "epoch": 0.5947372695525055,
+      "grad_norm": 3.7229174258197437,
+      "learning_rate": 1.86169079641001e-06,
+      "loss": 0.4411,
+      "step": 7323
+    },
+    {
+      "epoch": 0.594818484528547,
+      "grad_norm": 4.325040819420997,
+      "learning_rate": 1.861054992796138e-06,
+      "loss": 0.5971,
+      "step": 7324
+    },
+    {
+      "epoch": 0.5948996995045887,
+      "grad_norm": 5.429278035591401,
+      "learning_rate": 1.860419233393879e-06,
+      "loss": 0.5484,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5949809144806302,
+      "grad_norm": 4.384346726547922,
+      "learning_rate": 1.859783518247223e-06,
+      "loss": 0.5676,
+      "step": 7326
+    },
+    {
+      "epoch": 0.5950621294566718,
+      "grad_norm": 7.391929682556335,
+      "learning_rate": 1.8591478474001601e-06,
+      "loss": 0.4677,
+      "step": 7327
+    },
+    {
+      "epoch": 0.5951433444327134,
+      "grad_norm": 5.680231306618248,
+      "learning_rate": 1.858512220896675e-06,
+      "loss": 0.562,
+      "step": 7328
+    },
+    {
+      "epoch": 0.595224559408755,
+      "grad_norm": 4.0720446605312395,
+      "learning_rate": 1.857876638780748e-06,
+      "loss": 0.4176,
+      "step": 7329
+    },
+    {
+      "epoch": 0.5953057743847966,
+      "grad_norm": 4.5916594217289575,
+      "learning_rate": 1.85724110109636e-06,
+      "loss": 0.5309,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5953869893608381,
+      "grad_norm": 4.6780280469096365,
+      "learning_rate": 1.8566056078874858e-06,
+      "loss": 0.4923,
+      "step": 7331
+    },
+    {
+      "epoch": 0.5954682043368797,
+      "grad_norm": 3.6599980641941734,
+      "learning_rate": 1.8559701591980977e-06,
+      "loss": 0.5152,
+      "step": 7332
+    },
+    {
+      "epoch": 0.5955494193129213,
+      "grad_norm": 2.978800119941686,
+      "learning_rate": 1.8553347550721672e-06,
+      "loss": 0.4539,
+      "step": 7333
+    },
+    {
+      "epoch": 0.5956306342889629,
+      "grad_norm": 3.411250854476356,
+      "learning_rate": 1.8546993955536597e-06,
+      "loss": 0.64,
+      "step": 7334
+    },
+    {
+      "epoch": 0.5957118492650044,
+      "grad_norm": 4.069278334737661,
+      "learning_rate": 1.8540640806865379e-06,
+      "loss": 0.5401,
+      "step": 7335
+    },
+    {
+      "epoch": 0.595793064241046,
+      "grad_norm": 7.8976077553265505,
+      "learning_rate": 1.8534288105147644e-06,
+      "loss": 0.4868,
+      "step": 7336
+    },
+    {
+      "epoch": 0.5958742792170876,
+      "grad_norm": 3.08341050427021,
+      "learning_rate": 1.8527935850822947e-06,
+      "loss": 0.4693,
+      "step": 7337
+    },
+    {
+      "epoch": 0.5959554941931292,
+      "grad_norm": 4.0472231931258955,
+      "learning_rate": 1.8521584044330832e-06,
+      "loss": 0.5314,
+      "step": 7338
+    },
+    {
+      "epoch": 0.5960367091691708,
+      "grad_norm": 7.289070610294687,
+      "learning_rate": 1.851523268611082e-06,
+      "loss": 0.6015,
+      "step": 7339
+    },
+    {
+      "epoch": 0.5961179241452124,
+      "grad_norm": 5.48979102747638,
+      "learning_rate": 1.8508881776602386e-06,
+      "loss": 0.4459,
+      "step": 7340
+    },
+    {
+      "epoch": 0.596199139121254,
+      "grad_norm": 15.0231623389192,
+      "learning_rate": 1.850253131624497e-06,
+      "loss": 0.5274,
+      "step": 7341
+    },
+    {
+      "epoch": 0.5962803540972955,
+      "grad_norm": 4.569897243481455,
+      "learning_rate": 1.8496181305478014e-06,
+      "loss": 0.7543,
+      "step": 7342
+    },
+    {
+      "epoch": 0.5963615690733371,
+      "grad_norm": 5.818717995667628,
+      "learning_rate": 1.8489831744740887e-06,
+      "loss": 0.4809,
+      "step": 7343
+    },
+    {
+      "epoch": 0.5964427840493787,
+      "grad_norm": 7.4916301022433505,
+      "learning_rate": 1.8483482634472948e-06,
+      "loss": 0.4216,
+      "step": 7344
+    },
+    {
+      "epoch": 0.5965239990254203,
+      "grad_norm": 2.952985348960789,
+      "learning_rate": 1.8477133975113516e-06,
+      "loss": 0.4636,
+      "step": 7345
+    },
+    {
+      "epoch": 0.5966052140014618,
+      "grad_norm": 5.3497885917720875,
+      "learning_rate": 1.8470785767101898e-06,
+      "loss": 0.5137,
+      "step": 7346
+    },
+    {
+      "epoch": 0.5966864289775035,
+      "grad_norm": 4.056347500018793,
+      "learning_rate": 1.8464438010877348e-06,
+      "loss": 0.6904,
+      "step": 7347
+    },
+    {
+      "epoch": 0.596767643953545,
+      "grad_norm": 3.334261926906368,
+      "learning_rate": 1.845809070687909e-06,
+      "loss": 0.5172,
+      "step": 7348
+    },
+    {
+      "epoch": 0.5968488589295866,
+      "grad_norm": 8.367386875652794,
+      "learning_rate": 1.8451743855546345e-06,
+      "loss": 0.434,
+      "step": 7349
+    },
+    {
+      "epoch": 0.5969300739056282,
+      "grad_norm": 6.559459547573149,
+      "learning_rate": 1.8445397457318265e-06,
+      "loss": 0.5824,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5970112888816698,
+      "grad_norm": 6.094611931889219,
+      "learning_rate": 1.8439051512633984e-06,
+      "loss": 0.5153,
+      "step": 7351
+    },
+    {
+      "epoch": 0.5970925038577114,
+      "grad_norm": 25.80816640442915,
+      "learning_rate": 1.8432706021932627e-06,
+      "loss": 0.6593,
+      "step": 7352
+    },
+    {
+      "epoch": 0.5971737188337529,
+      "grad_norm": 4.977546000712229,
+      "learning_rate": 1.8426360985653248e-06,
+      "loss": 0.6459,
+      "step": 7353
+    },
+    {
+      "epoch": 0.5972549338097946,
+      "grad_norm": 5.821123685581579,
+      "learning_rate": 1.8420016404234897e-06,
+      "loss": 0.5861,
+      "step": 7354
+    },
+    {
+      "epoch": 0.5973361487858361,
+      "grad_norm": 6.3086144237425525,
+      "learning_rate": 1.8413672278116595e-06,
+      "loss": 0.4389,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5974173637618777,
+      "grad_norm": 5.044008936737754,
+      "learning_rate": 1.840732860773731e-06,
+      "loss": 0.4021,
+      "step": 7356
+    },
+    {
+      "epoch": 0.5974985787379192,
+      "grad_norm": 4.6734719943119005,
+      "learning_rate": 1.8400985393535986e-06,
+      "loss": 0.4824,
+      "step": 7357
+    },
+    {
+      "epoch": 0.5975797937139609,
+      "grad_norm": 3.603622687404737,
+      "learning_rate": 1.8394642635951563e-06,
+      "loss": 0.5207,
+      "step": 7358
+    },
+    {
+      "epoch": 0.5976610086900024,
+      "grad_norm": 3.7989831364132,
+      "learning_rate": 1.838830033542291e-06,
+      "loss": 0.6815,
+      "step": 7359
+    },
+    {
+      "epoch": 0.597742223666044,
+      "grad_norm": 2.3766163305666153,
+      "learning_rate": 1.8381958492388873e-06,
+      "loss": 0.4749,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5978234386420856,
+      "grad_norm": 3.0966906185261993,
+      "learning_rate": 1.837561710728828e-06,
+      "loss": 0.5948,
+      "step": 7361
+    },
+    {
+      "epoch": 0.5979046536181272,
+      "grad_norm": 8.468628755593905,
+      "learning_rate": 1.8369276180559933e-06,
+      "loss": 0.3566,
+      "step": 7362
+    },
+    {
+      "epoch": 0.5979858685941688,
+      "grad_norm": 9.11735062017391,
+      "learning_rate": 1.836293571264258e-06,
+      "loss": 0.4737,
+      "step": 7363
+    },
+    {
+      "epoch": 0.5980670835702103,
+      "grad_norm": 5.041492074022004,
+      "learning_rate": 1.835659570397494e-06,
+      "loss": 0.4476,
+      "step": 7364
+    },
+    {
+      "epoch": 0.598148298546252,
+      "grad_norm": 3.99736329645453,
+      "learning_rate": 1.8350256154995733e-06,
+      "loss": 0.6365,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5982295135222935,
+      "grad_norm": 8.810624359575181,
+      "learning_rate": 1.8343917066143597e-06,
+      "loss": 0.4232,
+      "step": 7366
+    },
+    {
+      "epoch": 0.5983107284983351,
+      "grad_norm": 4.925256449200388,
+      "learning_rate": 1.8337578437857169e-06,
+      "loss": 0.4499,
+      "step": 7367
+    },
+    {
+      "epoch": 0.5983919434743766,
+      "grad_norm": 3.5049195280397565,
+      "learning_rate": 1.8331240270575062e-06,
+      "loss": 0.6362,
+      "step": 7368
+    },
+    {
+      "epoch": 0.5984731584504183,
+      "grad_norm": 8.10123639728082,
+      "learning_rate": 1.8324902564735834e-06,
+      "loss": 0.4814,
+      "step": 7369
+    },
+    {
+      "epoch": 0.5985543734264598,
+      "grad_norm": 3.656092093713114,
+      "learning_rate": 1.831856532077801e-06,
+      "loss": 0.6061,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5986355884025014,
+      "grad_norm": 5.372488168989093,
+      "learning_rate": 1.831222853914012e-06,
+      "loss": 0.5401,
+      "step": 7371
+    },
+    {
+      "epoch": 0.598716803378543,
+      "grad_norm": 5.279583180258407,
+      "learning_rate": 1.830589222026062e-06,
+      "loss": 0.4336,
+      "step": 7372
+    },
+    {
+      "epoch": 0.5987980183545846,
+      "grad_norm": 8.322203353642944,
+      "learning_rate": 1.8299556364577936e-06,
+      "loss": 0.6477,
+      "step": 7373
+    },
+    {
+      "epoch": 0.5988792333306262,
+      "grad_norm": 4.359529690940866,
+      "learning_rate": 1.8293220972530498e-06,
+      "loss": 0.5287,
+      "step": 7374
+    },
+    {
+      "epoch": 0.5989604483066677,
+      "grad_norm": 4.599497389802597,
+      "learning_rate": 1.8286886044556678e-06,
+      "loss": 0.4167,
+      "step": 7375
+    },
+    {
+      "epoch": 0.5990416632827094,
+      "grad_norm": 4.944453273477391,
+      "learning_rate": 1.8280551581094808e-06,
+      "loss": 0.4743,
+      "step": 7376
+    },
+    {
+      "epoch": 0.5991228782587509,
+      "grad_norm": 6.818339130298646,
+      "learning_rate": 1.8274217582583207e-06,
+      "loss": 0.6592,
+      "step": 7377
+    },
+    {
+      "epoch": 0.5992040932347925,
+      "grad_norm": 7.668165030066059,
+      "learning_rate": 1.826788404946016e-06,
+      "loss": 0.5305,
+      "step": 7378
+    },
+    {
+      "epoch": 0.599285308210834,
+      "grad_norm": 4.13273227160322,
+      "learning_rate": 1.8261550982163904e-06,
+      "loss": 0.5133,
+      "step": 7379
+    },
+    {
+      "epoch": 0.5993665231868757,
+      "grad_norm": 2.7993562519365196,
+      "learning_rate": 1.825521838113265e-06,
+      "loss": 0.5664,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5994477381629172,
+      "grad_norm": 9.626646869624484,
+      "learning_rate": 1.8248886246804598e-06,
+      "loss": 0.4671,
+      "step": 7381
+    },
+    {
+      "epoch": 0.5995289531389588,
+      "grad_norm": 7.444176905061055,
+      "learning_rate": 1.8242554579617883e-06,
+      "loss": 0.5715,
+      "step": 7382
+    },
+    {
+      "epoch": 0.5996101681150005,
+      "grad_norm": 9.588196848439587,
+      "learning_rate": 1.8236223380010625e-06,
+      "loss": 0.4073,
+      "step": 7383
+    },
+    {
+      "epoch": 0.599691383091042,
+      "grad_norm": 6.866343301362604,
+      "learning_rate": 1.8229892648420922e-06,
+      "loss": 0.6084,
+      "step": 7384
+    },
+    {
+      "epoch": 0.5997725980670836,
+      "grad_norm": 5.01419995093504,
+      "learning_rate": 1.8223562385286809e-06,
+      "loss": 0.4774,
+      "step": 7385
+    },
+    {
+      "epoch": 0.5998538130431251,
+      "grad_norm": 5.956266278463153,
+      "learning_rate": 1.8217232591046313e-06,
+      "loss": 0.5421,
+      "step": 7386
+    },
+    {
+      "epoch": 0.5999350280191668,
+      "grad_norm": 5.610744689095254,
+      "learning_rate": 1.8210903266137434e-06,
+      "loss": 0.7168,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6000162429952083,
+      "grad_norm": 4.657987631786978,
+      "learning_rate": 1.8204574410998119e-06,
+      "loss": 0.5227,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6000974579712499,
+      "grad_norm": 4.434514515163585,
+      "learning_rate": 1.8198246026066279e-06,
+      "loss": 0.4822,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6001786729472914,
+      "grad_norm": 5.183467175096997,
+      "learning_rate": 1.819191811177982e-06,
+      "loss": 0.579,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6002598879233331,
+      "grad_norm": 4.579031661588377,
+      "learning_rate": 1.8185590668576602e-06,
+      "loss": 0.6555,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6003411028993746,
+      "grad_norm": 5.1765990102478066,
+      "learning_rate": 1.817926369689444e-06,
+      "loss": 0.541,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6004223178754162,
+      "grad_norm": 3.4939919762553977,
+      "learning_rate": 1.817293719717113e-06,
+      "loss": 0.5579,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6005035328514579,
+      "grad_norm": 4.962538720225719,
+      "learning_rate": 1.8166611169844444e-06,
+      "loss": 0.3529,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6005847478274994,
+      "grad_norm": 5.042600514532005,
+      "learning_rate": 1.8160285615352092e-06,
+      "loss": 0.4768,
+      "step": 7395
+    },
+    {
+      "epoch": 0.600665962803541,
+      "grad_norm": 3.8637771148189097,
+      "learning_rate": 1.8153960534131774e-06,
+      "loss": 0.6365,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6007471777795825,
+      "grad_norm": 4.22885301921691,
+      "learning_rate": 1.8147635926621162e-06,
+      "loss": 0.5797,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6008283927556242,
+      "grad_norm": 3.6962740741884095,
+      "learning_rate": 1.8141311793257876e-06,
+      "loss": 0.3554,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6009096077316657,
+      "grad_norm": 3.758699623360753,
+      "learning_rate": 1.813498813447951e-06,
+      "loss": 0.596,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6009908227077073,
+      "grad_norm": 3.201807519185477,
+      "learning_rate": 1.812866495072364e-06,
+      "loss": 0.5367,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6010720376837488,
+      "grad_norm": 4.447710834904164,
+      "learning_rate": 1.812234224242779e-06,
+      "loss": 0.4886,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6011532526597905,
+      "grad_norm": 4.642639006126207,
+      "learning_rate": 1.8116020010029448e-06,
+      "loss": 0.5678,
+      "step": 7402
+    },
+    {
+      "epoch": 0.601234467635832,
+      "grad_norm": 3.848966652372118,
+      "learning_rate": 1.8109698253966092e-06,
+      "loss": 0.4742,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6013156826118736,
+      "grad_norm": 4.125486033292355,
+      "learning_rate": 1.8103376974675157e-06,
+      "loss": 0.4872,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6013968975879153,
+      "grad_norm": 5.747061637911401,
+      "learning_rate": 1.8097056172594023e-06,
+      "loss": 0.4748,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6014781125639568,
+      "grad_norm": 4.074457822433304,
+      "learning_rate": 1.8090735848160079e-06,
+      "loss": 0.3921,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6015593275399984,
+      "grad_norm": 9.807247879545486,
+      "learning_rate": 1.808441600181065e-06,
+      "loss": 0.5085,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6016405425160399,
+      "grad_norm": 5.7128146988760475,
+      "learning_rate": 1.8078096633983023e-06,
+      "loss": 0.431,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6017217574920816,
+      "grad_norm": 21.8207651189527,
+      "learning_rate": 1.8071777745114477e-06,
+      "loss": 0.5174,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6018029724681231,
+      "grad_norm": 9.425825620792596,
+      "learning_rate": 1.8065459335642254e-06,
+      "loss": 0.4276,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6018841874441647,
+      "grad_norm": 4.023515489113687,
+      "learning_rate": 1.8059141406003532e-06,
+      "loss": 0.4419,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6019654024202062,
+      "grad_norm": 3.615004223601,
+      "learning_rate": 1.8052823956635496e-06,
+      "loss": 0.5893,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6020466173962479,
+      "grad_norm": 5.159465901388711,
+      "learning_rate": 1.8046506987975278e-06,
+      "loss": 0.429,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6021278323722894,
+      "grad_norm": 3.103825361755657,
+      "learning_rate": 1.804019050045998e-06,
+      "loss": 0.6653,
+      "step": 7414
+    },
+    {
+      "epoch": 0.602209047348331,
+      "grad_norm": 5.198427303718621,
+      "learning_rate": 1.8033874494526646e-06,
+      "loss": 0.5391,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6022902623243727,
+      "grad_norm": 7.073819537725478,
+      "learning_rate": 1.8027558970612347e-06,
+      "loss": 0.7089,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6023714773004142,
+      "grad_norm": 3.801629807198999,
+      "learning_rate": 1.8021243929154063e-06,
+      "loss": 0.3996,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6024526922764558,
+      "grad_norm": 24.873498159563677,
+      "learning_rate": 1.8014929370588757e-06,
+      "loss": 0.5828,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6025339072524973,
+      "grad_norm": 5.485907374005129,
+      "learning_rate": 1.8008615295353376e-06,
+      "loss": 0.5204,
+      "step": 7419
+    },
+    {
+      "epoch": 0.602615122228539,
+      "grad_norm": 3.9473453351114705,
+      "learning_rate": 1.8002301703884816e-06,
+      "loss": 0.4032,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6026963372045805,
+      "grad_norm": 4.470083941751573,
+      "learning_rate": 1.799598859661994e-06,
+      "loss": 0.4807,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6027775521806221,
+      "grad_norm": 5.031697404411579,
+      "learning_rate": 1.7989675973995585e-06,
+      "loss": 0.4221,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6028587671566636,
+      "grad_norm": 6.979484374974144,
+      "learning_rate": 1.7983363836448559e-06,
+      "loss": 0.3792,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6029399821327053,
+      "grad_norm": 5.457913677824293,
+      "learning_rate": 1.7977052184415606e-06,
+      "loss": 0.3426,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6030211971087468,
+      "grad_norm": 4.4056991242013455,
+      "learning_rate": 1.7970741018333482e-06,
+      "loss": 0.5527,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6031024120847884,
+      "grad_norm": 9.887330375392434,
+      "learning_rate": 1.7964430338638883e-06,
+      "loss": 0.392,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6031836270608301,
+      "grad_norm": 4.068954581333052,
+      "learning_rate": 1.7958120145768457e-06,
+      "loss": 0.4971,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6032648420368716,
+      "grad_norm": 4.434746948784985,
+      "learning_rate": 1.7951810440158853e-06,
+      "loss": 0.503,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6033460570129132,
+      "grad_norm": 4.082130196766388,
+      "learning_rate": 1.7945501222246673e-06,
+      "loss": 0.4411,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6034272719889547,
+      "grad_norm": 10.979428713512291,
+      "learning_rate": 1.793919249246846e-06,
+      "loss": 0.4922,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6035084869649964,
+      "grad_norm": 4.413393240856426,
+      "learning_rate": 1.7932884251260767e-06,
+      "loss": 0.5367,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6035897019410379,
+      "grad_norm": 3.815015398191507,
+      "learning_rate": 1.7926576499060078e-06,
+      "loss": 0.5533,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6036709169170795,
+      "grad_norm": 4.993930990302795,
+      "learning_rate": 1.7920269236302868e-06,
+      "loss": 0.4204,
+      "step": 7433
+    },
+    {
+      "epoch": 0.603752131893121,
+      "grad_norm": 4.907619542220186,
+      "learning_rate": 1.7913962463425544e-06,
+      "loss": 0.5547,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6038333468691627,
+      "grad_norm": 4.135750795376068,
+      "learning_rate": 1.7907656180864519e-06,
+      "loss": 0.6044,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6039145618452042,
+      "grad_norm": 15.259932160511177,
+      "learning_rate": 1.790135038905616e-06,
+      "loss": 0.5229,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6039957768212458,
+      "grad_norm": 5.126091226363883,
+      "learning_rate": 1.7895045088436772e-06,
+      "loss": 0.4935,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6040769917972875,
+      "grad_norm": 4.347192454488315,
+      "learning_rate": 1.7888740279442669e-06,
+      "loss": 0.4883,
+      "step": 7438
+    },
+    {
+      "epoch": 0.604158206773329,
+      "grad_norm": 8.459728931856809,
+      "learning_rate": 1.7882435962510102e-06,
+      "loss": 0.6231,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6042394217493706,
+      "grad_norm": 4.357956562088014,
+      "learning_rate": 1.7876132138075292e-06,
+      "loss": 0.6246,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6043206367254121,
+      "grad_norm": 5.116492859940842,
+      "learning_rate": 1.786982880657444e-06,
+      "loss": 0.3546,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6044018517014538,
+      "grad_norm": 4.281674880545698,
+      "learning_rate": 1.7863525968443705e-06,
+      "loss": 0.3998,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6044830666774953,
+      "grad_norm": 4.231714096569343,
+      "learning_rate": 1.785722362411919e-06,
+      "loss": 0.434,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6045642816535369,
+      "grad_norm": 4.3628558475147265,
+      "learning_rate": 1.7850921774037012e-06,
+      "loss": 0.4972,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6046454966295784,
+      "grad_norm": 3.8871791003489866,
+      "learning_rate": 1.7844620418633202e-06,
+      "loss": 0.5716,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6047267116056201,
+      "grad_norm": 5.078543576155098,
+      "learning_rate": 1.7838319558343786e-06,
+      "loss": 0.5441,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6048079265816616,
+      "grad_norm": 11.062361952259796,
+      "learning_rate": 1.7832019193604767e-06,
+      "loss": 0.5189,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6048891415577032,
+      "grad_norm": 5.15437180400868,
+      "learning_rate": 1.7825719324852075e-06,
+      "loss": 0.5409,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6049703565337449,
+      "grad_norm": 6.299357609539531,
+      "learning_rate": 1.7819419952521645e-06,
+      "loss": 0.4818,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6050515715097864,
+      "grad_norm": 3.001544405666737,
+      "learning_rate": 1.7813121077049336e-06,
+      "loss": 0.5361,
+      "step": 7450
+    },
+    {
+      "epoch": 0.605132786485828,
+      "grad_norm": 4.256801862178172,
+      "learning_rate": 1.7806822698871022e-06,
+      "loss": 0.6223,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6052140014618695,
+      "grad_norm": 3.963672219119554,
+      "learning_rate": 1.780052481842251e-06,
+      "loss": 0.4201,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6052952164379112,
+      "grad_norm": 3.995116455512054,
+      "learning_rate": 1.7794227436139569e-06,
+      "loss": 0.5345,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6053764314139527,
+      "grad_norm": 6.799955515987135,
+      "learning_rate": 1.778793055245796e-06,
+      "loss": 0.5772,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6054576463899943,
+      "grad_norm": 5.467231768671727,
+      "learning_rate": 1.7781634167813388e-06,
+      "loss": 0.6087,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6055388613660359,
+      "grad_norm": 5.970835497003089,
+      "learning_rate": 1.7775338282641525e-06,
+      "loss": 0.4178,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6056200763420775,
+      "grad_norm": 4.908729638196646,
+      "learning_rate": 1.776904289737802e-06,
+      "loss": 0.5252,
+      "step": 7457
+    },
+    {
+      "epoch": 0.605701291318119,
+      "grad_norm": 8.614648749798091,
+      "learning_rate": 1.7762748012458481e-06,
+      "loss": 0.527,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6057825062941606,
+      "grad_norm": 5.933581667639831,
+      "learning_rate": 1.7756453628318465e-06,
+      "loss": 0.4565,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6058637212702023,
+      "grad_norm": 3.8154357222510034,
+      "learning_rate": 1.7750159745393536e-06,
+      "loss": 0.4551,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6059449362462438,
+      "grad_norm": 7.8464108786714055,
+      "learning_rate": 1.7743866364119175e-06,
+      "loss": 0.473,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6060261512222854,
+      "grad_norm": 4.514141406557562,
+      "learning_rate": 1.7737573484930853e-06,
+      "loss": 0.6192,
+      "step": 7462
+    },
+    {
+      "epoch": 0.606107366198327,
+      "grad_norm": 4.9804191071501585,
+      "learning_rate": 1.7731281108264025e-06,
+      "loss": 0.5783,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6061885811743686,
+      "grad_norm": 3.8086232760432837,
+      "learning_rate": 1.7724989234554068e-06,
+      "loss": 0.4875,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6062697961504101,
+      "grad_norm": 4.453392068019001,
+      "learning_rate": 1.7718697864236344e-06,
+      "loss": 0.4404,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6063510111264517,
+      "grad_norm": 6.631970307210538,
+      "learning_rate": 1.771240699774621e-06,
+      "loss": 0.3888,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6064322261024933,
+      "grad_norm": 3.3473485905433815,
+      "learning_rate": 1.7706116635518933e-06,
+      "loss": 0.5312,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6065134410785349,
+      "grad_norm": 8.74164772322533,
+      "learning_rate": 1.7699826777989788e-06,
+      "loss": 0.4054,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6065946560545764,
+      "grad_norm": 4.144159190764816,
+      "learning_rate": 1.7693537425593984e-06,
+      "loss": 0.5257,
+      "step": 7469
+    },
+    {
+      "epoch": 0.606675871030618,
+      "grad_norm": 4.098313315005926,
+      "learning_rate": 1.7687248578766727e-06,
+      "loss": 0.3604,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6067570860066597,
+      "grad_norm": 6.610586143194924,
+      "learning_rate": 1.7680960237943174e-06,
+      "loss": 0.6578,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6068383009827012,
+      "grad_norm": 4.968803137717695,
+      "learning_rate": 1.7674672403558421e-06,
+      "loss": 0.4375,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6069195159587428,
+      "grad_norm": 7.6507424354276,
+      "learning_rate": 1.7668385076047584e-06,
+      "loss": 0.4943,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6070007309347843,
+      "grad_norm": 5.452374844312675,
+      "learning_rate": 1.7662098255845689e-06,
+      "loss": 0.6319,
+      "step": 7474
+    },
+    {
+      "epoch": 0.607081945910826,
+      "grad_norm": 10.145687214568056,
+      "learning_rate": 1.7655811943387758e-06,
+      "loss": 0.4869,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6071631608868675,
+      "grad_norm": 4.614412678031245,
+      "learning_rate": 1.764952613910878e-06,
+      "loss": 0.5788,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6072443758629091,
+      "grad_norm": 7.111983931763767,
+      "learning_rate": 1.7643240843443686e-06,
+      "loss": 0.505,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6073255908389507,
+      "grad_norm": 8.127180595467415,
+      "learning_rate": 1.7636956056827384e-06,
+      "loss": 0.6297,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6074068058149923,
+      "grad_norm": 6.123717134281918,
+      "learning_rate": 1.7630671779694768e-06,
+      "loss": 0.5801,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6074880207910338,
+      "grad_norm": 5.652896769901682,
+      "learning_rate": 1.7624388012480656e-06,
+      "loss": 0.4418,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6075692357670754,
+      "grad_norm": 5.923671663679729,
+      "learning_rate": 1.7618104755619852e-06,
+      "loss": 0.4421,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6076504507431171,
+      "grad_norm": 5.6081581939098495,
+      "learning_rate": 1.7611822009547143e-06,
+      "loss": 0.7917,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6077316657191586,
+      "grad_norm": 5.284990419872968,
+      "learning_rate": 1.7605539774697244e-06,
+      "loss": 0.567,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6078128806952002,
+      "grad_norm": 7.660780757932287,
+      "learning_rate": 1.7599258051504856e-06,
+      "loss": 0.5853,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6078940956712418,
+      "grad_norm": 4.687240171351279,
+      "learning_rate": 1.7592976840404652e-06,
+      "loss": 0.4899,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6079753106472834,
+      "grad_norm": 4.470994352643192,
+      "learning_rate": 1.7586696141831242e-06,
+      "loss": 0.4768,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6080565256233249,
+      "grad_norm": 5.24599057514584,
+      "learning_rate": 1.7580415956219229e-06,
+      "loss": 0.3794,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6081377405993665,
+      "grad_norm": 4.900299331302958,
+      "learning_rate": 1.7574136284003158e-06,
+      "loss": 0.5302,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6082189555754081,
+      "grad_norm": 4.555248134756306,
+      "learning_rate": 1.756785712561756e-06,
+      "loss": 0.4212,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6083001705514497,
+      "grad_norm": 7.786367658307978,
+      "learning_rate": 1.7561578481496917e-06,
+      "loss": 0.4278,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6083813855274912,
+      "grad_norm": 4.5786486381903515,
+      "learning_rate": 1.7555300352075662e-06,
+      "loss": 0.6145,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6084626005035328,
+      "grad_norm": 7.245725994286198,
+      "learning_rate": 1.7549022737788241e-06,
+      "loss": 0.4126,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6085438154795745,
+      "grad_norm": 3.67634983056109,
+      "learning_rate": 1.7542745639069004e-06,
+      "loss": 0.5058,
+      "step": 7493
+    },
+    {
+      "epoch": 0.608625030455616,
+      "grad_norm": 9.370780365887464,
+      "learning_rate": 1.7536469056352296e-06,
+      "loss": 0.3767,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6087062454316576,
+      "grad_norm": 2.966365312799223,
+      "learning_rate": 1.7530192990072436e-06,
+      "loss": 0.5582,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6087874604076992,
+      "grad_norm": 5.219714565139399,
+      "learning_rate": 1.7523917440663687e-06,
+      "loss": 0.409,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6088686753837408,
+      "grad_norm": 7.275963784070534,
+      "learning_rate": 1.7517642408560278e-06,
+      "loss": 0.516,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6089498903597823,
+      "grad_norm": 5.53725602534592,
+      "learning_rate": 1.7511367894196426e-06,
+      "loss": 0.5271,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6090311053358239,
+      "grad_norm": 6.420027579961015,
+      "learning_rate": 1.7505093898006275e-06,
+      "loss": 0.4234,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6091123203118655,
+      "grad_norm": 5.134669425660005,
+      "learning_rate": 1.749882042042396e-06,
+      "loss": 0.5215,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6091935352879071,
+      "grad_norm": 3.969108933592243,
+      "learning_rate": 1.749254746188358e-06,
+      "loss": 0.4116,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6092747502639486,
+      "grad_norm": 9.697561858550591,
+      "learning_rate": 1.7486275022819183e-06,
+      "loss": 0.4249,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6093559652399902,
+      "grad_norm": 6.498690544614313,
+      "learning_rate": 1.748000310366478e-06,
+      "loss": 0.5176,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6094371802160319,
+      "grad_norm": 3.9324281115521202,
+      "learning_rate": 1.7473731704854363e-06,
+      "loss": 0.566,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6095183951920734,
+      "grad_norm": 8.158895435967443,
+      "learning_rate": 1.7467460826821885e-06,
+      "loss": 0.4352,
+      "step": 7505
+    },
+    {
+      "epoch": 0.609599610168115,
+      "grad_norm": 6.125377617707914,
+      "learning_rate": 1.7461190470001252e-06,
+      "loss": 0.4992,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6096808251441566,
+      "grad_norm": 4.208156614795222,
+      "learning_rate": 1.7454920634826334e-06,
+      "loss": 0.4183,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6097620401201982,
+      "grad_norm": 11.55413920993596,
+      "learning_rate": 1.7448651321730985e-06,
+      "loss": 0.4898,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6098432550962397,
+      "grad_norm": 4.906701536886049,
+      "learning_rate": 1.7442382531148993e-06,
+      "loss": 0.4855,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6099244700722813,
+      "grad_norm": 4.012360366132999,
+      "learning_rate": 1.743611426351413e-06,
+      "loss": 0.4437,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6100056850483229,
+      "grad_norm": 4.507122998130638,
+      "learning_rate": 1.7429846519260139e-06,
+      "loss": 0.4269,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6100869000243645,
+      "grad_norm": 7.688714435664871,
+      "learning_rate": 1.7423579298820698e-06,
+      "loss": 0.4297,
+      "step": 7512
+    },
+    {
+      "epoch": 0.610168115000406,
+      "grad_norm": 4.5702359534664145,
+      "learning_rate": 1.7417312602629466e-06,
+      "loss": 0.4187,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6102493299764477,
+      "grad_norm": 7.677278210562309,
+      "learning_rate": 1.7411046431120082e-06,
+      "loss": 0.4983,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6103305449524893,
+      "grad_norm": 6.1185795533491545,
+      "learning_rate": 1.7404780784726113e-06,
+      "loss": 0.6269,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6104117599285308,
+      "grad_norm": 14.544008771708137,
+      "learning_rate": 1.7398515663881117e-06,
+      "loss": 0.595,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6104929749045724,
+      "grad_norm": 10.989546459503224,
+      "learning_rate": 1.7392251069018612e-06,
+      "loss": 0.6077,
+      "step": 7517
+    },
+    {
+      "epoch": 0.610574189880614,
+      "grad_norm": 6.706531652553325,
+      "learning_rate": 1.7385987000572072e-06,
+      "loss": 0.558,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6106554048566556,
+      "grad_norm": 3.482810028814553,
+      "learning_rate": 1.7379723458974923e-06,
+      "loss": 0.4935,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6107366198326971,
+      "grad_norm": 4.216977832082727,
+      "learning_rate": 1.737346044466059e-06,
+      "loss": 0.5477,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6108178348087387,
+      "grad_norm": 5.407693817919376,
+      "learning_rate": 1.7367197958062432e-06,
+      "loss": 0.561,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6108990497847803,
+      "grad_norm": 7.6500405315944215,
+      "learning_rate": 1.7360935999613777e-06,
+      "loss": 0.7106,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6109802647608219,
+      "grad_norm": 8.727840725862155,
+      "learning_rate": 1.7354674569747914e-06,
+      "loss": 0.4518,
+      "step": 7523
+    },
+    {
+      "epoch": 0.6110614797368634,
+      "grad_norm": 6.081313481209265,
+      "learning_rate": 1.7348413668898124e-06,
+      "loss": 0.4906,
+      "step": 7524
+    },
+    {
+      "epoch": 0.611142694712905,
+      "grad_norm": 4.575030134738263,
+      "learning_rate": 1.73421532974976e-06,
+      "loss": 0.5392,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6112239096889467,
+      "grad_norm": 3.827031756918253,
+      "learning_rate": 1.7335893455979538e-06,
+      "loss": 0.5111,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6113051246649882,
+      "grad_norm": 4.399973290074878,
+      "learning_rate": 1.7329634144777097e-06,
+      "loss": 0.7858,
+      "step": 7527
+    },
+    {
+      "epoch": 0.6113863396410298,
+      "grad_norm": 6.275043877297012,
+      "learning_rate": 1.7323375364323374e-06,
+      "loss": 0.4,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6114675546170714,
+      "grad_norm": 3.7917580182798964,
+      "learning_rate": 1.731711711505144e-06,
+      "loss": 0.5784,
+      "step": 7529
+    },
+    {
+      "epoch": 0.611548769593113,
+      "grad_norm": 5.449393065833403,
+      "learning_rate": 1.7310859397394356e-06,
+      "loss": 0.4097,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6116299845691545,
+      "grad_norm": 9.087763572815122,
+      "learning_rate": 1.7304602211785105e-06,
+      "loss": 0.4269,
+      "step": 7531
+    },
+    {
+      "epoch": 0.6117111995451961,
+      "grad_norm": 3.9239408883458196,
+      "learning_rate": 1.7298345558656643e-06,
+      "loss": 0.4189,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6117924145212377,
+      "grad_norm": 3.9512250145217327,
+      "learning_rate": 1.7292089438441912e-06,
+      "loss": 0.5171,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6118736294972793,
+      "grad_norm": 5.225768936804722,
+      "learning_rate": 1.7285833851573802e-06,
+      "loss": 0.5148,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6119548444733209,
+      "grad_norm": 8.877862953830741,
+      "learning_rate": 1.727957879848516e-06,
+      "loss": 0.6574,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6120360594493625,
+      "grad_norm": 3.7513803990699457,
+      "learning_rate": 1.72733242796088e-06,
+      "loss": 0.4649,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6121172744254041,
+      "grad_norm": 6.526118406530345,
+      "learning_rate": 1.7267070295377519e-06,
+      "loss": 0.5629,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6121984894014456,
+      "grad_norm": 8.83748922556791,
+      "learning_rate": 1.726081684622404e-06,
+      "loss": 0.483,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6122797043774872,
+      "grad_norm": 4.032292821062164,
+      "learning_rate": 1.7254563932581072e-06,
+      "loss": 0.6093,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6123609193535288,
+      "grad_norm": 19.261864290456085,
+      "learning_rate": 1.7248311554881297e-06,
+      "loss": 0.3226,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6124421343295704,
+      "grad_norm": 6.178514406690231,
+      "learning_rate": 1.7242059713557336e-06,
+      "loss": 0.4116,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6125233493056119,
+      "grad_norm": 6.272920000009786,
+      "learning_rate": 1.7235808409041775e-06,
+      "loss": 0.514,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6126045642816536,
+      "grad_norm": 3.700775958812784,
+      "learning_rate": 1.7229557641767191e-06,
+      "loss": 0.3956,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6126857792576951,
+      "grad_norm": 5.184180318926046,
+      "learning_rate": 1.7223307412166097e-06,
+      "loss": 0.5573,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6127669942337367,
+      "grad_norm": 7.17746460216716,
+      "learning_rate": 1.7217057720670955e-06,
+      "loss": 0.4188,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6128482092097783,
+      "grad_norm": 5.19044747938764,
+      "learning_rate": 1.7210808567714244e-06,
+      "loss": 0.6008,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6129294241858199,
+      "grad_norm": 6.637207987486905,
+      "learning_rate": 1.7204559953728355e-06,
+      "loss": 0.4163,
+      "step": 7547
+    },
+    {
+      "epoch": 0.6130106391618615,
+      "grad_norm": 4.123118258455451,
+      "learning_rate": 1.7198311879145652e-06,
+      "loss": 0.4583,
+      "step": 7548
+    },
+    {
+      "epoch": 0.613091854137903,
+      "grad_norm": 4.634690421136184,
+      "learning_rate": 1.719206434439848e-06,
+      "loss": 0.6119,
+      "step": 7549
+    },
+    {
+      "epoch": 0.6131730691139446,
+      "grad_norm": 5.330581154155666,
+      "learning_rate": 1.7185817349919137e-06,
+      "loss": 0.679,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6132542840899862,
+      "grad_norm": 4.733672202157837,
+      "learning_rate": 1.7179570896139869e-06,
+      "loss": 0.5463,
+      "step": 7551
+    },
+    {
+      "epoch": 0.6133354990660278,
+      "grad_norm": 3.417972262827162,
+      "learning_rate": 1.7173324983492912e-06,
+      "loss": 0.5625,
+      "step": 7552
+    },
+    {
+      "epoch": 0.6134167140420693,
+      "grad_norm": 5.5606216442015155,
+      "learning_rate": 1.7167079612410448e-06,
+      "loss": 0.541,
+      "step": 7553
+    },
+    {
+      "epoch": 0.613497929018111,
+      "grad_norm": 7.366074745822286,
+      "learning_rate": 1.7160834783324608e-06,
+      "loss": 0.5007,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6135791439941525,
+      "grad_norm": 4.4511321208599925,
+      "learning_rate": 1.7154590496667523e-06,
+      "loss": 0.4106,
+      "step": 7555
+    },
+    {
+      "epoch": 0.6136603589701941,
+      "grad_norm": 4.380414416940069,
+      "learning_rate": 1.7148346752871253e-06,
+      "loss": 0.5297,
+      "step": 7556
+    },
+    {
+      "epoch": 0.6137415739462357,
+      "grad_norm": 5.789276347920089,
+      "learning_rate": 1.7142103552367834e-06,
+      "loss": 0.4956,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6138227889222773,
+      "grad_norm": 5.435165522246494,
+      "learning_rate": 1.713586089558925e-06,
+      "loss": 0.5016,
+      "step": 7558
+    },
+    {
+      "epoch": 0.6139040038983189,
+      "grad_norm": 3.8853370254417796,
+      "learning_rate": 1.7129618782967488e-06,
+      "loss": 0.4796,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6139852188743604,
+      "grad_norm": 6.1117435269535765,
+      "learning_rate": 1.712337721493445e-06,
+      "loss": 0.3512,
+      "step": 7560
+    },
+    {
+      "epoch": 0.614066433850402,
+      "grad_norm": 5.140667780868619,
+      "learning_rate": 1.7117136191922013e-06,
+      "loss": 0.383,
+      "step": 7561
+    },
+    {
+      "epoch": 0.6141476488264436,
+      "grad_norm": 5.239177233517926,
+      "learning_rate": 1.7110895714362035e-06,
+      "loss": 0.4811,
+      "step": 7562
+    },
+    {
+      "epoch": 0.6142288638024852,
+      "grad_norm": 5.786445992631991,
+      "learning_rate": 1.710465578268633e-06,
+      "loss": 0.5776,
+      "step": 7563
+    },
+    {
+      "epoch": 0.6143100787785267,
+      "grad_norm": 4.264725420672594,
+      "learning_rate": 1.7098416397326647e-06,
+      "loss": 0.548,
+      "step": 7564
+    },
+    {
+      "epoch": 0.6143912937545684,
+      "grad_norm": 5.098547500018314,
+      "learning_rate": 1.7092177558714735e-06,
+      "loss": 0.4055,
+      "step": 7565
+    },
+    {
+      "epoch": 0.6144725087306099,
+      "grad_norm": 3.972712496253238,
+      "learning_rate": 1.7085939267282292e-06,
+      "loss": 0.5024,
+      "step": 7566
+    },
+    {
+      "epoch": 0.6145537237066515,
+      "grad_norm": 6.4566948580335835,
+      "learning_rate": 1.7079701523460957e-06,
+      "loss": 0.3837,
+      "step": 7567
+    },
+    {
+      "epoch": 0.6146349386826931,
+      "grad_norm": 4.440870142659438,
+      "learning_rate": 1.707346432768236e-06,
+      "loss": 0.5221,
+      "step": 7568
+    },
+    {
+      "epoch": 0.6147161536587347,
+      "grad_norm": 4.191872882758366,
+      "learning_rate": 1.706722768037809e-06,
+      "loss": 0.5073,
+      "step": 7569
+    },
+    {
+      "epoch": 0.6147973686347763,
+      "grad_norm": 5.091945928993117,
+      "learning_rate": 1.7060991581979668e-06,
+      "loss": 0.4276,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6148785836108178,
+      "grad_norm": 4.128925636949817,
+      "learning_rate": 1.7054756032918619e-06,
+      "loss": 0.5327,
+      "step": 7571
+    },
+    {
+      "epoch": 0.6149597985868595,
+      "grad_norm": 4.6497677030599265,
+      "learning_rate": 1.7048521033626406e-06,
+      "loss": 0.4233,
+      "step": 7572
+    },
+    {
+      "epoch": 0.615041013562901,
+      "grad_norm": 5.259086945381799,
+      "learning_rate": 1.7042286584534446e-06,
+      "loss": 0.5063,
+      "step": 7573
+    },
+    {
+      "epoch": 0.6151222285389426,
+      "grad_norm": 9.344094684965398,
+      "learning_rate": 1.703605268607415e-06,
+      "loss": 0.4642,
+      "step": 7574
+    },
+    {
+      "epoch": 0.6152034435149841,
+      "grad_norm": 6.059064695851907,
+      "learning_rate": 1.7029819338676851e-06,
+      "loss": 0.5159,
+      "step": 7575
+    },
+    {
+      "epoch": 0.6152846584910258,
+      "grad_norm": 7.950438579454252,
+      "learning_rate": 1.702358654277388e-06,
+      "loss": 0.4775,
+      "step": 7576
+    },
+    {
+      "epoch": 0.6153658734670673,
+      "grad_norm": 5.6577857752289695,
+      "learning_rate": 1.7017354298796495e-06,
+      "loss": 0.4521,
+      "step": 7577
+    },
+    {
+      "epoch": 0.6154470884431089,
+      "grad_norm": 6.694027039647233,
+      "learning_rate": 1.701112260717595e-06,
+      "loss": 0.5675,
+      "step": 7578
+    },
+    {
+      "epoch": 0.6155283034191505,
+      "grad_norm": 7.207274495499145,
+      "learning_rate": 1.7004891468343445e-06,
+      "loss": 0.4767,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6156095183951921,
+      "grad_norm": 10.710533616510515,
+      "learning_rate": 1.6998660882730127e-06,
+      "loss": 0.4575,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6156907333712337,
+      "grad_norm": 9.051943184654498,
+      "learning_rate": 1.6992430850767133e-06,
+      "loss": 0.4646,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6157719483472752,
+      "grad_norm": 6.318629144815555,
+      "learning_rate": 1.6986201372885551e-06,
+      "loss": 0.5112,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6158531633233169,
+      "grad_norm": 9.374156721440105,
+      "learning_rate": 1.6979972449516414e-06,
+      "loss": 0.5938,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6159343782993584,
+      "grad_norm": 5.954834061127918,
+      "learning_rate": 1.6973744081090737e-06,
+      "loss": 0.5195,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6160155932754,
+      "grad_norm": 4.686975998201248,
+      "learning_rate": 1.6967516268039502e-06,
+      "loss": 0.5192,
+      "step": 7585
+    },
+    {
+      "epoch": 0.6160968082514415,
+      "grad_norm": 5.255217548237502,
+      "learning_rate": 1.696128901079362e-06,
+      "loss": 0.5776,
+      "step": 7586
+    },
+    {
+      "epoch": 0.6161780232274832,
+      "grad_norm": 3.9588121129361915,
+      "learning_rate": 1.6955062309783993e-06,
+      "loss": 0.4449,
+      "step": 7587
+    },
+    {
+      "epoch": 0.6162592382035247,
+      "grad_norm": 2.910107855866022,
+      "learning_rate": 1.6948836165441487e-06,
+      "loss": 0.6015,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6163404531795663,
+      "grad_norm": 4.0433138018943735,
+      "learning_rate": 1.6942610578196898e-06,
+      "loss": 0.4804,
+      "step": 7589
+    },
+    {
+      "epoch": 0.616421668155608,
+      "grad_norm": 12.510674806813109,
+      "learning_rate": 1.6936385548481022e-06,
+      "loss": 0.3454,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6165028831316495,
+      "grad_norm": 4.596018618309467,
+      "learning_rate": 1.6930161076724586e-06,
+      "loss": 0.4479,
+      "step": 7591
+    },
+    {
+      "epoch": 0.6165840981076911,
+      "grad_norm": 4.171966658619839,
+      "learning_rate": 1.69239371633583e-06,
+      "loss": 0.5211,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6166653130837326,
+      "grad_norm": 8.771836978846274,
+      "learning_rate": 1.6917713808812808e-06,
+      "loss": 0.4352,
+      "step": 7593
+    },
+    {
+      "epoch": 0.6167465280597743,
+      "grad_norm": 4.139812780948147,
+      "learning_rate": 1.6911491013518752e-06,
+      "loss": 0.5914,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6168277430358158,
+      "grad_norm": 5.236782993816778,
+      "learning_rate": 1.6905268777906713e-06,
+      "loss": 0.5286,
+      "step": 7595
+    },
+    {
+      "epoch": 0.6169089580118574,
+      "grad_norm": 5.408332820716584,
+      "learning_rate": 1.6899047102407228e-06,
+      "loss": 0.3983,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6169901729878989,
+      "grad_norm": 3.729915257776433,
+      "learning_rate": 1.6892825987450811e-06,
+      "loss": 0.5177,
+      "step": 7597
+    },
+    {
+      "epoch": 0.6170713879639406,
+      "grad_norm": 4.845431694776476,
+      "learning_rate": 1.6886605433467937e-06,
+      "loss": 0.5994,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6171526029399821,
+      "grad_norm": 4.28204183668613,
+      "learning_rate": 1.6880385440889016e-06,
+      "loss": 0.4965,
+      "step": 7599
+    },
+    {
+      "epoch": 0.6172338179160237,
+      "grad_norm": 3.81321254807878,
+      "learning_rate": 1.6874166010144454e-06,
+      "loss": 0.5212,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6173150328920654,
+      "grad_norm": 4.863898183735516,
+      "learning_rate": 1.6867947141664606e-06,
+      "loss": 0.6291,
+      "step": 7601
+    },
+    {
+      "epoch": 0.6173962478681069,
+      "grad_norm": 3.908697856611701,
+      "learning_rate": 1.6861728835879764e-06,
+      "loss": 0.5934,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6174774628441485,
+      "grad_norm": 4.930386518701432,
+      "learning_rate": 1.685551109322023e-06,
+      "loss": 0.4434,
+      "step": 7603
+    },
+    {
+      "epoch": 0.61755867782019,
+      "grad_norm": 3.7606290186734586,
+      "learning_rate": 1.6849293914116215e-06,
+      "loss": 0.5494,
+      "step": 7604
+    },
+    {
+      "epoch": 0.6176398927962317,
+      "grad_norm": 4.342048831757059,
+      "learning_rate": 1.6843077298997924e-06,
+      "loss": 0.5169,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6177211077722732,
+      "grad_norm": 5.829444571059767,
+      "learning_rate": 1.6836861248295522e-06,
+      "loss": 0.5501,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6178023227483148,
+      "grad_norm": 4.7706445198897915,
+      "learning_rate": 1.6830645762439113e-06,
+      "loss": 0.4783,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6178835377243563,
+      "grad_norm": 8.032579213522444,
+      "learning_rate": 1.6824430841858773e-06,
+      "loss": 0.4552,
+      "step": 7608
+    },
+    {
+      "epoch": 0.617964752700398,
+      "grad_norm": 6.670715215785492,
+      "learning_rate": 1.6818216486984565e-06,
+      "loss": 0.4983,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6180459676764395,
+      "grad_norm": 6.128265762260737,
+      "learning_rate": 1.6812002698246468e-06,
+      "loss": 0.4088,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6181271826524811,
+      "grad_norm": 5.814013730186966,
+      "learning_rate": 1.6805789476074457e-06,
+      "loss": 0.5589,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6182083976285228,
+      "grad_norm": 6.00192645217164,
+      "learning_rate": 1.6799576820898433e-06,
+      "loss": 0.4175,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6182896126045643,
+      "grad_norm": 4.073976353754178,
+      "learning_rate": 1.6793364733148299e-06,
+      "loss": 0.3606,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6183708275806059,
+      "grad_norm": 4.36296820738616,
+      "learning_rate": 1.67871532132539e-06,
+      "loss": 0.6092,
+      "step": 7614
+    },
+    {
+      "epoch": 0.6184520425566474,
+      "grad_norm": 6.684933223291455,
+      "learning_rate": 1.6780942261645022e-06,
+      "loss": 0.4535,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6185332575326891,
+      "grad_norm": 4.124907550937014,
+      "learning_rate": 1.6774731878751443e-06,
+      "loss": 0.5198,
+      "step": 7616
+    },
+    {
+      "epoch": 0.6186144725087306,
+      "grad_norm": 2.736304544029034,
+      "learning_rate": 1.6768522065002895e-06,
+      "loss": 0.6509,
+      "step": 7617
+    },
+    {
+      "epoch": 0.6186956874847722,
+      "grad_norm": 3.054312384173202,
+      "learning_rate": 1.676231282082904e-06,
+      "loss": 0.5161,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6187769024608137,
+      "grad_norm": 8.402180136218686,
+      "learning_rate": 1.6756104146659557e-06,
+      "loss": 0.394,
+      "step": 7619
+    },
+    {
+      "epoch": 0.6188581174368554,
+      "grad_norm": 4.6791235959713315,
+      "learning_rate": 1.674989604292403e-06,
+      "loss": 0.6243,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6189393324128969,
+      "grad_norm": 9.604477306580446,
+      "learning_rate": 1.6743688510052025e-06,
+      "loss": 0.4754,
+      "step": 7621
+    },
+    {
+      "epoch": 0.6190205473889385,
+      "grad_norm": 9.89281790814999,
+      "learning_rate": 1.6737481548473094e-06,
+      "loss": 0.4893,
+      "step": 7622
+    },
+    {
+      "epoch": 0.6191017623649802,
+      "grad_norm": 3.3991457520981383,
+      "learning_rate": 1.6731275158616706e-06,
+      "loss": 0.457,
+      "step": 7623
+    },
+    {
+      "epoch": 0.6191829773410217,
+      "grad_norm": 2.80593016197633,
+      "learning_rate": 1.6725069340912306e-06,
+      "loss": 0.6278,
+      "step": 7624
+    },
+    {
+      "epoch": 0.6192641923170633,
+      "grad_norm": 5.418301042196,
+      "learning_rate": 1.6718864095789328e-06,
+      "loss": 0.4187,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6193454072931048,
+      "grad_norm": 13.06513067077772,
+      "learning_rate": 1.671265942367712e-06,
+      "loss": 0.4663,
+      "step": 7626
+    },
+    {
+      "epoch": 0.6194266222691465,
+      "grad_norm": 4.625063932915541,
+      "learning_rate": 1.6706455325005022e-06,
+      "loss": 0.4295,
+      "step": 7627
+    },
+    {
+      "epoch": 0.619507837245188,
+      "grad_norm": 3.5014064055098175,
+      "learning_rate": 1.6700251800202316e-06,
+      "loss": 0.4579,
+      "step": 7628
+    },
+    {
+      "epoch": 0.6195890522212296,
+      "grad_norm": 6.56770323902514,
+      "learning_rate": 1.6694048849698262e-06,
+      "loss": 0.6269,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6196702671972711,
+      "grad_norm": 4.07016650700423,
+      "learning_rate": 1.668784647392208e-06,
+      "loss": 0.6907,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6197514821733128,
+      "grad_norm": 5.113480533421879,
+      "learning_rate": 1.6681644673302915e-06,
+      "loss": 0.5111,
+      "step": 7631
+    },
+    {
+      "epoch": 0.6198326971493543,
+      "grad_norm": 4.254910794569682,
+      "learning_rate": 1.6675443448269924e-06,
+      "loss": 0.5239,
+      "step": 7632
+    },
+    {
+      "epoch": 0.6199139121253959,
+      "grad_norm": 5.408424083847769,
+      "learning_rate": 1.666924279925219e-06,
+      "loss": 0.5857,
+      "step": 7633
+    },
+    {
+      "epoch": 0.6199951271014376,
+      "grad_norm": 3.2011054128203766,
+      "learning_rate": 1.6663042726678752e-06,
+      "loss": 0.5784,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6200763420774791,
+      "grad_norm": 5.235602080991919,
+      "learning_rate": 1.6656843230978647e-06,
+      "loss": 0.3975,
+      "step": 7635
+    },
+    {
+      "epoch": 0.6201575570535207,
+      "grad_norm": 5.165250298024449,
+      "learning_rate": 1.6650644312580833e-06,
+      "loss": 0.5924,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6202387720295622,
+      "grad_norm": 4.340672727279216,
+      "learning_rate": 1.6644445971914235e-06,
+      "loss": 0.5221,
+      "step": 7637
+    },
+    {
+      "epoch": 0.6203199870056039,
+      "grad_norm": 6.841912089854011,
+      "learning_rate": 1.6638248209407767e-06,
+      "loss": 0.443,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6204012019816454,
+      "grad_norm": 4.081953440344846,
+      "learning_rate": 1.6632051025490265e-06,
+      "loss": 0.3495,
+      "step": 7639
+    },
+    {
+      "epoch": 0.620482416957687,
+      "grad_norm": 3.7772718364769142,
+      "learning_rate": 1.6625854420590538e-06,
+      "loss": 0.5034,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6205636319337285,
+      "grad_norm": 5.210380997814291,
+      "learning_rate": 1.6619658395137375e-06,
+      "loss": 0.6397,
+      "step": 7641
+    },
+    {
+      "epoch": 0.6206448469097702,
+      "grad_norm": 6.577887175557655,
+      "learning_rate": 1.6613462949559494e-06,
+      "loss": 0.4802,
+      "step": 7642
+    },
+    {
+      "epoch": 0.6207260618858117,
+      "grad_norm": 3.7841172613723657,
+      "learning_rate": 1.6607268084285587e-06,
+      "loss": 0.7127,
+      "step": 7643
+    },
+    {
+      "epoch": 0.6208072768618533,
+      "grad_norm": 8.73460364920805,
+      "learning_rate": 1.6601073799744322e-06,
+      "loss": 0.4822,
+      "step": 7644
+    },
+    {
+      "epoch": 0.620888491837895,
+      "grad_norm": 5.904973312911216,
+      "learning_rate": 1.6594880096364302e-06,
+      "loss": 0.5566,
+      "step": 7645
+    },
+    {
+      "epoch": 0.6209697068139365,
+      "grad_norm": 4.941626996148981,
+      "learning_rate": 1.6588686974574086e-06,
+      "loss": 0.3796,
+      "step": 7646
+    },
+    {
+      "epoch": 0.6210509217899781,
+      "grad_norm": 5.8291085465722325,
+      "learning_rate": 1.658249443480221e-06,
+      "loss": 0.4666,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6211321367660196,
+      "grad_norm": 5.909142689212799,
+      "learning_rate": 1.6576302477477185e-06,
+      "loss": 0.4453,
+      "step": 7648
+    },
+    {
+      "epoch": 0.6212133517420613,
+      "grad_norm": 5.951228309961236,
+      "learning_rate": 1.6570111103027436e-06,
+      "loss": 0.3766,
+      "step": 7649
+    },
+    {
+      "epoch": 0.6212945667181028,
+      "grad_norm": 6.681405040737811,
+      "learning_rate": 1.6563920311881382e-06,
+      "loss": 0.3992,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6213757816941444,
+      "grad_norm": 5.5260582323770135,
+      "learning_rate": 1.6557730104467407e-06,
+      "loss": 0.4259,
+      "step": 7651
+    },
+    {
+      "epoch": 0.621456996670186,
+      "grad_norm": 5.076655366887786,
+      "learning_rate": 1.6551540481213817e-06,
+      "loss": 0.5468,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6215382116462276,
+      "grad_norm": 5.8919929585227155,
+      "learning_rate": 1.6545351442548915e-06,
+      "loss": 0.4871,
+      "step": 7653
+    },
+    {
+      "epoch": 0.6216194266222691,
+      "grad_norm": 4.042380796567633,
+      "learning_rate": 1.6539162988900952e-06,
+      "loss": 0.4153,
+      "step": 7654
+    },
+    {
+      "epoch": 0.6217006415983107,
+      "grad_norm": 6.784897682604056,
+      "learning_rate": 1.6532975120698133e-06,
+      "loss": 0.4534,
+      "step": 7655
+    },
+    {
+      "epoch": 0.6217818565743524,
+      "grad_norm": 9.65234999480218,
+      "learning_rate": 1.6526787838368616e-06,
+      "loss": 0.4155,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6218630715503939,
+      "grad_norm": 17.08200480716379,
+      "learning_rate": 1.6520601142340549e-06,
+      "loss": 0.5921,
+      "step": 7657
+    },
+    {
+      "epoch": 0.6219442865264355,
+      "grad_norm": 5.175377372780213,
+      "learning_rate": 1.6514415033041997e-06,
+      "loss": 0.4526,
+      "step": 7658
+    },
+    {
+      "epoch": 0.622025501502477,
+      "grad_norm": 3.458058541785411,
+      "learning_rate": 1.6508229510901013e-06,
+      "loss": 0.4555,
+      "step": 7659
+    },
+    {
+      "epoch": 0.6221067164785187,
+      "grad_norm": 10.604824315531577,
+      "learning_rate": 1.6502044576345614e-06,
+      "loss": 0.452,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6221879314545602,
+      "grad_norm": 5.411425272158323,
+      "learning_rate": 1.6495860229803756e-06,
+      "loss": 0.6222,
+      "step": 7661
+    },
+    {
+      "epoch": 0.6222691464306018,
+      "grad_norm": 5.1740005505908995,
+      "learning_rate": 1.6489676471703352e-06,
+      "loss": 0.4145,
+      "step": 7662
+    },
+    {
+      "epoch": 0.6223503614066433,
+      "grad_norm": 8.085499523946917,
+      "learning_rate": 1.6483493302472302e-06,
+      "loss": 0.513,
+      "step": 7663
+    },
+    {
+      "epoch": 0.622431576382685,
+      "grad_norm": 10.292791376029198,
+      "learning_rate": 1.6477310722538447e-06,
+      "loss": 0.6126,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6225127913587265,
+      "grad_norm": 6.0268201310361995,
+      "learning_rate": 1.6471128732329579e-06,
+      "loss": 0.4292,
+      "step": 7665
+    },
+    {
+      "epoch": 0.6225940063347681,
+      "grad_norm": 5.022456342536074,
+      "learning_rate": 1.6464947332273459e-06,
+      "loss": 0.6152,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6226752213108098,
+      "grad_norm": 8.543190424818972,
+      "learning_rate": 1.6458766522797822e-06,
+      "loss": 0.6125,
+      "step": 7667
+    },
+    {
+      "epoch": 0.6227564362868513,
+      "grad_norm": 4.1156125031275135,
+      "learning_rate": 1.6452586304330333e-06,
+      "loss": 0.462,
+      "step": 7668
+    },
+    {
+      "epoch": 0.6228376512628929,
+      "grad_norm": 7.81665689239009,
+      "learning_rate": 1.6446406677298632e-06,
+      "loss": 0.4678,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6229188662389344,
+      "grad_norm": 6.949544264134675,
+      "learning_rate": 1.644022764213033e-06,
+      "loss": 0.4414,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6230000812149761,
+      "grad_norm": 8.196141138164219,
+      "learning_rate": 1.6434049199252966e-06,
+      "loss": 0.4455,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6230812961910176,
+      "grad_norm": 6.110724030611538,
+      "learning_rate": 1.6427871349094058e-06,
+      "loss": 0.3828,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6231625111670592,
+      "grad_norm": 4.134192350462065,
+      "learning_rate": 1.6421694092081097e-06,
+      "loss": 0.5077,
+      "step": 7673
+    },
+    {
+      "epoch": 0.6232437261431008,
+      "grad_norm": 4.263913374724854,
+      "learning_rate": 1.6415517428641504e-06,
+      "loss": 0.6757,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6233249411191424,
+      "grad_norm": 6.728733308443024,
+      "learning_rate": 1.640934135920266e-06,
+      "loss": 0.485,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6234061560951839,
+      "grad_norm": 6.332399518498263,
+      "learning_rate": 1.6403165884191935e-06,
+      "loss": 0.4495,
+      "step": 7676
+    },
+    {
+      "epoch": 0.6234873710712255,
+      "grad_norm": 5.132785407111167,
+      "learning_rate": 1.6396991004036638e-06,
+      "loss": 0.4718,
+      "step": 7677
+    },
+    {
+      "epoch": 0.6235685860472672,
+      "grad_norm": 4.651566891974008,
+      "learning_rate": 1.6390816719164022e-06,
+      "loss": 0.5763,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6236498010233087,
+      "grad_norm": 4.413484601105079,
+      "learning_rate": 1.6384643030001333e-06,
+      "loss": 0.722,
+      "step": 7679
+    },
+    {
+      "epoch": 0.6237310159993503,
+      "grad_norm": 8.71808254189496,
+      "learning_rate": 1.6378469936975752e-06,
+      "loss": 0.5068,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6238122309753918,
+      "grad_norm": 6.479840818255055,
+      "learning_rate": 1.6372297440514417e-06,
+      "loss": 0.4299,
+      "step": 7681
+    },
+    {
+      "epoch": 0.6238934459514335,
+      "grad_norm": 3.683453972916311,
+      "learning_rate": 1.6366125541044435e-06,
+      "loss": 0.545,
+      "step": 7682
+    },
+    {
+      "epoch": 0.623974660927475,
+      "grad_norm": 9.69410298079771,
+      "learning_rate": 1.6359954238992882e-06,
+      "loss": 0.3748,
+      "step": 7683
+    },
+    {
+      "epoch": 0.6240558759035166,
+      "grad_norm": 4.97119687493066,
+      "learning_rate": 1.6353783534786763e-06,
+      "loss": 0.4015,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6241370908795582,
+      "grad_norm": 7.7545114750163,
+      "learning_rate": 1.6347613428853059e-06,
+      "loss": 0.4756,
+      "step": 7685
+    },
+    {
+      "epoch": 0.6242183058555998,
+      "grad_norm": 3.5468434651588976,
+      "learning_rate": 1.634144392161872e-06,
+      "loss": 0.5798,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6242995208316413,
+      "grad_norm": 4.706606608660131,
+      "learning_rate": 1.6335275013510638e-06,
+      "loss": 0.5285,
+      "step": 7687
+    },
+    {
+      "epoch": 0.6243807358076829,
+      "grad_norm": 6.147626373688174,
+      "learning_rate": 1.632910670495566e-06,
+      "loss": 0.3421,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6244619507837246,
+      "grad_norm": 15.303934361158852,
+      "learning_rate": 1.6322938996380617e-06,
+      "loss": 0.46,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6245431657597661,
+      "grad_norm": 14.350203161237948,
+      "learning_rate": 1.6316771888212275e-06,
+      "loss": 0.4245,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6246243807358077,
+      "grad_norm": 4.190438584869176,
+      "learning_rate": 1.631060538087735e-06,
+      "loss": 0.566,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6247055957118492,
+      "grad_norm": 3.82815155289053,
+      "learning_rate": 1.6304439474802554e-06,
+      "loss": 0.4528,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6247868106878909,
+      "grad_norm": 6.351971066286823,
+      "learning_rate": 1.6298274170414524e-06,
+      "loss": 0.5233,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6248680256639324,
+      "grad_norm": 4.36724324752686,
+      "learning_rate": 1.6292109468139863e-06,
+      "loss": 0.4774,
+      "step": 7694
+    },
+    {
+      "epoch": 0.624949240639974,
+      "grad_norm": 6.655112537446748,
+      "learning_rate": 1.6285945368405146e-06,
+      "loss": 0.5333,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6250304556160156,
+      "grad_norm": 6.018071010184688,
+      "learning_rate": 1.6279781871636896e-06,
+      "loss": 0.4432,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6251116705920572,
+      "grad_norm": 5.306711764250664,
+      "learning_rate": 1.6273618978261576e-06,
+      "loss": 0.5138,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6251928855680987,
+      "grad_norm": 3.9883932484916325,
+      "learning_rate": 1.6267456688705647e-06,
+      "loss": 0.5547,
+      "step": 7698
+    },
+    {
+      "epoch": 0.6252741005441403,
+      "grad_norm": 6.6197356790826625,
+      "learning_rate": 1.6261295003395506e-06,
+      "loss": 0.4235,
+      "step": 7699
+    },
+    {
+      "epoch": 0.625355315520182,
+      "grad_norm": 3.6076841841661667,
+      "learning_rate": 1.6255133922757493e-06,
+      "loss": 0.4933,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6254365304962235,
+      "grad_norm": 4.929421788606065,
+      "learning_rate": 1.6248973447217926e-06,
+      "loss": 0.5483,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6255177454722651,
+      "grad_norm": 4.242914786049469,
+      "learning_rate": 1.6242813577203093e-06,
+      "loss": 0.4914,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6255989604483067,
+      "grad_norm": 4.388993067736639,
+      "learning_rate": 1.6236654313139213e-06,
+      "loss": 0.428,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6256801754243483,
+      "grad_norm": 6.684554645774749,
+      "learning_rate": 1.6230495655452466e-06,
+      "loss": 0.5199,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6257613904003898,
+      "grad_norm": 4.683366955828708,
+      "learning_rate": 1.6224337604569012e-06,
+      "loss": 0.4485,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6258426053764314,
+      "grad_norm": 4.786816490822921,
+      "learning_rate": 1.6218180160914959e-06,
+      "loss": 0.4682,
+      "step": 7706
+    },
+    {
+      "epoch": 0.625923820352473,
+      "grad_norm": 5.411731210706411,
+      "learning_rate": 1.6212023324916349e-06,
+      "loss": 0.423,
+      "step": 7707
+    },
+    {
+      "epoch": 0.6260050353285146,
+      "grad_norm": 4.143891933212883,
+      "learning_rate": 1.620586709699922e-06,
+      "loss": 0.7241,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6260862503045561,
+      "grad_norm": 5.695424075779182,
+      "learning_rate": 1.6199711477589553e-06,
+      "loss": 0.4617,
+      "step": 7709
+    },
+    {
+      "epoch": 0.6261674652805977,
+      "grad_norm": 7.132319497435146,
+      "learning_rate": 1.6193556467113264e-06,
+      "loss": 0.3636,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6262486802566394,
+      "grad_norm": 7.5139687849034225,
+      "learning_rate": 1.6187402065996267e-06,
+      "loss": 0.4507,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6263298952326809,
+      "grad_norm": 3.7215878058767675,
+      "learning_rate": 1.6181248274664413e-06,
+      "loss": 0.5443,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6264111102087225,
+      "grad_norm": 3.861480100387473,
+      "learning_rate": 1.617509509354349e-06,
+      "loss": 0.4875,
+      "step": 7713
+    },
+    {
+      "epoch": 0.626492325184764,
+      "grad_norm": 5.924413160412586,
+      "learning_rate": 1.616894252305929e-06,
+      "loss": 0.5528,
+      "step": 7714
+    },
+    {
+      "epoch": 0.6265735401608057,
+      "grad_norm": 6.36026377881963,
+      "learning_rate": 1.6162790563637538e-06,
+      "loss": 0.4045,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6266547551368472,
+      "grad_norm": 4.4919168460790475,
+      "learning_rate": 1.6156639215703896e-06,
+      "loss": 0.4531,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6267359701128888,
+      "grad_norm": 3.3752966345250583,
+      "learning_rate": 1.6150488479684022e-06,
+      "loss": 0.4291,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6268171850889304,
+      "grad_norm": 4.683411418815497,
+      "learning_rate": 1.6144338356003513e-06,
+      "loss": 0.5847,
+      "step": 7718
+    },
+    {
+      "epoch": 0.626898400064972,
+      "grad_norm": 6.26290414561336,
+      "learning_rate": 1.6138188845087926e-06,
+      "loss": 0.4307,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6269796150410135,
+      "grad_norm": 7.127187742803709,
+      "learning_rate": 1.613203994736276e-06,
+      "loss": 0.6484,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6270608300170551,
+      "grad_norm": 4.465338617099092,
+      "learning_rate": 1.61258916632535e-06,
+      "loss": 0.4218,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6271420449930968,
+      "grad_norm": 6.635578415699874,
+      "learning_rate": 1.6119743993185574e-06,
+      "loss": 0.5503,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6272232599691383,
+      "grad_norm": 9.293378886349172,
+      "learning_rate": 1.6113596937584358e-06,
+      "loss": 0.4743,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6273044749451799,
+      "grad_norm": 3.9282722976253432,
+      "learning_rate": 1.610745049687521e-06,
+      "loss": 0.4391,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6273856899212215,
+      "grad_norm": 5.900142462141413,
+      "learning_rate": 1.6101304671483425e-06,
+      "loss": 0.4308,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6274669048972631,
+      "grad_norm": 5.538553572045406,
+      "learning_rate": 1.6095159461834252e-06,
+      "loss": 0.3723,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6275481198733046,
+      "grad_norm": 5.156039599008567,
+      "learning_rate": 1.6089014868352925e-06,
+      "loss": 0.5306,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6276293348493462,
+      "grad_norm": 5.634533236318975,
+      "learning_rate": 1.608287089146461e-06,
+      "loss": 0.525,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6277105498253878,
+      "grad_norm": 5.598968838519203,
+      "learning_rate": 1.6076727531594428e-06,
+      "loss": 0.3932,
+      "step": 7729
+    },
+    {
+      "epoch": 0.6277917648014294,
+      "grad_norm": 5.654175409392677,
+      "learning_rate": 1.607058478916748e-06,
+      "loss": 0.4358,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6278729797774709,
+      "grad_norm": 7.282317038390533,
+      "learning_rate": 1.6064442664608808e-06,
+      "loss": 0.6522,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6279541947535126,
+      "grad_norm": 6.6301535369071924,
+      "learning_rate": 1.6058301158343408e-06,
+      "loss": 0.4913,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6280354097295542,
+      "grad_norm": 4.56677620849299,
+      "learning_rate": 1.6052160270796252e-06,
+      "loss": 0.4066,
+      "step": 7733
+    },
+    {
+      "epoch": 0.6281166247055957,
+      "grad_norm": 6.930735350423143,
+      "learning_rate": 1.6046020002392242e-06,
+      "loss": 0.6191,
+      "step": 7734
+    },
+    {
+      "epoch": 0.6281978396816373,
+      "grad_norm": 3.3385261614929584,
+      "learning_rate": 1.603988035355627e-06,
+      "loss": 0.4239,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6282790546576789,
+      "grad_norm": 6.108614723656081,
+      "learning_rate": 1.6033741324713143e-06,
+      "loss": 0.5148,
+      "step": 7736
+    },
+    {
+      "epoch": 0.6283602696337205,
+      "grad_norm": 5.083243850769626,
+      "learning_rate": 1.6027602916287665e-06,
+      "loss": 0.5385,
+      "step": 7737
+    },
+    {
+      "epoch": 0.628441484609762,
+      "grad_norm": 4.3659844044456735,
+      "learning_rate": 1.6021465128704592e-06,
+      "loss": 0.4434,
+      "step": 7738
+    },
+    {
+      "epoch": 0.6285226995858036,
+      "grad_norm": 3.246650947537016,
+      "learning_rate": 1.60153279623886e-06,
+      "loss": 0.5304,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6286039145618452,
+      "grad_norm": 6.965030627490836,
+      "learning_rate": 1.6009191417764366e-06,
+      "loss": 0.4603,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6286851295378868,
+      "grad_norm": 4.3257443108941835,
+      "learning_rate": 1.600305549525651e-06,
+      "loss": 0.4716,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6287663445139283,
+      "grad_norm": 5.777339582650028,
+      "learning_rate": 1.5996920195289586e-06,
+      "loss": 0.5073,
+      "step": 7742
+    },
+    {
+      "epoch": 0.62884755948997,
+      "grad_norm": 11.440189242706712,
+      "learning_rate": 1.5990785518288144e-06,
+      "loss": 0.6117,
+      "step": 7743
+    },
+    {
+      "epoch": 0.6289287744660116,
+      "grad_norm": 5.114208771227363,
+      "learning_rate": 1.5984651464676664e-06,
+      "loss": 0.4889,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6290099894420531,
+      "grad_norm": 5.705508333315021,
+      "learning_rate": 1.5978518034879583e-06,
+      "loss": 0.3973,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6290912044180947,
+      "grad_norm": 8.471787190410229,
+      "learning_rate": 1.5972385229321313e-06,
+      "loss": 0.5272,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6291724193941363,
+      "grad_norm": 4.82503142826085,
+      "learning_rate": 1.5966253048426212e-06,
+      "loss": 0.6467,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6292536343701779,
+      "grad_norm": 4.760084685878624,
+      "learning_rate": 1.596012149261858e-06,
+      "loss": 0.544,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6293348493462194,
+      "grad_norm": 4.772517813956784,
+      "learning_rate": 1.5953990562322708e-06,
+      "loss": 0.4272,
+      "step": 7749
+    },
+    {
+      "epoch": 0.629416064322261,
+      "grad_norm": 8.327385517646334,
+      "learning_rate": 1.5947860257962808e-06,
+      "loss": 0.4509,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6294972792983026,
+      "grad_norm": 4.076666918311391,
+      "learning_rate": 1.5941730579963065e-06,
+      "loss": 0.4959,
+      "step": 7751
+    },
+    {
+      "epoch": 0.6295784942743442,
+      "grad_norm": 6.179747838755071,
+      "learning_rate": 1.5935601528747635e-06,
+      "loss": 0.4375,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6296597092503857,
+      "grad_norm": 12.93986219429753,
+      "learning_rate": 1.5929473104740605e-06,
+      "loss": 0.5545,
+      "step": 7753
+    },
+    {
+      "epoch": 0.6297409242264274,
+      "grad_norm": 3.6144579855996346,
+      "learning_rate": 1.5923345308366033e-06,
+      "loss": 0.4124,
+      "step": 7754
+    },
+    {
+      "epoch": 0.629822139202469,
+      "grad_norm": 3.832036350634625,
+      "learning_rate": 1.591721814004792e-06,
+      "loss": 0.6357,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6299033541785105,
+      "grad_norm": 3.254583154769827,
+      "learning_rate": 1.5911091600210243e-06,
+      "loss": 0.582,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6299845691545521,
+      "grad_norm": 4.7973970186134505,
+      "learning_rate": 1.5904965689276935e-06,
+      "loss": 0.809,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6300657841305937,
+      "grad_norm": 4.07832937486864,
+      "learning_rate": 1.5898840407671854e-06,
+      "loss": 0.4779,
+      "step": 7758
+    },
+    {
+      "epoch": 0.6301469991066353,
+      "grad_norm": 11.649929406013856,
+      "learning_rate": 1.5892715755818855e-06,
+      "loss": 0.4955,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6302282140826768,
+      "grad_norm": 7.298974738788142,
+      "learning_rate": 1.588659173414173e-06,
+      "loss": 0.3887,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6303094290587185,
+      "grad_norm": 6.5690578488061115,
+      "learning_rate": 1.5880468343064215e-06,
+      "loss": 0.5319,
+      "step": 7761
+    },
+    {
+      "epoch": 0.63039064403476,
+      "grad_norm": 4.501194184210136,
+      "learning_rate": 1.5874345583010038e-06,
+      "loss": 0.5715,
+      "step": 7762
+    },
+    {
+      "epoch": 0.6304718590108016,
+      "grad_norm": 4.823495001127094,
+      "learning_rate": 1.5868223454402842e-06,
+      "loss": 0.5115,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6305530739868431,
+      "grad_norm": 4.765174608509368,
+      "learning_rate": 1.5862101957666251e-06,
+      "loss": 0.4864,
+      "step": 7764
+    },
+    {
+      "epoch": 0.6306342889628848,
+      "grad_norm": 8.316397458431478,
+      "learning_rate": 1.5855981093223851e-06,
+      "loss": 0.5312,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6307155039389264,
+      "grad_norm": 3.4813538772732096,
+      "learning_rate": 1.5849860861499161e-06,
+      "loss": 0.5063,
+      "step": 7766
+    },
+    {
+      "epoch": 0.6307967189149679,
+      "grad_norm": 5.709151262434003,
+      "learning_rate": 1.584374126291567e-06,
+      "loss": 0.6373,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6308779338910095,
+      "grad_norm": 6.071499004825764,
+      "learning_rate": 1.5837622297896832e-06,
+      "loss": 0.4196,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6309591488670511,
+      "grad_norm": 4.152234565584312,
+      "learning_rate": 1.5831503966866038e-06,
+      "loss": 0.4568,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6310403638430927,
+      "grad_norm": 5.124374169356126,
+      "learning_rate": 1.5825386270246649e-06,
+      "loss": 0.5482,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6311215788191342,
+      "grad_norm": 6.603032548895071,
+      "learning_rate": 1.5819269208461962e-06,
+      "loss": 0.4628,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6312027937951759,
+      "grad_norm": 3.929016081455086,
+      "learning_rate": 1.5813152781935264e-06,
+      "loss": 0.4962,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6312840087712174,
+      "grad_norm": 5.248036747944997,
+      "learning_rate": 1.5807036991089781e-06,
+      "loss": 0.5938,
+      "step": 7773
+    },
+    {
+      "epoch": 0.631365223747259,
+      "grad_norm": 4.602869975600319,
+      "learning_rate": 1.5800921836348671e-06,
+      "loss": 0.4762,
+      "step": 7774
+    },
+    {
+      "epoch": 0.6314464387233005,
+      "grad_norm": 3.998867630823634,
+      "learning_rate": 1.5794807318135097e-06,
+      "loss": 0.4337,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6315276536993422,
+      "grad_norm": 4.511923050001,
+      "learning_rate": 1.5788693436872132e-06,
+      "loss": 0.5222,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6316088686753838,
+      "grad_norm": 4.732480730008836,
+      "learning_rate": 1.5782580192982827e-06,
+      "loss": 0.5321,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6316900836514253,
+      "grad_norm": 6.209675234125353,
+      "learning_rate": 1.57764675868902e-06,
+      "loss": 0.4472,
+      "step": 7778
+    },
+    {
+      "epoch": 0.631771298627467,
+      "grad_norm": 4.538684305840299,
+      "learning_rate": 1.5770355619017198e-06,
+      "loss": 0.5362,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6318525136035085,
+      "grad_norm": 3.509225017297568,
+      "learning_rate": 1.5764244289786728e-06,
+      "loss": 0.4968,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6319337285795501,
+      "grad_norm": 6.521031887625569,
+      "learning_rate": 1.575813359962169e-06,
+      "loss": 0.3996,
+      "step": 7781
+    },
+    {
+      "epoch": 0.6320149435555916,
+      "grad_norm": 12.929846426438688,
+      "learning_rate": 1.5752023548944889e-06,
+      "loss": 0.4625,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6320961585316333,
+      "grad_norm": 9.52281730299787,
+      "learning_rate": 1.574591413817911e-06,
+      "loss": 0.5613,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6321773735076748,
+      "grad_norm": 3.924489403653561,
+      "learning_rate": 1.57398053677471e-06,
+      "loss": 0.5386,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6322585884837164,
+      "grad_norm": 3.8784061814789137,
+      "learning_rate": 1.5733697238071553e-06,
+      "loss": 0.4747,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6323398034597579,
+      "grad_norm": 4.881789075504623,
+      "learning_rate": 1.5727589749575107e-06,
+      "loss": 0.5493,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6324210184357996,
+      "grad_norm": 7.261971143003088,
+      "learning_rate": 1.5721482902680385e-06,
+      "loss": 0.5279,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6325022334118412,
+      "grad_norm": 6.077300198937425,
+      "learning_rate": 1.5715376697809937e-06,
+      "loss": 0.5173,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6325834483878827,
+      "grad_norm": 5.975876289320817,
+      "learning_rate": 1.570927113538629e-06,
+      "loss": 0.5638,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6326646633639244,
+      "grad_norm": 3.670648216770045,
+      "learning_rate": 1.5703166215831899e-06,
+      "loss": 0.4479,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6327458783399659,
+      "grad_norm": 3.6145233680649254,
+      "learning_rate": 1.5697061939569214e-06,
+      "loss": 0.5388,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6328270933160075,
+      "grad_norm": 4.822710807501337,
+      "learning_rate": 1.56909583070206e-06,
+      "loss": 0.6051,
+      "step": 7792
+    },
+    {
+      "epoch": 0.632908308292049,
+      "grad_norm": 6.545220131213222,
+      "learning_rate": 1.56848553186084e-06,
+      "loss": 0.4538,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6329895232680907,
+      "grad_norm": 6.080696544546954,
+      "learning_rate": 1.567875297475492e-06,
+      "loss": 0.4151,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6330707382441322,
+      "grad_norm": 9.899327382181253,
+      "learning_rate": 1.56726512758824e-06,
+      "loss": 0.5212,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6331519532201738,
+      "grad_norm": 4.273991889208243,
+      "learning_rate": 1.566655022241304e-06,
+      "loss": 0.5173,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6332331681962153,
+      "grad_norm": 5.170717612782167,
+      "learning_rate": 1.5660449814769021e-06,
+      "loss": 0.4797,
+      "step": 7797
+    },
+    {
+      "epoch": 0.633314383172257,
+      "grad_norm": 5.7053519794438685,
+      "learning_rate": 1.5654350053372442e-06,
+      "loss": 0.4382,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6333955981482986,
+      "grad_norm": 3.5081020926912667,
+      "learning_rate": 1.564825093864537e-06,
+      "loss": 0.4524,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6334768131243401,
+      "grad_norm": 9.01478070880551,
+      "learning_rate": 1.5642152471009849e-06,
+      "loss": 0.4047,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6335580281003818,
+      "grad_norm": 5.177815198718995,
+      "learning_rate": 1.563605465088785e-06,
+      "loss": 0.4582,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6336392430764233,
+      "grad_norm": 8.412137360840807,
+      "learning_rate": 1.5629957478701303e-06,
+      "loss": 0.4636,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6337204580524649,
+      "grad_norm": 5.581728871746328,
+      "learning_rate": 1.5623860954872116e-06,
+      "loss": 0.5674,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6338016730285064,
+      "grad_norm": 4.471083148690759,
+      "learning_rate": 1.5617765079822133e-06,
+      "loss": 0.4592,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6338828880045481,
+      "grad_norm": 3.475348192814086,
+      "learning_rate": 1.5611669853973141e-06,
+      "loss": 0.5273,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6339641029805896,
+      "grad_norm": 5.120001622959669,
+      "learning_rate": 1.5605575277746912e-06,
+      "loss": 0.4363,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6340453179566312,
+      "grad_norm": 4.543060760794114,
+      "learning_rate": 1.559948135156516e-06,
+      "loss": 0.4075,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6341265329326727,
+      "grad_norm": 4.1227313596641,
+      "learning_rate": 1.559338807584954e-06,
+      "loss": 0.4653,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6342077479087144,
+      "grad_norm": 4.831713571856866,
+      "learning_rate": 1.5587295451021678e-06,
+      "loss": 0.5367,
+      "step": 7809
+    },
+    {
+      "epoch": 0.634288962884756,
+      "grad_norm": 4.2655332482022414,
+      "learning_rate": 1.5581203477503166e-06,
+      "loss": 0.6405,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6343701778607975,
+      "grad_norm": 8.517022361410643,
+      "learning_rate": 1.5575112155715516e-06,
+      "loss": 0.408,
+      "step": 7811
+    },
+    {
+      "epoch": 0.6344513928368392,
+      "grad_norm": 4.14581190989536,
+      "learning_rate": 1.5569021486080223e-06,
+      "loss": 0.4273,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6345326078128807,
+      "grad_norm": 4.406741970957099,
+      "learning_rate": 1.5562931469018738e-06,
+      "loss": 0.688,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6346138227889223,
+      "grad_norm": 4.483314144569286,
+      "learning_rate": 1.555684210495245e-06,
+      "loss": 0.5208,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6346950377649638,
+      "grad_norm": 4.337838645295809,
+      "learning_rate": 1.5550753394302702e-06,
+      "loss": 0.5329,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6347762527410055,
+      "grad_norm": 6.548340782272945,
+      "learning_rate": 1.5544665337490822e-06,
+      "loss": 0.5038,
+      "step": 7816
+    },
+    {
+      "epoch": 0.634857467717047,
+      "grad_norm": 26.26368001727583,
+      "learning_rate": 1.5538577934938051e-06,
+      "loss": 0.415,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6349386826930886,
+      "grad_norm": 6.56905135630574,
+      "learning_rate": 1.5532491187065607e-06,
+      "loss": 0.4125,
+      "step": 7818
+    },
+    {
+      "epoch": 0.6350198976691301,
+      "grad_norm": 4.494023935708008,
+      "learning_rate": 1.5526405094294682e-06,
+      "loss": 0.4559,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6351011126451718,
+      "grad_norm": 7.264576312866752,
+      "learning_rate": 1.5520319657046384e-06,
+      "loss": 0.4749,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6351823276212134,
+      "grad_norm": 3.8422635582509037,
+      "learning_rate": 1.5514234875741785e-06,
+      "loss": 0.5266,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6352635425972549,
+      "grad_norm": 4.7814474085953815,
+      "learning_rate": 1.550815075080193e-06,
+      "loss": 0.5509,
+      "step": 7822
+    },
+    {
+      "epoch": 0.6353447575732966,
+      "grad_norm": 14.680542072418953,
+      "learning_rate": 1.5502067282647821e-06,
+      "loss": 0.4687,
+      "step": 7823
+    },
+    {
+      "epoch": 0.6354259725493381,
+      "grad_norm": 5.102982342242324,
+      "learning_rate": 1.5495984471700382e-06,
+      "loss": 0.5625,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6355071875253797,
+      "grad_norm": 6.340239866996015,
+      "learning_rate": 1.5489902318380512e-06,
+      "loss": 0.3983,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6355884025014212,
+      "grad_norm": 5.566203991765062,
+      "learning_rate": 1.5483820823109078e-06,
+      "loss": 0.5833,
+      "step": 7826
+    },
+    {
+      "epoch": 0.6356696174774629,
+      "grad_norm": 4.137320108495999,
+      "learning_rate": 1.5477739986306878e-06,
+      "loss": 0.6481,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6357508324535044,
+      "grad_norm": 4.613431700216779,
+      "learning_rate": 1.5471659808394669e-06,
+      "loss": 0.4265,
+      "step": 7828
+    },
+    {
+      "epoch": 0.635832047429546,
+      "grad_norm": 3.7043285271670507,
+      "learning_rate": 1.546558028979318e-06,
+      "loss": 0.3759,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6359132624055875,
+      "grad_norm": 5.051376359459655,
+      "learning_rate": 1.5459501430923073e-06,
+      "loss": 0.4925,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6359944773816292,
+      "grad_norm": 4.08262450327924,
+      "learning_rate": 1.5453423232204968e-06,
+      "loss": 0.4753,
+      "step": 7831
+    },
+    {
+      "epoch": 0.6360756923576708,
+      "grad_norm": 4.998294324944461,
+      "learning_rate": 1.5447345694059462e-06,
+      "loss": 0.4227,
+      "step": 7832
+    },
+    {
+      "epoch": 0.6361569073337123,
+      "grad_norm": 13.638204885625212,
+      "learning_rate": 1.5441268816907077e-06,
+      "loss": 0.5512,
+      "step": 7833
+    },
+    {
+      "epoch": 0.636238122309754,
+      "grad_norm": 22.49883181989276,
+      "learning_rate": 1.5435192601168293e-06,
+      "loss": 0.5912,
+      "step": 7834
+    },
+    {
+      "epoch": 0.6363193372857955,
+      "grad_norm": 5.137209289732004,
+      "learning_rate": 1.542911704726356e-06,
+      "loss": 0.425,
+      "step": 7835
+    },
+    {
+      "epoch": 0.6364005522618371,
+      "grad_norm": 5.554841196878051,
+      "learning_rate": 1.5423042155613283e-06,
+      "loss": 0.3496,
+      "step": 7836
+    },
+    {
+      "epoch": 0.6364817672378786,
+      "grad_norm": 6.808818397785924,
+      "learning_rate": 1.5416967926637793e-06,
+      "loss": 0.3677,
+      "step": 7837
+    },
+    {
+      "epoch": 0.6365629822139203,
+      "grad_norm": 10.986138436222886,
+      "learning_rate": 1.5410894360757408e-06,
+      "loss": 0.5055,
+      "step": 7838
+    },
+    {
+      "epoch": 0.6366441971899618,
+      "grad_norm": 7.0748129231448695,
+      "learning_rate": 1.540482145839239e-06,
+      "loss": 0.4956,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6367254121660034,
+      "grad_norm": 5.517324400634849,
+      "learning_rate": 1.5398749219962935e-06,
+      "loss": 0.5188,
+      "step": 7840
+    },
+    {
+      "epoch": 0.636806627142045,
+      "grad_norm": 7.369254822891894,
+      "learning_rate": 1.5392677645889225e-06,
+      "loss": 0.4918,
+      "step": 7841
+    },
+    {
+      "epoch": 0.6368878421180866,
+      "grad_norm": 4.443831510239287,
+      "learning_rate": 1.5386606736591381e-06,
+      "loss": 0.5523,
+      "step": 7842
+    },
+    {
+      "epoch": 0.6369690570941282,
+      "grad_norm": 5.484153254041072,
+      "learning_rate": 1.5380536492489468e-06,
+      "loss": 0.4526,
+      "step": 7843
+    },
+    {
+      "epoch": 0.6370502720701697,
+      "grad_norm": 3.6793655308252364,
+      "learning_rate": 1.5374466914003516e-06,
+      "loss": 0.5204,
+      "step": 7844
+    },
+    {
+      "epoch": 0.6371314870462114,
+      "grad_norm": 5.6124428134045,
+      "learning_rate": 1.536839800155352e-06,
+      "loss": 0.5365,
+      "step": 7845
+    },
+    {
+      "epoch": 0.6372127020222529,
+      "grad_norm": 4.415081736558317,
+      "learning_rate": 1.5362329755559402e-06,
+      "loss": 0.4503,
+      "step": 7846
+    },
+    {
+      "epoch": 0.6372939169982945,
+      "grad_norm": 8.36645209753681,
+      "learning_rate": 1.5356262176441051e-06,
+      "loss": 0.5873,
+      "step": 7847
+    },
+    {
+      "epoch": 0.637375131974336,
+      "grad_norm": 4.410444200701278,
+      "learning_rate": 1.5350195264618333e-06,
+      "loss": 0.4172,
+      "step": 7848
+    },
+    {
+      "epoch": 0.6374563469503777,
+      "grad_norm": 5.471308320004208,
+      "learning_rate": 1.5344129020511029e-06,
+      "loss": 0.378,
+      "step": 7849
+    },
+    {
+      "epoch": 0.6375375619264192,
+      "grad_norm": 2.964980011323985,
+      "learning_rate": 1.5338063444538887e-06,
+      "loss": 0.4858,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6376187769024608,
+      "grad_norm": 4.034123659647055,
+      "learning_rate": 1.533199853712162e-06,
+      "loss": 0.4601,
+      "step": 7851
+    },
+    {
+      "epoch": 0.6376999918785023,
+      "grad_norm": 5.360248478803197,
+      "learning_rate": 1.5325934298678896e-06,
+      "loss": 0.5198,
+      "step": 7852
+    },
+    {
+      "epoch": 0.637781206854544,
+      "grad_norm": 4.128851671530124,
+      "learning_rate": 1.5319870729630303e-06,
+      "loss": 0.5347,
+      "step": 7853
+    },
+    {
+      "epoch": 0.6378624218305856,
+      "grad_norm": 5.837367955901182,
+      "learning_rate": 1.5313807830395437e-06,
+      "loss": 0.4249,
+      "step": 7854
+    },
+    {
+      "epoch": 0.6379436368066271,
+      "grad_norm": 4.420374712053163,
+      "learning_rate": 1.5307745601393808e-06,
+      "loss": 0.5663,
+      "step": 7855
+    },
+    {
+      "epoch": 0.6380248517826688,
+      "grad_norm": 8.34561480709057,
+      "learning_rate": 1.5301684043044875e-06,
+      "loss": 0.569,
+      "step": 7856
+    },
+    {
+      "epoch": 0.6381060667587103,
+      "grad_norm": 11.076659493261435,
+      "learning_rate": 1.5295623155768086e-06,
+      "loss": 0.4581,
+      "step": 7857
+    },
+    {
+      "epoch": 0.6381872817347519,
+      "grad_norm": 5.281957951201278,
+      "learning_rate": 1.5289562939982822e-06,
+      "loss": 0.5524,
+      "step": 7858
+    },
+    {
+      "epoch": 0.6382684967107934,
+      "grad_norm": 4.075328659102114,
+      "learning_rate": 1.5283503396108401e-06,
+      "loss": 0.5083,
+      "step": 7859
+    },
+    {
+      "epoch": 0.6383497116868351,
+      "grad_norm": 4.739091105300953,
+      "learning_rate": 1.5277444524564117e-06,
+      "loss": 0.524,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6384309266628766,
+      "grad_norm": 12.502865910338423,
+      "learning_rate": 1.5271386325769227e-06,
+      "loss": 0.2867,
+      "step": 7861
+    },
+    {
+      "epoch": 0.6385121416389182,
+      "grad_norm": 4.833851282125769,
+      "learning_rate": 1.526532880014292e-06,
+      "loss": 0.6359,
+      "step": 7862
+    },
+    {
+      "epoch": 0.6385933566149598,
+      "grad_norm": 4.157660884304162,
+      "learning_rate": 1.5259271948104323e-06,
+      "loss": 0.4139,
+      "step": 7863
+    },
+    {
+      "epoch": 0.6386745715910014,
+      "grad_norm": 4.699011951945198,
+      "learning_rate": 1.5253215770072564e-06,
+      "loss": 0.4898,
+      "step": 7864
+    },
+    {
+      "epoch": 0.638755786567043,
+      "grad_norm": 5.911530191898309,
+      "learning_rate": 1.5247160266466693e-06,
+      "loss": 0.5325,
+      "step": 7865
+    },
+    {
+      "epoch": 0.6388370015430845,
+      "grad_norm": 3.680767366819412,
+      "learning_rate": 1.5241105437705706e-06,
+      "loss": 0.5321,
+      "step": 7866
+    },
+    {
+      "epoch": 0.6389182165191262,
+      "grad_norm": 6.398260678627941,
+      "learning_rate": 1.523505128420858e-06,
+      "loss": 0.4228,
+      "step": 7867
+    },
+    {
+      "epoch": 0.6389994314951677,
+      "grad_norm": 3.398702627971208,
+      "learning_rate": 1.522899780639423e-06,
+      "loss": 0.5505,
+      "step": 7868
+    },
+    {
+      "epoch": 0.6390806464712093,
+      "grad_norm": 6.202752332409573,
+      "learning_rate": 1.5222945004681504e-06,
+      "loss": 0.7697,
+      "step": 7869
+    },
+    {
+      "epoch": 0.6391618614472508,
+      "grad_norm": 5.874268780003323,
+      "learning_rate": 1.5216892879489253e-06,
+      "loss": 0.4696,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6392430764232925,
+      "grad_norm": 4.404120702182698,
+      "learning_rate": 1.521084143123624e-06,
+      "loss": 0.6132,
+      "step": 7871
+    },
+    {
+      "epoch": 0.639324291399334,
+      "grad_norm": 4.8946588555368695,
+      "learning_rate": 1.5204790660341178e-06,
+      "loss": 0.4298,
+      "step": 7872
+    },
+    {
+      "epoch": 0.6394055063753756,
+      "grad_norm": 4.660366909329026,
+      "learning_rate": 1.519874056722277e-06,
+      "loss": 0.6014,
+      "step": 7873
+    },
+    {
+      "epoch": 0.6394867213514172,
+      "grad_norm": 9.926138073456057,
+      "learning_rate": 1.5192691152299649e-06,
+      "loss": 0.5607,
+      "step": 7874
+    },
+    {
+      "epoch": 0.6395679363274588,
+      "grad_norm": 6.862534435729142,
+      "learning_rate": 1.5186642415990382e-06,
+      "loss": 0.4831,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6396491513035004,
+      "grad_norm": 5.109846150321177,
+      "learning_rate": 1.518059435871353e-06,
+      "loss": 0.3118,
+      "step": 7876
+    },
+    {
+      "epoch": 0.6397303662795419,
+      "grad_norm": 4.824987683441751,
+      "learning_rate": 1.5174546980887585e-06,
+      "loss": 0.407,
+      "step": 7877
+    },
+    {
+      "epoch": 0.6398115812555836,
+      "grad_norm": 3.088737083022684,
+      "learning_rate": 1.516850028293099e-06,
+      "loss": 0.6486,
+      "step": 7878
+    },
+    {
+      "epoch": 0.6398927962316251,
+      "grad_norm": 4.687914435447199,
+      "learning_rate": 1.516245426526213e-06,
+      "loss": 0.6395,
+      "step": 7879
+    },
+    {
+      "epoch": 0.6399740112076667,
+      "grad_norm": 6.45233418054512,
+      "learning_rate": 1.5156408928299377e-06,
+      "loss": 0.5865,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6400552261837082,
+      "grad_norm": 4.4953810242052,
+      "learning_rate": 1.5150364272461035e-06,
+      "loss": 0.4263,
+      "step": 7881
+    },
+    {
+      "epoch": 0.6401364411597499,
+      "grad_norm": 5.174417200023911,
+      "learning_rate": 1.5144320298165346e-06,
+      "loss": 0.4651,
+      "step": 7882
+    },
+    {
+      "epoch": 0.6402176561357914,
+      "grad_norm": 8.999793348428359,
+      "learning_rate": 1.5138277005830538e-06,
+      "loss": 0.4134,
+      "step": 7883
+    },
+    {
+      "epoch": 0.640298871111833,
+      "grad_norm": 4.942872220787506,
+      "learning_rate": 1.5132234395874773e-06,
+      "loss": 0.5613,
+      "step": 7884
+    },
+    {
+      "epoch": 0.6403800860878746,
+      "grad_norm": 3.7848669467000433,
+      "learning_rate": 1.5126192468716152e-06,
+      "loss": 0.4878,
+      "step": 7885
+    },
+    {
+      "epoch": 0.6404613010639162,
+      "grad_norm": 4.408963540747303,
+      "learning_rate": 1.5120151224772765e-06,
+      "loss": 0.6558,
+      "step": 7886
+    },
+    {
+      "epoch": 0.6405425160399578,
+      "grad_norm": 6.167053391776367,
+      "learning_rate": 1.5114110664462624e-06,
+      "loss": 0.5179,
+      "step": 7887
+    },
+    {
+      "epoch": 0.6406237310159993,
+      "grad_norm": 4.691809504965685,
+      "learning_rate": 1.5108070788203699e-06,
+      "loss": 0.5723,
+      "step": 7888
+    },
+    {
+      "epoch": 0.640704945992041,
+      "grad_norm": 4.8601836373152,
+      "learning_rate": 1.5102031596413927e-06,
+      "loss": 0.5001,
+      "step": 7889
+    },
+    {
+      "epoch": 0.6407861609680825,
+      "grad_norm": 4.771276325593275,
+      "learning_rate": 1.509599308951119e-06,
+      "loss": 0.4755,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6408673759441241,
+      "grad_norm": 7.6294635480914135,
+      "learning_rate": 1.5089955267913303e-06,
+      "loss": 0.3362,
+      "step": 7891
+    },
+    {
+      "epoch": 0.6409485909201657,
+      "grad_norm": 4.020044219945879,
+      "learning_rate": 1.5083918132038072e-06,
+      "loss": 0.3609,
+      "step": 7892
+    },
+    {
+      "epoch": 0.6410298058962073,
+      "grad_norm": 8.75122403334399,
+      "learning_rate": 1.5077881682303225e-06,
+      "loss": 0.4598,
+      "step": 7893
+    },
+    {
+      "epoch": 0.6411110208722488,
+      "grad_norm": 12.260107016009812,
+      "learning_rate": 1.5071845919126448e-06,
+      "loss": 0.5443,
+      "step": 7894
+    },
+    {
+      "epoch": 0.6411922358482904,
+      "grad_norm": 6.260976119003239,
+      "learning_rate": 1.5065810842925399e-06,
+      "loss": 0.456,
+      "step": 7895
+    },
+    {
+      "epoch": 0.641273450824332,
+      "grad_norm": 4.832778705163437,
+      "learning_rate": 1.5059776454117658e-06,
+      "loss": 0.4861,
+      "step": 7896
+    },
+    {
+      "epoch": 0.6413546658003736,
+      "grad_norm": 4.117223333471925,
+      "learning_rate": 1.505374275312078e-06,
+      "loss": 0.4146,
+      "step": 7897
+    },
+    {
+      "epoch": 0.6414358807764152,
+      "grad_norm": 7.715748444314561,
+      "learning_rate": 1.504770974035226e-06,
+      "loss": 0.4328,
+      "step": 7898
+    },
+    {
+      "epoch": 0.6415170957524567,
+      "grad_norm": 3.698130972907306,
+      "learning_rate": 1.5041677416229556e-06,
+      "loss": 0.5428,
+      "step": 7899
+    },
+    {
+      "epoch": 0.6415983107284984,
+      "grad_norm": 8.80799380441798,
+      "learning_rate": 1.5035645781170078e-06,
+      "loss": 0.5696,
+      "step": 7900
+    },
+    {
+      "epoch": 0.6416795257045399,
+      "grad_norm": 5.951094606026148,
+      "learning_rate": 1.502961483559116e-06,
+      "loss": 0.6523,
+      "step": 7901
+    },
+    {
+      "epoch": 0.6417607406805815,
+      "grad_norm": 5.956077850717039,
+      "learning_rate": 1.502358457991014e-06,
+      "loss": 0.4707,
+      "step": 7902
+    },
+    {
+      "epoch": 0.641841955656623,
+      "grad_norm": 17.140276946196106,
+      "learning_rate": 1.5017555014544273e-06,
+      "loss": 0.4322,
+      "step": 7903
+    },
+    {
+      "epoch": 0.6419231706326647,
+      "grad_norm": 4.87844418439407,
+      "learning_rate": 1.5011526139910754e-06,
+      "loss": 0.523,
+      "step": 7904
+    },
+    {
+      "epoch": 0.6420043856087062,
+      "grad_norm": 4.800050176008073,
+      "learning_rate": 1.5005497956426773e-06,
+      "loss": 0.4496,
+      "step": 7905
+    },
+    {
+      "epoch": 0.6420856005847478,
+      "grad_norm": 15.944468140204993,
+      "learning_rate": 1.4999470464509432e-06,
+      "loss": 0.4226,
+      "step": 7906
+    },
+    {
+      "epoch": 0.6421668155607894,
+      "grad_norm": 10.898961363880764,
+      "learning_rate": 1.4993443664575807e-06,
+      "loss": 0.4787,
+      "step": 7907
+    },
+    {
+      "epoch": 0.642248030536831,
+      "grad_norm": 3.386355939222327,
+      "learning_rate": 1.4987417557042928e-06,
+      "loss": 0.4948,
+      "step": 7908
+    },
+    {
+      "epoch": 0.6423292455128726,
+      "grad_norm": 4.044784764443736,
+      "learning_rate": 1.4981392142327761e-06,
+      "loss": 0.504,
+      "step": 7909
+    },
+    {
+      "epoch": 0.6424104604889141,
+      "grad_norm": 7.7128980212700915,
+      "learning_rate": 1.4975367420847225e-06,
+      "loss": 0.5462,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6424916754649558,
+      "grad_norm": 7.797675381291463,
+      "learning_rate": 1.4969343393018224e-06,
+      "loss": 0.4488,
+      "step": 7911
+    },
+    {
+      "epoch": 0.6425728904409973,
+      "grad_norm": 3.7094574496346198,
+      "learning_rate": 1.4963320059257565e-06,
+      "loss": 0.5137,
+      "step": 7912
+    },
+    {
+      "epoch": 0.6426541054170389,
+      "grad_norm": 6.476581873676165,
+      "learning_rate": 1.4957297419982047e-06,
+      "loss": 0.4704,
+      "step": 7913
+    },
+    {
+      "epoch": 0.6427353203930805,
+      "grad_norm": 4.73377261927438,
+      "learning_rate": 1.4951275475608387e-06,
+      "loss": 0.463,
+      "step": 7914
+    },
+    {
+      "epoch": 0.6428165353691221,
+      "grad_norm": 3.1667300398717386,
+      "learning_rate": 1.4945254226553288e-06,
+      "loss": 0.5192,
+      "step": 7915
+    },
+    {
+      "epoch": 0.6428977503451636,
+      "grad_norm": 4.733929299465015,
+      "learning_rate": 1.4939233673233387e-06,
+      "loss": 0.6137,
+      "step": 7916
+    },
+    {
+      "epoch": 0.6429789653212052,
+      "grad_norm": 5.618667686017774,
+      "learning_rate": 1.4933213816065257e-06,
+      "loss": 0.4966,
+      "step": 7917
+    },
+    {
+      "epoch": 0.6430601802972468,
+      "grad_norm": 3.847952094017611,
+      "learning_rate": 1.492719465546546e-06,
+      "loss": 0.4784,
+      "step": 7918
+    },
+    {
+      "epoch": 0.6431413952732884,
+      "grad_norm": 4.697213023502698,
+      "learning_rate": 1.492117619185049e-06,
+      "loss": 0.4471,
+      "step": 7919
+    },
+    {
+      "epoch": 0.64322261024933,
+      "grad_norm": 5.479689495939854,
+      "learning_rate": 1.4915158425636772e-06,
+      "loss": 0.5378,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6433038252253716,
+      "grad_norm": 3.3937052015697367,
+      "learning_rate": 1.4909141357240731e-06,
+      "loss": 0.4669,
+      "step": 7921
+    },
+    {
+      "epoch": 0.6433850402014132,
+      "grad_norm": 4.013372087007401,
+      "learning_rate": 1.4903124987078698e-06,
+      "loss": 0.4361,
+      "step": 7922
+    },
+    {
+      "epoch": 0.6434662551774547,
+      "grad_norm": 3.746362491526315,
+      "learning_rate": 1.4897109315566974e-06,
+      "loss": 0.4443,
+      "step": 7923
+    },
+    {
+      "epoch": 0.6435474701534963,
+      "grad_norm": 4.102308375643804,
+      "learning_rate": 1.4891094343121827e-06,
+      "loss": 0.446,
+      "step": 7924
+    },
+    {
+      "epoch": 0.6436286851295379,
+      "grad_norm": 3.8787498068648634,
+      "learning_rate": 1.488508007015944e-06,
+      "loss": 0.7058,
+      "step": 7925
+    },
+    {
+      "epoch": 0.6437099001055795,
+      "grad_norm": 7.193908220842526,
+      "learning_rate": 1.487906649709598e-06,
+      "loss": 0.4578,
+      "step": 7926
+    },
+    {
+      "epoch": 0.643791115081621,
+      "grad_norm": 7.588715391068836,
+      "learning_rate": 1.4873053624347567e-06,
+      "loss": 0.5096,
+      "step": 7927
+    },
+    {
+      "epoch": 0.6438723300576626,
+      "grad_norm": 5.508197098569508,
+      "learning_rate": 1.4867041452330238e-06,
+      "loss": 0.5221,
+      "step": 7928
+    },
+    {
+      "epoch": 0.6439535450337042,
+      "grad_norm": 3.6948692770323928,
+      "learning_rate": 1.4861029981460007e-06,
+      "loss": 0.4898,
+      "step": 7929
+    },
+    {
+      "epoch": 0.6440347600097458,
+      "grad_norm": 3.770725561207981,
+      "learning_rate": 1.4855019212152852e-06,
+      "loss": 0.4907,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6441159749857874,
+      "grad_norm": 6.044600128479535,
+      "learning_rate": 1.484900914482467e-06,
+      "loss": 0.3815,
+      "step": 7931
+    },
+    {
+      "epoch": 0.644197189961829,
+      "grad_norm": 5.141487441925867,
+      "learning_rate": 1.484299977989134e-06,
+      "loss": 0.5212,
+      "step": 7932
+    },
+    {
+      "epoch": 0.6442784049378706,
+      "grad_norm": 6.722063710271542,
+      "learning_rate": 1.4836991117768657e-06,
+      "loss": 0.465,
+      "step": 7933
+    },
+    {
+      "epoch": 0.6443596199139121,
+      "grad_norm": 6.60995189007181,
+      "learning_rate": 1.4830983158872414e-06,
+      "loss": 0.3926,
+      "step": 7934
+    },
+    {
+      "epoch": 0.6444408348899537,
+      "grad_norm": 5.159161950257783,
+      "learning_rate": 1.482497590361831e-06,
+      "loss": 0.5679,
+      "step": 7935
+    },
+    {
+      "epoch": 0.6445220498659953,
+      "grad_norm": 4.722697499573461,
+      "learning_rate": 1.4818969352422018e-06,
+      "loss": 0.4301,
+      "step": 7936
+    },
+    {
+      "epoch": 0.6446032648420369,
+      "grad_norm": 5.60413330763897,
+      "learning_rate": 1.4812963505699179e-06,
+      "loss": 0.5284,
+      "step": 7937
+    },
+    {
+      "epoch": 0.6446844798180784,
+      "grad_norm": 4.906758714354853,
+      "learning_rate": 1.4806958363865342e-06,
+      "loss": 0.4881,
+      "step": 7938
+    },
+    {
+      "epoch": 0.64476569479412,
+      "grad_norm": 5.416581495053993,
+      "learning_rate": 1.4800953927336036e-06,
+      "loss": 0.4121,
+      "step": 7939
+    },
+    {
+      "epoch": 0.6448469097701616,
+      "grad_norm": 3.1479449081387605,
+      "learning_rate": 1.4794950196526753e-06,
+      "loss": 0.4134,
+      "step": 7940
+    },
+    {
+      "epoch": 0.6449281247462032,
+      "grad_norm": 4.49719543896871,
+      "learning_rate": 1.4788947171852899e-06,
+      "loss": 0.3979,
+      "step": 7941
+    },
+    {
+      "epoch": 0.6450093397222448,
+      "grad_norm": 5.0629449929653205,
+      "learning_rate": 1.4782944853729856e-06,
+      "loss": 0.4008,
+      "step": 7942
+    },
+    {
+      "epoch": 0.6450905546982864,
+      "grad_norm": 4.398126234430731,
+      "learning_rate": 1.4776943242572966e-06,
+      "loss": 0.6205,
+      "step": 7943
+    },
+    {
+      "epoch": 0.645171769674328,
+      "grad_norm": 4.1699063676723425,
+      "learning_rate": 1.4770942338797491e-06,
+      "loss": 0.5619,
+      "step": 7944
+    },
+    {
+      "epoch": 0.6452529846503695,
+      "grad_norm": 83.64698952954932,
+      "learning_rate": 1.4764942142818667e-06,
+      "loss": 0.4946,
+      "step": 7945
+    },
+    {
+      "epoch": 0.6453341996264111,
+      "grad_norm": 5.2534053694927625,
+      "learning_rate": 1.475894265505169e-06,
+      "loss": 0.6233,
+      "step": 7946
+    },
+    {
+      "epoch": 0.6454154146024527,
+      "grad_norm": 3.941627067484926,
+      "learning_rate": 1.4752943875911673e-06,
+      "loss": 0.4209,
+      "step": 7947
+    },
+    {
+      "epoch": 0.6454966295784943,
+      "grad_norm": 9.922976413484275,
+      "learning_rate": 1.4746945805813707e-06,
+      "loss": 0.5124,
+      "step": 7948
+    },
+    {
+      "epoch": 0.6455778445545358,
+      "grad_norm": 6.9619380350835085,
+      "learning_rate": 1.4740948445172834e-06,
+      "loss": 0.3986,
+      "step": 7949
+    },
+    {
+      "epoch": 0.6456590595305775,
+      "grad_norm": 4.2939089133962325,
+      "learning_rate": 1.4734951794404035e-06,
+      "loss": 0.4537,
+      "step": 7950
+    },
+    {
+      "epoch": 0.645740274506619,
+      "grad_norm": 3.1416028655227293,
+      "learning_rate": 1.4728955853922238e-06,
+      "loss": 0.5554,
+      "step": 7951
+    },
+    {
+      "epoch": 0.6458214894826606,
+      "grad_norm": 17.016801772420305,
+      "learning_rate": 1.4722960624142336e-06,
+      "loss": 0.4172,
+      "step": 7952
+    },
+    {
+      "epoch": 0.6459027044587022,
+      "grad_norm": 5.911786929784765,
+      "learning_rate": 1.4716966105479175e-06,
+      "loss": 0.3805,
+      "step": 7953
+    },
+    {
+      "epoch": 0.6459839194347438,
+      "grad_norm": 4.538417255604256,
+      "learning_rate": 1.471097229834753e-06,
+      "loss": 0.4062,
+      "step": 7954
+    },
+    {
+      "epoch": 0.6460651344107854,
+      "grad_norm": 7.061375321676026,
+      "learning_rate": 1.4704979203162148e-06,
+      "loss": 0.4817,
+      "step": 7955
+    },
+    {
+      "epoch": 0.6461463493868269,
+      "grad_norm": 8.396993615373008,
+      "learning_rate": 1.4698986820337729e-06,
+      "loss": 0.3514,
+      "step": 7956
+    },
+    {
+      "epoch": 0.6462275643628685,
+      "grad_norm": 5.11600701972491,
+      "learning_rate": 1.4692995150288896e-06,
+      "loss": 0.4232,
+      "step": 7957
+    },
+    {
+      "epoch": 0.6463087793389101,
+      "grad_norm": 3.9201591191720446,
+      "learning_rate": 1.4687004193430248e-06,
+      "loss": 0.6237,
+      "step": 7958
+    },
+    {
+      "epoch": 0.6463899943149517,
+      "grad_norm": 4.347166743753607,
+      "learning_rate": 1.4681013950176338e-06,
+      "loss": 0.4223,
+      "step": 7959
+    },
+    {
+      "epoch": 0.6464712092909932,
+      "grad_norm": 4.5755798327245625,
+      "learning_rate": 1.4675024420941643e-06,
+      "loss": 0.4167,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6465524242670349,
+      "grad_norm": 3.768261457483465,
+      "learning_rate": 1.4669035606140613e-06,
+      "loss": 0.531,
+      "step": 7961
+    },
+    {
+      "epoch": 0.6466336392430764,
+      "grad_norm": 4.347537189926498,
+      "learning_rate": 1.4663047506187649e-06,
+      "loss": 0.4719,
+      "step": 7962
+    },
+    {
+      "epoch": 0.646714854219118,
+      "grad_norm": 4.256186132283668,
+      "learning_rate": 1.4657060121497095e-06,
+      "loss": 0.55,
+      "step": 7963
+    },
+    {
+      "epoch": 0.6467960691951596,
+      "grad_norm": 4.071283516303743,
+      "learning_rate": 1.4651073452483228e-06,
+      "loss": 0.762,
+      "step": 7964
+    },
+    {
+      "epoch": 0.6468772841712012,
+      "grad_norm": 10.936892803098575,
+      "learning_rate": 1.4645087499560313e-06,
+      "loss": 0.5069,
+      "step": 7965
+    },
+    {
+      "epoch": 0.6469584991472428,
+      "grad_norm": 4.622669916018039,
+      "learning_rate": 1.4639102263142546e-06,
+      "loss": 0.4825,
+      "step": 7966
+    },
+    {
+      "epoch": 0.6470397141232843,
+      "grad_norm": 5.995534947389931,
+      "learning_rate": 1.463311774364406e-06,
+      "loss": 0.5098,
+      "step": 7967
+    },
+    {
+      "epoch": 0.647120929099326,
+      "grad_norm": 81.25869393551474,
+      "learning_rate": 1.4627133941478958e-06,
+      "loss": 0.6188,
+      "step": 7968
+    },
+    {
+      "epoch": 0.6472021440753675,
+      "grad_norm": 5.489822227742221,
+      "learning_rate": 1.46211508570613e-06,
+      "loss": 0.4715,
+      "step": 7969
+    },
+    {
+      "epoch": 0.6472833590514091,
+      "grad_norm": 5.211788919279321,
+      "learning_rate": 1.4615168490805066e-06,
+      "loss": 0.3925,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6473645740274506,
+      "grad_norm": 4.441097732454784,
+      "learning_rate": 1.4609186843124208e-06,
+      "loss": 0.5827,
+      "step": 7971
+    },
+    {
+      "epoch": 0.6474457890034923,
+      "grad_norm": 4.918274629630136,
+      "learning_rate": 1.4603205914432638e-06,
+      "loss": 0.5865,
+      "step": 7972
+    },
+    {
+      "epoch": 0.6475270039795338,
+      "grad_norm": 5.381182030506248,
+      "learning_rate": 1.4597225705144189e-06,
+      "loss": 0.427,
+      "step": 7973
+    },
+    {
+      "epoch": 0.6476082189555754,
+      "grad_norm": 4.47373690879082,
+      "learning_rate": 1.459124621567266e-06,
+      "loss": 0.5468,
+      "step": 7974
+    },
+    {
+      "epoch": 0.647689433931617,
+      "grad_norm": 4.853763535922953,
+      "learning_rate": 1.4585267446431817e-06,
+      "loss": 0.4893,
+      "step": 7975
+    },
+    {
+      "epoch": 0.6477706489076586,
+      "grad_norm": 12.67331094044437,
+      "learning_rate": 1.4579289397835344e-06,
+      "loss": 0.6503,
+      "step": 7976
+    },
+    {
+      "epoch": 0.6478518638837002,
+      "grad_norm": 3.5376573305444725,
+      "learning_rate": 1.4573312070296885e-06,
+      "loss": 0.48,
+      "step": 7977
+    },
+    {
+      "epoch": 0.6479330788597417,
+      "grad_norm": 6.986996329380094,
+      "learning_rate": 1.4567335464230062e-06,
+      "loss": 0.4793,
+      "step": 7978
+    },
+    {
+      "epoch": 0.6480142938357834,
+      "grad_norm": 6.098834362052517,
+      "learning_rate": 1.4561359580048394e-06,
+      "loss": 0.567,
+      "step": 7979
+    },
+    {
+      "epoch": 0.6480955088118249,
+      "grad_norm": 5.728827122709186,
+      "learning_rate": 1.4555384418165405e-06,
+      "loss": 0.5839,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6481767237878665,
+      "grad_norm": 6.108126965155767,
+      "learning_rate": 1.4549409978994543e-06,
+      "loss": 0.721,
+      "step": 7981
+    },
+    {
+      "epoch": 0.648257938763908,
+      "grad_norm": 12.175353754173727,
+      "learning_rate": 1.45434362629492e-06,
+      "loss": 0.3989,
+      "step": 7982
+    },
+    {
+      "epoch": 0.6483391537399497,
+      "grad_norm": 5.726270390332232,
+      "learning_rate": 1.453746327044272e-06,
+      "loss": 0.5943,
+      "step": 7983
+    },
+    {
+      "epoch": 0.6484203687159912,
+      "grad_norm": 8.228427503586767,
+      "learning_rate": 1.4531491001888421e-06,
+      "loss": 0.3702,
+      "step": 7984
+    },
+    {
+      "epoch": 0.6485015836920328,
+      "grad_norm": 6.718803197230006,
+      "learning_rate": 1.4525519457699527e-06,
+      "loss": 0.4701,
+      "step": 7985
+    },
+    {
+      "epoch": 0.6485827986680744,
+      "grad_norm": 6.288459331865035,
+      "learning_rate": 1.451954863828926e-06,
+      "loss": 0.454,
+      "step": 7986
+    },
+    {
+      "epoch": 0.648664013644116,
+      "grad_norm": 5.184071201153414,
+      "learning_rate": 1.4513578544070753e-06,
+      "loss": 0.4614,
+      "step": 7987
+    },
+    {
+      "epoch": 0.6487452286201576,
+      "grad_norm": 6.12187597162872,
+      "learning_rate": 1.4507609175457121e-06,
+      "loss": 0.4276,
+      "step": 7988
+    },
+    {
+      "epoch": 0.6488264435961991,
+      "grad_norm": 6.050006695871517,
+      "learning_rate": 1.4501640532861405e-06,
+      "loss": 0.4831,
+      "step": 7989
+    },
+    {
+      "epoch": 0.6489076585722408,
+      "grad_norm": 6.399026819639625,
+      "learning_rate": 1.4495672616696594e-06,
+      "loss": 0.3898,
+      "step": 7990
+    },
+    {
+      "epoch": 0.6489888735482823,
+      "grad_norm": 5.876464671442483,
+      "learning_rate": 1.448970542737565e-06,
+      "loss": 0.5073,
+      "step": 7991
+    },
+    {
+      "epoch": 0.6490700885243239,
+      "grad_norm": 4.813231306921125,
+      "learning_rate": 1.4483738965311455e-06,
+      "loss": 0.5262,
+      "step": 7992
+    },
+    {
+      "epoch": 0.6491513035003654,
+      "grad_norm": 3.7301648640363716,
+      "learning_rate": 1.4477773230916872e-06,
+      "loss": 0.4813,
+      "step": 7993
+    },
+    {
+      "epoch": 0.6492325184764071,
+      "grad_norm": 5.647909285913068,
+      "learning_rate": 1.44718082246047e-06,
+      "loss": 0.442,
+      "step": 7994
+    },
+    {
+      "epoch": 0.6493137334524486,
+      "grad_norm": 5.414289909475496,
+      "learning_rate": 1.4465843946787683e-06,
+      "loss": 0.3569,
+      "step": 7995
+    },
+    {
+      "epoch": 0.6493949484284902,
+      "grad_norm": 7.022409413689571,
+      "learning_rate": 1.44598803978785e-06,
+      "loss": 0.485,
+      "step": 7996
+    },
+    {
+      "epoch": 0.6494761634045318,
+      "grad_norm": 5.325732469754769,
+      "learning_rate": 1.4453917578289823e-06,
+      "loss": 0.6726,
+      "step": 7997
+    },
+    {
+      "epoch": 0.6495573783805734,
+      "grad_norm": 6.073037858190773,
+      "learning_rate": 1.4447955488434223e-06,
+      "loss": 0.5305,
+      "step": 7998
+    },
+    {
+      "epoch": 0.649638593356615,
+      "grad_norm": 7.0964517666533515,
+      "learning_rate": 1.4441994128724258e-06,
+      "loss": 0.4777,
+      "step": 7999
+    },
+    {
+      "epoch": 0.6497198083326565,
+      "grad_norm": 5.760466887618525,
+      "learning_rate": 1.443603349957243e-06,
+      "loss": 0.5041,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6498010233086982,
+      "grad_norm": 4.591683516314596,
+      "learning_rate": 1.4430073601391175e-06,
+      "loss": 0.5213,
+      "step": 8001
+    },
+    {
+      "epoch": 0.6498822382847397,
+      "grad_norm": 5.711693932834374,
+      "learning_rate": 1.442411443459289e-06,
+      "loss": 0.5668,
+      "step": 8002
+    },
+    {
+      "epoch": 0.6499634532607813,
+      "grad_norm": 5.284351713643368,
+      "learning_rate": 1.44181559995899e-06,
+      "loss": 0.6332,
+      "step": 8003
+    },
+    {
+      "epoch": 0.6500446682368228,
+      "grad_norm": 7.634523671563961,
+      "learning_rate": 1.4412198296794516e-06,
+      "loss": 0.4598,
+      "step": 8004
+    },
+    {
+      "epoch": 0.6501258832128645,
+      "grad_norm": 4.421291562696544,
+      "learning_rate": 1.4406241326618981e-06,
+      "loss": 0.585,
+      "step": 8005
+    },
+    {
+      "epoch": 0.650207098188906,
+      "grad_norm": 4.721051742091286,
+      "learning_rate": 1.4400285089475468e-06,
+      "loss": 0.7653,
+      "step": 8006
+    },
+    {
+      "epoch": 0.6502883131649476,
+      "grad_norm": 5.351925711601577,
+      "learning_rate": 1.4394329585776143e-06,
+      "loss": 0.3979,
+      "step": 8007
+    },
+    {
+      "epoch": 0.6503695281409893,
+      "grad_norm": 4.447327868885781,
+      "learning_rate": 1.4388374815933078e-06,
+      "loss": 0.4846,
+      "step": 8008
+    },
+    {
+      "epoch": 0.6504507431170308,
+      "grad_norm": 4.016961136441612,
+      "learning_rate": 1.4382420780358306e-06,
+      "loss": 0.4739,
+      "step": 8009
+    },
+    {
+      "epoch": 0.6505319580930724,
+      "grad_norm": 5.414485646946576,
+      "learning_rate": 1.4376467479463832e-06,
+      "loss": 0.3597,
+      "step": 8010
+    },
+    {
+      "epoch": 0.6506131730691139,
+      "grad_norm": 11.24686014696604,
+      "learning_rate": 1.4370514913661576e-06,
+      "loss": 0.5058,
+      "step": 8011
+    },
+    {
+      "epoch": 0.6506943880451556,
+      "grad_norm": 3.574284491275758,
+      "learning_rate": 1.436456308336343e-06,
+      "loss": 0.6707,
+      "step": 8012
+    },
+    {
+      "epoch": 0.6507756030211971,
+      "grad_norm": 4.818858068808228,
+      "learning_rate": 1.4358611988981242e-06,
+      "loss": 0.4094,
+      "step": 8013
+    },
+    {
+      "epoch": 0.6508568179972387,
+      "grad_norm": 4.129816073543707,
+      "learning_rate": 1.4352661630926783e-06,
+      "loss": 0.5239,
+      "step": 8014
+    },
+    {
+      "epoch": 0.6509380329732802,
+      "grad_norm": 7.2250869844226155,
+      "learning_rate": 1.4346712009611786e-06,
+      "loss": 0.4763,
+      "step": 8015
+    },
+    {
+      "epoch": 0.6510192479493219,
+      "grad_norm": 6.519568020534992,
+      "learning_rate": 1.434076312544794e-06,
+      "loss": 0.4563,
+      "step": 8016
+    },
+    {
+      "epoch": 0.6511004629253634,
+      "grad_norm": 5.982613565742739,
+      "learning_rate": 1.4334814978846863e-06,
+      "loss": 0.4024,
+      "step": 8017
+    },
+    {
+      "epoch": 0.651181677901405,
+      "grad_norm": 6.189908706083811,
+      "learning_rate": 1.4328867570220148e-06,
+      "loss": 0.4997,
+      "step": 8018
+    },
+    {
+      "epoch": 0.6512628928774467,
+      "grad_norm": 6.213197790148053,
+      "learning_rate": 1.4322920899979327e-06,
+      "loss": 0.5354,
+      "step": 8019
+    },
+    {
+      "epoch": 0.6513441078534882,
+      "grad_norm": 4.184480691904144,
+      "learning_rate": 1.4316974968535873e-06,
+      "loss": 0.3249,
+      "step": 8020
+    },
+    {
+      "epoch": 0.6514253228295298,
+      "grad_norm": 4.810772139069584,
+      "learning_rate": 1.4311029776301216e-06,
+      "loss": 0.4938,
+      "step": 8021
+    },
+    {
+      "epoch": 0.6515065378055713,
+      "grad_norm": 6.385985541045838,
+      "learning_rate": 1.4305085323686714e-06,
+      "loss": 0.4544,
+      "step": 8022
+    },
+    {
+      "epoch": 0.651587752781613,
+      "grad_norm": 4.1470201913398155,
+      "learning_rate": 1.4299141611103717e-06,
+      "loss": 0.5221,
+      "step": 8023
+    },
+    {
+      "epoch": 0.6516689677576545,
+      "grad_norm": 4.404511419426839,
+      "learning_rate": 1.4293198638963476e-06,
+      "loss": 0.5142,
+      "step": 8024
+    },
+    {
+      "epoch": 0.6517501827336961,
+      "grad_norm": 4.771427690672761,
+      "learning_rate": 1.4287256407677225e-06,
+      "loss": 0.6298,
+      "step": 8025
+    },
+    {
+      "epoch": 0.6518313977097376,
+      "grad_norm": 5.110969640062156,
+      "learning_rate": 1.4281314917656144e-06,
+      "loss": 0.4597,
+      "step": 8026
+    },
+    {
+      "epoch": 0.6519126126857793,
+      "grad_norm": 4.07743528288326,
+      "learning_rate": 1.4275374169311345e-06,
+      "loss": 0.5317,
+      "step": 8027
+    },
+    {
+      "epoch": 0.6519938276618208,
+      "grad_norm": 3.381461916822027,
+      "learning_rate": 1.426943416305388e-06,
+      "loss": 0.6096,
+      "step": 8028
+    },
+    {
+      "epoch": 0.6520750426378624,
+      "grad_norm": 6.120291184514056,
+      "learning_rate": 1.4263494899294794e-06,
+      "loss": 0.3527,
+      "step": 8029
+    },
+    {
+      "epoch": 0.6521562576139041,
+      "grad_norm": 6.654443944669531,
+      "learning_rate": 1.4257556378445025e-06,
+      "loss": 0.479,
+      "step": 8030
+    },
+    {
+      "epoch": 0.6522374725899456,
+      "grad_norm": 5.964757640753549,
+      "learning_rate": 1.4251618600915503e-06,
+      "loss": 0.4986,
+      "step": 8031
+    },
+    {
+      "epoch": 0.6523186875659872,
+      "grad_norm": 4.555104801980313,
+      "learning_rate": 1.4245681567117097e-06,
+      "loss": 0.4979,
+      "step": 8032
+    },
+    {
+      "epoch": 0.6523999025420287,
+      "grad_norm": 4.9235533317425,
+      "learning_rate": 1.4239745277460614e-06,
+      "loss": 0.4747,
+      "step": 8033
+    },
+    {
+      "epoch": 0.6524811175180704,
+      "grad_norm": 6.579534318663932,
+      "learning_rate": 1.4233809732356798e-06,
+      "loss": 0.5299,
+      "step": 8034
+    },
+    {
+      "epoch": 0.6525623324941119,
+      "grad_norm": 5.083100688498042,
+      "learning_rate": 1.4227874932216378e-06,
+      "loss": 0.3675,
+      "step": 8035
+    },
+    {
+      "epoch": 0.6526435474701535,
+      "grad_norm": 23.029068662665217,
+      "learning_rate": 1.4221940877450006e-06,
+      "loss": 0.5651,
+      "step": 8036
+    },
+    {
+      "epoch": 0.652724762446195,
+      "grad_norm": 5.851390054577819,
+      "learning_rate": 1.4216007568468272e-06,
+      "loss": 0.52,
+      "step": 8037
+    },
+    {
+      "epoch": 0.6528059774222367,
+      "grad_norm": 3.2181868045877704,
+      "learning_rate": 1.4210075005681737e-06,
+      "loss": 0.5162,
+      "step": 8038
+    },
+    {
+      "epoch": 0.6528871923982782,
+      "grad_norm": 7.483188427836994,
+      "learning_rate": 1.420414318950092e-06,
+      "loss": 0.4435,
+      "step": 8039
+    },
+    {
+      "epoch": 0.6529684073743198,
+      "grad_norm": 4.4784108288995315,
+      "learning_rate": 1.4198212120336255e-06,
+      "loss": 0.441,
+      "step": 8040
+    },
+    {
+      "epoch": 0.6530496223503615,
+      "grad_norm": 6.391569173443781,
+      "learning_rate": 1.4192281798598133e-06,
+      "loss": 0.492,
+      "step": 8041
+    },
+    {
+      "epoch": 0.653130837326403,
+      "grad_norm": 5.75063446601935,
+      "learning_rate": 1.4186352224696926e-06,
+      "loss": 0.5992,
+      "step": 8042
+    },
+    {
+      "epoch": 0.6532120523024446,
+      "grad_norm": 6.211326632664685,
+      "learning_rate": 1.4180423399042902e-06,
+      "loss": 0.4642,
+      "step": 8043
+    },
+    {
+      "epoch": 0.6532932672784861,
+      "grad_norm": 11.997397806821628,
+      "learning_rate": 1.4174495322046316e-06,
+      "loss": 0.5696,
+      "step": 8044
+    },
+    {
+      "epoch": 0.6533744822545278,
+      "grad_norm": 7.193039419958394,
+      "learning_rate": 1.4168567994117375e-06,
+      "loss": 0.4744,
+      "step": 8045
+    },
+    {
+      "epoch": 0.6534556972305693,
+      "grad_norm": 4.494283159785974,
+      "learning_rate": 1.41626414156662e-06,
+      "loss": 0.3842,
+      "step": 8046
+    },
+    {
+      "epoch": 0.6535369122066109,
+      "grad_norm": 5.857275012139722,
+      "learning_rate": 1.4156715587102875e-06,
+      "loss": 0.4835,
+      "step": 8047
+    },
+    {
+      "epoch": 0.6536181271826524,
+      "grad_norm": 6.4200372823639675,
+      "learning_rate": 1.4150790508837453e-06,
+      "loss": 0.4392,
+      "step": 8048
+    },
+    {
+      "epoch": 0.6536993421586941,
+      "grad_norm": 3.22921759731994,
+      "learning_rate": 1.4144866181279908e-06,
+      "loss": 0.429,
+      "step": 8049
+    },
+    {
+      "epoch": 0.6537805571347356,
+      "grad_norm": 4.691847016227414,
+      "learning_rate": 1.4138942604840167e-06,
+      "loss": 0.4329,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6538617721107772,
+      "grad_norm": 5.045037115274969,
+      "learning_rate": 1.4133019779928115e-06,
+      "loss": 0.4094,
+      "step": 8051
+    },
+    {
+      "epoch": 0.6539429870868189,
+      "grad_norm": 3.3537113111805317,
+      "learning_rate": 1.4127097706953591e-06,
+      "loss": 0.6154,
+      "step": 8052
+    },
+    {
+      "epoch": 0.6540242020628604,
+      "grad_norm": 6.660882673507178,
+      "learning_rate": 1.4121176386326352e-06,
+      "loss": 0.3856,
+      "step": 8053
+    },
+    {
+      "epoch": 0.654105417038902,
+      "grad_norm": 4.011343252532216,
+      "learning_rate": 1.4115255818456138e-06,
+      "loss": 0.5311,
+      "step": 8054
+    },
+    {
+      "epoch": 0.6541866320149435,
+      "grad_norm": 3.3532310150002513,
+      "learning_rate": 1.4109336003752619e-06,
+      "loss": 0.4813,
+      "step": 8055
+    },
+    {
+      "epoch": 0.6542678469909852,
+      "grad_norm": 4.86729352164219,
+      "learning_rate": 1.4103416942625397e-06,
+      "loss": 0.4928,
+      "step": 8056
+    },
+    {
+      "epoch": 0.6543490619670267,
+      "grad_norm": 3.96263482227693,
+      "learning_rate": 1.4097498635484057e-06,
+      "loss": 0.447,
+      "step": 8057
+    },
+    {
+      "epoch": 0.6544302769430683,
+      "grad_norm": 5.204864204379556,
+      "learning_rate": 1.4091581082738122e-06,
+      "loss": 0.3607,
+      "step": 8058
+    },
+    {
+      "epoch": 0.6545114919191098,
+      "grad_norm": 4.292415805341833,
+      "learning_rate": 1.4085664284797041e-06,
+      "loss": 0.448,
+      "step": 8059
+    },
+    {
+      "epoch": 0.6545927068951515,
+      "grad_norm": 7.880826154296216,
+      "learning_rate": 1.407974824207022e-06,
+      "loss": 0.6328,
+      "step": 8060
+    },
+    {
+      "epoch": 0.654673921871193,
+      "grad_norm": 5.310250315901305,
+      "learning_rate": 1.4073832954967032e-06,
+      "loss": 0.4029,
+      "step": 8061
+    },
+    {
+      "epoch": 0.6547551368472346,
+      "grad_norm": 6.728553995913191,
+      "learning_rate": 1.406791842389677e-06,
+      "loss": 0.5177,
+      "step": 8062
+    },
+    {
+      "epoch": 0.6548363518232763,
+      "grad_norm": 5.981311174156778,
+      "learning_rate": 1.4062004649268696e-06,
+      "loss": 0.4751,
+      "step": 8063
+    },
+    {
+      "epoch": 0.6549175667993178,
+      "grad_norm": 6.024059197477494,
+      "learning_rate": 1.405609163149202e-06,
+      "loss": 0.4305,
+      "step": 8064
+    },
+    {
+      "epoch": 0.6549987817753594,
+      "grad_norm": 7.798448146594784,
+      "learning_rate": 1.4050179370975886e-06,
+      "loss": 0.4975,
+      "step": 8065
+    },
+    {
+      "epoch": 0.6550799967514009,
+      "grad_norm": 4.4502104987754105,
+      "learning_rate": 1.4044267868129374e-06,
+      "loss": 0.4111,
+      "step": 8066
+    },
+    {
+      "epoch": 0.6551612117274426,
+      "grad_norm": 3.807078310048636,
+      "learning_rate": 1.4038357123361556e-06,
+      "loss": 0.5769,
+      "step": 8067
+    },
+    {
+      "epoch": 0.6552424267034841,
+      "grad_norm": 4.154767587614258,
+      "learning_rate": 1.4032447137081414e-06,
+      "loss": 0.5461,
+      "step": 8068
+    },
+    {
+      "epoch": 0.6553236416795257,
+      "grad_norm": 5.8170559826468935,
+      "learning_rate": 1.4026537909697873e-06,
+      "loss": 0.5508,
+      "step": 8069
+    },
+    {
+      "epoch": 0.6554048566555672,
+      "grad_norm": 6.732343822655561,
+      "learning_rate": 1.4020629441619831e-06,
+      "loss": 0.4174,
+      "step": 8070
+    },
+    {
+      "epoch": 0.6554860716316089,
+      "grad_norm": 15.814471928424329,
+      "learning_rate": 1.4014721733256137e-06,
+      "loss": 0.3694,
+      "step": 8071
+    },
+    {
+      "epoch": 0.6555672866076504,
+      "grad_norm": 5.728680709044823,
+      "learning_rate": 1.4008814785015548e-06,
+      "loss": 0.5148,
+      "step": 8072
+    },
+    {
+      "epoch": 0.655648501583692,
+      "grad_norm": 5.735426011218967,
+      "learning_rate": 1.4002908597306817e-06,
+      "loss": 0.3364,
+      "step": 8073
+    },
+    {
+      "epoch": 0.6557297165597337,
+      "grad_norm": 3.8279642749863916,
+      "learning_rate": 1.3997003170538608e-06,
+      "loss": 0.4943,
+      "step": 8074
+    },
+    {
+      "epoch": 0.6558109315357752,
+      "grad_norm": 8.191610658404883,
+      "learning_rate": 1.3991098505119537e-06,
+      "loss": 0.6124,
+      "step": 8075
+    },
+    {
+      "epoch": 0.6558921465118168,
+      "grad_norm": 22.562198626943964,
+      "learning_rate": 1.3985194601458192e-06,
+      "loss": 0.6547,
+      "step": 8076
+    },
+    {
+      "epoch": 0.6559733614878583,
+      "grad_norm": 5.841079008969222,
+      "learning_rate": 1.3979291459963087e-06,
+      "loss": 0.2968,
+      "step": 8077
+    },
+    {
+      "epoch": 0.6560545764639,
+      "grad_norm": 3.714067743777654,
+      "learning_rate": 1.397338908104269e-06,
+      "loss": 0.43,
+      "step": 8078
+    },
+    {
+      "epoch": 0.6561357914399415,
+      "grad_norm": 4.366280468082913,
+      "learning_rate": 1.3967487465105401e-06,
+      "loss": 0.643,
+      "step": 8079
+    },
+    {
+      "epoch": 0.6562170064159831,
+      "grad_norm": 4.043858317470031,
+      "learning_rate": 1.3961586612559602e-06,
+      "loss": 0.6495,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6562982213920246,
+      "grad_norm": 4.59900089958901,
+      "learning_rate": 1.3955686523813588e-06,
+      "loss": 0.3702,
+      "step": 8081
+    },
+    {
+      "epoch": 0.6563794363680663,
+      "grad_norm": 6.449912994740261,
+      "learning_rate": 1.3949787199275606e-06,
+      "loss": 0.5347,
+      "step": 8082
+    },
+    {
+      "epoch": 0.6564606513441078,
+      "grad_norm": 5.373213623100858,
+      "learning_rate": 1.3943888639353866e-06,
+      "loss": 0.4742,
+      "step": 8083
+    },
+    {
+      "epoch": 0.6565418663201494,
+      "grad_norm": 5.296870374063085,
+      "learning_rate": 1.3937990844456528e-06,
+      "loss": 0.4213,
+      "step": 8084
+    },
+    {
+      "epoch": 0.6566230812961911,
+      "grad_norm": 3.8879241039555734,
+      "learning_rate": 1.393209381499167e-06,
+      "loss": 0.3738,
+      "step": 8085
+    },
+    {
+      "epoch": 0.6567042962722326,
+      "grad_norm": 4.289208191857055,
+      "learning_rate": 1.3926197551367355e-06,
+      "loss": 0.4136,
+      "step": 8086
+    },
+    {
+      "epoch": 0.6567855112482742,
+      "grad_norm": 5.234771066049035,
+      "learning_rate": 1.3920302053991564e-06,
+      "loss": 0.3542,
+      "step": 8087
+    },
+    {
+      "epoch": 0.6568667262243157,
+      "grad_norm": 5.571643681394787,
+      "learning_rate": 1.3914407323272216e-06,
+      "loss": 0.4513,
+      "step": 8088
+    },
+    {
+      "epoch": 0.6569479412003574,
+      "grad_norm": 7.213207108596739,
+      "learning_rate": 1.3908513359617217e-06,
+      "loss": 0.5061,
+      "step": 8089
+    },
+    {
+      "epoch": 0.6570291561763989,
+      "grad_norm": 4.780947638090955,
+      "learning_rate": 1.39026201634344e-06,
+      "loss": 0.3634,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6571103711524405,
+      "grad_norm": 3.0166870867114053,
+      "learning_rate": 1.3896727735131538e-06,
+      "loss": 0.6613,
+      "step": 8091
+    },
+    {
+      "epoch": 0.657191586128482,
+      "grad_norm": 3.8519737800090974,
+      "learning_rate": 1.3890836075116343e-06,
+      "loss": 0.4642,
+      "step": 8092
+    },
+    {
+      "epoch": 0.6572728011045237,
+      "grad_norm": 4.334939110181824,
+      "learning_rate": 1.3884945183796505e-06,
+      "loss": 0.5912,
+      "step": 8093
+    },
+    {
+      "epoch": 0.6573540160805652,
+      "grad_norm": 6.294374675194089,
+      "learning_rate": 1.3879055061579635e-06,
+      "loss": 0.7303,
+      "step": 8094
+    },
+    {
+      "epoch": 0.6574352310566068,
+      "grad_norm": 3.7598306069040546,
+      "learning_rate": 1.3873165708873286e-06,
+      "loss": 0.5085,
+      "step": 8095
+    },
+    {
+      "epoch": 0.6575164460326485,
+      "grad_norm": 6.863418465636099,
+      "learning_rate": 1.3867277126084989e-06,
+      "loss": 0.5744,
+      "step": 8096
+    },
+    {
+      "epoch": 0.65759766100869,
+      "grad_norm": 7.054233776542322,
+      "learning_rate": 1.3861389313622197e-06,
+      "loss": 0.5576,
+      "step": 8097
+    },
+    {
+      "epoch": 0.6576788759847316,
+      "grad_norm": 4.4365820824808555,
+      "learning_rate": 1.3855502271892313e-06,
+      "loss": 0.423,
+      "step": 8098
+    },
+    {
+      "epoch": 0.6577600909607731,
+      "grad_norm": 6.862928291636716,
+      "learning_rate": 1.3849616001302696e-06,
+      "loss": 0.4621,
+      "step": 8099
+    },
+    {
+      "epoch": 0.6578413059368148,
+      "grad_norm": 5.667668412588925,
+      "learning_rate": 1.3843730502260639e-06,
+      "loss": 0.4926,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6579225209128563,
+      "grad_norm": 4.227345644846646,
+      "learning_rate": 1.3837845775173375e-06,
+      "loss": 0.6139,
+      "step": 8101
+    },
+    {
+      "epoch": 0.6580037358888979,
+      "grad_norm": 4.744805542732661,
+      "learning_rate": 1.383196182044811e-06,
+      "loss": 0.5144,
+      "step": 8102
+    },
+    {
+      "epoch": 0.6580849508649395,
+      "grad_norm": 3.999861099199686,
+      "learning_rate": 1.3826078638491994e-06,
+      "loss": 0.561,
+      "step": 8103
+    },
+    {
+      "epoch": 0.6581661658409811,
+      "grad_norm": 4.070333113027189,
+      "learning_rate": 1.3820196229712085e-06,
+      "loss": 0.4381,
+      "step": 8104
+    },
+    {
+      "epoch": 0.6582473808170226,
+      "grad_norm": 6.884502103701524,
+      "learning_rate": 1.3814314594515443e-06,
+      "loss": 0.6255,
+      "step": 8105
+    },
+    {
+      "epoch": 0.6583285957930642,
+      "grad_norm": 4.918735322058878,
+      "learning_rate": 1.3808433733309028e-06,
+      "loss": 0.548,
+      "step": 8106
+    },
+    {
+      "epoch": 0.6584098107691059,
+      "grad_norm": 16.27622530232076,
+      "learning_rate": 1.380255364649976e-06,
+      "loss": 0.3939,
+      "step": 8107
+    },
+    {
+      "epoch": 0.6584910257451474,
+      "grad_norm": 10.25631338164915,
+      "learning_rate": 1.3796674334494529e-06,
+      "loss": 0.581,
+      "step": 8108
+    },
+    {
+      "epoch": 0.658572240721189,
+      "grad_norm": 5.903422438123663,
+      "learning_rate": 1.3790795797700129e-06,
+      "loss": 0.5879,
+      "step": 8109
+    },
+    {
+      "epoch": 0.6586534556972305,
+      "grad_norm": 5.425671652540845,
+      "learning_rate": 1.3784918036523346e-06,
+      "loss": 0.4939,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6587346706732722,
+      "grad_norm": 6.259094931538714,
+      "learning_rate": 1.377904105137087e-06,
+      "loss": 0.5689,
+      "step": 8111
+    },
+    {
+      "epoch": 0.6588158856493137,
+      "grad_norm": 5.861839266316476,
+      "learning_rate": 1.3773164842649377e-06,
+      "loss": 0.4444,
+      "step": 8112
+    },
+    {
+      "epoch": 0.6588971006253553,
+      "grad_norm": 6.629921332417931,
+      "learning_rate": 1.376728941076546e-06,
+      "loss": 0.3919,
+      "step": 8113
+    },
+    {
+      "epoch": 0.6589783156013969,
+      "grad_norm": 6.496651649704183,
+      "learning_rate": 1.3761414756125658e-06,
+      "loss": 0.5354,
+      "step": 8114
+    },
+    {
+      "epoch": 0.6590595305774385,
+      "grad_norm": 4.3858502738087655,
+      "learning_rate": 1.3755540879136474e-06,
+      "loss": 0.5213,
+      "step": 8115
+    },
+    {
+      "epoch": 0.65914074555348,
+      "grad_norm": 5.705183500815472,
+      "learning_rate": 1.3749667780204365e-06,
+      "loss": 0.5628,
+      "step": 8116
+    },
+    {
+      "epoch": 0.6592219605295216,
+      "grad_norm": 3.5828316864822236,
+      "learning_rate": 1.3743795459735692e-06,
+      "loss": 0.518,
+      "step": 8117
+    },
+    {
+      "epoch": 0.6593031755055633,
+      "grad_norm": 4.1843405148298025,
+      "learning_rate": 1.373792391813681e-06,
+      "loss": 0.6037,
+      "step": 8118
+    },
+    {
+      "epoch": 0.6593843904816048,
+      "grad_norm": 5.543134932829188,
+      "learning_rate": 1.3732053155813987e-06,
+      "loss": 0.3632,
+      "step": 8119
+    },
+    {
+      "epoch": 0.6594656054576464,
+      "grad_norm": 3.9215981353205938,
+      "learning_rate": 1.3726183173173441e-06,
+      "loss": 0.4956,
+      "step": 8120
+    },
+    {
+      "epoch": 0.659546820433688,
+      "grad_norm": 6.56518519854498,
+      "learning_rate": 1.3720313970621369e-06,
+      "loss": 0.479,
+      "step": 8121
+    },
+    {
+      "epoch": 0.6596280354097296,
+      "grad_norm": 8.721756176365712,
+      "learning_rate": 1.3714445548563856e-06,
+      "loss": 0.5228,
+      "step": 8122
+    },
+    {
+      "epoch": 0.6597092503857711,
+      "grad_norm": 6.414976174126078,
+      "learning_rate": 1.3708577907406988e-06,
+      "loss": 0.5035,
+      "step": 8123
+    },
+    {
+      "epoch": 0.6597904653618127,
+      "grad_norm": 5.440110547817854,
+      "learning_rate": 1.3702711047556777e-06,
+      "loss": 0.4776,
+      "step": 8124
+    },
+    {
+      "epoch": 0.6598716803378543,
+      "grad_norm": 5.614130528625265,
+      "learning_rate": 1.3696844969419174e-06,
+      "loss": 0.3457,
+      "step": 8125
+    },
+    {
+      "epoch": 0.6599528953138959,
+      "grad_norm": 4.022888633337103,
+      "learning_rate": 1.3690979673400067e-06,
+      "loss": 0.445,
+      "step": 8126
+    },
+    {
+      "epoch": 0.6600341102899374,
+      "grad_norm": 3.741552498785602,
+      "learning_rate": 1.3685115159905325e-06,
+      "loss": 0.3523,
+      "step": 8127
+    },
+    {
+      "epoch": 0.660115325265979,
+      "grad_norm": 5.950815202545946,
+      "learning_rate": 1.3679251429340717e-06,
+      "loss": 0.738,
+      "step": 8128
+    },
+    {
+      "epoch": 0.6601965402420207,
+      "grad_norm": 9.51105805660359,
+      "learning_rate": 1.367338848211201e-06,
+      "loss": 0.4257,
+      "step": 8129
+    },
+    {
+      "epoch": 0.6602777552180622,
+      "grad_norm": 2.9131542456623314,
+      "learning_rate": 1.3667526318624862e-06,
+      "loss": 0.4203,
+      "step": 8130
+    },
+    {
+      "epoch": 0.6603589701941038,
+      "grad_norm": 6.754300006832807,
+      "learning_rate": 1.366166493928493e-06,
+      "loss": 0.6132,
+      "step": 8131
+    },
+    {
+      "epoch": 0.6604401851701454,
+      "grad_norm": 11.724662577057464,
+      "learning_rate": 1.3655804344497775e-06,
+      "loss": 0.4595,
+      "step": 8132
+    },
+    {
+      "epoch": 0.660521400146187,
+      "grad_norm": 5.387659360154549,
+      "learning_rate": 1.364994453466891e-06,
+      "loss": 0.4081,
+      "step": 8133
+    },
+    {
+      "epoch": 0.6606026151222285,
+      "grad_norm": 5.107977084416256,
+      "learning_rate": 1.3644085510203813e-06,
+      "loss": 0.4898,
+      "step": 8134
+    },
+    {
+      "epoch": 0.6606838300982701,
+      "grad_norm": 3.8833833190727183,
+      "learning_rate": 1.363822727150791e-06,
+      "loss": 0.5417,
+      "step": 8135
+    },
+    {
+      "epoch": 0.6607650450743117,
+      "grad_norm": 7.8141538363601795,
+      "learning_rate": 1.363236981898654e-06,
+      "loss": 0.4968,
+      "step": 8136
+    },
+    {
+      "epoch": 0.6608462600503533,
+      "grad_norm": 3.590897276111095,
+      "learning_rate": 1.3626513153045024e-06,
+      "loss": 0.5024,
+      "step": 8137
+    },
+    {
+      "epoch": 0.6609274750263948,
+      "grad_norm": 3.612475255145642,
+      "learning_rate": 1.3620657274088606e-06,
+      "loss": 0.4954,
+      "step": 8138
+    },
+    {
+      "epoch": 0.6610086900024364,
+      "grad_norm": 14.567400918143866,
+      "learning_rate": 1.3614802182522469e-06,
+      "loss": 0.668,
+      "step": 8139
+    },
+    {
+      "epoch": 0.6610899049784781,
+      "grad_norm": 11.017893005217958,
+      "learning_rate": 1.3608947878751777e-06,
+      "loss": 0.4346,
+      "step": 8140
+    },
+    {
+      "epoch": 0.6611711199545196,
+      "grad_norm": 3.7950075970163617,
+      "learning_rate": 1.3603094363181596e-06,
+      "loss": 0.5231,
+      "step": 8141
+    },
+    {
+      "epoch": 0.6612523349305612,
+      "grad_norm": 7.66686320728164,
+      "learning_rate": 1.3597241636216965e-06,
+      "loss": 0.6696,
+      "step": 8142
+    },
+    {
+      "epoch": 0.6613335499066028,
+      "grad_norm": 5.608044387753557,
+      "learning_rate": 1.3591389698262875e-06,
+      "loss": 0.5376,
+      "step": 8143
+    },
+    {
+      "epoch": 0.6614147648826444,
+      "grad_norm": 4.03444037697512,
+      "learning_rate": 1.3585538549724242e-06,
+      "loss": 0.526,
+      "step": 8144
+    },
+    {
+      "epoch": 0.6614959798586859,
+      "grad_norm": 5.082198135850972,
+      "learning_rate": 1.3579688191005926e-06,
+      "loss": 0.5056,
+      "step": 8145
+    },
+    {
+      "epoch": 0.6615771948347275,
+      "grad_norm": 3.3444467710171346,
+      "learning_rate": 1.3573838622512743e-06,
+      "loss": 0.6148,
+      "step": 8146
+    },
+    {
+      "epoch": 0.6616584098107691,
+      "grad_norm": 4.212117525236555,
+      "learning_rate": 1.3567989844649448e-06,
+      "loss": 0.5947,
+      "step": 8147
+    },
+    {
+      "epoch": 0.6617396247868107,
+      "grad_norm": 5.234823635180376,
+      "learning_rate": 1.3562141857820765e-06,
+      "loss": 0.5863,
+      "step": 8148
+    },
+    {
+      "epoch": 0.6618208397628522,
+      "grad_norm": 9.716406215450048,
+      "learning_rate": 1.3556294662431325e-06,
+      "loss": 0.4493,
+      "step": 8149
+    },
+    {
+      "epoch": 0.6619020547388939,
+      "grad_norm": 5.722283978029896,
+      "learning_rate": 1.3550448258885734e-06,
+      "loss": 0.4364,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6619832697149355,
+      "grad_norm": 4.763814608798664,
+      "learning_rate": 1.3544602647588528e-06,
+      "loss": 0.742,
+      "step": 8151
+    },
+    {
+      "epoch": 0.662064484690977,
+      "grad_norm": 5.80232988441981,
+      "learning_rate": 1.3538757828944188e-06,
+      "loss": 0.4423,
+      "step": 8152
+    },
+    {
+      "epoch": 0.6621456996670186,
+      "grad_norm": 7.5957193433728145,
+      "learning_rate": 1.353291380335715e-06,
+      "loss": 0.5588,
+      "step": 8153
+    },
+    {
+      "epoch": 0.6622269146430602,
+      "grad_norm": 4.336543165787254,
+      "learning_rate": 1.3527070571231786e-06,
+      "loss": 0.5969,
+      "step": 8154
+    },
+    {
+      "epoch": 0.6623081296191018,
+      "grad_norm": 4.034182275364077,
+      "learning_rate": 1.3521228132972414e-06,
+      "loss": 0.4803,
+      "step": 8155
+    },
+    {
+      "epoch": 0.6623893445951433,
+      "grad_norm": 6.6167843827609865,
+      "learning_rate": 1.3515386488983317e-06,
+      "loss": 0.4521,
+      "step": 8156
+    },
+    {
+      "epoch": 0.662470559571185,
+      "grad_norm": 7.161780111206926,
+      "learning_rate": 1.3509545639668691e-06,
+      "loss": 0.3106,
+      "step": 8157
+    },
+    {
+      "epoch": 0.6625517745472265,
+      "grad_norm": 5.4096275250827945,
+      "learning_rate": 1.3503705585432687e-06,
+      "loss": 0.4119,
+      "step": 8158
+    },
+    {
+      "epoch": 0.6626329895232681,
+      "grad_norm": 7.137614931386881,
+      "learning_rate": 1.349786632667942e-06,
+      "loss": 0.6203,
+      "step": 8159
+    },
+    {
+      "epoch": 0.6627142044993096,
+      "grad_norm": 4.038451931034832,
+      "learning_rate": 1.3492027863812924e-06,
+      "loss": 0.749,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6627954194753513,
+      "grad_norm": 4.101256157779943,
+      "learning_rate": 1.3486190197237189e-06,
+      "loss": 0.5766,
+      "step": 8161
+    },
+    {
+      "epoch": 0.6628766344513929,
+      "grad_norm": 3.827717281308639,
+      "learning_rate": 1.348035332735617e-06,
+      "loss": 0.3621,
+      "step": 8162
+    },
+    {
+      "epoch": 0.6629578494274344,
+      "grad_norm": 3.2797754851269856,
+      "learning_rate": 1.3474517254573731e-06,
+      "loss": 0.6257,
+      "step": 8163
+    },
+    {
+      "epoch": 0.663039064403476,
+      "grad_norm": 7.6698914171017245,
+      "learning_rate": 1.3468681979293702e-06,
+      "loss": 0.413,
+      "step": 8164
+    },
+    {
+      "epoch": 0.6631202793795176,
+      "grad_norm": 6.602624984901768,
+      "learning_rate": 1.3462847501919843e-06,
+      "loss": 0.3934,
+      "step": 8165
+    },
+    {
+      "epoch": 0.6632014943555592,
+      "grad_norm": 7.1560351882255215,
+      "learning_rate": 1.3457013822855886e-06,
+      "loss": 0.7646,
+      "step": 8166
+    },
+    {
+      "epoch": 0.6632827093316007,
+      "grad_norm": 6.274238416078604,
+      "learning_rate": 1.345118094250547e-06,
+      "loss": 0.408,
+      "step": 8167
+    },
+    {
+      "epoch": 0.6633639243076423,
+      "grad_norm": 4.400441017779448,
+      "learning_rate": 1.3445348861272217e-06,
+      "loss": 0.4495,
+      "step": 8168
+    },
+    {
+      "epoch": 0.6634451392836839,
+      "grad_norm": 4.941239230227576,
+      "learning_rate": 1.3439517579559675e-06,
+      "loss": 0.3428,
+      "step": 8169
+    },
+    {
+      "epoch": 0.6635263542597255,
+      "grad_norm": 6.964343826754793,
+      "learning_rate": 1.3433687097771337e-06,
+      "loss": 0.4209,
+      "step": 8170
+    },
+    {
+      "epoch": 0.663607569235767,
+      "grad_norm": 5.512709601692857,
+      "learning_rate": 1.3427857416310626e-06,
+      "loss": 0.5785,
+      "step": 8171
+    },
+    {
+      "epoch": 0.6636887842118087,
+      "grad_norm": 4.294567753627923,
+      "learning_rate": 1.3422028535580947e-06,
+      "loss": 0.4546,
+      "step": 8172
+    },
+    {
+      "epoch": 0.6637699991878503,
+      "grad_norm": 4.0701024523531135,
+      "learning_rate": 1.3416200455985607e-06,
+      "loss": 0.8187,
+      "step": 8173
+    },
+    {
+      "epoch": 0.6638512141638918,
+      "grad_norm": 7.655445454850961,
+      "learning_rate": 1.3410373177927893e-06,
+      "loss": 0.4059,
+      "step": 8174
+    },
+    {
+      "epoch": 0.6639324291399334,
+      "grad_norm": 2.7798324043491838,
+      "learning_rate": 1.3404546701811022e-06,
+      "loss": 0.5913,
+      "step": 8175
+    },
+    {
+      "epoch": 0.664013644115975,
+      "grad_norm": 3.646058875181944,
+      "learning_rate": 1.3398721028038155e-06,
+      "loss": 0.5026,
+      "step": 8176
+    },
+    {
+      "epoch": 0.6640948590920166,
+      "grad_norm": 4.496311517930436,
+      "learning_rate": 1.3392896157012386e-06,
+      "loss": 0.6419,
+      "step": 8177
+    },
+    {
+      "epoch": 0.6641760740680581,
+      "grad_norm": 17.84153072314891,
+      "learning_rate": 1.3387072089136776e-06,
+      "loss": 0.5958,
+      "step": 8178
+    },
+    {
+      "epoch": 0.6642572890440998,
+      "grad_norm": 3.8720341467833834,
+      "learning_rate": 1.3381248824814326e-06,
+      "loss": 0.5721,
+      "step": 8179
+    },
+    {
+      "epoch": 0.6643385040201413,
+      "grad_norm": 4.7847332539123935,
+      "learning_rate": 1.337542636444795e-06,
+      "loss": 0.44,
+      "step": 8180
+    },
+    {
+      "epoch": 0.6644197189961829,
+      "grad_norm": 6.021015620083501,
+      "learning_rate": 1.3369604708440548e-06,
+      "loss": 0.3977,
+      "step": 8181
+    },
+    {
+      "epoch": 0.6645009339722244,
+      "grad_norm": 11.600335989741517,
+      "learning_rate": 1.3363783857194957e-06,
+      "loss": 0.469,
+      "step": 8182
+    },
+    {
+      "epoch": 0.6645821489482661,
+      "grad_norm": 3.859966985311204,
+      "learning_rate": 1.3357963811113938e-06,
+      "loss": 0.4263,
+      "step": 8183
+    },
+    {
+      "epoch": 0.6646633639243077,
+      "grad_norm": 18.782716428504518,
+      "learning_rate": 1.3352144570600203e-06,
+      "loss": 0.6374,
+      "step": 8184
+    },
+    {
+      "epoch": 0.6647445789003492,
+      "grad_norm": 3.387091226384912,
+      "learning_rate": 1.3346326136056425e-06,
+      "loss": 0.4308,
+      "step": 8185
+    },
+    {
+      "epoch": 0.6648257938763908,
+      "grad_norm": 4.252596910764206,
+      "learning_rate": 1.3340508507885194e-06,
+      "loss": 0.5107,
+      "step": 8186
+    },
+    {
+      "epoch": 0.6649070088524324,
+      "grad_norm": 7.6038955526056995,
+      "learning_rate": 1.3334691686489064e-06,
+      "loss": 0.4946,
+      "step": 8187
+    },
+    {
+      "epoch": 0.664988223828474,
+      "grad_norm": 4.976169303513984,
+      "learning_rate": 1.3328875672270547e-06,
+      "loss": 0.4381,
+      "step": 8188
+    },
+    {
+      "epoch": 0.6650694388045155,
+      "grad_norm": 4.667554378601018,
+      "learning_rate": 1.332306046563206e-06,
+      "loss": 0.6222,
+      "step": 8189
+    },
+    {
+      "epoch": 0.6651506537805572,
+      "grad_norm": 4.379049488125857,
+      "learning_rate": 1.3317246066975981e-06,
+      "loss": 0.6358,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6652318687565987,
+      "grad_norm": 3.794618143261633,
+      "learning_rate": 1.3311432476704655e-06,
+      "loss": 0.4243,
+      "step": 8191
+    },
+    {
+      "epoch": 0.6653130837326403,
+      "grad_norm": 5.072184639847683,
+      "learning_rate": 1.3305619695220332e-06,
+      "loss": 0.5623,
+      "step": 8192
+    },
+    {
+      "epoch": 0.6653942987086818,
+      "grad_norm": 7.59051423468418,
+      "learning_rate": 1.3299807722925231e-06,
+      "loss": 0.4993,
+      "step": 8193
+    },
+    {
+      "epoch": 0.6654755136847235,
+      "grad_norm": 8.819021574798183,
+      "learning_rate": 1.3293996560221526e-06,
+      "loss": 0.4178,
+      "step": 8194
+    },
+    {
+      "epoch": 0.6655567286607651,
+      "grad_norm": 6.287522877063349,
+      "learning_rate": 1.3288186207511303e-06,
+      "loss": 0.3488,
+      "step": 8195
+    },
+    {
+      "epoch": 0.6656379436368066,
+      "grad_norm": 5.563934142204648,
+      "learning_rate": 1.3282376665196603e-06,
+      "loss": 0.4812,
+      "step": 8196
+    },
+    {
+      "epoch": 0.6657191586128482,
+      "grad_norm": 3.033804351653404,
+      "learning_rate": 1.327656793367943e-06,
+      "loss": 0.4407,
+      "step": 8197
+    },
+    {
+      "epoch": 0.6658003735888898,
+      "grad_norm": 3.998217455418668,
+      "learning_rate": 1.3270760013361713e-06,
+      "loss": 0.4281,
+      "step": 8198
+    },
+    {
+      "epoch": 0.6658815885649314,
+      "grad_norm": 4.64706121281191,
+      "learning_rate": 1.3264952904645317e-06,
+      "loss": 0.6847,
+      "step": 8199
+    },
+    {
+      "epoch": 0.6659628035409729,
+      "grad_norm": 6.436231964294674,
+      "learning_rate": 1.325914660793207e-06,
+      "loss": 0.4931,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6660440185170146,
+      "grad_norm": 3.707584274007003,
+      "learning_rate": 1.3253341123623756e-06,
+      "loss": 0.5129,
+      "step": 8201
+    },
+    {
+      "epoch": 0.6661252334930561,
+      "grad_norm": 6.262622978288038,
+      "learning_rate": 1.3247536452122064e-06,
+      "loss": 0.5491,
+      "step": 8202
+    },
+    {
+      "epoch": 0.6662064484690977,
+      "grad_norm": 4.982037820278029,
+      "learning_rate": 1.3241732593828644e-06,
+      "loss": 0.4305,
+      "step": 8203
+    },
+    {
+      "epoch": 0.6662876634451392,
+      "grad_norm": 6.381634581129901,
+      "learning_rate": 1.3235929549145105e-06,
+      "loss": 0.4896,
+      "step": 8204
+    },
+    {
+      "epoch": 0.6663688784211809,
+      "grad_norm": 7.969785119936784,
+      "learning_rate": 1.3230127318472972e-06,
+      "loss": 0.5373,
+      "step": 8205
+    },
+    {
+      "epoch": 0.6664500933972225,
+      "grad_norm": 6.0137267480338945,
+      "learning_rate": 1.3224325902213736e-06,
+      "loss": 0.476,
+      "step": 8206
+    },
+    {
+      "epoch": 0.666531308373264,
+      "grad_norm": 6.68986461696678,
+      "learning_rate": 1.3218525300768837e-06,
+      "loss": 0.4782,
+      "step": 8207
+    },
+    {
+      "epoch": 0.6666125233493057,
+      "grad_norm": 6.316034448019237,
+      "learning_rate": 1.3212725514539635e-06,
+      "loss": 0.4462,
+      "step": 8208
+    },
+    {
+      "epoch": 0.6666937383253472,
+      "grad_norm": 5.228047836489257,
+      "learning_rate": 1.3206926543927435e-06,
+      "loss": 0.3812,
+      "step": 8209
+    },
+    {
+      "epoch": 0.6667749533013888,
+      "grad_norm": 6.369850283074102,
+      "learning_rate": 1.320112838933351e-06,
+      "loss": 0.469,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6668561682774303,
+      "grad_norm": 4.859894104230745,
+      "learning_rate": 1.3195331051159058e-06,
+      "loss": 0.4444,
+      "step": 8211
+    },
+    {
+      "epoch": 0.666937383253472,
+      "grad_norm": 4.0624444842962735,
+      "learning_rate": 1.3189534529805212e-06,
+      "loss": 0.4554,
+      "step": 8212
+    },
+    {
+      "epoch": 0.6670185982295135,
+      "grad_norm": 3.4817004946398287,
+      "learning_rate": 1.318373882567307e-06,
+      "loss": 0.5201,
+      "step": 8213
+    },
+    {
+      "epoch": 0.6670998132055551,
+      "grad_norm": 4.704982903669908,
+      "learning_rate": 1.3177943939163677e-06,
+      "loss": 0.4444,
+      "step": 8214
+    },
+    {
+      "epoch": 0.6671810281815966,
+      "grad_norm": 5.794219397597491,
+      "learning_rate": 1.3172149870677985e-06,
+      "loss": 0.4768,
+      "step": 8215
+    },
+    {
+      "epoch": 0.6672622431576383,
+      "grad_norm": 6.208905832820019,
+      "learning_rate": 1.3166356620616932e-06,
+      "loss": 0.5479,
+      "step": 8216
+    },
+    {
+      "epoch": 0.6673434581336799,
+      "grad_norm": 8.198766362247154,
+      "learning_rate": 1.3160564189381376e-06,
+      "loss": 0.4532,
+      "step": 8217
+    },
+    {
+      "epoch": 0.6674246731097214,
+      "grad_norm": 4.331763449279288,
+      "learning_rate": 1.3154772577372104e-06,
+      "loss": 0.4955,
+      "step": 8218
+    },
+    {
+      "epoch": 0.667505888085763,
+      "grad_norm": 5.10837125089982,
+      "learning_rate": 1.3148981784989884e-06,
+      "loss": 0.4523,
+      "step": 8219
+    },
+    {
+      "epoch": 0.6675871030618046,
+      "grad_norm": 3.5163841336361807,
+      "learning_rate": 1.3143191812635408e-06,
+      "loss": 0.6125,
+      "step": 8220
+    },
+    {
+      "epoch": 0.6676683180378462,
+      "grad_norm": 11.844210364989475,
+      "learning_rate": 1.3137402660709314e-06,
+      "loss": 0.4425,
+      "step": 8221
+    },
+    {
+      "epoch": 0.6677495330138877,
+      "grad_norm": 7.996601522252169,
+      "learning_rate": 1.3131614329612158e-06,
+      "loss": 0.6123,
+      "step": 8222
+    },
+    {
+      "epoch": 0.6678307479899294,
+      "grad_norm": 6.114487257718311,
+      "learning_rate": 1.3125826819744493e-06,
+      "loss": 0.5059,
+      "step": 8223
+    },
+    {
+      "epoch": 0.6679119629659709,
+      "grad_norm": 4.43142805188028,
+      "learning_rate": 1.3120040131506767e-06,
+      "loss": 0.4303,
+      "step": 8224
+    },
+    {
+      "epoch": 0.6679931779420125,
+      "grad_norm": 5.687962653675712,
+      "learning_rate": 1.3114254265299379e-06,
+      "loss": 0.4586,
+      "step": 8225
+    },
+    {
+      "epoch": 0.668074392918054,
+      "grad_norm": 4.1918238610968475,
+      "learning_rate": 1.310846922152269e-06,
+      "loss": 0.5445,
+      "step": 8226
+    },
+    {
+      "epoch": 0.6681556078940957,
+      "grad_norm": 5.179655298264121,
+      "learning_rate": 1.310268500057701e-06,
+      "loss": 0.5172,
+      "step": 8227
+    },
+    {
+      "epoch": 0.6682368228701373,
+      "grad_norm": 7.779574674479231,
+      "learning_rate": 1.309690160286255e-06,
+      "loss": 0.5864,
+      "step": 8228
+    },
+    {
+      "epoch": 0.6683180378461788,
+      "grad_norm": 6.122052156829049,
+      "learning_rate": 1.3091119028779514e-06,
+      "loss": 0.548,
+      "step": 8229
+    },
+    {
+      "epoch": 0.6683992528222205,
+      "grad_norm": 4.789126863334509,
+      "learning_rate": 1.308533727872801e-06,
+      "loss": 0.4313,
+      "step": 8230
+    },
+    {
+      "epoch": 0.668480467798262,
+      "grad_norm": 5.766521551851652,
+      "learning_rate": 1.3079556353108106e-06,
+      "loss": 0.532,
+      "step": 8231
+    },
+    {
+      "epoch": 0.6685616827743036,
+      "grad_norm": 6.05863385781591,
+      "learning_rate": 1.307377625231981e-06,
+      "loss": 0.443,
+      "step": 8232
+    },
+    {
+      "epoch": 0.6686428977503451,
+      "grad_norm": 4.181674552131915,
+      "learning_rate": 1.3067996976763086e-06,
+      "loss": 0.5801,
+      "step": 8233
+    },
+    {
+      "epoch": 0.6687241127263868,
+      "grad_norm": 3.206551391730835,
+      "learning_rate": 1.3062218526837828e-06,
+      "loss": 0.5704,
+      "step": 8234
+    },
+    {
+      "epoch": 0.6688053277024283,
+      "grad_norm": 10.62932913375947,
+      "learning_rate": 1.3056440902943856e-06,
+      "loss": 0.4123,
+      "step": 8235
+    },
+    {
+      "epoch": 0.6688865426784699,
+      "grad_norm": 4.544980879348178,
+      "learning_rate": 1.305066410548097e-06,
+      "loss": 0.5647,
+      "step": 8236
+    },
+    {
+      "epoch": 0.6689677576545114,
+      "grad_norm": 5.011004929662773,
+      "learning_rate": 1.304488813484889e-06,
+      "loss": 0.5001,
+      "step": 8237
+    },
+    {
+      "epoch": 0.6690489726305531,
+      "grad_norm": 5.610021787019349,
+      "learning_rate": 1.303911299144727e-06,
+      "loss": 0.4763,
+      "step": 8238
+    },
+    {
+      "epoch": 0.6691301876065947,
+      "grad_norm": 87.36388987621034,
+      "learning_rate": 1.3033338675675726e-06,
+      "loss": 0.4336,
+      "step": 8239
+    },
+    {
+      "epoch": 0.6692114025826362,
+      "grad_norm": 3.9673278123195566,
+      "learning_rate": 1.3027565187933828e-06,
+      "loss": 0.4406,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6692926175586779,
+      "grad_norm": 4.0150982480618165,
+      "learning_rate": 1.3021792528621041e-06,
+      "loss": 0.4505,
+      "step": 8241
+    },
+    {
+      "epoch": 0.6693738325347194,
+      "grad_norm": 3.8387858911160473,
+      "learning_rate": 1.3016020698136827e-06,
+      "loss": 0.4101,
+      "step": 8242
+    },
+    {
+      "epoch": 0.669455047510761,
+      "grad_norm": 3.2719359158581245,
+      "learning_rate": 1.3010249696880558e-06,
+      "loss": 0.5057,
+      "step": 8243
+    },
+    {
+      "epoch": 0.6695362624868025,
+      "grad_norm": 8.689632620129073,
+      "learning_rate": 1.3004479525251545e-06,
+      "loss": 0.3984,
+      "step": 8244
+    },
+    {
+      "epoch": 0.6696174774628442,
+      "grad_norm": 4.334649919352921,
+      "learning_rate": 1.2998710183649066e-06,
+      "loss": 0.48,
+      "step": 8245
+    },
+    {
+      "epoch": 0.6696986924388857,
+      "grad_norm": 3.604389609859541,
+      "learning_rate": 1.2992941672472332e-06,
+      "loss": 0.4935,
+      "step": 8246
+    },
+    {
+      "epoch": 0.6697799074149273,
+      "grad_norm": 4.775702711045651,
+      "learning_rate": 1.2987173992120478e-06,
+      "loss": 0.4603,
+      "step": 8247
+    },
+    {
+      "epoch": 0.6698611223909688,
+      "grad_norm": 5.179644996464377,
+      "learning_rate": 1.2981407142992618e-06,
+      "loss": 0.4972,
+      "step": 8248
+    },
+    {
+      "epoch": 0.6699423373670105,
+      "grad_norm": 4.832968999586265,
+      "learning_rate": 1.2975641125487777e-06,
+      "loss": 0.5587,
+      "step": 8249
+    },
+    {
+      "epoch": 0.6700235523430521,
+      "grad_norm": 4.3612144381710065,
+      "learning_rate": 1.2969875940004923e-06,
+      "loss": 0.335,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6701047673190936,
+      "grad_norm": 6.055517461311078,
+      "learning_rate": 1.2964111586942996e-06,
+      "loss": 0.6306,
+      "step": 8251
+    },
+    {
+      "epoch": 0.6701859822951353,
+      "grad_norm": 3.996269903879433,
+      "learning_rate": 1.2958348066700833e-06,
+      "loss": 0.5962,
+      "step": 8252
+    },
+    {
+      "epoch": 0.6702671972711768,
+      "grad_norm": 6.099178781598084,
+      "learning_rate": 1.2952585379677268e-06,
+      "loss": 0.453,
+      "step": 8253
+    },
+    {
+      "epoch": 0.6703484122472184,
+      "grad_norm": 6.116289571948397,
+      "learning_rate": 1.2946823526271023e-06,
+      "loss": 0.4299,
+      "step": 8254
+    },
+    {
+      "epoch": 0.6704296272232599,
+      "grad_norm": 5.092428454455709,
+      "learning_rate": 1.2941062506880811e-06,
+      "loss": 0.5551,
+      "step": 8255
+    },
+    {
+      "epoch": 0.6705108421993016,
+      "grad_norm": 5.682161639107994,
+      "learning_rate": 1.2935302321905252e-06,
+      "loss": 0.6384,
+      "step": 8256
+    },
+    {
+      "epoch": 0.6705920571753431,
+      "grad_norm": 4.3807459106680575,
+      "learning_rate": 1.292954297174291e-06,
+      "loss": 0.6603,
+      "step": 8257
+    },
+    {
+      "epoch": 0.6706732721513847,
+      "grad_norm": 3.7416878890359517,
+      "learning_rate": 1.2923784456792314e-06,
+      "loss": 0.4831,
+      "step": 8258
+    },
+    {
+      "epoch": 0.6707544871274262,
+      "grad_norm": 3.7742295359579874,
+      "learning_rate": 1.291802677745193e-06,
+      "loss": 0.4317,
+      "step": 8259
+    },
+    {
+      "epoch": 0.6708357021034679,
+      "grad_norm": 4.136648173116798,
+      "learning_rate": 1.2912269934120142e-06,
+      "loss": 0.5397,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6709169170795095,
+      "grad_norm": 10.177523748838187,
+      "learning_rate": 1.2906513927195308e-06,
+      "loss": 0.4942,
+      "step": 8261
+    },
+    {
+      "epoch": 0.670998132055551,
+      "grad_norm": 5.041065114037551,
+      "learning_rate": 1.290075875707571e-06,
+      "loss": 0.3985,
+      "step": 8262
+    },
+    {
+      "epoch": 0.6710793470315927,
+      "grad_norm": 6.204016855298799,
+      "learning_rate": 1.2895004424159557e-06,
+      "loss": 0.4884,
+      "step": 8263
+    },
+    {
+      "epoch": 0.6711605620076342,
+      "grad_norm": 6.451950713200213,
+      "learning_rate": 1.2889250928845038e-06,
+      "loss": 0.3397,
+      "step": 8264
+    },
+    {
+      "epoch": 0.6712417769836758,
+      "grad_norm": 5.789167573594882,
+      "learning_rate": 1.2883498271530265e-06,
+      "loss": 0.5267,
+      "step": 8265
+    },
+    {
+      "epoch": 0.6713229919597173,
+      "grad_norm": 3.020900821630963,
+      "learning_rate": 1.2877746452613277e-06,
+      "loss": 0.5771,
+      "step": 8266
+    },
+    {
+      "epoch": 0.671404206935759,
+      "grad_norm": 7.136575144733493,
+      "learning_rate": 1.2871995472492088e-06,
+      "loss": 0.4145,
+      "step": 8267
+    },
+    {
+      "epoch": 0.6714854219118005,
+      "grad_norm": 6.86909884452662,
+      "learning_rate": 1.2866245331564627e-06,
+      "loss": 0.4899,
+      "step": 8268
+    },
+    {
+      "epoch": 0.6715666368878421,
+      "grad_norm": 4.392832664582703,
+      "learning_rate": 1.2860496030228763e-06,
+      "loss": 0.7517,
+      "step": 8269
+    },
+    {
+      "epoch": 0.6716478518638836,
+      "grad_norm": 4.036722831992974,
+      "learning_rate": 1.2854747568882336e-06,
+      "loss": 0.4853,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6717290668399253,
+      "grad_norm": 4.039548813778698,
+      "learning_rate": 1.2848999947923089e-06,
+      "loss": 0.3944,
+      "step": 8271
+    },
+    {
+      "epoch": 0.6718102818159669,
+      "grad_norm": 4.0410979346726625,
+      "learning_rate": 1.2843253167748745e-06,
+      "loss": 0.4949,
+      "step": 8272
+    },
+    {
+      "epoch": 0.6718914967920084,
+      "grad_norm": 4.170268917322463,
+      "learning_rate": 1.2837507228756934e-06,
+      "loss": 0.7473,
+      "step": 8273
+    },
+    {
+      "epoch": 0.6719727117680501,
+      "grad_norm": 5.339351144770748,
+      "learning_rate": 1.2831762131345265e-06,
+      "loss": 0.3525,
+      "step": 8274
+    },
+    {
+      "epoch": 0.6720539267440916,
+      "grad_norm": 4.558326943954262,
+      "learning_rate": 1.2826017875911257e-06,
+      "loss": 0.5092,
+      "step": 8275
+    },
+    {
+      "epoch": 0.6721351417201332,
+      "grad_norm": 6.121469512209196,
+      "learning_rate": 1.2820274462852373e-06,
+      "loss": 0.3959,
+      "step": 8276
+    },
+    {
+      "epoch": 0.6722163566961747,
+      "grad_norm": 6.247553742372245,
+      "learning_rate": 1.2814531892566034e-06,
+      "loss": 0.4574,
+      "step": 8277
+    },
+    {
+      "epoch": 0.6722975716722164,
+      "grad_norm": 3.743110767471212,
+      "learning_rate": 1.2808790165449609e-06,
+      "loss": 0.4817,
+      "step": 8278
+    },
+    {
+      "epoch": 0.6723787866482579,
+      "grad_norm": 3.847254853806929,
+      "learning_rate": 1.280304928190037e-06,
+      "loss": 0.4439,
+      "step": 8279
+    },
+    {
+      "epoch": 0.6724600016242995,
+      "grad_norm": 5.111654340531649,
+      "learning_rate": 1.2797309242315584e-06,
+      "loss": 0.5166,
+      "step": 8280
+    },
+    {
+      "epoch": 0.672541216600341,
+      "grad_norm": 5.402468457889438,
+      "learning_rate": 1.2791570047092413e-06,
+      "loss": 0.4379,
+      "step": 8281
+    },
+    {
+      "epoch": 0.6726224315763827,
+      "grad_norm": 4.64148894732247,
+      "learning_rate": 1.2785831696627975e-06,
+      "loss": 0.4985,
+      "step": 8282
+    },
+    {
+      "epoch": 0.6727036465524243,
+      "grad_norm": 4.977759194515003,
+      "learning_rate": 1.2780094191319348e-06,
+      "loss": 0.3378,
+      "step": 8283
+    },
+    {
+      "epoch": 0.6727848615284658,
+      "grad_norm": 5.750546485168828,
+      "learning_rate": 1.2774357531563522e-06,
+      "loss": 0.5315,
+      "step": 8284
+    },
+    {
+      "epoch": 0.6728660765045075,
+      "grad_norm": 5.462788945858097,
+      "learning_rate": 1.276862171775745e-06,
+      "loss": 0.611,
+      "step": 8285
+    },
+    {
+      "epoch": 0.672947291480549,
+      "grad_norm": 9.040739646165454,
+      "learning_rate": 1.2762886750298033e-06,
+      "loss": 0.4435,
+      "step": 8286
+    },
+    {
+      "epoch": 0.6730285064565906,
+      "grad_norm": 5.419042990213366,
+      "learning_rate": 1.275715262958209e-06,
+      "loss": 0.585,
+      "step": 8287
+    },
+    {
+      "epoch": 0.6731097214326321,
+      "grad_norm": 8.622423592062997,
+      "learning_rate": 1.275141935600639e-06,
+      "loss": 0.4452,
+      "step": 8288
+    },
+    {
+      "epoch": 0.6731909364086738,
+      "grad_norm": 3.39665996833734,
+      "learning_rate": 1.2745686929967632e-06,
+      "loss": 0.5745,
+      "step": 8289
+    },
+    {
+      "epoch": 0.6732721513847153,
+      "grad_norm": 9.640446040935274,
+      "learning_rate": 1.2739955351862488e-06,
+      "loss": 0.4325,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6733533663607569,
+      "grad_norm": 3.375245810027421,
+      "learning_rate": 1.2734224622087556e-06,
+      "loss": 0.742,
+      "step": 8291
+    },
+    {
+      "epoch": 0.6734345813367985,
+      "grad_norm": 6.070864539301688,
+      "learning_rate": 1.2728494741039354e-06,
+      "loss": 0.4282,
+      "step": 8292
+    },
+    {
+      "epoch": 0.6735157963128401,
+      "grad_norm": 3.1473742664954956,
+      "learning_rate": 1.2722765709114382e-06,
+      "loss": 0.5101,
+      "step": 8293
+    },
+    {
+      "epoch": 0.6735970112888817,
+      "grad_norm": 5.797541760307167,
+      "learning_rate": 1.2717037526709048e-06,
+      "loss": 0.4927,
+      "step": 8294
+    },
+    {
+      "epoch": 0.6736782262649232,
+      "grad_norm": 3.9968403294119192,
+      "learning_rate": 1.2711310194219695e-06,
+      "loss": 0.5272,
+      "step": 8295
+    },
+    {
+      "epoch": 0.6737594412409649,
+      "grad_norm": 8.686594608534724,
+      "learning_rate": 1.2705583712042654e-06,
+      "loss": 0.3865,
+      "step": 8296
+    },
+    {
+      "epoch": 0.6738406562170064,
+      "grad_norm": 5.299840930228158,
+      "learning_rate": 1.2699858080574141e-06,
+      "loss": 0.6049,
+      "step": 8297
+    },
+    {
+      "epoch": 0.673921871193048,
+      "grad_norm": 5.383383957246399,
+      "learning_rate": 1.2694133300210354e-06,
+      "loss": 0.5769,
+      "step": 8298
+    },
+    {
+      "epoch": 0.6740030861690895,
+      "grad_norm": 5.48276722761316,
+      "learning_rate": 1.2688409371347422e-06,
+      "loss": 0.5609,
+      "step": 8299
+    },
+    {
+      "epoch": 0.6740843011451312,
+      "grad_norm": 4.890248642899915,
+      "learning_rate": 1.2682686294381403e-06,
+      "loss": 0.5895,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6741655161211727,
+      "grad_norm": 5.286027918182194,
+      "learning_rate": 1.2676964069708294e-06,
+      "loss": 0.406,
+      "step": 8301
+    },
+    {
+      "epoch": 0.6742467310972143,
+      "grad_norm": 4.144062037889405,
+      "learning_rate": 1.2671242697724061e-06,
+      "loss": 0.6805,
+      "step": 8302
+    },
+    {
+      "epoch": 0.6743279460732559,
+      "grad_norm": 3.10497422112379,
+      "learning_rate": 1.266552217882458e-06,
+      "loss": 0.4172,
+      "step": 8303
+    },
+    {
+      "epoch": 0.6744091610492975,
+      "grad_norm": 4.7159885122704495,
+      "learning_rate": 1.265980251340568e-06,
+      "loss": 0.3766,
+      "step": 8304
+    },
+    {
+      "epoch": 0.6744903760253391,
+      "grad_norm": 6.029787369464984,
+      "learning_rate": 1.265408370186315e-06,
+      "loss": 0.5068,
+      "step": 8305
+    },
+    {
+      "epoch": 0.6745715910013806,
+      "grad_norm": 9.477273387771245,
+      "learning_rate": 1.2648365744592683e-06,
+      "loss": 0.5335,
+      "step": 8306
+    },
+    {
+      "epoch": 0.6746528059774223,
+      "grad_norm": 5.274931921378115,
+      "learning_rate": 1.264264864198994e-06,
+      "loss": 0.3893,
+      "step": 8307
+    },
+    {
+      "epoch": 0.6747340209534638,
+      "grad_norm": 6.694925532405975,
+      "learning_rate": 1.2636932394450502e-06,
+      "loss": 0.3245,
+      "step": 8308
+    },
+    {
+      "epoch": 0.6748152359295054,
+      "grad_norm": 3.4329112146537994,
+      "learning_rate": 1.2631217002369917e-06,
+      "loss": 0.5724,
+      "step": 8309
+    },
+    {
+      "epoch": 0.674896450905547,
+      "grad_norm": 5.211887506168933,
+      "learning_rate": 1.2625502466143646e-06,
+      "loss": 0.6977,
+      "step": 8310
+    },
+    {
+      "epoch": 0.6749776658815886,
+      "grad_norm": 3.3614053784547258,
+      "learning_rate": 1.2619788786167113e-06,
+      "loss": 0.4674,
+      "step": 8311
+    },
+    {
+      "epoch": 0.6750588808576301,
+      "grad_norm": 3.329679930838226,
+      "learning_rate": 1.2614075962835688e-06,
+      "loss": 0.4857,
+      "step": 8312
+    },
+    {
+      "epoch": 0.6751400958336717,
+      "grad_norm": 5.737443755358583,
+      "learning_rate": 1.2608363996544654e-06,
+      "loss": 0.5106,
+      "step": 8313
+    },
+    {
+      "epoch": 0.6752213108097133,
+      "grad_norm": 6.125567742013193,
+      "learning_rate": 1.2602652887689237e-06,
+      "loss": 0.4443,
+      "step": 8314
+    },
+    {
+      "epoch": 0.6753025257857549,
+      "grad_norm": 9.237383534994807,
+      "learning_rate": 1.2596942636664638e-06,
+      "loss": 0.4483,
+      "step": 8315
+    },
+    {
+      "epoch": 0.6753837407617965,
+      "grad_norm": 5.106127682199849,
+      "learning_rate": 1.2591233243865958e-06,
+      "loss": 0.4114,
+      "step": 8316
+    },
+    {
+      "epoch": 0.675464955737838,
+      "grad_norm": 6.956467222846337,
+      "learning_rate": 1.2585524709688268e-06,
+      "loss": 0.5567,
+      "step": 8317
+    },
+    {
+      "epoch": 0.6755461707138797,
+      "grad_norm": 6.165264566742239,
+      "learning_rate": 1.257981703452657e-06,
+      "loss": 0.4624,
+      "step": 8318
+    },
+    {
+      "epoch": 0.6756273856899212,
+      "grad_norm": 3.908854775644603,
+      "learning_rate": 1.2574110218775804e-06,
+      "loss": 0.479,
+      "step": 8319
+    },
+    {
+      "epoch": 0.6757086006659628,
+      "grad_norm": 5.299167034989398,
+      "learning_rate": 1.2568404262830836e-06,
+      "loss": 0.4723,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6757898156420044,
+      "grad_norm": 5.113207491776007,
+      "learning_rate": 1.256269916708651e-06,
+      "loss": 0.4358,
+      "step": 8321
+    },
+    {
+      "epoch": 0.675871030618046,
+      "grad_norm": 6.490258586098834,
+      "learning_rate": 1.2556994931937565e-06,
+      "loss": 0.4825,
+      "step": 8322
+    },
+    {
+      "epoch": 0.6759522455940875,
+      "grad_norm": 5.544338315054307,
+      "learning_rate": 1.2551291557778721e-06,
+      "loss": 0.4548,
+      "step": 8323
+    },
+    {
+      "epoch": 0.6760334605701291,
+      "grad_norm": 4.16562191240731,
+      "learning_rate": 1.2545589045004627e-06,
+      "loss": 0.5272,
+      "step": 8324
+    },
+    {
+      "epoch": 0.6761146755461707,
+      "grad_norm": 8.317648547442525,
+      "learning_rate": 1.2539887394009855e-06,
+      "loss": 0.5807,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6761958905222123,
+      "grad_norm": 4.863432965134022,
+      "learning_rate": 1.2534186605188933e-06,
+      "loss": 0.4716,
+      "step": 8326
+    },
+    {
+      "epoch": 0.6762771054982539,
+      "grad_norm": 4.473812223631319,
+      "learning_rate": 1.2528486678936313e-06,
+      "loss": 0.3962,
+      "step": 8327
+    },
+    {
+      "epoch": 0.6763583204742954,
+      "grad_norm": 7.282308721393281,
+      "learning_rate": 1.2522787615646421e-06,
+      "loss": 0.4288,
+      "step": 8328
+    },
+    {
+      "epoch": 0.6764395354503371,
+      "grad_norm": 6.136375474933964,
+      "learning_rate": 1.251708941571358e-06,
+      "loss": 0.4208,
+      "step": 8329
+    },
+    {
+      "epoch": 0.6765207504263786,
+      "grad_norm": 8.250435251721406,
+      "learning_rate": 1.2511392079532087e-06,
+      "loss": 0.3731,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6766019654024202,
+      "grad_norm": 8.8618136486422,
+      "learning_rate": 1.2505695607496176e-06,
+      "loss": 0.4431,
+      "step": 8331
+    },
+    {
+      "epoch": 0.6766831803784618,
+      "grad_norm": 6.809802423773869,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.437,
+      "step": 8332
+    },
+    {
+      "epoch": 0.6767643953545034,
+      "grad_norm": 4.412176686499365,
+      "learning_rate": 1.2494305257437669e-06,
+      "loss": 0.6663,
+      "step": 8333
+    },
+    {
+      "epoch": 0.6768456103305449,
+      "grad_norm": 8.259279196472452,
+      "learning_rate": 1.2488611380203234e-06,
+      "loss": 0.5233,
+      "step": 8334
+    },
+    {
+      "epoch": 0.6769268253065865,
+      "grad_norm": 5.681125220205126,
+      "learning_rate": 1.2482918368690666e-06,
+      "loss": 0.4985,
+      "step": 8335
+    },
+    {
+      "epoch": 0.6770080402826281,
+      "grad_norm": 8.25956808119689,
+      "learning_rate": 1.24772262232939e-06,
+      "loss": 0.5764,
+      "step": 8336
+    },
+    {
+      "epoch": 0.6770892552586697,
+      "grad_norm": 5.608911686082642,
+      "learning_rate": 1.2471534944406813e-06,
+      "loss": 0.5972,
+      "step": 8337
+    },
+    {
+      "epoch": 0.6771704702347113,
+      "grad_norm": 3.3213708352263054,
+      "learning_rate": 1.2465844532423201e-06,
+      "loss": 0.4957,
+      "step": 8338
+    },
+    {
+      "epoch": 0.6772516852107529,
+      "grad_norm": 7.167255021326034,
+      "learning_rate": 1.2460154987736806e-06,
+      "loss": 0.5549,
+      "step": 8339
+    },
+    {
+      "epoch": 0.6773329001867945,
+      "grad_norm": 4.442760200949274,
+      "learning_rate": 1.2454466310741326e-06,
+      "loss": 0.427,
+      "step": 8340
+    },
+    {
+      "epoch": 0.677414115162836,
+      "grad_norm": 5.768478872426691,
+      "learning_rate": 1.244877850183038e-06,
+      "loss": 0.4285,
+      "step": 8341
+    },
+    {
+      "epoch": 0.6774953301388776,
+      "grad_norm": 4.225339833775357,
+      "learning_rate": 1.2443091561397527e-06,
+      "loss": 0.5469,
+      "step": 8342
+    },
+    {
+      "epoch": 0.6775765451149192,
+      "grad_norm": 6.322105186178602,
+      "learning_rate": 1.2437405489836282e-06,
+      "loss": 0.4678,
+      "step": 8343
+    },
+    {
+      "epoch": 0.6776577600909608,
+      "grad_norm": 12.80035724783234,
+      "learning_rate": 1.2431720287540097e-06,
+      "loss": 0.4633,
+      "step": 8344
+    },
+    {
+      "epoch": 0.6777389750670023,
+      "grad_norm": 6.865099951483512,
+      "learning_rate": 1.2426035954902356e-06,
+      "loss": 0.5027,
+      "step": 8345
+    },
+    {
+      "epoch": 0.677820190043044,
+      "grad_norm": 4.2665462756693975,
+      "learning_rate": 1.2420352492316368e-06,
+      "loss": 0.4114,
+      "step": 8346
+    },
+    {
+      "epoch": 0.6779014050190855,
+      "grad_norm": 4.207313035828698,
+      "learning_rate": 1.2414669900175423e-06,
+      "loss": 0.4601,
+      "step": 8347
+    },
+    {
+      "epoch": 0.6779826199951271,
+      "grad_norm": 5.802065214743892,
+      "learning_rate": 1.2408988178872699e-06,
+      "loss": 0.3487,
+      "step": 8348
+    },
+    {
+      "epoch": 0.6780638349711687,
+      "grad_norm": 5.071217203345165,
+      "learning_rate": 1.240330732880136e-06,
+      "loss": 0.4018,
+      "step": 8349
+    },
+    {
+      "epoch": 0.6781450499472103,
+      "grad_norm": 4.658634987558885,
+      "learning_rate": 1.2397627350354494e-06,
+      "loss": 0.5006,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6782262649232519,
+      "grad_norm": 5.989526021591514,
+      "learning_rate": 1.2391948243925119e-06,
+      "loss": 0.3748,
+      "step": 8351
+    },
+    {
+      "epoch": 0.6783074798992934,
+      "grad_norm": 4.543912419314957,
+      "learning_rate": 1.238627000990619e-06,
+      "loss": 0.3414,
+      "step": 8352
+    },
+    {
+      "epoch": 0.678388694875335,
+      "grad_norm": 4.935095050500174,
+      "learning_rate": 1.2380592648690629e-06,
+      "loss": 0.4755,
+      "step": 8353
+    },
+    {
+      "epoch": 0.6784699098513766,
+      "grad_norm": 3.9622208310484024,
+      "learning_rate": 1.2374916160671268e-06,
+      "loss": 0.4053,
+      "step": 8354
+    },
+    {
+      "epoch": 0.6785511248274182,
+      "grad_norm": 4.65626323141216,
+      "learning_rate": 1.2369240546240881e-06,
+      "loss": 0.5236,
+      "step": 8355
+    },
+    {
+      "epoch": 0.6786323398034597,
+      "grad_norm": 5.707169042711556,
+      "learning_rate": 1.2363565805792202e-06,
+      "loss": 0.4848,
+      "step": 8356
+    },
+    {
+      "epoch": 0.6787135547795013,
+      "grad_norm": 5.695511088930679,
+      "learning_rate": 1.2357891939717903e-06,
+      "loss": 0.4024,
+      "step": 8357
+    },
+    {
+      "epoch": 0.6787947697555429,
+      "grad_norm": 5.28913545779317,
+      "learning_rate": 1.2352218948410563e-06,
+      "loss": 0.4305,
+      "step": 8358
+    },
+    {
+      "epoch": 0.6788759847315845,
+      "grad_norm": 3.680524773252283,
+      "learning_rate": 1.2346546832262743e-06,
+      "loss": 0.5757,
+      "step": 8359
+    },
+    {
+      "epoch": 0.6789571997076261,
+      "grad_norm": 4.937519409498853,
+      "learning_rate": 1.2340875591666917e-06,
+      "loss": 0.5091,
+      "step": 8360
+    },
+    {
+      "epoch": 0.6790384146836677,
+      "grad_norm": 6.252444692643531,
+      "learning_rate": 1.2335205227015494e-06,
+      "loss": 0.3707,
+      "step": 8361
+    },
+    {
+      "epoch": 0.6791196296597093,
+      "grad_norm": 6.749516651304727,
+      "learning_rate": 1.2329535738700838e-06,
+      "loss": 0.4521,
+      "step": 8362
+    },
+    {
+      "epoch": 0.6792008446357508,
+      "grad_norm": 6.554192643461585,
+      "learning_rate": 1.232386712711526e-06,
+      "loss": 0.4651,
+      "step": 8363
+    },
+    {
+      "epoch": 0.6792820596117924,
+      "grad_norm": 6.372502999598231,
+      "learning_rate": 1.2318199392650993e-06,
+      "loss": 0.5727,
+      "step": 8364
+    },
+    {
+      "epoch": 0.679363274587834,
+      "grad_norm": 7.402271446804732,
+      "learning_rate": 1.23125325357002e-06,
+      "loss": 0.3813,
+      "step": 8365
+    },
+    {
+      "epoch": 0.6794444895638756,
+      "grad_norm": 4.56010296625159,
+      "learning_rate": 1.2306866556655016e-06,
+      "loss": 0.3856,
+      "step": 8366
+    },
+    {
+      "epoch": 0.6795257045399171,
+      "grad_norm": 5.070138865153647,
+      "learning_rate": 1.2301201455907492e-06,
+      "loss": 0.3799,
+      "step": 8367
+    },
+    {
+      "epoch": 0.6796069195159588,
+      "grad_norm": 7.907043621920382,
+      "learning_rate": 1.2295537233849608e-06,
+      "loss": 0.4336,
+      "step": 8368
+    },
+    {
+      "epoch": 0.6796881344920003,
+      "grad_norm": 4.083402933054175,
+      "learning_rate": 1.2289873890873311e-06,
+      "loss": 0.4805,
+      "step": 8369
+    },
+    {
+      "epoch": 0.6797693494680419,
+      "grad_norm": 3.3225519411749134,
+      "learning_rate": 1.2284211427370483e-06,
+      "loss": 0.5801,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6798505644440835,
+      "grad_norm": 5.893743990272938,
+      "learning_rate": 1.2278549843732915e-06,
+      "loss": 0.4576,
+      "step": 8371
+    },
+    {
+      "epoch": 0.6799317794201251,
+      "grad_norm": 4.351833235603589,
+      "learning_rate": 1.2272889140352382e-06,
+      "loss": 0.5446,
+      "step": 8372
+    },
+    {
+      "epoch": 0.6800129943961667,
+      "grad_norm": 6.1661761775892945,
+      "learning_rate": 1.2267229317620564e-06,
+      "loss": 0.4169,
+      "step": 8373
+    },
+    {
+      "epoch": 0.6800942093722082,
+      "grad_norm": 3.8116371519348404,
+      "learning_rate": 1.2261570375929077e-06,
+      "loss": 0.4341,
+      "step": 8374
+    },
+    {
+      "epoch": 0.6801754243482498,
+      "grad_norm": 8.305006268325108,
+      "learning_rate": 1.2255912315669507e-06,
+      "loss": 0.5365,
+      "step": 8375
+    },
+    {
+      "epoch": 0.6802566393242914,
+      "grad_norm": 3.87785735314886,
+      "learning_rate": 1.2250255137233363e-06,
+      "loss": 0.3122,
+      "step": 8376
+    },
+    {
+      "epoch": 0.680337854300333,
+      "grad_norm": 5.364139835040124,
+      "learning_rate": 1.224459884101209e-06,
+      "loss": 0.4254,
+      "step": 8377
+    },
+    {
+      "epoch": 0.6804190692763745,
+      "grad_norm": 5.505941463868022,
+      "learning_rate": 1.2238943427397059e-06,
+      "loss": 0.5725,
+      "step": 8378
+    },
+    {
+      "epoch": 0.6805002842524162,
+      "grad_norm": 3.2670918728153495,
+      "learning_rate": 1.2233288896779617e-06,
+      "loss": 0.4677,
+      "step": 8379
+    },
+    {
+      "epoch": 0.6805814992284577,
+      "grad_norm": 5.650997362758408,
+      "learning_rate": 1.2227635249551014e-06,
+      "loss": 0.4548,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6806627142044993,
+      "grad_norm": 4.85029252716426,
+      "learning_rate": 1.2221982486102446e-06,
+      "loss": 0.533,
+      "step": 8381
+    },
+    {
+      "epoch": 0.6807439291805409,
+      "grad_norm": 6.520961416314167,
+      "learning_rate": 1.2216330606825063e-06,
+      "loss": 0.5872,
+      "step": 8382
+    },
+    {
+      "epoch": 0.6808251441565825,
+      "grad_norm": 4.358523321449324,
+      "learning_rate": 1.2210679612109957e-06,
+      "loss": 0.463,
+      "step": 8383
+    },
+    {
+      "epoch": 0.6809063591326241,
+      "grad_norm": 3.3319401584526256,
+      "learning_rate": 1.2205029502348123e-06,
+      "loss": 0.4915,
+      "step": 8384
+    },
+    {
+      "epoch": 0.6809875741086656,
+      "grad_norm": 5.916515103224009,
+      "learning_rate": 1.2199380277930542e-06,
+      "loss": 0.3504,
+      "step": 8385
+    },
+    {
+      "epoch": 0.6810687890847072,
+      "grad_norm": 3.720584300258901,
+      "learning_rate": 1.2193731939248098e-06,
+      "loss": 0.5314,
+      "step": 8386
+    },
+    {
+      "epoch": 0.6811500040607488,
+      "grad_norm": 6.706752035654295,
+      "learning_rate": 1.218808448669162e-06,
+      "loss": 0.4826,
+      "step": 8387
+    },
+    {
+      "epoch": 0.6812312190367904,
+      "grad_norm": 7.136044672639504,
+      "learning_rate": 1.218243792065189e-06,
+      "loss": 0.3701,
+      "step": 8388
+    },
+    {
+      "epoch": 0.6813124340128319,
+      "grad_norm": 15.62272184000503,
+      "learning_rate": 1.2176792241519628e-06,
+      "loss": 0.511,
+      "step": 8389
+    },
+    {
+      "epoch": 0.6813936489888736,
+      "grad_norm": 5.7775925878468275,
+      "learning_rate": 1.2171147449685469e-06,
+      "loss": 0.4626,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6814748639649151,
+      "grad_norm": 8.76813533021635,
+      "learning_rate": 1.2165503545540017e-06,
+      "loss": 0.3721,
+      "step": 8391
+    },
+    {
+      "epoch": 0.6815560789409567,
+      "grad_norm": 2.7821535648469307,
+      "learning_rate": 1.2159860529473796e-06,
+      "loss": 0.5439,
+      "step": 8392
+    },
+    {
+      "epoch": 0.6816372939169983,
+      "grad_norm": 8.296103570527087,
+      "learning_rate": 1.2154218401877263e-06,
+      "loss": 0.4945,
+      "step": 8393
+    },
+    {
+      "epoch": 0.6817185088930399,
+      "grad_norm": 3.680639071610561,
+      "learning_rate": 1.214857716314083e-06,
+      "loss": 0.5059,
+      "step": 8394
+    },
+    {
+      "epoch": 0.6817997238690815,
+      "grad_norm": 3.5629979371942997,
+      "learning_rate": 1.2142936813654848e-06,
+      "loss": 0.4787,
+      "step": 8395
+    },
+    {
+      "epoch": 0.681880938845123,
+      "grad_norm": 13.02646445886847,
+      "learning_rate": 1.21372973538096e-06,
+      "loss": 0.5225,
+      "step": 8396
+    },
+    {
+      "epoch": 0.6819621538211647,
+      "grad_norm": 6.585522603186978,
+      "learning_rate": 1.2131658783995285e-06,
+      "loss": 0.3445,
+      "step": 8397
+    },
+    {
+      "epoch": 0.6820433687972062,
+      "grad_norm": 4.999188751543709,
+      "learning_rate": 1.212602110460209e-06,
+      "loss": 0.6751,
+      "step": 8398
+    },
+    {
+      "epoch": 0.6821245837732478,
+      "grad_norm": 5.137382890011477,
+      "learning_rate": 1.2120384316020098e-06,
+      "loss": 0.464,
+      "step": 8399
+    },
+    {
+      "epoch": 0.6822057987492893,
+      "grad_norm": 3.57144580039995,
+      "learning_rate": 1.2114748418639339e-06,
+      "loss": 0.3924,
+      "step": 8400
+    },
+    {
+      "epoch": 0.682287013725331,
+      "grad_norm": 6.448424519995378,
+      "learning_rate": 1.2109113412849792e-06,
+      "loss": 0.5204,
+      "step": 8401
+    },
+    {
+      "epoch": 0.6823682287013725,
+      "grad_norm": 4.80310339792193,
+      "learning_rate": 1.2103479299041388e-06,
+      "loss": 0.5435,
+      "step": 8402
+    },
+    {
+      "epoch": 0.6824494436774141,
+      "grad_norm": 11.6534555123514,
+      "learning_rate": 1.209784607760395e-06,
+      "loss": 0.514,
+      "step": 8403
+    },
+    {
+      "epoch": 0.6825306586534557,
+      "grad_norm": 14.006283811833011,
+      "learning_rate": 1.209221374892729e-06,
+      "loss": 0.5202,
+      "step": 8404
+    },
+    {
+      "epoch": 0.6826118736294973,
+      "grad_norm": 3.410546012906801,
+      "learning_rate": 1.2086582313401125e-06,
+      "loss": 0.6359,
+      "step": 8405
+    },
+    {
+      "epoch": 0.6826930886055389,
+      "grad_norm": 4.650396850410136,
+      "learning_rate": 1.208095177141511e-06,
+      "loss": 0.6307,
+      "step": 8406
+    },
+    {
+      "epoch": 0.6827743035815804,
+      "grad_norm": 4.067175768365543,
+      "learning_rate": 1.2075322123358857e-06,
+      "loss": 0.5884,
+      "step": 8407
+    },
+    {
+      "epoch": 0.682855518557622,
+      "grad_norm": 7.834411232708035,
+      "learning_rate": 1.2069693369621924e-06,
+      "loss": 0.4241,
+      "step": 8408
+    },
+    {
+      "epoch": 0.6829367335336636,
+      "grad_norm": 5.025485826761079,
+      "learning_rate": 1.2064065510593765e-06,
+      "loss": 0.4524,
+      "step": 8409
+    },
+    {
+      "epoch": 0.6830179485097052,
+      "grad_norm": 7.200545827802387,
+      "learning_rate": 1.205843854666382e-06,
+      "loss": 0.6125,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6830991634857467,
+      "grad_norm": 5.911187736116562,
+      "learning_rate": 1.2052812478221437e-06,
+      "loss": 0.6645,
+      "step": 8411
+    },
+    {
+      "epoch": 0.6831803784617884,
+      "grad_norm": 7.850988389705168,
+      "learning_rate": 1.2047187305655898e-06,
+      "loss": 0.4591,
+      "step": 8412
+    },
+    {
+      "epoch": 0.6832615934378299,
+      "grad_norm": 4.511296495523972,
+      "learning_rate": 1.2041563029356454e-06,
+      "loss": 0.6595,
+      "step": 8413
+    },
+    {
+      "epoch": 0.6833428084138715,
+      "grad_norm": 5.3954091188162145,
+      "learning_rate": 1.203593964971226e-06,
+      "loss": 0.5137,
+      "step": 8414
+    },
+    {
+      "epoch": 0.6834240233899131,
+      "grad_norm": 3.303858814227981,
+      "learning_rate": 1.2030317167112438e-06,
+      "loss": 0.5769,
+      "step": 8415
+    },
+    {
+      "epoch": 0.6835052383659547,
+      "grad_norm": 3.763109647115813,
+      "learning_rate": 1.2024695581946016e-06,
+      "loss": 0.4826,
+      "step": 8416
+    },
+    {
+      "epoch": 0.6835864533419963,
+      "grad_norm": 17.92152213202901,
+      "learning_rate": 1.2019074894602005e-06,
+      "loss": 0.3603,
+      "step": 8417
+    },
+    {
+      "epoch": 0.6836676683180378,
+      "grad_norm": 6.268708930822519,
+      "learning_rate": 1.2013455105469304e-06,
+      "loss": 0.4264,
+      "step": 8418
+    },
+    {
+      "epoch": 0.6837488832940795,
+      "grad_norm": 4.52113050826669,
+      "learning_rate": 1.2007836214936773e-06,
+      "loss": 0.5712,
+      "step": 8419
+    },
+    {
+      "epoch": 0.683830098270121,
+      "grad_norm": 5.948974609805699,
+      "learning_rate": 1.2002218223393213e-06,
+      "loss": 0.5343,
+      "step": 8420
+    },
+    {
+      "epoch": 0.6839113132461626,
+      "grad_norm": 5.13305028563167,
+      "learning_rate": 1.1996601131227376e-06,
+      "loss": 0.6497,
+      "step": 8421
+    },
+    {
+      "epoch": 0.6839925282222041,
+      "grad_norm": 6.071521774254982,
+      "learning_rate": 1.1990984938827907e-06,
+      "loss": 0.6081,
+      "step": 8422
+    },
+    {
+      "epoch": 0.6840737431982458,
+      "grad_norm": 4.917477348975693,
+      "learning_rate": 1.1985369646583442e-06,
+      "loss": 0.5464,
+      "step": 8423
+    },
+    {
+      "epoch": 0.6841549581742873,
+      "grad_norm": 4.304996098971284,
+      "learning_rate": 1.1979755254882519e-06,
+      "loss": 0.4677,
+      "step": 8424
+    },
+    {
+      "epoch": 0.6842361731503289,
+      "grad_norm": 7.848305808110954,
+      "learning_rate": 1.1974141764113617e-06,
+      "loss": 0.3471,
+      "step": 8425
+    },
+    {
+      "epoch": 0.6843173881263706,
+      "grad_norm": 5.645856868747358,
+      "learning_rate": 1.1968529174665173e-06,
+      "loss": 0.4901,
+      "step": 8426
+    },
+    {
+      "epoch": 0.6843986031024121,
+      "grad_norm": 5.273055048309659,
+      "learning_rate": 1.1962917486925532e-06,
+      "loss": 0.5054,
+      "step": 8427
+    },
+    {
+      "epoch": 0.6844798180784537,
+      "grad_norm": 4.090778363630613,
+      "learning_rate": 1.1957306701283002e-06,
+      "loss": 0.3776,
+      "step": 8428
+    },
+    {
+      "epoch": 0.6845610330544952,
+      "grad_norm": 6.031487669992356,
+      "learning_rate": 1.1951696818125835e-06,
+      "loss": 0.5705,
+      "step": 8429
+    },
+    {
+      "epoch": 0.6846422480305369,
+      "grad_norm": 5.241073571056581,
+      "learning_rate": 1.1946087837842188e-06,
+      "loss": 0.3882,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6847234630065784,
+      "grad_norm": 6.683258623229494,
+      "learning_rate": 1.1940479760820177e-06,
+      "loss": 0.486,
+      "step": 8431
+    },
+    {
+      "epoch": 0.68480467798262,
+      "grad_norm": 5.092498342825334,
+      "learning_rate": 1.1934872587447838e-06,
+      "loss": 0.4291,
+      "step": 8432
+    },
+    {
+      "epoch": 0.6848858929586615,
+      "grad_norm": 2.9158897378436746,
+      "learning_rate": 1.1929266318113172e-06,
+      "loss": 0.5162,
+      "step": 8433
+    },
+    {
+      "epoch": 0.6849671079347032,
+      "grad_norm": 6.199252123306311,
+      "learning_rate": 1.192366095320411e-06,
+      "loss": 0.4291,
+      "step": 8434
+    },
+    {
+      "epoch": 0.6850483229107447,
+      "grad_norm": 6.833595326483498,
+      "learning_rate": 1.1918056493108493e-06,
+      "loss": 0.398,
+      "step": 8435
+    },
+    {
+      "epoch": 0.6851295378867863,
+      "grad_norm": 4.488586364488279,
+      "learning_rate": 1.1912452938214142e-06,
+      "loss": 0.5889,
+      "step": 8436
+    },
+    {
+      "epoch": 0.685210752862828,
+      "grad_norm": 7.767681900407471,
+      "learning_rate": 1.1906850288908783e-06,
+      "loss": 0.5559,
+      "step": 8437
+    },
+    {
+      "epoch": 0.6852919678388695,
+      "grad_norm": 4.340043418686966,
+      "learning_rate": 1.1901248545580082e-06,
+      "loss": 0.4359,
+      "step": 8438
+    },
+    {
+      "epoch": 0.6853731828149111,
+      "grad_norm": 4.93465575158808,
+      "learning_rate": 1.1895647708615665e-06,
+      "loss": 0.4433,
+      "step": 8439
+    },
+    {
+      "epoch": 0.6854543977909526,
+      "grad_norm": 4.964788137277078,
+      "learning_rate": 1.1890047778403063e-06,
+      "loss": 0.4608,
+      "step": 8440
+    },
+    {
+      "epoch": 0.6855356127669943,
+      "grad_norm": 4.5014832518222905,
+      "learning_rate": 1.1884448755329772e-06,
+      "loss": 0.6192,
+      "step": 8441
+    },
+    {
+      "epoch": 0.6856168277430358,
+      "grad_norm": 3.814073027261063,
+      "learning_rate": 1.1878850639783224e-06,
+      "loss": 0.65,
+      "step": 8442
+    },
+    {
+      "epoch": 0.6856980427190774,
+      "grad_norm": 4.698487957966528,
+      "learning_rate": 1.1873253432150769e-06,
+      "loss": 0.4156,
+      "step": 8443
+    },
+    {
+      "epoch": 0.6857792576951189,
+      "grad_norm": 2.6979949526342173,
+      "learning_rate": 1.1867657132819693e-06,
+      "loss": 0.4989,
+      "step": 8444
+    },
+    {
+      "epoch": 0.6858604726711606,
+      "grad_norm": 4.663840811404981,
+      "learning_rate": 1.1862061742177253e-06,
+      "loss": 0.3499,
+      "step": 8445
+    },
+    {
+      "epoch": 0.6859416876472021,
+      "grad_norm": 4.603946365527777,
+      "learning_rate": 1.1856467260610597e-06,
+      "loss": 0.4283,
+      "step": 8446
+    },
+    {
+      "epoch": 0.6860229026232437,
+      "grad_norm": 6.004017761597916,
+      "learning_rate": 1.1850873688506847e-06,
+      "loss": 0.4741,
+      "step": 8447
+    },
+    {
+      "epoch": 0.6861041175992854,
+      "grad_norm": 4.91811982386808,
+      "learning_rate": 1.1845281026253055e-06,
+      "loss": 0.4488,
+      "step": 8448
+    },
+    {
+      "epoch": 0.6861853325753269,
+      "grad_norm": 4.124631666496995,
+      "learning_rate": 1.1839689274236197e-06,
+      "loss": 0.4576,
+      "step": 8449
+    },
+    {
+      "epoch": 0.6862665475513685,
+      "grad_norm": 4.967951805934966,
+      "learning_rate": 1.183409843284319e-06,
+      "loss": 0.4204,
+      "step": 8450
+    },
+    {
+      "epoch": 0.68634776252741,
+      "grad_norm": 3.6946567527289083,
+      "learning_rate": 1.1828508502460884e-06,
+      "loss": 0.501,
+      "step": 8451
+    },
+    {
+      "epoch": 0.6864289775034517,
+      "grad_norm": 4.3257591185306214,
+      "learning_rate": 1.1822919483476089e-06,
+      "loss": 0.5587,
+      "step": 8452
+    },
+    {
+      "epoch": 0.6865101924794932,
+      "grad_norm": 13.53355986373059,
+      "learning_rate": 1.1817331376275518e-06,
+      "loss": 0.3784,
+      "step": 8453
+    },
+    {
+      "epoch": 0.6865914074555348,
+      "grad_norm": 4.686471250690919,
+      "learning_rate": 1.181174418124585e-06,
+      "loss": 0.529,
+      "step": 8454
+    },
+    {
+      "epoch": 0.6866726224315763,
+      "grad_norm": 13.019567449891584,
+      "learning_rate": 1.1806157898773694e-06,
+      "loss": 0.521,
+      "step": 8455
+    },
+    {
+      "epoch": 0.686753837407618,
+      "grad_norm": 5.024504582924606,
+      "learning_rate": 1.1800572529245581e-06,
+      "loss": 0.4579,
+      "step": 8456
+    },
+    {
+      "epoch": 0.6868350523836595,
+      "grad_norm": 5.622740484481374,
+      "learning_rate": 1.1794988073047986e-06,
+      "loss": 0.4946,
+      "step": 8457
+    },
+    {
+      "epoch": 0.6869162673597011,
+      "grad_norm": 3.383472516465761,
+      "learning_rate": 1.1789404530567338e-06,
+      "loss": 0.454,
+      "step": 8458
+    },
+    {
+      "epoch": 0.6869974823357428,
+      "grad_norm": 39.06057548040847,
+      "learning_rate": 1.178382190218997e-06,
+      "loss": 0.4916,
+      "step": 8459
+    },
+    {
+      "epoch": 0.6870786973117843,
+      "grad_norm": 5.13942966732467,
+      "learning_rate": 1.1778240188302181e-06,
+      "loss": 0.5374,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6871599122878259,
+      "grad_norm": 5.2491128634756805,
+      "learning_rate": 1.177265938929021e-06,
+      "loss": 0.4787,
+      "step": 8461
+    },
+    {
+      "epoch": 0.6872411272638674,
+      "grad_norm": 3.883090224748362,
+      "learning_rate": 1.1767079505540198e-06,
+      "loss": 0.4962,
+      "step": 8462
+    },
+    {
+      "epoch": 0.6873223422399091,
+      "grad_norm": 4.459207504966209,
+      "learning_rate": 1.1761500537438246e-06,
+      "loss": 0.4901,
+      "step": 8463
+    },
+    {
+      "epoch": 0.6874035572159506,
+      "grad_norm": 5.526598548112002,
+      "learning_rate": 1.1755922485370397e-06,
+      "loss": 0.6049,
+      "step": 8464
+    },
+    {
+      "epoch": 0.6874847721919922,
+      "grad_norm": 3.9962235924628,
+      "learning_rate": 1.1750345349722611e-06,
+      "loss": 0.6439,
+      "step": 8465
+    },
+    {
+      "epoch": 0.6875659871680337,
+      "grad_norm": 5.430777327111154,
+      "learning_rate": 1.1744769130880814e-06,
+      "loss": 0.5233,
+      "step": 8466
+    },
+    {
+      "epoch": 0.6876472021440754,
+      "grad_norm": 5.478225172555752,
+      "learning_rate": 1.1739193829230833e-06,
+      "loss": 0.4564,
+      "step": 8467
+    },
+    {
+      "epoch": 0.6877284171201169,
+      "grad_norm": 14.175270639083902,
+      "learning_rate": 1.1733619445158465e-06,
+      "loss": 0.5561,
+      "step": 8468
+    },
+    {
+      "epoch": 0.6878096320961585,
+      "grad_norm": 4.726662335713709,
+      "learning_rate": 1.1728045979049421e-06,
+      "loss": 0.439,
+      "step": 8469
+    },
+    {
+      "epoch": 0.6878908470722002,
+      "grad_norm": 6.585567352748735,
+      "learning_rate": 1.1722473431289344e-06,
+      "loss": 0.4297,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6879720620482417,
+      "grad_norm": 7.151511500781442,
+      "learning_rate": 1.1716901802263845e-06,
+      "loss": 0.4433,
+      "step": 8471
+    },
+    {
+      "epoch": 0.6880532770242833,
+      "grad_norm": 13.543001844242685,
+      "learning_rate": 1.171133109235843e-06,
+      "loss": 0.3943,
+      "step": 8472
+    },
+    {
+      "epoch": 0.6881344920003248,
+      "grad_norm": 6.360919578390468,
+      "learning_rate": 1.1705761301958576e-06,
+      "loss": 0.3842,
+      "step": 8473
+    },
+    {
+      "epoch": 0.6882157069763665,
+      "grad_norm": 5.289828859594367,
+      "learning_rate": 1.170019243144969e-06,
+      "loss": 0.3475,
+      "step": 8474
+    },
+    {
+      "epoch": 0.688296921952408,
+      "grad_norm": 6.436765945910618,
+      "learning_rate": 1.16946244812171e-06,
+      "loss": 0.5738,
+      "step": 8475
+    },
+    {
+      "epoch": 0.6883781369284496,
+      "grad_norm": 9.265057636256202,
+      "learning_rate": 1.1689057451646072e-06,
+      "loss": 0.5015,
+      "step": 8476
+    },
+    {
+      "epoch": 0.6884593519044911,
+      "grad_norm": 5.62906939860262,
+      "learning_rate": 1.1683491343121825e-06,
+      "loss": 0.6312,
+      "step": 8477
+    },
+    {
+      "epoch": 0.6885405668805328,
+      "grad_norm": 3.608605687615943,
+      "learning_rate": 1.1677926156029495e-06,
+      "loss": 0.5705,
+      "step": 8478
+    },
+    {
+      "epoch": 0.6886217818565743,
+      "grad_norm": 7.326218578037191,
+      "learning_rate": 1.1672361890754165e-06,
+      "loss": 0.481,
+      "step": 8479
+    },
+    {
+      "epoch": 0.6887029968326159,
+      "grad_norm": 3.101458446740605,
+      "learning_rate": 1.1666798547680871e-06,
+      "loss": 0.5597,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6887842118086576,
+      "grad_norm": 4.57887880856938,
+      "learning_rate": 1.166123612719455e-06,
+      "loss": 0.6213,
+      "step": 8481
+    },
+    {
+      "epoch": 0.6888654267846991,
+      "grad_norm": 10.086781507059271,
+      "learning_rate": 1.1655674629680083e-06,
+      "loss": 0.5009,
+      "step": 8482
+    },
+    {
+      "epoch": 0.6889466417607407,
+      "grad_norm": 5.0242271976636514,
+      "learning_rate": 1.165011405552232e-06,
+      "loss": 0.7435,
+      "step": 8483
+    },
+    {
+      "epoch": 0.6890278567367822,
+      "grad_norm": 7.2774357651661274,
+      "learning_rate": 1.164455440510601e-06,
+      "loss": 0.4607,
+      "step": 8484
+    },
+    {
+      "epoch": 0.6891090717128239,
+      "grad_norm": 4.136550631258544,
+      "learning_rate": 1.1638995678815843e-06,
+      "loss": 0.5113,
+      "step": 8485
+    },
+    {
+      "epoch": 0.6891902866888654,
+      "grad_norm": 5.117305452537372,
+      "learning_rate": 1.1633437877036462e-06,
+      "loss": 0.4338,
+      "step": 8486
+    },
+    {
+      "epoch": 0.689271501664907,
+      "grad_norm": 3.714646489723773,
+      "learning_rate": 1.162788100015245e-06,
+      "loss": 0.5338,
+      "step": 8487
+    },
+    {
+      "epoch": 0.6893527166409485,
+      "grad_norm": 3.9854040499315877,
+      "learning_rate": 1.1622325048548303e-06,
+      "loss": 0.6151,
+      "step": 8488
+    },
+    {
+      "epoch": 0.6894339316169902,
+      "grad_norm": 4.8147850700984804,
+      "learning_rate": 1.1616770022608447e-06,
+      "loss": 0.5104,
+      "step": 8489
+    },
+    {
+      "epoch": 0.6895151465930317,
+      "grad_norm": 12.49347709660076,
+      "learning_rate": 1.161121592271729e-06,
+      "loss": 0.4588,
+      "step": 8490
+    },
+    {
+      "epoch": 0.6895963615690733,
+      "grad_norm": 4.382600842190844,
+      "learning_rate": 1.1605662749259123e-06,
+      "loss": 0.5458,
+      "step": 8491
+    },
+    {
+      "epoch": 0.689677576545115,
+      "grad_norm": 3.4100489659901823,
+      "learning_rate": 1.1600110502618204e-06,
+      "loss": 0.5746,
+      "step": 8492
+    },
+    {
+      "epoch": 0.6897587915211565,
+      "grad_norm": 5.056416537418844,
+      "learning_rate": 1.1594559183178727e-06,
+      "loss": 0.4043,
+      "step": 8493
+    },
+    {
+      "epoch": 0.6898400064971981,
+      "grad_norm": 4.3441729287251105,
+      "learning_rate": 1.158900879132481e-06,
+      "loss": 0.5067,
+      "step": 8494
+    },
+    {
+      "epoch": 0.6899212214732396,
+      "grad_norm": 3.6975358584627447,
+      "learning_rate": 1.1583459327440496e-06,
+      "loss": 0.6198,
+      "step": 8495
+    },
+    {
+      "epoch": 0.6900024364492813,
+      "grad_norm": 5.765909988508904,
+      "learning_rate": 1.1577910791909802e-06,
+      "loss": 0.557,
+      "step": 8496
+    },
+    {
+      "epoch": 0.6900836514253228,
+      "grad_norm": 6.346348844345966,
+      "learning_rate": 1.1572363185116648e-06,
+      "loss": 0.5423,
+      "step": 8497
+    },
+    {
+      "epoch": 0.6901648664013644,
+      "grad_norm": 6.761851681144281,
+      "learning_rate": 1.1566816507444884e-06,
+      "loss": 0.7113,
+      "step": 8498
+    },
+    {
+      "epoch": 0.690246081377406,
+      "grad_norm": 4.9827548056260484,
+      "learning_rate": 1.1561270759278326e-06,
+      "loss": 0.4832,
+      "step": 8499
+    },
+    {
+      "epoch": 0.6903272963534476,
+      "grad_norm": 7.290496908601368,
+      "learning_rate": 1.1555725941000715e-06,
+      "loss": 0.4773,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6904085113294891,
+      "grad_norm": 4.93347902991029,
+      "learning_rate": 1.1550182052995706e-06,
+      "loss": 0.4249,
+      "step": 8501
+    },
+    {
+      "epoch": 0.6904897263055307,
+      "grad_norm": 4.857856203684281,
+      "learning_rate": 1.154463909564693e-06,
+      "loss": 0.4705,
+      "step": 8502
+    },
+    {
+      "epoch": 0.6905709412815724,
+      "grad_norm": 16.432653414970492,
+      "learning_rate": 1.1539097069337913e-06,
+      "loss": 0.4833,
+      "step": 8503
+    },
+    {
+      "epoch": 0.6906521562576139,
+      "grad_norm": 4.517576424813998,
+      "learning_rate": 1.1533555974452128e-06,
+      "loss": 0.5461,
+      "step": 8504
+    },
+    {
+      "epoch": 0.6907333712336555,
+      "grad_norm": 6.387090563981134,
+      "learning_rate": 1.1528015811373004e-06,
+      "loss": 0.44,
+      "step": 8505
+    },
+    {
+      "epoch": 0.690814586209697,
+      "grad_norm": 7.168256261694638,
+      "learning_rate": 1.1522476580483893e-06,
+      "loss": 0.4357,
+      "step": 8506
+    },
+    {
+      "epoch": 0.6908958011857387,
+      "grad_norm": 5.267349807599555,
+      "learning_rate": 1.1516938282168074e-06,
+      "loss": 0.5764,
+      "step": 8507
+    },
+    {
+      "epoch": 0.6909770161617802,
+      "grad_norm": 5.446388451950695,
+      "learning_rate": 1.151140091680876e-06,
+      "loss": 0.4762,
+      "step": 8508
+    },
+    {
+      "epoch": 0.6910582311378218,
+      "grad_norm": 7.776144758103636,
+      "learning_rate": 1.1505864484789122e-06,
+      "loss": 0.4635,
+      "step": 8509
+    },
+    {
+      "epoch": 0.6911394461138634,
+      "grad_norm": 4.652490283523999,
+      "learning_rate": 1.1500328986492246e-06,
+      "loss": 0.4744,
+      "step": 8510
+    },
+    {
+      "epoch": 0.691220661089905,
+      "grad_norm": 5.281761321834011,
+      "learning_rate": 1.149479442230115e-06,
+      "loss": 0.5802,
+      "step": 8511
+    },
+    {
+      "epoch": 0.6913018760659465,
+      "grad_norm": 4.121830278672146,
+      "learning_rate": 1.1489260792598803e-06,
+      "loss": 0.4812,
+      "step": 8512
+    },
+    {
+      "epoch": 0.6913830910419881,
+      "grad_norm": 4.197838163994761,
+      "learning_rate": 1.1483728097768116e-06,
+      "loss": 0.4427,
+      "step": 8513
+    },
+    {
+      "epoch": 0.6914643060180298,
+      "grad_norm": 4.757540075293028,
+      "learning_rate": 1.14781963381919e-06,
+      "loss": 0.4649,
+      "step": 8514
+    },
+    {
+      "epoch": 0.6915455209940713,
+      "grad_norm": 4.228790147362145,
+      "learning_rate": 1.1472665514252943e-06,
+      "loss": 0.4501,
+      "step": 8515
+    },
+    {
+      "epoch": 0.6916267359701129,
+      "grad_norm": 3.198795799968073,
+      "learning_rate": 1.146713562633394e-06,
+      "loss": 0.4265,
+      "step": 8516
+    },
+    {
+      "epoch": 0.6917079509461544,
+      "grad_norm": 3.293645191890074,
+      "learning_rate": 1.1461606674817518e-06,
+      "loss": 0.5258,
+      "step": 8517
+    },
+    {
+      "epoch": 0.6917891659221961,
+      "grad_norm": 4.377855525825729,
+      "learning_rate": 1.1456078660086266e-06,
+      "loss": 0.4102,
+      "step": 8518
+    },
+    {
+      "epoch": 0.6918703808982376,
+      "grad_norm": 9.573511509775727,
+      "learning_rate": 1.1450551582522702e-06,
+      "loss": 0.3486,
+      "step": 8519
+    },
+    {
+      "epoch": 0.6919515958742792,
+      "grad_norm": 3.560198474639802,
+      "learning_rate": 1.1445025442509258e-06,
+      "loss": 0.4863,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6920328108503208,
+      "grad_norm": 6.303365456673007,
+      "learning_rate": 1.1439500240428304e-06,
+      "loss": 0.5363,
+      "step": 8521
+    },
+    {
+      "epoch": 0.6921140258263624,
+      "grad_norm": 5.438200231090006,
+      "learning_rate": 1.1433975976662172e-06,
+      "loss": 0.5908,
+      "step": 8522
+    },
+    {
+      "epoch": 0.6921952408024039,
+      "grad_norm": 7.217503219294745,
+      "learning_rate": 1.1428452651593102e-06,
+      "loss": 0.5535,
+      "step": 8523
+    },
+    {
+      "epoch": 0.6922764557784455,
+      "grad_norm": 3.113332201559408,
+      "learning_rate": 1.142293026560328e-06,
+      "loss": 0.4252,
+      "step": 8524
+    },
+    {
+      "epoch": 0.6923576707544872,
+      "grad_norm": 3.7670638203585396,
+      "learning_rate": 1.1417408819074835e-06,
+      "loss": 0.5357,
+      "step": 8525
+    },
+    {
+      "epoch": 0.6924388857305287,
+      "grad_norm": 9.156807084218489,
+      "learning_rate": 1.1411888312389815e-06,
+      "loss": 0.5814,
+      "step": 8526
+    },
+    {
+      "epoch": 0.6925201007065703,
+      "grad_norm": 7.477340025472298,
+      "learning_rate": 1.1406368745930201e-06,
+      "loss": 0.5775,
+      "step": 8527
+    },
+    {
+      "epoch": 0.6926013156826119,
+      "grad_norm": 5.777723223269901,
+      "learning_rate": 1.140085012007794e-06,
+      "loss": 0.4531,
+      "step": 8528
+    },
+    {
+      "epoch": 0.6926825306586535,
+      "grad_norm": 17.932544473465622,
+      "learning_rate": 1.1395332435214873e-06,
+      "loss": 0.5404,
+      "step": 8529
+    },
+    {
+      "epoch": 0.692763745634695,
+      "grad_norm": 6.420149798086578,
+      "learning_rate": 1.138981569172279e-06,
+      "loss": 0.4261,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6928449606107366,
+      "grad_norm": 9.184826721296508,
+      "learning_rate": 1.1384299889983432e-06,
+      "loss": 0.4474,
+      "step": 8531
+    },
+    {
+      "epoch": 0.6929261755867782,
+      "grad_norm": 8.376036553312167,
+      "learning_rate": 1.1378785030378473e-06,
+      "loss": 0.5644,
+      "step": 8532
+    },
+    {
+      "epoch": 0.6930073905628198,
+      "grad_norm": 4.928837681497556,
+      "learning_rate": 1.137327111328949e-06,
+      "loss": 0.3391,
+      "step": 8533
+    },
+    {
+      "epoch": 0.6930886055388613,
+      "grad_norm": 4.407213199355116,
+      "learning_rate": 1.1367758139098037e-06,
+      "loss": 0.453,
+      "step": 8534
+    },
+    {
+      "epoch": 0.693169820514903,
+      "grad_norm": 4.884167064220795,
+      "learning_rate": 1.1362246108185571e-06,
+      "loss": 0.4148,
+      "step": 8535
+    },
+    {
+      "epoch": 0.6932510354909446,
+      "grad_norm": 5.275613682866477,
+      "learning_rate": 1.135673502093349e-06,
+      "loss": 0.6555,
+      "step": 8536
+    },
+    {
+      "epoch": 0.6933322504669861,
+      "grad_norm": 4.72395706382707,
+      "learning_rate": 1.1351224877723137e-06,
+      "loss": 0.4664,
+      "step": 8537
+    },
+    {
+      "epoch": 0.6934134654430277,
+      "grad_norm": 6.683252761194182,
+      "learning_rate": 1.1345715678935802e-06,
+      "loss": 0.4335,
+      "step": 8538
+    },
+    {
+      "epoch": 0.6934946804190693,
+      "grad_norm": 3.6056858456464704,
+      "learning_rate": 1.1340207424952673e-06,
+      "loss": 0.4784,
+      "step": 8539
+    },
+    {
+      "epoch": 0.6935758953951109,
+      "grad_norm": 7.002949745318934,
+      "learning_rate": 1.133470011615489e-06,
+      "loss": 0.5434,
+      "step": 8540
+    },
+    {
+      "epoch": 0.6936571103711524,
+      "grad_norm": 6.659133719702008,
+      "learning_rate": 1.1329193752923543e-06,
+      "loss": 0.5708,
+      "step": 8541
+    },
+    {
+      "epoch": 0.693738325347194,
+      "grad_norm": 12.35767990560376,
+      "learning_rate": 1.1323688335639637e-06,
+      "loss": 0.4835,
+      "step": 8542
+    },
+    {
+      "epoch": 0.6938195403232356,
+      "grad_norm": 3.77158348926073,
+      "learning_rate": 1.131818386468411e-06,
+      "loss": 0.4171,
+      "step": 8543
+    },
+    {
+      "epoch": 0.6939007552992772,
+      "grad_norm": 5.1758054265529205,
+      "learning_rate": 1.1312680340437848e-06,
+      "loss": 0.4684,
+      "step": 8544
+    },
+    {
+      "epoch": 0.6939819702753187,
+      "grad_norm": 4.915722528576737,
+      "learning_rate": 1.130717776328168e-06,
+      "loss": 0.2615,
+      "step": 8545
+    },
+    {
+      "epoch": 0.6940631852513603,
+      "grad_norm": 3.1585274055113914,
+      "learning_rate": 1.130167613359633e-06,
+      "loss": 0.5899,
+      "step": 8546
+    },
+    {
+      "epoch": 0.694144400227402,
+      "grad_norm": 4.7926238667126055,
+      "learning_rate": 1.1296175451762504e-06,
+      "loss": 0.3803,
+      "step": 8547
+    },
+    {
+      "epoch": 0.6942256152034435,
+      "grad_norm": 4.742004176848653,
+      "learning_rate": 1.129067571816081e-06,
+      "loss": 0.5899,
+      "step": 8548
+    },
+    {
+      "epoch": 0.6943068301794851,
+      "grad_norm": 4.740522772470059,
+      "learning_rate": 1.128517693317179e-06,
+      "loss": 0.5,
+      "step": 8549
+    },
+    {
+      "epoch": 0.6943880451555267,
+      "grad_norm": 3.0796548456527457,
+      "learning_rate": 1.1279679097175944e-06,
+      "loss": 0.4379,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6944692601315683,
+      "grad_norm": 4.872361456956494,
+      "learning_rate": 1.12741822105537e-06,
+      "loss": 0.4375,
+      "step": 8551
+    },
+    {
+      "epoch": 0.6945504751076098,
+      "grad_norm": 3.8126089904616625,
+      "learning_rate": 1.1268686273685391e-06,
+      "loss": 0.41,
+      "step": 8552
+    },
+    {
+      "epoch": 0.6946316900836514,
+      "grad_norm": 4.918083532251475,
+      "learning_rate": 1.1263191286951333e-06,
+      "loss": 0.7237,
+      "step": 8553
+    },
+    {
+      "epoch": 0.694712905059693,
+      "grad_norm": 4.980898147720527,
+      "learning_rate": 1.1257697250731735e-06,
+      "loss": 0.5379,
+      "step": 8554
+    },
+    {
+      "epoch": 0.6947941200357346,
+      "grad_norm": 3.3037282568689204,
+      "learning_rate": 1.1252204165406753e-06,
+      "loss": 0.5215,
+      "step": 8555
+    },
+    {
+      "epoch": 0.6948753350117761,
+      "grad_norm": 4.960611865059338,
+      "learning_rate": 1.1246712031356486e-06,
+      "loss": 0.5758,
+      "step": 8556
+    },
+    {
+      "epoch": 0.6949565499878178,
+      "grad_norm": 7.319288707757544,
+      "learning_rate": 1.1241220848960952e-06,
+      "loss": 0.4719,
+      "step": 8557
+    },
+    {
+      "epoch": 0.6950377649638594,
+      "grad_norm": 3.819399015745177,
+      "learning_rate": 1.1235730618600126e-06,
+      "loss": 0.4936,
+      "step": 8558
+    },
+    {
+      "epoch": 0.6951189799399009,
+      "grad_norm": 3.538376081354946,
+      "learning_rate": 1.1230241340653888e-06,
+      "loss": 0.5413,
+      "step": 8559
+    },
+    {
+      "epoch": 0.6952001949159425,
+      "grad_norm": 6.249841085120328,
+      "learning_rate": 1.122475301550208e-06,
+      "loss": 0.5515,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6952814098919841,
+      "grad_norm": 8.110595035089235,
+      "learning_rate": 1.121926564352446e-06,
+      "loss": 0.6036,
+      "step": 8561
+    },
+    {
+      "epoch": 0.6953626248680257,
+      "grad_norm": 4.635874973816129,
+      "learning_rate": 1.1213779225100715e-06,
+      "loss": 0.7324,
+      "step": 8562
+    },
+    {
+      "epoch": 0.6954438398440672,
+      "grad_norm": 6.322869921153947,
+      "learning_rate": 1.1208293760610486e-06,
+      "loss": 0.52,
+      "step": 8563
+    },
+    {
+      "epoch": 0.6955250548201088,
+      "grad_norm": 7.6067666576253545,
+      "learning_rate": 1.1202809250433345e-06,
+      "loss": 0.3976,
+      "step": 8564
+    },
+    {
+      "epoch": 0.6956062697961504,
+      "grad_norm": 4.351083583151557,
+      "learning_rate": 1.1197325694948774e-06,
+      "loss": 0.5236,
+      "step": 8565
+    },
+    {
+      "epoch": 0.695687484772192,
+      "grad_norm": 5.090734435411529,
+      "learning_rate": 1.1191843094536225e-06,
+      "loss": 0.4614,
+      "step": 8566
+    },
+    {
+      "epoch": 0.6957686997482335,
+      "grad_norm": 6.001665213669127,
+      "learning_rate": 1.1186361449575055e-06,
+      "loss": 0.3699,
+      "step": 8567
+    },
+    {
+      "epoch": 0.6958499147242752,
+      "grad_norm": 6.350310371588752,
+      "learning_rate": 1.1180880760444558e-06,
+      "loss": 0.6122,
+      "step": 8568
+    },
+    {
+      "epoch": 0.6959311297003168,
+      "grad_norm": 4.918916151218767,
+      "learning_rate": 1.117540102752398e-06,
+      "loss": 0.4498,
+      "step": 8569
+    },
+    {
+      "epoch": 0.6960123446763583,
+      "grad_norm": 4.2337365050726286,
+      "learning_rate": 1.116992225119248e-06,
+      "loss": 0.3585,
+      "step": 8570
+    },
+    {
+      "epoch": 0.6960935596523999,
+      "grad_norm": 4.164412182555689,
+      "learning_rate": 1.1164444431829163e-06,
+      "loss": 0.5052,
+      "step": 8571
+    },
+    {
+      "epoch": 0.6961747746284415,
+      "grad_norm": 4.915037193176119,
+      "learning_rate": 1.1158967569813079e-06,
+      "loss": 0.4742,
+      "step": 8572
+    },
+    {
+      "epoch": 0.6962559896044831,
+      "grad_norm": 7.126882606950106,
+      "learning_rate": 1.1153491665523186e-06,
+      "loss": 0.5764,
+      "step": 8573
+    },
+    {
+      "epoch": 0.6963372045805246,
+      "grad_norm": 6.194568824850657,
+      "learning_rate": 1.1148016719338387e-06,
+      "loss": 0.3955,
+      "step": 8574
+    },
+    {
+      "epoch": 0.6964184195565662,
+      "grad_norm": 4.582251362036855,
+      "learning_rate": 1.1142542731637513e-06,
+      "loss": 0.2946,
+      "step": 8575
+    },
+    {
+      "epoch": 0.6964996345326078,
+      "grad_norm": 4.565291424432009,
+      "learning_rate": 1.1137069702799341e-06,
+      "loss": 0.5785,
+      "step": 8576
+    },
+    {
+      "epoch": 0.6965808495086494,
+      "grad_norm": 6.268014709041092,
+      "learning_rate": 1.1131597633202587e-06,
+      "loss": 0.5502,
+      "step": 8577
+    },
+    {
+      "epoch": 0.6966620644846909,
+      "grad_norm": 4.232016340106684,
+      "learning_rate": 1.1126126523225869e-06,
+      "loss": 0.4602,
+      "step": 8578
+    },
+    {
+      "epoch": 0.6967432794607326,
+      "grad_norm": 5.324130939463538,
+      "learning_rate": 1.112065637324778e-06,
+      "loss": 0.4341,
+      "step": 8579
+    },
+    {
+      "epoch": 0.6968244944367742,
+      "grad_norm": 3.6701229318957034,
+      "learning_rate": 1.1115187183646814e-06,
+      "loss": 0.5489,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6969057094128157,
+      "grad_norm": 4.1075695508244365,
+      "learning_rate": 1.1109718954801398e-06,
+      "loss": 0.5336,
+      "step": 8581
+    },
+    {
+      "epoch": 0.6969869243888573,
+      "grad_norm": 4.14721530411858,
+      "learning_rate": 1.110425168708993e-06,
+      "loss": 0.5781,
+      "step": 8582
+    },
+    {
+      "epoch": 0.6970681393648989,
+      "grad_norm": 5.832775679972837,
+      "learning_rate": 1.1098785380890696e-06,
+      "loss": 0.427,
+      "step": 8583
+    },
+    {
+      "epoch": 0.6971493543409405,
+      "grad_norm": 3.33845605592306,
+      "learning_rate": 1.1093320036581936e-06,
+      "loss": 0.4442,
+      "step": 8584
+    },
+    {
+      "epoch": 0.697230569316982,
+      "grad_norm": 5.037481188891807,
+      "learning_rate": 1.1087855654541843e-06,
+      "loss": 0.368,
+      "step": 8585
+    },
+    {
+      "epoch": 0.6973117842930237,
+      "grad_norm": 4.499749726051026,
+      "learning_rate": 1.1082392235148509e-06,
+      "loss": 0.4573,
+      "step": 8586
+    },
+    {
+      "epoch": 0.6973929992690652,
+      "grad_norm": 4.154523490167273,
+      "learning_rate": 1.1076929778779965e-06,
+      "loss": 0.4741,
+      "step": 8587
+    },
+    {
+      "epoch": 0.6974742142451068,
+      "grad_norm": 6.620859889971063,
+      "learning_rate": 1.1071468285814201e-06,
+      "loss": 0.5117,
+      "step": 8588
+    },
+    {
+      "epoch": 0.6975554292211483,
+      "grad_norm": 5.322737847031269,
+      "learning_rate": 1.106600775662911e-06,
+      "loss": 0.428,
+      "step": 8589
+    },
+    {
+      "epoch": 0.69763664419719,
+      "grad_norm": 6.543366704108589,
+      "learning_rate": 1.1060548191602535e-06,
+      "loss": 0.5052,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6977178591732316,
+      "grad_norm": 3.978052554444944,
+      "learning_rate": 1.105508959111226e-06,
+      "loss": 0.4051,
+      "step": 8591
+    },
+    {
+      "epoch": 0.6977990741492731,
+      "grad_norm": 4.013135595680058,
+      "learning_rate": 1.1049631955535985e-06,
+      "loss": 0.5263,
+      "step": 8592
+    },
+    {
+      "epoch": 0.6978802891253147,
+      "grad_norm": 7.931439768063147,
+      "learning_rate": 1.1044175285251348e-06,
+      "loss": 0.4994,
+      "step": 8593
+    },
+    {
+      "epoch": 0.6979615041013563,
+      "grad_norm": 7.9626002926512465,
+      "learning_rate": 1.1038719580635913e-06,
+      "loss": 0.5575,
+      "step": 8594
+    },
+    {
+      "epoch": 0.6980427190773979,
+      "grad_norm": 5.399616750795122,
+      "learning_rate": 1.103326484206719e-06,
+      "loss": 0.492,
+      "step": 8595
+    },
+    {
+      "epoch": 0.6981239340534394,
+      "grad_norm": 5.8829565085948765,
+      "learning_rate": 1.1027811069922634e-06,
+      "loss": 0.5324,
+      "step": 8596
+    },
+    {
+      "epoch": 0.698205149029481,
+      "grad_norm": 4.044823741608967,
+      "learning_rate": 1.1022358264579593e-06,
+      "loss": 0.4975,
+      "step": 8597
+    },
+    {
+      "epoch": 0.6982863640055226,
+      "grad_norm": 5.039936723673592,
+      "learning_rate": 1.1016906426415397e-06,
+      "loss": 0.6072,
+      "step": 8598
+    },
+    {
+      "epoch": 0.6983675789815642,
+      "grad_norm": 12.620558241769995,
+      "learning_rate": 1.1011455555807272e-06,
+      "loss": 0.4333,
+      "step": 8599
+    },
+    {
+      "epoch": 0.6984487939576057,
+      "grad_norm": 4.360858885346072,
+      "learning_rate": 1.1006005653132376e-06,
+      "loss": 0.4666,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6985300089336474,
+      "grad_norm": 4.573558742038945,
+      "learning_rate": 1.100055671876784e-06,
+      "loss": 0.5247,
+      "step": 8601
+    },
+    {
+      "epoch": 0.698611223909689,
+      "grad_norm": 5.213935252357303,
+      "learning_rate": 1.0995108753090677e-06,
+      "loss": 0.4746,
+      "step": 8602
+    },
+    {
+      "epoch": 0.6986924388857305,
+      "grad_norm": 5.309900569340087,
+      "learning_rate": 1.0989661756477869e-06,
+      "loss": 0.4982,
+      "step": 8603
+    },
+    {
+      "epoch": 0.6987736538617721,
+      "grad_norm": 6.957485365514494,
+      "learning_rate": 1.0984215729306328e-06,
+      "loss": 0.6216,
+      "step": 8604
+    },
+    {
+      "epoch": 0.6988548688378137,
+      "grad_norm": 4.191127343883717,
+      "learning_rate": 1.097877067195288e-06,
+      "loss": 0.5468,
+      "step": 8605
+    },
+    {
+      "epoch": 0.6989360838138553,
+      "grad_norm": 4.214027815939813,
+      "learning_rate": 1.0973326584794286e-06,
+      "loss": 0.5753,
+      "step": 8606
+    },
+    {
+      "epoch": 0.6990172987898968,
+      "grad_norm": 5.627295117665019,
+      "learning_rate": 1.0967883468207265e-06,
+      "loss": 0.3873,
+      "step": 8607
+    },
+    {
+      "epoch": 0.6990985137659385,
+      "grad_norm": 5.494066620160367,
+      "learning_rate": 1.0962441322568437e-06,
+      "loss": 0.4377,
+      "step": 8608
+    },
+    {
+      "epoch": 0.69917972874198,
+      "grad_norm": 4.933052728810609,
+      "learning_rate": 1.0957000148254387e-06,
+      "loss": 0.4346,
+      "step": 8609
+    },
+    {
+      "epoch": 0.6992609437180216,
+      "grad_norm": 4.67963560490583,
+      "learning_rate": 1.0951559945641592e-06,
+      "loss": 0.4196,
+      "step": 8610
+    },
+    {
+      "epoch": 0.6993421586940631,
+      "grad_norm": 4.555476078942722,
+      "learning_rate": 1.094612071510651e-06,
+      "loss": 0.5331,
+      "step": 8611
+    },
+    {
+      "epoch": 0.6994233736701048,
+      "grad_norm": 4.542359966546809,
+      "learning_rate": 1.0940682457025498e-06,
+      "loss": 0.3957,
+      "step": 8612
+    },
+    {
+      "epoch": 0.6995045886461464,
+      "grad_norm": 4.4445515035232495,
+      "learning_rate": 1.0935245171774842e-06,
+      "loss": 0.6409,
+      "step": 8613
+    },
+    {
+      "epoch": 0.6995858036221879,
+      "grad_norm": 5.499967666697894,
+      "learning_rate": 1.092980885973079e-06,
+      "loss": 0.5317,
+      "step": 8614
+    },
+    {
+      "epoch": 0.6996670185982296,
+      "grad_norm": 4.090307339205739,
+      "learning_rate": 1.0924373521269492e-06,
+      "loss": 0.3992,
+      "step": 8615
+    },
+    {
+      "epoch": 0.6997482335742711,
+      "grad_norm": 5.094031142055631,
+      "learning_rate": 1.091893915676705e-06,
+      "loss": 0.8069,
+      "step": 8616
+    },
+    {
+      "epoch": 0.6998294485503127,
+      "grad_norm": 5.102016167710116,
+      "learning_rate": 1.0913505766599506e-06,
+      "loss": 0.5136,
+      "step": 8617
+    },
+    {
+      "epoch": 0.6999106635263542,
+      "grad_norm": 6.840823358662633,
+      "learning_rate": 1.090807335114281e-06,
+      "loss": 0.4567,
+      "step": 8618
+    },
+    {
+      "epoch": 0.6999918785023959,
+      "grad_norm": 6.604394180582011,
+      "learning_rate": 1.0902641910772852e-06,
+      "loss": 0.411,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7000730934784374,
+      "grad_norm": 5.660730236893839,
+      "learning_rate": 1.0897211445865472e-06,
+      "loss": 0.432,
+      "step": 8620
+    },
+    {
+      "epoch": 0.700154308454479,
+      "grad_norm": 4.702707757036113,
+      "learning_rate": 1.089178195679641e-06,
+      "loss": 0.3665,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7002355234305205,
+      "grad_norm": 4.96553057734693,
+      "learning_rate": 1.0886353443941373e-06,
+      "loss": 0.4387,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7003167384065622,
+      "grad_norm": 8.264862786380114,
+      "learning_rate": 1.088092590767599e-06,
+      "loss": 0.3632,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7003979533826038,
+      "grad_norm": 6.893091911108661,
+      "learning_rate": 1.0875499348375812e-06,
+      "loss": 0.5675,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7004791683586453,
+      "grad_norm": 4.387426391179371,
+      "learning_rate": 1.0870073766416315e-06,
+      "loss": 0.4464,
+      "step": 8625
+    },
+    {
+      "epoch": 0.700560383334687,
+      "grad_norm": 5.027871465737512,
+      "learning_rate": 1.0864649162172941e-06,
+      "loss": 0.6075,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7006415983107285,
+      "grad_norm": 6.143259232572587,
+      "learning_rate": 1.0859225536021034e-06,
+      "loss": 0.5228,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7007228132867701,
+      "grad_norm": 6.464771014837394,
+      "learning_rate": 1.0853802888335874e-06,
+      "loss": 0.6185,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7008040282628116,
+      "grad_norm": 4.205829517254945,
+      "learning_rate": 1.0848381219492684e-06,
+      "loss": 0.498,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7008852432388533,
+      "grad_norm": 3.9072366994873855,
+      "learning_rate": 1.0842960529866627e-06,
+      "loss": 0.4887,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7009664582148948,
+      "grad_norm": 5.509854679596712,
+      "learning_rate": 1.0837540819832779e-06,
+      "loss": 0.4346,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7010476731909364,
+      "grad_norm": 4.3762550261485,
+      "learning_rate": 1.0832122089766143e-06,
+      "loss": 0.4646,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7011288881669779,
+      "grad_norm": 4.21313805277909,
+      "learning_rate": 1.082670434004168e-06,
+      "loss": 0.605,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7012101031430196,
+      "grad_norm": 5.0290295451057245,
+      "learning_rate": 1.0821287571034261e-06,
+      "loss": 0.5164,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7012913181190612,
+      "grad_norm": 3.852504953053529,
+      "learning_rate": 1.0815871783118701e-06,
+      "loss": 0.5563,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7013725330951027,
+      "grad_norm": 4.4031006382832265,
+      "learning_rate": 1.0810456976669753e-06,
+      "loss": 0.4816,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7014537480711444,
+      "grad_norm": 5.020999898610326,
+      "learning_rate": 1.0805043152062086e-06,
+      "loss": 0.5762,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7015349630471859,
+      "grad_norm": 12.438678796300241,
+      "learning_rate": 1.07996303096703e-06,
+      "loss": 0.4189,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7016161780232275,
+      "grad_norm": 3.8964414970182455,
+      "learning_rate": 1.0794218449868948e-06,
+      "loss": 0.4803,
+      "step": 8639
+    },
+    {
+      "epoch": 0.701697392999269,
+      "grad_norm": 6.66037981965594,
+      "learning_rate": 1.07888075730325e-06,
+      "loss": 0.4304,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7017786079753107,
+      "grad_norm": 9.601550182923392,
+      "learning_rate": 1.0783397679535343e-06,
+      "loss": 0.5567,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7018598229513522,
+      "grad_norm": 5.837088176132076,
+      "learning_rate": 1.077798876975183e-06,
+      "loss": 0.3856,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7019410379273938,
+      "grad_norm": 4.127084436289391,
+      "learning_rate": 1.0772580844056232e-06,
+      "loss": 0.5247,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7020222529034353,
+      "grad_norm": 3.613384254118512,
+      "learning_rate": 1.0767173902822733e-06,
+      "loss": 0.3272,
+      "step": 8644
+    },
+    {
+      "epoch": 0.702103467879477,
+      "grad_norm": 2.5823659536071637,
+      "learning_rate": 1.0761767946425482e-06,
+      "loss": 0.5089,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7021846828555186,
+      "grad_norm": 4.879797467260154,
+      "learning_rate": 1.0756362975238539e-06,
+      "loss": 0.8247,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7022658978315601,
+      "grad_norm": 6.519914984541137,
+      "learning_rate": 1.0750958989635879e-06,
+      "loss": 0.4224,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7023471128076018,
+      "grad_norm": 3.9501597340899903,
+      "learning_rate": 1.074555598999145e-06,
+      "loss": 0.571,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7024283277836433,
+      "grad_norm": 6.770723225323952,
+      "learning_rate": 1.0740153976679114e-06,
+      "loss": 0.4328,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7025095427596849,
+      "grad_norm": 13.76466570186757,
+      "learning_rate": 1.073475295007265e-06,
+      "loss": 0.424,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7025907577357264,
+      "grad_norm": 5.062420028751364,
+      "learning_rate": 1.0729352910545779e-06,
+      "loss": 0.5359,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7026719727117681,
+      "grad_norm": 8.4588766125872,
+      "learning_rate": 1.0723953858472167e-06,
+      "loss": 0.4192,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7027531876878096,
+      "grad_norm": 4.594691832051315,
+      "learning_rate": 1.0718555794225385e-06,
+      "loss": 0.4356,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7028344026638512,
+      "grad_norm": 5.910635422121678,
+      "learning_rate": 1.071315871817896e-06,
+      "loss": 0.4946,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7029156176398927,
+      "grad_norm": 7.200371807778343,
+      "learning_rate": 1.0707762630706345e-06,
+      "loss": 0.6541,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7029968326159344,
+      "grad_norm": 4.327245345198459,
+      "learning_rate": 1.0702367532180919e-06,
+      "loss": 0.5585,
+      "step": 8656
+    },
+    {
+      "epoch": 0.703078047591976,
+      "grad_norm": 5.70898727396328,
+      "learning_rate": 1.0696973422975978e-06,
+      "loss": 0.4336,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7031592625680175,
+      "grad_norm": 3.3122955544251,
+      "learning_rate": 1.0691580303464791e-06,
+      "loss": 0.7113,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7032404775440592,
+      "grad_norm": 6.853811044628905,
+      "learning_rate": 1.068618817402052e-06,
+      "loss": 0.5942,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7033216925201007,
+      "grad_norm": 6.076640488626101,
+      "learning_rate": 1.0680797035016264e-06,
+      "loss": 0.657,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7034029074961423,
+      "grad_norm": 6.2346543123925136,
+      "learning_rate": 1.0675406886825065e-06,
+      "loss": 0.5554,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7034841224721838,
+      "grad_norm": 4.087802951829346,
+      "learning_rate": 1.0670017729819911e-06,
+      "loss": 0.6035,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7035653374482255,
+      "grad_norm": 4.8340326949666,
+      "learning_rate": 1.066462956437369e-06,
+      "loss": 0.4762,
+      "step": 8663
+    },
+    {
+      "epoch": 0.703646552424267,
+      "grad_norm": 5.486314431686029,
+      "learning_rate": 1.0659242390859224e-06,
+      "loss": 0.474,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7037277674003086,
+      "grad_norm": 5.968816148722602,
+      "learning_rate": 1.0653856209649297e-06,
+      "loss": 0.4181,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7038089823763501,
+      "grad_norm": 4.467813005237297,
+      "learning_rate": 1.0648471021116584e-06,
+      "loss": 0.5303,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7038901973523918,
+      "grad_norm": 3.4998502366425757,
+      "learning_rate": 1.0643086825633723e-06,
+      "loss": 0.5941,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7039714123284334,
+      "grad_norm": 3.138792298225771,
+      "learning_rate": 1.0637703623573278e-06,
+      "loss": 0.6115,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7040526273044749,
+      "grad_norm": 4.420119383519892,
+      "learning_rate": 1.0632321415307734e-06,
+      "loss": 0.4483,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7041338422805166,
+      "grad_norm": 5.331801055403152,
+      "learning_rate": 1.0626940201209497e-06,
+      "loss": 0.3817,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7042150572565581,
+      "grad_norm": 5.750793644056218,
+      "learning_rate": 1.062155998165094e-06,
+      "loss": 0.5817,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7042962722325997,
+      "grad_norm": 3.4756270753695104,
+      "learning_rate": 1.0616180757004333e-06,
+      "loss": 0.6701,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7043774872086412,
+      "grad_norm": 6.270791519856763,
+      "learning_rate": 1.0610802527641883e-06,
+      "loss": 0.4016,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7044587021846829,
+      "grad_norm": 5.4889815502837775,
+      "learning_rate": 1.0605425293935748e-06,
+      "loss": 0.431,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7045399171607244,
+      "grad_norm": 4.255319581514907,
+      "learning_rate": 1.0600049056258008e-06,
+      "loss": 0.623,
+      "step": 8675
+    },
+    {
+      "epoch": 0.704621132136766,
+      "grad_norm": 4.619912538008321,
+      "learning_rate": 1.0594673814980652e-06,
+      "loss": 0.4224,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7047023471128075,
+      "grad_norm": 4.4624099848947445,
+      "learning_rate": 1.058929957047564e-06,
+      "loss": 0.4773,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7047835620888492,
+      "grad_norm": 3.568435548062833,
+      "learning_rate": 1.0583926323114829e-06,
+      "loss": 0.5004,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7048647770648908,
+      "grad_norm": 5.199277112036462,
+      "learning_rate": 1.057855407327001e-06,
+      "loss": 0.5205,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7049459920409323,
+      "grad_norm": 4.806067495687931,
+      "learning_rate": 1.0573182821312927e-06,
+      "loss": 0.4348,
+      "step": 8680
+    },
+    {
+      "epoch": 0.705027207016974,
+      "grad_norm": 4.581743151795791,
+      "learning_rate": 1.056781256761525e-06,
+      "loss": 0.5282,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7051084219930155,
+      "grad_norm": 4.187102501676517,
+      "learning_rate": 1.0562443312548558e-06,
+      "loss": 0.3811,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7051896369690571,
+      "grad_norm": 10.94900482763677,
+      "learning_rate": 1.0557075056484373e-06,
+      "loss": 0.4645,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7052708519450986,
+      "grad_norm": 3.286189799041891,
+      "learning_rate": 1.0551707799794164e-06,
+      "loss": 0.515,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7053520669211403,
+      "grad_norm": 5.1687388035544535,
+      "learning_rate": 1.054634154284931e-06,
+      "loss": 0.3609,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7054332818971818,
+      "grad_norm": 6.765257196850052,
+      "learning_rate": 1.0540976286021115e-06,
+      "loss": 0.5426,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7055144968732234,
+      "grad_norm": 5.1893140007432885,
+      "learning_rate": 1.053561202968084e-06,
+      "loss": 0.4086,
+      "step": 8687
+    },
+    {
+      "epoch": 0.705595711849265,
+      "grad_norm": 4.075572539458918,
+      "learning_rate": 1.053024877419967e-06,
+      "loss": 0.4706,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7056769268253066,
+      "grad_norm": 5.198722303974706,
+      "learning_rate": 1.0524886519948693e-06,
+      "loss": 0.5246,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7057581418013482,
+      "grad_norm": 5.742252264604989,
+      "learning_rate": 1.0519525267298972e-06,
+      "loss": 0.3365,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7058393567773897,
+      "grad_norm": 4.08553333166632,
+      "learning_rate": 1.0514165016621464e-06,
+      "loss": 0.6085,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7059205717534314,
+      "grad_norm": 5.7021732693041995,
+      "learning_rate": 1.0508805768287061e-06,
+      "loss": 0.5561,
+      "step": 8692
+    },
+    {
+      "epoch": 0.7060017867294729,
+      "grad_norm": 11.500144742392935,
+      "learning_rate": 1.050344752266661e-06,
+      "loss": 0.4041,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7060830017055145,
+      "grad_norm": 5.728188208611933,
+      "learning_rate": 1.0498090280130873e-06,
+      "loss": 0.5148,
+      "step": 8694
+    },
+    {
+      "epoch": 0.706164216681556,
+      "grad_norm": 3.8194604173660442,
+      "learning_rate": 1.0492734041050532e-06,
+      "loss": 0.5706,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7062454316575977,
+      "grad_norm": 4.238106988426028,
+      "learning_rate": 1.0487378805796225e-06,
+      "loss": 0.4247,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7063266466336392,
+      "grad_norm": 4.981104011286187,
+      "learning_rate": 1.0482024574738498e-06,
+      "loss": 0.4802,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7064078616096808,
+      "grad_norm": 5.353614712971508,
+      "learning_rate": 1.0476671348247834e-06,
+      "loss": 0.3993,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7064890765857224,
+      "grad_norm": 5.533743790772518,
+      "learning_rate": 1.047131912669464e-06,
+      "loss": 0.391,
+      "step": 8699
+    },
+    {
+      "epoch": 0.706570291561764,
+      "grad_norm": 5.99441767333963,
+      "learning_rate": 1.0465967910449274e-06,
+      "loss": 0.4343,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7066515065378056,
+      "grad_norm": 10.472893307184739,
+      "learning_rate": 1.046061769988201e-06,
+      "loss": 0.3472,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7067327215138471,
+      "grad_norm": 6.614917885036994,
+      "learning_rate": 1.045526849536305e-06,
+      "loss": 0.3806,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7068139364898888,
+      "grad_norm": 6.161723653557723,
+      "learning_rate": 1.0449920297262542e-06,
+      "loss": 0.5681,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7068951514659303,
+      "grad_norm": 4.341342853033297,
+      "learning_rate": 1.0444573105950543e-06,
+      "loss": 0.4967,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7069763664419719,
+      "grad_norm": 3.964088413604789,
+      "learning_rate": 1.0439226921797042e-06,
+      "loss": 0.5641,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7070575814180134,
+      "grad_norm": 7.436857002923769,
+      "learning_rate": 1.0433881745171976e-06,
+      "loss": 0.3931,
+      "step": 8706
+    },
+    {
+      "epoch": 0.7071387963940551,
+      "grad_norm": 21.38007720377836,
+      "learning_rate": 1.042853757644521e-06,
+      "loss": 0.5358,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7072200113700966,
+      "grad_norm": 9.571044595510465,
+      "learning_rate": 1.0423194415986518e-06,
+      "loss": 0.394,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7073012263461382,
+      "grad_norm": 3.4522432755903045,
+      "learning_rate": 1.0417852264165637e-06,
+      "loss": 0.7711,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7073824413221798,
+      "grad_norm": 4.437337860525526,
+      "learning_rate": 1.0412511121352201e-06,
+      "loss": 0.5022,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7074636562982214,
+      "grad_norm": 4.832102307573993,
+      "learning_rate": 1.0407170987915786e-06,
+      "loss": 0.3741,
+      "step": 8711
+    },
+    {
+      "epoch": 0.707544871274263,
+      "grad_norm": 4.424143419021889,
+      "learning_rate": 1.0401831864225915e-06,
+      "loss": 0.4477,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7076260862503045,
+      "grad_norm": 4.167265583846372,
+      "learning_rate": 1.0396493750652008e-06,
+      "loss": 0.6794,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7077073012263462,
+      "grad_norm": 5.174243065313732,
+      "learning_rate": 1.039115664756345e-06,
+      "loss": 0.4642,
+      "step": 8714
+    },
+    {
+      "epoch": 0.7077885162023877,
+      "grad_norm": 4.436027938464365,
+      "learning_rate": 1.0385820555329543e-06,
+      "loss": 0.5378,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7078697311784293,
+      "grad_norm": 6.048625139053751,
+      "learning_rate": 1.0380485474319507e-06,
+      "loss": 0.4807,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7079509461544709,
+      "grad_norm": 4.409054430942545,
+      "learning_rate": 1.0375151404902507e-06,
+      "loss": 0.4596,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7080321611305125,
+      "grad_norm": 6.231874069745442,
+      "learning_rate": 1.0369818347447617e-06,
+      "loss": 0.396,
+      "step": 8718
+    },
+    {
+      "epoch": 0.708113376106554,
+      "grad_norm": 6.571192019178439,
+      "learning_rate": 1.0364486302323868e-06,
+      "loss": 0.6371,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7081945910825956,
+      "grad_norm": 6.2588184473383555,
+      "learning_rate": 1.035915526990022e-06,
+      "loss": 0.4857,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7082758060586372,
+      "grad_norm": 4.086683961249212,
+      "learning_rate": 1.0353825250545533e-06,
+      "loss": 0.4409,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7083570210346788,
+      "grad_norm": 4.896811289532637,
+      "learning_rate": 1.0348496244628633e-06,
+      "loss": 0.4143,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7084382360107204,
+      "grad_norm": 6.411254993382916,
+      "learning_rate": 1.0343168252518252e-06,
+      "loss": 0.4197,
+      "step": 8723
+    },
+    {
+      "epoch": 0.708519450986762,
+      "grad_norm": 6.5354079259733915,
+      "learning_rate": 1.0337841274583046e-06,
+      "loss": 0.6165,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7086006659628036,
+      "grad_norm": 5.167176814080369,
+      "learning_rate": 1.0332515311191627e-06,
+      "loss": 0.5354,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7086818809388451,
+      "grad_norm": 7.4442577463893835,
+      "learning_rate": 1.032719036271253e-06,
+      "loss": 0.5054,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7087630959148867,
+      "grad_norm": 4.039848430781059,
+      "learning_rate": 1.0321866429514199e-06,
+      "loss": 0.5241,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7088443108909283,
+      "grad_norm": 3.2336915820436736,
+      "learning_rate": 1.0316543511965035e-06,
+      "loss": 0.5754,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7089255258669699,
+      "grad_norm": 4.7161410228484195,
+      "learning_rate": 1.031122161043335e-06,
+      "loss": 0.377,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7090067408430114,
+      "grad_norm": 3.6485307474977624,
+      "learning_rate": 1.030590072528738e-06,
+      "loss": 0.4437,
+      "step": 8730
+    },
+    {
+      "epoch": 0.709087955819053,
+      "grad_norm": 4.868323212713634,
+      "learning_rate": 1.030058085689532e-06,
+      "loss": 0.4561,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7091691707950946,
+      "grad_norm": 6.225865954562011,
+      "learning_rate": 1.0295262005625262e-06,
+      "loss": 0.413,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7092503857711362,
+      "grad_norm": 4.393355867203619,
+      "learning_rate": 1.028994417184525e-06,
+      "loss": 0.5137,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7093316007471778,
+      "grad_norm": 4.03858492787502,
+      "learning_rate": 1.0284627355923257e-06,
+      "loss": 0.5499,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7094128157232193,
+      "grad_norm": 7.2081095740889225,
+      "learning_rate": 1.0279311558227174e-06,
+      "loss": 0.4145,
+      "step": 8735
+    },
+    {
+      "epoch": 0.709494030699261,
+      "grad_norm": 4.52684696449661,
+      "learning_rate": 1.027399677912482e-06,
+      "loss": 0.4802,
+      "step": 8736
+    },
+    {
+      "epoch": 0.7095752456753025,
+      "grad_norm": 4.627236682606212,
+      "learning_rate": 1.0268683018983944e-06,
+      "loss": 0.4124,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7096564606513441,
+      "grad_norm": 4.60387331373005,
+      "learning_rate": 1.026337027817224e-06,
+      "loss": 0.4463,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7097376756273857,
+      "grad_norm": 4.049679102453057,
+      "learning_rate": 1.0258058557057328e-06,
+      "loss": 0.4391,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7098188906034273,
+      "grad_norm": 3.932889064119171,
+      "learning_rate": 1.0252747856006735e-06,
+      "loss": 0.4598,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7099001055794688,
+      "grad_norm": 5.445177988596398,
+      "learning_rate": 1.0247438175387946e-06,
+      "loss": 0.3975,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7099813205555104,
+      "grad_norm": 4.336033272668492,
+      "learning_rate": 1.0242129515568364e-06,
+      "loss": 0.6375,
+      "step": 8742
+    },
+    {
+      "epoch": 0.710062535531552,
+      "grad_norm": 5.942395150551147,
+      "learning_rate": 1.0236821876915303e-06,
+      "loss": 0.5762,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7101437505075936,
+      "grad_norm": 7.955186448658773,
+      "learning_rate": 1.0231515259796046e-06,
+      "loss": 0.4297,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7102249654836352,
+      "grad_norm": 4.589645135923486,
+      "learning_rate": 1.022620966457776e-06,
+      "loss": 0.5287,
+      "step": 8745
+    },
+    {
+      "epoch": 0.7103061804596768,
+      "grad_norm": 10.369934951756198,
+      "learning_rate": 1.0220905091627581e-06,
+      "loss": 0.3939,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7103873954357184,
+      "grad_norm": 4.830026444654436,
+      "learning_rate": 1.0215601541312556e-06,
+      "loss": 0.4922,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7104686104117599,
+      "grad_norm": 4.89390413686288,
+      "learning_rate": 1.0210299013999662e-06,
+      "loss": 0.4854,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7105498253878015,
+      "grad_norm": 5.6081171633262015,
+      "learning_rate": 1.0204997510055793e-06,
+      "loss": 0.4383,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7106310403638431,
+      "grad_norm": 7.598002540248723,
+      "learning_rate": 1.0199697029847804e-06,
+      "loss": 0.5754,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7107122553398847,
+      "grad_norm": 6.496257353872819,
+      "learning_rate": 1.0194397573742442e-06,
+      "loss": 0.3061,
+      "step": 8751
+    },
+    {
+      "epoch": 0.7107934703159262,
+      "grad_norm": 7.92240606328125,
+      "learning_rate": 1.0189099142106421e-06,
+      "loss": 0.4262,
+      "step": 8752
+    },
+    {
+      "epoch": 0.7108746852919678,
+      "grad_norm": 27.70458593561067,
+      "learning_rate": 1.0183801735306342e-06,
+      "loss": 0.4334,
+      "step": 8753
+    },
+    {
+      "epoch": 0.7109559002680094,
+      "grad_norm": 4.449530450041586,
+      "learning_rate": 1.0178505353708779e-06,
+      "loss": 0.5682,
+      "step": 8754
+    },
+    {
+      "epoch": 0.711037115244051,
+      "grad_norm": 5.978122767767535,
+      "learning_rate": 1.0173209997680203e-06,
+      "loss": 0.5643,
+      "step": 8755
+    },
+    {
+      "epoch": 0.7111183302200926,
+      "grad_norm": 4.850468611110746,
+      "learning_rate": 1.0167915667587019e-06,
+      "loss": 0.47,
+      "step": 8756
+    },
+    {
+      "epoch": 0.7111995451961342,
+      "grad_norm": 4.3524338310305835,
+      "learning_rate": 1.016262236379558e-06,
+      "loss": 0.5099,
+      "step": 8757
+    },
+    {
+      "epoch": 0.7112807601721758,
+      "grad_norm": 6.685288002285286,
+      "learning_rate": 1.015733008667214e-06,
+      "loss": 0.6591,
+      "step": 8758
+    },
+    {
+      "epoch": 0.7113619751482173,
+      "grad_norm": 7.509285220548554,
+      "learning_rate": 1.0152038836582903e-06,
+      "loss": 0.5526,
+      "step": 8759
+    },
+    {
+      "epoch": 0.7114431901242589,
+      "grad_norm": 3.6292963710683366,
+      "learning_rate": 1.0146748613894005e-06,
+      "loss": 0.4786,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7115244051003005,
+      "grad_norm": 4.730030440457719,
+      "learning_rate": 1.0141459418971496e-06,
+      "loss": 0.409,
+      "step": 8761
+    },
+    {
+      "epoch": 0.7116056200763421,
+      "grad_norm": 3.5510572572185977,
+      "learning_rate": 1.0136171252181348e-06,
+      "loss": 0.5859,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7116868350523836,
+      "grad_norm": 7.350261219519113,
+      "learning_rate": 1.0130884113889491e-06,
+      "loss": 0.4996,
+      "step": 8763
+    },
+    {
+      "epoch": 0.7117680500284252,
+      "grad_norm": 4.1865654744906635,
+      "learning_rate": 1.0125598004461752e-06,
+      "loss": 0.395,
+      "step": 8764
+    },
+    {
+      "epoch": 0.7118492650044668,
+      "grad_norm": 8.196706389670227,
+      "learning_rate": 1.012031292426391e-06,
+      "loss": 0.4005,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7119304799805084,
+      "grad_norm": 5.9136835481706855,
+      "learning_rate": 1.011502887366167e-06,
+      "loss": 0.4292,
+      "step": 8766
+    },
+    {
+      "epoch": 0.71201169495655,
+      "grad_norm": 5.09214562271127,
+      "learning_rate": 1.0109745853020655e-06,
+      "loss": 0.4259,
+      "step": 8767
+    },
+    {
+      "epoch": 0.7120929099325916,
+      "grad_norm": 8.352854444325088,
+      "learning_rate": 1.0104463862706414e-06,
+      "loss": 0.3907,
+      "step": 8768
+    },
+    {
+      "epoch": 0.7121741249086332,
+      "grad_norm": 4.539279498681904,
+      "learning_rate": 1.0099182903084448e-06,
+      "loss": 0.5434,
+      "step": 8769
+    },
+    {
+      "epoch": 0.7122553398846747,
+      "grad_norm": 6.108837746410746,
+      "learning_rate": 1.0093902974520165e-06,
+      "loss": 0.4994,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7123365548607163,
+      "grad_norm": 6.827549453211144,
+      "learning_rate": 1.0088624077378897e-06,
+      "loss": 0.4736,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7124177698367579,
+      "grad_norm": 5.0626039477295315,
+      "learning_rate": 1.0083346212025923e-06,
+      "loss": 0.6442,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7124989848127995,
+      "grad_norm": 4.104258023545044,
+      "learning_rate": 1.0078069378826458e-06,
+      "loss": 0.579,
+      "step": 8773
+    },
+    {
+      "epoch": 0.712580199788841,
+      "grad_norm": 3.7706839195991297,
+      "learning_rate": 1.0072793578145618e-06,
+      "loss": 0.5327,
+      "step": 8774
+    },
+    {
+      "epoch": 0.7126614147648827,
+      "grad_norm": 7.155111652149057,
+      "learning_rate": 1.0067518810348453e-06,
+      "loss": 0.5282,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7127426297409242,
+      "grad_norm": 7.637395079270046,
+      "learning_rate": 1.0062245075799966e-06,
+      "loss": 0.5045,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7128238447169658,
+      "grad_norm": 6.413393628245962,
+      "learning_rate": 1.0056972374865054e-06,
+      "loss": 0.6264,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7129050596930074,
+      "grad_norm": 5.7064526408846135,
+      "learning_rate": 1.0051700707908569e-06,
+      "loss": 0.7116,
+      "step": 8778
+    },
+    {
+      "epoch": 0.712986274669049,
+      "grad_norm": 5.902226546762958,
+      "learning_rate": 1.0046430075295287e-06,
+      "loss": 0.6247,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7130674896450906,
+      "grad_norm": 5.831122076718099,
+      "learning_rate": 1.0041160477389909e-06,
+      "loss": 0.4892,
+      "step": 8780
+    },
+    {
+      "epoch": 0.7131487046211321,
+      "grad_norm": 5.215468338547022,
+      "learning_rate": 1.0035891914557044e-06,
+      "loss": 0.5625,
+      "step": 8781
+    },
+    {
+      "epoch": 0.7132299195971737,
+      "grad_norm": 7.447326637422664,
+      "learning_rate": 1.0030624387161273e-06,
+      "loss": 0.4898,
+      "step": 8782
+    },
+    {
+      "epoch": 0.7133111345732153,
+      "grad_norm": 4.541675365615177,
+      "learning_rate": 1.002535789556707e-06,
+      "loss": 0.5263,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7133923495492569,
+      "grad_norm": 5.008687031433551,
+      "learning_rate": 1.0020092440138833e-06,
+      "loss": 0.57,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7134735645252984,
+      "grad_norm": 4.218282129108528,
+      "learning_rate": 1.0014828021240932e-06,
+      "loss": 0.3873,
+      "step": 8785
+    },
+    {
+      "epoch": 0.71355477950134,
+      "grad_norm": 7.9910976481847875,
+      "learning_rate": 1.0009564639237627e-06,
+      "loss": 0.3644,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7136359944773816,
+      "grad_norm": 4.348847062904255,
+      "learning_rate": 1.0004302294493104e-06,
+      "loss": 0.4983,
+      "step": 8787
+    },
+    {
+      "epoch": 0.7137172094534232,
+      "grad_norm": 6.4121947565646815,
+      "learning_rate": 9.999040987371505e-07,
+      "loss": 0.4464,
+      "step": 8788
+    },
+    {
+      "epoch": 0.7137984244294648,
+      "grad_norm": 4.829469885568656,
+      "learning_rate": 9.993780718236882e-07,
+      "loss": 0.3603,
+      "step": 8789
+    },
+    {
+      "epoch": 0.7138796394055064,
+      "grad_norm": 3.668444891519201,
+      "learning_rate": 9.988521487453203e-07,
+      "loss": 0.5219,
+      "step": 8790
+    },
+    {
+      "epoch": 0.713960854381548,
+      "grad_norm": 4.4085220444412245,
+      "learning_rate": 9.98326329538439e-07,
+      "loss": 0.48,
+      "step": 8791
+    },
+    {
+      "epoch": 0.7140420693575895,
+      "grad_norm": 6.942288338011259,
+      "learning_rate": 9.978006142394292e-07,
+      "loss": 0.6356,
+      "step": 8792
+    },
+    {
+      "epoch": 0.7141232843336311,
+      "grad_norm": 3.9011079567183353,
+      "learning_rate": 9.972750028846665e-07,
+      "loss": 0.4649,
+      "step": 8793
+    },
+    {
+      "epoch": 0.7142044993096727,
+      "grad_norm": 5.691178624558273,
+      "learning_rate": 9.967494955105197e-07,
+      "loss": 0.5606,
+      "step": 8794
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 4.779174640506254,
+      "learning_rate": 9.962240921533528e-07,
+      "loss": 0.4898,
+      "step": 8795
+    },
+    {
+      "epoch": 0.7143669292617558,
+      "grad_norm": 5.8645704612877445,
+      "learning_rate": 9.956987928495193e-07,
+      "loss": 0.4236,
+      "step": 8796
+    },
+    {
+      "epoch": 0.7144481442377975,
+      "grad_norm": 4.402062224686781,
+      "learning_rate": 9.951735976353677e-07,
+      "loss": 0.4852,
+      "step": 8797
+    },
+    {
+      "epoch": 0.714529359213839,
+      "grad_norm": 5.0026387695604875,
+      "learning_rate": 9.946485065472402e-07,
+      "loss": 0.357,
+      "step": 8798
+    },
+    {
+      "epoch": 0.7146105741898806,
+      "grad_norm": 7.498003830377879,
+      "learning_rate": 9.941235196214687e-07,
+      "loss": 0.5172,
+      "step": 8799
+    },
+    {
+      "epoch": 0.7146917891659222,
+      "grad_norm": 3.3478530645853675,
+      "learning_rate": 9.935986368943796e-07,
+      "loss": 0.4808,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7147730041419638,
+      "grad_norm": 6.779585993646702,
+      "learning_rate": 9.930738584022925e-07,
+      "loss": 0.5855,
+      "step": 8801
+    },
+    {
+      "epoch": 0.7148542191180054,
+      "grad_norm": 4.424567878048419,
+      "learning_rate": 9.925491841815197e-07,
+      "loss": 0.4775,
+      "step": 8802
+    },
+    {
+      "epoch": 0.7149354340940469,
+      "grad_norm": 4.168929836677361,
+      "learning_rate": 9.92024614268364e-07,
+      "loss": 0.611,
+      "step": 8803
+    },
+    {
+      "epoch": 0.7150166490700886,
+      "grad_norm": 4.398671369510578,
+      "learning_rate": 9.915001486991243e-07,
+      "loss": 0.6159,
+      "step": 8804
+    },
+    {
+      "epoch": 0.7150978640461301,
+      "grad_norm": 4.041777750287028,
+      "learning_rate": 9.909757875100914e-07,
+      "loss": 0.4175,
+      "step": 8805
+    },
+    {
+      "epoch": 0.7151790790221717,
+      "grad_norm": 7.847660489045761,
+      "learning_rate": 9.904515307375478e-07,
+      "loss": 0.5913,
+      "step": 8806
+    },
+    {
+      "epoch": 0.7152602939982132,
+      "grad_norm": 4.250630000320927,
+      "learning_rate": 9.899273784177681e-07,
+      "loss": 0.3363,
+      "step": 8807
+    },
+    {
+      "epoch": 0.7153415089742549,
+      "grad_norm": 8.197363212404545,
+      "learning_rate": 9.894033305870229e-07,
+      "loss": 0.4725,
+      "step": 8808
+    },
+    {
+      "epoch": 0.7154227239502964,
+      "grad_norm": 3.91482587309766,
+      "learning_rate": 9.888793872815716e-07,
+      "loss": 0.4924,
+      "step": 8809
+    },
+    {
+      "epoch": 0.715503938926338,
+      "grad_norm": 4.933671181822034,
+      "learning_rate": 9.883555485376688e-07,
+      "loss": 0.6437,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7155851539023796,
+      "grad_norm": 5.499179017080881,
+      "learning_rate": 9.878318143915633e-07,
+      "loss": 0.3706,
+      "step": 8811
+    },
+    {
+      "epoch": 0.7156663688784212,
+      "grad_norm": 7.743811775235963,
+      "learning_rate": 9.873081848794926e-07,
+      "loss": 0.4138,
+      "step": 8812
+    },
+    {
+      "epoch": 0.7157475838544628,
+      "grad_norm": 4.332868158341539,
+      "learning_rate": 9.867846600376892e-07,
+      "loss": 0.5267,
+      "step": 8813
+    },
+    {
+      "epoch": 0.7158287988305043,
+      "grad_norm": 4.072077480048703,
+      "learning_rate": 9.862612399023797e-07,
+      "loss": 0.6121,
+      "step": 8814
+    },
+    {
+      "epoch": 0.715910013806546,
+      "grad_norm": 6.275203078249558,
+      "learning_rate": 9.85737924509781e-07,
+      "loss": 0.4842,
+      "step": 8815
+    },
+    {
+      "epoch": 0.7159912287825875,
+      "grad_norm": 6.229912521186827,
+      "learning_rate": 9.852147138961026e-07,
+      "loss": 0.3927,
+      "step": 8816
+    },
+    {
+      "epoch": 0.7160724437586291,
+      "grad_norm": 5.84959994620228,
+      "learning_rate": 9.846916080975493e-07,
+      "loss": 0.4515,
+      "step": 8817
+    },
+    {
+      "epoch": 0.7161536587346706,
+      "grad_norm": 8.565273476837879,
+      "learning_rate": 9.841686071503178e-07,
+      "loss": 0.4579,
+      "step": 8818
+    },
+    {
+      "epoch": 0.7162348737107123,
+      "grad_norm": 5.7709811538417854,
+      "learning_rate": 9.836457110905956e-07,
+      "loss": 0.7484,
+      "step": 8819
+    },
+    {
+      "epoch": 0.7163160886867538,
+      "grad_norm": 5.3290551767188035,
+      "learning_rate": 9.831229199545659e-07,
+      "loss": 0.3563,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7163973036627954,
+      "grad_norm": 6.90902242367057,
+      "learning_rate": 9.82600233778402e-07,
+      "loss": 0.5667,
+      "step": 8821
+    },
+    {
+      "epoch": 0.716478518638837,
+      "grad_norm": 5.0511931716125105,
+      "learning_rate": 9.820776525982703e-07,
+      "loss": 0.4209,
+      "step": 8822
+    },
+    {
+      "epoch": 0.7165597336148786,
+      "grad_norm": 4.8369460147576095,
+      "learning_rate": 9.815551764503317e-07,
+      "loss": 0.4668,
+      "step": 8823
+    },
+    {
+      "epoch": 0.7166409485909202,
+      "grad_norm": 4.486543984473168,
+      "learning_rate": 9.810328053707394e-07,
+      "loss": 0.8039,
+      "step": 8824
+    },
+    {
+      "epoch": 0.7167221635669617,
+      "grad_norm": 4.014813431510159,
+      "learning_rate": 9.805105393956378e-07,
+      "loss": 0.471,
+      "step": 8825
+    },
+    {
+      "epoch": 0.7168033785430034,
+      "grad_norm": 7.519778869225568,
+      "learning_rate": 9.799883785611647e-07,
+      "loss": 0.4735,
+      "step": 8826
+    },
+    {
+      "epoch": 0.7168845935190449,
+      "grad_norm": 5.5537814970796315,
+      "learning_rate": 9.794663229034518e-07,
+      "loss": 0.4381,
+      "step": 8827
+    },
+    {
+      "epoch": 0.7169658084950865,
+      "grad_norm": 19.344716592490663,
+      "learning_rate": 9.78944372458622e-07,
+      "loss": 0.4718,
+      "step": 8828
+    },
+    {
+      "epoch": 0.717047023471128,
+      "grad_norm": 5.884182711703903,
+      "learning_rate": 9.784225272627908e-07,
+      "loss": 0.5988,
+      "step": 8829
+    },
+    {
+      "epoch": 0.7171282384471697,
+      "grad_norm": 4.891986811172294,
+      "learning_rate": 9.77900787352068e-07,
+      "loss": 0.3812,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7172094534232112,
+      "grad_norm": 10.276801237301445,
+      "learning_rate": 9.773791527625557e-07,
+      "loss": 0.6522,
+      "step": 8831
+    },
+    {
+      "epoch": 0.7172906683992528,
+      "grad_norm": 6.1468764794545505,
+      "learning_rate": 9.76857623530347e-07,
+      "loss": 0.5199,
+      "step": 8832
+    },
+    {
+      "epoch": 0.7173718833752945,
+      "grad_norm": 3.8035905037114097,
+      "learning_rate": 9.763361996915302e-07,
+      "loss": 0.5275,
+      "step": 8833
+    },
+    {
+      "epoch": 0.717453098351336,
+      "grad_norm": 6.897927489006048,
+      "learning_rate": 9.75814881282185e-07,
+      "loss": 0.5616,
+      "step": 8834
+    },
+    {
+      "epoch": 0.7175343133273776,
+      "grad_norm": 3.453903520221075,
+      "learning_rate": 9.752936683383822e-07,
+      "loss": 0.4339,
+      "step": 8835
+    },
+    {
+      "epoch": 0.7176155283034191,
+      "grad_norm": 4.99189702123876,
+      "learning_rate": 9.747725608961881e-07,
+      "loss": 0.5881,
+      "step": 8836
+    },
+    {
+      "epoch": 0.7176967432794608,
+      "grad_norm": 5.00035982299125,
+      "learning_rate": 9.742515589916615e-07,
+      "loss": 0.4863,
+      "step": 8837
+    },
+    {
+      "epoch": 0.7177779582555023,
+      "grad_norm": 4.906966447423704,
+      "learning_rate": 9.737306626608514e-07,
+      "loss": 0.5093,
+      "step": 8838
+    },
+    {
+      "epoch": 0.7178591732315439,
+      "grad_norm": 5.968171309664626,
+      "learning_rate": 9.732098719398025e-07,
+      "loss": 0.5323,
+      "step": 8839
+    },
+    {
+      "epoch": 0.7179403882075854,
+      "grad_norm": 4.720388953914238,
+      "learning_rate": 9.726891868645502e-07,
+      "loss": 0.4468,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7180216031836271,
+      "grad_norm": 3.842805699155025,
+      "learning_rate": 9.721686074711228e-07,
+      "loss": 0.5385,
+      "step": 8841
+    },
+    {
+      "epoch": 0.7181028181596686,
+      "grad_norm": 9.473267130229601,
+      "learning_rate": 9.716481337955411e-07,
+      "loss": 0.4456,
+      "step": 8842
+    },
+    {
+      "epoch": 0.7181840331357102,
+      "grad_norm": 5.317653998758775,
+      "learning_rate": 9.711277658738197e-07,
+      "loss": 0.6716,
+      "step": 8843
+    },
+    {
+      "epoch": 0.7182652481117519,
+      "grad_norm": 4.085741117429908,
+      "learning_rate": 9.706075037419666e-07,
+      "loss": 0.553,
+      "step": 8844
+    },
+    {
+      "epoch": 0.7183464630877934,
+      "grad_norm": 21.53419743501485,
+      "learning_rate": 9.700873474359786e-07,
+      "loss": 0.4472,
+      "step": 8845
+    },
+    {
+      "epoch": 0.718427678063835,
+      "grad_norm": 6.994061607271889,
+      "learning_rate": 9.695672969918508e-07,
+      "loss": 0.4094,
+      "step": 8846
+    },
+    {
+      "epoch": 0.7185088930398765,
+      "grad_norm": 4.554212460035874,
+      "learning_rate": 9.69047352445566e-07,
+      "loss": 0.4225,
+      "step": 8847
+    },
+    {
+      "epoch": 0.7185901080159182,
+      "grad_norm": 7.821491253661923,
+      "learning_rate": 9.68527513833101e-07,
+      "loss": 0.4877,
+      "step": 8848
+    },
+    {
+      "epoch": 0.7186713229919597,
+      "grad_norm": 3.1283429984031783,
+      "learning_rate": 9.68007781190427e-07,
+      "loss": 0.3561,
+      "step": 8849
+    },
+    {
+      "epoch": 0.7187525379680013,
+      "grad_norm": 5.628658133939152,
+      "learning_rate": 9.674881545535073e-07,
+      "loss": 0.5745,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7188337529440428,
+      "grad_norm": 11.726251049043716,
+      "learning_rate": 9.669686339582959e-07,
+      "loss": 0.6452,
+      "step": 8851
+    },
+    {
+      "epoch": 0.7189149679200845,
+      "grad_norm": 4.111507695592982,
+      "learning_rate": 9.664492194407425e-07,
+      "loss": 0.559,
+      "step": 8852
+    },
+    {
+      "epoch": 0.718996182896126,
+      "grad_norm": 6.507482959837951,
+      "learning_rate": 9.659299110367868e-07,
+      "loss": 0.4079,
+      "step": 8853
+    },
+    {
+      "epoch": 0.7190773978721676,
+      "grad_norm": 6.32505374062163,
+      "learning_rate": 9.654107087823613e-07,
+      "loss": 0.6939,
+      "step": 8854
+    },
+    {
+      "epoch": 0.7191586128482093,
+      "grad_norm": 7.74320454513088,
+      "learning_rate": 9.64891612713393e-07,
+      "loss": 0.6168,
+      "step": 8855
+    },
+    {
+      "epoch": 0.7192398278242508,
+      "grad_norm": 9.529588740791155,
+      "learning_rate": 9.643726228658017e-07,
+      "loss": 0.4821,
+      "step": 8856
+    },
+    {
+      "epoch": 0.7193210428002924,
+      "grad_norm": 5.534280027657455,
+      "learning_rate": 9.638537392754968e-07,
+      "loss": 0.4415,
+      "step": 8857
+    },
+    {
+      "epoch": 0.7194022577763339,
+      "grad_norm": 5.287392525426979,
+      "learning_rate": 9.63334961978384e-07,
+      "loss": 0.4648,
+      "step": 8858
+    },
+    {
+      "epoch": 0.7194834727523756,
+      "grad_norm": 5.411950885514694,
+      "learning_rate": 9.628162910103595e-07,
+      "loss": 0.4528,
+      "step": 8859
+    },
+    {
+      "epoch": 0.7195646877284171,
+      "grad_norm": 5.234971772019722,
+      "learning_rate": 9.62297726407312e-07,
+      "loss": 0.5357,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7196459027044587,
+      "grad_norm": 3.5786127468446147,
+      "learning_rate": 9.617792682051228e-07,
+      "loss": 0.4252,
+      "step": 8861
+    },
+    {
+      "epoch": 0.7197271176805002,
+      "grad_norm": 4.004340303264486,
+      "learning_rate": 9.612609164396672e-07,
+      "loss": 0.4677,
+      "step": 8862
+    },
+    {
+      "epoch": 0.7198083326565419,
+      "grad_norm": 5.758524558297638,
+      "learning_rate": 9.607426711468135e-07,
+      "loss": 0.4476,
+      "step": 8863
+    },
+    {
+      "epoch": 0.7198895476325834,
+      "grad_norm": 8.535570347156773,
+      "learning_rate": 9.602245323624195e-07,
+      "loss": 0.4546,
+      "step": 8864
+    },
+    {
+      "epoch": 0.719970762608625,
+      "grad_norm": 4.001330283022169,
+      "learning_rate": 9.597065001223397e-07,
+      "loss": 0.4607,
+      "step": 8865
+    },
+    {
+      "epoch": 0.7200519775846667,
+      "grad_norm": 4.377728948837983,
+      "learning_rate": 9.591885744624183e-07,
+      "loss": 0.3572,
+      "step": 8866
+    },
+    {
+      "epoch": 0.7201331925607082,
+      "grad_norm": 3.8588903316964442,
+      "learning_rate": 9.586707554184918e-07,
+      "loss": 0.4552,
+      "step": 8867
+    },
+    {
+      "epoch": 0.7202144075367498,
+      "grad_norm": 6.457542883089068,
+      "learning_rate": 9.581530430263919e-07,
+      "loss": 0.4036,
+      "step": 8868
+    },
+    {
+      "epoch": 0.7202956225127913,
+      "grad_norm": 6.505717474355659,
+      "learning_rate": 9.57635437321942e-07,
+      "loss": 0.2751,
+      "step": 8869
+    },
+    {
+      "epoch": 0.720376837488833,
+      "grad_norm": 7.655308168808172,
+      "learning_rate": 9.571179383409561e-07,
+      "loss": 0.4116,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7204580524648745,
+      "grad_norm": 3.1031621847917634,
+      "learning_rate": 9.566005461192444e-07,
+      "loss": 0.4976,
+      "step": 8871
+    },
+    {
+      "epoch": 0.7205392674409161,
+      "grad_norm": 6.088083430309052,
+      "learning_rate": 9.560832606926064e-07,
+      "loss": 0.4635,
+      "step": 8872
+    },
+    {
+      "epoch": 0.7206204824169576,
+      "grad_norm": 6.004449008237206,
+      "learning_rate": 9.55566082096835e-07,
+      "loss": 0.5112,
+      "step": 8873
+    },
+    {
+      "epoch": 0.7207016973929993,
+      "grad_norm": 9.024846945792124,
+      "learning_rate": 9.550490103677176e-07,
+      "loss": 0.3386,
+      "step": 8874
+    },
+    {
+      "epoch": 0.7207829123690408,
+      "grad_norm": 4.2699585973289915,
+      "learning_rate": 9.54532045541031e-07,
+      "loss": 0.4822,
+      "step": 8875
+    },
+    {
+      "epoch": 0.7208641273450824,
+      "grad_norm": 4.631080829221704,
+      "learning_rate": 9.54015187652548e-07,
+      "loss": 0.5532,
+      "step": 8876
+    },
+    {
+      "epoch": 0.7209453423211241,
+      "grad_norm": 6.8091971004552585,
+      "learning_rate": 9.534984367380329e-07,
+      "loss": 0.5322,
+      "step": 8877
+    },
+    {
+      "epoch": 0.7210265572971656,
+      "grad_norm": 3.589150990709112,
+      "learning_rate": 9.529817928332411e-07,
+      "loss": 0.523,
+      "step": 8878
+    },
+    {
+      "epoch": 0.7211077722732072,
+      "grad_norm": 8.421640744360598,
+      "learning_rate": 9.524652559739217e-07,
+      "loss": 0.3827,
+      "step": 8879
+    },
+    {
+      "epoch": 0.7211889872492487,
+      "grad_norm": 5.462634830370036,
+      "learning_rate": 9.519488261958157e-07,
+      "loss": 0.4339,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7212702022252904,
+      "grad_norm": 6.652542616373296,
+      "learning_rate": 9.514325035346577e-07,
+      "loss": 0.5582,
+      "step": 8881
+    },
+    {
+      "epoch": 0.7213514172013319,
+      "grad_norm": 11.329982918292163,
+      "learning_rate": 9.509162880261757e-07,
+      "loss": 0.4287,
+      "step": 8882
+    },
+    {
+      "epoch": 0.7214326321773735,
+      "grad_norm": 4.346970981846385,
+      "learning_rate": 9.504001797060875e-07,
+      "loss": 0.3969,
+      "step": 8883
+    },
+    {
+      "epoch": 0.721513847153415,
+      "grad_norm": 10.533676550458823,
+      "learning_rate": 9.498841786101065e-07,
+      "loss": 0.3896,
+      "step": 8884
+    },
+    {
+      "epoch": 0.7215950621294567,
+      "grad_norm": 6.522166801681026,
+      "learning_rate": 9.493682847739363e-07,
+      "loss": 0.4312,
+      "step": 8885
+    },
+    {
+      "epoch": 0.7216762771054982,
+      "grad_norm": 15.683352653559364,
+      "learning_rate": 9.488524982332734e-07,
+      "loss": 0.4908,
+      "step": 8886
+    },
+    {
+      "epoch": 0.7217574920815398,
+      "grad_norm": 4.885082924071923,
+      "learning_rate": 9.483368190238093e-07,
+      "loss": 0.604,
+      "step": 8887
+    },
+    {
+      "epoch": 0.7218387070575815,
+      "grad_norm": 6.215187035651405,
+      "learning_rate": 9.478212471812242e-07,
+      "loss": 0.6215,
+      "step": 8888
+    },
+    {
+      "epoch": 0.721919922033623,
+      "grad_norm": 6.199409295607022,
+      "learning_rate": 9.473057827411941e-07,
+      "loss": 0.46,
+      "step": 8889
+    },
+    {
+      "epoch": 0.7220011370096646,
+      "grad_norm": 5.210688836874202,
+      "learning_rate": 9.467904257393873e-07,
+      "loss": 0.4037,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7220823519857061,
+      "grad_norm": 5.580114042556584,
+      "learning_rate": 9.462751762114625e-07,
+      "loss": 0.5964,
+      "step": 8891
+    },
+    {
+      "epoch": 0.7221635669617478,
+      "grad_norm": 5.621455969727306,
+      "learning_rate": 9.45760034193072e-07,
+      "loss": 0.5682,
+      "step": 8892
+    },
+    {
+      "epoch": 0.7222447819377893,
+      "grad_norm": 11.91702063843931,
+      "learning_rate": 9.45244999719862e-07,
+      "loss": 0.3895,
+      "step": 8893
+    },
+    {
+      "epoch": 0.7223259969138309,
+      "grad_norm": 4.231827622581398,
+      "learning_rate": 9.447300728274689e-07,
+      "loss": 0.3848,
+      "step": 8894
+    },
+    {
+      "epoch": 0.7224072118898724,
+      "grad_norm": 4.914560874009442,
+      "learning_rate": 9.442152535515245e-07,
+      "loss": 0.4818,
+      "step": 8895
+    },
+    {
+      "epoch": 0.7224884268659141,
+      "grad_norm": 3.3941269423081386,
+      "learning_rate": 9.437005419276496e-07,
+      "loss": 0.5125,
+      "step": 8896
+    },
+    {
+      "epoch": 0.7225696418419557,
+      "grad_norm": 7.1114894868029515,
+      "learning_rate": 9.431859379914615e-07,
+      "loss": 0.5125,
+      "step": 8897
+    },
+    {
+      "epoch": 0.7226508568179972,
+      "grad_norm": 7.590278377377921,
+      "learning_rate": 9.426714417785673e-07,
+      "loss": 0.4268,
+      "step": 8898
+    },
+    {
+      "epoch": 0.7227320717940389,
+      "grad_norm": 8.118555760980742,
+      "learning_rate": 9.421570533245663e-07,
+      "loss": 0.6289,
+      "step": 8899
+    },
+    {
+      "epoch": 0.7228132867700804,
+      "grad_norm": 4.8344990060798025,
+      "learning_rate": 9.416427726650535e-07,
+      "loss": 0.4768,
+      "step": 8900
+    },
+    {
+      "epoch": 0.722894501746122,
+      "grad_norm": 3.812357829292625,
+      "learning_rate": 9.411285998356124e-07,
+      "loss": 0.588,
+      "step": 8901
+    },
+    {
+      "epoch": 0.7229757167221635,
+      "grad_norm": 4.3624783009325805,
+      "learning_rate": 9.406145348718218e-07,
+      "loss": 0.6037,
+      "step": 8902
+    },
+    {
+      "epoch": 0.7230569316982052,
+      "grad_norm": 7.735840837751049,
+      "learning_rate": 9.401005778092537e-07,
+      "loss": 0.5546,
+      "step": 8903
+    },
+    {
+      "epoch": 0.7231381466742467,
+      "grad_norm": 4.2176648353846184,
+      "learning_rate": 9.395867286834695e-07,
+      "loss": 0.4084,
+      "step": 8904
+    },
+    {
+      "epoch": 0.7232193616502883,
+      "grad_norm": 10.500833125072006,
+      "learning_rate": 9.390729875300247e-07,
+      "loss": 0.3893,
+      "step": 8905
+    },
+    {
+      "epoch": 0.7233005766263298,
+      "grad_norm": 5.323683099575961,
+      "learning_rate": 9.38559354384469e-07,
+      "loss": 0.48,
+      "step": 8906
+    },
+    {
+      "epoch": 0.7233817916023715,
+      "grad_norm": 4.623752481785489,
+      "learning_rate": 9.38045829282341e-07,
+      "loss": 0.6171,
+      "step": 8907
+    },
+    {
+      "epoch": 0.7234630065784131,
+      "grad_norm": 7.855386690391926,
+      "learning_rate": 9.375324122591753e-07,
+      "loss": 0.5059,
+      "step": 8908
+    },
+    {
+      "epoch": 0.7235442215544546,
+      "grad_norm": 4.500533084765437,
+      "learning_rate": 9.370191033504982e-07,
+      "loss": 0.5279,
+      "step": 8909
+    },
+    {
+      "epoch": 0.7236254365304963,
+      "grad_norm": 7.958925600183833,
+      "learning_rate": 9.365059025918274e-07,
+      "loss": 0.4437,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7237066515065378,
+      "grad_norm": 3.95735207574078,
+      "learning_rate": 9.359928100186724e-07,
+      "loss": 0.5215,
+      "step": 8911
+    },
+    {
+      "epoch": 0.7237878664825794,
+      "grad_norm": 11.501156203158045,
+      "learning_rate": 9.354798256665384e-07,
+      "loss": 0.474,
+      "step": 8912
+    },
+    {
+      "epoch": 0.7238690814586209,
+      "grad_norm": 4.487504865241946,
+      "learning_rate": 9.349669495709208e-07,
+      "loss": 0.3921,
+      "step": 8913
+    },
+    {
+      "epoch": 0.7239502964346626,
+      "grad_norm": 8.160345585558959,
+      "learning_rate": 9.344541817673061e-07,
+      "loss": 0.5789,
+      "step": 8914
+    },
+    {
+      "epoch": 0.7240315114107041,
+      "grad_norm": 5.808899053157705,
+      "learning_rate": 9.339415222911766e-07,
+      "loss": 0.5123,
+      "step": 8915
+    },
+    {
+      "epoch": 0.7241127263867457,
+      "grad_norm": 3.2770410237558165,
+      "learning_rate": 9.334289711780062e-07,
+      "loss": 0.7287,
+      "step": 8916
+    },
+    {
+      "epoch": 0.7241939413627873,
+      "grad_norm": 6.106039871460762,
+      "learning_rate": 9.329165284632602e-07,
+      "loss": 0.4535,
+      "step": 8917
+    },
+    {
+      "epoch": 0.7242751563388289,
+      "grad_norm": 4.5914206885164,
+      "learning_rate": 9.324041941823961e-07,
+      "loss": 0.4845,
+      "step": 8918
+    },
+    {
+      "epoch": 0.7243563713148705,
+      "grad_norm": 8.623594737117811,
+      "learning_rate": 9.318919683708661e-07,
+      "loss": 0.5425,
+      "step": 8919
+    },
+    {
+      "epoch": 0.724437586290912,
+      "grad_norm": 5.503813897391588,
+      "learning_rate": 9.313798510641117e-07,
+      "loss": 0.507,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7245188012669537,
+      "grad_norm": 27.071015075508583,
+      "learning_rate": 9.308678422975701e-07,
+      "loss": 0.4499,
+      "step": 8921
+    },
+    {
+      "epoch": 0.7246000162429952,
+      "grad_norm": 5.264809337126668,
+      "learning_rate": 9.303559421066699e-07,
+      "loss": 0.4108,
+      "step": 8922
+    },
+    {
+      "epoch": 0.7246812312190368,
+      "grad_norm": 4.739116238844419,
+      "learning_rate": 9.298441505268316e-07,
+      "loss": 0.459,
+      "step": 8923
+    },
+    {
+      "epoch": 0.7247624461950783,
+      "grad_norm": 7.57503395698494,
+      "learning_rate": 9.29332467593467e-07,
+      "loss": 0.4787,
+      "step": 8924
+    },
+    {
+      "epoch": 0.72484366117112,
+      "grad_norm": 7.320416604446665,
+      "learning_rate": 9.28820893341984e-07,
+      "loss": 0.4384,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7249248761471615,
+      "grad_norm": 8.330410195117771,
+      "learning_rate": 9.28309427807779e-07,
+      "loss": 0.5397,
+      "step": 8926
+    },
+    {
+      "epoch": 0.7250060911232031,
+      "grad_norm": 8.653115313732869,
+      "learning_rate": 9.277980710262432e-07,
+      "loss": 0.5418,
+      "step": 8927
+    },
+    {
+      "epoch": 0.7250873060992447,
+      "grad_norm": 5.712763483431208,
+      "learning_rate": 9.272868230327614e-07,
+      "loss": 0.4397,
+      "step": 8928
+    },
+    {
+      "epoch": 0.7251685210752863,
+      "grad_norm": 7.20946070132443,
+      "learning_rate": 9.267756838627079e-07,
+      "loss": 0.619,
+      "step": 8929
+    },
+    {
+      "epoch": 0.7252497360513279,
+      "grad_norm": 3.836932313032167,
+      "learning_rate": 9.262646535514499e-07,
+      "loss": 0.5411,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7253309510273694,
+      "grad_norm": 7.397925686014042,
+      "learning_rate": 9.257537321343499e-07,
+      "loss": 0.5804,
+      "step": 8931
+    },
+    {
+      "epoch": 0.7254121660034111,
+      "grad_norm": 7.414052669733408,
+      "learning_rate": 9.252429196467603e-07,
+      "loss": 0.5053,
+      "step": 8932
+    },
+    {
+      "epoch": 0.7254933809794526,
+      "grad_norm": 6.780735171626088,
+      "learning_rate": 9.247322161240252e-07,
+      "loss": 0.4038,
+      "step": 8933
+    },
+    {
+      "epoch": 0.7255745959554942,
+      "grad_norm": 4.08445905007205,
+      "learning_rate": 9.242216216014838e-07,
+      "loss": 0.6769,
+      "step": 8934
+    },
+    {
+      "epoch": 0.7256558109315357,
+      "grad_norm": 5.6918275065058275,
+      "learning_rate": 9.237111361144674e-07,
+      "loss": 0.3345,
+      "step": 8935
+    },
+    {
+      "epoch": 0.7257370259075774,
+      "grad_norm": 5.320211088421992,
+      "learning_rate": 9.232007596982978e-07,
+      "loss": 0.5265,
+      "step": 8936
+    },
+    {
+      "epoch": 0.7258182408836189,
+      "grad_norm": 5.5679637549611165,
+      "learning_rate": 9.226904923882901e-07,
+      "loss": 0.5231,
+      "step": 8937
+    },
+    {
+      "epoch": 0.7258994558596605,
+      "grad_norm": 9.0059520761274,
+      "learning_rate": 9.22180334219753e-07,
+      "loss": 0.4251,
+      "step": 8938
+    },
+    {
+      "epoch": 0.7259806708357021,
+      "grad_norm": 7.189840208919893,
+      "learning_rate": 9.216702852279857e-07,
+      "loss": 0.5836,
+      "step": 8939
+    },
+    {
+      "epoch": 0.7260618858117437,
+      "grad_norm": 8.24631932538056,
+      "learning_rate": 9.211603454482812e-07,
+      "loss": 0.5427,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7261431007877853,
+      "grad_norm": 11.91801174885722,
+      "learning_rate": 9.206505149159259e-07,
+      "loss": 0.5369,
+      "step": 8941
+    },
+    {
+      "epoch": 0.7262243157638268,
+      "grad_norm": 4.788846104575723,
+      "learning_rate": 9.201407936661963e-07,
+      "loss": 0.6227,
+      "step": 8942
+    },
+    {
+      "epoch": 0.7263055307398685,
+      "grad_norm": 4.572117460437812,
+      "learning_rate": 9.196311817343618e-07,
+      "loss": 0.4761,
+      "step": 8943
+    },
+    {
+      "epoch": 0.72638674571591,
+      "grad_norm": 6.432401948520424,
+      "learning_rate": 9.191216791556864e-07,
+      "loss": 0.494,
+      "step": 8944
+    },
+    {
+      "epoch": 0.7264679606919516,
+      "grad_norm": 3.666688290179581,
+      "learning_rate": 9.18612285965424e-07,
+      "loss": 0.4724,
+      "step": 8945
+    },
+    {
+      "epoch": 0.7265491756679932,
+      "grad_norm": 3.558045855054741,
+      "learning_rate": 9.18103002198821e-07,
+      "loss": 0.6309,
+      "step": 8946
+    },
+    {
+      "epoch": 0.7266303906440348,
+      "grad_norm": 5.292568438307435,
+      "learning_rate": 9.175938278911184e-07,
+      "loss": 0.5463,
+      "step": 8947
+    },
+    {
+      "epoch": 0.7267116056200763,
+      "grad_norm": 8.253942436158207,
+      "learning_rate": 9.170847630775489e-07,
+      "loss": 0.4128,
+      "step": 8948
+    },
+    {
+      "epoch": 0.7267928205961179,
+      "grad_norm": 12.693589172331533,
+      "learning_rate": 9.165758077933365e-07,
+      "loss": 0.5094,
+      "step": 8949
+    },
+    {
+      "epoch": 0.7268740355721595,
+      "grad_norm": 4.3211523742949876,
+      "learning_rate": 9.160669620736973e-07,
+      "loss": 0.5676,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7269552505482011,
+      "grad_norm": 11.192017775213207,
+      "learning_rate": 9.15558225953842e-07,
+      "loss": 0.5242,
+      "step": 8951
+    },
+    {
+      "epoch": 0.7270364655242427,
+      "grad_norm": 25.015456026993192,
+      "learning_rate": 9.150495994689712e-07,
+      "loss": 0.4027,
+      "step": 8952
+    },
+    {
+      "epoch": 0.7271176805002842,
+      "grad_norm": 5.152075474730206,
+      "learning_rate": 9.145410826542797e-07,
+      "loss": 0.4545,
+      "step": 8953
+    },
+    {
+      "epoch": 0.7271988954763259,
+      "grad_norm": 6.684848268615437,
+      "learning_rate": 9.140326755449555e-07,
+      "loss": 0.4933,
+      "step": 8954
+    },
+    {
+      "epoch": 0.7272801104523674,
+      "grad_norm": 5.088644961893354,
+      "learning_rate": 9.135243781761763e-07,
+      "loss": 0.4792,
+      "step": 8955
+    },
+    {
+      "epoch": 0.727361325428409,
+      "grad_norm": 3.810037495766052,
+      "learning_rate": 9.130161905831131e-07,
+      "loss": 0.5671,
+      "step": 8956
+    },
+    {
+      "epoch": 0.7274425404044506,
+      "grad_norm": 8.122117526021382,
+      "learning_rate": 9.125081128009314e-07,
+      "loss": 0.4428,
+      "step": 8957
+    },
+    {
+      "epoch": 0.7275237553804922,
+      "grad_norm": 6.6487819565990325,
+      "learning_rate": 9.120001448647867e-07,
+      "loss": 0.4813,
+      "step": 8958
+    },
+    {
+      "epoch": 0.7276049703565337,
+      "grad_norm": 6.992838856608024,
+      "learning_rate": 9.114922868098267e-07,
+      "loss": 0.4352,
+      "step": 8959
+    },
+    {
+      "epoch": 0.7276861853325753,
+      "grad_norm": 4.483263285402361,
+      "learning_rate": 9.109845386711932e-07,
+      "loss": 0.5618,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7277674003086169,
+      "grad_norm": 4.2451531883975235,
+      "learning_rate": 9.104769004840208e-07,
+      "loss": 0.5705,
+      "step": 8961
+    },
+    {
+      "epoch": 0.7278486152846585,
+      "grad_norm": 3.5682877453830186,
+      "learning_rate": 9.099693722834336e-07,
+      "loss": 0.3645,
+      "step": 8962
+    },
+    {
+      "epoch": 0.7279298302607001,
+      "grad_norm": 4.835617927031836,
+      "learning_rate": 9.094619541045516e-07,
+      "loss": 0.6062,
+      "step": 8963
+    },
+    {
+      "epoch": 0.7280110452367416,
+      "grad_norm": 4.723308760739116,
+      "learning_rate": 9.089546459824846e-07,
+      "loss": 0.441,
+      "step": 8964
+    },
+    {
+      "epoch": 0.7280922602127833,
+      "grad_norm": 5.594812767897018,
+      "learning_rate": 9.084474479523347e-07,
+      "loss": 0.4843,
+      "step": 8965
+    },
+    {
+      "epoch": 0.7281734751888248,
+      "grad_norm": 5.966924171312079,
+      "learning_rate": 9.079403600491982e-07,
+      "loss": 0.4151,
+      "step": 8966
+    },
+    {
+      "epoch": 0.7282546901648664,
+      "grad_norm": 4.336193418701047,
+      "learning_rate": 9.074333823081638e-07,
+      "loss": 0.5172,
+      "step": 8967
+    },
+    {
+      "epoch": 0.728335905140908,
+      "grad_norm": 6.063173388232329,
+      "learning_rate": 9.069265147643109e-07,
+      "loss": 0.4559,
+      "step": 8968
+    },
+    {
+      "epoch": 0.7284171201169496,
+      "grad_norm": 4.628275139282452,
+      "learning_rate": 9.064197574527112e-07,
+      "loss": 0.5929,
+      "step": 8969
+    },
+    {
+      "epoch": 0.7284983350929911,
+      "grad_norm": 3.7566474590811447,
+      "learning_rate": 9.059131104084309e-07,
+      "loss": 0.6199,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7285795500690327,
+      "grad_norm": 5.357753313412571,
+      "learning_rate": 9.054065736665268e-07,
+      "loss": 0.519,
+      "step": 8971
+    },
+    {
+      "epoch": 0.7286607650450743,
+      "grad_norm": 5.601299513674894,
+      "learning_rate": 9.049001472620481e-07,
+      "loss": 0.6267,
+      "step": 8972
+    },
+    {
+      "epoch": 0.7287419800211159,
+      "grad_norm": 6.85651492063141,
+      "learning_rate": 9.043938312300368e-07,
+      "loss": 0.4893,
+      "step": 8973
+    },
+    {
+      "epoch": 0.7288231949971575,
+      "grad_norm": 6.764798797003058,
+      "learning_rate": 9.038876256055288e-07,
+      "loss": 0.3582,
+      "step": 8974
+    },
+    {
+      "epoch": 0.728904409973199,
+      "grad_norm": 6.409476783815426,
+      "learning_rate": 9.033815304235488e-07,
+      "loss": 0.412,
+      "step": 8975
+    },
+    {
+      "epoch": 0.7289856249492407,
+      "grad_norm": 18.649132110530104,
+      "learning_rate": 9.028755457191179e-07,
+      "loss": 0.3983,
+      "step": 8976
+    },
+    {
+      "epoch": 0.7290668399252822,
+      "grad_norm": 6.513807711538606,
+      "learning_rate": 9.023696715272468e-07,
+      "loss": 0.3894,
+      "step": 8977
+    },
+    {
+      "epoch": 0.7291480549013238,
+      "grad_norm": 6.299055569156435,
+      "learning_rate": 9.018639078829378e-07,
+      "loss": 0.4124,
+      "step": 8978
+    },
+    {
+      "epoch": 0.7292292698773654,
+      "grad_norm": 3.514477375819592,
+      "learning_rate": 9.013582548211885e-07,
+      "loss": 0.568,
+      "step": 8979
+    },
+    {
+      "epoch": 0.729310484853407,
+      "grad_norm": 3.3460958135923318,
+      "learning_rate": 9.008527123769883e-07,
+      "loss": 0.4293,
+      "step": 8980
+    },
+    {
+      "epoch": 0.7293916998294485,
+      "grad_norm": 4.436979913495132,
+      "learning_rate": 9.003472805853161e-07,
+      "loss": 0.3699,
+      "step": 8981
+    },
+    {
+      "epoch": 0.7294729148054901,
+      "grad_norm": 5.276006150098665,
+      "learning_rate": 8.998419594811467e-07,
+      "loss": 0.5529,
+      "step": 8982
+    },
+    {
+      "epoch": 0.7295541297815317,
+      "grad_norm": 4.525908461414848,
+      "learning_rate": 8.993367490994451e-07,
+      "loss": 0.4417,
+      "step": 8983
+    },
+    {
+      "epoch": 0.7296353447575733,
+      "grad_norm": 6.180441338511432,
+      "learning_rate": 8.988316494751683e-07,
+      "loss": 0.4299,
+      "step": 8984
+    },
+    {
+      "epoch": 0.7297165597336149,
+      "grad_norm": 3.4284667198540983,
+      "learning_rate": 8.983266606432672e-07,
+      "loss": 0.5197,
+      "step": 8985
+    },
+    {
+      "epoch": 0.7297977747096565,
+      "grad_norm": 4.57248481351034,
+      "learning_rate": 8.978217826386853e-07,
+      "loss": 0.5724,
+      "step": 8986
+    },
+    {
+      "epoch": 0.7298789896856981,
+      "grad_norm": 3.3815176907013984,
+      "learning_rate": 8.973170154963567e-07,
+      "loss": 0.5855,
+      "step": 8987
+    },
+    {
+      "epoch": 0.7299602046617396,
+      "grad_norm": 4.911910066001664,
+      "learning_rate": 8.968123592512076e-07,
+      "loss": 0.5791,
+      "step": 8988
+    },
+    {
+      "epoch": 0.7300414196377812,
+      "grad_norm": 6.250038591037473,
+      "learning_rate": 8.963078139381595e-07,
+      "loss": 0.4979,
+      "step": 8989
+    },
+    {
+      "epoch": 0.7301226346138228,
+      "grad_norm": 4.472456064967235,
+      "learning_rate": 8.958033795921231e-07,
+      "loss": 0.4881,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7302038495898644,
+      "grad_norm": 6.163691073085615,
+      "learning_rate": 8.952990562480021e-07,
+      "loss": 0.3618,
+      "step": 8991
+    },
+    {
+      "epoch": 0.7302850645659059,
+      "grad_norm": 6.473412100775378,
+      "learning_rate": 8.947948439406934e-07,
+      "loss": 0.6202,
+      "step": 8992
+    },
+    {
+      "epoch": 0.7303662795419475,
+      "grad_norm": 6.060384183610972,
+      "learning_rate": 8.94290742705087e-07,
+      "loss": 0.4327,
+      "step": 8993
+    },
+    {
+      "epoch": 0.7304474945179891,
+      "grad_norm": 7.291283308794519,
+      "learning_rate": 8.937867525760622e-07,
+      "loss": 0.6409,
+      "step": 8994
+    },
+    {
+      "epoch": 0.7305287094940307,
+      "grad_norm": 4.822306787639834,
+      "learning_rate": 8.932828735884944e-07,
+      "loss": 0.6233,
+      "step": 8995
+    },
+    {
+      "epoch": 0.7306099244700723,
+      "grad_norm": 4.5942352939506055,
+      "learning_rate": 8.927791057772481e-07,
+      "loss": 0.4074,
+      "step": 8996
+    },
+    {
+      "epoch": 0.7306911394461139,
+      "grad_norm": 5.662519945223263,
+      "learning_rate": 8.922754491771807e-07,
+      "loss": 0.4292,
+      "step": 8997
+    },
+    {
+      "epoch": 0.7307723544221555,
+      "grad_norm": 5.270491463874796,
+      "learning_rate": 8.917719038231437e-07,
+      "loss": 0.5461,
+      "step": 8998
+    },
+    {
+      "epoch": 0.730853569398197,
+      "grad_norm": 4.428969136605445,
+      "learning_rate": 8.912684697499801e-07,
+      "loss": 0.4651,
+      "step": 8999
+    },
+    {
+      "epoch": 0.7309347843742386,
+      "grad_norm": 6.961674772191147,
+      "learning_rate": 8.907651469925236e-07,
+      "loss": 0.4143,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7310159993502802,
+      "grad_norm": 4.726617406201288,
+      "learning_rate": 8.902619355856032e-07,
+      "loss": 0.4823,
+      "step": 9001
+    },
+    {
+      "epoch": 0.7310972143263218,
+      "grad_norm": 5.037522415851766,
+      "learning_rate": 8.897588355640371e-07,
+      "loss": 0.4849,
+      "step": 9002
+    },
+    {
+      "epoch": 0.7311784293023633,
+      "grad_norm": 4.702839968633091,
+      "learning_rate": 8.892558469626375e-07,
+      "loss": 0.4989,
+      "step": 9003
+    },
+    {
+      "epoch": 0.731259644278405,
+      "grad_norm": 4.121133343260713,
+      "learning_rate": 8.887529698162079e-07,
+      "loss": 0.4176,
+      "step": 9004
+    },
+    {
+      "epoch": 0.7313408592544465,
+      "grad_norm": 5.799397892585551,
+      "learning_rate": 8.882502041595454e-07,
+      "loss": 0.4902,
+      "step": 9005
+    },
+    {
+      "epoch": 0.7314220742304881,
+      "grad_norm": 10.129599048058408,
+      "learning_rate": 8.877475500274393e-07,
+      "loss": 0.5977,
+      "step": 9006
+    },
+    {
+      "epoch": 0.7315032892065297,
+      "grad_norm": 6.254376447594514,
+      "learning_rate": 8.872450074546696e-07,
+      "loss": 0.4533,
+      "step": 9007
+    },
+    {
+      "epoch": 0.7315845041825713,
+      "grad_norm": 7.497553888308231,
+      "learning_rate": 8.867425764760104e-07,
+      "loss": 0.4203,
+      "step": 9008
+    },
+    {
+      "epoch": 0.7316657191586129,
+      "grad_norm": 7.377753478494686,
+      "learning_rate": 8.862402571262272e-07,
+      "loss": 0.5435,
+      "step": 9009
+    },
+    {
+      "epoch": 0.7317469341346544,
+      "grad_norm": 3.964499330804677,
+      "learning_rate": 8.857380494400764e-07,
+      "loss": 0.6699,
+      "step": 9010
+    },
+    {
+      "epoch": 0.731828149110696,
+      "grad_norm": 4.556955979542769,
+      "learning_rate": 8.852359534523091e-07,
+      "loss": 0.4287,
+      "step": 9011
+    },
+    {
+      "epoch": 0.7319093640867376,
+      "grad_norm": 5.283260885622843,
+      "learning_rate": 8.847339691976689e-07,
+      "loss": 0.6336,
+      "step": 9012
+    },
+    {
+      "epoch": 0.7319905790627792,
+      "grad_norm": 8.239303453615642,
+      "learning_rate": 8.842320967108886e-07,
+      "loss": 0.5459,
+      "step": 9013
+    },
+    {
+      "epoch": 0.7320717940388207,
+      "grad_norm": 9.181392373672194,
+      "learning_rate": 8.837303360266966e-07,
+      "loss": 0.4328,
+      "step": 9014
+    },
+    {
+      "epoch": 0.7321530090148624,
+      "grad_norm": 9.020007347941661,
+      "learning_rate": 8.832286871798113e-07,
+      "loss": 0.4695,
+      "step": 9015
+    },
+    {
+      "epoch": 0.7322342239909039,
+      "grad_norm": 11.695346831333792,
+      "learning_rate": 8.827271502049434e-07,
+      "loss": 0.4455,
+      "step": 9016
+    },
+    {
+      "epoch": 0.7323154389669455,
+      "grad_norm": 5.166457906877146,
+      "learning_rate": 8.822257251367983e-07,
+      "loss": 0.4783,
+      "step": 9017
+    },
+    {
+      "epoch": 0.7323966539429871,
+      "grad_norm": 3.9452559803067575,
+      "learning_rate": 8.817244120100702e-07,
+      "loss": 0.4167,
+      "step": 9018
+    },
+    {
+      "epoch": 0.7324778689190287,
+      "grad_norm": 5.030514569697522,
+      "learning_rate": 8.812232108594482e-07,
+      "loss": 0.3934,
+      "step": 9019
+    },
+    {
+      "epoch": 0.7325590838950703,
+      "grad_norm": 10.567800558944477,
+      "learning_rate": 8.807221217196135e-07,
+      "loss": 0.5891,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7326402988711118,
+      "grad_norm": 7.700337486933208,
+      "learning_rate": 8.802211446252379e-07,
+      "loss": 0.5597,
+      "step": 9021
+    },
+    {
+      "epoch": 0.7327215138471534,
+      "grad_norm": 11.713465789679601,
+      "learning_rate": 8.797202796109869e-07,
+      "loss": 0.5172,
+      "step": 9022
+    },
+    {
+      "epoch": 0.732802728823195,
+      "grad_norm": 7.145291394301389,
+      "learning_rate": 8.792195267115163e-07,
+      "loss": 0.4569,
+      "step": 9023
+    },
+    {
+      "epoch": 0.7328839437992366,
+      "grad_norm": 4.598031815516408,
+      "learning_rate": 8.787188859614768e-07,
+      "loss": 0.6056,
+      "step": 9024
+    },
+    {
+      "epoch": 0.7329651587752781,
+      "grad_norm": 5.275503679314213,
+      "learning_rate": 8.782183573955105e-07,
+      "loss": 0.4652,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7330463737513198,
+      "grad_norm": 6.295243362632932,
+      "learning_rate": 8.777179410482498e-07,
+      "loss": 0.6455,
+      "step": 9026
+    },
+    {
+      "epoch": 0.7331275887273613,
+      "grad_norm": 4.4474547984782395,
+      "learning_rate": 8.772176369543229e-07,
+      "loss": 0.3385,
+      "step": 9027
+    },
+    {
+      "epoch": 0.7332088037034029,
+      "grad_norm": 3.172334169434078,
+      "learning_rate": 8.767174451483468e-07,
+      "loss": 0.5206,
+      "step": 9028
+    },
+    {
+      "epoch": 0.7332900186794445,
+      "grad_norm": 4.6456662676555815,
+      "learning_rate": 8.762173656649317e-07,
+      "loss": 0.6199,
+      "step": 9029
+    },
+    {
+      "epoch": 0.7333712336554861,
+      "grad_norm": 7.265881502649945,
+      "learning_rate": 8.757173985386819e-07,
+      "loss": 0.4957,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7334524486315277,
+      "grad_norm": 3.4982261171160833,
+      "learning_rate": 8.752175438041908e-07,
+      "loss": 0.5664,
+      "step": 9031
+    },
+    {
+      "epoch": 0.7335336636075692,
+      "grad_norm": 3.6827798694601754,
+      "learning_rate": 8.747178014960467e-07,
+      "loss": 0.4881,
+      "step": 9032
+    },
+    {
+      "epoch": 0.7336148785836109,
+      "grad_norm": 7.960332905031216,
+      "learning_rate": 8.742181716488302e-07,
+      "loss": 0.3276,
+      "step": 9033
+    },
+    {
+      "epoch": 0.7336960935596524,
+      "grad_norm": 4.4974178890676075,
+      "learning_rate": 8.737186542971115e-07,
+      "loss": 0.4636,
+      "step": 9034
+    },
+    {
+      "epoch": 0.733777308535694,
+      "grad_norm": 4.122814660625721,
+      "learning_rate": 8.732192494754541e-07,
+      "loss": 0.3476,
+      "step": 9035
+    },
+    {
+      "epoch": 0.7338585235117355,
+      "grad_norm": 4.213370810352235,
+      "learning_rate": 8.727199572184161e-07,
+      "loss": 0.5909,
+      "step": 9036
+    },
+    {
+      "epoch": 0.7339397384877772,
+      "grad_norm": 10.14479503528053,
+      "learning_rate": 8.722207775605437e-07,
+      "loss": 0.4764,
+      "step": 9037
+    },
+    {
+      "epoch": 0.7340209534638187,
+      "grad_norm": 4.624804246473112,
+      "learning_rate": 8.717217105363798e-07,
+      "loss": 0.5015,
+      "step": 9038
+    },
+    {
+      "epoch": 0.7341021684398603,
+      "grad_norm": 5.183928674417439,
+      "learning_rate": 8.712227561804548e-07,
+      "loss": 0.4768,
+      "step": 9039
+    },
+    {
+      "epoch": 0.734183383415902,
+      "grad_norm": 7.897979278065682,
+      "learning_rate": 8.707239145272958e-07,
+      "loss": 0.5025,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7342645983919435,
+      "grad_norm": 16.857325758216245,
+      "learning_rate": 8.702251856114191e-07,
+      "loss": 0.4793,
+      "step": 9041
+    },
+    {
+      "epoch": 0.7343458133679851,
+      "grad_norm": 6.305732473052533,
+      "learning_rate": 8.697265694673334e-07,
+      "loss": 0.4738,
+      "step": 9042
+    },
+    {
+      "epoch": 0.7344270283440266,
+      "grad_norm": 4.796761988404685,
+      "learning_rate": 8.692280661295419e-07,
+      "loss": 0.334,
+      "step": 9043
+    },
+    {
+      "epoch": 0.7345082433200683,
+      "grad_norm": 5.71378461104089,
+      "learning_rate": 8.687296756325364e-07,
+      "loss": 0.5107,
+      "step": 9044
+    },
+    {
+      "epoch": 0.7345894582961098,
+      "grad_norm": 10.227432767121579,
+      "learning_rate": 8.68231398010804e-07,
+      "loss": 0.6141,
+      "step": 9045
+    },
+    {
+      "epoch": 0.7346706732721514,
+      "grad_norm": 6.276424519094259,
+      "learning_rate": 8.677332332988236e-07,
+      "loss": 0.4475,
+      "step": 9046
+    },
+    {
+      "epoch": 0.7347518882481929,
+      "grad_norm": 5.843493313646698,
+      "learning_rate": 8.672351815310651e-07,
+      "loss": 0.4045,
+      "step": 9047
+    },
+    {
+      "epoch": 0.7348331032242346,
+      "grad_norm": 7.797951199210248,
+      "learning_rate": 8.667372427419895e-07,
+      "loss": 0.4015,
+      "step": 9048
+    },
+    {
+      "epoch": 0.7349143182002761,
+      "grad_norm": 7.633315491295807,
+      "learning_rate": 8.66239416966054e-07,
+      "loss": 0.4516,
+      "step": 9049
+    },
+    {
+      "epoch": 0.7349955331763177,
+      "grad_norm": 5.0774721644669025,
+      "learning_rate": 8.657417042377034e-07,
+      "loss": 0.5822,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7350767481523593,
+      "grad_norm": 3.3621612548870967,
+      "learning_rate": 8.652441045913775e-07,
+      "loss": 0.5267,
+      "step": 9051
+    },
+    {
+      "epoch": 0.7351579631284009,
+      "grad_norm": 6.871554047086195,
+      "learning_rate": 8.647466180615085e-07,
+      "loss": 0.5536,
+      "step": 9052
+    },
+    {
+      "epoch": 0.7352391781044425,
+      "grad_norm": 6.572776544165536,
+      "learning_rate": 8.642492446825193e-07,
+      "loss": 0.5135,
+      "step": 9053
+    },
+    {
+      "epoch": 0.735320393080484,
+      "grad_norm": 5.770321723227472,
+      "learning_rate": 8.637519844888245e-07,
+      "loss": 0.4804,
+      "step": 9054
+    },
+    {
+      "epoch": 0.7354016080565257,
+      "grad_norm": 10.272898706445313,
+      "learning_rate": 8.632548375148333e-07,
+      "loss": 0.5766,
+      "step": 9055
+    },
+    {
+      "epoch": 0.7354828230325672,
+      "grad_norm": 7.89177634147062,
+      "learning_rate": 8.627578037949441e-07,
+      "loss": 0.6254,
+      "step": 9056
+    },
+    {
+      "epoch": 0.7355640380086088,
+      "grad_norm": 6.509795135122351,
+      "learning_rate": 8.62260883363551e-07,
+      "loss": 0.4276,
+      "step": 9057
+    },
+    {
+      "epoch": 0.7356452529846503,
+      "grad_norm": 5.327797962449124,
+      "learning_rate": 8.617640762550361e-07,
+      "loss": 0.4104,
+      "step": 9058
+    },
+    {
+      "epoch": 0.735726467960692,
+      "grad_norm": 10.83596502585155,
+      "learning_rate": 8.612673825037776e-07,
+      "loss": 0.6383,
+      "step": 9059
+    },
+    {
+      "epoch": 0.7358076829367335,
+      "grad_norm": 3.2930515840707257,
+      "learning_rate": 8.607708021441436e-07,
+      "loss": 0.5538,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7358888979127751,
+      "grad_norm": 6.123300825914036,
+      "learning_rate": 8.602743352104936e-07,
+      "loss": 0.5416,
+      "step": 9061
+    },
+    {
+      "epoch": 0.7359701128888168,
+      "grad_norm": 4.126241878471722,
+      "learning_rate": 8.597779817371824e-07,
+      "loss": 0.4344,
+      "step": 9062
+    },
+    {
+      "epoch": 0.7360513278648583,
+      "grad_norm": 4.001896575407721,
+      "learning_rate": 8.592817417585534e-07,
+      "loss": 0.5276,
+      "step": 9063
+    },
+    {
+      "epoch": 0.7361325428408999,
+      "grad_norm": 4.410919527265607,
+      "learning_rate": 8.587856153089444e-07,
+      "loss": 0.5128,
+      "step": 9064
+    },
+    {
+      "epoch": 0.7362137578169414,
+      "grad_norm": 4.6300581291782885,
+      "learning_rate": 8.582896024226855e-07,
+      "loss": 0.5539,
+      "step": 9065
+    },
+    {
+      "epoch": 0.7362949727929831,
+      "grad_norm": 5.804508870075955,
+      "learning_rate": 8.577937031340975e-07,
+      "loss": 0.4621,
+      "step": 9066
+    },
+    {
+      "epoch": 0.7363761877690246,
+      "grad_norm": 3.9641442301684653,
+      "learning_rate": 8.572979174774934e-07,
+      "loss": 0.4759,
+      "step": 9067
+    },
+    {
+      "epoch": 0.7364574027450662,
+      "grad_norm": 3.935697034995573,
+      "learning_rate": 8.568022454871802e-07,
+      "loss": 0.5874,
+      "step": 9068
+    },
+    {
+      "epoch": 0.7365386177211077,
+      "grad_norm": 5.777703937802822,
+      "learning_rate": 8.563066871974543e-07,
+      "loss": 0.4736,
+      "step": 9069
+    },
+    {
+      "epoch": 0.7366198326971494,
+      "grad_norm": 3.7700387674835176,
+      "learning_rate": 8.558112426426062e-07,
+      "loss": 0.4482,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7367010476731909,
+      "grad_norm": 4.042064333038825,
+      "learning_rate": 8.553159118569196e-07,
+      "loss": 0.3638,
+      "step": 9071
+    },
+    {
+      "epoch": 0.7367822626492325,
+      "grad_norm": 5.379208558888325,
+      "learning_rate": 8.548206948746673e-07,
+      "loss": 0.359,
+      "step": 9072
+    },
+    {
+      "epoch": 0.7368634776252742,
+      "grad_norm": 5.80342467828285,
+      "learning_rate": 8.543255917301163e-07,
+      "loss": 0.5691,
+      "step": 9073
+    },
+    {
+      "epoch": 0.7369446926013157,
+      "grad_norm": 9.897018766300208,
+      "learning_rate": 8.538306024575235e-07,
+      "loss": 0.5253,
+      "step": 9074
+    },
+    {
+      "epoch": 0.7370259075773573,
+      "grad_norm": 5.1025049749314055,
+      "learning_rate": 8.533357270911419e-07,
+      "loss": 0.4075,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7371071225533988,
+      "grad_norm": 8.535318151291682,
+      "learning_rate": 8.52840965665212e-07,
+      "loss": 0.4842,
+      "step": 9076
+    },
+    {
+      "epoch": 0.7371883375294405,
+      "grad_norm": 8.277801039325004,
+      "learning_rate": 8.523463182139699e-07,
+      "loss": 0.4553,
+      "step": 9077
+    },
+    {
+      "epoch": 0.737269552505482,
+      "grad_norm": 4.521612840769156,
+      "learning_rate": 8.518517847716435e-07,
+      "loss": 0.627,
+      "step": 9078
+    },
+    {
+      "epoch": 0.7373507674815236,
+      "grad_norm": 3.8960485335075603,
+      "learning_rate": 8.513573653724508e-07,
+      "loss": 0.5943,
+      "step": 9079
+    },
+    {
+      "epoch": 0.7374319824575651,
+      "grad_norm": 4.606452564353779,
+      "learning_rate": 8.508630600506021e-07,
+      "loss": 0.6701,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7375131974336068,
+      "grad_norm": 7.052763751543803,
+      "learning_rate": 8.503688688403028e-07,
+      "loss": 0.5393,
+      "step": 9081
+    },
+    {
+      "epoch": 0.7375944124096483,
+      "grad_norm": 4.538323480054813,
+      "learning_rate": 8.498747917757464e-07,
+      "loss": 0.434,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7376756273856899,
+      "grad_norm": 3.679103857972724,
+      "learning_rate": 8.49380828891121e-07,
+      "loss": 0.561,
+      "step": 9083
+    },
+    {
+      "epoch": 0.7377568423617316,
+      "grad_norm": 9.599527757034656,
+      "learning_rate": 8.488869802206073e-07,
+      "loss": 0.4462,
+      "step": 9084
+    },
+    {
+      "epoch": 0.7378380573377731,
+      "grad_norm": 5.51962161938916,
+      "learning_rate": 8.483932457983765e-07,
+      "loss": 0.5055,
+      "step": 9085
+    },
+    {
+      "epoch": 0.7379192723138147,
+      "grad_norm": 7.720325892853574,
+      "learning_rate": 8.478996256585909e-07,
+      "loss": 0.3904,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7380004872898562,
+      "grad_norm": 4.8150352416660125,
+      "learning_rate": 8.474061198354086e-07,
+      "loss": 0.3425,
+      "step": 9087
+    },
+    {
+      "epoch": 0.7380817022658979,
+      "grad_norm": 4.123061421099957,
+      "learning_rate": 8.469127283629766e-07,
+      "loss": 0.4812,
+      "step": 9088
+    },
+    {
+      "epoch": 0.7381629172419394,
+      "grad_norm": 4.2839585199281975,
+      "learning_rate": 8.464194512754339e-07,
+      "loss": 0.6729,
+      "step": 9089
+    },
+    {
+      "epoch": 0.738244132217981,
+      "grad_norm": 10.07664977406758,
+      "learning_rate": 8.459262886069139e-07,
+      "loss": 0.4227,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7383253471940225,
+      "grad_norm": 5.667705109796953,
+      "learning_rate": 8.454332403915416e-07,
+      "loss": 0.4117,
+      "step": 9091
+    },
+    {
+      "epoch": 0.7384065621700642,
+      "grad_norm": 5.360148423288729,
+      "learning_rate": 8.44940306663432e-07,
+      "loss": 0.4129,
+      "step": 9092
+    },
+    {
+      "epoch": 0.7384877771461057,
+      "grad_norm": 4.840577845593679,
+      "learning_rate": 8.444474874566935e-07,
+      "loss": 0.43,
+      "step": 9093
+    },
+    {
+      "epoch": 0.7385689921221473,
+      "grad_norm": 5.228586184130316,
+      "learning_rate": 8.439547828054276e-07,
+      "loss": 0.5745,
+      "step": 9094
+    },
+    {
+      "epoch": 0.738650207098189,
+      "grad_norm": 5.804073223258344,
+      "learning_rate": 8.434621927437253e-07,
+      "loss": 0.6014,
+      "step": 9095
+    },
+    {
+      "epoch": 0.7387314220742305,
+      "grad_norm": 6.711353565911964,
+      "learning_rate": 8.429697173056726e-07,
+      "loss": 0.4756,
+      "step": 9096
+    },
+    {
+      "epoch": 0.7388126370502721,
+      "grad_norm": 5.514320352974961,
+      "learning_rate": 8.42477356525346e-07,
+      "loss": 0.5374,
+      "step": 9097
+    },
+    {
+      "epoch": 0.7388938520263136,
+      "grad_norm": 3.835666081477301,
+      "learning_rate": 8.419851104368143e-07,
+      "loss": 0.3875,
+      "step": 9098
+    },
+    {
+      "epoch": 0.7389750670023553,
+      "grad_norm": 12.210910906950325,
+      "learning_rate": 8.414929790741371e-07,
+      "loss": 0.4524,
+      "step": 9099
+    },
+    {
+      "epoch": 0.7390562819783968,
+      "grad_norm": 5.875608947253129,
+      "learning_rate": 8.410009624713691e-07,
+      "loss": 0.5712,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7391374969544384,
+      "grad_norm": 4.500274285625198,
+      "learning_rate": 8.405090606625547e-07,
+      "loss": 0.6175,
+      "step": 9101
+    },
+    {
+      "epoch": 0.7392187119304799,
+      "grad_norm": 7.944159885650859,
+      "learning_rate": 8.400172736817294e-07,
+      "loss": 0.5565,
+      "step": 9102
+    },
+    {
+      "epoch": 0.7392999269065216,
+      "grad_norm": 10.979049183670785,
+      "learning_rate": 8.395256015629233e-07,
+      "loss": 0.4571,
+      "step": 9103
+    },
+    {
+      "epoch": 0.7393811418825631,
+      "grad_norm": 6.5837339811285815,
+      "learning_rate": 8.390340443401588e-07,
+      "loss": 0.4318,
+      "step": 9104
+    },
+    {
+      "epoch": 0.7394623568586047,
+      "grad_norm": 6.045591348963159,
+      "learning_rate": 8.385426020474468e-07,
+      "loss": 0.4703,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7395435718346464,
+      "grad_norm": 11.087294536746336,
+      "learning_rate": 8.380512747187944e-07,
+      "loss": 0.3586,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7396247868106879,
+      "grad_norm": 5.179365997405158,
+      "learning_rate": 8.375600623881983e-07,
+      "loss": 0.3442,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7397060017867295,
+      "grad_norm": 5.5937991392764355,
+      "learning_rate": 8.370689650896465e-07,
+      "loss": 0.4803,
+      "step": 9108
+    },
+    {
+      "epoch": 0.739787216762771,
+      "grad_norm": 5.345681951321185,
+      "learning_rate": 8.365779828571214e-07,
+      "loss": 0.5371,
+      "step": 9109
+    },
+    {
+      "epoch": 0.7398684317388127,
+      "grad_norm": 10.438164567037536,
+      "learning_rate": 8.360871157245973e-07,
+      "loss": 0.4857,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7399496467148542,
+      "grad_norm": 12.570837392968885,
+      "learning_rate": 8.355963637260387e-07,
+      "loss": 0.4761,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7400308616908958,
+      "grad_norm": 5.293881936064758,
+      "learning_rate": 8.351057268954019e-07,
+      "loss": 0.6403,
+      "step": 9112
+    },
+    {
+      "epoch": 0.7401120766669373,
+      "grad_norm": 8.074420587248548,
+      "learning_rate": 8.346152052666385e-07,
+      "loss": 0.4348,
+      "step": 9113
+    },
+    {
+      "epoch": 0.740193291642979,
+      "grad_norm": 4.833671703517281,
+      "learning_rate": 8.341247988736889e-07,
+      "loss": 0.5021,
+      "step": 9114
+    },
+    {
+      "epoch": 0.7402745066190205,
+      "grad_norm": 5.173257511917773,
+      "learning_rate": 8.336345077504851e-07,
+      "loss": 0.4408,
+      "step": 9115
+    },
+    {
+      "epoch": 0.7403557215950621,
+      "grad_norm": 4.696108390642895,
+      "learning_rate": 8.331443319309557e-07,
+      "loss": 0.4978,
+      "step": 9116
+    },
+    {
+      "epoch": 0.7404369365711038,
+      "grad_norm": 6.333906802407053,
+      "learning_rate": 8.326542714490172e-07,
+      "loss": 0.4354,
+      "step": 9117
+    },
+    {
+      "epoch": 0.7405181515471453,
+      "grad_norm": 4.361831325871427,
+      "learning_rate": 8.321643263385776e-07,
+      "loss": 0.4368,
+      "step": 9118
+    },
+    {
+      "epoch": 0.7405993665231869,
+      "grad_norm": 4.6714291210784085,
+      "learning_rate": 8.316744966335408e-07,
+      "loss": 0.6296,
+      "step": 9119
+    },
+    {
+      "epoch": 0.7406805814992284,
+      "grad_norm": 4.933127806189035,
+      "learning_rate": 8.31184782367799e-07,
+      "loss": 0.5809,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7407617964752701,
+      "grad_norm": 4.025591077863814,
+      "learning_rate": 8.306951835752378e-07,
+      "loss": 0.6013,
+      "step": 9121
+    },
+    {
+      "epoch": 0.7408430114513116,
+      "grad_norm": 11.41144080285039,
+      "learning_rate": 8.302057002897349e-07,
+      "loss": 0.5021,
+      "step": 9122
+    },
+    {
+      "epoch": 0.7409242264273532,
+      "grad_norm": 4.627955700774956,
+      "learning_rate": 8.297163325451612e-07,
+      "loss": 0.5904,
+      "step": 9123
+    },
+    {
+      "epoch": 0.7410054414033947,
+      "grad_norm": 7.878246043105683,
+      "learning_rate": 8.292270803753765e-07,
+      "loss": 0.457,
+      "step": 9124
+    },
+    {
+      "epoch": 0.7410866563794364,
+      "grad_norm": 4.436045720835732,
+      "learning_rate": 8.287379438142365e-07,
+      "loss": 0.5311,
+      "step": 9125
+    },
+    {
+      "epoch": 0.7411678713554779,
+      "grad_norm": 5.747604125886795,
+      "learning_rate": 8.282489228955856e-07,
+      "loss": 0.4093,
+      "step": 9126
+    },
+    {
+      "epoch": 0.7412490863315195,
+      "grad_norm": 26.14780023838032,
+      "learning_rate": 8.277600176532608e-07,
+      "loss": 0.4506,
+      "step": 9127
+    },
+    {
+      "epoch": 0.7413303013075612,
+      "grad_norm": 5.408380773620181,
+      "learning_rate": 8.272712281210926e-07,
+      "loss": 0.5179,
+      "step": 9128
+    },
+    {
+      "epoch": 0.7414115162836027,
+      "grad_norm": 3.881322907472951,
+      "learning_rate": 8.267825543329033e-07,
+      "loss": 0.5395,
+      "step": 9129
+    },
+    {
+      "epoch": 0.7414927312596443,
+      "grad_norm": 4.612767077216341,
+      "learning_rate": 8.262939963225058e-07,
+      "loss": 0.4454,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7415739462356858,
+      "grad_norm": 20.881362818292562,
+      "learning_rate": 8.258055541237054e-07,
+      "loss": 0.4422,
+      "step": 9131
+    },
+    {
+      "epoch": 0.7416551612117275,
+      "grad_norm": 10.331638585893913,
+      "learning_rate": 8.253172277703006e-07,
+      "loss": 0.4213,
+      "step": 9132
+    },
+    {
+      "epoch": 0.741736376187769,
+      "grad_norm": 7.471565014697083,
+      "learning_rate": 8.248290172960804e-07,
+      "loss": 0.4839,
+      "step": 9133
+    },
+    {
+      "epoch": 0.7418175911638106,
+      "grad_norm": 6.969777149983467,
+      "learning_rate": 8.24340922734826e-07,
+      "loss": 0.4021,
+      "step": 9134
+    },
+    {
+      "epoch": 0.7418988061398522,
+      "grad_norm": 6.91819887817937,
+      "learning_rate": 8.238529441203111e-07,
+      "loss": 0.3838,
+      "step": 9135
+    },
+    {
+      "epoch": 0.7419800211158938,
+      "grad_norm": 3.8948651109688703,
+      "learning_rate": 8.233650814863026e-07,
+      "loss": 0.4054,
+      "step": 9136
+    },
+    {
+      "epoch": 0.7420612360919353,
+      "grad_norm": 4.018293826197548,
+      "learning_rate": 8.228773348665561e-07,
+      "loss": 0.4144,
+      "step": 9137
+    },
+    {
+      "epoch": 0.7421424510679769,
+      "grad_norm": 3.9662330274627355,
+      "learning_rate": 8.223897042948228e-07,
+      "loss": 0.4907,
+      "step": 9138
+    },
+    {
+      "epoch": 0.7422236660440186,
+      "grad_norm": 10.723781510737455,
+      "learning_rate": 8.219021898048435e-07,
+      "loss": 0.5779,
+      "step": 9139
+    },
+    {
+      "epoch": 0.7423048810200601,
+      "grad_norm": 4.79787389560584,
+      "learning_rate": 8.214147914303505e-07,
+      "loss": 0.4803,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7423860959961017,
+      "grad_norm": 3.222506085754893,
+      "learning_rate": 8.209275092050701e-07,
+      "loss": 0.5073,
+      "step": 9141
+    },
+    {
+      "epoch": 0.7424673109721432,
+      "grad_norm": 7.127778601731678,
+      "learning_rate": 8.204403431627206e-07,
+      "loss": 0.4199,
+      "step": 9142
+    },
+    {
+      "epoch": 0.7425485259481849,
+      "grad_norm": 6.285963377503724,
+      "learning_rate": 8.199532933370094e-07,
+      "loss": 0.3786,
+      "step": 9143
+    },
+    {
+      "epoch": 0.7426297409242264,
+      "grad_norm": 8.21223183631982,
+      "learning_rate": 8.194663597616398e-07,
+      "loss": 0.5224,
+      "step": 9144
+    },
+    {
+      "epoch": 0.742710955900268,
+      "grad_norm": 9.976782891147915,
+      "learning_rate": 8.18979542470304e-07,
+      "loss": 0.5302,
+      "step": 9145
+    },
+    {
+      "epoch": 0.7427921708763096,
+      "grad_norm": 6.028309491334264,
+      "learning_rate": 8.184928414966873e-07,
+      "loss": 0.4914,
+      "step": 9146
+    },
+    {
+      "epoch": 0.7428733858523512,
+      "grad_norm": 13.676857941967034,
+      "learning_rate": 8.180062568744657e-07,
+      "loss": 0.5282,
+      "step": 9147
+    },
+    {
+      "epoch": 0.7429546008283927,
+      "grad_norm": 17.032146668354567,
+      "learning_rate": 8.175197886373093e-07,
+      "loss": 0.4902,
+      "step": 9148
+    },
+    {
+      "epoch": 0.7430358158044343,
+      "grad_norm": 5.4622799815547936,
+      "learning_rate": 8.170334368188798e-07,
+      "loss": 0.3934,
+      "step": 9149
+    },
+    {
+      "epoch": 0.743117030780476,
+      "grad_norm": 8.517487460268729,
+      "learning_rate": 8.16547201452829e-07,
+      "loss": 0.5146,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7431982457565175,
+      "grad_norm": 5.038501260406377,
+      "learning_rate": 8.160610825728029e-07,
+      "loss": 0.5049,
+      "step": 9151
+    },
+    {
+      "epoch": 0.7432794607325591,
+      "grad_norm": 4.455377373168075,
+      "learning_rate": 8.155750802124379e-07,
+      "loss": 0.5832,
+      "step": 9152
+    },
+    {
+      "epoch": 0.7433606757086006,
+      "grad_norm": 3.9339441762573863,
+      "learning_rate": 8.150891944053615e-07,
+      "loss": 0.4169,
+      "step": 9153
+    },
+    {
+      "epoch": 0.7434418906846423,
+      "grad_norm": 3.517549583636786,
+      "learning_rate": 8.146034251851959e-07,
+      "loss": 0.5344,
+      "step": 9154
+    },
+    {
+      "epoch": 0.7435231056606838,
+      "grad_norm": 4.5372086466348085,
+      "learning_rate": 8.141177725855543e-07,
+      "loss": 0.5208,
+      "step": 9155
+    },
+    {
+      "epoch": 0.7436043206367254,
+      "grad_norm": 4.458967507222066,
+      "learning_rate": 8.136322366400396e-07,
+      "loss": 0.4537,
+      "step": 9156
+    },
+    {
+      "epoch": 0.743685535612767,
+      "grad_norm": 3.5505929046873104,
+      "learning_rate": 8.131468173822499e-07,
+      "loss": 0.3824,
+      "step": 9157
+    },
+    {
+      "epoch": 0.7437667505888086,
+      "grad_norm": 4.049572608845729,
+      "learning_rate": 8.126615148457728e-07,
+      "loss": 0.4316,
+      "step": 9158
+    },
+    {
+      "epoch": 0.7438479655648501,
+      "grad_norm": 6.329253089634774,
+      "learning_rate": 8.121763290641879e-07,
+      "loss": 0.3844,
+      "step": 9159
+    },
+    {
+      "epoch": 0.7439291805408917,
+      "grad_norm": 5.7289617458205235,
+      "learning_rate": 8.116912600710694e-07,
+      "loss": 0.5157,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7440103955169334,
+      "grad_norm": 7.461558435663731,
+      "learning_rate": 8.112063078999794e-07,
+      "loss": 0.4108,
+      "step": 9161
+    },
+    {
+      "epoch": 0.7440916104929749,
+      "grad_norm": 6.260601884198078,
+      "learning_rate": 8.107214725844753e-07,
+      "loss": 0.4063,
+      "step": 9162
+    },
+    {
+      "epoch": 0.7441728254690165,
+      "grad_norm": 7.096380076386157,
+      "learning_rate": 8.102367541581055e-07,
+      "loss": 0.375,
+      "step": 9163
+    },
+    {
+      "epoch": 0.744254040445058,
+      "grad_norm": 4.583892134072468,
+      "learning_rate": 8.097521526544094e-07,
+      "loss": 0.5004,
+      "step": 9164
+    },
+    {
+      "epoch": 0.7443352554210997,
+      "grad_norm": 4.285669547146368,
+      "learning_rate": 8.092676681069189e-07,
+      "loss": 0.4534,
+      "step": 9165
+    },
+    {
+      "epoch": 0.7444164703971412,
+      "grad_norm": 5.84102693467306,
+      "learning_rate": 8.087833005491568e-07,
+      "loss": 0.6221,
+      "step": 9166
+    },
+    {
+      "epoch": 0.7444976853731828,
+      "grad_norm": 4.581219331241057,
+      "learning_rate": 8.082990500146398e-07,
+      "loss": 0.5261,
+      "step": 9167
+    },
+    {
+      "epoch": 0.7445789003492244,
+      "grad_norm": 6.253344718728393,
+      "learning_rate": 8.078149165368762e-07,
+      "loss": 0.4019,
+      "step": 9168
+    },
+    {
+      "epoch": 0.744660115325266,
+      "grad_norm": 4.961021707376709,
+      "learning_rate": 8.073309001493637e-07,
+      "loss": 0.5145,
+      "step": 9169
+    },
+    {
+      "epoch": 0.7447413303013075,
+      "grad_norm": 60.59463933382603,
+      "learning_rate": 8.068470008855953e-07,
+      "loss": 0.4394,
+      "step": 9170
+    },
+    {
+      "epoch": 0.7448225452773491,
+      "grad_norm": 5.570185363264054,
+      "learning_rate": 8.063632187790538e-07,
+      "loss": 0.4144,
+      "step": 9171
+    },
+    {
+      "epoch": 0.7449037602533908,
+      "grad_norm": 6.346150984764303,
+      "learning_rate": 8.05879553863213e-07,
+      "loss": 0.4612,
+      "step": 9172
+    },
+    {
+      "epoch": 0.7449849752294323,
+      "grad_norm": 5.7858397897115745,
+      "learning_rate": 8.053960061715421e-07,
+      "loss": 0.5154,
+      "step": 9173
+    },
+    {
+      "epoch": 0.7450661902054739,
+      "grad_norm": 9.04581065503539,
+      "learning_rate": 8.049125757374978e-07,
+      "loss": 0.4543,
+      "step": 9174
+    },
+    {
+      "epoch": 0.7451474051815155,
+      "grad_norm": 4.5831119885995575,
+      "learning_rate": 8.044292625945327e-07,
+      "loss": 0.5,
+      "step": 9175
+    },
+    {
+      "epoch": 0.7452286201575571,
+      "grad_norm": 4.73034396532834,
+      "learning_rate": 8.039460667760892e-07,
+      "loss": 0.6158,
+      "step": 9176
+    },
+    {
+      "epoch": 0.7453098351335986,
+      "grad_norm": 7.297415511486316,
+      "learning_rate": 8.034629883156019e-07,
+      "loss": 0.4434,
+      "step": 9177
+    },
+    {
+      "epoch": 0.7453910501096402,
+      "grad_norm": 7.087502286901482,
+      "learning_rate": 8.029800272464963e-07,
+      "loss": 0.3851,
+      "step": 9178
+    },
+    {
+      "epoch": 0.7454722650856818,
+      "grad_norm": 21.28250462127868,
+      "learning_rate": 8.024971836021922e-07,
+      "loss": 0.3312,
+      "step": 9179
+    },
+    {
+      "epoch": 0.7455534800617234,
+      "grad_norm": 4.028786793175537,
+      "learning_rate": 8.020144574160984e-07,
+      "loss": 0.5862,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7456346950377649,
+      "grad_norm": 8.089625182959997,
+      "learning_rate": 8.015318487216184e-07,
+      "loss": 0.4179,
+      "step": 9181
+    },
+    {
+      "epoch": 0.7457159100138065,
+      "grad_norm": 10.703076445335201,
+      "learning_rate": 8.010493575521444e-07,
+      "loss": 0.5247,
+      "step": 9182
+    },
+    {
+      "epoch": 0.7457971249898482,
+      "grad_norm": 5.996668534107049,
+      "learning_rate": 8.005669839410643e-07,
+      "loss": 0.6692,
+      "step": 9183
+    },
+    {
+      "epoch": 0.7458783399658897,
+      "grad_norm": 4.264762756411719,
+      "learning_rate": 8.00084727921755e-07,
+      "loss": 0.7248,
+      "step": 9184
+    },
+    {
+      "epoch": 0.7459595549419313,
+      "grad_norm": 6.337959117205082,
+      "learning_rate": 7.996025895275846e-07,
+      "loss": 0.4624,
+      "step": 9185
+    },
+    {
+      "epoch": 0.7460407699179729,
+      "grad_norm": 4.73694838657318,
+      "learning_rate": 7.991205687919163e-07,
+      "loss": 0.5874,
+      "step": 9186
+    },
+    {
+      "epoch": 0.7461219848940145,
+      "grad_norm": 4.444846677665557,
+      "learning_rate": 7.986386657481032e-07,
+      "loss": 0.4305,
+      "step": 9187
+    },
+    {
+      "epoch": 0.746203199870056,
+      "grad_norm": 4.371046177977482,
+      "learning_rate": 7.981568804294895e-07,
+      "loss": 0.4158,
+      "step": 9188
+    },
+    {
+      "epoch": 0.7462844148460976,
+      "grad_norm": 5.23612698457551,
+      "learning_rate": 7.976752128694134e-07,
+      "loss": 0.5297,
+      "step": 9189
+    },
+    {
+      "epoch": 0.7463656298221392,
+      "grad_norm": 5.6064882601787325,
+      "learning_rate": 7.971936631012033e-07,
+      "loss": 0.4,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7464468447981808,
+      "grad_norm": 4.736891493880583,
+      "learning_rate": 7.96712231158179e-07,
+      "loss": 0.465,
+      "step": 9191
+    },
+    {
+      "epoch": 0.7465280597742223,
+      "grad_norm": 16.20683819170043,
+      "learning_rate": 7.962309170736546e-07,
+      "loss": 0.538,
+      "step": 9192
+    },
+    {
+      "epoch": 0.746609274750264,
+      "grad_norm": 4.072174973984312,
+      "learning_rate": 7.957497208809328e-07,
+      "loss": 0.5302,
+      "step": 9193
+    },
+    {
+      "epoch": 0.7466904897263056,
+      "grad_norm": 3.329371490211698,
+      "learning_rate": 7.952686426133105e-07,
+      "loss": 0.5949,
+      "step": 9194
+    },
+    {
+      "epoch": 0.7467717047023471,
+      "grad_norm": 10.181237849562704,
+      "learning_rate": 7.947876823040771e-07,
+      "loss": 0.5202,
+      "step": 9195
+    },
+    {
+      "epoch": 0.7468529196783887,
+      "grad_norm": 4.700711476652391,
+      "learning_rate": 7.943068399865111e-07,
+      "loss": 0.3624,
+      "step": 9196
+    },
+    {
+      "epoch": 0.7469341346544303,
+      "grad_norm": 4.764991430944277,
+      "learning_rate": 7.93826115693884e-07,
+      "loss": 0.3774,
+      "step": 9197
+    },
+    {
+      "epoch": 0.7470153496304719,
+      "grad_norm": 7.85688314418302,
+      "learning_rate": 7.933455094594602e-07,
+      "loss": 0.3546,
+      "step": 9198
+    },
+    {
+      "epoch": 0.7470965646065134,
+      "grad_norm": 4.120541672766695,
+      "learning_rate": 7.928650213164945e-07,
+      "loss": 0.6123,
+      "step": 9199
+    },
+    {
+      "epoch": 0.747177779582555,
+      "grad_norm": 7.327958848564803,
+      "learning_rate": 7.92384651298235e-07,
+      "loss": 0.5695,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7472589945585966,
+      "grad_norm": 5.719769810682618,
+      "learning_rate": 7.919043994379194e-07,
+      "loss": 0.4224,
+      "step": 9201
+    },
+    {
+      "epoch": 0.7473402095346382,
+      "grad_norm": 4.094688049104328,
+      "learning_rate": 7.914242657687804e-07,
+      "loss": 0.5687,
+      "step": 9202
+    },
+    {
+      "epoch": 0.7474214245106797,
+      "grad_norm": 8.405367942793522,
+      "learning_rate": 7.909442503240395e-07,
+      "loss": 0.429,
+      "step": 9203
+    },
+    {
+      "epoch": 0.7475026394867214,
+      "grad_norm": 6.01950206469752,
+      "learning_rate": 7.904643531369108e-07,
+      "loss": 0.4256,
+      "step": 9204
+    },
+    {
+      "epoch": 0.747583854462763,
+      "grad_norm": 7.930996919339839,
+      "learning_rate": 7.899845742406017e-07,
+      "loss": 0.4707,
+      "step": 9205
+    },
+    {
+      "epoch": 0.7476650694388045,
+      "grad_norm": 16.262367161009937,
+      "learning_rate": 7.895049136683095e-07,
+      "loss": 0.411,
+      "step": 9206
+    },
+    {
+      "epoch": 0.7477462844148461,
+      "grad_norm": 3.3605717819153527,
+      "learning_rate": 7.890253714532245e-07,
+      "loss": 0.4796,
+      "step": 9207
+    },
+    {
+      "epoch": 0.7478274993908877,
+      "grad_norm": 4.04997390894807,
+      "learning_rate": 7.885459476285292e-07,
+      "loss": 0.5579,
+      "step": 9208
+    },
+    {
+      "epoch": 0.7479087143669293,
+      "grad_norm": 4.435766819763639,
+      "learning_rate": 7.880666422273969e-07,
+      "loss": 0.3457,
+      "step": 9209
+    },
+    {
+      "epoch": 0.7479899293429708,
+      "grad_norm": 5.1365435166693505,
+      "learning_rate": 7.875874552829918e-07,
+      "loss": 0.5797,
+      "step": 9210
+    },
+    {
+      "epoch": 0.7480711443190124,
+      "grad_norm": 5.591130735882937,
+      "learning_rate": 7.871083868284726e-07,
+      "loss": 0.5462,
+      "step": 9211
+    },
+    {
+      "epoch": 0.748152359295054,
+      "grad_norm": 8.16910174430336,
+      "learning_rate": 7.866294368969871e-07,
+      "loss": 0.4643,
+      "step": 9212
+    },
+    {
+      "epoch": 0.7482335742710956,
+      "grad_norm": 7.038360149078053,
+      "learning_rate": 7.861506055216764e-07,
+      "loss": 0.5222,
+      "step": 9213
+    },
+    {
+      "epoch": 0.7483147892471371,
+      "grad_norm": 5.659068513220842,
+      "learning_rate": 7.856718927356743e-07,
+      "loss": 0.5721,
+      "step": 9214
+    },
+    {
+      "epoch": 0.7483960042231788,
+      "grad_norm": 7.479766446282722,
+      "learning_rate": 7.851932985721042e-07,
+      "loss": 0.5484,
+      "step": 9215
+    },
+    {
+      "epoch": 0.7484772191992204,
+      "grad_norm": 7.0753754777583575,
+      "learning_rate": 7.847148230640825e-07,
+      "loss": 0.3887,
+      "step": 9216
+    },
+    {
+      "epoch": 0.7485584341752619,
+      "grad_norm": 6.933072652515654,
+      "learning_rate": 7.842364662447161e-07,
+      "loss": 0.4933,
+      "step": 9217
+    },
+    {
+      "epoch": 0.7486396491513035,
+      "grad_norm": 4.082184303346161,
+      "learning_rate": 7.837582281471065e-07,
+      "loss": 0.4314,
+      "step": 9218
+    },
+    {
+      "epoch": 0.7487208641273451,
+      "grad_norm": 5.120311127792642,
+      "learning_rate": 7.832801088043438e-07,
+      "loss": 0.3486,
+      "step": 9219
+    },
+    {
+      "epoch": 0.7488020791033867,
+      "grad_norm": 5.389253627233221,
+      "learning_rate": 7.828021082495118e-07,
+      "loss": 0.5207,
+      "step": 9220
+    },
+    {
+      "epoch": 0.7488832940794282,
+      "grad_norm": 5.865053171467202,
+      "learning_rate": 7.823242265156866e-07,
+      "loss": 0.3661,
+      "step": 9221
+    },
+    {
+      "epoch": 0.7489645090554699,
+      "grad_norm": 4.518682135966551,
+      "learning_rate": 7.818464636359344e-07,
+      "loss": 0.6257,
+      "step": 9222
+    },
+    {
+      "epoch": 0.7490457240315114,
+      "grad_norm": 5.19395467966571,
+      "learning_rate": 7.813688196433125e-07,
+      "loss": 0.4796,
+      "step": 9223
+    },
+    {
+      "epoch": 0.749126939007553,
+      "grad_norm": 4.850339059401898,
+      "learning_rate": 7.808912945708738e-07,
+      "loss": 0.4585,
+      "step": 9224
+    },
+    {
+      "epoch": 0.7492081539835945,
+      "grad_norm": 5.013059286403891,
+      "learning_rate": 7.804138884516583e-07,
+      "loss": 0.3884,
+      "step": 9225
+    },
+    {
+      "epoch": 0.7492893689596362,
+      "grad_norm": 4.841130061593439,
+      "learning_rate": 7.799366013187007e-07,
+      "loss": 0.4484,
+      "step": 9226
+    },
+    {
+      "epoch": 0.7493705839356778,
+      "grad_norm": 5.230772273048009,
+      "learning_rate": 7.794594332050282e-07,
+      "loss": 0.4274,
+      "step": 9227
+    },
+    {
+      "epoch": 0.7494517989117193,
+      "grad_norm": 7.075947384663193,
+      "learning_rate": 7.789823841436567e-07,
+      "loss": 0.4516,
+      "step": 9228
+    },
+    {
+      "epoch": 0.749533013887761,
+      "grad_norm": 6.992483714956555,
+      "learning_rate": 7.785054541675954e-07,
+      "loss": 0.3975,
+      "step": 9229
+    },
+    {
+      "epoch": 0.7496142288638025,
+      "grad_norm": 4.561345669683036,
+      "learning_rate": 7.780286433098464e-07,
+      "loss": 0.7948,
+      "step": 9230
+    },
+    {
+      "epoch": 0.7496954438398441,
+      "grad_norm": 4.478198203586605,
+      "learning_rate": 7.775519516034019e-07,
+      "loss": 0.4752,
+      "step": 9231
+    },
+    {
+      "epoch": 0.7497766588158856,
+      "grad_norm": 3.9360754902947894,
+      "learning_rate": 7.770753790812455e-07,
+      "loss": 0.5631,
+      "step": 9232
+    },
+    {
+      "epoch": 0.7498578737919273,
+      "grad_norm": 6.792888436524307,
+      "learning_rate": 7.765989257763545e-07,
+      "loss": 0.4384,
+      "step": 9233
+    },
+    {
+      "epoch": 0.7499390887679688,
+      "grad_norm": 5.737542594203367,
+      "learning_rate": 7.761225917216978e-07,
+      "loss": 0.4428,
+      "step": 9234
+    },
+    {
+      "epoch": 0.7500203037440104,
+      "grad_norm": 10.689284036969283,
+      "learning_rate": 7.75646376950234e-07,
+      "loss": 0.4604,
+      "step": 9235
+    },
+    {
+      "epoch": 0.7501015187200519,
+      "grad_norm": 22.015199504327047,
+      "learning_rate": 7.751702814949145e-07,
+      "loss": 0.5013,
+      "step": 9236
+    },
+    {
+      "epoch": 0.7501827336960936,
+      "grad_norm": 9.731543225761502,
+      "learning_rate": 7.746943053886835e-07,
+      "loss": 0.4218,
+      "step": 9237
+    },
+    {
+      "epoch": 0.7502639486721352,
+      "grad_norm": 6.82059578534053,
+      "learning_rate": 7.742184486644746e-07,
+      "loss": 0.5271,
+      "step": 9238
+    },
+    {
+      "epoch": 0.7503451636481767,
+      "grad_norm": 4.580942490166404,
+      "learning_rate": 7.737427113552157e-07,
+      "loss": 0.5683,
+      "step": 9239
+    },
+    {
+      "epoch": 0.7504263786242183,
+      "grad_norm": 3.654932263444829,
+      "learning_rate": 7.732670934938257e-07,
+      "loss": 0.4838,
+      "step": 9240
+    },
+    {
+      "epoch": 0.7505075936002599,
+      "grad_norm": 5.579207786127818,
+      "learning_rate": 7.727915951132145e-07,
+      "loss": 0.478,
+      "step": 9241
+    },
+    {
+      "epoch": 0.7505888085763015,
+      "grad_norm": 4.861880759749431,
+      "learning_rate": 7.723162162462827e-07,
+      "loss": 0.5107,
+      "step": 9242
+    },
+    {
+      "epoch": 0.750670023552343,
+      "grad_norm": 3.336615456803476,
+      "learning_rate": 7.718409569259261e-07,
+      "loss": 0.439,
+      "step": 9243
+    },
+    {
+      "epoch": 0.7507512385283847,
+      "grad_norm": 3.594363219165978,
+      "learning_rate": 7.713658171850289e-07,
+      "loss": 0.5717,
+      "step": 9244
+    },
+    {
+      "epoch": 0.7508324535044262,
+      "grad_norm": 6.352249538029256,
+      "learning_rate": 7.708907970564672e-07,
+      "loss": 0.4124,
+      "step": 9245
+    },
+    {
+      "epoch": 0.7509136684804678,
+      "grad_norm": 5.815211228408834,
+      "learning_rate": 7.704158965731126e-07,
+      "loss": 0.6313,
+      "step": 9246
+    },
+    {
+      "epoch": 0.7509948834565093,
+      "grad_norm": 4.251212567683483,
+      "learning_rate": 7.699411157678241e-07,
+      "loss": 0.4166,
+      "step": 9247
+    },
+    {
+      "epoch": 0.751076098432551,
+      "grad_norm": 4.650261805037805,
+      "learning_rate": 7.694664546734534e-07,
+      "loss": 0.5619,
+      "step": 9248
+    },
+    {
+      "epoch": 0.7511573134085926,
+      "grad_norm": 7.169544455959594,
+      "learning_rate": 7.689919133228462e-07,
+      "loss": 0.3819,
+      "step": 9249
+    },
+    {
+      "epoch": 0.7512385283846341,
+      "grad_norm": 5.930269929305272,
+      "learning_rate": 7.685174917488375e-07,
+      "loss": 0.45,
+      "step": 9250
+    },
+    {
+      "epoch": 0.7513197433606758,
+      "grad_norm": 4.069763452953068,
+      "learning_rate": 7.680431899842538e-07,
+      "loss": 0.4904,
+      "step": 9251
+    },
+    {
+      "epoch": 0.7514009583367173,
+      "grad_norm": 6.375436578366889,
+      "learning_rate": 7.67569008061915e-07,
+      "loss": 0.42,
+      "step": 9252
+    },
+    {
+      "epoch": 0.7514821733127589,
+      "grad_norm": 4.507854006209749,
+      "learning_rate": 7.670949460146329e-07,
+      "loss": 0.5423,
+      "step": 9253
+    },
+    {
+      "epoch": 0.7515633882888004,
+      "grad_norm": 4.6919308246764615,
+      "learning_rate": 7.666210038752092e-07,
+      "loss": 0.4196,
+      "step": 9254
+    },
+    {
+      "epoch": 0.7516446032648421,
+      "grad_norm": 7.409437183391873,
+      "learning_rate": 7.661471816764377e-07,
+      "loss": 0.5427,
+      "step": 9255
+    },
+    {
+      "epoch": 0.7517258182408836,
+      "grad_norm": 3.4419438252999446,
+      "learning_rate": 7.656734794511056e-07,
+      "loss": 0.4405,
+      "step": 9256
+    },
+    {
+      "epoch": 0.7518070332169252,
+      "grad_norm": 12.815965237950453,
+      "learning_rate": 7.65199897231989e-07,
+      "loss": 0.4739,
+      "step": 9257
+    },
+    {
+      "epoch": 0.7518882481929667,
+      "grad_norm": 6.654081291463507,
+      "learning_rate": 7.647264350518582e-07,
+      "loss": 0.4545,
+      "step": 9258
+    },
+    {
+      "epoch": 0.7519694631690084,
+      "grad_norm": 4.719867564770338,
+      "learning_rate": 7.642530929434752e-07,
+      "loss": 0.4012,
+      "step": 9259
+    },
+    {
+      "epoch": 0.75205067814505,
+      "grad_norm": 5.37752756201268,
+      "learning_rate": 7.637798709395919e-07,
+      "loss": 0.5197,
+      "step": 9260
+    },
+    {
+      "epoch": 0.7521318931210915,
+      "grad_norm": 5.47385742990871,
+      "learning_rate": 7.633067690729517e-07,
+      "loss": 0.367,
+      "step": 9261
+    },
+    {
+      "epoch": 0.7522131080971332,
+      "grad_norm": 4.4733368861841045,
+      "learning_rate": 7.628337873762928e-07,
+      "loss": 0.4252,
+      "step": 9262
+    },
+    {
+      "epoch": 0.7522943230731747,
+      "grad_norm": 6.510631608383568,
+      "learning_rate": 7.62360925882342e-07,
+      "loss": 0.4848,
+      "step": 9263
+    },
+    {
+      "epoch": 0.7523755380492163,
+      "grad_norm": 8.2985790641125,
+      "learning_rate": 7.618881846238177e-07,
+      "loss": 0.3856,
+      "step": 9264
+    },
+    {
+      "epoch": 0.7524567530252578,
+      "grad_norm": 8.29108898122942,
+      "learning_rate": 7.614155636334325e-07,
+      "loss": 0.284,
+      "step": 9265
+    },
+    {
+      "epoch": 0.7525379680012995,
+      "grad_norm": 14.447143227318355,
+      "learning_rate": 7.609430629438896e-07,
+      "loss": 0.4909,
+      "step": 9266
+    },
+    {
+      "epoch": 0.752619182977341,
+      "grad_norm": 9.786187786821417,
+      "learning_rate": 7.604706825878822e-07,
+      "loss": 0.5262,
+      "step": 9267
+    },
+    {
+      "epoch": 0.7527003979533826,
+      "grad_norm": 6.314038643222093,
+      "learning_rate": 7.59998422598098e-07,
+      "loss": 0.4241,
+      "step": 9268
+    },
+    {
+      "epoch": 0.7527816129294241,
+      "grad_norm": 3.701696525893301,
+      "learning_rate": 7.595262830072142e-07,
+      "loss": 0.4264,
+      "step": 9269
+    },
+    {
+      "epoch": 0.7528628279054658,
+      "grad_norm": 6.147998028522049,
+      "learning_rate": 7.590542638478992e-07,
+      "loss": 0.5874,
+      "step": 9270
+    },
+    {
+      "epoch": 0.7529440428815074,
+      "grad_norm": 5.173315947912089,
+      "learning_rate": 7.585823651528157e-07,
+      "loss": 0.5389,
+      "step": 9271
+    },
+    {
+      "epoch": 0.7530252578575489,
+      "grad_norm": 5.039146817437555,
+      "learning_rate": 7.581105869546168e-07,
+      "loss": 0.3099,
+      "step": 9272
+    },
+    {
+      "epoch": 0.7531064728335906,
+      "grad_norm": 6.7461838507335505,
+      "learning_rate": 7.576389292859465e-07,
+      "loss": 0.5686,
+      "step": 9273
+    },
+    {
+      "epoch": 0.7531876878096321,
+      "grad_norm": 4.776535888289513,
+      "learning_rate": 7.5716739217944e-07,
+      "loss": 0.5206,
+      "step": 9274
+    },
+    {
+      "epoch": 0.7532689027856737,
+      "grad_norm": 5.586476353122101,
+      "learning_rate": 7.566959756677272e-07,
+      "loss": 0.4071,
+      "step": 9275
+    },
+    {
+      "epoch": 0.7533501177617152,
+      "grad_norm": 6.574615954432533,
+      "learning_rate": 7.562246797834266e-07,
+      "loss": 0.4455,
+      "step": 9276
+    },
+    {
+      "epoch": 0.7534313327377569,
+      "grad_norm": 3.8731896502314855,
+      "learning_rate": 7.557535045591485e-07,
+      "loss": 0.5776,
+      "step": 9277
+    },
+    {
+      "epoch": 0.7535125477137984,
+      "grad_norm": 4.1035501788561906,
+      "learning_rate": 7.552824500274963e-07,
+      "loss": 0.4607,
+      "step": 9278
+    },
+    {
+      "epoch": 0.75359376268984,
+      "grad_norm": 3.37747957490476,
+      "learning_rate": 7.548115162210659e-07,
+      "loss": 0.5089,
+      "step": 9279
+    },
+    {
+      "epoch": 0.7536749776658815,
+      "grad_norm": 5.307633133304405,
+      "learning_rate": 7.543407031724415e-07,
+      "loss": 0.7355,
+      "step": 9280
+    },
+    {
+      "epoch": 0.7537561926419232,
+      "grad_norm": 5.0482613355790695,
+      "learning_rate": 7.538700109142022e-07,
+      "loss": 0.5037,
+      "step": 9281
+    },
+    {
+      "epoch": 0.7538374076179648,
+      "grad_norm": 5.458953154438223,
+      "learning_rate": 7.533994394789171e-07,
+      "loss": 0.4972,
+      "step": 9282
+    },
+    {
+      "epoch": 0.7539186225940063,
+      "grad_norm": 8.54249426936385,
+      "learning_rate": 7.529289888991462e-07,
+      "loss": 0.5581,
+      "step": 9283
+    },
+    {
+      "epoch": 0.753999837570048,
+      "grad_norm": 7.083450524943488,
+      "learning_rate": 7.524586592074432e-07,
+      "loss": 0.4335,
+      "step": 9284
+    },
+    {
+      "epoch": 0.7540810525460895,
+      "grad_norm": 3.984907663593172,
+      "learning_rate": 7.519884504363525e-07,
+      "loss": 0.5737,
+      "step": 9285
+    },
+    {
+      "epoch": 0.7541622675221311,
+      "grad_norm": 11.241450055916163,
+      "learning_rate": 7.515183626184095e-07,
+      "loss": 0.4297,
+      "step": 9286
+    },
+    {
+      "epoch": 0.7542434824981726,
+      "grad_norm": 5.7339323931383035,
+      "learning_rate": 7.510483957861428e-07,
+      "loss": 0.5938,
+      "step": 9287
+    },
+    {
+      "epoch": 0.7543246974742143,
+      "grad_norm": 4.59002005480556,
+      "learning_rate": 7.505785499720708e-07,
+      "loss": 0.3925,
+      "step": 9288
+    },
+    {
+      "epoch": 0.7544059124502558,
+      "grad_norm": 4.295589622295897,
+      "learning_rate": 7.501088252087046e-07,
+      "loss": 0.5777,
+      "step": 9289
+    },
+    {
+      "epoch": 0.7544871274262974,
+      "grad_norm": 6.677871787585794,
+      "learning_rate": 7.496392215285456e-07,
+      "loss": 0.5826,
+      "step": 9290
+    },
+    {
+      "epoch": 0.7545683424023389,
+      "grad_norm": 6.4999364694050366,
+      "learning_rate": 7.49169738964089e-07,
+      "loss": 0.6462,
+      "step": 9291
+    },
+    {
+      "epoch": 0.7546495573783806,
+      "grad_norm": 4.2437533825373706,
+      "learning_rate": 7.487003775478208e-07,
+      "loss": 0.4204,
+      "step": 9292
+    },
+    {
+      "epoch": 0.7547307723544222,
+      "grad_norm": 6.9886624820729475,
+      "learning_rate": 7.482311373122173e-07,
+      "loss": 0.4936,
+      "step": 9293
+    },
+    {
+      "epoch": 0.7548119873304637,
+      "grad_norm": 9.509767514085462,
+      "learning_rate": 7.477620182897485e-07,
+      "loss": 0.5231,
+      "step": 9294
+    },
+    {
+      "epoch": 0.7548932023065054,
+      "grad_norm": 6.97430223457955,
+      "learning_rate": 7.472930205128748e-07,
+      "loss": 0.614,
+      "step": 9295
+    },
+    {
+      "epoch": 0.7549744172825469,
+      "grad_norm": 6.802220975306286,
+      "learning_rate": 7.46824144014047e-07,
+      "loss": 0.4198,
+      "step": 9296
+    },
+    {
+      "epoch": 0.7550556322585885,
+      "grad_norm": 5.419177843717326,
+      "learning_rate": 7.4635538882571e-07,
+      "loss": 0.5168,
+      "step": 9297
+    },
+    {
+      "epoch": 0.75513684723463,
+      "grad_norm": 4.5756834213715365,
+      "learning_rate": 7.458867549802998e-07,
+      "loss": 0.4739,
+      "step": 9298
+    },
+    {
+      "epoch": 0.7552180622106717,
+      "grad_norm": 10.932740692189409,
+      "learning_rate": 7.454182425102418e-07,
+      "loss": 0.3756,
+      "step": 9299
+    },
+    {
+      "epoch": 0.7552992771867132,
+      "grad_norm": 7.458052877323758,
+      "learning_rate": 7.449498514479564e-07,
+      "loss": 0.4295,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7553804921627548,
+      "grad_norm": 7.572226076657359,
+      "learning_rate": 7.444815818258527e-07,
+      "loss": 0.5706,
+      "step": 9301
+    },
+    {
+      "epoch": 0.7554617071387963,
+      "grad_norm": 7.895090561106518,
+      "learning_rate": 7.440134336763316e-07,
+      "loss": 0.7025,
+      "step": 9302
+    },
+    {
+      "epoch": 0.755542922114838,
+      "grad_norm": 7.629608085828636,
+      "learning_rate": 7.435454070317885e-07,
+      "loss": 0.4952,
+      "step": 9303
+    },
+    {
+      "epoch": 0.7556241370908796,
+      "grad_norm": 5.105832653252446,
+      "learning_rate": 7.430775019246064e-07,
+      "loss": 0.3872,
+      "step": 9304
+    },
+    {
+      "epoch": 0.7557053520669211,
+      "grad_norm": 5.491484533703616,
+      "learning_rate": 7.426097183871636e-07,
+      "loss": 0.3978,
+      "step": 9305
+    },
+    {
+      "epoch": 0.7557865670429628,
+      "grad_norm": 6.946802048002631,
+      "learning_rate": 7.421420564518267e-07,
+      "loss": 0.6585,
+      "step": 9306
+    },
+    {
+      "epoch": 0.7558677820190043,
+      "grad_norm": 3.9872313781539206,
+      "learning_rate": 7.41674516150957e-07,
+      "loss": 0.6047,
+      "step": 9307
+    },
+    {
+      "epoch": 0.7559489969950459,
+      "grad_norm": 7.213117113503657,
+      "learning_rate": 7.412070975169047e-07,
+      "loss": 0.583,
+      "step": 9308
+    },
+    {
+      "epoch": 0.7560302119710874,
+      "grad_norm": 5.24742138382295,
+      "learning_rate": 7.407398005820123e-07,
+      "loss": 0.4034,
+      "step": 9309
+    },
+    {
+      "epoch": 0.7561114269471291,
+      "grad_norm": 4.211946251001222,
+      "learning_rate": 7.402726253786152e-07,
+      "loss": 0.4058,
+      "step": 9310
+    },
+    {
+      "epoch": 0.7561926419231706,
+      "grad_norm": 4.415557038084326,
+      "learning_rate": 7.398055719390399e-07,
+      "loss": 0.5176,
+      "step": 9311
+    },
+    {
+      "epoch": 0.7562738568992122,
+      "grad_norm": 4.229473804773957,
+      "learning_rate": 7.39338640295602e-07,
+      "loss": 0.533,
+      "step": 9312
+    },
+    {
+      "epoch": 0.7563550718752537,
+      "grad_norm": 6.390722780847675,
+      "learning_rate": 7.388718304806133e-07,
+      "loss": 0.4943,
+      "step": 9313
+    },
+    {
+      "epoch": 0.7564362868512954,
+      "grad_norm": 3.9879443116206224,
+      "learning_rate": 7.384051425263733e-07,
+      "loss": 0.6664,
+      "step": 9314
+    },
+    {
+      "epoch": 0.756517501827337,
+      "grad_norm": 4.396256081895415,
+      "learning_rate": 7.379385764651737e-07,
+      "loss": 0.3858,
+      "step": 9315
+    },
+    {
+      "epoch": 0.7565987168033785,
+      "grad_norm": 9.650686792105313,
+      "learning_rate": 7.374721323292985e-07,
+      "loss": 0.6082,
+      "step": 9316
+    },
+    {
+      "epoch": 0.7566799317794202,
+      "grad_norm": 4.52159603222765,
+      "learning_rate": 7.370058101510249e-07,
+      "loss": 0.4969,
+      "step": 9317
+    },
+    {
+      "epoch": 0.7567611467554617,
+      "grad_norm": 4.592670810869141,
+      "learning_rate": 7.365396099626176e-07,
+      "loss": 0.6284,
+      "step": 9318
+    },
+    {
+      "epoch": 0.7568423617315033,
+      "grad_norm": 14.987779633029737,
+      "learning_rate": 7.360735317963374e-07,
+      "loss": 0.4282,
+      "step": 9319
+    },
+    {
+      "epoch": 0.7569235767075448,
+      "grad_norm": 5.108452215084867,
+      "learning_rate": 7.356075756844333e-07,
+      "loss": 0.5101,
+      "step": 9320
+    },
+    {
+      "epoch": 0.7570047916835865,
+      "grad_norm": 6.336779399554521,
+      "learning_rate": 7.351417416591461e-07,
+      "loss": 0.4218,
+      "step": 9321
+    },
+    {
+      "epoch": 0.757086006659628,
+      "grad_norm": 3.2420955010959633,
+      "learning_rate": 7.346760297527109e-07,
+      "loss": 0.4596,
+      "step": 9322
+    },
+    {
+      "epoch": 0.7571672216356696,
+      "grad_norm": 14.31442060248585,
+      "learning_rate": 7.342104399973507e-07,
+      "loss": 0.6222,
+      "step": 9323
+    },
+    {
+      "epoch": 0.7572484366117112,
+      "grad_norm": 4.841998000446461,
+      "learning_rate": 7.337449724252837e-07,
+      "loss": 0.3906,
+      "step": 9324
+    },
+    {
+      "epoch": 0.7573296515877528,
+      "grad_norm": 6.189602667818374,
+      "learning_rate": 7.332796270687159e-07,
+      "loss": 0.5206,
+      "step": 9325
+    },
+    {
+      "epoch": 0.7574108665637944,
+      "grad_norm": 3.4330814806419747,
+      "learning_rate": 7.328144039598487e-07,
+      "loss": 0.4608,
+      "step": 9326
+    },
+    {
+      "epoch": 0.7574920815398359,
+      "grad_norm": 8.246364139369895,
+      "learning_rate": 7.323493031308718e-07,
+      "loss": 0.3708,
+      "step": 9327
+    },
+    {
+      "epoch": 0.7575732965158776,
+      "grad_norm": 11.977800366186257,
+      "learning_rate": 7.318843246139673e-07,
+      "loss": 0.4251,
+      "step": 9328
+    },
+    {
+      "epoch": 0.7576545114919191,
+      "grad_norm": 4.480080892757191,
+      "learning_rate": 7.314194684413098e-07,
+      "loss": 0.5398,
+      "step": 9329
+    },
+    {
+      "epoch": 0.7577357264679607,
+      "grad_norm": 5.761110269898293,
+      "learning_rate": 7.309547346450658e-07,
+      "loss": 0.486,
+      "step": 9330
+    },
+    {
+      "epoch": 0.7578169414440022,
+      "grad_norm": 4.670162793343675,
+      "learning_rate": 7.304901232573908e-07,
+      "loss": 0.4607,
+      "step": 9331
+    },
+    {
+      "epoch": 0.7578981564200439,
+      "grad_norm": 7.987770600523061,
+      "learning_rate": 7.300256343104351e-07,
+      "loss": 0.4614,
+      "step": 9332
+    },
+    {
+      "epoch": 0.7579793713960854,
+      "grad_norm": 5.479915469263117,
+      "learning_rate": 7.295612678363382e-07,
+      "loss": 0.3639,
+      "step": 9333
+    },
+    {
+      "epoch": 0.758060586372127,
+      "grad_norm": 25.580608798332342,
+      "learning_rate": 7.290970238672307e-07,
+      "loss": 0.3886,
+      "step": 9334
+    },
+    {
+      "epoch": 0.7581418013481686,
+      "grad_norm": 3.80358687651955,
+      "learning_rate": 7.286329024352376e-07,
+      "loss": 0.5083,
+      "step": 9335
+    },
+    {
+      "epoch": 0.7582230163242102,
+      "grad_norm": 3.6188064823034747,
+      "learning_rate": 7.281689035724718e-07,
+      "loss": 0.4203,
+      "step": 9336
+    },
+    {
+      "epoch": 0.7583042313002518,
+      "grad_norm": 6.662801100557623,
+      "learning_rate": 7.277050273110408e-07,
+      "loss": 0.428,
+      "step": 9337
+    },
+    {
+      "epoch": 0.7583854462762933,
+      "grad_norm": 4.198195373303361,
+      "learning_rate": 7.272412736830431e-07,
+      "loss": 0.5499,
+      "step": 9338
+    },
+    {
+      "epoch": 0.758466661252335,
+      "grad_norm": 7.26119245955303,
+      "learning_rate": 7.26777642720567e-07,
+      "loss": 0.5308,
+      "step": 9339
+    },
+    {
+      "epoch": 0.7585478762283765,
+      "grad_norm": 6.6134602398961295,
+      "learning_rate": 7.263141344556924e-07,
+      "loss": 0.4514,
+      "step": 9340
+    },
+    {
+      "epoch": 0.7586290912044181,
+      "grad_norm": 3.9208951218117427,
+      "learning_rate": 7.258507489204935e-07,
+      "loss": 0.3762,
+      "step": 9341
+    },
+    {
+      "epoch": 0.7587103061804596,
+      "grad_norm": 7.76615893471226,
+      "learning_rate": 7.253874861470325e-07,
+      "loss": 0.5326,
+      "step": 9342
+    },
+    {
+      "epoch": 0.7587915211565013,
+      "grad_norm": 7.24244627028924,
+      "learning_rate": 7.24924346167366e-07,
+      "loss": 0.4045,
+      "step": 9343
+    },
+    {
+      "epoch": 0.7588727361325428,
+      "grad_norm": 3.373479639826732,
+      "learning_rate": 7.244613290135396e-07,
+      "loss": 0.4882,
+      "step": 9344
+    },
+    {
+      "epoch": 0.7589539511085844,
+      "grad_norm": 6.628941975131967,
+      "learning_rate": 7.239984347175932e-07,
+      "loss": 0.3442,
+      "step": 9345
+    },
+    {
+      "epoch": 0.759035166084626,
+      "grad_norm": 3.784372823794077,
+      "learning_rate": 7.235356633115559e-07,
+      "loss": 0.5905,
+      "step": 9346
+    },
+    {
+      "epoch": 0.7591163810606676,
+      "grad_norm": 4.27156618640782,
+      "learning_rate": 7.230730148274478e-07,
+      "loss": 0.3702,
+      "step": 9347
+    },
+    {
+      "epoch": 0.7591975960367092,
+      "grad_norm": 3.9436102310547327,
+      "learning_rate": 7.226104892972838e-07,
+      "loss": 0.3701,
+      "step": 9348
+    },
+    {
+      "epoch": 0.7592788110127507,
+      "grad_norm": 3.9798514443021196,
+      "learning_rate": 7.221480867530664e-07,
+      "loss": 0.4805,
+      "step": 9349
+    },
+    {
+      "epoch": 0.7593600259887924,
+      "grad_norm": 3.87348251530302,
+      "learning_rate": 7.216858072267924e-07,
+      "loss": 0.5473,
+      "step": 9350
+    },
+    {
+      "epoch": 0.7594412409648339,
+      "grad_norm": 5.915769976523819,
+      "learning_rate": 7.212236507504494e-07,
+      "loss": 0.5054,
+      "step": 9351
+    },
+    {
+      "epoch": 0.7595224559408755,
+      "grad_norm": 3.8732533424148605,
+      "learning_rate": 7.207616173560158e-07,
+      "loss": 0.4806,
+      "step": 9352
+    },
+    {
+      "epoch": 0.759603670916917,
+      "grad_norm": 6.11681123799206,
+      "learning_rate": 7.202997070754613e-07,
+      "loss": 0.46,
+      "step": 9353
+    },
+    {
+      "epoch": 0.7596848858929587,
+      "grad_norm": 8.365104395894681,
+      "learning_rate": 7.198379199407488e-07,
+      "loss": 0.5624,
+      "step": 9354
+    },
+    {
+      "epoch": 0.7597661008690002,
+      "grad_norm": 6.4866744161399685,
+      "learning_rate": 7.193762559838299e-07,
+      "loss": 0.5142,
+      "step": 9355
+    },
+    {
+      "epoch": 0.7598473158450418,
+      "grad_norm": 6.4484042472139675,
+      "learning_rate": 7.189147152366504e-07,
+      "loss": 0.3683,
+      "step": 9356
+    },
+    {
+      "epoch": 0.7599285308210834,
+      "grad_norm": 4.5783151146826935,
+      "learning_rate": 7.184532977311471e-07,
+      "loss": 0.4517,
+      "step": 9357
+    },
+    {
+      "epoch": 0.760009745797125,
+      "grad_norm": 4.142951500808283,
+      "learning_rate": 7.179920034992469e-07,
+      "loss": 0.4783,
+      "step": 9358
+    },
+    {
+      "epoch": 0.7600909607731666,
+      "grad_norm": 3.327558136006068,
+      "learning_rate": 7.175308325728689e-07,
+      "loss": 0.6403,
+      "step": 9359
+    },
+    {
+      "epoch": 0.7601721757492081,
+      "grad_norm": 7.732587779509069,
+      "learning_rate": 7.170697849839229e-07,
+      "loss": 0.4863,
+      "step": 9360
+    },
+    {
+      "epoch": 0.7602533907252498,
+      "grad_norm": 3.4261727551662866,
+      "learning_rate": 7.166088607643123e-07,
+      "loss": 0.4994,
+      "step": 9361
+    },
+    {
+      "epoch": 0.7603346057012913,
+      "grad_norm": 7.420461816144346,
+      "learning_rate": 7.161480599459297e-07,
+      "loss": 0.3715,
+      "step": 9362
+    },
+    {
+      "epoch": 0.7604158206773329,
+      "grad_norm": 4.177142639059444,
+      "learning_rate": 7.156873825606603e-07,
+      "loss": 0.5635,
+      "step": 9363
+    },
+    {
+      "epoch": 0.7604970356533745,
+      "grad_norm": 6.465161714985024,
+      "learning_rate": 7.152268286403813e-07,
+      "loss": 0.4047,
+      "step": 9364
+    },
+    {
+      "epoch": 0.7605782506294161,
+      "grad_norm": 6.173359562536338,
+      "learning_rate": 7.147663982169601e-07,
+      "loss": 0.4028,
+      "step": 9365
+    },
+    {
+      "epoch": 0.7606594656054576,
+      "grad_norm": 3.9455176413170583,
+      "learning_rate": 7.143060913222552e-07,
+      "loss": 0.6058,
+      "step": 9366
+    },
+    {
+      "epoch": 0.7607406805814992,
+      "grad_norm": 10.642099423017335,
+      "learning_rate": 7.138459079881188e-07,
+      "loss": 0.706,
+      "step": 9367
+    },
+    {
+      "epoch": 0.7608218955575408,
+      "grad_norm": 4.19417453033392,
+      "learning_rate": 7.133858482463918e-07,
+      "loss": 0.6098,
+      "step": 9368
+    },
+    {
+      "epoch": 0.7609031105335824,
+      "grad_norm": 4.253649965044964,
+      "learning_rate": 7.129259121289086e-07,
+      "loss": 0.4453,
+      "step": 9369
+    },
+    {
+      "epoch": 0.760984325509624,
+      "grad_norm": 3.668621601717135,
+      "learning_rate": 7.124660996674951e-07,
+      "loss": 0.5733,
+      "step": 9370
+    },
+    {
+      "epoch": 0.7610655404856655,
+      "grad_norm": 6.048356569124348,
+      "learning_rate": 7.12006410893967e-07,
+      "loss": 0.4285,
+      "step": 9371
+    },
+    {
+      "epoch": 0.7611467554617072,
+      "grad_norm": 4.317558436319294,
+      "learning_rate": 7.115468458401317e-07,
+      "loss": 0.6383,
+      "step": 9372
+    },
+    {
+      "epoch": 0.7612279704377487,
+      "grad_norm": 7.296251737717159,
+      "learning_rate": 7.110874045377902e-07,
+      "loss": 0.551,
+      "step": 9373
+    },
+    {
+      "epoch": 0.7613091854137903,
+      "grad_norm": 5.66808374273542,
+      "learning_rate": 7.106280870187326e-07,
+      "loss": 0.4665,
+      "step": 9374
+    },
+    {
+      "epoch": 0.7613904003898319,
+      "grad_norm": 3.897913983532991,
+      "learning_rate": 7.101688933147397e-07,
+      "loss": 0.5898,
+      "step": 9375
+    },
+    {
+      "epoch": 0.7614716153658735,
+      "grad_norm": 5.913213010761257,
+      "learning_rate": 7.097098234575883e-07,
+      "loss": 0.379,
+      "step": 9376
+    },
+    {
+      "epoch": 0.761552830341915,
+      "grad_norm": 8.586921010693551,
+      "learning_rate": 7.092508774790424e-07,
+      "loss": 0.3769,
+      "step": 9377
+    },
+    {
+      "epoch": 0.7616340453179566,
+      "grad_norm": 6.837790290025689,
+      "learning_rate": 7.087920554108582e-07,
+      "loss": 0.5177,
+      "step": 9378
+    },
+    {
+      "epoch": 0.7617152602939982,
+      "grad_norm": 4.809392070744878,
+      "learning_rate": 7.083333572847831e-07,
+      "loss": 0.4528,
+      "step": 9379
+    },
+    {
+      "epoch": 0.7617964752700398,
+      "grad_norm": 3.467857419045606,
+      "learning_rate": 7.078747831325583e-07,
+      "loss": 0.5457,
+      "step": 9380
+    },
+    {
+      "epoch": 0.7618776902460814,
+      "grad_norm": 4.495338193171126,
+      "learning_rate": 7.074163329859129e-07,
+      "loss": 0.3799,
+      "step": 9381
+    },
+    {
+      "epoch": 0.761958905222123,
+      "grad_norm": 4.00025930868917,
+      "learning_rate": 7.069580068765702e-07,
+      "loss": 0.3478,
+      "step": 9382
+    },
+    {
+      "epoch": 0.7620401201981646,
+      "grad_norm": 10.827446952884547,
+      "learning_rate": 7.064998048362448e-07,
+      "loss": 0.4943,
+      "step": 9383
+    },
+    {
+      "epoch": 0.7621213351742061,
+      "grad_norm": 5.064306842835891,
+      "learning_rate": 7.060417268966408e-07,
+      "loss": 0.4528,
+      "step": 9384
+    },
+    {
+      "epoch": 0.7622025501502477,
+      "grad_norm": 4.468211696665271,
+      "learning_rate": 7.055837730894541e-07,
+      "loss": 0.4532,
+      "step": 9385
+    },
+    {
+      "epoch": 0.7622837651262893,
+      "grad_norm": 6.800829039851487,
+      "learning_rate": 7.051259434463745e-07,
+      "loss": 0.347,
+      "step": 9386
+    },
+    {
+      "epoch": 0.7623649801023309,
+      "grad_norm": 6.208378537952931,
+      "learning_rate": 7.046682379990794e-07,
+      "loss": 0.4928,
+      "step": 9387
+    },
+    {
+      "epoch": 0.7624461950783724,
+      "grad_norm": 4.2620464174120425,
+      "learning_rate": 7.042106567792406e-07,
+      "loss": 0.5339,
+      "step": 9388
+    },
+    {
+      "epoch": 0.762527410054414,
+      "grad_norm": 5.564383633638113,
+      "learning_rate": 7.03753199818521e-07,
+      "loss": 0.4415,
+      "step": 9389
+    },
+    {
+      "epoch": 0.7626086250304556,
+      "grad_norm": 6.375756161119263,
+      "learning_rate": 7.032958671485734e-07,
+      "loss": 0.4055,
+      "step": 9390
+    },
+    {
+      "epoch": 0.7626898400064972,
+      "grad_norm": 6.484850680260223,
+      "learning_rate": 7.028386588010421e-07,
+      "loss": 0.5106,
+      "step": 9391
+    },
+    {
+      "epoch": 0.7627710549825388,
+      "grad_norm": 6.127381595989031,
+      "learning_rate": 7.023815748075651e-07,
+      "loss": 0.324,
+      "step": 9392
+    },
+    {
+      "epoch": 0.7628522699585804,
+      "grad_norm": 4.069418782141938,
+      "learning_rate": 7.019246151997694e-07,
+      "loss": 0.4221,
+      "step": 9393
+    },
+    {
+      "epoch": 0.762933484934622,
+      "grad_norm": 3.966681143690813,
+      "learning_rate": 7.014677800092734e-07,
+      "loss": 0.5576,
+      "step": 9394
+    },
+    {
+      "epoch": 0.7630146999106635,
+      "grad_norm": 6.497751818986687,
+      "learning_rate": 7.010110692676886e-07,
+      "loss": 0.4702,
+      "step": 9395
+    },
+    {
+      "epoch": 0.7630959148867051,
+      "grad_norm": 6.139900453454705,
+      "learning_rate": 7.005544830066172e-07,
+      "loss": 0.3902,
+      "step": 9396
+    },
+    {
+      "epoch": 0.7631771298627467,
+      "grad_norm": 5.005470766044911,
+      "learning_rate": 7.000980212576522e-07,
+      "loss": 0.4043,
+      "step": 9397
+    },
+    {
+      "epoch": 0.7632583448387883,
+      "grad_norm": 13.852454716609795,
+      "learning_rate": 6.996416840523776e-07,
+      "loss": 0.5992,
+      "step": 9398
+    },
+    {
+      "epoch": 0.7633395598148298,
+      "grad_norm": 4.487243506708304,
+      "learning_rate": 6.991854714223711e-07,
+      "loss": 0.4657,
+      "step": 9399
+    },
+    {
+      "epoch": 0.7634207747908714,
+      "grad_norm": 5.202920899583729,
+      "learning_rate": 6.987293833991984e-07,
+      "loss": 0.6265,
+      "step": 9400
+    },
+    {
+      "epoch": 0.763501989766913,
+      "grad_norm": 17.066516911339072,
+      "learning_rate": 6.982734200144192e-07,
+      "loss": 0.4925,
+      "step": 9401
+    },
+    {
+      "epoch": 0.7635832047429546,
+      "grad_norm": 3.6601744793382305,
+      "learning_rate": 6.978175812995847e-07,
+      "loss": 0.4923,
+      "step": 9402
+    },
+    {
+      "epoch": 0.7636644197189962,
+      "grad_norm": 6.162261867859467,
+      "learning_rate": 6.973618672862357e-07,
+      "loss": 0.3402,
+      "step": 9403
+    },
+    {
+      "epoch": 0.7637456346950378,
+      "grad_norm": 3.383805533605749,
+      "learning_rate": 6.969062780059041e-07,
+      "loss": 0.4892,
+      "step": 9404
+    },
+    {
+      "epoch": 0.7638268496710794,
+      "grad_norm": 5.034659251576166,
+      "learning_rate": 6.964508134901162e-07,
+      "loss": 0.4553,
+      "step": 9405
+    },
+    {
+      "epoch": 0.7639080646471209,
+      "grad_norm": 4.219393214106526,
+      "learning_rate": 6.959954737703872e-07,
+      "loss": 0.3549,
+      "step": 9406
+    },
+    {
+      "epoch": 0.7639892796231625,
+      "grad_norm": 7.125686847574229,
+      "learning_rate": 6.955402588782229e-07,
+      "loss": 0.4162,
+      "step": 9407
+    },
+    {
+      "epoch": 0.7640704945992041,
+      "grad_norm": 4.00897406634947,
+      "learning_rate": 6.950851688451224e-07,
+      "loss": 0.4606,
+      "step": 9408
+    },
+    {
+      "epoch": 0.7641517095752457,
+      "grad_norm": 5.368532207744499,
+      "learning_rate": 6.94630203702577e-07,
+      "loss": 0.6832,
+      "step": 9409
+    },
+    {
+      "epoch": 0.7642329245512872,
+      "grad_norm": 4.61391210253197,
+      "learning_rate": 6.941753634820658e-07,
+      "loss": 0.4458,
+      "step": 9410
+    },
+    {
+      "epoch": 0.7643141395273289,
+      "grad_norm": 10.571864997786388,
+      "learning_rate": 6.93720648215063e-07,
+      "loss": 0.3812,
+      "step": 9411
+    },
+    {
+      "epoch": 0.7643953545033704,
+      "grad_norm": 6.071699040314471,
+      "learning_rate": 6.932660579330317e-07,
+      "loss": 0.3526,
+      "step": 9412
+    },
+    {
+      "epoch": 0.764476569479412,
+      "grad_norm": 5.121163071457247,
+      "learning_rate": 6.928115926674265e-07,
+      "loss": 0.5468,
+      "step": 9413
+    },
+    {
+      "epoch": 0.7645577844554536,
+      "grad_norm": 4.5789958968091025,
+      "learning_rate": 6.923572524496946e-07,
+      "loss": 0.4688,
+      "step": 9414
+    },
+    {
+      "epoch": 0.7646389994314952,
+      "grad_norm": 4.61496141221473,
+      "learning_rate": 6.919030373112748e-07,
+      "loss": 0.538,
+      "step": 9415
+    },
+    {
+      "epoch": 0.7647202144075368,
+      "grad_norm": 5.530010793903644,
+      "learning_rate": 6.914489472835959e-07,
+      "loss": 0.3921,
+      "step": 9416
+    },
+    {
+      "epoch": 0.7648014293835783,
+      "grad_norm": 4.053580386916616,
+      "learning_rate": 6.909949823980772e-07,
+      "loss": 0.4411,
+      "step": 9417
+    },
+    {
+      "epoch": 0.76488264435962,
+      "grad_norm": 5.361420424547812,
+      "learning_rate": 6.905411426861322e-07,
+      "loss": 0.4674,
+      "step": 9418
+    },
+    {
+      "epoch": 0.7649638593356615,
+      "grad_norm": 6.357458261429573,
+      "learning_rate": 6.900874281791639e-07,
+      "loss": 0.4666,
+      "step": 9419
+    },
+    {
+      "epoch": 0.7650450743117031,
+      "grad_norm": 6.195545553218025,
+      "learning_rate": 6.89633838908566e-07,
+      "loss": 0.4103,
+      "step": 9420
+    },
+    {
+      "epoch": 0.7651262892877446,
+      "grad_norm": 5.87439782429603,
+      "learning_rate": 6.891803749057255e-07,
+      "loss": 0.415,
+      "step": 9421
+    },
+    {
+      "epoch": 0.7652075042637863,
+      "grad_norm": 3.5630192542232497,
+      "learning_rate": 6.887270362020199e-07,
+      "loss": 0.4371,
+      "step": 9422
+    },
+    {
+      "epoch": 0.7652887192398278,
+      "grad_norm": 6.836068046454834,
+      "learning_rate": 6.882738228288166e-07,
+      "loss": 0.3986,
+      "step": 9423
+    },
+    {
+      "epoch": 0.7653699342158694,
+      "grad_norm": 6.446997290383441,
+      "learning_rate": 6.87820734817477e-07,
+      "loss": 0.4896,
+      "step": 9424
+    },
+    {
+      "epoch": 0.765451149191911,
+      "grad_norm": 4.358114968607592,
+      "learning_rate": 6.873677721993518e-07,
+      "loss": 0.514,
+      "step": 9425
+    },
+    {
+      "epoch": 0.7655323641679526,
+      "grad_norm": 8.732219383382438,
+      "learning_rate": 6.86914935005783e-07,
+      "loss": 0.5265,
+      "step": 9426
+    },
+    {
+      "epoch": 0.7656135791439942,
+      "grad_norm": 4.310805179732659,
+      "learning_rate": 6.864622232681048e-07,
+      "loss": 0.4076,
+      "step": 9427
+    },
+    {
+      "epoch": 0.7656947941200357,
+      "grad_norm": 4.045216285940855,
+      "learning_rate": 6.860096370176436e-07,
+      "loss": 0.5051,
+      "step": 9428
+    },
+    {
+      "epoch": 0.7657760090960773,
+      "grad_norm": 7.3004627003203435,
+      "learning_rate": 6.855571762857144e-07,
+      "loss": 0.5137,
+      "step": 9429
+    },
+    {
+      "epoch": 0.7658572240721189,
+      "grad_norm": 4.355642197018315,
+      "learning_rate": 6.851048411036265e-07,
+      "loss": 0.4596,
+      "step": 9430
+    },
+    {
+      "epoch": 0.7659384390481605,
+      "grad_norm": 4.107096289312408,
+      "learning_rate": 6.846526315026783e-07,
+      "loss": 0.5929,
+      "step": 9431
+    },
+    {
+      "epoch": 0.766019654024202,
+      "grad_norm": 4.55847735238035,
+      "learning_rate": 6.842005475141606e-07,
+      "loss": 0.5062,
+      "step": 9432
+    },
+    {
+      "epoch": 0.7661008690002437,
+      "grad_norm": 4.084279279942593,
+      "learning_rate": 6.837485891693541e-07,
+      "loss": 0.5043,
+      "step": 9433
+    },
+    {
+      "epoch": 0.7661820839762852,
+      "grad_norm": 13.22441226336881,
+      "learning_rate": 6.83296756499533e-07,
+      "loss": 0.4094,
+      "step": 9434
+    },
+    {
+      "epoch": 0.7662632989523268,
+      "grad_norm": 4.334505119567936,
+      "learning_rate": 6.828450495359623e-07,
+      "loss": 0.4747,
+      "step": 9435
+    },
+    {
+      "epoch": 0.7663445139283684,
+      "grad_norm": 5.930691768024282,
+      "learning_rate": 6.823934683098963e-07,
+      "loss": 0.6257,
+      "step": 9436
+    },
+    {
+      "epoch": 0.76642572890441,
+      "grad_norm": 3.3704120793280197,
+      "learning_rate": 6.819420128525834e-07,
+      "loss": 0.3181,
+      "step": 9437
+    },
+    {
+      "epoch": 0.7665069438804516,
+      "grad_norm": 3.842942498154423,
+      "learning_rate": 6.814906831952611e-07,
+      "loss": 0.4773,
+      "step": 9438
+    },
+    {
+      "epoch": 0.7665881588564931,
+      "grad_norm": 8.895568429556429,
+      "learning_rate": 6.810394793691585e-07,
+      "loss": 0.3918,
+      "step": 9439
+    },
+    {
+      "epoch": 0.7666693738325348,
+      "grad_norm": 6.8916662279772565,
+      "learning_rate": 6.805884014054975e-07,
+      "loss": 0.4543,
+      "step": 9440
+    },
+    {
+      "epoch": 0.7667505888085763,
+      "grad_norm": 4.932759072758281,
+      "learning_rate": 6.801374493354907e-07,
+      "loss": 0.555,
+      "step": 9441
+    },
+    {
+      "epoch": 0.7668318037846179,
+      "grad_norm": 3.165718589290986,
+      "learning_rate": 6.796866231903402e-07,
+      "loss": 0.5693,
+      "step": 9442
+    },
+    {
+      "epoch": 0.7669130187606594,
+      "grad_norm": 5.661881562952782,
+      "learning_rate": 6.792359230012418e-07,
+      "loss": 0.3955,
+      "step": 9443
+    },
+    {
+      "epoch": 0.7669942337367011,
+      "grad_norm": 5.857023328763498,
+      "learning_rate": 6.787853487993817e-07,
+      "loss": 0.2898,
+      "step": 9444
+    },
+    {
+      "epoch": 0.7670754487127426,
+      "grad_norm": 6.46410699371152,
+      "learning_rate": 6.783349006159359e-07,
+      "loss": 0.326,
+      "step": 9445
+    },
+    {
+      "epoch": 0.7671566636887842,
+      "grad_norm": 6.9335177221214614,
+      "learning_rate": 6.778845784820739e-07,
+      "loss": 0.4844,
+      "step": 9446
+    },
+    {
+      "epoch": 0.7672378786648258,
+      "grad_norm": 5.439529483084923,
+      "learning_rate": 6.774343824289567e-07,
+      "loss": 0.6114,
+      "step": 9447
+    },
+    {
+      "epoch": 0.7673190936408674,
+      "grad_norm": 5.645470096440396,
+      "learning_rate": 6.769843124877343e-07,
+      "loss": 0.5204,
+      "step": 9448
+    },
+    {
+      "epoch": 0.767400308616909,
+      "grad_norm": 24.655262909769625,
+      "learning_rate": 6.765343686895484e-07,
+      "loss": 0.4941,
+      "step": 9449
+    },
+    {
+      "epoch": 0.7674815235929505,
+      "grad_norm": 6.241765051887523,
+      "learning_rate": 6.760845510655345e-07,
+      "loss": 0.3861,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7675627385689922,
+      "grad_norm": 8.035908170373448,
+      "learning_rate": 6.756348596468168e-07,
+      "loss": 0.5332,
+      "step": 9451
+    },
+    {
+      "epoch": 0.7676439535450337,
+      "grad_norm": 4.6392581562003405,
+      "learning_rate": 6.751852944645107e-07,
+      "loss": 0.4359,
+      "step": 9452
+    },
+    {
+      "epoch": 0.7677251685210753,
+      "grad_norm": 6.355510758018541,
+      "learning_rate": 6.747358555497244e-07,
+      "loss": 0.4006,
+      "step": 9453
+    },
+    {
+      "epoch": 0.7678063834971168,
+      "grad_norm": 5.191185254070647,
+      "learning_rate": 6.742865429335576e-07,
+      "loss": 0.6034,
+      "step": 9454
+    },
+    {
+      "epoch": 0.7678875984731585,
+      "grad_norm": 4.336515041136279,
+      "learning_rate": 6.738373566470991e-07,
+      "loss": 0.3565,
+      "step": 9455
+    },
+    {
+      "epoch": 0.7679688134492,
+      "grad_norm": 7.028294785077958,
+      "learning_rate": 6.733882967214312e-07,
+      "loss": 0.4001,
+      "step": 9456
+    },
+    {
+      "epoch": 0.7680500284252416,
+      "grad_norm": 3.345393934309023,
+      "learning_rate": 6.729393631876257e-07,
+      "loss": 0.4984,
+      "step": 9457
+    },
+    {
+      "epoch": 0.7681312434012832,
+      "grad_norm": 4.618093066401069,
+      "learning_rate": 6.724905560767464e-07,
+      "loss": 0.3027,
+      "step": 9458
+    },
+    {
+      "epoch": 0.7682124583773248,
+      "grad_norm": 5.511844154772235,
+      "learning_rate": 6.720418754198485e-07,
+      "loss": 0.3492,
+      "step": 9459
+    },
+    {
+      "epoch": 0.7682936733533664,
+      "grad_norm": 4.512587186638037,
+      "learning_rate": 6.715933212479791e-07,
+      "loss": 0.3126,
+      "step": 9460
+    },
+    {
+      "epoch": 0.7683748883294079,
+      "grad_norm": 3.8768216178706996,
+      "learning_rate": 6.711448935921744e-07,
+      "loss": 0.4618,
+      "step": 9461
+    },
+    {
+      "epoch": 0.7684561033054496,
+      "grad_norm": 4.849694261108834,
+      "learning_rate": 6.706965924834649e-07,
+      "loss": 0.3315,
+      "step": 9462
+    },
+    {
+      "epoch": 0.7685373182814911,
+      "grad_norm": 4.801051258082802,
+      "learning_rate": 6.702484179528699e-07,
+      "loss": 0.5456,
+      "step": 9463
+    },
+    {
+      "epoch": 0.7686185332575327,
+      "grad_norm": 9.352209783795086,
+      "learning_rate": 6.698003700313993e-07,
+      "loss": 0.427,
+      "step": 9464
+    },
+    {
+      "epoch": 0.7686997482335742,
+      "grad_norm": 7.643828934266029,
+      "learning_rate": 6.69352448750058e-07,
+      "loss": 0.7298,
+      "step": 9465
+    },
+    {
+      "epoch": 0.7687809632096159,
+      "grad_norm": 5.286073745430146,
+      "learning_rate": 6.689046541398378e-07,
+      "loss": 0.5492,
+      "step": 9466
+    },
+    {
+      "epoch": 0.7688621781856574,
+      "grad_norm": 7.104403560909407,
+      "learning_rate": 6.684569862317255e-07,
+      "loss": 0.4082,
+      "step": 9467
+    },
+    {
+      "epoch": 0.768943393161699,
+      "grad_norm": 21.80499840324673,
+      "learning_rate": 6.680094450566957e-07,
+      "loss": 0.4536,
+      "step": 9468
+    },
+    {
+      "epoch": 0.7690246081377407,
+      "grad_norm": 9.81351894485181,
+      "learning_rate": 6.675620306457172e-07,
+      "loss": 0.4041,
+      "step": 9469
+    },
+    {
+      "epoch": 0.7691058231137822,
+      "grad_norm": 5.220882981566686,
+      "learning_rate": 6.671147430297481e-07,
+      "loss": 0.4332,
+      "step": 9470
+    },
+    {
+      "epoch": 0.7691870380898238,
+      "grad_norm": 5.429857205093562,
+      "learning_rate": 6.666675822397378e-07,
+      "loss": 0.4061,
+      "step": 9471
+    },
+    {
+      "epoch": 0.7692682530658653,
+      "grad_norm": 8.052028153479819,
+      "learning_rate": 6.662205483066281e-07,
+      "loss": 0.326,
+      "step": 9472
+    },
+    {
+      "epoch": 0.769349468041907,
+      "grad_norm": 4.655903640053554,
+      "learning_rate": 6.65773641261352e-07,
+      "loss": 0.469,
+      "step": 9473
+    },
+    {
+      "epoch": 0.7694306830179485,
+      "grad_norm": 6.279574614073089,
+      "learning_rate": 6.653268611348315e-07,
+      "loss": 0.3736,
+      "step": 9474
+    },
+    {
+      "epoch": 0.7695118979939901,
+      "grad_norm": 4.349805162366064,
+      "learning_rate": 6.64880207957983e-07,
+      "loss": 0.4281,
+      "step": 9475
+    },
+    {
+      "epoch": 0.7695931129700316,
+      "grad_norm": 4.485659687329684,
+      "learning_rate": 6.644336817617122e-07,
+      "loss": 0.4795,
+      "step": 9476
+    },
+    {
+      "epoch": 0.7696743279460733,
+      "grad_norm": 5.1817758651078885,
+      "learning_rate": 6.63987282576915e-07,
+      "loss": 0.4077,
+      "step": 9477
+    },
+    {
+      "epoch": 0.7697555429221148,
+      "grad_norm": 3.6533998487921298,
+      "learning_rate": 6.635410104344819e-07,
+      "loss": 0.5877,
+      "step": 9478
+    },
+    {
+      "epoch": 0.7698367578981564,
+      "grad_norm": 5.823071091152919,
+      "learning_rate": 6.630948653652905e-07,
+      "loss": 0.4759,
+      "step": 9479
+    },
+    {
+      "epoch": 0.769917972874198,
+      "grad_norm": 6.866234496364005,
+      "learning_rate": 6.62648847400213e-07,
+      "loss": 0.4428,
+      "step": 9480
+    },
+    {
+      "epoch": 0.7699991878502396,
+      "grad_norm": 5.054351987364898,
+      "learning_rate": 6.622029565701118e-07,
+      "loss": 0.4417,
+      "step": 9481
+    },
+    {
+      "epoch": 0.7700804028262812,
+      "grad_norm": 4.79929266732889,
+      "learning_rate": 6.617571929058397e-07,
+      "loss": 0.4682,
+      "step": 9482
+    },
+    {
+      "epoch": 0.7701616178023227,
+      "grad_norm": 4.783027204950755,
+      "learning_rate": 6.613115564382403e-07,
+      "loss": 0.402,
+      "step": 9483
+    },
+    {
+      "epoch": 0.7702428327783644,
+      "grad_norm": 11.930675421175312,
+      "learning_rate": 6.608660471981509e-07,
+      "loss": 0.4791,
+      "step": 9484
+    },
+    {
+      "epoch": 0.7703240477544059,
+      "grad_norm": 6.054291067646692,
+      "learning_rate": 6.604206652163967e-07,
+      "loss": 0.4235,
+      "step": 9485
+    },
+    {
+      "epoch": 0.7704052627304475,
+      "grad_norm": 7.228203275275898,
+      "learning_rate": 6.599754105237974e-07,
+      "loss": 0.4832,
+      "step": 9486
+    },
+    {
+      "epoch": 0.770486477706489,
+      "grad_norm": 3.7391869318420285,
+      "learning_rate": 6.595302831511607e-07,
+      "loss": 0.537,
+      "step": 9487
+    },
+    {
+      "epoch": 0.7705676926825307,
+      "grad_norm": 5.637019086673706,
+      "learning_rate": 6.590852831292885e-07,
+      "loss": 0.6264,
+      "step": 9488
+    },
+    {
+      "epoch": 0.7706489076585722,
+      "grad_norm": 7.91711166378353,
+      "learning_rate": 6.586404104889721e-07,
+      "loss": 0.477,
+      "step": 9489
+    },
+    {
+      "epoch": 0.7707301226346138,
+      "grad_norm": 5.980422311755343,
+      "learning_rate": 6.58195665260993e-07,
+      "loss": 0.433,
+      "step": 9490
+    },
+    {
+      "epoch": 0.7708113376106555,
+      "grad_norm": 10.886812979952378,
+      "learning_rate": 6.577510474761272e-07,
+      "loss": 0.4323,
+      "step": 9491
+    },
+    {
+      "epoch": 0.770892552586697,
+      "grad_norm": 5.242170156688564,
+      "learning_rate": 6.573065571651383e-07,
+      "loss": 0.3652,
+      "step": 9492
+    },
+    {
+      "epoch": 0.7709737675627386,
+      "grad_norm": 5.032479849055025,
+      "learning_rate": 6.56862194358783e-07,
+      "loss": 0.4055,
+      "step": 9493
+    },
+    {
+      "epoch": 0.7710549825387801,
+      "grad_norm": 9.525750208818708,
+      "learning_rate": 6.5641795908781e-07,
+      "loss": 0.5108,
+      "step": 9494
+    },
+    {
+      "epoch": 0.7711361975148218,
+      "grad_norm": 4.492371142698525,
+      "learning_rate": 6.559738513829572e-07,
+      "loss": 0.679,
+      "step": 9495
+    },
+    {
+      "epoch": 0.7712174124908633,
+      "grad_norm": 5.27190852592138,
+      "learning_rate": 6.555298712749538e-07,
+      "loss": 0.5479,
+      "step": 9496
+    },
+    {
+      "epoch": 0.7712986274669049,
+      "grad_norm": 5.955473036514189,
+      "learning_rate": 6.550860187945227e-07,
+      "loss": 0.4585,
+      "step": 9497
+    },
+    {
+      "epoch": 0.7713798424429464,
+      "grad_norm": 5.382341493930171,
+      "learning_rate": 6.546422939723738e-07,
+      "loss": 0.4537,
+      "step": 9498
+    },
+    {
+      "epoch": 0.7714610574189881,
+      "grad_norm": 5.842094824500748,
+      "learning_rate": 6.541986968392119e-07,
+      "loss": 0.4458,
+      "step": 9499
+    },
+    {
+      "epoch": 0.7715422723950296,
+      "grad_norm": 4.8906020204774165,
+      "learning_rate": 6.537552274257322e-07,
+      "loss": 0.4932,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7716234873710712,
+      "grad_norm": 8.934213855225288,
+      "learning_rate": 6.533118857626194e-07,
+      "loss": 0.5223,
+      "step": 9501
+    },
+    {
+      "epoch": 0.7717047023471129,
+      "grad_norm": 4.746748520348421,
+      "learning_rate": 6.52868671880551e-07,
+      "loss": 0.5565,
+      "step": 9502
+    },
+    {
+      "epoch": 0.7717859173231544,
+      "grad_norm": 8.532705058823186,
+      "learning_rate": 6.524255858101938e-07,
+      "loss": 0.5495,
+      "step": 9503
+    },
+    {
+      "epoch": 0.771867132299196,
+      "grad_norm": 10.865485946925974,
+      "learning_rate": 6.519826275822086e-07,
+      "loss": 0.5132,
+      "step": 9504
+    },
+    {
+      "epoch": 0.7719483472752375,
+      "grad_norm": 3.470727903343789,
+      "learning_rate": 6.515397972272444e-07,
+      "loss": 0.4055,
+      "step": 9505
+    },
+    {
+      "epoch": 0.7720295622512792,
+      "grad_norm": 4.190361234303148,
+      "learning_rate": 6.510970947759434e-07,
+      "loss": 0.4745,
+      "step": 9506
+    },
+    {
+      "epoch": 0.7721107772273207,
+      "grad_norm": 4.1269789461455515,
+      "learning_rate": 6.50654520258939e-07,
+      "loss": 0.4198,
+      "step": 9507
+    },
+    {
+      "epoch": 0.7721919922033623,
+      "grad_norm": 5.796517123286705,
+      "learning_rate": 6.502120737068543e-07,
+      "loss": 0.5517,
+      "step": 9508
+    },
+    {
+      "epoch": 0.7722732071794038,
+      "grad_norm": 3.9798948730265002,
+      "learning_rate": 6.497697551503032e-07,
+      "loss": 0.4771,
+      "step": 9509
+    },
+    {
+      "epoch": 0.7723544221554455,
+      "grad_norm": 3.3670043032275445,
+      "learning_rate": 6.493275646198941e-07,
+      "loss": 0.507,
+      "step": 9510
+    },
+    {
+      "epoch": 0.772435637131487,
+      "grad_norm": 4.789052775696503,
+      "learning_rate": 6.488855021462218e-07,
+      "loss": 0.651,
+      "step": 9511
+    },
+    {
+      "epoch": 0.7725168521075286,
+      "grad_norm": 4.574449211149501,
+      "learning_rate": 6.484435677598761e-07,
+      "loss": 0.3833,
+      "step": 9512
+    },
+    {
+      "epoch": 0.7725980670835703,
+      "grad_norm": 5.744708779873541,
+      "learning_rate": 6.480017614914369e-07,
+      "loss": 0.52,
+      "step": 9513
+    },
+    {
+      "epoch": 0.7726792820596118,
+      "grad_norm": 6.185709557761946,
+      "learning_rate": 6.475600833714743e-07,
+      "loss": 0.5064,
+      "step": 9514
+    },
+    {
+      "epoch": 0.7727604970356534,
+      "grad_norm": 6.093796766128401,
+      "learning_rate": 6.471185334305491e-07,
+      "loss": 0.5049,
+      "step": 9515
+    },
+    {
+      "epoch": 0.7728417120116949,
+      "grad_norm": 5.325643999422,
+      "learning_rate": 6.466771116992162e-07,
+      "loss": 0.4234,
+      "step": 9516
+    },
+    {
+      "epoch": 0.7729229269877366,
+      "grad_norm": 5.854284079602375,
+      "learning_rate": 6.462358182080175e-07,
+      "loss": 0.504,
+      "step": 9517
+    },
+    {
+      "epoch": 0.7730041419637781,
+      "grad_norm": 6.5341398235283865,
+      "learning_rate": 6.457946529874895e-07,
+      "loss": 0.6357,
+      "step": 9518
+    },
+    {
+      "epoch": 0.7730853569398197,
+      "grad_norm": 3.7271101675803027,
+      "learning_rate": 6.453536160681592e-07,
+      "loss": 0.484,
+      "step": 9519
+    },
+    {
+      "epoch": 0.7731665719158612,
+      "grad_norm": 3.610155546754207,
+      "learning_rate": 6.449127074805428e-07,
+      "loss": 0.464,
+      "step": 9520
+    },
+    {
+      "epoch": 0.7732477868919029,
+      "grad_norm": 4.384516099481292,
+      "learning_rate": 6.444719272551491e-07,
+      "loss": 0.4131,
+      "step": 9521
+    },
+    {
+      "epoch": 0.7733290018679444,
+      "grad_norm": 7.395867107582111,
+      "learning_rate": 6.440312754224773e-07,
+      "loss": 0.489,
+      "step": 9522
+    },
+    {
+      "epoch": 0.773410216843986,
+      "grad_norm": 3.52191839672315,
+      "learning_rate": 6.435907520130191e-07,
+      "loss": 0.3974,
+      "step": 9523
+    },
+    {
+      "epoch": 0.7734914318200277,
+      "grad_norm": 5.735438136851437,
+      "learning_rate": 6.431503570572554e-07,
+      "loss": 0.3681,
+      "step": 9524
+    },
+    {
+      "epoch": 0.7735726467960692,
+      "grad_norm": 4.138098595663441,
+      "learning_rate": 6.427100905856598e-07,
+      "loss": 0.46,
+      "step": 9525
+    },
+    {
+      "epoch": 0.7736538617721108,
+      "grad_norm": 4.627439119735813,
+      "learning_rate": 6.422699526286969e-07,
+      "loss": 0.5792,
+      "step": 9526
+    },
+    {
+      "epoch": 0.7737350767481523,
+      "grad_norm": 4.326175948208715,
+      "learning_rate": 6.418299432168215e-07,
+      "loss": 0.6043,
+      "step": 9527
+    },
+    {
+      "epoch": 0.773816291724194,
+      "grad_norm": 4.7243209100130334,
+      "learning_rate": 6.413900623804792e-07,
+      "loss": 0.3815,
+      "step": 9528
+    },
+    {
+      "epoch": 0.7738975067002355,
+      "grad_norm": 6.9628262078532535,
+      "learning_rate": 6.409503101501086e-07,
+      "loss": 0.5016,
+      "step": 9529
+    },
+    {
+      "epoch": 0.7739787216762771,
+      "grad_norm": 3.781446710484677,
+      "learning_rate": 6.405106865561367e-07,
+      "loss": 0.6054,
+      "step": 9530
+    },
+    {
+      "epoch": 0.7740599366523186,
+      "grad_norm": 8.30986430715804,
+      "learning_rate": 6.400711916289846e-07,
+      "loss": 0.443,
+      "step": 9531
+    },
+    {
+      "epoch": 0.7741411516283603,
+      "grad_norm": 4.838062384790877,
+      "learning_rate": 6.396318253990628e-07,
+      "loss": 0.4772,
+      "step": 9532
+    },
+    {
+      "epoch": 0.7742223666044018,
+      "grad_norm": 6.597678306183141,
+      "learning_rate": 6.391925878967728e-07,
+      "loss": 0.4941,
+      "step": 9533
+    },
+    {
+      "epoch": 0.7743035815804434,
+      "grad_norm": 4.6205393788485845,
+      "learning_rate": 6.387534791525072e-07,
+      "loss": 0.5037,
+      "step": 9534
+    },
+    {
+      "epoch": 0.7743847965564851,
+      "grad_norm": 4.269745258902035,
+      "learning_rate": 6.383144991966508e-07,
+      "loss": 0.7124,
+      "step": 9535
+    },
+    {
+      "epoch": 0.7744660115325266,
+      "grad_norm": 7.2941195850877145,
+      "learning_rate": 6.378756480595782e-07,
+      "loss": 0.4697,
+      "step": 9536
+    },
+    {
+      "epoch": 0.7745472265085682,
+      "grad_norm": 8.30813779565925,
+      "learning_rate": 6.374369257716548e-07,
+      "loss": 0.4272,
+      "step": 9537
+    },
+    {
+      "epoch": 0.7746284414846097,
+      "grad_norm": 5.268713325301926,
+      "learning_rate": 6.369983323632389e-07,
+      "loss": 0.3696,
+      "step": 9538
+    },
+    {
+      "epoch": 0.7747096564606514,
+      "grad_norm": 6.353077785001505,
+      "learning_rate": 6.365598678646793e-07,
+      "loss": 0.499,
+      "step": 9539
+    },
+    {
+      "epoch": 0.7747908714366929,
+      "grad_norm": 7.787690788411613,
+      "learning_rate": 6.361215323063144e-07,
+      "loss": 0.517,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7748720864127345,
+      "grad_norm": 6.712726861513404,
+      "learning_rate": 6.356833257184747e-07,
+      "loss": 0.3541,
+      "step": 9541
+    },
+    {
+      "epoch": 0.774953301388776,
+      "grad_norm": 7.036101207768536,
+      "learning_rate": 6.352452481314825e-07,
+      "loss": 0.5433,
+      "step": 9542
+    },
+    {
+      "epoch": 0.7750345163648177,
+      "grad_norm": 4.440795097972874,
+      "learning_rate": 6.348072995756497e-07,
+      "loss": 0.4672,
+      "step": 9543
+    },
+    {
+      "epoch": 0.7751157313408592,
+      "grad_norm": 6.623356794323855,
+      "learning_rate": 6.3436948008128e-07,
+      "loss": 0.479,
+      "step": 9544
+    },
+    {
+      "epoch": 0.7751969463169008,
+      "grad_norm": 3.870234328714397,
+      "learning_rate": 6.339317896786693e-07,
+      "loss": 0.5124,
+      "step": 9545
+    },
+    {
+      "epoch": 0.7752781612929425,
+      "grad_norm": 7.935205106400605,
+      "learning_rate": 6.33494228398103e-07,
+      "loss": 0.5108,
+      "step": 9546
+    },
+    {
+      "epoch": 0.775359376268984,
+      "grad_norm": 4.477009003593662,
+      "learning_rate": 6.33056796269857e-07,
+      "loss": 0.5893,
+      "step": 9547
+    },
+    {
+      "epoch": 0.7754405912450256,
+      "grad_norm": 6.221717457872238,
+      "learning_rate": 6.326194933242006e-07,
+      "loss": 0.5316,
+      "step": 9548
+    },
+    {
+      "epoch": 0.7755218062210671,
+      "grad_norm": 5.507373178005721,
+      "learning_rate": 6.321823195913924e-07,
+      "loss": 0.5283,
+      "step": 9549
+    },
+    {
+      "epoch": 0.7756030211971088,
+      "grad_norm": 5.749251136653241,
+      "learning_rate": 6.317452751016815e-07,
+      "loss": 0.5322,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7756842361731503,
+      "grad_norm": 4.194072441757716,
+      "learning_rate": 6.313083598853101e-07,
+      "loss": 0.5895,
+      "step": 9551
+    },
+    {
+      "epoch": 0.7757654511491919,
+      "grad_norm": 5.534135119543315,
+      "learning_rate": 6.308715739725108e-07,
+      "loss": 0.3903,
+      "step": 9552
+    },
+    {
+      "epoch": 0.7758466661252335,
+      "grad_norm": 5.890217163199354,
+      "learning_rate": 6.30434917393506e-07,
+      "loss": 0.4641,
+      "step": 9553
+    },
+    {
+      "epoch": 0.7759278811012751,
+      "grad_norm": 4.610182903829491,
+      "learning_rate": 6.299983901785109e-07,
+      "loss": 0.5238,
+      "step": 9554
+    },
+    {
+      "epoch": 0.7760090960773166,
+      "grad_norm": 5.475473491313324,
+      "learning_rate": 6.295619923577303e-07,
+      "loss": 0.4373,
+      "step": 9555
+    },
+    {
+      "epoch": 0.7760903110533582,
+      "grad_norm": 3.7085653093083364,
+      "learning_rate": 6.291257239613599e-07,
+      "loss": 0.5475,
+      "step": 9556
+    },
+    {
+      "epoch": 0.7761715260293999,
+      "grad_norm": 5.431007075610515,
+      "learning_rate": 6.286895850195882e-07,
+      "loss": 0.521,
+      "step": 9557
+    },
+    {
+      "epoch": 0.7762527410054414,
+      "grad_norm": 7.326528629894263,
+      "learning_rate": 6.28253575562594e-07,
+      "loss": 0.5119,
+      "step": 9558
+    },
+    {
+      "epoch": 0.776333955981483,
+      "grad_norm": 6.69328051954586,
+      "learning_rate": 6.278176956205462e-07,
+      "loss": 0.468,
+      "step": 9559
+    },
+    {
+      "epoch": 0.7764151709575245,
+      "grad_norm": 7.6181081748087776,
+      "learning_rate": 6.273819452236049e-07,
+      "loss": 0.4013,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7764963859335662,
+      "grad_norm": 5.517368857197591,
+      "learning_rate": 6.269463244019231e-07,
+      "loss": 0.4919,
+      "step": 9561
+    },
+    {
+      "epoch": 0.7765776009096077,
+      "grad_norm": 5.2478286059509145,
+      "learning_rate": 6.265108331856423e-07,
+      "loss": 0.4883,
+      "step": 9562
+    },
+    {
+      "epoch": 0.7766588158856493,
+      "grad_norm": 8.241994509786712,
+      "learning_rate": 6.260754716048961e-07,
+      "loss": 0.4527,
+      "step": 9563
+    },
+    {
+      "epoch": 0.7767400308616909,
+      "grad_norm": 5.796864584164602,
+      "learning_rate": 6.256402396898095e-07,
+      "loss": 0.4505,
+      "step": 9564
+    },
+    {
+      "epoch": 0.7768212458377325,
+      "grad_norm": 5.288449361519552,
+      "learning_rate": 6.252051374704992e-07,
+      "loss": 0.5593,
+      "step": 9565
+    },
+    {
+      "epoch": 0.776902460813774,
+      "grad_norm": 4.215256696565076,
+      "learning_rate": 6.247701649770707e-07,
+      "loss": 0.4833,
+      "step": 9566
+    },
+    {
+      "epoch": 0.7769836757898156,
+      "grad_norm": 4.625763848409002,
+      "learning_rate": 6.243353222396229e-07,
+      "loss": 0.4169,
+      "step": 9567
+    },
+    {
+      "epoch": 0.7770648907658573,
+      "grad_norm": 7.386761076675003,
+      "learning_rate": 6.239006092882438e-07,
+      "loss": 0.5572,
+      "step": 9568
+    },
+    {
+      "epoch": 0.7771461057418988,
+      "grad_norm": 3.960429753595598,
+      "learning_rate": 6.234660261530126e-07,
+      "loss": 0.448,
+      "step": 9569
+    },
+    {
+      "epoch": 0.7772273207179404,
+      "grad_norm": 6.218816667980998,
+      "learning_rate": 6.23031572864001e-07,
+      "loss": 0.4399,
+      "step": 9570
+    },
+    {
+      "epoch": 0.777308535693982,
+      "grad_norm": 11.37451355769962,
+      "learning_rate": 6.225972494512719e-07,
+      "loss": 0.4474,
+      "step": 9571
+    },
+    {
+      "epoch": 0.7773897506700236,
+      "grad_norm": 4.63095111539401,
+      "learning_rate": 6.22163055944876e-07,
+      "loss": 0.4599,
+      "step": 9572
+    },
+    {
+      "epoch": 0.7774709656460651,
+      "grad_norm": 4.637853908423072,
+      "learning_rate": 6.217289923748592e-07,
+      "loss": 0.3144,
+      "step": 9573
+    },
+    {
+      "epoch": 0.7775521806221067,
+      "grad_norm": 7.081685491650372,
+      "learning_rate": 6.212950587712557e-07,
+      "loss": 0.6146,
+      "step": 9574
+    },
+    {
+      "epoch": 0.7776333955981483,
+      "grad_norm": 4.122348761014315,
+      "learning_rate": 6.20861255164091e-07,
+      "loss": 0.5314,
+      "step": 9575
+    },
+    {
+      "epoch": 0.7777146105741899,
+      "grad_norm": 5.74857146041133,
+      "learning_rate": 6.204275815833807e-07,
+      "loss": 0.4767,
+      "step": 9576
+    },
+    {
+      "epoch": 0.7777958255502314,
+      "grad_norm": 5.874273219932277,
+      "learning_rate": 6.19994038059136e-07,
+      "loss": 0.4707,
+      "step": 9577
+    },
+    {
+      "epoch": 0.777877040526273,
+      "grad_norm": 6.134515006712459,
+      "learning_rate": 6.19560624621354e-07,
+      "loss": 0.5144,
+      "step": 9578
+    },
+    {
+      "epoch": 0.7779582555023147,
+      "grad_norm": 5.019484047549089,
+      "learning_rate": 6.191273413000237e-07,
+      "loss": 0.3622,
+      "step": 9579
+    },
+    {
+      "epoch": 0.7780394704783562,
+      "grad_norm": 4.813176127586792,
+      "learning_rate": 6.186941881251279e-07,
+      "loss": 0.3609,
+      "step": 9580
+    },
+    {
+      "epoch": 0.7781206854543978,
+      "grad_norm": 6.957964558660779,
+      "learning_rate": 6.182611651266376e-07,
+      "loss": 0.4625,
+      "step": 9581
+    },
+    {
+      "epoch": 0.7782019004304394,
+      "grad_norm": 4.786938388125974,
+      "learning_rate": 6.17828272334515e-07,
+      "loss": 0.4196,
+      "step": 9582
+    },
+    {
+      "epoch": 0.778283115406481,
+      "grad_norm": 14.455541969389369,
+      "learning_rate": 6.173955097787149e-07,
+      "loss": 0.3105,
+      "step": 9583
+    },
+    {
+      "epoch": 0.7783643303825225,
+      "grad_norm": 6.015474094967977,
+      "learning_rate": 6.169628774891826e-07,
+      "loss": 0.6127,
+      "step": 9584
+    },
+    {
+      "epoch": 0.7784455453585641,
+      "grad_norm": 7.69183168267593,
+      "learning_rate": 6.165303754958524e-07,
+      "loss": 0.5329,
+      "step": 9585
+    },
+    {
+      "epoch": 0.7785267603346057,
+      "grad_norm": 7.494666404519941,
+      "learning_rate": 6.160980038286529e-07,
+      "loss": 0.3892,
+      "step": 9586
+    },
+    {
+      "epoch": 0.7786079753106473,
+      "grad_norm": 3.6352807891152508,
+      "learning_rate": 6.156657625175011e-07,
+      "loss": 0.4718,
+      "step": 9587
+    },
+    {
+      "epoch": 0.7786891902866888,
+      "grad_norm": 3.6374040934237066,
+      "learning_rate": 6.152336515923052e-07,
+      "loss": 0.49,
+      "step": 9588
+    },
+    {
+      "epoch": 0.7787704052627304,
+      "grad_norm": 4.790151735557977,
+      "learning_rate": 6.148016710829654e-07,
+      "loss": 0.5964,
+      "step": 9589
+    },
+    {
+      "epoch": 0.7788516202387721,
+      "grad_norm": 6.6959826045900614,
+      "learning_rate": 6.143698210193738e-07,
+      "loss": 0.7207,
+      "step": 9590
+    },
+    {
+      "epoch": 0.7789328352148136,
+      "grad_norm": 5.021956045573845,
+      "learning_rate": 6.139381014314108e-07,
+      "loss": 0.4336,
+      "step": 9591
+    },
+    {
+      "epoch": 0.7790140501908552,
+      "grad_norm": 8.574991586907696,
+      "learning_rate": 6.135065123489486e-07,
+      "loss": 0.4282,
+      "step": 9592
+    },
+    {
+      "epoch": 0.7790952651668968,
+      "grad_norm": 4.869738833833407,
+      "learning_rate": 6.130750538018524e-07,
+      "loss": 0.5189,
+      "step": 9593
+    },
+    {
+      "epoch": 0.7791764801429384,
+      "grad_norm": 5.876692632158238,
+      "learning_rate": 6.12643725819976e-07,
+      "loss": 0.4978,
+      "step": 9594
+    },
+    {
+      "epoch": 0.7792576951189799,
+      "grad_norm": 7.302404610138221,
+      "learning_rate": 6.122125284331646e-07,
+      "loss": 0.5031,
+      "step": 9595
+    },
+    {
+      "epoch": 0.7793389100950215,
+      "grad_norm": 6.239496444525358,
+      "learning_rate": 6.117814616712548e-07,
+      "loss": 0.4399,
+      "step": 9596
+    },
+    {
+      "epoch": 0.7794201250710631,
+      "grad_norm": 4.341820954785043,
+      "learning_rate": 6.113505255640756e-07,
+      "loss": 0.4661,
+      "step": 9597
+    },
+    {
+      "epoch": 0.7795013400471047,
+      "grad_norm": 4.82168987167205,
+      "learning_rate": 6.109197201414438e-07,
+      "loss": 0.4285,
+      "step": 9598
+    },
+    {
+      "epoch": 0.7795825550231462,
+      "grad_norm": 5.620360996597726,
+      "learning_rate": 6.104890454331702e-07,
+      "loss": 0.6044,
+      "step": 9599
+    },
+    {
+      "epoch": 0.7796637699991879,
+      "grad_norm": 4.809738135444571,
+      "learning_rate": 6.100585014690547e-07,
+      "loss": 0.6426,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7797449849752295,
+      "grad_norm": 3.621342250605208,
+      "learning_rate": 6.096280882788874e-07,
+      "loss": 0.5404,
+      "step": 9601
+    },
+    {
+      "epoch": 0.779826199951271,
+      "grad_norm": 8.468527887913863,
+      "learning_rate": 6.091978058924522e-07,
+      "loss": 0.3656,
+      "step": 9602
+    },
+    {
+      "epoch": 0.7799074149273126,
+      "grad_norm": 3.464489359474417,
+      "learning_rate": 6.087676543395224e-07,
+      "loss": 0.6374,
+      "step": 9603
+    },
+    {
+      "epoch": 0.7799886299033542,
+      "grad_norm": 5.245897271466931,
+      "learning_rate": 6.083376336498608e-07,
+      "loss": 0.5772,
+      "step": 9604
+    },
+    {
+      "epoch": 0.7800698448793958,
+      "grad_norm": 6.62332037882836,
+      "learning_rate": 6.079077438532246e-07,
+      "loss": 0.5533,
+      "step": 9605
+    },
+    {
+      "epoch": 0.7801510598554373,
+      "grad_norm": 5.690296444200083,
+      "learning_rate": 6.074779849793585e-07,
+      "loss": 0.5408,
+      "step": 9606
+    },
+    {
+      "epoch": 0.780232274831479,
+      "grad_norm": 4.804271179305116,
+      "learning_rate": 6.07048357057999e-07,
+      "loss": 0.558,
+      "step": 9607
+    },
+    {
+      "epoch": 0.7803134898075205,
+      "grad_norm": 6.074710914942616,
+      "learning_rate": 6.066188601188757e-07,
+      "loss": 0.5146,
+      "step": 9608
+    },
+    {
+      "epoch": 0.7803947047835621,
+      "grad_norm": 6.506228847573674,
+      "learning_rate": 6.061894941917062e-07,
+      "loss": 0.5551,
+      "step": 9609
+    },
+    {
+      "epoch": 0.7804759197596036,
+      "grad_norm": 5.035663800312012,
+      "learning_rate": 6.057602593062015e-07,
+      "loss": 0.4042,
+      "step": 9610
+    },
+    {
+      "epoch": 0.7805571347356453,
+      "grad_norm": 3.7689993752107234,
+      "learning_rate": 6.053311554920607e-07,
+      "loss": 0.4688,
+      "step": 9611
+    },
+    {
+      "epoch": 0.7806383497116869,
+      "grad_norm": 6.095711004984003,
+      "learning_rate": 6.049021827789774e-07,
+      "loss": 0.3031,
+      "step": 9612
+    },
+    {
+      "epoch": 0.7807195646877284,
+      "grad_norm": 5.01460034158621,
+      "learning_rate": 6.044733411966336e-07,
+      "loss": 0.4798,
+      "step": 9613
+    },
+    {
+      "epoch": 0.78080077966377,
+      "grad_norm": 8.19175212779431,
+      "learning_rate": 6.040446307747019e-07,
+      "loss": 0.5674,
+      "step": 9614
+    },
+    {
+      "epoch": 0.7808819946398116,
+      "grad_norm": 4.961605893602875,
+      "learning_rate": 6.036160515428475e-07,
+      "loss": 0.4449,
+      "step": 9615
+    },
+    {
+      "epoch": 0.7809632096158532,
+      "grad_norm": 6.915675610043426,
+      "learning_rate": 6.031876035307263e-07,
+      "loss": 0.4569,
+      "step": 9616
+    },
+    {
+      "epoch": 0.7810444245918947,
+      "grad_norm": 5.568377529756666,
+      "learning_rate": 6.027592867679838e-07,
+      "loss": 0.3962,
+      "step": 9617
+    },
+    {
+      "epoch": 0.7811256395679363,
+      "grad_norm": 6.129156630011155,
+      "learning_rate": 6.023311012842581e-07,
+      "loss": 0.5745,
+      "step": 9618
+    },
+    {
+      "epoch": 0.7812068545439779,
+      "grad_norm": 5.108160511033742,
+      "learning_rate": 6.019030471091772e-07,
+      "loss": 0.3951,
+      "step": 9619
+    },
+    {
+      "epoch": 0.7812880695200195,
+      "grad_norm": 5.561892142462487,
+      "learning_rate": 6.014751242723591e-07,
+      "loss": 0.5265,
+      "step": 9620
+    },
+    {
+      "epoch": 0.781369284496061,
+      "grad_norm": 3.7511127265919244,
+      "learning_rate": 6.010473328034153e-07,
+      "loss": 0.5185,
+      "step": 9621
+    },
+    {
+      "epoch": 0.7814504994721027,
+      "grad_norm": 5.167066450794399,
+      "learning_rate": 6.006196727319452e-07,
+      "loss": 0.4383,
+      "step": 9622
+    },
+    {
+      "epoch": 0.7815317144481443,
+      "grad_norm": 7.96179114023694,
+      "learning_rate": 6.001921440875414e-07,
+      "loss": 0.3846,
+      "step": 9623
+    },
+    {
+      "epoch": 0.7816129294241858,
+      "grad_norm": 5.371452420693181,
+      "learning_rate": 5.997647468997875e-07,
+      "loss": 0.6281,
+      "step": 9624
+    },
+    {
+      "epoch": 0.7816941444002274,
+      "grad_norm": 5.8732462009916,
+      "learning_rate": 5.99337481198256e-07,
+      "loss": 0.5173,
+      "step": 9625
+    },
+    {
+      "epoch": 0.781775359376269,
+      "grad_norm": 5.56845044238569,
+      "learning_rate": 5.989103470125113e-07,
+      "loss": 0.5523,
+      "step": 9626
+    },
+    {
+      "epoch": 0.7818565743523106,
+      "grad_norm": 6.424489397897517,
+      "learning_rate": 5.984833443721097e-07,
+      "loss": 0.3735,
+      "step": 9627
+    },
+    {
+      "epoch": 0.7819377893283521,
+      "grad_norm": 4.1718433726191275,
+      "learning_rate": 5.980564733065963e-07,
+      "loss": 0.4501,
+      "step": 9628
+    },
+    {
+      "epoch": 0.7820190043043938,
+      "grad_norm": 5.012432986771707,
+      "learning_rate": 5.976297338455101e-07,
+      "loss": 0.5626,
+      "step": 9629
+    },
+    {
+      "epoch": 0.7821002192804353,
+      "grad_norm": 13.705889019793014,
+      "learning_rate": 5.972031260183772e-07,
+      "loss": 0.5116,
+      "step": 9630
+    },
+    {
+      "epoch": 0.7821814342564769,
+      "grad_norm": 5.262747778591087,
+      "learning_rate": 5.967766498547181e-07,
+      "loss": 0.4009,
+      "step": 9631
+    },
+    {
+      "epoch": 0.7822626492325184,
+      "grad_norm": 4.146136341129552,
+      "learning_rate": 5.963503053840425e-07,
+      "loss": 0.4744,
+      "step": 9632
+    },
+    {
+      "epoch": 0.7823438642085601,
+      "grad_norm": 6.971385385680695,
+      "learning_rate": 5.959240926358501e-07,
+      "loss": 0.4348,
+      "step": 9633
+    },
+    {
+      "epoch": 0.7824250791846017,
+      "grad_norm": 4.565877029510119,
+      "learning_rate": 5.954980116396336e-07,
+      "loss": 0.6681,
+      "step": 9634
+    },
+    {
+      "epoch": 0.7825062941606432,
+      "grad_norm": 4.089174310097666,
+      "learning_rate": 5.950720624248749e-07,
+      "loss": 0.484,
+      "step": 9635
+    },
+    {
+      "epoch": 0.7825875091366848,
+      "grad_norm": 5.870091842126987,
+      "learning_rate": 5.946462450210477e-07,
+      "loss": 0.4509,
+      "step": 9636
+    },
+    {
+      "epoch": 0.7826687241127264,
+      "grad_norm": 17.233940029392116,
+      "learning_rate": 5.942205594576173e-07,
+      "loss": 0.5734,
+      "step": 9637
+    },
+    {
+      "epoch": 0.782749939088768,
+      "grad_norm": 12.113335333152698,
+      "learning_rate": 5.937950057640376e-07,
+      "loss": 0.4828,
+      "step": 9638
+    },
+    {
+      "epoch": 0.7828311540648095,
+      "grad_norm": 7.210552071495601,
+      "learning_rate": 5.933695839697548e-07,
+      "loss": 0.4928,
+      "step": 9639
+    },
+    {
+      "epoch": 0.7829123690408512,
+      "grad_norm": 4.338185170282578,
+      "learning_rate": 5.929442941042066e-07,
+      "loss": 0.578,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7829935840168927,
+      "grad_norm": 6.223280812892159,
+      "learning_rate": 5.925191361968194e-07,
+      "loss": 0.4616,
+      "step": 9641
+    },
+    {
+      "epoch": 0.7830747989929343,
+      "grad_norm": 3.9906953099976397,
+      "learning_rate": 5.920941102770128e-07,
+      "loss": 0.539,
+      "step": 9642
+    },
+    {
+      "epoch": 0.7831560139689758,
+      "grad_norm": 4.895008165041246,
+      "learning_rate": 5.916692163741972e-07,
+      "loss": 0.5437,
+      "step": 9643
+    },
+    {
+      "epoch": 0.7832372289450175,
+      "grad_norm": 7.001022857461772,
+      "learning_rate": 5.91244454517772e-07,
+      "loss": 0.4387,
+      "step": 9644
+    },
+    {
+      "epoch": 0.7833184439210591,
+      "grad_norm": 8.865782780347201,
+      "learning_rate": 5.908198247371289e-07,
+      "loss": 0.4938,
+      "step": 9645
+    },
+    {
+      "epoch": 0.7833996588971006,
+      "grad_norm": 3.951135802126336,
+      "learning_rate": 5.903953270616486e-07,
+      "loss": 0.4084,
+      "step": 9646
+    },
+    {
+      "epoch": 0.7834808738731422,
+      "grad_norm": 6.755133249113319,
+      "learning_rate": 5.899709615207055e-07,
+      "loss": 0.5784,
+      "step": 9647
+    },
+    {
+      "epoch": 0.7835620888491838,
+      "grad_norm": 7.946028938098305,
+      "learning_rate": 5.895467281436637e-07,
+      "loss": 0.5064,
+      "step": 9648
+    },
+    {
+      "epoch": 0.7836433038252254,
+      "grad_norm": 5.014225943190868,
+      "learning_rate": 5.891226269598768e-07,
+      "loss": 0.6636,
+      "step": 9649
+    },
+    {
+      "epoch": 0.7837245188012669,
+      "grad_norm": 5.034789109316721,
+      "learning_rate": 5.886986579986917e-07,
+      "loss": 0.4543,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7838057337773086,
+      "grad_norm": 5.031634092156704,
+      "learning_rate": 5.882748212894441e-07,
+      "loss": 0.5141,
+      "step": 9651
+    },
+    {
+      "epoch": 0.7838869487533501,
+      "grad_norm": 3.651635786477034,
+      "learning_rate": 5.878511168614601e-07,
+      "loss": 0.5764,
+      "step": 9652
+    },
+    {
+      "epoch": 0.7839681637293917,
+      "grad_norm": 4.577152340868677,
+      "learning_rate": 5.874275447440599e-07,
+      "loss": 0.4339,
+      "step": 9653
+    },
+    {
+      "epoch": 0.7840493787054332,
+      "grad_norm": 6.2520933191135875,
+      "learning_rate": 5.870041049665507e-07,
+      "loss": 0.5523,
+      "step": 9654
+    },
+    {
+      "epoch": 0.7841305936814749,
+      "grad_norm": 4.385286574382671,
+      "learning_rate": 5.86580797558233e-07,
+      "loss": 0.4017,
+      "step": 9655
+    },
+    {
+      "epoch": 0.7842118086575165,
+      "grad_norm": 3.6357368477226233,
+      "learning_rate": 5.861576225483984e-07,
+      "loss": 0.4594,
+      "step": 9656
+    },
+    {
+      "epoch": 0.784293023633558,
+      "grad_norm": 5.476827776987912,
+      "learning_rate": 5.857345799663272e-07,
+      "loss": 0.3793,
+      "step": 9657
+    },
+    {
+      "epoch": 0.7843742386095996,
+      "grad_norm": 5.500782599304935,
+      "learning_rate": 5.853116698412913e-07,
+      "loss": 0.4516,
+      "step": 9658
+    },
+    {
+      "epoch": 0.7844554535856412,
+      "grad_norm": 4.882146338011125,
+      "learning_rate": 5.848888922025553e-07,
+      "loss": 0.5161,
+      "step": 9659
+    },
+    {
+      "epoch": 0.7845366685616828,
+      "grad_norm": 5.737418334120377,
+      "learning_rate": 5.844662470793716e-07,
+      "loss": 0.4623,
+      "step": 9660
+    },
+    {
+      "epoch": 0.7846178835377243,
+      "grad_norm": 5.3911366434680685,
+      "learning_rate": 5.840437345009859e-07,
+      "loss": 0.5734,
+      "step": 9661
+    },
+    {
+      "epoch": 0.784699098513766,
+      "grad_norm": 4.290906281542307,
+      "learning_rate": 5.83621354496634e-07,
+      "loss": 0.5312,
+      "step": 9662
+    },
+    {
+      "epoch": 0.7847803134898075,
+      "grad_norm": 5.746160856248301,
+      "learning_rate": 5.831991070955426e-07,
+      "loss": 0.4219,
+      "step": 9663
+    },
+    {
+      "epoch": 0.7848615284658491,
+      "grad_norm": 5.04902095963468,
+      "learning_rate": 5.827769923269283e-07,
+      "loss": 0.5044,
+      "step": 9664
+    },
+    {
+      "epoch": 0.7849427434418906,
+      "grad_norm": 10.70114453686239,
+      "learning_rate": 5.823550102199985e-07,
+      "loss": 0.4453,
+      "step": 9665
+    },
+    {
+      "epoch": 0.7850239584179323,
+      "grad_norm": 4.150840098492769,
+      "learning_rate": 5.819331608039538e-07,
+      "loss": 0.4819,
+      "step": 9666
+    },
+    {
+      "epoch": 0.7851051733939739,
+      "grad_norm": 7.526924883890027,
+      "learning_rate": 5.815114441079825e-07,
+      "loss": 0.5268,
+      "step": 9667
+    },
+    {
+      "epoch": 0.7851863883700154,
+      "grad_norm": 6.803056853476825,
+      "learning_rate": 5.810898601612657e-07,
+      "loss": 0.6169,
+      "step": 9668
+    },
+    {
+      "epoch": 0.785267603346057,
+      "grad_norm": 6.656336422437745,
+      "learning_rate": 5.806684089929756e-07,
+      "loss": 0.4574,
+      "step": 9669
+    },
+    {
+      "epoch": 0.7853488183220986,
+      "grad_norm": 4.099463565565569,
+      "learning_rate": 5.802470906322738e-07,
+      "loss": 0.5343,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7854300332981402,
+      "grad_norm": 9.804778473842848,
+      "learning_rate": 5.798259051083124e-07,
+      "loss": 0.4658,
+      "step": 9671
+    },
+    {
+      "epoch": 0.7855112482741817,
+      "grad_norm": 5.831632381423531,
+      "learning_rate": 5.794048524502366e-07,
+      "loss": 0.3633,
+      "step": 9672
+    },
+    {
+      "epoch": 0.7855924632502234,
+      "grad_norm": 4.512592207772697,
+      "learning_rate": 5.789839326871799e-07,
+      "loss": 0.5674,
+      "step": 9673
+    },
+    {
+      "epoch": 0.7856736782262649,
+      "grad_norm": 7.058643124685738,
+      "learning_rate": 5.785631458482679e-07,
+      "loss": 0.4566,
+      "step": 9674
+    },
+    {
+      "epoch": 0.7857548932023065,
+      "grad_norm": 9.677711342457231,
+      "learning_rate": 5.781424919626183e-07,
+      "loss": 0.5513,
+      "step": 9675
+    },
+    {
+      "epoch": 0.785836108178348,
+      "grad_norm": 7.654772369572434,
+      "learning_rate": 5.777219710593365e-07,
+      "loss": 0.4374,
+      "step": 9676
+    },
+    {
+      "epoch": 0.7859173231543897,
+      "grad_norm": 5.412971281189123,
+      "learning_rate": 5.773015831675204e-07,
+      "loss": 0.518,
+      "step": 9677
+    },
+    {
+      "epoch": 0.7859985381304313,
+      "grad_norm": 7.042856826106027,
+      "learning_rate": 5.768813283162597e-07,
+      "loss": 0.4422,
+      "step": 9678
+    },
+    {
+      "epoch": 0.7860797531064728,
+      "grad_norm": 8.697761416414153,
+      "learning_rate": 5.764612065346328e-07,
+      "loss": 0.3727,
+      "step": 9679
+    },
+    {
+      "epoch": 0.7861609680825145,
+      "grad_norm": 11.951299771554737,
+      "learning_rate": 5.760412178517099e-07,
+      "loss": 0.3518,
+      "step": 9680
+    },
+    {
+      "epoch": 0.786242183058556,
+      "grad_norm": 5.170910615421707,
+      "learning_rate": 5.75621362296552e-07,
+      "loss": 0.7013,
+      "step": 9681
+    },
+    {
+      "epoch": 0.7863233980345976,
+      "grad_norm": 15.391790408942455,
+      "learning_rate": 5.752016398982122e-07,
+      "loss": 0.4973,
+      "step": 9682
+    },
+    {
+      "epoch": 0.7864046130106391,
+      "grad_norm": 6.4992455503524145,
+      "learning_rate": 5.747820506857318e-07,
+      "loss": 0.5486,
+      "step": 9683
+    },
+    {
+      "epoch": 0.7864858279866808,
+      "grad_norm": 5.514244182216503,
+      "learning_rate": 5.74362594688144e-07,
+      "loss": 0.434,
+      "step": 9684
+    },
+    {
+      "epoch": 0.7865670429627223,
+      "grad_norm": 4.203294166143777,
+      "learning_rate": 5.739432719344737e-07,
+      "loss": 0.4048,
+      "step": 9685
+    },
+    {
+      "epoch": 0.7866482579387639,
+      "grad_norm": 5.401361035345732,
+      "learning_rate": 5.73524082453735e-07,
+      "loss": 0.4279,
+      "step": 9686
+    },
+    {
+      "epoch": 0.7867294729148054,
+      "grad_norm": 4.158513777957124,
+      "learning_rate": 5.731050262749341e-07,
+      "loss": 0.4841,
+      "step": 9687
+    },
+    {
+      "epoch": 0.7868106878908471,
+      "grad_norm": 7.879580004180549,
+      "learning_rate": 5.726861034270681e-07,
+      "loss": 0.3681,
+      "step": 9688
+    },
+    {
+      "epoch": 0.7868919028668887,
+      "grad_norm": 3.9835235897565107,
+      "learning_rate": 5.722673139391236e-07,
+      "loss": 0.4581,
+      "step": 9689
+    },
+    {
+      "epoch": 0.7869731178429302,
+      "grad_norm": 5.506433493815765,
+      "learning_rate": 5.718486578400775e-07,
+      "loss": 0.4241,
+      "step": 9690
+    },
+    {
+      "epoch": 0.7870543328189719,
+      "grad_norm": 6.423378795375947,
+      "learning_rate": 5.714301351589008e-07,
+      "loss": 0.6751,
+      "step": 9691
+    },
+    {
+      "epoch": 0.7871355477950134,
+      "grad_norm": 4.6037563920362095,
+      "learning_rate": 5.710117459245518e-07,
+      "loss": 0.6218,
+      "step": 9692
+    },
+    {
+      "epoch": 0.787216762771055,
+      "grad_norm": 5.535444733765645,
+      "learning_rate": 5.705934901659804e-07,
+      "loss": 0.609,
+      "step": 9693
+    },
+    {
+      "epoch": 0.7872979777470965,
+      "grad_norm": 6.072040161253787,
+      "learning_rate": 5.70175367912128e-07,
+      "loss": 0.3255,
+      "step": 9694
+    },
+    {
+      "epoch": 0.7873791927231382,
+      "grad_norm": 3.930200276565603,
+      "learning_rate": 5.697573791919275e-07,
+      "loss": 0.3883,
+      "step": 9695
+    },
+    {
+      "epoch": 0.7874604076991797,
+      "grad_norm": 13.405961625131877,
+      "learning_rate": 5.693395240343e-07,
+      "loss": 0.469,
+      "step": 9696
+    },
+    {
+      "epoch": 0.7875416226752213,
+      "grad_norm": 3.8674362850461357,
+      "learning_rate": 5.689218024681603e-07,
+      "loss": 0.4996,
+      "step": 9697
+    },
+    {
+      "epoch": 0.7876228376512628,
+      "grad_norm": 6.411118676070839,
+      "learning_rate": 5.685042145224118e-07,
+      "loss": 0.3813,
+      "step": 9698
+    },
+    {
+      "epoch": 0.7877040526273045,
+      "grad_norm": 3.36119530692789,
+      "learning_rate": 5.680867602259485e-07,
+      "loss": 0.449,
+      "step": 9699
+    },
+    {
+      "epoch": 0.7877852676033461,
+      "grad_norm": 5.637861341121901,
+      "learning_rate": 5.676694396076568e-07,
+      "loss": 0.5444,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7878664825793876,
+      "grad_norm": 8.011959392935337,
+      "learning_rate": 5.672522526964141e-07,
+      "loss": 0.4141,
+      "step": 9701
+    },
+    {
+      "epoch": 0.7879476975554293,
+      "grad_norm": 4.254934260559165,
+      "learning_rate": 5.668351995210866e-07,
+      "loss": 0.5489,
+      "step": 9702
+    },
+    {
+      "epoch": 0.7880289125314708,
+      "grad_norm": 4.2266389347699995,
+      "learning_rate": 5.664182801105314e-07,
+      "loss": 0.4893,
+      "step": 9703
+    },
+    {
+      "epoch": 0.7881101275075124,
+      "grad_norm": 9.12782592389075,
+      "learning_rate": 5.660014944935985e-07,
+      "loss": 0.5584,
+      "step": 9704
+    },
+    {
+      "epoch": 0.7881913424835539,
+      "grad_norm": 4.637756752610576,
+      "learning_rate": 5.655848426991267e-07,
+      "loss": 0.4832,
+      "step": 9705
+    },
+    {
+      "epoch": 0.7882725574595956,
+      "grad_norm": 5.685502111243551,
+      "learning_rate": 5.651683247559445e-07,
+      "loss": 0.3528,
+      "step": 9706
+    },
+    {
+      "epoch": 0.7883537724356371,
+      "grad_norm": 6.879039707986046,
+      "learning_rate": 5.647519406928758e-07,
+      "loss": 0.3939,
+      "step": 9707
+    },
+    {
+      "epoch": 0.7884349874116787,
+      "grad_norm": 4.090096342166239,
+      "learning_rate": 5.643356905387307e-07,
+      "loss": 0.6541,
+      "step": 9708
+    },
+    {
+      "epoch": 0.7885162023877202,
+      "grad_norm": 5.705199569403093,
+      "learning_rate": 5.639195743223105e-07,
+      "loss": 0.4791,
+      "step": 9709
+    },
+    {
+      "epoch": 0.7885974173637619,
+      "grad_norm": 6.503352654339194,
+      "learning_rate": 5.635035920724102e-07,
+      "loss": 0.4197,
+      "step": 9710
+    },
+    {
+      "epoch": 0.7886786323398035,
+      "grad_norm": 3.9110515788245204,
+      "learning_rate": 5.630877438178126e-07,
+      "loss": 0.6805,
+      "step": 9711
+    },
+    {
+      "epoch": 0.788759847315845,
+      "grad_norm": 6.1493240468982275,
+      "learning_rate": 5.626720295872911e-07,
+      "loss": 0.4074,
+      "step": 9712
+    },
+    {
+      "epoch": 0.7888410622918867,
+      "grad_norm": 3.943608994312866,
+      "learning_rate": 5.622564494096122e-07,
+      "loss": 0.4565,
+      "step": 9713
+    },
+    {
+      "epoch": 0.7889222772679282,
+      "grad_norm": 7.779376580705911,
+      "learning_rate": 5.618410033135325e-07,
+      "loss": 0.5792,
+      "step": 9714
+    },
+    {
+      "epoch": 0.7890034922439698,
+      "grad_norm": 3.3797672800758134,
+      "learning_rate": 5.614256913277968e-07,
+      "loss": 0.3237,
+      "step": 9715
+    },
+    {
+      "epoch": 0.7890847072200113,
+      "grad_norm": 5.330674659642353,
+      "learning_rate": 5.610105134811444e-07,
+      "loss": 0.4209,
+      "step": 9716
+    },
+    {
+      "epoch": 0.789165922196053,
+      "grad_norm": 4.322221219959529,
+      "learning_rate": 5.605954698023023e-07,
+      "loss": 0.4035,
+      "step": 9717
+    },
+    {
+      "epoch": 0.7892471371720945,
+      "grad_norm": 9.488626438377107,
+      "learning_rate": 5.601805603199889e-07,
+      "loss": 0.3763,
+      "step": 9718
+    },
+    {
+      "epoch": 0.7893283521481361,
+      "grad_norm": 5.681079603298455,
+      "learning_rate": 5.597657850629145e-07,
+      "loss": 0.5129,
+      "step": 9719
+    },
+    {
+      "epoch": 0.7894095671241776,
+      "grad_norm": 4.177871819811436,
+      "learning_rate": 5.593511440597799e-07,
+      "loss": 0.4432,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7894907821002193,
+      "grad_norm": 7.924561057142348,
+      "learning_rate": 5.589366373392754e-07,
+      "loss": 0.4561,
+      "step": 9721
+    },
+    {
+      "epoch": 0.7895719970762609,
+      "grad_norm": 3.32290128813931,
+      "learning_rate": 5.58522264930082e-07,
+      "loss": 0.5691,
+      "step": 9722
+    },
+    {
+      "epoch": 0.7896532120523024,
+      "grad_norm": 7.189607322217405,
+      "learning_rate": 5.581080268608733e-07,
+      "loss": 0.5209,
+      "step": 9723
+    },
+    {
+      "epoch": 0.7897344270283441,
+      "grad_norm": 6.847085179945522,
+      "learning_rate": 5.576939231603118e-07,
+      "loss": 0.533,
+      "step": 9724
+    },
+    {
+      "epoch": 0.7898156420043856,
+      "grad_norm": 5.441399177870424,
+      "learning_rate": 5.572799538570506e-07,
+      "loss": 0.3767,
+      "step": 9725
+    },
+    {
+      "epoch": 0.7898968569804272,
+      "grad_norm": 5.540677770063936,
+      "learning_rate": 5.56866118979735e-07,
+      "loss": 0.5119,
+      "step": 9726
+    },
+    {
+      "epoch": 0.7899780719564687,
+      "grad_norm": 8.524810246072764,
+      "learning_rate": 5.564524185570008e-07,
+      "loss": 0.5606,
+      "step": 9727
+    },
+    {
+      "epoch": 0.7900592869325104,
+      "grad_norm": 3.4073724271339327,
+      "learning_rate": 5.560388526174723e-07,
+      "loss": 0.6541,
+      "step": 9728
+    },
+    {
+      "epoch": 0.7901405019085519,
+      "grad_norm": 4.045724757251306,
+      "learning_rate": 5.556254211897677e-07,
+      "loss": 0.481,
+      "step": 9729
+    },
+    {
+      "epoch": 0.7902217168845935,
+      "grad_norm": 6.886529740016336,
+      "learning_rate": 5.552121243024935e-07,
+      "loss": 0.456,
+      "step": 9730
+    },
+    {
+      "epoch": 0.790302931860635,
+      "grad_norm": 4.136880017146439,
+      "learning_rate": 5.54798961984247e-07,
+      "loss": 0.4138,
+      "step": 9731
+    },
+    {
+      "epoch": 0.7903841468366767,
+      "grad_norm": 5.315316660423206,
+      "learning_rate": 5.543859342636177e-07,
+      "loss": 0.4179,
+      "step": 9732
+    },
+    {
+      "epoch": 0.7904653618127183,
+      "grad_norm": 9.808582227554735,
+      "learning_rate": 5.539730411691851e-07,
+      "loss": 0.4608,
+      "step": 9733
+    },
+    {
+      "epoch": 0.7905465767887598,
+      "grad_norm": 3.7605530791600725,
+      "learning_rate": 5.535602827295189e-07,
+      "loss": 0.4465,
+      "step": 9734
+    },
+    {
+      "epoch": 0.7906277917648015,
+      "grad_norm": 5.688896710047547,
+      "learning_rate": 5.53147658973179e-07,
+      "loss": 0.3428,
+      "step": 9735
+    },
+    {
+      "epoch": 0.790709006740843,
+      "grad_norm": 11.37494567313138,
+      "learning_rate": 5.527351699287184e-07,
+      "loss": 0.4684,
+      "step": 9736
+    },
+    {
+      "epoch": 0.7907902217168846,
+      "grad_norm": 6.335210866254778,
+      "learning_rate": 5.523228156246782e-07,
+      "loss": 0.5216,
+      "step": 9737
+    },
+    {
+      "epoch": 0.7908714366929261,
+      "grad_norm": 3.543537249807746,
+      "learning_rate": 5.519105960895904e-07,
+      "loss": 0.6721,
+      "step": 9738
+    },
+    {
+      "epoch": 0.7909526516689678,
+      "grad_norm": 3.746923394479714,
+      "learning_rate": 5.514985113519794e-07,
+      "loss": 0.5059,
+      "step": 9739
+    },
+    {
+      "epoch": 0.7910338666450093,
+      "grad_norm": 3.243317942178259,
+      "learning_rate": 5.510865614403599e-07,
+      "loss": 0.5028,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7911150816210509,
+      "grad_norm": 5.231091301479268,
+      "learning_rate": 5.506747463832348e-07,
+      "loss": 0.506,
+      "step": 9741
+    },
+    {
+      "epoch": 0.7911962965970925,
+      "grad_norm": 3.037936622624663,
+      "learning_rate": 5.502630662091016e-07,
+      "loss": 0.4508,
+      "step": 9742
+    },
+    {
+      "epoch": 0.7912775115731341,
+      "grad_norm": 5.011070638322092,
+      "learning_rate": 5.498515209464453e-07,
+      "loss": 0.5612,
+      "step": 9743
+    },
+    {
+      "epoch": 0.7913587265491757,
+      "grad_norm": 57.93538906369309,
+      "learning_rate": 5.49440110623742e-07,
+      "loss": 0.4245,
+      "step": 9744
+    },
+    {
+      "epoch": 0.7914399415252172,
+      "grad_norm": 4.391572013936799,
+      "learning_rate": 5.490288352694598e-07,
+      "loss": 0.4482,
+      "step": 9745
+    },
+    {
+      "epoch": 0.7915211565012589,
+      "grad_norm": 4.558873733467807,
+      "learning_rate": 5.486176949120575e-07,
+      "loss": 0.6039,
+      "step": 9746
+    },
+    {
+      "epoch": 0.7916023714773004,
+      "grad_norm": 7.119700732402204,
+      "learning_rate": 5.482066895799825e-07,
+      "loss": 0.467,
+      "step": 9747
+    },
+    {
+      "epoch": 0.791683586453342,
+      "grad_norm": 9.302916508581008,
+      "learning_rate": 5.477958193016758e-07,
+      "loss": 0.589,
+      "step": 9748
+    },
+    {
+      "epoch": 0.7917648014293835,
+      "grad_norm": 3.810258704896879,
+      "learning_rate": 5.473850841055664e-07,
+      "loss": 0.5026,
+      "step": 9749
+    },
+    {
+      "epoch": 0.7918460164054252,
+      "grad_norm": 3.970336831775221,
+      "learning_rate": 5.469744840200741e-07,
+      "loss": 0.4808,
+      "step": 9750
+    },
+    {
+      "epoch": 0.7919272313814667,
+      "grad_norm": 4.344838984297026,
+      "learning_rate": 5.465640190736124e-07,
+      "loss": 0.5038,
+      "step": 9751
+    },
+    {
+      "epoch": 0.7920084463575083,
+      "grad_norm": 4.912118996595724,
+      "learning_rate": 5.461536892945812e-07,
+      "loss": 0.4581,
+      "step": 9752
+    },
+    {
+      "epoch": 0.7920896613335499,
+      "grad_norm": 5.652749892181998,
+      "learning_rate": 5.457434947113749e-07,
+      "loss": 0.4009,
+      "step": 9753
+    },
+    {
+      "epoch": 0.7921708763095915,
+      "grad_norm": 17.57147989196119,
+      "learning_rate": 5.453334353523754e-07,
+      "loss": 0.4287,
+      "step": 9754
+    },
+    {
+      "epoch": 0.7922520912856331,
+      "grad_norm": 3.898615950581783,
+      "learning_rate": 5.449235112459577e-07,
+      "loss": 0.4907,
+      "step": 9755
+    },
+    {
+      "epoch": 0.7923333062616746,
+      "grad_norm": 6.661559536211671,
+      "learning_rate": 5.445137224204861e-07,
+      "loss": 0.6232,
+      "step": 9756
+    },
+    {
+      "epoch": 0.7924145212377163,
+      "grad_norm": 4.6703720755130265,
+      "learning_rate": 5.441040689043148e-07,
+      "loss": 0.3485,
+      "step": 9757
+    },
+    {
+      "epoch": 0.7924957362137578,
+      "grad_norm": 7.3100697341525995,
+      "learning_rate": 5.436945507257907e-07,
+      "loss": 0.485,
+      "step": 9758
+    },
+    {
+      "epoch": 0.7925769511897994,
+      "grad_norm": 5.877076789096071,
+      "learning_rate": 5.432851679132506e-07,
+      "loss": 0.5487,
+      "step": 9759
+    },
+    {
+      "epoch": 0.792658166165841,
+      "grad_norm": 6.67706951439948,
+      "learning_rate": 5.428759204950204e-07,
+      "loss": 0.8112,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7927393811418826,
+      "grad_norm": 5.439821280458846,
+      "learning_rate": 5.424668084994195e-07,
+      "loss": 0.5156,
+      "step": 9761
+    },
+    {
+      "epoch": 0.7928205961179241,
+      "grad_norm": 4.572295130779405,
+      "learning_rate": 5.420578319547551e-07,
+      "loss": 0.3297,
+      "step": 9762
+    },
+    {
+      "epoch": 0.7929018110939657,
+      "grad_norm": 5.618034848926441,
+      "learning_rate": 5.416489908893258e-07,
+      "loss": 0.5123,
+      "step": 9763
+    },
+    {
+      "epoch": 0.7929830260700073,
+      "grad_norm": 5.221377474566307,
+      "learning_rate": 5.412402853314227e-07,
+      "loss": 0.3917,
+      "step": 9764
+    },
+    {
+      "epoch": 0.7930642410460489,
+      "grad_norm": 10.356233466887792,
+      "learning_rate": 5.408317153093245e-07,
+      "loss": 0.6027,
+      "step": 9765
+    },
+    {
+      "epoch": 0.7931454560220905,
+      "grad_norm": 4.822402093871006,
+      "learning_rate": 5.404232808513027e-07,
+      "loss": 0.5062,
+      "step": 9766
+    },
+    {
+      "epoch": 0.793226670998132,
+      "grad_norm": 4.342762770694148,
+      "learning_rate": 5.400149819856199e-07,
+      "loss": 0.4948,
+      "step": 9767
+    },
+    {
+      "epoch": 0.7933078859741737,
+      "grad_norm": 5.561331791979438,
+      "learning_rate": 5.396068187405273e-07,
+      "loss": 0.4032,
+      "step": 9768
+    },
+    {
+      "epoch": 0.7933891009502152,
+      "grad_norm": 8.670840918114324,
+      "learning_rate": 5.391987911442667e-07,
+      "loss": 0.613,
+      "step": 9769
+    },
+    {
+      "epoch": 0.7934703159262568,
+      "grad_norm": 4.613030257259587,
+      "learning_rate": 5.387908992250731e-07,
+      "loss": 0.3999,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7935515309022984,
+      "grad_norm": 5.329616352899454,
+      "learning_rate": 5.383831430111691e-07,
+      "loss": 0.4021,
+      "step": 9771
+    },
+    {
+      "epoch": 0.79363274587834,
+      "grad_norm": 5.051235158011703,
+      "learning_rate": 5.379755225307707e-07,
+      "loss": 0.6117,
+      "step": 9772
+    },
+    {
+      "epoch": 0.7937139608543815,
+      "grad_norm": 4.393658723431004,
+      "learning_rate": 5.375680378120812e-07,
+      "loss": 0.5466,
+      "step": 9773
+    },
+    {
+      "epoch": 0.7937951758304231,
+      "grad_norm": 8.092758743122241,
+      "learning_rate": 5.371606888832984e-07,
+      "loss": 0.4675,
+      "step": 9774
+    },
+    {
+      "epoch": 0.7938763908064647,
+      "grad_norm": 4.672078958128614,
+      "learning_rate": 5.367534757726079e-07,
+      "loss": 0.5704,
+      "step": 9775
+    },
+    {
+      "epoch": 0.7939576057825063,
+      "grad_norm": 5.017160554885112,
+      "learning_rate": 5.363463985081854e-07,
+      "loss": 0.5743,
+      "step": 9776
+    },
+    {
+      "epoch": 0.7940388207585479,
+      "grad_norm": 3.292901108068312,
+      "learning_rate": 5.359394571182e-07,
+      "loss": 0.5656,
+      "step": 9777
+    },
+    {
+      "epoch": 0.7941200357345894,
+      "grad_norm": 5.959333091031521,
+      "learning_rate": 5.355326516308102e-07,
+      "loss": 0.484,
+      "step": 9778
+    },
+    {
+      "epoch": 0.7942012507106311,
+      "grad_norm": 3.555927641182116,
+      "learning_rate": 5.351259820741633e-07,
+      "loss": 0.5468,
+      "step": 9779
+    },
+    {
+      "epoch": 0.7942824656866726,
+      "grad_norm": 9.33015466014825,
+      "learning_rate": 5.347194484764001e-07,
+      "loss": 0.5981,
+      "step": 9780
+    },
+    {
+      "epoch": 0.7943636806627142,
+      "grad_norm": 5.675486153355801,
+      "learning_rate": 5.343130508656502e-07,
+      "loss": 0.5224,
+      "step": 9781
+    },
+    {
+      "epoch": 0.7944448956387558,
+      "grad_norm": 4.362949089225925,
+      "learning_rate": 5.339067892700331e-07,
+      "loss": 0.4188,
+      "step": 9782
+    },
+    {
+      "epoch": 0.7945261106147974,
+      "grad_norm": 4.529916181314969,
+      "learning_rate": 5.335006637176612e-07,
+      "loss": 0.4768,
+      "step": 9783
+    },
+    {
+      "epoch": 0.7946073255908389,
+      "grad_norm": 6.84665053039972,
+      "learning_rate": 5.330946742366356e-07,
+      "loss": 0.4384,
+      "step": 9784
+    },
+    {
+      "epoch": 0.7946885405668805,
+      "grad_norm": 5.58529418435716,
+      "learning_rate": 5.326888208550485e-07,
+      "loss": 0.3712,
+      "step": 9785
+    },
+    {
+      "epoch": 0.7947697555429221,
+      "grad_norm": 5.963815630595506,
+      "learning_rate": 5.32283103600984e-07,
+      "loss": 0.5674,
+      "step": 9786
+    },
+    {
+      "epoch": 0.7948509705189637,
+      "grad_norm": 5.054597490597483,
+      "learning_rate": 5.318775225025147e-07,
+      "loss": 0.5324,
+      "step": 9787
+    },
+    {
+      "epoch": 0.7949321854950053,
+      "grad_norm": 5.080349599593879,
+      "learning_rate": 5.314720775877046e-07,
+      "loss": 0.4342,
+      "step": 9788
+    },
+    {
+      "epoch": 0.7950134004710468,
+      "grad_norm": 4.048899362069842,
+      "learning_rate": 5.31066768884608e-07,
+      "loss": 0.3672,
+      "step": 9789
+    },
+    {
+      "epoch": 0.7950946154470885,
+      "grad_norm": 33.20617438804645,
+      "learning_rate": 5.306615964212705e-07,
+      "loss": 0.5476,
+      "step": 9790
+    },
+    {
+      "epoch": 0.79517583042313,
+      "grad_norm": 7.450168497762338,
+      "learning_rate": 5.302565602257285e-07,
+      "loss": 0.6486,
+      "step": 9791
+    },
+    {
+      "epoch": 0.7952570453991716,
+      "grad_norm": 7.689105107124658,
+      "learning_rate": 5.298516603260071e-07,
+      "loss": 0.3788,
+      "step": 9792
+    },
+    {
+      "epoch": 0.7953382603752132,
+      "grad_norm": 4.276239675857061,
+      "learning_rate": 5.294468967501248e-07,
+      "loss": 0.4255,
+      "step": 9793
+    },
+    {
+      "epoch": 0.7954194753512548,
+      "grad_norm": 5.366407144652406,
+      "learning_rate": 5.29042269526088e-07,
+      "loss": 0.5787,
+      "step": 9794
+    },
+    {
+      "epoch": 0.7955006903272963,
+      "grad_norm": 3.8340051916962503,
+      "learning_rate": 5.286377786818944e-07,
+      "loss": 0.6228,
+      "step": 9795
+    },
+    {
+      "epoch": 0.7955819053033379,
+      "grad_norm": 4.50626750141108,
+      "learning_rate": 5.282334242455339e-07,
+      "loss": 0.5358,
+      "step": 9796
+    },
+    {
+      "epoch": 0.7956631202793795,
+      "grad_norm": 6.044552408083624,
+      "learning_rate": 5.278292062449844e-07,
+      "loss": 0.5658,
+      "step": 9797
+    },
+    {
+      "epoch": 0.7957443352554211,
+      "grad_norm": 4.404044522106774,
+      "learning_rate": 5.274251247082163e-07,
+      "loss": 0.5208,
+      "step": 9798
+    },
+    {
+      "epoch": 0.7958255502314627,
+      "grad_norm": 7.333443652190031,
+      "learning_rate": 5.270211796631905e-07,
+      "loss": 0.3324,
+      "step": 9799
+    },
+    {
+      "epoch": 0.7959067652075043,
+      "grad_norm": 4.223546024424326,
+      "learning_rate": 5.266173711378572e-07,
+      "loss": 0.5734,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7959879801835459,
+      "grad_norm": 5.341487545794119,
+      "learning_rate": 5.262136991601572e-07,
+      "loss": 0.385,
+      "step": 9801
+    },
+    {
+      "epoch": 0.7960691951595874,
+      "grad_norm": 4.803777217199065,
+      "learning_rate": 5.258101637580238e-07,
+      "loss": 0.4812,
+      "step": 9802
+    },
+    {
+      "epoch": 0.796150410135629,
+      "grad_norm": 6.650426970575839,
+      "learning_rate": 5.254067649593781e-07,
+      "loss": 0.4336,
+      "step": 9803
+    },
+    {
+      "epoch": 0.7962316251116706,
+      "grad_norm": 5.441981893530827,
+      "learning_rate": 5.250035027921338e-07,
+      "loss": 0.3946,
+      "step": 9804
+    },
+    {
+      "epoch": 0.7963128400877122,
+      "grad_norm": 5.107486747720455,
+      "learning_rate": 5.246003772841953e-07,
+      "loss": 0.4776,
+      "step": 9805
+    },
+    {
+      "epoch": 0.7963940550637537,
+      "grad_norm": 3.5626105502835497,
+      "learning_rate": 5.24197388463456e-07,
+      "loss": 0.328,
+      "step": 9806
+    },
+    {
+      "epoch": 0.7964752700397953,
+      "grad_norm": 6.074463700483569,
+      "learning_rate": 5.237945363578006e-07,
+      "loss": 0.4275,
+      "step": 9807
+    },
+    {
+      "epoch": 0.7965564850158369,
+      "grad_norm": 12.182536694266675,
+      "learning_rate": 5.233918209951039e-07,
+      "loss": 0.3317,
+      "step": 9808
+    },
+    {
+      "epoch": 0.7966376999918785,
+      "grad_norm": 6.106295220054585,
+      "learning_rate": 5.229892424032326e-07,
+      "loss": 0.4446,
+      "step": 9809
+    },
+    {
+      "epoch": 0.7967189149679201,
+      "grad_norm": 4.589038843159026,
+      "learning_rate": 5.225868006100421e-07,
+      "loss": 0.4815,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7968001299439617,
+      "grad_norm": 13.701424714841433,
+      "learning_rate": 5.221844956433794e-07,
+      "loss": 0.3585,
+      "step": 9811
+    },
+    {
+      "epoch": 0.7968813449200033,
+      "grad_norm": 4.331772567167185,
+      "learning_rate": 5.21782327531083e-07,
+      "loss": 0.426,
+      "step": 9812
+    },
+    {
+      "epoch": 0.7969625598960448,
+      "grad_norm": 5.814768311032915,
+      "learning_rate": 5.213802963009798e-07,
+      "loss": 0.5894,
+      "step": 9813
+    },
+    {
+      "epoch": 0.7970437748720864,
+      "grad_norm": 4.7885091116888505,
+      "learning_rate": 5.209784019808877e-07,
+      "loss": 0.401,
+      "step": 9814
+    },
+    {
+      "epoch": 0.797124989848128,
+      "grad_norm": 3.7476316440320403,
+      "learning_rate": 5.205766445986174e-07,
+      "loss": 0.4015,
+      "step": 9815
+    },
+    {
+      "epoch": 0.7972062048241696,
+      "grad_norm": 4.916810203123699,
+      "learning_rate": 5.201750241819664e-07,
+      "loss": 0.3485,
+      "step": 9816
+    },
+    {
+      "epoch": 0.7972874198002111,
+      "grad_norm": 4.506048343346604,
+      "learning_rate": 5.197735407587257e-07,
+      "loss": 0.4305,
+      "step": 9817
+    },
+    {
+      "epoch": 0.7973686347762527,
+      "grad_norm": 3.516529711458559,
+      "learning_rate": 5.193721943566762e-07,
+      "loss": 0.5148,
+      "step": 9818
+    },
+    {
+      "epoch": 0.7974498497522943,
+      "grad_norm": 4.681381707890613,
+      "learning_rate": 5.189709850035887e-07,
+      "loss": 0.4342,
+      "step": 9819
+    },
+    {
+      "epoch": 0.7975310647283359,
+      "grad_norm": 5.7757941641398665,
+      "learning_rate": 5.185699127272243e-07,
+      "loss": 0.3261,
+      "step": 9820
+    },
+    {
+      "epoch": 0.7976122797043775,
+      "grad_norm": 3.0498451518722196,
+      "learning_rate": 5.181689775553355e-07,
+      "loss": 0.5548,
+      "step": 9821
+    },
+    {
+      "epoch": 0.7976934946804191,
+      "grad_norm": 7.950302605138208,
+      "learning_rate": 5.17768179515665e-07,
+      "loss": 0.3438,
+      "step": 9822
+    },
+    {
+      "epoch": 0.7977747096564607,
+      "grad_norm": 6.18464348555484,
+      "learning_rate": 5.173675186359451e-07,
+      "loss": 0.4751,
+      "step": 9823
+    },
+    {
+      "epoch": 0.7978559246325022,
+      "grad_norm": 7.611226667202139,
+      "learning_rate": 5.169669949438996e-07,
+      "loss": 0.3245,
+      "step": 9824
+    },
+    {
+      "epoch": 0.7979371396085438,
+      "grad_norm": 6.064790144719224,
+      "learning_rate": 5.165666084672439e-07,
+      "loss": 0.4096,
+      "step": 9825
+    },
+    {
+      "epoch": 0.7980183545845854,
+      "grad_norm": 4.770411960870846,
+      "learning_rate": 5.161663592336815e-07,
+      "loss": 0.6826,
+      "step": 9826
+    },
+    {
+      "epoch": 0.798099569560627,
+      "grad_norm": 7.597577840704956,
+      "learning_rate": 5.157662472709075e-07,
+      "loss": 0.5609,
+      "step": 9827
+    },
+    {
+      "epoch": 0.7981807845366685,
+      "grad_norm": 8.243596223621436,
+      "learning_rate": 5.153662726066083e-07,
+      "loss": 0.4814,
+      "step": 9828
+    },
+    {
+      "epoch": 0.7982619995127102,
+      "grad_norm": 9.676688199336681,
+      "learning_rate": 5.149664352684586e-07,
+      "loss": 0.5136,
+      "step": 9829
+    },
+    {
+      "epoch": 0.7983432144887517,
+      "grad_norm": 5.795108480675768,
+      "learning_rate": 5.14566735284126e-07,
+      "loss": 0.3747,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7984244294647933,
+      "grad_norm": 6.471676661405976,
+      "learning_rate": 5.141671726812683e-07,
+      "loss": 0.5808,
+      "step": 9831
+    },
+    {
+      "epoch": 0.7985056444408349,
+      "grad_norm": 4.438123959791211,
+      "learning_rate": 5.137677474875324e-07,
+      "loss": 0.5517,
+      "step": 9832
+    },
+    {
+      "epoch": 0.7985868594168765,
+      "grad_norm": 7.424549295374417,
+      "learning_rate": 5.133684597305557e-07,
+      "loss": 0.3537,
+      "step": 9833
+    },
+    {
+      "epoch": 0.7986680743929181,
+      "grad_norm": 5.275736853110713,
+      "learning_rate": 5.129693094379684e-07,
+      "loss": 0.3372,
+      "step": 9834
+    },
+    {
+      "epoch": 0.7987492893689596,
+      "grad_norm": 3.2450613994463056,
+      "learning_rate": 5.125702966373883e-07,
+      "loss": 0.3272,
+      "step": 9835
+    },
+    {
+      "epoch": 0.7988305043450012,
+      "grad_norm": 12.566077839285736,
+      "learning_rate": 5.121714213564249e-07,
+      "loss": 0.4256,
+      "step": 9836
+    },
+    {
+      "epoch": 0.7989117193210428,
+      "grad_norm": 3.457305067918944,
+      "learning_rate": 5.117726836226786e-07,
+      "loss": 0.4961,
+      "step": 9837
+    },
+    {
+      "epoch": 0.7989929342970844,
+      "grad_norm": 5.900089910314316,
+      "learning_rate": 5.113740834637407e-07,
+      "loss": 0.3549,
+      "step": 9838
+    },
+    {
+      "epoch": 0.7990741492731259,
+      "grad_norm": 8.54709344532498,
+      "learning_rate": 5.109756209071908e-07,
+      "loss": 0.469,
+      "step": 9839
+    },
+    {
+      "epoch": 0.7991553642491676,
+      "grad_norm": 3.7791884771682147,
+      "learning_rate": 5.105772959806021e-07,
+      "loss": 0.5095,
+      "step": 9840
+    },
+    {
+      "epoch": 0.7992365792252091,
+      "grad_norm": 6.060813178732931,
+      "learning_rate": 5.101791087115354e-07,
+      "loss": 0.5591,
+      "step": 9841
+    },
+    {
+      "epoch": 0.7993177942012507,
+      "grad_norm": 3.1018063396354605,
+      "learning_rate": 5.097810591275429e-07,
+      "loss": 0.3985,
+      "step": 9842
+    },
+    {
+      "epoch": 0.7993990091772923,
+      "grad_norm": 5.327968286749991,
+      "learning_rate": 5.093831472561681e-07,
+      "loss": 0.4773,
+      "step": 9843
+    },
+    {
+      "epoch": 0.7994802241533339,
+      "grad_norm": 7.284888223378119,
+      "learning_rate": 5.089853731249448e-07,
+      "loss": 0.4411,
+      "step": 9844
+    },
+    {
+      "epoch": 0.7995614391293755,
+      "grad_norm": 9.03698387870564,
+      "learning_rate": 5.085877367613964e-07,
+      "loss": 0.4432,
+      "step": 9845
+    },
+    {
+      "epoch": 0.799642654105417,
+      "grad_norm": 4.655253921964928,
+      "learning_rate": 5.081902381930365e-07,
+      "loss": 0.5223,
+      "step": 9846
+    },
+    {
+      "epoch": 0.7997238690814586,
+      "grad_norm": 6.893829661308056,
+      "learning_rate": 5.077928774473714e-07,
+      "loss": 0.3228,
+      "step": 9847
+    },
+    {
+      "epoch": 0.7998050840575002,
+      "grad_norm": 4.268542033494732,
+      "learning_rate": 5.073956545518949e-07,
+      "loss": 0.4231,
+      "step": 9848
+    },
+    {
+      "epoch": 0.7998862990335418,
+      "grad_norm": 14.574424300264582,
+      "learning_rate": 5.069985695340931e-07,
+      "loss": 0.4657,
+      "step": 9849
+    },
+    {
+      "epoch": 0.7999675140095833,
+      "grad_norm": 5.388314959320381,
+      "learning_rate": 5.066016224214435e-07,
+      "loss": 0.6228,
+      "step": 9850
+    },
+    {
+      "epoch": 0.800048728985625,
+      "grad_norm": 5.9704183797863575,
+      "learning_rate": 5.062048132414116e-07,
+      "loss": 0.4077,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8001299439616665,
+      "grad_norm": 7.054787471974678,
+      "learning_rate": 5.058081420214538e-07,
+      "loss": 0.394,
+      "step": 9852
+    },
+    {
+      "epoch": 0.8002111589377081,
+      "grad_norm": 9.477648614150743,
+      "learning_rate": 5.054116087890196e-07,
+      "loss": 0.5202,
+      "step": 9853
+    },
+    {
+      "epoch": 0.8002923739137497,
+      "grad_norm": 5.337615316923885,
+      "learning_rate": 5.050152135715453e-07,
+      "loss": 0.4712,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8003735888897913,
+      "grad_norm": 4.045630006380179,
+      "learning_rate": 5.046189563964595e-07,
+      "loss": 0.4629,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8004548038658329,
+      "grad_norm": 4.540887627720557,
+      "learning_rate": 5.042228372911815e-07,
+      "loss": 0.4685,
+      "step": 9856
+    },
+    {
+      "epoch": 0.8005360188418744,
+      "grad_norm": 5.474722039970322,
+      "learning_rate": 5.038268562831214e-07,
+      "loss": 0.4574,
+      "step": 9857
+    },
+    {
+      "epoch": 0.800617233817916,
+      "grad_norm": 4.262273306813277,
+      "learning_rate": 5.034310133996772e-07,
+      "loss": 0.6693,
+      "step": 9858
+    },
+    {
+      "epoch": 0.8006984487939576,
+      "grad_norm": 4.349294849404306,
+      "learning_rate": 5.030353086682413e-07,
+      "loss": 0.4987,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8007796637699992,
+      "grad_norm": 3.749623526864484,
+      "learning_rate": 5.02639742116193e-07,
+      "loss": 0.5335,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8008608787460407,
+      "grad_norm": 3.6511910926652926,
+      "learning_rate": 5.022443137709032e-07,
+      "loss": 0.5024,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8009420937220824,
+      "grad_norm": 7.167220329931862,
+      "learning_rate": 5.018490236597337e-07,
+      "loss": 0.7358,
+      "step": 9862
+    },
+    {
+      "epoch": 0.8010233086981239,
+      "grad_norm": 4.748703110219761,
+      "learning_rate": 5.014538718100373e-07,
+      "loss": 0.5198,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8011045236741655,
+      "grad_norm": 6.53063347027075,
+      "learning_rate": 5.01058858249156e-07,
+      "loss": 0.5499,
+      "step": 9864
+    },
+    {
+      "epoch": 0.8011857386502071,
+      "grad_norm": 4.349589259638556,
+      "learning_rate": 5.006639830044219e-07,
+      "loss": 0.4342,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8012669536262487,
+      "grad_norm": 4.143371415401599,
+      "learning_rate": 5.002692461031591e-07,
+      "loss": 0.5895,
+      "step": 9866
+    },
+    {
+      "epoch": 0.8013481686022903,
+      "grad_norm": 7.7254724395274055,
+      "learning_rate": 4.998746475726815e-07,
+      "loss": 0.5336,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8014293835783318,
+      "grad_norm": 5.175670084157273,
+      "learning_rate": 4.994801874402918e-07,
+      "loss": 0.6204,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8015105985543735,
+      "grad_norm": 6.735892978669596,
+      "learning_rate": 4.990858657332856e-07,
+      "loss": 0.3911,
+      "step": 9869
+    },
+    {
+      "epoch": 0.801591813530415,
+      "grad_norm": 5.104643492256398,
+      "learning_rate": 4.986916824789484e-07,
+      "loss": 0.5035,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8016730285064566,
+      "grad_norm": 3.5531328145211476,
+      "learning_rate": 4.982976377045546e-07,
+      "loss": 0.4066,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8017542434824981,
+      "grad_norm": 4.292790999663058,
+      "learning_rate": 4.979037314373708e-07,
+      "loss": 0.7253,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8018354584585398,
+      "grad_norm": 3.9126112690178934,
+      "learning_rate": 4.975099637046529e-07,
+      "loss": 0.5173,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8019166734345813,
+      "grad_norm": 4.952119344921789,
+      "learning_rate": 4.971163345336469e-07,
+      "loss": 0.546,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8019978884106229,
+      "grad_norm": 4.007491175094282,
+      "learning_rate": 4.967228439515903e-07,
+      "loss": 0.5014,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8020791033866645,
+      "grad_norm": 7.4996807066549405,
+      "learning_rate": 4.963294919857115e-07,
+      "loss": 0.4807,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8021603183627061,
+      "grad_norm": 3.8718307958429308,
+      "learning_rate": 4.959362786632274e-07,
+      "loss": 0.3412,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8022415333387477,
+      "grad_norm": 14.309366650475116,
+      "learning_rate": 4.955432040113459e-07,
+      "loss": 0.4632,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8023227483147892,
+      "grad_norm": 6.072346159252271,
+      "learning_rate": 4.95150268057267e-07,
+      "loss": 0.3375,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8024039632908309,
+      "grad_norm": 3.9060655122044956,
+      "learning_rate": 4.947574708281788e-07,
+      "loss": 0.4548,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8024851782668724,
+      "grad_norm": 4.543113575494821,
+      "learning_rate": 4.943648123512607e-07,
+      "loss": 0.5799,
+      "step": 9881
+    },
+    {
+      "epoch": 0.802566393242914,
+      "grad_norm": 8.154618850928907,
+      "learning_rate": 4.939722926536825e-07,
+      "loss": 0.3684,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8026476082189555,
+      "grad_norm": 3.3302480665945673,
+      "learning_rate": 4.935799117626058e-07,
+      "loss": 0.5121,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8027288231949972,
+      "grad_norm": 5.010095824925431,
+      "learning_rate": 4.931876697051797e-07,
+      "loss": 0.4584,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8028100381710387,
+      "grad_norm": 8.290730780611321,
+      "learning_rate": 4.927955665085466e-07,
+      "loss": 0.5904,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8028912531470803,
+      "grad_norm": 6.530421557967707,
+      "learning_rate": 4.924036021998372e-07,
+      "loss": 0.6501,
+      "step": 9886
+    },
+    {
+      "epoch": 0.802972468123122,
+      "grad_norm": 3.933371390397304,
+      "learning_rate": 4.92011776806173e-07,
+      "loss": 0.4631,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8030536830991635,
+      "grad_norm": 4.630936993671216,
+      "learning_rate": 4.916200903546664e-07,
+      "loss": 0.4694,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8031348980752051,
+      "grad_norm": 8.920541461832672,
+      "learning_rate": 4.912285428724214e-07,
+      "loss": 0.4291,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8032161130512466,
+      "grad_norm": 4.683558801065991,
+      "learning_rate": 4.908371343865289e-07,
+      "loss": 0.5715,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8032973280272883,
+      "grad_norm": 4.064081828276734,
+      "learning_rate": 4.904458649240742e-07,
+      "loss": 0.6187,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8033785430033298,
+      "grad_norm": 13.844068111838718,
+      "learning_rate": 4.900547345121304e-07,
+      "loss": 0.4852,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8034597579793714,
+      "grad_norm": 7.334487582827237,
+      "learning_rate": 4.896637431777607e-07,
+      "loss": 0.5241,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8035409729554129,
+      "grad_norm": 4.075794059330378,
+      "learning_rate": 4.89272890948021e-07,
+      "loss": 0.4231,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8036221879314546,
+      "grad_norm": 3.1839585809732913,
+      "learning_rate": 4.88882177849955e-07,
+      "loss": 0.3873,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8037034029074961,
+      "grad_norm": 3.9421773033440233,
+      "learning_rate": 4.884916039105994e-07,
+      "loss": 0.4194,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8037846178835377,
+      "grad_norm": 4.937157672903017,
+      "learning_rate": 4.881011691569781e-07,
+      "loss": 0.3529,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8038658328595794,
+      "grad_norm": 9.136038936527344,
+      "learning_rate": 4.877108736161091e-07,
+      "loss": 0.5727,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8039470478356209,
+      "grad_norm": 4.5845637025408,
+      "learning_rate": 4.873207173149974e-07,
+      "loss": 0.608,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8040282628116625,
+      "grad_norm": 5.642715408471168,
+      "learning_rate": 4.869307002806397e-07,
+      "loss": 0.471,
+      "step": 9900
+    },
+    {
+      "epoch": 0.804109477787704,
+      "grad_norm": 6.2233329502733294,
+      "learning_rate": 4.865408225400234e-07,
+      "loss": 0.51,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8041906927637457,
+      "grad_norm": 7.993507725893066,
+      "learning_rate": 4.861510841201266e-07,
+      "loss": 0.3892,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8042719077397872,
+      "grad_norm": 4.96226956095677,
+      "learning_rate": 4.857614850479161e-07,
+      "loss": 0.5246,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8043531227158288,
+      "grad_norm": 11.603584675552659,
+      "learning_rate": 4.853720253503514e-07,
+      "loss": 0.4751,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8044343376918703,
+      "grad_norm": 4.342808900661657,
+      "learning_rate": 4.849827050543801e-07,
+      "loss": 0.5307,
+      "step": 9905
+    },
+    {
+      "epoch": 0.804515552667912,
+      "grad_norm": 6.60599795238213,
+      "learning_rate": 4.845935241869409e-07,
+      "loss": 0.4751,
+      "step": 9906
+    },
+    {
+      "epoch": 0.8045967676439535,
+      "grad_norm": 4.155426698257137,
+      "learning_rate": 4.842044827749632e-07,
+      "loss": 0.511,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8046779826199951,
+      "grad_norm": 5.569956854051911,
+      "learning_rate": 4.838155808453676e-07,
+      "loss": 0.6412,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8047591975960368,
+      "grad_norm": 8.778391998701604,
+      "learning_rate": 4.834268184250626e-07,
+      "loss": 0.3876,
+      "step": 9909
+    },
+    {
+      "epoch": 0.8048404125720783,
+      "grad_norm": 6.809131934976907,
+      "learning_rate": 4.830381955409497e-07,
+      "loss": 0.438,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8049216275481199,
+      "grad_norm": 6.2072199104772885,
+      "learning_rate": 4.826497122199191e-07,
+      "loss": 0.5043,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8050028425241614,
+      "grad_norm": 9.779692232356227,
+      "learning_rate": 4.822613684888519e-07,
+      "loss": 0.4799,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8050840575002031,
+      "grad_norm": 5.12858071171813,
+      "learning_rate": 4.818731643746186e-07,
+      "loss": 0.5888,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8051652724762446,
+      "grad_norm": 3.86454745630488,
+      "learning_rate": 4.814850999040816e-07,
+      "loss": 0.551,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8052464874522862,
+      "grad_norm": 4.189470752357414,
+      "learning_rate": 4.810971751040932e-07,
+      "loss": 0.4416,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8053277024283277,
+      "grad_norm": 8.768793347385872,
+      "learning_rate": 4.80709390001495e-07,
+      "loss": 0.3583,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8054089174043694,
+      "grad_norm": 5.814244672682726,
+      "learning_rate": 4.803217446231206e-07,
+      "loss": 0.5113,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8054901323804109,
+      "grad_norm": 4.585923445910094,
+      "learning_rate": 4.799342389957925e-07,
+      "loss": 0.4861,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8055713473564525,
+      "grad_norm": 5.831696175264736,
+      "learning_rate": 4.795468731463232e-07,
+      "loss": 0.5333,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8056525623324942,
+      "grad_norm": 4.199244928525552,
+      "learning_rate": 4.791596471015175e-07,
+      "loss": 0.5757,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8057337773085357,
+      "grad_norm": 4.550642371506264,
+      "learning_rate": 4.787725608881694e-07,
+      "loss": 0.5091,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8058149922845773,
+      "grad_norm": 6.904305095897743,
+      "learning_rate": 4.783856145330624e-07,
+      "loss": 0.3853,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8058962072606188,
+      "grad_norm": 5.386276895757226,
+      "learning_rate": 4.779988080629722e-07,
+      "loss": 0.547,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8059774222366605,
+      "grad_norm": 5.592802881169542,
+      "learning_rate": 4.776121415046634e-07,
+      "loss": 0.6014,
+      "step": 9924
+    },
+    {
+      "epoch": 0.806058637212702,
+      "grad_norm": 6.5204999954371825,
+      "learning_rate": 4.772256148848903e-07,
+      "loss": 0.5031,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8061398521887436,
+      "grad_norm": 4.641348135597619,
+      "learning_rate": 4.768392282303999e-07,
+      "loss": 0.4754,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8062210671647851,
+      "grad_norm": 6.611205228762905,
+      "learning_rate": 4.7645298156792667e-07,
+      "loss": 0.6335,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8063022821408268,
+      "grad_norm": 4.080391885424033,
+      "learning_rate": 4.7606687492419785e-07,
+      "loss": 0.5242,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8063834971168683,
+      "grad_norm": 12.801912551363564,
+      "learning_rate": 4.7568090832593033e-07,
+      "loss": 0.4403,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8064647120929099,
+      "grad_norm": 6.118657536304109,
+      "learning_rate": 4.752950817998303e-07,
+      "loss": 0.505,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8065459270689516,
+      "grad_norm": 5.362943206753638,
+      "learning_rate": 4.7490939537259527e-07,
+      "loss": 0.4619,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8066271420449931,
+      "grad_norm": 6.975033167992144,
+      "learning_rate": 4.745238490709117e-07,
+      "loss": 0.3267,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8067083570210347,
+      "grad_norm": 4.258954432113065,
+      "learning_rate": 4.741384429214579e-07,
+      "loss": 0.4316,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8067895719970762,
+      "grad_norm": 3.8337908412938972,
+      "learning_rate": 4.7375317695090295e-07,
+      "loss": 0.5411,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8068707869731179,
+      "grad_norm": 3.8911807697106213,
+      "learning_rate": 4.7336805118590375e-07,
+      "loss": 0.4551,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8069520019491594,
+      "grad_norm": 3.9612535233371764,
+      "learning_rate": 4.729830656531101e-07,
+      "loss": 0.5383,
+      "step": 9936
+    },
+    {
+      "epoch": 0.807033216925201,
+      "grad_norm": 4.951917020944689,
+      "learning_rate": 4.725982203791607e-07,
+      "loss": 0.6442,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8071144319012425,
+      "grad_norm": 9.313306419072807,
+      "learning_rate": 4.7221351539068374e-07,
+      "loss": 0.4088,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8071956468772842,
+      "grad_norm": 5.792989399998622,
+      "learning_rate": 4.7182895071430036e-07,
+      "loss": 0.4883,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8072768618533257,
+      "grad_norm": 6.612743860989568,
+      "learning_rate": 4.7144452637661875e-07,
+      "loss": 0.522,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8073580768293673,
+      "grad_norm": 5.3380622522214916,
+      "learning_rate": 4.7106024240424014e-07,
+      "loss": 0.5033,
+      "step": 9941
+    },
+    {
+      "epoch": 0.807439291805409,
+      "grad_norm": 10.643012463139256,
+      "learning_rate": 4.706760988237555e-07,
+      "loss": 0.4336,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8075205067814505,
+      "grad_norm": 6.794117107098659,
+      "learning_rate": 4.702920956617446e-07,
+      "loss": 0.4817,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8076017217574921,
+      "grad_norm": 10.634630209323328,
+      "learning_rate": 4.6990823294477795e-07,
+      "loss": 0.3929,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8076829367335336,
+      "grad_norm": 6.387003003213154,
+      "learning_rate": 4.695245106994181e-07,
+      "loss": 0.4227,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8077641517095753,
+      "grad_norm": 5.885749845817612,
+      "learning_rate": 4.691409289522156e-07,
+      "loss": 0.4674,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8078453666856168,
+      "grad_norm": 5.744010803243097,
+      "learning_rate": 4.6875748772971244e-07,
+      "loss": 0.6015,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8079265816616584,
+      "grad_norm": 3.2133616441041704,
+      "learning_rate": 4.683741870584413e-07,
+      "loss": 0.4788,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8080077966377,
+      "grad_norm": 6.807257172917265,
+      "learning_rate": 4.679910269649246e-07,
+      "loss": 0.5901,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8080890116137416,
+      "grad_norm": 3.9285075156101823,
+      "learning_rate": 4.676080074756745e-07,
+      "loss": 0.5877,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8081702265897831,
+      "grad_norm": 5.276676978089501,
+      "learning_rate": 4.6722512861719304e-07,
+      "loss": 0.5472,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8082514415658247,
+      "grad_norm": 5.815263402068283,
+      "learning_rate": 4.6684239041597524e-07,
+      "loss": 0.6121,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8083326565418664,
+      "grad_norm": 8.366540950724946,
+      "learning_rate": 4.6645979289850316e-07,
+      "loss": 0.3672,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8084138715179079,
+      "grad_norm": 4.915630857560634,
+      "learning_rate": 4.66077336091251e-07,
+      "loss": 0.4365,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8084950864939495,
+      "grad_norm": 14.896314047238715,
+      "learning_rate": 4.6569502002068336e-07,
+      "loss": 0.3652,
+      "step": 9955
+    },
+    {
+      "epoch": 0.808576301469991,
+      "grad_norm": 5.45782242106025,
+      "learning_rate": 4.6531284471325375e-07,
+      "loss": 0.4285,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8086575164460327,
+      "grad_norm": 4.550708998274069,
+      "learning_rate": 4.649308101954064e-07,
+      "loss": 0.5137,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8087387314220742,
+      "grad_norm": 4.5179592494433125,
+      "learning_rate": 4.645489164935774e-07,
+      "loss": 0.4994,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8088199463981158,
+      "grad_norm": 8.186932272267061,
+      "learning_rate": 4.641671636341899e-07,
+      "loss": 0.4727,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8089011613741574,
+      "grad_norm": 7.7788457747580635,
+      "learning_rate": 4.637855516436604e-07,
+      "loss": 0.4331,
+      "step": 9960
+    },
+    {
+      "epoch": 0.808982376350199,
+      "grad_norm": 9.523483750496569,
+      "learning_rate": 4.634040805483947e-07,
+      "loss": 0.4042,
+      "step": 9961
+    },
+    {
+      "epoch": 0.8090635913262405,
+      "grad_norm": 4.262182822939964,
+      "learning_rate": 4.6302275037478804e-07,
+      "loss": 0.6089,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8091448063022821,
+      "grad_norm": 5.01145075636782,
+      "learning_rate": 4.6264156114922605e-07,
+      "loss": 0.6344,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8092260212783238,
+      "grad_norm": 6.573486459689827,
+      "learning_rate": 4.622605128980862e-07,
+      "loss": 0.3984,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8093072362543653,
+      "grad_norm": 6.380655374815469,
+      "learning_rate": 4.61879605647734e-07,
+      "loss": 0.3601,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8093884512304069,
+      "grad_norm": 11.839284886283753,
+      "learning_rate": 4.6149883942452595e-07,
+      "loss": 0.626,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8094696662064484,
+      "grad_norm": 6.373110585879932,
+      "learning_rate": 4.6111821425480956e-07,
+      "loss": 0.7225,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8095508811824901,
+      "grad_norm": 7.7580649386331455,
+      "learning_rate": 4.6073773016492267e-07,
+      "loss": 0.5471,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8096320961585316,
+      "grad_norm": 4.824098525076564,
+      "learning_rate": 4.603573871811923e-07,
+      "loss": 0.5152,
+      "step": 9969
+    },
+    {
+      "epoch": 0.8097133111345732,
+      "grad_norm": 2.881153799640226,
+      "learning_rate": 4.5997718532993535e-07,
+      "loss": 0.5088,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8097945261106148,
+      "grad_norm": 6.514585014658824,
+      "learning_rate": 4.5959712463746144e-07,
+      "loss": 0.4321,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8098757410866564,
+      "grad_norm": 7.85731483603247,
+      "learning_rate": 4.5921720513006697e-07,
+      "loss": 0.3917,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8099569560626979,
+      "grad_norm": 5.1598024248620495,
+      "learning_rate": 4.588374268340412e-07,
+      "loss": 0.5726,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8100381710387395,
+      "grad_norm": 4.512666934916995,
+      "learning_rate": 4.584577897756634e-07,
+      "loss": 0.5283,
+      "step": 9974
+    },
+    {
+      "epoch": 0.8101193860147812,
+      "grad_norm": 8.411566127602082,
+      "learning_rate": 4.58078293981202e-07,
+      "loss": 0.4524,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8102006009908227,
+      "grad_norm": 8.15739766766053,
+      "learning_rate": 4.5769893947691517e-07,
+      "loss": 0.4705,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8102818159668643,
+      "grad_norm": 6.8035924289379235,
+      "learning_rate": 4.5731972628905357e-07,
+      "loss": 0.4252,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8103630309429058,
+      "grad_norm": 4.227519769293889,
+      "learning_rate": 4.5694065444385564e-07,
+      "loss": 0.5061,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8104442459189475,
+      "grad_norm": 6.581692909390787,
+      "learning_rate": 4.5656172396755156e-07,
+      "loss": 0.4785,
+      "step": 9979
+    },
+    {
+      "epoch": 0.810525460894989,
+      "grad_norm": 4.606311339058521,
+      "learning_rate": 4.561829348863622e-07,
+      "loss": 0.5214,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8106066758710306,
+      "grad_norm": 6.277335985489802,
+      "learning_rate": 4.55804287226497e-07,
+      "loss": 0.2991,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8106878908470722,
+      "grad_norm": 5.904104493248545,
+      "learning_rate": 4.5542578101415576e-07,
+      "loss": 0.4301,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8107691058231138,
+      "grad_norm": 3.728461827087679,
+      "learning_rate": 4.550474162755303e-07,
+      "loss": 0.6069,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8108503207991553,
+      "grad_norm": 4.995689618250689,
+      "learning_rate": 4.546691930368008e-07,
+      "loss": 0.4381,
+      "step": 9984
+    },
+    {
+      "epoch": 0.8109315357751969,
+      "grad_norm": 6.939274659064955,
+      "learning_rate": 4.5429111132413773e-07,
+      "loss": 0.5178,
+      "step": 9985
+    },
+    {
+      "epoch": 0.8110127507512386,
+      "grad_norm": 6.822025362372183,
+      "learning_rate": 4.539131711637032e-07,
+      "loss": 0.3963,
+      "step": 9986
+    },
+    {
+      "epoch": 0.8110939657272801,
+      "grad_norm": 6.644364425158026,
+      "learning_rate": 4.535353725816488e-07,
+      "loss": 0.504,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8111751807033217,
+      "grad_norm": 5.480802203291707,
+      "learning_rate": 4.5315771560411617e-07,
+      "loss": 0.4743,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8112563956793633,
+      "grad_norm": 5.912237232094158,
+      "learning_rate": 4.5278020025723596e-07,
+      "loss": 0.4656,
+      "step": 9989
+    },
+    {
+      "epoch": 0.8113376106554049,
+      "grad_norm": 5.0776992221582935,
+      "learning_rate": 4.524028265671318e-07,
+      "loss": 0.4281,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8114188256314464,
+      "grad_norm": 5.3270257232616025,
+      "learning_rate": 4.5202559455991473e-07,
+      "loss": 0.4505,
+      "step": 9991
+    },
+    {
+      "epoch": 0.811500040607488,
+      "grad_norm": 6.190730407880878,
+      "learning_rate": 4.516485042616878e-07,
+      "loss": 0.4437,
+      "step": 9992
+    },
+    {
+      "epoch": 0.8115812555835296,
+      "grad_norm": 5.542928858101199,
+      "learning_rate": 4.512715556985442e-07,
+      "loss": 0.4011,
+      "step": 9993
+    },
+    {
+      "epoch": 0.8116624705595712,
+      "grad_norm": 7.642739194941673,
+      "learning_rate": 4.508947488965662e-07,
+      "loss": 0.4349,
+      "step": 9994
+    },
+    {
+      "epoch": 0.8117436855356127,
+      "grad_norm": 7.8072925191375395,
+      "learning_rate": 4.505180838818263e-07,
+      "loss": 0.5984,
+      "step": 9995
+    },
+    {
+      "epoch": 0.8118249005116543,
+      "grad_norm": 4.656078397528526,
+      "learning_rate": 4.501415606803888e-07,
+      "loss": 0.4646,
+      "step": 9996
+    },
+    {
+      "epoch": 0.811906115487696,
+      "grad_norm": 4.525381591463549,
+      "learning_rate": 4.4976517931830637e-07,
+      "loss": 0.4135,
+      "step": 9997
+    },
+    {
+      "epoch": 0.8119873304637375,
+      "grad_norm": 5.890129159202458,
+      "learning_rate": 4.4938893982162253e-07,
+      "loss": 0.5532,
+      "step": 9998
+    },
+    {
+      "epoch": 0.8120685454397791,
+      "grad_norm": 6.903081683256655,
+      "learning_rate": 4.4901284221637113e-07,
+      "loss": 0.5935,
+      "step": 9999
+    },
+    {
+      "epoch": 0.8121497604158207,
+      "grad_norm": 19.18119241326282,
+      "learning_rate": 4.48636886528577e-07,
+      "loss": 0.4778,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8122309753918623,
+      "grad_norm": 6.157895091798137,
+      "learning_rate": 4.482610727842532e-07,
+      "loss": 0.69,
+      "step": 10001
+    },
+    {
+      "epoch": 0.8123121903679038,
+      "grad_norm": 5.599909239938712,
+      "learning_rate": 4.47885401009405e-07,
+      "loss": 0.4685,
+      "step": 10002
+    },
+    {
+      "epoch": 0.8123934053439454,
+      "grad_norm": 6.484452032525161,
+      "learning_rate": 4.475098712300263e-07,
+      "loss": 0.5125,
+      "step": 10003
+    },
+    {
+      "epoch": 0.812474620319987,
+      "grad_norm": 8.535162150811436,
+      "learning_rate": 4.4713448347210114e-07,
+      "loss": 0.3792,
+      "step": 10004
+    },
+    {
+      "epoch": 0.8125558352960286,
+      "grad_norm": 13.187357588897884,
+      "learning_rate": 4.4675923776160533e-07,
+      "loss": 0.4935,
+      "step": 10005
+    },
+    {
+      "epoch": 0.8126370502720701,
+      "grad_norm": 7.720879346905562,
+      "learning_rate": 4.463841341245043e-07,
+      "loss": 0.6142,
+      "step": 10006
+    },
+    {
+      "epoch": 0.8127182652481117,
+      "grad_norm": 4.332760377433552,
+      "learning_rate": 4.460091725867524e-07,
+      "loss": 0.436,
+      "step": 10007
+    },
+    {
+      "epoch": 0.8127994802241534,
+      "grad_norm": 6.514614871896753,
+      "learning_rate": 4.456343531742946e-07,
+      "loss": 0.4141,
+      "step": 10008
+    },
+    {
+      "epoch": 0.8128806952001949,
+      "grad_norm": 5.3062919238625605,
+      "learning_rate": 4.4525967591306757e-07,
+      "loss": 0.4301,
+      "step": 10009
+    },
+    {
+      "epoch": 0.8129619101762365,
+      "grad_norm": 9.134817655594835,
+      "learning_rate": 4.448851408289964e-07,
+      "loss": 0.5738,
+      "step": 10010
+    },
+    {
+      "epoch": 0.8130431251522781,
+      "grad_norm": 4.873683627050256,
+      "learning_rate": 4.4451074794799627e-07,
+      "loss": 0.304,
+      "step": 10011
+    },
+    {
+      "epoch": 0.8131243401283197,
+      "grad_norm": 3.4718602879812437,
+      "learning_rate": 4.4413649729597386e-07,
+      "loss": 0.3923,
+      "step": 10012
+    },
+    {
+      "epoch": 0.8132055551043612,
+      "grad_norm": 4.570485265053012,
+      "learning_rate": 4.43762388898826e-07,
+      "loss": 0.4372,
+      "step": 10013
+    },
+    {
+      "epoch": 0.8132867700804028,
+      "grad_norm": 3.9431671288156664,
+      "learning_rate": 4.4338842278243784e-07,
+      "loss": 0.4041,
+      "step": 10014
+    },
+    {
+      "epoch": 0.8133679850564444,
+      "grad_norm": 9.625205916829707,
+      "learning_rate": 4.4301459897268695e-07,
+      "loss": 0.4428,
+      "step": 10015
+    },
+    {
+      "epoch": 0.813449200032486,
+      "grad_norm": 7.135399774793007,
+      "learning_rate": 4.426409174954391e-07,
+      "loss": 0.5457,
+      "step": 10016
+    },
+    {
+      "epoch": 0.8135304150085275,
+      "grad_norm": 7.241634226766735,
+      "learning_rate": 4.4226737837655106e-07,
+      "loss": 0.4478,
+      "step": 10017
+    },
+    {
+      "epoch": 0.8136116299845692,
+      "grad_norm": 5.474906261114699,
+      "learning_rate": 4.418939816418699e-07,
+      "loss": 0.4813,
+      "step": 10018
+    },
+    {
+      "epoch": 0.8136928449606108,
+      "grad_norm": 7.248288071920974,
+      "learning_rate": 4.4152072731723336e-07,
+      "loss": 0.5229,
+      "step": 10019
+    },
+    {
+      "epoch": 0.8137740599366523,
+      "grad_norm": 4.68601406629194,
+      "learning_rate": 4.411476154284683e-07,
+      "loss": 0.4383,
+      "step": 10020
+    },
+    {
+      "epoch": 0.8138552749126939,
+      "grad_norm": 8.548985633110162,
+      "learning_rate": 4.407746460013912e-07,
+      "loss": 0.434,
+      "step": 10021
+    },
+    {
+      "epoch": 0.8139364898887355,
+      "grad_norm": 3.598258827038912,
+      "learning_rate": 4.404018190618109e-07,
+      "loss": 0.6293,
+      "step": 10022
+    },
+    {
+      "epoch": 0.8140177048647771,
+      "grad_norm": 4.9162560687761525,
+      "learning_rate": 4.4002913463552457e-07,
+      "loss": 0.6338,
+      "step": 10023
+    },
+    {
+      "epoch": 0.8140989198408186,
+      "grad_norm": 5.176333776815743,
+      "learning_rate": 4.39656592748319e-07,
+      "loss": 0.4134,
+      "step": 10024
+    },
+    {
+      "epoch": 0.8141801348168602,
+      "grad_norm": 5.309773965001737,
+      "learning_rate": 4.392841934259731e-07,
+      "loss": 0.4122,
+      "step": 10025
+    },
+    {
+      "epoch": 0.8142613497929018,
+      "grad_norm": 4.204661866235209,
+      "learning_rate": 4.3891193669425567e-07,
+      "loss": 0.5778,
+      "step": 10026
+    },
+    {
+      "epoch": 0.8143425647689434,
+      "grad_norm": 5.654465245751199,
+      "learning_rate": 4.3853982257892335e-07,
+      "loss": 0.5514,
+      "step": 10027
+    },
+    {
+      "epoch": 0.8144237797449849,
+      "grad_norm": 4.506512234014459,
+      "learning_rate": 4.3816785110572554e-07,
+      "loss": 0.537,
+      "step": 10028
+    },
+    {
+      "epoch": 0.8145049947210266,
+      "grad_norm": 3.728172094291727,
+      "learning_rate": 4.3779602230040075e-07,
+      "loss": 0.4989,
+      "step": 10029
+    },
+    {
+      "epoch": 0.8145862096970682,
+      "grad_norm": 7.112854813373821,
+      "learning_rate": 4.3742433618867623e-07,
+      "loss": 0.4539,
+      "step": 10030
+    },
+    {
+      "epoch": 0.8146674246731097,
+      "grad_norm": 4.9529582527943745,
+      "learning_rate": 4.370527927962717e-07,
+      "loss": 0.6694,
+      "step": 10031
+    },
+    {
+      "epoch": 0.8147486396491513,
+      "grad_norm": 5.619899087759572,
+      "learning_rate": 4.366813921488966e-07,
+      "loss": 0.5285,
+      "step": 10032
+    },
+    {
+      "epoch": 0.8148298546251929,
+      "grad_norm": 5.707386464605084,
+      "learning_rate": 4.363101342722484e-07,
+      "loss": 0.4412,
+      "step": 10033
+    },
+    {
+      "epoch": 0.8149110696012345,
+      "grad_norm": 4.86389548489138,
+      "learning_rate": 4.359390191920176e-07,
+      "loss": 0.5749,
+      "step": 10034
+    },
+    {
+      "epoch": 0.814992284577276,
+      "grad_norm": 5.994879397108235,
+      "learning_rate": 4.35568046933883e-07,
+      "loss": 0.3602,
+      "step": 10035
+    },
+    {
+      "epoch": 0.8150734995533176,
+      "grad_norm": 3.0944128092213523,
+      "learning_rate": 4.3519721752351305e-07,
+      "loss": 0.448,
+      "step": 10036
+    },
+    {
+      "epoch": 0.8151547145293592,
+      "grad_norm": 6.293183554065846,
+      "learning_rate": 4.3482653098656764e-07,
+      "loss": 0.5155,
+      "step": 10037
+    },
+    {
+      "epoch": 0.8152359295054008,
+      "grad_norm": 6.726079260676008,
+      "learning_rate": 4.3445598734869725e-07,
+      "loss": 0.5044,
+      "step": 10038
+    },
+    {
+      "epoch": 0.8153171444814423,
+      "grad_norm": 5.929353123479193,
+      "learning_rate": 4.340855866355409e-07,
+      "loss": 0.3846,
+      "step": 10039
+    },
+    {
+      "epoch": 0.815398359457484,
+      "grad_norm": 4.346247885330788,
+      "learning_rate": 4.3371532887272747e-07,
+      "loss": 0.4431,
+      "step": 10040
+    },
+    {
+      "epoch": 0.8154795744335256,
+      "grad_norm": 6.564210241083292,
+      "learning_rate": 4.333452140858782e-07,
+      "loss": 0.4179,
+      "step": 10041
+    },
+    {
+      "epoch": 0.8155607894095671,
+      "grad_norm": 6.160516432704893,
+      "learning_rate": 4.3297524230060257e-07,
+      "loss": 0.4109,
+      "step": 10042
+    },
+    {
+      "epoch": 0.8156420043856087,
+      "grad_norm": 4.046212520398475,
+      "learning_rate": 4.326054135425001e-07,
+      "loss": 0.4727,
+      "step": 10043
+    },
+    {
+      "epoch": 0.8157232193616503,
+      "grad_norm": 4.21989184407862,
+      "learning_rate": 4.322357278371614e-07,
+      "loss": 0.5261,
+      "step": 10044
+    },
+    {
+      "epoch": 0.8158044343376919,
+      "grad_norm": 5.065472356667533,
+      "learning_rate": 4.3186618521016745e-07,
+      "loss": 0.4124,
+      "step": 10045
+    },
+    {
+      "epoch": 0.8158856493137334,
+      "grad_norm": 6.317472489620383,
+      "learning_rate": 4.314967856870872e-07,
+      "loss": 0.439,
+      "step": 10046
+    },
+    {
+      "epoch": 0.815966864289775,
+      "grad_norm": 6.622990894266353,
+      "learning_rate": 4.31127529293483e-07,
+      "loss": 0.2964,
+      "step": 10047
+    },
+    {
+      "epoch": 0.8160480792658166,
+      "grad_norm": 5.343326462409636,
+      "learning_rate": 4.3075841605490414e-07,
+      "loss": 0.4928,
+      "step": 10048
+    },
+    {
+      "epoch": 0.8161292942418582,
+      "grad_norm": 7.146378891333632,
+      "learning_rate": 4.3038944599689105e-07,
+      "loss": 0.3974,
+      "step": 10049
+    },
+    {
+      "epoch": 0.8162105092178997,
+      "grad_norm": 4.603775445290206,
+      "learning_rate": 4.300206191449749e-07,
+      "loss": 0.3021,
+      "step": 10050
+    },
+    {
+      "epoch": 0.8162917241939414,
+      "grad_norm": 6.382432355081491,
+      "learning_rate": 4.2965193552467753e-07,
+      "loss": 0.4456,
+      "step": 10051
+    },
+    {
+      "epoch": 0.816372939169983,
+      "grad_norm": 5.191619470887934,
+      "learning_rate": 4.292833951615083e-07,
+      "loss": 0.3823,
+      "step": 10052
+    },
+    {
+      "epoch": 0.8164541541460245,
+      "grad_norm": 3.9888610071010224,
+      "learning_rate": 4.289149980809698e-07,
+      "loss": 0.6601,
+      "step": 10053
+    },
+    {
+      "epoch": 0.8165353691220661,
+      "grad_norm": 8.500802989835256,
+      "learning_rate": 4.2854674430855224e-07,
+      "loss": 0.4692,
+      "step": 10054
+    },
+    {
+      "epoch": 0.8166165840981077,
+      "grad_norm": 4.251250937241956,
+      "learning_rate": 4.281786338697369e-07,
+      "loss": 0.4196,
+      "step": 10055
+    },
+    {
+      "epoch": 0.8166977990741493,
+      "grad_norm": 5.93357022678489,
+      "learning_rate": 4.278106667899945e-07,
+      "loss": 0.7008,
+      "step": 10056
+    },
+    {
+      "epoch": 0.8167790140501908,
+      "grad_norm": 4.588729939434148,
+      "learning_rate": 4.274428430947872e-07,
+      "loss": 0.4552,
+      "step": 10057
+    },
+    {
+      "epoch": 0.8168602290262325,
+      "grad_norm": 6.368147028406131,
+      "learning_rate": 4.270751628095668e-07,
+      "loss": 0.3992,
+      "step": 10058
+    },
+    {
+      "epoch": 0.816941444002274,
+      "grad_norm": 6.6517461490094565,
+      "learning_rate": 4.2670762595977356e-07,
+      "loss": 0.4615,
+      "step": 10059
+    },
+    {
+      "epoch": 0.8170226589783156,
+      "grad_norm": 8.295784988993061,
+      "learning_rate": 4.2634023257084074e-07,
+      "loss": 0.5174,
+      "step": 10060
+    },
+    {
+      "epoch": 0.8171038739543571,
+      "grad_norm": 5.523298295622445,
+      "learning_rate": 4.259729826681891e-07,
+      "loss": 0.3388,
+      "step": 10061
+    },
+    {
+      "epoch": 0.8171850889303988,
+      "grad_norm": 3.8441813036836137,
+      "learning_rate": 4.2560587627722973e-07,
+      "loss": 0.4553,
+      "step": 10062
+    },
+    {
+      "epoch": 0.8172663039064404,
+      "grad_norm": 6.951945297996189,
+      "learning_rate": 4.2523891342336506e-07,
+      "loss": 0.5312,
+      "step": 10063
+    },
+    {
+      "epoch": 0.8173475188824819,
+      "grad_norm": 4.871582562337795,
+      "learning_rate": 4.2487209413198784e-07,
+      "loss": 0.3208,
+      "step": 10064
+    },
+    {
+      "epoch": 0.8174287338585235,
+      "grad_norm": 7.346570602660871,
+      "learning_rate": 4.245054184284786e-07,
+      "loss": 0.4115,
+      "step": 10065
+    },
+    {
+      "epoch": 0.8175099488345651,
+      "grad_norm": 4.392641843655781,
+      "learning_rate": 4.2413888633821064e-07,
+      "loss": 0.5406,
+      "step": 10066
+    },
+    {
+      "epoch": 0.8175911638106067,
+      "grad_norm": 7.23578723028704,
+      "learning_rate": 4.237724978865454e-07,
+      "loss": 0.4598,
+      "step": 10067
+    },
+    {
+      "epoch": 0.8176723787866482,
+      "grad_norm": 3.553356832932114,
+      "learning_rate": 4.234062530988342e-07,
+      "loss": 0.5109,
+      "step": 10068
+    },
+    {
+      "epoch": 0.8177535937626899,
+      "grad_norm": 5.0655040941198255,
+      "learning_rate": 4.2304015200042095e-07,
+      "loss": 0.6022,
+      "step": 10069
+    },
+    {
+      "epoch": 0.8178348087387314,
+      "grad_norm": 4.3230856397717226,
+      "learning_rate": 4.2267419461663626e-07,
+      "loss": 0.4315,
+      "step": 10070
+    },
+    {
+      "epoch": 0.817916023714773,
+      "grad_norm": 4.6434028931530715,
+      "learning_rate": 4.223083809728032e-07,
+      "loss": 0.489,
+      "step": 10071
+    },
+    {
+      "epoch": 0.8179972386908145,
+      "grad_norm": 3.9428152943512673,
+      "learning_rate": 4.219427110942348e-07,
+      "loss": 0.5609,
+      "step": 10072
+    },
+    {
+      "epoch": 0.8180784536668562,
+      "grad_norm": 3.9140488822139243,
+      "learning_rate": 4.215771850062328e-07,
+      "loss": 0.5247,
+      "step": 10073
+    },
+    {
+      "epoch": 0.8181596686428978,
+      "grad_norm": 5.179289360939204,
+      "learning_rate": 4.2121180273408976e-07,
+      "loss": 0.4456,
+      "step": 10074
+    },
+    {
+      "epoch": 0.8182408836189393,
+      "grad_norm": 6.433424694858643,
+      "learning_rate": 4.2084656430308765e-07,
+      "loss": 0.4248,
+      "step": 10075
+    },
+    {
+      "epoch": 0.818322098594981,
+      "grad_norm": 3.365422337217589,
+      "learning_rate": 4.204814697384993e-07,
+      "loss": 0.4796,
+      "step": 10076
+    },
+    {
+      "epoch": 0.8184033135710225,
+      "grad_norm": 7.511774642841241,
+      "learning_rate": 4.2011651906558815e-07,
+      "loss": 0.6421,
+      "step": 10077
+    },
+    {
+      "epoch": 0.8184845285470641,
+      "grad_norm": 8.546191005512076,
+      "learning_rate": 4.1975171230960563e-07,
+      "loss": 0.3865,
+      "step": 10078
+    },
+    {
+      "epoch": 0.8185657435231056,
+      "grad_norm": 4.417370933628543,
+      "learning_rate": 4.193870494957958e-07,
+      "loss": 0.453,
+      "step": 10079
+    },
+    {
+      "epoch": 0.8186469584991473,
+      "grad_norm": 6.377398869498588,
+      "learning_rate": 4.190225306493906e-07,
+      "loss": 0.4526,
+      "step": 10080
+    },
+    {
+      "epoch": 0.8187281734751888,
+      "grad_norm": 4.132468182015889,
+      "learning_rate": 4.186581557956124e-07,
+      "loss": 0.5708,
+      "step": 10081
+    },
+    {
+      "epoch": 0.8188093884512304,
+      "grad_norm": 5.42105998479533,
+      "learning_rate": 4.1829392495967485e-07,
+      "loss": 0.5129,
+      "step": 10082
+    },
+    {
+      "epoch": 0.8188906034272719,
+      "grad_norm": 5.495216782158462,
+      "learning_rate": 4.1792983816677987e-07,
+      "loss": 0.5831,
+      "step": 10083
+    },
+    {
+      "epoch": 0.8189718184033136,
+      "grad_norm": 3.395888728224807,
+      "learning_rate": 4.175658954421208e-07,
+      "loss": 0.5072,
+      "step": 10084
+    },
+    {
+      "epoch": 0.8190530333793552,
+      "grad_norm": 5.212070130995405,
+      "learning_rate": 4.172020968108814e-07,
+      "loss": 0.5714,
+      "step": 10085
+    },
+    {
+      "epoch": 0.8191342483553967,
+      "grad_norm": 5.41209526251742,
+      "learning_rate": 4.168384422982338e-07,
+      "loss": 0.5516,
+      "step": 10086
+    },
+    {
+      "epoch": 0.8192154633314384,
+      "grad_norm": 5.766740901549195,
+      "learning_rate": 4.164749319293404e-07,
+      "loss": 0.54,
+      "step": 10087
+    },
+    {
+      "epoch": 0.8192966783074799,
+      "grad_norm": 12.522728877358098,
+      "learning_rate": 4.1611156572935545e-07,
+      "loss": 0.5872,
+      "step": 10088
+    },
+    {
+      "epoch": 0.8193778932835215,
+      "grad_norm": 7.055171169509863,
+      "learning_rate": 4.1574834372342053e-07,
+      "loss": 0.3578,
+      "step": 10089
+    },
+    {
+      "epoch": 0.819459108259563,
+      "grad_norm": 10.127142685814183,
+      "learning_rate": 4.153852659366697e-07,
+      "loss": 0.4256,
+      "step": 10090
+    },
+    {
+      "epoch": 0.8195403232356047,
+      "grad_norm": 6.833361098639403,
+      "learning_rate": 4.1502233239422624e-07,
+      "loss": 0.5271,
+      "step": 10091
+    },
+    {
+      "epoch": 0.8196215382116462,
+      "grad_norm": 3.7527822726714057,
+      "learning_rate": 4.14659543121203e-07,
+      "loss": 0.4952,
+      "step": 10092
+    },
+    {
+      "epoch": 0.8197027531876878,
+      "grad_norm": 8.853385017010872,
+      "learning_rate": 4.1429689814270284e-07,
+      "loss": 0.4493,
+      "step": 10093
+    },
+    {
+      "epoch": 0.8197839681637293,
+      "grad_norm": 4.804086833503905,
+      "learning_rate": 4.139343974838181e-07,
+      "loss": 0.4145,
+      "step": 10094
+    },
+    {
+      "epoch": 0.819865183139771,
+      "grad_norm": 2.9666099986165473,
+      "learning_rate": 4.135720411696334e-07,
+      "loss": 0.3967,
+      "step": 10095
+    },
+    {
+      "epoch": 0.8199463981158126,
+      "grad_norm": 6.498314301540783,
+      "learning_rate": 4.132098292252204e-07,
+      "loss": 0.3838,
+      "step": 10096
+    },
+    {
+      "epoch": 0.8200276130918541,
+      "grad_norm": 5.829495180353205,
+      "learning_rate": 4.128477616756432e-07,
+      "loss": 0.4869,
+      "step": 10097
+    },
+    {
+      "epoch": 0.8201088280678958,
+      "grad_norm": 5.33074506215202,
+      "learning_rate": 4.124858385459554e-07,
+      "loss": 0.4165,
+      "step": 10098
+    },
+    {
+      "epoch": 0.8201900430439373,
+      "grad_norm": 4.833962665019804,
+      "learning_rate": 4.1212405986119975e-07,
+      "loss": 0.546,
+      "step": 10099
+    },
+    {
+      "epoch": 0.8202712580199789,
+      "grad_norm": 4.555452441900101,
+      "learning_rate": 4.117624256464084e-07,
+      "loss": 0.4589,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8203524729960204,
+      "grad_norm": 3.2866233085274623,
+      "learning_rate": 4.114009359266061e-07,
+      "loss": 0.482,
+      "step": 10101
+    },
+    {
+      "epoch": 0.8204336879720621,
+      "grad_norm": 5.633628638636074,
+      "learning_rate": 4.1103959072680446e-07,
+      "loss": 0.3467,
+      "step": 10102
+    },
+    {
+      "epoch": 0.8205149029481036,
+      "grad_norm": 2.6695213124316703,
+      "learning_rate": 4.106783900720074e-07,
+      "loss": 0.5319,
+      "step": 10103
+    },
+    {
+      "epoch": 0.8205961179241452,
+      "grad_norm": 6.398738658758001,
+      "learning_rate": 4.1031733398720906e-07,
+      "loss": 0.4174,
+      "step": 10104
+    },
+    {
+      "epoch": 0.8206773329001867,
+      "grad_norm": 10.457002673876866,
+      "learning_rate": 4.099564224973915e-07,
+      "loss": 0.4107,
+      "step": 10105
+    },
+    {
+      "epoch": 0.8207585478762284,
+      "grad_norm": 3.806681811308465,
+      "learning_rate": 4.0959565562752767e-07,
+      "loss": 0.5535,
+      "step": 10106
+    },
+    {
+      "epoch": 0.82083976285227,
+      "grad_norm": 3.11228162293575,
+      "learning_rate": 4.092350334025816e-07,
+      "loss": 0.5205,
+      "step": 10107
+    },
+    {
+      "epoch": 0.8209209778283115,
+      "grad_norm": 6.448242144302385,
+      "learning_rate": 4.0887455584750547e-07,
+      "loss": 0.5066,
+      "step": 10108
+    },
+    {
+      "epoch": 0.8210021928043532,
+      "grad_norm": 4.566144679361284,
+      "learning_rate": 4.0851422298724354e-07,
+      "loss": 0.6172,
+      "step": 10109
+    },
+    {
+      "epoch": 0.8210834077803947,
+      "grad_norm": 5.487995399366438,
+      "learning_rate": 4.081540348467278e-07,
+      "loss": 0.4025,
+      "step": 10110
+    },
+    {
+      "epoch": 0.8211646227564363,
+      "grad_norm": 5.156550511751783,
+      "learning_rate": 4.0779399145088247e-07,
+      "loss": 0.4171,
+      "step": 10111
+    },
+    {
+      "epoch": 0.8212458377324778,
+      "grad_norm": 5.693851091075354,
+      "learning_rate": 4.074340928246201e-07,
+      "loss": 0.4858,
+      "step": 10112
+    },
+    {
+      "epoch": 0.8213270527085195,
+      "grad_norm": 10.430588612793592,
+      "learning_rate": 4.0707433899284333e-07,
+      "loss": 0.4029,
+      "step": 10113
+    },
+    {
+      "epoch": 0.821408267684561,
+      "grad_norm": 6.3587031868115975,
+      "learning_rate": 4.067147299804458e-07,
+      "loss": 0.4363,
+      "step": 10114
+    },
+    {
+      "epoch": 0.8214894826606026,
+      "grad_norm": 6.255964758565154,
+      "learning_rate": 4.063552658123102e-07,
+      "loss": 0.5734,
+      "step": 10115
+    },
+    {
+      "epoch": 0.8215706976366441,
+      "grad_norm": 5.605733474613555,
+      "learning_rate": 4.0599594651330956e-07,
+      "loss": 0.5963,
+      "step": 10116
+    },
+    {
+      "epoch": 0.8216519126126858,
+      "grad_norm": 4.079641751194994,
+      "learning_rate": 4.0563677210830763e-07,
+      "loss": 0.5508,
+      "step": 10117
+    },
+    {
+      "epoch": 0.8217331275887274,
+      "grad_norm": 6.640081496102607,
+      "learning_rate": 4.0527774262215687e-07,
+      "loss": 0.3412,
+      "step": 10118
+    },
+    {
+      "epoch": 0.8218143425647689,
+      "grad_norm": 5.1973586672338845,
+      "learning_rate": 4.049188580796995e-07,
+      "loss": 0.3926,
+      "step": 10119
+    },
+    {
+      "epoch": 0.8218955575408106,
+      "grad_norm": 6.292394610949972,
+      "learning_rate": 4.0456011850576985e-07,
+      "loss": 0.3678,
+      "step": 10120
+    },
+    {
+      "epoch": 0.8219767725168521,
+      "grad_norm": 7.307142444437324,
+      "learning_rate": 4.0420152392518926e-07,
+      "loss": 0.4533,
+      "step": 10121
+    },
+    {
+      "epoch": 0.8220579874928937,
+      "grad_norm": 4.1068923254830345,
+      "learning_rate": 4.038430743627714e-07,
+      "loss": 0.535,
+      "step": 10122
+    },
+    {
+      "epoch": 0.8221392024689352,
+      "grad_norm": 10.080633399311306,
+      "learning_rate": 4.0348476984331977e-07,
+      "loss": 0.5241,
+      "step": 10123
+    },
+    {
+      "epoch": 0.8222204174449769,
+      "grad_norm": 5.11781610289055,
+      "learning_rate": 4.031266103916262e-07,
+      "loss": 0.6826,
+      "step": 10124
+    },
+    {
+      "epoch": 0.8223016324210184,
+      "grad_norm": 10.246403450559914,
+      "learning_rate": 4.0276859603247317e-07,
+      "loss": 0.5244,
+      "step": 10125
+    },
+    {
+      "epoch": 0.82238284739706,
+      "grad_norm": 5.888899275267595,
+      "learning_rate": 4.0241072679063437e-07,
+      "loss": 0.5099,
+      "step": 10126
+    },
+    {
+      "epoch": 0.8224640623731015,
+      "grad_norm": 4.555118993832531,
+      "learning_rate": 4.02053002690872e-07,
+      "loss": 0.4915,
+      "step": 10127
+    },
+    {
+      "epoch": 0.8225452773491432,
+      "grad_norm": 6.444960720915754,
+      "learning_rate": 4.016954237579382e-07,
+      "loss": 0.6126,
+      "step": 10128
+    },
+    {
+      "epoch": 0.8226264923251848,
+      "grad_norm": 5.028533921984391,
+      "learning_rate": 4.013379900165756e-07,
+      "loss": 0.4286,
+      "step": 10129
+    },
+    {
+      "epoch": 0.8227077073012263,
+      "grad_norm": 9.57264830043398,
+      "learning_rate": 4.009807014915179e-07,
+      "loss": 0.634,
+      "step": 10130
+    },
+    {
+      "epoch": 0.822788922277268,
+      "grad_norm": 3.535035189189592,
+      "learning_rate": 4.006235582074866e-07,
+      "loss": 0.519,
+      "step": 10131
+    },
+    {
+      "epoch": 0.8228701372533095,
+      "grad_norm": 5.20695512400172,
+      "learning_rate": 4.002665601891939e-07,
+      "loss": 0.4938,
+      "step": 10132
+    },
+    {
+      "epoch": 0.8229513522293511,
+      "grad_norm": 4.97816750794527,
+      "learning_rate": 3.9990970746134283e-07,
+      "loss": 0.3337,
+      "step": 10133
+    },
+    {
+      "epoch": 0.8230325672053926,
+      "grad_norm": 4.2871354698006945,
+      "learning_rate": 3.99553000048625e-07,
+      "loss": 0.4538,
+      "step": 10134
+    },
+    {
+      "epoch": 0.8231137821814343,
+      "grad_norm": 17.838493735891614,
+      "learning_rate": 3.991964379757232e-07,
+      "loss": 0.5731,
+      "step": 10135
+    },
+    {
+      "epoch": 0.8231949971574758,
+      "grad_norm": 5.170687504793261,
+      "learning_rate": 3.988400212673099e-07,
+      "loss": 0.5173,
+      "step": 10136
+    },
+    {
+      "epoch": 0.8232762121335174,
+      "grad_norm": 6.426083270641829,
+      "learning_rate": 3.9848374994804734e-07,
+      "loss": 0.5226,
+      "step": 10137
+    },
+    {
+      "epoch": 0.823357427109559,
+      "grad_norm": 3.7713552353157476,
+      "learning_rate": 3.9812762404258605e-07,
+      "loss": 0.4277,
+      "step": 10138
+    },
+    {
+      "epoch": 0.8234386420856006,
+      "grad_norm": 6.423516094367269,
+      "learning_rate": 3.977716435755702e-07,
+      "loss": 0.4386,
+      "step": 10139
+    },
+    {
+      "epoch": 0.8235198570616422,
+      "grad_norm": 5.7394083306414325,
+      "learning_rate": 3.9741580857163036e-07,
+      "loss": 0.38,
+      "step": 10140
+    },
+    {
+      "epoch": 0.8236010720376837,
+      "grad_norm": 5.863968524558383,
+      "learning_rate": 3.9706011905538827e-07,
+      "loss": 0.3539,
+      "step": 10141
+    },
+    {
+      "epoch": 0.8236822870137254,
+      "grad_norm": 7.197622618993387,
+      "learning_rate": 3.9670457505145643e-07,
+      "loss": 0.6156,
+      "step": 10142
+    },
+    {
+      "epoch": 0.8237635019897669,
+      "grad_norm": 9.307547582177493,
+      "learning_rate": 3.963491765844371e-07,
+      "loss": 0.5406,
+      "step": 10143
+    },
+    {
+      "epoch": 0.8238447169658085,
+      "grad_norm": 3.8307158333938185,
+      "learning_rate": 3.959939236789212e-07,
+      "loss": 0.431,
+      "step": 10144
+    },
+    {
+      "epoch": 0.82392593194185,
+      "grad_norm": 5.112333570814876,
+      "learning_rate": 3.9563881635948984e-07,
+      "loss": 0.4988,
+      "step": 10145
+    },
+    {
+      "epoch": 0.8240071469178917,
+      "grad_norm": 5.863847427204812,
+      "learning_rate": 3.9528385465071594e-07,
+      "loss": 0.6043,
+      "step": 10146
+    },
+    {
+      "epoch": 0.8240883618939332,
+      "grad_norm": 5.626678027585737,
+      "learning_rate": 3.949290385771595e-07,
+      "loss": 0.431,
+      "step": 10147
+    },
+    {
+      "epoch": 0.8241695768699748,
+      "grad_norm": 4.081512795131598,
+      "learning_rate": 3.945743681633729e-07,
+      "loss": 0.5401,
+      "step": 10148
+    },
+    {
+      "epoch": 0.8242507918460164,
+      "grad_norm": 4.533747792384772,
+      "learning_rate": 3.9421984343389756e-07,
+      "loss": 0.5523,
+      "step": 10149
+    },
+    {
+      "epoch": 0.824332006822058,
+      "grad_norm": 3.743850075034431,
+      "learning_rate": 3.9386546441326444e-07,
+      "loss": 0.4706,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8244132217980996,
+      "grad_norm": 23.913229954006653,
+      "learning_rate": 3.9351123112599393e-07,
+      "loss": 0.6842,
+      "step": 10151
+    },
+    {
+      "epoch": 0.8244944367741411,
+      "grad_norm": 7.100866725720028,
+      "learning_rate": 3.931571435965986e-07,
+      "loss": 0.377,
+      "step": 10152
+    },
+    {
+      "epoch": 0.8245756517501828,
+      "grad_norm": 8.290123785692964,
+      "learning_rate": 3.9280320184957864e-07,
+      "loss": 0.5232,
+      "step": 10153
+    },
+    {
+      "epoch": 0.8246568667262243,
+      "grad_norm": 4.89062234331338,
+      "learning_rate": 3.9244940590942413e-07,
+      "loss": 0.4422,
+      "step": 10154
+    },
+    {
+      "epoch": 0.8247380817022659,
+      "grad_norm": 7.550251223417641,
+      "learning_rate": 3.9209575580061663e-07,
+      "loss": 0.2946,
+      "step": 10155
+    },
+    {
+      "epoch": 0.8248192966783074,
+      "grad_norm": 10.05450235709694,
+      "learning_rate": 3.9174225154762766e-07,
+      "loss": 0.6482,
+      "step": 10156
+    },
+    {
+      "epoch": 0.8249005116543491,
+      "grad_norm": 5.365029681588859,
+      "learning_rate": 3.9138889317491656e-07,
+      "loss": 0.5071,
+      "step": 10157
+    },
+    {
+      "epoch": 0.8249817266303906,
+      "grad_norm": 5.609914610484028,
+      "learning_rate": 3.9103568070693485e-07,
+      "loss": 0.5451,
+      "step": 10158
+    },
+    {
+      "epoch": 0.8250629416064322,
+      "grad_norm": 4.514680074383672,
+      "learning_rate": 3.906826141681225e-07,
+      "loss": 0.532,
+      "step": 10159
+    },
+    {
+      "epoch": 0.8251441565824738,
+      "grad_norm": 10.206347105991203,
+      "learning_rate": 3.903296935829093e-07,
+      "loss": 0.5132,
+      "step": 10160
+    },
+    {
+      "epoch": 0.8252253715585154,
+      "grad_norm": 6.858018717895062,
+      "learning_rate": 3.8997691897571577e-07,
+      "loss": 0.4462,
+      "step": 10161
+    },
+    {
+      "epoch": 0.825306586534557,
+      "grad_norm": 5.173227278123777,
+      "learning_rate": 3.896242903709532e-07,
+      "loss": 0.5436,
+      "step": 10162
+    },
+    {
+      "epoch": 0.8253878015105985,
+      "grad_norm": 11.450687404556978,
+      "learning_rate": 3.8927180779302076e-07,
+      "loss": 0.4657,
+      "step": 10163
+    },
+    {
+      "epoch": 0.8254690164866402,
+      "grad_norm": 4.609261708122115,
+      "learning_rate": 3.889194712663075e-07,
+      "loss": 0.4993,
+      "step": 10164
+    },
+    {
+      "epoch": 0.8255502314626817,
+      "grad_norm": 3.5084080535675026,
+      "learning_rate": 3.885672808151947e-07,
+      "loss": 0.4571,
+      "step": 10165
+    },
+    {
+      "epoch": 0.8256314464387233,
+      "grad_norm": 21.30482869821724,
+      "learning_rate": 3.882152364640518e-07,
+      "loss": 0.3259,
+      "step": 10166
+    },
+    {
+      "epoch": 0.8257126614147648,
+      "grad_norm": 4.848731610369329,
+      "learning_rate": 3.878633382372371e-07,
+      "loss": 0.5129,
+      "step": 10167
+    },
+    {
+      "epoch": 0.8257938763908065,
+      "grad_norm": 4.713993311348796,
+      "learning_rate": 3.875115861591014e-07,
+      "loss": 0.4037,
+      "step": 10168
+    },
+    {
+      "epoch": 0.825875091366848,
+      "grad_norm": 6.885554383643488,
+      "learning_rate": 3.871599802539841e-07,
+      "loss": 0.2991,
+      "step": 10169
+    },
+    {
+      "epoch": 0.8259563063428896,
+      "grad_norm": 11.560653980725935,
+      "learning_rate": 3.868085205462135e-07,
+      "loss": 0.4264,
+      "step": 10170
+    },
+    {
+      "epoch": 0.8260375213189312,
+      "grad_norm": 18.94305982548915,
+      "learning_rate": 3.8645720706010997e-07,
+      "loss": 0.4572,
+      "step": 10171
+    },
+    {
+      "epoch": 0.8261187362949728,
+      "grad_norm": 4.3868517626424754,
+      "learning_rate": 3.8610603981998204e-07,
+      "loss": 0.4054,
+      "step": 10172
+    },
+    {
+      "epoch": 0.8261999512710144,
+      "grad_norm": 10.379799445814022,
+      "learning_rate": 3.85755018850128e-07,
+      "loss": 0.4782,
+      "step": 10173
+    },
+    {
+      "epoch": 0.8262811662470559,
+      "grad_norm": 5.464482264230086,
+      "learning_rate": 3.854041441748371e-07,
+      "loss": 0.4246,
+      "step": 10174
+    },
+    {
+      "epoch": 0.8263623812230976,
+      "grad_norm": 5.259022216933499,
+      "learning_rate": 3.8505341581838854e-07,
+      "loss": 0.4936,
+      "step": 10175
+    },
+    {
+      "epoch": 0.8264435961991391,
+      "grad_norm": 3.834344996242316,
+      "learning_rate": 3.8470283380504987e-07,
+      "loss": 0.47,
+      "step": 10176
+    },
+    {
+      "epoch": 0.8265248111751807,
+      "grad_norm": 9.45059948818576,
+      "learning_rate": 3.8435239815908077e-07,
+      "loss": 0.4509,
+      "step": 10177
+    },
+    {
+      "epoch": 0.8266060261512223,
+      "grad_norm": 4.5939757403211585,
+      "learning_rate": 3.8400210890472883e-07,
+      "loss": 0.3479,
+      "step": 10178
+    },
+    {
+      "epoch": 0.8266872411272639,
+      "grad_norm": 5.823737388475345,
+      "learning_rate": 3.836519660662313e-07,
+      "loss": 0.3906,
+      "step": 10179
+    },
+    {
+      "epoch": 0.8267684561033054,
+      "grad_norm": 4.6299433684789095,
+      "learning_rate": 3.8330196966781723e-07,
+      "loss": 0.5363,
+      "step": 10180
+    },
+    {
+      "epoch": 0.826849671079347,
+      "grad_norm": 6.306388306208667,
+      "learning_rate": 3.829521197337052e-07,
+      "loss": 0.433,
+      "step": 10181
+    },
+    {
+      "epoch": 0.8269308860553886,
+      "grad_norm": 5.650157408413531,
+      "learning_rate": 3.8260241628810203e-07,
+      "loss": 0.5346,
+      "step": 10182
+    },
+    {
+      "epoch": 0.8270121010314302,
+      "grad_norm": 4.139343726577053,
+      "learning_rate": 3.8225285935520493e-07,
+      "loss": 0.5034,
+      "step": 10183
+    },
+    {
+      "epoch": 0.8270933160074718,
+      "grad_norm": 4.27842292241429,
+      "learning_rate": 3.8190344895920246e-07,
+      "loss": 0.4205,
+      "step": 10184
+    },
+    {
+      "epoch": 0.8271745309835133,
+      "grad_norm": 4.60689072938466,
+      "learning_rate": 3.815541851242713e-07,
+      "loss": 0.5121,
+      "step": 10185
+    },
+    {
+      "epoch": 0.827255745959555,
+      "grad_norm": 14.92261155989774,
+      "learning_rate": 3.812050678745785e-07,
+      "loss": 0.3954,
+      "step": 10186
+    },
+    {
+      "epoch": 0.8273369609355965,
+      "grad_norm": 4.53044644344942,
+      "learning_rate": 3.808560972342812e-07,
+      "loss": 0.5392,
+      "step": 10187
+    },
+    {
+      "epoch": 0.8274181759116381,
+      "grad_norm": 5.540716412331475,
+      "learning_rate": 3.8050727322752726e-07,
+      "loss": 0.5574,
+      "step": 10188
+    },
+    {
+      "epoch": 0.8274993908876797,
+      "grad_norm": 8.417967136559637,
+      "learning_rate": 3.8015859587845233e-07,
+      "loss": 0.3885,
+      "step": 10189
+    },
+    {
+      "epoch": 0.8275806058637213,
+      "grad_norm": 53.802360428112415,
+      "learning_rate": 3.798100652111839e-07,
+      "loss": 0.4015,
+      "step": 10190
+    },
+    {
+      "epoch": 0.8276618208397628,
+      "grad_norm": 4.877120364803925,
+      "learning_rate": 3.7946168124983776e-07,
+      "loss": 0.3768,
+      "step": 10191
+    },
+    {
+      "epoch": 0.8277430358158044,
+      "grad_norm": 9.495267916803392,
+      "learning_rate": 3.791134440185201e-07,
+      "loss": 0.5635,
+      "step": 10192
+    },
+    {
+      "epoch": 0.827824250791846,
+      "grad_norm": 5.733855636295819,
+      "learning_rate": 3.787653535413277e-07,
+      "loss": 0.4829,
+      "step": 10193
+    },
+    {
+      "epoch": 0.8279054657678876,
+      "grad_norm": 5.873682615843861,
+      "learning_rate": 3.784174098423465e-07,
+      "loss": 0.5228,
+      "step": 10194
+    },
+    {
+      "epoch": 0.8279866807439292,
+      "grad_norm": 4.833822246575063,
+      "learning_rate": 3.780696129456521e-07,
+      "loss": 0.5421,
+      "step": 10195
+    },
+    {
+      "epoch": 0.8280678957199707,
+      "grad_norm": 5.208263826542225,
+      "learning_rate": 3.7772196287531066e-07,
+      "loss": 0.5473,
+      "step": 10196
+    },
+    {
+      "epoch": 0.8281491106960124,
+      "grad_norm": 8.378333393225612,
+      "learning_rate": 3.773744596553774e-07,
+      "loss": 0.4702,
+      "step": 10197
+    },
+    {
+      "epoch": 0.8282303256720539,
+      "grad_norm": 3.4663578157064583,
+      "learning_rate": 3.7702710330989765e-07,
+      "loss": 0.5652,
+      "step": 10198
+    },
+    {
+      "epoch": 0.8283115406480955,
+      "grad_norm": 6.827267196321197,
+      "learning_rate": 3.766798938629063e-07,
+      "loss": 0.4027,
+      "step": 10199
+    },
+    {
+      "epoch": 0.8283927556241371,
+      "grad_norm": 9.444190295092335,
+      "learning_rate": 3.7633283133842845e-07,
+      "loss": 0.5462,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8284739706001787,
+      "grad_norm": 6.929524668677438,
+      "learning_rate": 3.7598591576048e-07,
+      "loss": 0.5488,
+      "step": 10201
+    },
+    {
+      "epoch": 0.8285551855762202,
+      "grad_norm": 5.6038151431290935,
+      "learning_rate": 3.756391471530646e-07,
+      "loss": 0.556,
+      "step": 10202
+    },
+    {
+      "epoch": 0.8286364005522618,
+      "grad_norm": 4.185345966650429,
+      "learning_rate": 3.7529252554017765e-07,
+      "loss": 0.5079,
+      "step": 10203
+    },
+    {
+      "epoch": 0.8287176155283034,
+      "grad_norm": 6.844019960274571,
+      "learning_rate": 3.7494605094580305e-07,
+      "loss": 0.427,
+      "step": 10204
+    },
+    {
+      "epoch": 0.828798830504345,
+      "grad_norm": 5.36417941344362,
+      "learning_rate": 3.7459972339391445e-07,
+      "loss": 0.797,
+      "step": 10205
+    },
+    {
+      "epoch": 0.8288800454803866,
+      "grad_norm": 5.328404842131845,
+      "learning_rate": 3.742535429084765e-07,
+      "loss": 0.3947,
+      "step": 10206
+    },
+    {
+      "epoch": 0.8289612604564282,
+      "grad_norm": 5.223058995984122,
+      "learning_rate": 3.739075095134437e-07,
+      "loss": 0.4567,
+      "step": 10207
+    },
+    {
+      "epoch": 0.8290424754324698,
+      "grad_norm": 4.706033163133879,
+      "learning_rate": 3.735616232327582e-07,
+      "loss": 0.4547,
+      "step": 10208
+    },
+    {
+      "epoch": 0.8291236904085113,
+      "grad_norm": 5.7396745415353845,
+      "learning_rate": 3.732158840903552e-07,
+      "loss": 0.5567,
+      "step": 10209
+    },
+    {
+      "epoch": 0.8292049053845529,
+      "grad_norm": 4.958530850319359,
+      "learning_rate": 3.728702921101571e-07,
+      "loss": 0.5436,
+      "step": 10210
+    },
+    {
+      "epoch": 0.8292861203605945,
+      "grad_norm": 4.861206028038226,
+      "learning_rate": 3.725248473160764e-07,
+      "loss": 0.5996,
+      "step": 10211
+    },
+    {
+      "epoch": 0.8293673353366361,
+      "grad_norm": 4.682688015412765,
+      "learning_rate": 3.721795497320174e-07,
+      "loss": 0.5646,
+      "step": 10212
+    },
+    {
+      "epoch": 0.8294485503126776,
+      "grad_norm": 5.262659812480152,
+      "learning_rate": 3.718343993818718e-07,
+      "loss": 0.5358,
+      "step": 10213
+    },
+    {
+      "epoch": 0.8295297652887192,
+      "grad_norm": 9.851823169517866,
+      "learning_rate": 3.7148939628952246e-07,
+      "loss": 0.4423,
+      "step": 10214
+    },
+    {
+      "epoch": 0.8296109802647608,
+      "grad_norm": 5.776223337431074,
+      "learning_rate": 3.7114454047884247e-07,
+      "loss": 0.3799,
+      "step": 10215
+    },
+    {
+      "epoch": 0.8296921952408024,
+      "grad_norm": 8.26924554342392,
+      "learning_rate": 3.707998319736936e-07,
+      "loss": 0.4953,
+      "step": 10216
+    },
+    {
+      "epoch": 0.829773410216844,
+      "grad_norm": 5.982564098065872,
+      "learning_rate": 3.7045527079792753e-07,
+      "loss": 0.593,
+      "step": 10217
+    },
+    {
+      "epoch": 0.8298546251928856,
+      "grad_norm": 5.7895130886280635,
+      "learning_rate": 3.7011085697538587e-07,
+      "loss": 0.4198,
+      "step": 10218
+    },
+    {
+      "epoch": 0.8299358401689272,
+      "grad_norm": 4.460178263466723,
+      "learning_rate": 3.6976659052990056e-07,
+      "loss": 0.6089,
+      "step": 10219
+    },
+    {
+      "epoch": 0.8300170551449687,
+      "grad_norm": 5.700291972867315,
+      "learning_rate": 3.694224714852937e-07,
+      "loss": 0.405,
+      "step": 10220
+    },
+    {
+      "epoch": 0.8300982701210103,
+      "grad_norm": 7.327954495107345,
+      "learning_rate": 3.6907849986537516e-07,
+      "loss": 0.4625,
+      "step": 10221
+    },
+    {
+      "epoch": 0.8301794850970519,
+      "grad_norm": 5.605599034700318,
+      "learning_rate": 3.687346756939475e-07,
+      "loss": 0.4958,
+      "step": 10222
+    },
+    {
+      "epoch": 0.8302607000730935,
+      "grad_norm": 4.754692475561762,
+      "learning_rate": 3.6839099899480033e-07,
+      "loss": 0.3742,
+      "step": 10223
+    },
+    {
+      "epoch": 0.830341915049135,
+      "grad_norm": 3.9081424392387327,
+      "learning_rate": 3.680474697917144e-07,
+      "loss": 0.6079,
+      "step": 10224
+    },
+    {
+      "epoch": 0.8304231300251766,
+      "grad_norm": 4.983983156364833,
+      "learning_rate": 3.677040881084609e-07,
+      "loss": 0.4596,
+      "step": 10225
+    },
+    {
+      "epoch": 0.8305043450012182,
+      "grad_norm": 6.779953311673761,
+      "learning_rate": 3.6736085396879896e-07,
+      "loss": 0.368,
+      "step": 10226
+    },
+    {
+      "epoch": 0.8305855599772598,
+      "grad_norm": 3.867084328558502,
+      "learning_rate": 3.6701776739647893e-07,
+      "loss": 0.5043,
+      "step": 10227
+    },
+    {
+      "epoch": 0.8306667749533014,
+      "grad_norm": 5.763370281929799,
+      "learning_rate": 3.666748284152413e-07,
+      "loss": 0.6709,
+      "step": 10228
+    },
+    {
+      "epoch": 0.830747989929343,
+      "grad_norm": 5.039857359159814,
+      "learning_rate": 3.663320370488152e-07,
+      "loss": 0.4427,
+      "step": 10229
+    },
+    {
+      "epoch": 0.8308292049053846,
+      "grad_norm": 3.9366691516556678,
+      "learning_rate": 3.659893933209191e-07,
+      "loss": 0.5353,
+      "step": 10230
+    },
+    {
+      "epoch": 0.8309104198814261,
+      "grad_norm": 16.823744196747114,
+      "learning_rate": 3.6564689725526377e-07,
+      "loss": 0.455,
+      "step": 10231
+    },
+    {
+      "epoch": 0.8309916348574677,
+      "grad_norm": 3.5430059865653996,
+      "learning_rate": 3.6530454887554636e-07,
+      "loss": 0.6269,
+      "step": 10232
+    },
+    {
+      "epoch": 0.8310728498335093,
+      "grad_norm": 4.922946722027889,
+      "learning_rate": 3.649623482054565e-07,
+      "loss": 0.4448,
+      "step": 10233
+    },
+    {
+      "epoch": 0.8311540648095509,
+      "grad_norm": 5.418331879921831,
+      "learning_rate": 3.6462029526867335e-07,
+      "loss": 0.3769,
+      "step": 10234
+    },
+    {
+      "epoch": 0.8312352797855924,
+      "grad_norm": 6.912181302002352,
+      "learning_rate": 3.642783900888644e-07,
+      "loss": 0.4543,
+      "step": 10235
+    },
+    {
+      "epoch": 0.831316494761634,
+      "grad_norm": 7.405882641111188,
+      "learning_rate": 3.639366326896876e-07,
+      "loss": 0.4988,
+      "step": 10236
+    },
+    {
+      "epoch": 0.8313977097376756,
+      "grad_norm": 6.548258048190026,
+      "learning_rate": 3.635950230947902e-07,
+      "loss": 0.4483,
+      "step": 10237
+    },
+    {
+      "epoch": 0.8314789247137172,
+      "grad_norm": 7.4684554933443925,
+      "learning_rate": 3.632535613278107e-07,
+      "loss": 0.4406,
+      "step": 10238
+    },
+    {
+      "epoch": 0.8315601396897588,
+      "grad_norm": 3.8512394389587703,
+      "learning_rate": 3.629122474123767e-07,
+      "loss": 0.4396,
+      "step": 10239
+    },
+    {
+      "epoch": 0.8316413546658004,
+      "grad_norm": 6.464656562967151,
+      "learning_rate": 3.6257108137210396e-07,
+      "loss": 0.37,
+      "step": 10240
+    },
+    {
+      "epoch": 0.831722569641842,
+      "grad_norm": 5.940372384619711,
+      "learning_rate": 3.622300632306011e-07,
+      "loss": 0.3829,
+      "step": 10241
+    },
+    {
+      "epoch": 0.8318037846178835,
+      "grad_norm": 7.830539757818605,
+      "learning_rate": 3.6188919301146375e-07,
+      "loss": 0.4114,
+      "step": 10242
+    },
+    {
+      "epoch": 0.8318849995939251,
+      "grad_norm": 4.839330335943199,
+      "learning_rate": 3.615484707382777e-07,
+      "loss": 0.5915,
+      "step": 10243
+    },
+    {
+      "epoch": 0.8319662145699667,
+      "grad_norm": 3.688765962182601,
+      "learning_rate": 3.6120789643462053e-07,
+      "loss": 0.5409,
+      "step": 10244
+    },
+    {
+      "epoch": 0.8320474295460083,
+      "grad_norm": 6.537020216802781,
+      "learning_rate": 3.608674701240572e-07,
+      "loss": 0.3567,
+      "step": 10245
+    },
+    {
+      "epoch": 0.8321286445220498,
+      "grad_norm": 4.589211449298252,
+      "learning_rate": 3.605271918301434e-07,
+      "loss": 0.5638,
+      "step": 10246
+    },
+    {
+      "epoch": 0.8322098594980915,
+      "grad_norm": 6.29504528521005,
+      "learning_rate": 3.601870615764258e-07,
+      "loss": 0.5068,
+      "step": 10247
+    },
+    {
+      "epoch": 0.832291074474133,
+      "grad_norm": 4.262119801386439,
+      "learning_rate": 3.5984707938643864e-07,
+      "loss": 0.4094,
+      "step": 10248
+    },
+    {
+      "epoch": 0.8323722894501746,
+      "grad_norm": 4.841197784154483,
+      "learning_rate": 3.5950724528370615e-07,
+      "loss": 0.4699,
+      "step": 10249
+    },
+    {
+      "epoch": 0.8324535044262162,
+      "grad_norm": 5.42937376505632,
+      "learning_rate": 3.591675592917449e-07,
+      "loss": 0.4166,
+      "step": 10250
+    },
+    {
+      "epoch": 0.8325347194022578,
+      "grad_norm": 6.296856237155061,
+      "learning_rate": 3.5882802143405755e-07,
+      "loss": 0.449,
+      "step": 10251
+    },
+    {
+      "epoch": 0.8326159343782994,
+      "grad_norm": 7.636038816244131,
+      "learning_rate": 3.584886317341396e-07,
+      "loss": 0.4694,
+      "step": 10252
+    },
+    {
+      "epoch": 0.8326971493543409,
+      "grad_norm": 21.61523367073273,
+      "learning_rate": 3.58149390215474e-07,
+      "loss": 0.5688,
+      "step": 10253
+    },
+    {
+      "epoch": 0.8327783643303825,
+      "grad_norm": 5.877406518099175,
+      "learning_rate": 3.5781029690153567e-07,
+      "loss": 0.5595,
+      "step": 10254
+    },
+    {
+      "epoch": 0.8328595793064241,
+      "grad_norm": 4.889479170900449,
+      "learning_rate": 3.574713518157874e-07,
+      "loss": 0.5255,
+      "step": 10255
+    },
+    {
+      "epoch": 0.8329407942824657,
+      "grad_norm": 6.633095888789909,
+      "learning_rate": 3.571325549816818e-07,
+      "loss": 0.3982,
+      "step": 10256
+    },
+    {
+      "epoch": 0.8330220092585072,
+      "grad_norm": 4.06737934876213,
+      "learning_rate": 3.56793906422663e-07,
+      "loss": 0.5486,
+      "step": 10257
+    },
+    {
+      "epoch": 0.8331032242345489,
+      "grad_norm": 4.0549895036371,
+      "learning_rate": 3.564554061621625e-07,
+      "loss": 0.4597,
+      "step": 10258
+    },
+    {
+      "epoch": 0.8331844392105904,
+      "grad_norm": 8.500590717560147,
+      "learning_rate": 3.5611705422360335e-07,
+      "loss": 0.491,
+      "step": 10259
+    },
+    {
+      "epoch": 0.833265654186632,
+      "grad_norm": 5.841348510557204,
+      "learning_rate": 3.557788506303986e-07,
+      "loss": 0.3571,
+      "step": 10260
+    },
+    {
+      "epoch": 0.8333468691626736,
+      "grad_norm": 8.428998076590059,
+      "learning_rate": 3.5544079540594884e-07,
+      "loss": 0.5252,
+      "step": 10261
+    },
+    {
+      "epoch": 0.8334280841387152,
+      "grad_norm": 6.0161944380779255,
+      "learning_rate": 3.551028885736457e-07,
+      "loss": 0.5061,
+      "step": 10262
+    },
+    {
+      "epoch": 0.8335092991147568,
+      "grad_norm": 26.06934398233002,
+      "learning_rate": 3.5476513015687136e-07,
+      "loss": 0.4221,
+      "step": 10263
+    },
+    {
+      "epoch": 0.8335905140907983,
+      "grad_norm": 6.725770937675891,
+      "learning_rate": 3.5442752017899625e-07,
+      "loss": 0.5557,
+      "step": 10264
+    },
+    {
+      "epoch": 0.83367172906684,
+      "grad_norm": 5.63216646803754,
+      "learning_rate": 3.5409005866338134e-07,
+      "loss": 0.4044,
+      "step": 10265
+    },
+    {
+      "epoch": 0.8337529440428815,
+      "grad_norm": 5.540976364008194,
+      "learning_rate": 3.537527456333778e-07,
+      "loss": 0.4748,
+      "step": 10266
+    },
+    {
+      "epoch": 0.8338341590189231,
+      "grad_norm": 4.118260128090111,
+      "learning_rate": 3.5341558111232547e-07,
+      "loss": 0.5088,
+      "step": 10267
+    },
+    {
+      "epoch": 0.8339153739949646,
+      "grad_norm": 4.454821659286079,
+      "learning_rate": 3.5307856512355354e-07,
+      "loss": 0.408,
+      "step": 10268
+    },
+    {
+      "epoch": 0.8339965889710063,
+      "grad_norm": 7.906091232376149,
+      "learning_rate": 3.527416976903833e-07,
+      "loss": 0.4369,
+      "step": 10269
+    },
+    {
+      "epoch": 0.8340778039470479,
+      "grad_norm": 4.946330213556263,
+      "learning_rate": 3.5240497883612333e-07,
+      "loss": 0.5383,
+      "step": 10270
+    },
+    {
+      "epoch": 0.8341590189230894,
+      "grad_norm": 3.921389705656694,
+      "learning_rate": 3.5206840858407225e-07,
+      "loss": 0.5159,
+      "step": 10271
+    },
+    {
+      "epoch": 0.834240233899131,
+      "grad_norm": 9.466775431263171,
+      "learning_rate": 3.517319869575195e-07,
+      "loss": 0.3524,
+      "step": 10272
+    },
+    {
+      "epoch": 0.8343214488751726,
+      "grad_norm": 7.781642113894127,
+      "learning_rate": 3.5139571397974416e-07,
+      "loss": 0.5091,
+      "step": 10273
+    },
+    {
+      "epoch": 0.8344026638512142,
+      "grad_norm": 5.031639260527779,
+      "learning_rate": 3.5105958967401404e-07,
+      "loss": 0.3777,
+      "step": 10274
+    },
+    {
+      "epoch": 0.8344838788272557,
+      "grad_norm": 4.936172284543063,
+      "learning_rate": 3.5072361406358696e-07,
+      "loss": 0.5142,
+      "step": 10275
+    },
+    {
+      "epoch": 0.8345650938032974,
+      "grad_norm": 5.62972700250256,
+      "learning_rate": 3.5038778717171123e-07,
+      "loss": 0.5169,
+      "step": 10276
+    },
+    {
+      "epoch": 0.8346463087793389,
+      "grad_norm": 4.876718013101768,
+      "learning_rate": 3.500521090216233e-07,
+      "loss": 0.5267,
+      "step": 10277
+    },
+    {
+      "epoch": 0.8347275237553805,
+      "grad_norm": 5.004111684696393,
+      "learning_rate": 3.497165796365512e-07,
+      "loss": 0.6079,
+      "step": 10278
+    },
+    {
+      "epoch": 0.834808738731422,
+      "grad_norm": 5.227016363738208,
+      "learning_rate": 3.4938119903971195e-07,
+      "loss": 0.3386,
+      "step": 10279
+    },
+    {
+      "epoch": 0.8348899537074637,
+      "grad_norm": 4.70841750512664,
+      "learning_rate": 3.49045967254312e-07,
+      "loss": 0.4524,
+      "step": 10280
+    },
+    {
+      "epoch": 0.8349711686835053,
+      "grad_norm": 4.398665081558391,
+      "learning_rate": 3.487108843035467e-07,
+      "loss": 0.4655,
+      "step": 10281
+    },
+    {
+      "epoch": 0.8350523836595468,
+      "grad_norm": 4.351497987088346,
+      "learning_rate": 3.4837595021060296e-07,
+      "loss": 0.4796,
+      "step": 10282
+    },
+    {
+      "epoch": 0.8351335986355884,
+      "grad_norm": 4.803405743264782,
+      "learning_rate": 3.480411649986565e-07,
+      "loss": 0.3933,
+      "step": 10283
+    },
+    {
+      "epoch": 0.83521481361163,
+      "grad_norm": 4.927789971086148,
+      "learning_rate": 3.477065286908715e-07,
+      "loss": 0.407,
+      "step": 10284
+    },
+    {
+      "epoch": 0.8352960285876716,
+      "grad_norm": 9.580044334289088,
+      "learning_rate": 3.4737204131040397e-07,
+      "loss": 0.5534,
+      "step": 10285
+    },
+    {
+      "epoch": 0.8353772435637131,
+      "grad_norm": 7.407853324575902,
+      "learning_rate": 3.470377028803992e-07,
+      "loss": 0.4651,
+      "step": 10286
+    },
+    {
+      "epoch": 0.8354584585397548,
+      "grad_norm": 4.484117272427251,
+      "learning_rate": 3.46703513423991e-07,
+      "loss": 0.5232,
+      "step": 10287
+    },
+    {
+      "epoch": 0.8355396735157963,
+      "grad_norm": 4.330327720269298,
+      "learning_rate": 3.4636947296430274e-07,
+      "loss": 0.4451,
+      "step": 10288
+    },
+    {
+      "epoch": 0.8356208884918379,
+      "grad_norm": 5.965597862522399,
+      "learning_rate": 3.460355815244498e-07,
+      "loss": 0.5906,
+      "step": 10289
+    },
+    {
+      "epoch": 0.8357021034678794,
+      "grad_norm": 8.22500497243567,
+      "learning_rate": 3.457018391275341e-07,
+      "loss": 0.3985,
+      "step": 10290
+    },
+    {
+      "epoch": 0.8357833184439211,
+      "grad_norm": 5.544151542569365,
+      "learning_rate": 3.4536824579665007e-07,
+      "loss": 0.4722,
+      "step": 10291
+    },
+    {
+      "epoch": 0.8358645334199627,
+      "grad_norm": 5.097926053416765,
+      "learning_rate": 3.4503480155488044e-07,
+      "loss": 0.5343,
+      "step": 10292
+    },
+    {
+      "epoch": 0.8359457483960042,
+      "grad_norm": 11.549757808229108,
+      "learning_rate": 3.447015064252976e-07,
+      "loss": 0.3995,
+      "step": 10293
+    },
+    {
+      "epoch": 0.8360269633720459,
+      "grad_norm": 4.072804987437322,
+      "learning_rate": 3.443683604309633e-07,
+      "loss": 0.4749,
+      "step": 10294
+    },
+    {
+      "epoch": 0.8361081783480874,
+      "grad_norm": 5.09382553179545,
+      "learning_rate": 3.4403536359493034e-07,
+      "loss": 0.4529,
+      "step": 10295
+    },
+    {
+      "epoch": 0.836189393324129,
+      "grad_norm": 4.720065781209744,
+      "learning_rate": 3.437025159402399e-07,
+      "loss": 0.58,
+      "step": 10296
+    },
+    {
+      "epoch": 0.8362706083001705,
+      "grad_norm": 4.694987719383441,
+      "learning_rate": 3.43369817489923e-07,
+      "loss": 0.4097,
+      "step": 10297
+    },
+    {
+      "epoch": 0.8363518232762122,
+      "grad_norm": 5.4566621726960145,
+      "learning_rate": 3.430372682670008e-07,
+      "loss": 0.4231,
+      "step": 10298
+    },
+    {
+      "epoch": 0.8364330382522537,
+      "grad_norm": 8.027749216651072,
+      "learning_rate": 3.4270486829448476e-07,
+      "loss": 0.4286,
+      "step": 10299
+    },
+    {
+      "epoch": 0.8365142532282953,
+      "grad_norm": 5.684585620616569,
+      "learning_rate": 3.423726175953737e-07,
+      "loss": 0.4205,
+      "step": 10300
+    },
+    {
+      "epoch": 0.8365954682043368,
+      "grad_norm": 3.9316108403026178,
+      "learning_rate": 3.4204051619265905e-07,
+      "loss": 0.4315,
+      "step": 10301
+    },
+    {
+      "epoch": 0.8366766831803785,
+      "grad_norm": 5.5134529395223915,
+      "learning_rate": 3.4170856410931986e-07,
+      "loss": 0.3852,
+      "step": 10302
+    },
+    {
+      "epoch": 0.8367578981564201,
+      "grad_norm": 6.0566636394615285,
+      "learning_rate": 3.41376761368325e-07,
+      "loss": 0.544,
+      "step": 10303
+    },
+    {
+      "epoch": 0.8368391131324616,
+      "grad_norm": 5.68024482168099,
+      "learning_rate": 3.4104510799263356e-07,
+      "loss": 0.4424,
+      "step": 10304
+    },
+    {
+      "epoch": 0.8369203281085033,
+      "grad_norm": 8.931124333462172,
+      "learning_rate": 3.407136040051953e-07,
+      "loss": 0.4358,
+      "step": 10305
+    },
+    {
+      "epoch": 0.8370015430845448,
+      "grad_norm": 6.202340802104501,
+      "learning_rate": 3.40382249428948e-07,
+      "loss": 0.6703,
+      "step": 10306
+    },
+    {
+      "epoch": 0.8370827580605864,
+      "grad_norm": 5.36293568403821,
+      "learning_rate": 3.400510442868185e-07,
+      "loss": 0.5796,
+      "step": 10307
+    },
+    {
+      "epoch": 0.8371639730366279,
+      "grad_norm": 5.206542300226407,
+      "learning_rate": 3.3971998860172605e-07,
+      "loss": 0.5902,
+      "step": 10308
+    },
+    {
+      "epoch": 0.8372451880126696,
+      "grad_norm": 3.560373786193406,
+      "learning_rate": 3.393890823965768e-07,
+      "loss": 0.4523,
+      "step": 10309
+    },
+    {
+      "epoch": 0.8373264029887111,
+      "grad_norm": 5.239048746596322,
+      "learning_rate": 3.390583256942681e-07,
+      "loss": 0.5946,
+      "step": 10310
+    },
+    {
+      "epoch": 0.8374076179647527,
+      "grad_norm": 4.838192804018576,
+      "learning_rate": 3.3872771851768737e-07,
+      "loss": 0.4644,
+      "step": 10311
+    },
+    {
+      "epoch": 0.8374888329407942,
+      "grad_norm": 7.90289534921277,
+      "learning_rate": 3.383972608897099e-07,
+      "loss": 0.4878,
+      "step": 10312
+    },
+    {
+      "epoch": 0.8375700479168359,
+      "grad_norm": 3.8504253330468092,
+      "learning_rate": 3.3806695283320145e-07,
+      "loss": 0.3547,
+      "step": 10313
+    },
+    {
+      "epoch": 0.8376512628928775,
+      "grad_norm": 11.90866871558082,
+      "learning_rate": 3.377367943710183e-07,
+      "loss": 0.4281,
+      "step": 10314
+    },
+    {
+      "epoch": 0.837732477868919,
+      "grad_norm": 4.797876474078643,
+      "learning_rate": 3.374067855260055e-07,
+      "loss": 0.4421,
+      "step": 10315
+    },
+    {
+      "epoch": 0.8378136928449607,
+      "grad_norm": 4.661745082017848,
+      "learning_rate": 3.370769263209975e-07,
+      "loss": 0.3877,
+      "step": 10316
+    },
+    {
+      "epoch": 0.8378949078210022,
+      "grad_norm": 4.545258248925443,
+      "learning_rate": 3.3674721677881853e-07,
+      "loss": 0.5798,
+      "step": 10317
+    },
+    {
+      "epoch": 0.8379761227970438,
+      "grad_norm": 6.305655433495236,
+      "learning_rate": 3.364176569222843e-07,
+      "loss": 0.5197,
+      "step": 10318
+    },
+    {
+      "epoch": 0.8380573377730853,
+      "grad_norm": 5.2525205813121945,
+      "learning_rate": 3.360882467741969e-07,
+      "loss": 0.4127,
+      "step": 10319
+    },
+    {
+      "epoch": 0.838138552749127,
+      "grad_norm": 4.370136047848057,
+      "learning_rate": 3.35758986357351e-07,
+      "loss": 0.6077,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8382197677251685,
+      "grad_norm": 10.028481598001118,
+      "learning_rate": 3.354298756945293e-07,
+      "loss": 0.3249,
+      "step": 10321
+    },
+    {
+      "epoch": 0.8383009827012101,
+      "grad_norm": 4.972555184812446,
+      "learning_rate": 3.351009148085038e-07,
+      "loss": 0.5867,
+      "step": 10322
+    },
+    {
+      "epoch": 0.8383821976772516,
+      "grad_norm": 3.9492347620594224,
+      "learning_rate": 3.347721037220372e-07,
+      "loss": 0.4575,
+      "step": 10323
+    },
+    {
+      "epoch": 0.8384634126532933,
+      "grad_norm": 24.137740084826742,
+      "learning_rate": 3.344434424578824e-07,
+      "loss": 0.5544,
+      "step": 10324
+    },
+    {
+      "epoch": 0.8385446276293349,
+      "grad_norm": 7.000852410859473,
+      "learning_rate": 3.3411493103878036e-07,
+      "loss": 0.4401,
+      "step": 10325
+    },
+    {
+      "epoch": 0.8386258426053764,
+      "grad_norm": 6.577988762799303,
+      "learning_rate": 3.3378656948746176e-07,
+      "loss": 0.4975,
+      "step": 10326
+    },
+    {
+      "epoch": 0.8387070575814181,
+      "grad_norm": 5.455565928416991,
+      "learning_rate": 3.334583578266487e-07,
+      "loss": 0.489,
+      "step": 10327
+    },
+    {
+      "epoch": 0.8387882725574596,
+      "grad_norm": 4.806945898095566,
+      "learning_rate": 3.3313029607905087e-07,
+      "loss": 0.394,
+      "step": 10328
+    },
+    {
+      "epoch": 0.8388694875335012,
+      "grad_norm": 5.601127834993915,
+      "learning_rate": 3.328023842673678e-07,
+      "loss": 0.5171,
+      "step": 10329
+    },
+    {
+      "epoch": 0.8389507025095427,
+      "grad_norm": 8.23233143842366,
+      "learning_rate": 3.324746224142902e-07,
+      "loss": 0.3461,
+      "step": 10330
+    },
+    {
+      "epoch": 0.8390319174855844,
+      "grad_norm": 5.534926193999312,
+      "learning_rate": 3.321470105424979e-07,
+      "loss": 0.4056,
+      "step": 10331
+    },
+    {
+      "epoch": 0.8391131324616259,
+      "grad_norm": 4.70029893588722,
+      "learning_rate": 3.3181954867465864e-07,
+      "loss": 0.5831,
+      "step": 10332
+    },
+    {
+      "epoch": 0.8391943474376675,
+      "grad_norm": 4.310642438062851,
+      "learning_rate": 3.314922368334322e-07,
+      "loss": 0.4913,
+      "step": 10333
+    },
+    {
+      "epoch": 0.839275562413709,
+      "grad_norm": 3.3683032390188012,
+      "learning_rate": 3.3116507504146633e-07,
+      "loss": 0.5738,
+      "step": 10334
+    },
+    {
+      "epoch": 0.8393567773897507,
+      "grad_norm": 4.726658134420679,
+      "learning_rate": 3.3083806332139837e-07,
+      "loss": 0.3906,
+      "step": 10335
+    },
+    {
+      "epoch": 0.8394379923657923,
+      "grad_norm": 4.793027997624646,
+      "learning_rate": 3.305112016958562e-07,
+      "loss": 0.3441,
+      "step": 10336
+    },
+    {
+      "epoch": 0.8395192073418338,
+      "grad_norm": 15.53691157648671,
+      "learning_rate": 3.3018449018745765e-07,
+      "loss": 0.4705,
+      "step": 10337
+    },
+    {
+      "epoch": 0.8396004223178755,
+      "grad_norm": 5.292837088334354,
+      "learning_rate": 3.298579288188081e-07,
+      "loss": 0.5973,
+      "step": 10338
+    },
+    {
+      "epoch": 0.839681637293917,
+      "grad_norm": 7.13356115130351,
+      "learning_rate": 3.2953151761250526e-07,
+      "loss": 0.4081,
+      "step": 10339
+    },
+    {
+      "epoch": 0.8397628522699586,
+      "grad_norm": 4.657968605497615,
+      "learning_rate": 3.292052565911344e-07,
+      "loss": 0.3476,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8398440672460001,
+      "grad_norm": 5.343433901408026,
+      "learning_rate": 3.288791457772708e-07,
+      "loss": 0.4242,
+      "step": 10341
+    },
+    {
+      "epoch": 0.8399252822220418,
+      "grad_norm": 7.943551058264865,
+      "learning_rate": 3.2855318519347924e-07,
+      "loss": 0.3954,
+      "step": 10342
+    },
+    {
+      "epoch": 0.8400064971980833,
+      "grad_norm": 4.253480461006163,
+      "learning_rate": 3.282273748623152e-07,
+      "loss": 0.6146,
+      "step": 10343
+    },
+    {
+      "epoch": 0.8400877121741249,
+      "grad_norm": 3.649658017480646,
+      "learning_rate": 3.279017148063235e-07,
+      "loss": 0.4316,
+      "step": 10344
+    },
+    {
+      "epoch": 0.8401689271501664,
+      "grad_norm": 6.247382983100074,
+      "learning_rate": 3.275762050480369e-07,
+      "loss": 0.6303,
+      "step": 10345
+    },
+    {
+      "epoch": 0.8402501421262081,
+      "grad_norm": 5.081757914347488,
+      "learning_rate": 3.272508456099799e-07,
+      "loss": 0.3951,
+      "step": 10346
+    },
+    {
+      "epoch": 0.8403313571022497,
+      "grad_norm": 4.2849275181266515,
+      "learning_rate": 3.269256365146653e-07,
+      "loss": 0.4885,
+      "step": 10347
+    },
+    {
+      "epoch": 0.8404125720782912,
+      "grad_norm": 5.408392743532504,
+      "learning_rate": 3.2660057778459513e-07,
+      "loss": 0.5593,
+      "step": 10348
+    },
+    {
+      "epoch": 0.8404937870543329,
+      "grad_norm": 16.566283426785315,
+      "learning_rate": 3.262756694422628e-07,
+      "loss": 0.4464,
+      "step": 10349
+    },
+    {
+      "epoch": 0.8405750020303744,
+      "grad_norm": 9.235201670008626,
+      "learning_rate": 3.2595091151015e-07,
+      "loss": 0.4379,
+      "step": 10350
+    },
+    {
+      "epoch": 0.840656217006416,
+      "grad_norm": 3.706535071305654,
+      "learning_rate": 3.2562630401072796e-07,
+      "loss": 0.3997,
+      "step": 10351
+    },
+    {
+      "epoch": 0.8407374319824575,
+      "grad_norm": 6.291131527025278,
+      "learning_rate": 3.2530184696645846e-07,
+      "loss": 0.3929,
+      "step": 10352
+    },
+    {
+      "epoch": 0.8408186469584992,
+      "grad_norm": 6.059066502473035,
+      "learning_rate": 3.249775403997915e-07,
+      "loss": 0.5271,
+      "step": 10353
+    },
+    {
+      "epoch": 0.8408998619345407,
+      "grad_norm": 5.828503261223924,
+      "learning_rate": 3.24653384333167e-07,
+      "loss": 0.4112,
+      "step": 10354
+    },
+    {
+      "epoch": 0.8409810769105823,
+      "grad_norm": 5.607819748297354,
+      "learning_rate": 3.243293787890162e-07,
+      "loss": 0.5535,
+      "step": 10355
+    },
+    {
+      "epoch": 0.8410622918866238,
+      "grad_norm": 4.947445738957033,
+      "learning_rate": 3.2400552378975744e-07,
+      "loss": 0.5297,
+      "step": 10356
+    },
+    {
+      "epoch": 0.8411435068626655,
+      "grad_norm": 12.539916176915135,
+      "learning_rate": 3.236818193577998e-07,
+      "loss": 0.5609,
+      "step": 10357
+    },
+    {
+      "epoch": 0.8412247218387071,
+      "grad_norm": 7.269085157428185,
+      "learning_rate": 3.233582655155429e-07,
+      "loss": 0.4367,
+      "step": 10358
+    },
+    {
+      "epoch": 0.8413059368147486,
+      "grad_norm": 7.3313861075084725,
+      "learning_rate": 3.2303486228537436e-07,
+      "loss": 0.5678,
+      "step": 10359
+    },
+    {
+      "epoch": 0.8413871517907903,
+      "grad_norm": 6.798734073899363,
+      "learning_rate": 3.227116096896718e-07,
+      "loss": 0.4094,
+      "step": 10360
+    },
+    {
+      "epoch": 0.8414683667668318,
+      "grad_norm": 5.446235513662214,
+      "learning_rate": 3.223885077508024e-07,
+      "loss": 0.5598,
+      "step": 10361
+    },
+    {
+      "epoch": 0.8415495817428734,
+      "grad_norm": 5.464432211318365,
+      "learning_rate": 3.220655564911232e-07,
+      "loss": 0.5195,
+      "step": 10362
+    },
+    {
+      "epoch": 0.8416307967189149,
+      "grad_norm": 3.4734065266272767,
+      "learning_rate": 3.217427559329814e-07,
+      "loss": 0.4573,
+      "step": 10363
+    },
+    {
+      "epoch": 0.8417120116949566,
+      "grad_norm": 4.612719279186468,
+      "learning_rate": 3.2142010609871236e-07,
+      "loss": 0.4274,
+      "step": 10364
+    },
+    {
+      "epoch": 0.8417932266709981,
+      "grad_norm": 5.548612820561558,
+      "learning_rate": 3.2109760701064227e-07,
+      "loss": 0.4346,
+      "step": 10365
+    },
+    {
+      "epoch": 0.8418744416470397,
+      "grad_norm": 7.436228501727278,
+      "learning_rate": 3.207752586910862e-07,
+      "loss": 0.4898,
+      "step": 10366
+    },
+    {
+      "epoch": 0.8419556566230812,
+      "grad_norm": 7.589666228758146,
+      "learning_rate": 3.2045306116234824e-07,
+      "loss": 0.5607,
+      "step": 10367
+    },
+    {
+      "epoch": 0.8420368715991229,
+      "grad_norm": 5.181080777503049,
+      "learning_rate": 3.2013101444672345e-07,
+      "loss": 0.4397,
+      "step": 10368
+    },
+    {
+      "epoch": 0.8421180865751645,
+      "grad_norm": 7.40645836332802,
+      "learning_rate": 3.198091185664964e-07,
+      "loss": 0.5543,
+      "step": 10369
+    },
+    {
+      "epoch": 0.842199301551206,
+      "grad_norm": 4.3886565655110035,
+      "learning_rate": 3.194873735439391e-07,
+      "loss": 0.394,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8422805165272477,
+      "grad_norm": 5.3852633366108655,
+      "learning_rate": 3.1916577940131585e-07,
+      "loss": 0.6436,
+      "step": 10371
+    },
+    {
+      "epoch": 0.8423617315032892,
+      "grad_norm": 4.336066931433411,
+      "learning_rate": 3.188443361608787e-07,
+      "loss": 0.5485,
+      "step": 10372
+    },
+    {
+      "epoch": 0.8424429464793308,
+      "grad_norm": 7.090559367854015,
+      "learning_rate": 3.185230438448694e-07,
+      "loss": 0.5496,
+      "step": 10373
+    },
+    {
+      "epoch": 0.8425241614553723,
+      "grad_norm": 5.001322891880096,
+      "learning_rate": 3.182019024755209e-07,
+      "loss": 0.5178,
+      "step": 10374
+    },
+    {
+      "epoch": 0.842605376431414,
+      "grad_norm": 8.355308016625518,
+      "learning_rate": 3.1788091207505285e-07,
+      "loss": 0.5462,
+      "step": 10375
+    },
+    {
+      "epoch": 0.8426865914074555,
+      "grad_norm": 5.941932747298044,
+      "learning_rate": 3.175600726656772e-07,
+      "loss": 0.3143,
+      "step": 10376
+    },
+    {
+      "epoch": 0.8427678063834971,
+      "grad_norm": 13.808992644334168,
+      "learning_rate": 3.172393842695948e-07,
+      "loss": 0.5841,
+      "step": 10377
+    },
+    {
+      "epoch": 0.8428490213595387,
+      "grad_norm": 4.737960132276757,
+      "learning_rate": 3.169188469089945e-07,
+      "loss": 0.6052,
+      "step": 10378
+    },
+    {
+      "epoch": 0.8429302363355803,
+      "grad_norm": 6.490398809975363,
+      "learning_rate": 3.165984606060565e-07,
+      "loss": 0.4312,
+      "step": 10379
+    },
+    {
+      "epoch": 0.8430114513116219,
+      "grad_norm": 6.469336726754756,
+      "learning_rate": 3.1627822538294883e-07,
+      "loss": 0.381,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8430926662876634,
+      "grad_norm": 4.644733440027643,
+      "learning_rate": 3.159581412618309e-07,
+      "loss": 0.488,
+      "step": 10381
+    },
+    {
+      "epoch": 0.8431738812637051,
+      "grad_norm": 5.02222613398405,
+      "learning_rate": 3.1563820826485127e-07,
+      "loss": 0.4021,
+      "step": 10382
+    },
+    {
+      "epoch": 0.8432550962397466,
+      "grad_norm": 5.6081490476021285,
+      "learning_rate": 3.153184264141465e-07,
+      "loss": 0.4034,
+      "step": 10383
+    },
+    {
+      "epoch": 0.8433363112157882,
+      "grad_norm": 12.62989950175683,
+      "learning_rate": 3.1499879573184486e-07,
+      "loss": 0.4546,
+      "step": 10384
+    },
+    {
+      "epoch": 0.8434175261918297,
+      "grad_norm": 4.275321566511599,
+      "learning_rate": 3.146793162400627e-07,
+      "loss": 0.3961,
+      "step": 10385
+    },
+    {
+      "epoch": 0.8434987411678714,
+      "grad_norm": 5.910594667196357,
+      "learning_rate": 3.143599879609055e-07,
+      "loss": 0.4436,
+      "step": 10386
+    },
+    {
+      "epoch": 0.8435799561439129,
+      "grad_norm": 4.218477387154232,
+      "learning_rate": 3.1404081091647027e-07,
+      "loss": 0.5959,
+      "step": 10387
+    },
+    {
+      "epoch": 0.8436611711199545,
+      "grad_norm": 7.34082449004889,
+      "learning_rate": 3.1372178512884154e-07,
+      "loss": 0.6363,
+      "step": 10388
+    },
+    {
+      "epoch": 0.843742386095996,
+      "grad_norm": 5.370383548087424,
+      "learning_rate": 3.1340291062009446e-07,
+      "loss": 0.5443,
+      "step": 10389
+    },
+    {
+      "epoch": 0.8438236010720377,
+      "grad_norm": 3.265960200811652,
+      "learning_rate": 3.130841874122942e-07,
+      "loss": 0.4581,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8439048160480793,
+      "grad_norm": 8.155289410608034,
+      "learning_rate": 3.1276561552749415e-07,
+      "loss": 0.5283,
+      "step": 10391
+    },
+    {
+      "epoch": 0.8439860310241208,
+      "grad_norm": 4.910342387320026,
+      "learning_rate": 3.1244719498773693e-07,
+      "loss": 0.4737,
+      "step": 10392
+    },
+    {
+      "epoch": 0.8440672460001625,
+      "grad_norm": 7.838611981148996,
+      "learning_rate": 3.1212892581505697e-07,
+      "loss": 0.5164,
+      "step": 10393
+    },
+    {
+      "epoch": 0.844148460976204,
+      "grad_norm": 3.137926778602015,
+      "learning_rate": 3.118108080314758e-07,
+      "loss": 0.6133,
+      "step": 10394
+    },
+    {
+      "epoch": 0.8442296759522456,
+      "grad_norm": 6.8077151459524,
+      "learning_rate": 3.1149284165900627e-07,
+      "loss": 0.3985,
+      "step": 10395
+    },
+    {
+      "epoch": 0.8443108909282871,
+      "grad_norm": 4.575748337907764,
+      "learning_rate": 3.111750267196492e-07,
+      "loss": 0.6132,
+      "step": 10396
+    },
+    {
+      "epoch": 0.8443921059043288,
+      "grad_norm": 3.4885659741848833,
+      "learning_rate": 3.1085736323539647e-07,
+      "loss": 0.584,
+      "step": 10397
+    },
+    {
+      "epoch": 0.8444733208803703,
+      "grad_norm": 5.967423076508702,
+      "learning_rate": 3.1053985122822844e-07,
+      "loss": 0.6819,
+      "step": 10398
+    },
+    {
+      "epoch": 0.8445545358564119,
+      "grad_norm": 4.9263444027598124,
+      "learning_rate": 3.1022249072011455e-07,
+      "loss": 0.5207,
+      "step": 10399
+    },
+    {
+      "epoch": 0.8446357508324535,
+      "grad_norm": 7.840385643435142,
+      "learning_rate": 3.0990528173301557e-07,
+      "loss": 0.4711,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8447169658084951,
+      "grad_norm": 6.013199743602192,
+      "learning_rate": 3.095882242888795e-07,
+      "loss": 0.4817,
+      "step": 10401
+    },
+    {
+      "epoch": 0.8447981807845367,
+      "grad_norm": 6.574555556543331,
+      "learning_rate": 3.09271318409646e-07,
+      "loss": 0.4581,
+      "step": 10402
+    },
+    {
+      "epoch": 0.8448793957605782,
+      "grad_norm": 3.106518820205016,
+      "learning_rate": 3.089545641172434e-07,
+      "loss": 0.549,
+      "step": 10403
+    },
+    {
+      "epoch": 0.8449606107366199,
+      "grad_norm": 4.4853372930685715,
+      "learning_rate": 3.086379614335891e-07,
+      "loss": 0.3649,
+      "step": 10404
+    },
+    {
+      "epoch": 0.8450418257126614,
+      "grad_norm": 6.692195853008346,
+      "learning_rate": 3.083215103805895e-07,
+      "loss": 0.5595,
+      "step": 10405
+    },
+    {
+      "epoch": 0.845123040688703,
+      "grad_norm": 3.961593687618352,
+      "learning_rate": 3.080052109801429e-07,
+      "loss": 0.4905,
+      "step": 10406
+    },
+    {
+      "epoch": 0.8452042556647446,
+      "grad_norm": 5.657024124857219,
+      "learning_rate": 3.0768906325413404e-07,
+      "loss": 0.304,
+      "step": 10407
+    },
+    {
+      "epoch": 0.8452854706407862,
+      "grad_norm": 4.947735022977894,
+      "learning_rate": 3.073730672244393e-07,
+      "loss": 0.5874,
+      "step": 10408
+    },
+    {
+      "epoch": 0.8453666856168277,
+      "grad_norm": 4.804603939591969,
+      "learning_rate": 3.0705722291292457e-07,
+      "loss": 0.5255,
+      "step": 10409
+    },
+    {
+      "epoch": 0.8454479005928693,
+      "grad_norm": 8.27849842792289,
+      "learning_rate": 3.067415303414442e-07,
+      "loss": 0.5203,
+      "step": 10410
+    },
+    {
+      "epoch": 0.8455291155689109,
+      "grad_norm": 10.32641648689632,
+      "learning_rate": 3.0642598953184164e-07,
+      "loss": 0.4933,
+      "step": 10411
+    },
+    {
+      "epoch": 0.8456103305449525,
+      "grad_norm": 3.764527737605284,
+      "learning_rate": 3.0611060050595166e-07,
+      "loss": 0.5453,
+      "step": 10412
+    },
+    {
+      "epoch": 0.8456915455209941,
+      "grad_norm": 5.190043324766475,
+      "learning_rate": 3.057953632855973e-07,
+      "loss": 0.5636,
+      "step": 10413
+    },
+    {
+      "epoch": 0.8457727604970356,
+      "grad_norm": 4.262101482407405,
+      "learning_rate": 3.0548027789259057e-07,
+      "loss": 0.5088,
+      "step": 10414
+    },
+    {
+      "epoch": 0.8458539754730773,
+      "grad_norm": 4.496146141780251,
+      "learning_rate": 3.05165344348734e-07,
+      "loss": 0.4839,
+      "step": 10415
+    },
+    {
+      "epoch": 0.8459351904491188,
+      "grad_norm": 5.9108085665989645,
+      "learning_rate": 3.0485056267582054e-07,
+      "loss": 0.6495,
+      "step": 10416
+    },
+    {
+      "epoch": 0.8460164054251604,
+      "grad_norm": 5.4963646098218355,
+      "learning_rate": 3.0453593289563015e-07,
+      "loss": 0.4643,
+      "step": 10417
+    },
+    {
+      "epoch": 0.846097620401202,
+      "grad_norm": 11.437063436049456,
+      "learning_rate": 3.0422145502993355e-07,
+      "loss": 0.4108,
+      "step": 10418
+    },
+    {
+      "epoch": 0.8461788353772436,
+      "grad_norm": 6.788003957398547,
+      "learning_rate": 3.0390712910049166e-07,
+      "loss": 0.3901,
+      "step": 10419
+    },
+    {
+      "epoch": 0.8462600503532851,
+      "grad_norm": 4.716129840076831,
+      "learning_rate": 3.035929551290534e-07,
+      "loss": 0.3811,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8463412653293267,
+      "grad_norm": 4.429133558600282,
+      "learning_rate": 3.0327893313735814e-07,
+      "loss": 0.3569,
+      "step": 10421
+    },
+    {
+      "epoch": 0.8464224803053683,
+      "grad_norm": 5.2938219799131865,
+      "learning_rate": 3.0296506314713534e-07,
+      "loss": 0.4368,
+      "step": 10422
+    },
+    {
+      "epoch": 0.8465036952814099,
+      "grad_norm": 4.19248350579702,
+      "learning_rate": 3.0265134518010274e-07,
+      "loss": 0.5749,
+      "step": 10423
+    },
+    {
+      "epoch": 0.8465849102574515,
+      "grad_norm": 8.957100181001996,
+      "learning_rate": 3.0233777925796683e-07,
+      "loss": 0.4146,
+      "step": 10424
+    },
+    {
+      "epoch": 0.846666125233493,
+      "grad_norm": 4.999603068660096,
+      "learning_rate": 3.020243654024266e-07,
+      "loss": 0.3864,
+      "step": 10425
+    },
+    {
+      "epoch": 0.8467473402095347,
+      "grad_norm": 18.067549039426147,
+      "learning_rate": 3.017111036351672e-07,
+      "loss": 0.4653,
+      "step": 10426
+    },
+    {
+      "epoch": 0.8468285551855762,
+      "grad_norm": 6.599994188483718,
+      "learning_rate": 3.01397993977865e-07,
+      "loss": 0.4479,
+      "step": 10427
+    },
+    {
+      "epoch": 0.8469097701616178,
+      "grad_norm": 5.582107888533948,
+      "learning_rate": 3.010850364521853e-07,
+      "loss": 0.5203,
+      "step": 10428
+    },
+    {
+      "epoch": 0.8469909851376594,
+      "grad_norm": 4.2550759235692555,
+      "learning_rate": 3.007722310797842e-07,
+      "loss": 0.463,
+      "step": 10429
+    },
+    {
+      "epoch": 0.847072200113701,
+      "grad_norm": 4.496695788396444,
+      "learning_rate": 3.004595778823055e-07,
+      "loss": 0.6158,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8471534150897425,
+      "grad_norm": 4.611503155703553,
+      "learning_rate": 3.0014707688138244e-07,
+      "loss": 0.504,
+      "step": 10431
+    },
+    {
+      "epoch": 0.8472346300657841,
+      "grad_norm": 5.19185166511966,
+      "learning_rate": 2.9983472809863996e-07,
+      "loss": 0.5022,
+      "step": 10432
+    },
+    {
+      "epoch": 0.8473158450418257,
+      "grad_norm": 4.311886074274771,
+      "learning_rate": 2.995225315556891e-07,
+      "loss": 0.5097,
+      "step": 10433
+    },
+    {
+      "epoch": 0.8473970600178673,
+      "grad_norm": 5.401400848063594,
+      "learning_rate": 2.992104872741336e-07,
+      "loss": 0.4966,
+      "step": 10434
+    },
+    {
+      "epoch": 0.8474782749939089,
+      "grad_norm": 4.369156116773576,
+      "learning_rate": 2.9889859527556517e-07,
+      "loss": 0.5447,
+      "step": 10435
+    },
+    {
+      "epoch": 0.8475594899699505,
+      "grad_norm": 4.166723822178224,
+      "learning_rate": 2.985868555815646e-07,
+      "loss": 0.5355,
+      "step": 10436
+    },
+    {
+      "epoch": 0.8476407049459921,
+      "grad_norm": 6.75124539957616,
+      "learning_rate": 2.9827526821370274e-07,
+      "loss": 0.6095,
+      "step": 10437
+    },
+    {
+      "epoch": 0.8477219199220336,
+      "grad_norm": 31.475529417596338,
+      "learning_rate": 2.9796383319353997e-07,
+      "loss": 0.4393,
+      "step": 10438
+    },
+    {
+      "epoch": 0.8478031348980752,
+      "grad_norm": 5.167153691651564,
+      "learning_rate": 2.976525505426253e-07,
+      "loss": 0.546,
+      "step": 10439
+    },
+    {
+      "epoch": 0.8478843498741168,
+      "grad_norm": 5.678318484729858,
+      "learning_rate": 2.9734142028249867e-07,
+      "loss": 0.5173,
+      "step": 10440
+    },
+    {
+      "epoch": 0.8479655648501584,
+      "grad_norm": 3.293870862181777,
+      "learning_rate": 2.970304424346887e-07,
+      "loss": 0.4497,
+      "step": 10441
+    },
+    {
+      "epoch": 0.8480467798261999,
+      "grad_norm": 7.221626978308798,
+      "learning_rate": 2.9671961702071314e-07,
+      "loss": 0.4407,
+      "step": 10442
+    },
+    {
+      "epoch": 0.8481279948022415,
+      "grad_norm": 17.363810182049495,
+      "learning_rate": 2.9640894406207875e-07,
+      "loss": 0.5855,
+      "step": 10443
+    },
+    {
+      "epoch": 0.8482092097782831,
+      "grad_norm": 74.21346545661238,
+      "learning_rate": 2.960984235802836e-07,
+      "loss": 0.5255,
+      "step": 10444
+    },
+    {
+      "epoch": 0.8482904247543247,
+      "grad_norm": 5.175532691743031,
+      "learning_rate": 2.957880555968137e-07,
+      "loss": 0.6829,
+      "step": 10445
+    },
+    {
+      "epoch": 0.8483716397303663,
+      "grad_norm": 6.761383142028467,
+      "learning_rate": 2.95477840133144e-07,
+      "loss": 0.4693,
+      "step": 10446
+    },
+    {
+      "epoch": 0.8484528547064079,
+      "grad_norm": 3.230874015786307,
+      "learning_rate": 2.951677772107406e-07,
+      "loss": 0.4897,
+      "step": 10447
+    },
+    {
+      "epoch": 0.8485340696824495,
+      "grad_norm": 4.35547207648812,
+      "learning_rate": 2.9485786685105876e-07,
+      "loss": 0.3499,
+      "step": 10448
+    },
+    {
+      "epoch": 0.848615284658491,
+      "grad_norm": 4.106258252858656,
+      "learning_rate": 2.945481090755417e-07,
+      "loss": 0.5062,
+      "step": 10449
+    },
+    {
+      "epoch": 0.8486964996345326,
+      "grad_norm": 7.391399136928735,
+      "learning_rate": 2.942385039056231e-07,
+      "loss": 0.5064,
+      "step": 10450
+    },
+    {
+      "epoch": 0.8487777146105742,
+      "grad_norm": 4.053993280844301,
+      "learning_rate": 2.939290513627266e-07,
+      "loss": 0.5534,
+      "step": 10451
+    },
+    {
+      "epoch": 0.8488589295866158,
+      "grad_norm": 7.209839450243032,
+      "learning_rate": 2.936197514682637e-07,
+      "loss": 0.5145,
+      "step": 10452
+    },
+    {
+      "epoch": 0.8489401445626573,
+      "grad_norm": 4.068133193825577,
+      "learning_rate": 2.933106042436368e-07,
+      "loss": 0.5653,
+      "step": 10453
+    },
+    {
+      "epoch": 0.849021359538699,
+      "grad_norm": 7.446752715397537,
+      "learning_rate": 2.930016097102378e-07,
+      "loss": 0.4394,
+      "step": 10454
+    },
+    {
+      "epoch": 0.8491025745147405,
+      "grad_norm": 8.151515842737028,
+      "learning_rate": 2.9269276788944726e-07,
+      "loss": 0.4768,
+      "step": 10455
+    },
+    {
+      "epoch": 0.8491837894907821,
+      "grad_norm": 4.210590324590446,
+      "learning_rate": 2.923840788026347e-07,
+      "loss": 0.5837,
+      "step": 10456
+    },
+    {
+      "epoch": 0.8492650044668237,
+      "grad_norm": 4.153598656728726,
+      "learning_rate": 2.9207554247116047e-07,
+      "loss": 0.4728,
+      "step": 10457
+    },
+    {
+      "epoch": 0.8493462194428653,
+      "grad_norm": 4.033479881839166,
+      "learning_rate": 2.917671589163737e-07,
+      "loss": 0.5434,
+      "step": 10458
+    },
+    {
+      "epoch": 0.8494274344189069,
+      "grad_norm": 4.9531557828779045,
+      "learning_rate": 2.9145892815961194e-07,
+      "loss": 0.3945,
+      "step": 10459
+    },
+    {
+      "epoch": 0.8495086493949484,
+      "grad_norm": 5.232476687979792,
+      "learning_rate": 2.911508502222041e-07,
+      "loss": 0.6322,
+      "step": 10460
+    },
+    {
+      "epoch": 0.84958986437099,
+      "grad_norm": 2.92678651363929,
+      "learning_rate": 2.908429251254674e-07,
+      "loss": 0.4652,
+      "step": 10461
+    },
+    {
+      "epoch": 0.8496710793470316,
+      "grad_norm": 6.325678937933874,
+      "learning_rate": 2.90535152890708e-07,
+      "loss": 0.4732,
+      "step": 10462
+    },
+    {
+      "epoch": 0.8497522943230732,
+      "grad_norm": 5.685417644326699,
+      "learning_rate": 2.902275335392232e-07,
+      "loss": 0.6602,
+      "step": 10463
+    },
+    {
+      "epoch": 0.8498335092991147,
+      "grad_norm": 4.163035284682951,
+      "learning_rate": 2.8992006709229803e-07,
+      "loss": 0.3988,
+      "step": 10464
+    },
+    {
+      "epoch": 0.8499147242751564,
+      "grad_norm": 4.57874278542473,
+      "learning_rate": 2.8961275357120704e-07,
+      "loss": 0.4362,
+      "step": 10465
+    },
+    {
+      "epoch": 0.8499959392511979,
+      "grad_norm": 5.572554269055086,
+      "learning_rate": 2.893055929972152e-07,
+      "loss": 0.4934,
+      "step": 10466
+    },
+    {
+      "epoch": 0.8500771542272395,
+      "grad_norm": 10.65165020002116,
+      "learning_rate": 2.8899858539157694e-07,
+      "loss": 0.5137,
+      "step": 10467
+    },
+    {
+      "epoch": 0.8501583692032811,
+      "grad_norm": 9.776998737199031,
+      "learning_rate": 2.886917307755349e-07,
+      "loss": 0.4341,
+      "step": 10468
+    },
+    {
+      "epoch": 0.8502395841793227,
+      "grad_norm": 6.586810382710384,
+      "learning_rate": 2.8838502917032136e-07,
+      "loss": 0.4922,
+      "step": 10469
+    },
+    {
+      "epoch": 0.8503207991553643,
+      "grad_norm": 5.154039748490755,
+      "learning_rate": 2.880784805971595e-07,
+      "loss": 0.3545,
+      "step": 10470
+    },
+    {
+      "epoch": 0.8504020141314058,
+      "grad_norm": 3.9568608557183516,
+      "learning_rate": 2.8777208507726056e-07,
+      "loss": 0.4876,
+      "step": 10471
+    },
+    {
+      "epoch": 0.8504832291074474,
+      "grad_norm": 4.288857280998075,
+      "learning_rate": 2.874658426318244e-07,
+      "loss": 0.4853,
+      "step": 10472
+    },
+    {
+      "epoch": 0.850564444083489,
+      "grad_norm": 4.368402834061044,
+      "learning_rate": 2.871597532820425e-07,
+      "loss": 0.4401,
+      "step": 10473
+    },
+    {
+      "epoch": 0.8506456590595306,
+      "grad_norm": 4.033924407915728,
+      "learning_rate": 2.86853817049095e-07,
+      "loss": 0.5619,
+      "step": 10474
+    },
+    {
+      "epoch": 0.8507268740355721,
+      "grad_norm": 5.6815820407123585,
+      "learning_rate": 2.865480339541496e-07,
+      "loss": 0.4642,
+      "step": 10475
+    },
+    {
+      "epoch": 0.8508080890116138,
+      "grad_norm": 3.948873040109591,
+      "learning_rate": 2.8624240401836647e-07,
+      "loss": 0.4468,
+      "step": 10476
+    },
+    {
+      "epoch": 0.8508893039876553,
+      "grad_norm": 3.2353533973872084,
+      "learning_rate": 2.859369272628928e-07,
+      "loss": 0.532,
+      "step": 10477
+    },
+    {
+      "epoch": 0.8509705189636969,
+      "grad_norm": 3.308823715003122,
+      "learning_rate": 2.856316037088655e-07,
+      "loss": 0.4157,
+      "step": 10478
+    },
+    {
+      "epoch": 0.8510517339397385,
+      "grad_norm": 13.811883074908797,
+      "learning_rate": 2.8532643337741195e-07,
+      "loss": 0.4894,
+      "step": 10479
+    },
+    {
+      "epoch": 0.8511329489157801,
+      "grad_norm": 3.2224568567146945,
+      "learning_rate": 2.8502141628964836e-07,
+      "loss": 0.4306,
+      "step": 10480
+    },
+    {
+      "epoch": 0.8512141638918217,
+      "grad_norm": 3.9763076127585313,
+      "learning_rate": 2.8471655246668007e-07,
+      "loss": 0.5575,
+      "step": 10481
+    },
+    {
+      "epoch": 0.8512953788678632,
+      "grad_norm": 6.740981385096001,
+      "learning_rate": 2.844118419296024e-07,
+      "loss": 0.4055,
+      "step": 10482
+    },
+    {
+      "epoch": 0.8513765938439048,
+      "grad_norm": 20.280874148042784,
+      "learning_rate": 2.841072846994994e-07,
+      "loss": 0.3875,
+      "step": 10483
+    },
+    {
+      "epoch": 0.8514578088199464,
+      "grad_norm": 9.735322905143065,
+      "learning_rate": 2.8380288079744494e-07,
+      "loss": 0.3133,
+      "step": 10484
+    },
+    {
+      "epoch": 0.851539023795988,
+      "grad_norm": 4.313636812248141,
+      "learning_rate": 2.8349863024450143e-07,
+      "loss": 0.4564,
+      "step": 10485
+    },
+    {
+      "epoch": 0.8516202387720295,
+      "grad_norm": 6.764374322762755,
+      "learning_rate": 2.8319453306172225e-07,
+      "loss": 0.396,
+      "step": 10486
+    },
+    {
+      "epoch": 0.8517014537480712,
+      "grad_norm": 5.159578175212937,
+      "learning_rate": 2.8289058927014944e-07,
+      "loss": 0.5508,
+      "step": 10487
+    },
+    {
+      "epoch": 0.8517826687241127,
+      "grad_norm": 9.297209132292254,
+      "learning_rate": 2.8258679889081346e-07,
+      "loss": 0.4523,
+      "step": 10488
+    },
+    {
+      "epoch": 0.8518638837001543,
+      "grad_norm": 3.958399617586093,
+      "learning_rate": 2.8228316194473607e-07,
+      "loss": 0.4515,
+      "step": 10489
+    },
+    {
+      "epoch": 0.8519450986761959,
+      "grad_norm": 6.059112536774565,
+      "learning_rate": 2.8197967845292687e-07,
+      "loss": 0.5989,
+      "step": 10490
+    },
+    {
+      "epoch": 0.8520263136522375,
+      "grad_norm": 7.811507918796633,
+      "learning_rate": 2.8167634843638434e-07,
+      "loss": 0.3805,
+      "step": 10491
+    },
+    {
+      "epoch": 0.8521075286282791,
+      "grad_norm": 6.346071562908303,
+      "learning_rate": 2.8137317191609864e-07,
+      "loss": 0.3511,
+      "step": 10492
+    },
+    {
+      "epoch": 0.8521887436043206,
+      "grad_norm": 8.889395270642837,
+      "learning_rate": 2.810701489130477e-07,
+      "loss": 0.3602,
+      "step": 10493
+    },
+    {
+      "epoch": 0.8522699585803623,
+      "grad_norm": 5.77337789424289,
+      "learning_rate": 2.807672794481986e-07,
+      "loss": 0.3249,
+      "step": 10494
+    },
+    {
+      "epoch": 0.8523511735564038,
+      "grad_norm": 5.856871874864799,
+      "learning_rate": 2.804645635425091e-07,
+      "loss": 0.4549,
+      "step": 10495
+    },
+    {
+      "epoch": 0.8524323885324454,
+      "grad_norm": 3.9708905797147924,
+      "learning_rate": 2.801620012169251e-07,
+      "loss": 0.4752,
+      "step": 10496
+    },
+    {
+      "epoch": 0.8525136035084869,
+      "grad_norm": 5.58347667715412,
+      "learning_rate": 2.7985959249238165e-07,
+      "loss": 0.5097,
+      "step": 10497
+    },
+    {
+      "epoch": 0.8525948184845286,
+      "grad_norm": 5.356550864085213,
+      "learning_rate": 2.7955733738980443e-07,
+      "loss": 0.3949,
+      "step": 10498
+    },
+    {
+      "epoch": 0.8526760334605701,
+      "grad_norm": 5.227735405567097,
+      "learning_rate": 2.792552359301087e-07,
+      "loss": 0.4579,
+      "step": 10499
+    },
+    {
+      "epoch": 0.8527572484366117,
+      "grad_norm": 10.79645556327445,
+      "learning_rate": 2.789532881341969e-07,
+      "loss": 0.5133,
+      "step": 10500
+    },
+    {
+      "epoch": 0.8528384634126533,
+      "grad_norm": 9.986241964497967,
+      "learning_rate": 2.786514940229634e-07,
+      "loss": 0.4597,
+      "step": 10501
+    },
+    {
+      "epoch": 0.8529196783886949,
+      "grad_norm": 5.528139225360127,
+      "learning_rate": 2.7834985361728987e-07,
+      "loss": 0.312,
+      "step": 10502
+    },
+    {
+      "epoch": 0.8530008933647365,
+      "grad_norm": 5.4346809701316685,
+      "learning_rate": 2.7804836693804905e-07,
+      "loss": 0.6169,
+      "step": 10503
+    },
+    {
+      "epoch": 0.853082108340778,
+      "grad_norm": 4.421996352718119,
+      "learning_rate": 2.7774703400610086e-07,
+      "loss": 0.5074,
+      "step": 10504
+    },
+    {
+      "epoch": 0.8531633233168197,
+      "grad_norm": 12.764502226614324,
+      "learning_rate": 2.7744585484229674e-07,
+      "loss": 0.2709,
+      "step": 10505
+    },
+    {
+      "epoch": 0.8532445382928612,
+      "grad_norm": 5.85671022027329,
+      "learning_rate": 2.771448294674775e-07,
+      "loss": 0.4455,
+      "step": 10506
+    },
+    {
+      "epoch": 0.8533257532689028,
+      "grad_norm": 8.599669196683427,
+      "learning_rate": 2.768439579024712e-07,
+      "loss": 0.5653,
+      "step": 10507
+    },
+    {
+      "epoch": 0.8534069682449443,
+      "grad_norm": 6.732667207212168,
+      "learning_rate": 2.7654324016809757e-07,
+      "loss": 0.4152,
+      "step": 10508
+    },
+    {
+      "epoch": 0.853488183220986,
+      "grad_norm": 4.779828616945185,
+      "learning_rate": 2.7624267628516445e-07,
+      "loss": 0.4589,
+      "step": 10509
+    },
+    {
+      "epoch": 0.8535693981970275,
+      "grad_norm": 11.113748713116502,
+      "learning_rate": 2.759422662744682e-07,
+      "loss": 0.3779,
+      "step": 10510
+    },
+    {
+      "epoch": 0.8536506131730691,
+      "grad_norm": 4.481185872009129,
+      "learning_rate": 2.7564201015679664e-07,
+      "loss": 0.4627,
+      "step": 10511
+    },
+    {
+      "epoch": 0.8537318281491107,
+      "grad_norm": 4.723745824336893,
+      "learning_rate": 2.7534190795292626e-07,
+      "loss": 0.5168,
+      "step": 10512
+    },
+    {
+      "epoch": 0.8538130431251523,
+      "grad_norm": 4.5772924241630895,
+      "learning_rate": 2.750419596836215e-07,
+      "loss": 0.5351,
+      "step": 10513
+    },
+    {
+      "epoch": 0.8538942581011939,
+      "grad_norm": 4.890888601625891,
+      "learning_rate": 2.7474216536963803e-07,
+      "loss": 0.358,
+      "step": 10514
+    },
+    {
+      "epoch": 0.8539754730772354,
+      "grad_norm": 9.34358416833541,
+      "learning_rate": 2.744425250317201e-07,
+      "loss": 0.3687,
+      "step": 10515
+    },
+    {
+      "epoch": 0.8540566880532771,
+      "grad_norm": 4.739733132295761,
+      "learning_rate": 2.7414303869059994e-07,
+      "loss": 0.5659,
+      "step": 10516
+    },
+    {
+      "epoch": 0.8541379030293186,
+      "grad_norm": 4.5314061539867145,
+      "learning_rate": 2.7384370636700187e-07,
+      "loss": 0.4997,
+      "step": 10517
+    },
+    {
+      "epoch": 0.8542191180053602,
+      "grad_norm": 10.166709743253572,
+      "learning_rate": 2.735445280816373e-07,
+      "loss": 0.3698,
+      "step": 10518
+    },
+    {
+      "epoch": 0.8543003329814017,
+      "grad_norm": 6.233844850352306,
+      "learning_rate": 2.7324550385520844e-07,
+      "loss": 0.4164,
+      "step": 10519
+    },
+    {
+      "epoch": 0.8543815479574434,
+      "grad_norm": 4.1121635178190665,
+      "learning_rate": 2.72946633708405e-07,
+      "loss": 0.5892,
+      "step": 10520
+    },
+    {
+      "epoch": 0.8544627629334849,
+      "grad_norm": 10.454864690447497,
+      "learning_rate": 2.726479176619087e-07,
+      "loss": 0.5313,
+      "step": 10521
+    },
+    {
+      "epoch": 0.8545439779095265,
+      "grad_norm": 3.298486331376736,
+      "learning_rate": 2.723493557363885e-07,
+      "loss": 0.6448,
+      "step": 10522
+    },
+    {
+      "epoch": 0.8546251928855682,
+      "grad_norm": 4.677659267745939,
+      "learning_rate": 2.720509479525027e-07,
+      "loss": 0.5786,
+      "step": 10523
+    },
+    {
+      "epoch": 0.8547064078616097,
+      "grad_norm": 4.621472290562395,
+      "learning_rate": 2.7175269433089984e-07,
+      "loss": 0.4443,
+      "step": 10524
+    },
+    {
+      "epoch": 0.8547876228376513,
+      "grad_norm": 4.075558864406144,
+      "learning_rate": 2.7145459489221845e-07,
+      "loss": 0.4766,
+      "step": 10525
+    },
+    {
+      "epoch": 0.8548688378136928,
+      "grad_norm": 3.9376155282925427,
+      "learning_rate": 2.7115664965708387e-07,
+      "loss": 0.5365,
+      "step": 10526
+    },
+    {
+      "epoch": 0.8549500527897345,
+      "grad_norm": 4.9988126907241135,
+      "learning_rate": 2.708588586461139e-07,
+      "loss": 0.5849,
+      "step": 10527
+    },
+    {
+      "epoch": 0.855031267765776,
+      "grad_norm": 4.260322647357954,
+      "learning_rate": 2.7056122187991306e-07,
+      "loss": 0.5862,
+      "step": 10528
+    },
+    {
+      "epoch": 0.8551124827418176,
+      "grad_norm": 4.653669694516675,
+      "learning_rate": 2.7026373937907636e-07,
+      "loss": 0.6337,
+      "step": 10529
+    },
+    {
+      "epoch": 0.8551936977178591,
+      "grad_norm": 16.425837953856785,
+      "learning_rate": 2.6996641116418863e-07,
+      "loss": 0.3838,
+      "step": 10530
+    },
+    {
+      "epoch": 0.8552749126939008,
+      "grad_norm": 8.7333194311729,
+      "learning_rate": 2.696692372558224e-07,
+      "loss": 0.3394,
+      "step": 10531
+    },
+    {
+      "epoch": 0.8553561276699423,
+      "grad_norm": 6.271008188155812,
+      "learning_rate": 2.6937221767454086e-07,
+      "loss": 0.4557,
+      "step": 10532
+    },
+    {
+      "epoch": 0.8554373426459839,
+      "grad_norm": 4.33027681898429,
+      "learning_rate": 2.690753524408973e-07,
+      "loss": 0.4263,
+      "step": 10533
+    },
+    {
+      "epoch": 0.8555185576220256,
+      "grad_norm": 8.869308357187672,
+      "learning_rate": 2.6877864157543204e-07,
+      "loss": 0.4403,
+      "step": 10534
+    },
+    {
+      "epoch": 0.8555997725980671,
+      "grad_norm": 2.978988832273135,
+      "learning_rate": 2.684820850986758e-07,
+      "loss": 0.5313,
+      "step": 10535
+    },
+    {
+      "epoch": 0.8556809875741087,
+      "grad_norm": 8.360644599601576,
+      "learning_rate": 2.6818568303114967e-07,
+      "loss": 0.422,
+      "step": 10536
+    },
+    {
+      "epoch": 0.8557622025501502,
+      "grad_norm": 11.386127329062182,
+      "learning_rate": 2.67889435393362e-07,
+      "loss": 0.3573,
+      "step": 10537
+    },
+    {
+      "epoch": 0.8558434175261919,
+      "grad_norm": 5.2062587599165475,
+      "learning_rate": 2.6759334220581273e-07,
+      "loss": 0.6315,
+      "step": 10538
+    },
+    {
+      "epoch": 0.8559246325022334,
+      "grad_norm": 4.445713015076414,
+      "learning_rate": 2.6729740348898886e-07,
+      "loss": 0.4583,
+      "step": 10539
+    },
+    {
+      "epoch": 0.856005847478275,
+      "grad_norm": 5.948969244528773,
+      "learning_rate": 2.670016192633687e-07,
+      "loss": 0.4175,
+      "step": 10540
+    },
+    {
+      "epoch": 0.8560870624543165,
+      "grad_norm": 7.049516828972635,
+      "learning_rate": 2.667059895494184e-07,
+      "loss": 0.4651,
+      "step": 10541
+    },
+    {
+      "epoch": 0.8561682774303582,
+      "grad_norm": 12.341170101279898,
+      "learning_rate": 2.6641051436759353e-07,
+      "loss": 0.397,
+      "step": 10542
+    },
+    {
+      "epoch": 0.8562494924063997,
+      "grad_norm": 4.024117199833696,
+      "learning_rate": 2.6611519373834076e-07,
+      "loss": 0.5794,
+      "step": 10543
+    },
+    {
+      "epoch": 0.8563307073824413,
+      "grad_norm": 9.173352219765476,
+      "learning_rate": 2.6582002768209326e-07,
+      "loss": 0.4269,
+      "step": 10544
+    },
+    {
+      "epoch": 0.856411922358483,
+      "grad_norm": 6.424827613404485,
+      "learning_rate": 2.6552501621927544e-07,
+      "loss": 0.5124,
+      "step": 10545
+    },
+    {
+      "epoch": 0.8564931373345245,
+      "grad_norm": 4.601328438451911,
+      "learning_rate": 2.6523015937030136e-07,
+      "loss": 0.4117,
+      "step": 10546
+    },
+    {
+      "epoch": 0.8565743523105661,
+      "grad_norm": 5.439497421746022,
+      "learning_rate": 2.649354571555729e-07,
+      "loss": 0.3709,
+      "step": 10547
+    },
+    {
+      "epoch": 0.8566555672866076,
+      "grad_norm": 5.6866827995608755,
+      "learning_rate": 2.6464090959548135e-07,
+      "loss": 0.4981,
+      "step": 10548
+    },
+    {
+      "epoch": 0.8567367822626493,
+      "grad_norm": 8.576641000068408,
+      "learning_rate": 2.6434651671040894e-07,
+      "loss": 0.4026,
+      "step": 10549
+    },
+    {
+      "epoch": 0.8568179972386908,
+      "grad_norm": 4.32299331991155,
+      "learning_rate": 2.6405227852072504e-07,
+      "loss": 0.5568,
+      "step": 10550
+    },
+    {
+      "epoch": 0.8568992122147324,
+      "grad_norm": 6.515668343832238,
+      "learning_rate": 2.637581950467896e-07,
+      "loss": 0.7008,
+      "step": 10551
+    },
+    {
+      "epoch": 0.8569804271907739,
+      "grad_norm": 4.115866997613363,
+      "learning_rate": 2.634642663089529e-07,
+      "loss": 0.5308,
+      "step": 10552
+    },
+    {
+      "epoch": 0.8570616421668156,
+      "grad_norm": 5.714701058196075,
+      "learning_rate": 2.6317049232755185e-07,
+      "loss": 0.4151,
+      "step": 10553
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 6.258820706442769,
+      "learning_rate": 2.628768731229142e-07,
+      "loss": 0.3542,
+      "step": 10554
+    },
+    {
+      "epoch": 0.8572240721188987,
+      "grad_norm": 7.447014078813493,
+      "learning_rate": 2.6258340871535753e-07,
+      "loss": 0.3555,
+      "step": 10555
+    },
+    {
+      "epoch": 0.8573052870949404,
+      "grad_norm": 4.687623344122876,
+      "learning_rate": 2.6229009912518754e-07,
+      "loss": 0.3792,
+      "step": 10556
+    },
+    {
+      "epoch": 0.8573865020709819,
+      "grad_norm": 4.765486433369644,
+      "learning_rate": 2.619969443726994e-07,
+      "loss": 0.5262,
+      "step": 10557
+    },
+    {
+      "epoch": 0.8574677170470235,
+      "grad_norm": 4.429572901564648,
+      "learning_rate": 2.6170394447817824e-07,
+      "loss": 0.3549,
+      "step": 10558
+    },
+    {
+      "epoch": 0.857548932023065,
+      "grad_norm": 5.054699713881342,
+      "learning_rate": 2.6141109946189874e-07,
+      "loss": 0.6322,
+      "step": 10559
+    },
+    {
+      "epoch": 0.8576301469991067,
+      "grad_norm": 5.352387291113756,
+      "learning_rate": 2.611184093441232e-07,
+      "loss": 0.4139,
+      "step": 10560
+    },
+    {
+      "epoch": 0.8577113619751482,
+      "grad_norm": 7.045189522256243,
+      "learning_rate": 2.608258741451045e-07,
+      "loss": 0.6806,
+      "step": 10561
+    },
+    {
+      "epoch": 0.8577925769511898,
+      "grad_norm": 4.4588250538525935,
+      "learning_rate": 2.605334938850851e-07,
+      "loss": 0.5799,
+      "step": 10562
+    },
+    {
+      "epoch": 0.8578737919272313,
+      "grad_norm": 4.741068158876179,
+      "learning_rate": 2.6024126858429503e-07,
+      "loss": 0.4641,
+      "step": 10563
+    },
+    {
+      "epoch": 0.857955006903273,
+      "grad_norm": 6.387856164555808,
+      "learning_rate": 2.599491982629554e-07,
+      "loss": 0.4024,
+      "step": 10564
+    },
+    {
+      "epoch": 0.8580362218793145,
+      "grad_norm": 11.111320718396847,
+      "learning_rate": 2.596572829412766e-07,
+      "loss": 0.5682,
+      "step": 10565
+    },
+    {
+      "epoch": 0.8581174368553561,
+      "grad_norm": 3.9432874147413366,
+      "learning_rate": 2.59365522639457e-07,
+      "loss": 0.4614,
+      "step": 10566
+    },
+    {
+      "epoch": 0.8581986518313978,
+      "grad_norm": 4.644099435339523,
+      "learning_rate": 2.590739173776841e-07,
+      "loss": 0.5058,
+      "step": 10567
+    },
+    {
+      "epoch": 0.8582798668074393,
+      "grad_norm": 6.2728581984877865,
+      "learning_rate": 2.5878246717613684e-07,
+      "loss": 0.3661,
+      "step": 10568
+    },
+    {
+      "epoch": 0.8583610817834809,
+      "grad_norm": 4.322705528890997,
+      "learning_rate": 2.5849117205498096e-07,
+      "loss": 0.6458,
+      "step": 10569
+    },
+    {
+      "epoch": 0.8584422967595224,
+      "grad_norm": 3.749986780079194,
+      "learning_rate": 2.582000320343728e-07,
+      "loss": 0.6983,
+      "step": 10570
+    },
+    {
+      "epoch": 0.8585235117355641,
+      "grad_norm": 7.654907749368039,
+      "learning_rate": 2.579090471344584e-07,
+      "loss": 0.4619,
+      "step": 10571
+    },
+    {
+      "epoch": 0.8586047267116056,
+      "grad_norm": 7.384217550856351,
+      "learning_rate": 2.576182173753719e-07,
+      "loss": 0.4562,
+      "step": 10572
+    },
+    {
+      "epoch": 0.8586859416876472,
+      "grad_norm": 6.675039979352513,
+      "learning_rate": 2.5732754277723703e-07,
+      "loss": 0.5025,
+      "step": 10573
+    },
+    {
+      "epoch": 0.8587671566636887,
+      "grad_norm": 4.025631310651602,
+      "learning_rate": 2.5703702336016654e-07,
+      "loss": 0.4889,
+      "step": 10574
+    },
+    {
+      "epoch": 0.8588483716397304,
+      "grad_norm": 4.195667201107953,
+      "learning_rate": 2.567466591442638e-07,
+      "loss": 0.3889,
+      "step": 10575
+    },
+    {
+      "epoch": 0.8589295866157719,
+      "grad_norm": 5.444485183310965,
+      "learning_rate": 2.5645645014961947e-07,
+      "loss": 0.4681,
+      "step": 10576
+    },
+    {
+      "epoch": 0.8590108015918135,
+      "grad_norm": 4.670632142056175,
+      "learning_rate": 2.561663963963151e-07,
+      "loss": 0.558,
+      "step": 10577
+    },
+    {
+      "epoch": 0.8590920165678552,
+      "grad_norm": 5.0298060033672645,
+      "learning_rate": 2.558764979044212e-07,
+      "loss": 0.5758,
+      "step": 10578
+    },
+    {
+      "epoch": 0.8591732315438967,
+      "grad_norm": 3.3846837426199845,
+      "learning_rate": 2.555867546939969e-07,
+      "loss": 0.4184,
+      "step": 10579
+    },
+    {
+      "epoch": 0.8592544465199383,
+      "grad_norm": 3.2790619896944957,
+      "learning_rate": 2.5529716678509007e-07,
+      "loss": 0.3676,
+      "step": 10580
+    },
+    {
+      "epoch": 0.8593356614959798,
+      "grad_norm": 8.215528066018926,
+      "learning_rate": 2.5500773419774e-07,
+      "loss": 0.5089,
+      "step": 10581
+    },
+    {
+      "epoch": 0.8594168764720215,
+      "grad_norm": 7.985001045047634,
+      "learning_rate": 2.547184569519728e-07,
+      "loss": 0.5952,
+      "step": 10582
+    },
+    {
+      "epoch": 0.859498091448063,
+      "grad_norm": 9.533391213718112,
+      "learning_rate": 2.5442933506780536e-07,
+      "loss": 0.4201,
+      "step": 10583
+    },
+    {
+      "epoch": 0.8595793064241046,
+      "grad_norm": 3.854806058643972,
+      "learning_rate": 2.541403685652438e-07,
+      "loss": 0.5651,
+      "step": 10584
+    },
+    {
+      "epoch": 0.8596605214001461,
+      "grad_norm": 10.475802887731723,
+      "learning_rate": 2.53851557464283e-07,
+      "loss": 0.4844,
+      "step": 10585
+    },
+    {
+      "epoch": 0.8597417363761878,
+      "grad_norm": 6.07171775635679,
+      "learning_rate": 2.535629017849062e-07,
+      "loss": 0.4356,
+      "step": 10586
+    },
+    {
+      "epoch": 0.8598229513522293,
+      "grad_norm": 9.000523149669734,
+      "learning_rate": 2.532744015470878e-07,
+      "loss": 0.3976,
+      "step": 10587
+    },
+    {
+      "epoch": 0.8599041663282709,
+      "grad_norm": 3.9919734198421164,
+      "learning_rate": 2.529860567707904e-07,
+      "loss": 0.4646,
+      "step": 10588
+    },
+    {
+      "epoch": 0.8599853813043126,
+      "grad_norm": 5.113963141533617,
+      "learning_rate": 2.5269786747596504e-07,
+      "loss": 0.5287,
+      "step": 10589
+    },
+    {
+      "epoch": 0.8600665962803541,
+      "grad_norm": 6.0753816648158345,
+      "learning_rate": 2.5240983368255365e-07,
+      "loss": 0.3916,
+      "step": 10590
+    },
+    {
+      "epoch": 0.8601478112563957,
+      "grad_norm": 5.777553433923641,
+      "learning_rate": 2.52121955410487e-07,
+      "loss": 0.7061,
+      "step": 10591
+    },
+    {
+      "epoch": 0.8602290262324372,
+      "grad_norm": 6.304746787186769,
+      "learning_rate": 2.518342326796844e-07,
+      "loss": 0.4157,
+      "step": 10592
+    },
+    {
+      "epoch": 0.8603102412084789,
+      "grad_norm": 8.184175586338531,
+      "learning_rate": 2.515466655100543e-07,
+      "loss": 0.4792,
+      "step": 10593
+    },
+    {
+      "epoch": 0.8603914561845204,
+      "grad_norm": 7.804180717198223,
+      "learning_rate": 2.5125925392149533e-07,
+      "loss": 0.4817,
+      "step": 10594
+    },
+    {
+      "epoch": 0.860472671160562,
+      "grad_norm": 7.390682075003671,
+      "learning_rate": 2.5097199793389456e-07,
+      "loss": 0.3667,
+      "step": 10595
+    },
+    {
+      "epoch": 0.8605538861366036,
+      "grad_norm": 4.844008089075363,
+      "learning_rate": 2.506848975671283e-07,
+      "loss": 0.4751,
+      "step": 10596
+    },
+    {
+      "epoch": 0.8606351011126452,
+      "grad_norm": 15.0313162783177,
+      "learning_rate": 2.5039795284106354e-07,
+      "loss": 0.47,
+      "step": 10597
+    },
+    {
+      "epoch": 0.8607163160886867,
+      "grad_norm": 7.827029414244254,
+      "learning_rate": 2.5011116377555463e-07,
+      "loss": 0.3705,
+      "step": 10598
+    },
+    {
+      "epoch": 0.8607975310647283,
+      "grad_norm": 4.165034621143684,
+      "learning_rate": 2.4982453039044536e-07,
+      "loss": 0.5141,
+      "step": 10599
+    },
+    {
+      "epoch": 0.86087874604077,
+      "grad_norm": 9.403450320448428,
+      "learning_rate": 2.495380527055699e-07,
+      "loss": 0.5018,
+      "step": 10600
+    },
+    {
+      "epoch": 0.8609599610168115,
+      "grad_norm": 3.669900811182216,
+      "learning_rate": 2.49251730740751e-07,
+      "loss": 0.5024,
+      "step": 10601
+    },
+    {
+      "epoch": 0.8610411759928531,
+      "grad_norm": 8.666914052435182,
+      "learning_rate": 2.4896556451579985e-07,
+      "loss": 0.3806,
+      "step": 10602
+    },
+    {
+      "epoch": 0.8611223909688946,
+      "grad_norm": 6.5382282595398955,
+      "learning_rate": 2.4867955405051826e-07,
+      "loss": 0.436,
+      "step": 10603
+    },
+    {
+      "epoch": 0.8612036059449363,
+      "grad_norm": 6.652949951974725,
+      "learning_rate": 2.483936993646971e-07,
+      "loss": 0.4528,
+      "step": 10604
+    },
+    {
+      "epoch": 0.8612848209209778,
+      "grad_norm": 7.183747958159308,
+      "learning_rate": 2.48108000478115e-07,
+      "loss": 0.4787,
+      "step": 10605
+    },
+    {
+      "epoch": 0.8613660358970194,
+      "grad_norm": 10.226892509366555,
+      "learning_rate": 2.4782245741054175e-07,
+      "loss": 0.5123,
+      "step": 10606
+    },
+    {
+      "epoch": 0.861447250873061,
+      "grad_norm": 6.451808047428951,
+      "learning_rate": 2.475370701817348e-07,
+      "loss": 0.3932,
+      "step": 10607
+    },
+    {
+      "epoch": 0.8615284658491026,
+      "grad_norm": 4.898759701665131,
+      "learning_rate": 2.4725183881144114e-07,
+      "loss": 0.5491,
+      "step": 10608
+    },
+    {
+      "epoch": 0.8616096808251441,
+      "grad_norm": 31.142791051653955,
+      "learning_rate": 2.4696676331939786e-07,
+      "loss": 0.4651,
+      "step": 10609
+    },
+    {
+      "epoch": 0.8616908958011857,
+      "grad_norm": 3.5669616787142697,
+      "learning_rate": 2.46681843725331e-07,
+      "loss": 0.5049,
+      "step": 10610
+    },
+    {
+      "epoch": 0.8617721107772274,
+      "grad_norm": 5.205643471968962,
+      "learning_rate": 2.4639708004895515e-07,
+      "loss": 0.4391,
+      "step": 10611
+    },
+    {
+      "epoch": 0.8618533257532689,
+      "grad_norm": 5.483632769288808,
+      "learning_rate": 2.4611247230997366e-07,
+      "loss": 0.5294,
+      "step": 10612
+    },
+    {
+      "epoch": 0.8619345407293105,
+      "grad_norm": 7.848858229369638,
+      "learning_rate": 2.458280205280811e-07,
+      "loss": 0.4576,
+      "step": 10613
+    },
+    {
+      "epoch": 0.862015755705352,
+      "grad_norm": 4.862179433393447,
+      "learning_rate": 2.455437247229595e-07,
+      "loss": 0.5552,
+      "step": 10614
+    },
+    {
+      "epoch": 0.8620969706813937,
+      "grad_norm": 7.004075796989462,
+      "learning_rate": 2.4525958491428026e-07,
+      "loss": 0.573,
+      "step": 10615
+    },
+    {
+      "epoch": 0.8621781856574352,
+      "grad_norm": 27.404691438629225,
+      "learning_rate": 2.4497560112170444e-07,
+      "loss": 0.5041,
+      "step": 10616
+    },
+    {
+      "epoch": 0.8622594006334768,
+      "grad_norm": 3.9710080611622955,
+      "learning_rate": 2.446917733648829e-07,
+      "loss": 0.6183,
+      "step": 10617
+    },
+    {
+      "epoch": 0.8623406156095184,
+      "grad_norm": 7.736011644339195,
+      "learning_rate": 2.444081016634545e-07,
+      "loss": 0.6269,
+      "step": 10618
+    },
+    {
+      "epoch": 0.86242183058556,
+      "grad_norm": 3.296357608833365,
+      "learning_rate": 2.4412458603704806e-07,
+      "loss": 0.6128,
+      "step": 10619
+    },
+    {
+      "epoch": 0.8625030455616015,
+      "grad_norm": 4.26650357660672,
+      "learning_rate": 2.438412265052814e-07,
+      "loss": 0.4254,
+      "step": 10620
+    },
+    {
+      "epoch": 0.8625842605376431,
+      "grad_norm": 7.332608647465775,
+      "learning_rate": 2.4355802308776073e-07,
+      "loss": 0.4871,
+      "step": 10621
+    },
+    {
+      "epoch": 0.8626654755136848,
+      "grad_norm": 4.755644735563533,
+      "learning_rate": 2.4327497580408285e-07,
+      "loss": 0.5566,
+      "step": 10622
+    },
+    {
+      "epoch": 0.8627466904897263,
+      "grad_norm": 3.8321180175370824,
+      "learning_rate": 2.4299208467383347e-07,
+      "loss": 0.4417,
+      "step": 10623
+    },
+    {
+      "epoch": 0.8628279054657679,
+      "grad_norm": 4.133812851532381,
+      "learning_rate": 2.427093497165864e-07,
+      "loss": 0.4393,
+      "step": 10624
+    },
+    {
+      "epoch": 0.8629091204418095,
+      "grad_norm": 4.292765424503046,
+      "learning_rate": 2.4242677095190623e-07,
+      "loss": 0.5653,
+      "step": 10625
+    },
+    {
+      "epoch": 0.8629903354178511,
+      "grad_norm": 5.735964641093858,
+      "learning_rate": 2.4214434839934545e-07,
+      "loss": 0.3592,
+      "step": 10626
+    },
+    {
+      "epoch": 0.8630715503938926,
+      "grad_norm": 8.285033867612887,
+      "learning_rate": 2.418620820784462e-07,
+      "loss": 0.5062,
+      "step": 10627
+    },
+    {
+      "epoch": 0.8631527653699342,
+      "grad_norm": 4.512095996980739,
+      "learning_rate": 2.4157997200873945e-07,
+      "loss": 0.3959,
+      "step": 10628
+    },
+    {
+      "epoch": 0.8632339803459758,
+      "grad_norm": 5.106558948145591,
+      "learning_rate": 2.4129801820974604e-07,
+      "loss": 0.4257,
+      "step": 10629
+    },
+    {
+      "epoch": 0.8633151953220174,
+      "grad_norm": 8.542776714887058,
+      "learning_rate": 2.410162207009761e-07,
+      "loss": 0.4211,
+      "step": 10630
+    },
+    {
+      "epoch": 0.8633964102980589,
+      "grad_norm": 4.654592367436493,
+      "learning_rate": 2.4073457950192806e-07,
+      "loss": 0.4189,
+      "step": 10631
+    },
+    {
+      "epoch": 0.8634776252741005,
+      "grad_norm": 6.844838480590244,
+      "learning_rate": 2.404530946320904e-07,
+      "loss": 0.5268,
+      "step": 10632
+    },
+    {
+      "epoch": 0.8635588402501422,
+      "grad_norm": 5.39856003020653,
+      "learning_rate": 2.401717661109401e-07,
+      "loss": 0.5319,
+      "step": 10633
+    },
+    {
+      "epoch": 0.8636400552261837,
+      "grad_norm": 6.044733492723436,
+      "learning_rate": 2.398905939579432e-07,
+      "loss": 0.4763,
+      "step": 10634
+    },
+    {
+      "epoch": 0.8637212702022253,
+      "grad_norm": 3.216826620946339,
+      "learning_rate": 2.396095781925556e-07,
+      "loss": 0.4624,
+      "step": 10635
+    },
+    {
+      "epoch": 0.8638024851782669,
+      "grad_norm": 4.322010048301707,
+      "learning_rate": 2.3932871883422276e-07,
+      "loss": 0.3858,
+      "step": 10636
+    },
+    {
+      "epoch": 0.8638837001543085,
+      "grad_norm": 6.581640577707594,
+      "learning_rate": 2.3904801590237783e-07,
+      "loss": 0.3858,
+      "step": 10637
+    },
+    {
+      "epoch": 0.86396491513035,
+      "grad_norm": 5.430902885102804,
+      "learning_rate": 2.3876746941644464e-07,
+      "loss": 0.547,
+      "step": 10638
+    },
+    {
+      "epoch": 0.8640461301063916,
+      "grad_norm": 9.539650370477867,
+      "learning_rate": 2.384870793958349e-07,
+      "loss": 0.4549,
+      "step": 10639
+    },
+    {
+      "epoch": 0.8641273450824332,
+      "grad_norm": 6.254836784455941,
+      "learning_rate": 2.3820684585995012e-07,
+      "loss": 0.5427,
+      "step": 10640
+    },
+    {
+      "epoch": 0.8642085600584748,
+      "grad_norm": 7.936489779390335,
+      "learning_rate": 2.379267688281814e-07,
+      "loss": 0.394,
+      "step": 10641
+    },
+    {
+      "epoch": 0.8642897750345163,
+      "grad_norm": 4.9403538410304,
+      "learning_rate": 2.3764684831990874e-07,
+      "loss": 0.5883,
+      "step": 10642
+    },
+    {
+      "epoch": 0.864370990010558,
+      "grad_norm": 5.718951894452898,
+      "learning_rate": 2.3736708435450033e-07,
+      "loss": 0.4907,
+      "step": 10643
+    },
+    {
+      "epoch": 0.8644522049865996,
+      "grad_norm": 4.730922288439977,
+      "learning_rate": 2.370874769513154e-07,
+      "loss": 0.4677,
+      "step": 10644
+    },
+    {
+      "epoch": 0.8645334199626411,
+      "grad_norm": 7.22516165081707,
+      "learning_rate": 2.3680802612970068e-07,
+      "loss": 0.4901,
+      "step": 10645
+    },
+    {
+      "epoch": 0.8646146349386827,
+      "grad_norm": 5.535751530022896,
+      "learning_rate": 2.365287319089929e-07,
+      "loss": 0.4482,
+      "step": 10646
+    },
+    {
+      "epoch": 0.8646958499147243,
+      "grad_norm": 4.330430490684841,
+      "learning_rate": 2.362495943085172e-07,
+      "loss": 0.3961,
+      "step": 10647
+    },
+    {
+      "epoch": 0.8647770648907659,
+      "grad_norm": 5.262984242628663,
+      "learning_rate": 2.3597061334758864e-07,
+      "loss": 0.4269,
+      "step": 10648
+    },
+    {
+      "epoch": 0.8648582798668074,
+      "grad_norm": 3.5730441803712014,
+      "learning_rate": 2.3569178904551181e-07,
+      "loss": 0.4865,
+      "step": 10649
+    },
+    {
+      "epoch": 0.864939494842849,
+      "grad_norm": 4.35982646863664,
+      "learning_rate": 2.3541312142157934e-07,
+      "loss": 0.62,
+      "step": 10650
+    },
+    {
+      "epoch": 0.8650207098188906,
+      "grad_norm": 5.501060251786152,
+      "learning_rate": 2.3513461049507385e-07,
+      "loss": 0.309,
+      "step": 10651
+    },
+    {
+      "epoch": 0.8651019247949322,
+      "grad_norm": 18.5509951879782,
+      "learning_rate": 2.3485625628526688e-07,
+      "loss": 0.3867,
+      "step": 10652
+    },
+    {
+      "epoch": 0.8651831397709737,
+      "grad_norm": 4.778384806825011,
+      "learning_rate": 2.3457805881141854e-07,
+      "loss": 0.4658,
+      "step": 10653
+    },
+    {
+      "epoch": 0.8652643547470154,
+      "grad_norm": 7.32649293368322,
+      "learning_rate": 2.3430001809277873e-07,
+      "loss": 0.483,
+      "step": 10654
+    },
+    {
+      "epoch": 0.865345569723057,
+      "grad_norm": 6.199897652234001,
+      "learning_rate": 2.340221341485871e-07,
+      "loss": 0.435,
+      "step": 10655
+    },
+    {
+      "epoch": 0.8654267846990985,
+      "grad_norm": 4.754389895318495,
+      "learning_rate": 2.3374440699807072e-07,
+      "loss": 0.5084,
+      "step": 10656
+    },
+    {
+      "epoch": 0.8655079996751401,
+      "grad_norm": 4.649740673308111,
+      "learning_rate": 2.334668366604481e-07,
+      "loss": 0.5881,
+      "step": 10657
+    },
+    {
+      "epoch": 0.8655892146511817,
+      "grad_norm": 6.638508068101479,
+      "learning_rate": 2.3318942315492477e-07,
+      "loss": 0.38,
+      "step": 10658
+    },
+    {
+      "epoch": 0.8656704296272233,
+      "grad_norm": 3.952860937759617,
+      "learning_rate": 2.3291216650069587e-07,
+      "loss": 0.4626,
+      "step": 10659
+    },
+    {
+      "epoch": 0.8657516446032648,
+      "grad_norm": 6.6956489482132,
+      "learning_rate": 2.3263506671694747e-07,
+      "loss": 0.4982,
+      "step": 10660
+    },
+    {
+      "epoch": 0.8658328595793064,
+      "grad_norm": 3.478866841645943,
+      "learning_rate": 2.323581238228517e-07,
+      "loss": 0.4391,
+      "step": 10661
+    },
+    {
+      "epoch": 0.865914074555348,
+      "grad_norm": 4.525751483245147,
+      "learning_rate": 2.3208133783757302e-07,
+      "loss": 0.5142,
+      "step": 10662
+    },
+    {
+      "epoch": 0.8659952895313896,
+      "grad_norm": 4.621330522104981,
+      "learning_rate": 2.3180470878026275e-07,
+      "loss": 0.5449,
+      "step": 10663
+    },
+    {
+      "epoch": 0.8660765045074311,
+      "grad_norm": 4.973654556392699,
+      "learning_rate": 2.3152823667006248e-07,
+      "loss": 0.4468,
+      "step": 10664
+    },
+    {
+      "epoch": 0.8661577194834728,
+      "grad_norm": 5.396466731566874,
+      "learning_rate": 2.3125192152610277e-07,
+      "loss": 0.633,
+      "step": 10665
+    },
+    {
+      "epoch": 0.8662389344595144,
+      "grad_norm": 4.918829274592111,
+      "learning_rate": 2.3097576336750248e-07,
+      "loss": 0.4153,
+      "step": 10666
+    },
+    {
+      "epoch": 0.8663201494355559,
+      "grad_norm": 6.0092199166872,
+      "learning_rate": 2.3069976221337054e-07,
+      "loss": 0.7968,
+      "step": 10667
+    },
+    {
+      "epoch": 0.8664013644115975,
+      "grad_norm": 13.224667725694793,
+      "learning_rate": 2.304239180828055e-07,
+      "loss": 0.6541,
+      "step": 10668
+    },
+    {
+      "epoch": 0.8664825793876391,
+      "grad_norm": 4.332528390591684,
+      "learning_rate": 2.3014823099489326e-07,
+      "loss": 0.4286,
+      "step": 10669
+    },
+    {
+      "epoch": 0.8665637943636807,
+      "grad_norm": 8.678328469216439,
+      "learning_rate": 2.2987270096871072e-07,
+      "loss": 0.457,
+      "step": 10670
+    },
+    {
+      "epoch": 0.8666450093397222,
+      "grad_norm": 6.543536578014087,
+      "learning_rate": 2.2959732802332296e-07,
+      "loss": 0.3446,
+      "step": 10671
+    },
+    {
+      "epoch": 0.8667262243157638,
+      "grad_norm": 5.10238287168412,
+      "learning_rate": 2.2932211217778388e-07,
+      "loss": 0.6514,
+      "step": 10672
+    },
+    {
+      "epoch": 0.8668074392918054,
+      "grad_norm": 4.902163867519704,
+      "learning_rate": 2.2904705345113743e-07,
+      "loss": 0.6377,
+      "step": 10673
+    },
+    {
+      "epoch": 0.866888654267847,
+      "grad_norm": 4.914208356693128,
+      "learning_rate": 2.287721518624156e-07,
+      "loss": 0.5384,
+      "step": 10674
+    },
+    {
+      "epoch": 0.8669698692438885,
+      "grad_norm": 4.695162916684552,
+      "learning_rate": 2.2849740743064063e-07,
+      "loss": 0.3763,
+      "step": 10675
+    },
+    {
+      "epoch": 0.8670510842199302,
+      "grad_norm": 5.516447784018673,
+      "learning_rate": 2.282228201748238e-07,
+      "loss": 0.4784,
+      "step": 10676
+    },
+    {
+      "epoch": 0.8671322991959718,
+      "grad_norm": 4.784027348664785,
+      "learning_rate": 2.2794839011396453e-07,
+      "loss": 0.3715,
+      "step": 10677
+    },
+    {
+      "epoch": 0.8672135141720133,
+      "grad_norm": 4.524507243360301,
+      "learning_rate": 2.2767411726705157e-07,
+      "loss": 0.66,
+      "step": 10678
+    },
+    {
+      "epoch": 0.8672947291480549,
+      "grad_norm": 6.2607763304203194,
+      "learning_rate": 2.2740000165306393e-07,
+      "loss": 0.4065,
+      "step": 10679
+    },
+    {
+      "epoch": 0.8673759441240965,
+      "grad_norm": 6.045389943291753,
+      "learning_rate": 2.2712604329096833e-07,
+      "loss": 0.5403,
+      "step": 10680
+    },
+    {
+      "epoch": 0.8674571591001381,
+      "grad_norm": 4.945546166899713,
+      "learning_rate": 2.2685224219972185e-07,
+      "loss": 0.6084,
+      "step": 10681
+    },
+    {
+      "epoch": 0.8675383740761796,
+      "grad_norm": 7.020025507043163,
+      "learning_rate": 2.2657859839826934e-07,
+      "loss": 0.4117,
+      "step": 10682
+    },
+    {
+      "epoch": 0.8676195890522213,
+      "grad_norm": 3.378493580413376,
+      "learning_rate": 2.2630511190554621e-07,
+      "loss": 0.3219,
+      "step": 10683
+    },
+    {
+      "epoch": 0.8677008040282628,
+      "grad_norm": 14.896543415086516,
+      "learning_rate": 2.260317827404762e-07,
+      "loss": 0.3889,
+      "step": 10684
+    },
+    {
+      "epoch": 0.8677820190043044,
+      "grad_norm": 4.531796982257481,
+      "learning_rate": 2.2575861092197143e-07,
+      "loss": 0.6512,
+      "step": 10685
+    },
+    {
+      "epoch": 0.8678632339803459,
+      "grad_norm": 7.006910779412939,
+      "learning_rate": 2.254855964689351e-07,
+      "loss": 0.5622,
+      "step": 10686
+    },
+    {
+      "epoch": 0.8679444489563876,
+      "grad_norm": 5.778562835427389,
+      "learning_rate": 2.2521273940025705e-07,
+      "loss": 0.5571,
+      "step": 10687
+    },
+    {
+      "epoch": 0.8680256639324292,
+      "grad_norm": 5.550998523968033,
+      "learning_rate": 2.2494003973481864e-07,
+      "loss": 0.481,
+      "step": 10688
+    },
+    {
+      "epoch": 0.8681068789084707,
+      "grad_norm": 5.8427464651275836,
+      "learning_rate": 2.2466749749148919e-07,
+      "loss": 0.4879,
+      "step": 10689
+    },
+    {
+      "epoch": 0.8681880938845123,
+      "grad_norm": 4.136202285236531,
+      "learning_rate": 2.2439511268912666e-07,
+      "loss": 0.3003,
+      "step": 10690
+    },
+    {
+      "epoch": 0.8682693088605539,
+      "grad_norm": 7.410829723701502,
+      "learning_rate": 2.2412288534657878e-07,
+      "loss": 0.556,
+      "step": 10691
+    },
+    {
+      "epoch": 0.8683505238365955,
+      "grad_norm": 5.668425487445851,
+      "learning_rate": 2.2385081548268268e-07,
+      "loss": 0.5302,
+      "step": 10692
+    },
+    {
+      "epoch": 0.868431738812637,
+      "grad_norm": 5.0755641679869505,
+      "learning_rate": 2.2357890311626328e-07,
+      "loss": 0.4406,
+      "step": 10693
+    },
+    {
+      "epoch": 0.8685129537886787,
+      "grad_norm": 4.6801666968241795,
+      "learning_rate": 2.2330714826613586e-07,
+      "loss": 0.3976,
+      "step": 10694
+    },
+    {
+      "epoch": 0.8685941687647202,
+      "grad_norm": 5.777100294603406,
+      "learning_rate": 2.2303555095110507e-07,
+      "loss": 0.4059,
+      "step": 10695
+    },
+    {
+      "epoch": 0.8686753837407618,
+      "grad_norm": 5.314872055548993,
+      "learning_rate": 2.2276411118996366e-07,
+      "loss": 0.4719,
+      "step": 10696
+    },
+    {
+      "epoch": 0.8687565987168033,
+      "grad_norm": 4.410426094256012,
+      "learning_rate": 2.22492829001493e-07,
+      "loss": 0.4898,
+      "step": 10697
+    },
+    {
+      "epoch": 0.868837813692845,
+      "grad_norm": 4.7733912494094755,
+      "learning_rate": 2.2222170440446557e-07,
+      "loss": 0.4248,
+      "step": 10698
+    },
+    {
+      "epoch": 0.8689190286688866,
+      "grad_norm": 6.898790830573995,
+      "learning_rate": 2.219507374176408e-07,
+      "loss": 0.4269,
+      "step": 10699
+    },
+    {
+      "epoch": 0.8690002436449281,
+      "grad_norm": 5.2231893468594635,
+      "learning_rate": 2.2167992805976896e-07,
+      "loss": 0.5232,
+      "step": 10700
+    },
+    {
+      "epoch": 0.8690814586209697,
+      "grad_norm": 7.520502969621975,
+      "learning_rate": 2.2140927634958788e-07,
+      "loss": 0.3991,
+      "step": 10701
+    },
+    {
+      "epoch": 0.8691626735970113,
+      "grad_norm": 4.652004187169794,
+      "learning_rate": 2.2113878230582615e-07,
+      "loss": 0.5095,
+      "step": 10702
+    },
+    {
+      "epoch": 0.8692438885730529,
+      "grad_norm": 3.840996235904483,
+      "learning_rate": 2.2086844594719993e-07,
+      "loss": 0.5209,
+      "step": 10703
+    },
+    {
+      "epoch": 0.8693251035490944,
+      "grad_norm": 5.428257976832266,
+      "learning_rate": 2.205982672924145e-07,
+      "loss": 0.4342,
+      "step": 10704
+    },
+    {
+      "epoch": 0.8694063185251361,
+      "grad_norm": 12.646126359342446,
+      "learning_rate": 2.203282463601661e-07,
+      "loss": 0.5407,
+      "step": 10705
+    },
+    {
+      "epoch": 0.8694875335011776,
+      "grad_norm": 8.151893101320535,
+      "learning_rate": 2.2005838316913746e-07,
+      "loss": 0.3985,
+      "step": 10706
+    },
+    {
+      "epoch": 0.8695687484772192,
+      "grad_norm": 3.6426553923017337,
+      "learning_rate": 2.1978867773800205e-07,
+      "loss": 0.6897,
+      "step": 10707
+    },
+    {
+      "epoch": 0.8696499634532607,
+      "grad_norm": 5.392872165186957,
+      "learning_rate": 2.1951913008542297e-07,
+      "loss": 0.318,
+      "step": 10708
+    },
+    {
+      "epoch": 0.8697311784293024,
+      "grad_norm": 6.652109485767294,
+      "learning_rate": 2.1924974023005086e-07,
+      "loss": 0.396,
+      "step": 10709
+    },
+    {
+      "epoch": 0.869812393405344,
+      "grad_norm": 6.329696488288848,
+      "learning_rate": 2.189805081905255e-07,
+      "loss": 0.3915,
+      "step": 10710
+    },
+    {
+      "epoch": 0.8698936083813855,
+      "grad_norm": 3.238598078513282,
+      "learning_rate": 2.1871143398547735e-07,
+      "loss": 0.5392,
+      "step": 10711
+    },
+    {
+      "epoch": 0.8699748233574272,
+      "grad_norm": 6.202895386014103,
+      "learning_rate": 2.184425176335239e-07,
+      "loss": 0.535,
+      "step": 10712
+    },
+    {
+      "epoch": 0.8700560383334687,
+      "grad_norm": 7.1313443357236075,
+      "learning_rate": 2.1817375915327342e-07,
+      "loss": 0.3804,
+      "step": 10713
+    },
+    {
+      "epoch": 0.8701372533095103,
+      "grad_norm": 4.686543627354921,
+      "learning_rate": 2.1790515856332268e-07,
+      "loss": 0.485,
+      "step": 10714
+    },
+    {
+      "epoch": 0.8702184682855518,
+      "grad_norm": 4.348632743659829,
+      "learning_rate": 2.1763671588225705e-07,
+      "loss": 0.4805,
+      "step": 10715
+    },
+    {
+      "epoch": 0.8702996832615935,
+      "grad_norm": 4.845358239649171,
+      "learning_rate": 2.173684311286517e-07,
+      "loss": 0.4082,
+      "step": 10716
+    },
+    {
+      "epoch": 0.870380898237635,
+      "grad_norm": 5.562307404701293,
+      "learning_rate": 2.1710030432106982e-07,
+      "loss": 0.5136,
+      "step": 10717
+    },
+    {
+      "epoch": 0.8704621132136766,
+      "grad_norm": 15.50221360580663,
+      "learning_rate": 2.1683233547806494e-07,
+      "loss": 0.4594,
+      "step": 10718
+    },
+    {
+      "epoch": 0.8705433281897181,
+      "grad_norm": 8.184235968979692,
+      "learning_rate": 2.1656452461817883e-07,
+      "loss": 0.5134,
+      "step": 10719
+    },
+    {
+      "epoch": 0.8706245431657598,
+      "grad_norm": 5.086338805312825,
+      "learning_rate": 2.162968717599423e-07,
+      "loss": 0.6206,
+      "step": 10720
+    },
+    {
+      "epoch": 0.8707057581418014,
+      "grad_norm": 6.866907411125338,
+      "learning_rate": 2.1602937692187685e-07,
+      "loss": 0.3373,
+      "step": 10721
+    },
+    {
+      "epoch": 0.8707869731178429,
+      "grad_norm": 6.472579080212052,
+      "learning_rate": 2.1576204012249053e-07,
+      "loss": 0.4828,
+      "step": 10722
+    },
+    {
+      "epoch": 0.8708681880938846,
+      "grad_norm": 6.9868588515529915,
+      "learning_rate": 2.1549486138028125e-07,
+      "loss": 0.537,
+      "step": 10723
+    },
+    {
+      "epoch": 0.8709494030699261,
+      "grad_norm": 4.648588890106703,
+      "learning_rate": 2.152278407137376e-07,
+      "loss": 0.4446,
+      "step": 10724
+    },
+    {
+      "epoch": 0.8710306180459677,
+      "grad_norm": 5.2378357215552915,
+      "learning_rate": 2.1496097814133503e-07,
+      "loss": 0.4319,
+      "step": 10725
+    },
+    {
+      "epoch": 0.8711118330220092,
+      "grad_norm": 4.349980838222139,
+      "learning_rate": 2.146942736815391e-07,
+      "loss": 0.3548,
+      "step": 10726
+    },
+    {
+      "epoch": 0.8711930479980509,
+      "grad_norm": 3.1974318652567026,
+      "learning_rate": 2.1442772735280532e-07,
+      "loss": 0.574,
+      "step": 10727
+    },
+    {
+      "epoch": 0.8712742629740924,
+      "grad_norm": 6.948883443858628,
+      "learning_rate": 2.1416133917357668e-07,
+      "loss": 0.5418,
+      "step": 10728
+    },
+    {
+      "epoch": 0.871355477950134,
+      "grad_norm": 7.741555115120562,
+      "learning_rate": 2.1389510916228513e-07,
+      "loss": 0.5892,
+      "step": 10729
+    },
+    {
+      "epoch": 0.8714366929261755,
+      "grad_norm": 4.815397769745156,
+      "learning_rate": 2.136290373373534e-07,
+      "loss": 0.6456,
+      "step": 10730
+    },
+    {
+      "epoch": 0.8715179079022172,
+      "grad_norm": 4.733327642778111,
+      "learning_rate": 2.1336312371719182e-07,
+      "loss": 0.5199,
+      "step": 10731
+    },
+    {
+      "epoch": 0.8715991228782588,
+      "grad_norm": 6.863949439219225,
+      "learning_rate": 2.130973683201998e-07,
+      "loss": 0.489,
+      "step": 10732
+    },
+    {
+      "epoch": 0.8716803378543003,
+      "grad_norm": 6.500756310540661,
+      "learning_rate": 2.128317711647665e-07,
+      "loss": 0.4344,
+      "step": 10733
+    },
+    {
+      "epoch": 0.871761552830342,
+      "grad_norm": 7.69706976013433,
+      "learning_rate": 2.125663322692706e-07,
+      "loss": 0.5088,
+      "step": 10734
+    },
+    {
+      "epoch": 0.8718427678063835,
+      "grad_norm": 6.331349228694531,
+      "learning_rate": 2.1230105165207848e-07,
+      "loss": 0.6559,
+      "step": 10735
+    },
+    {
+      "epoch": 0.8719239827824251,
+      "grad_norm": 5.041214163229692,
+      "learning_rate": 2.120359293315455e-07,
+      "loss": 0.6072,
+      "step": 10736
+    },
+    {
+      "epoch": 0.8720051977584666,
+      "grad_norm": 5.582839372557744,
+      "learning_rate": 2.1177096532601777e-07,
+      "loss": 0.4588,
+      "step": 10737
+    },
+    {
+      "epoch": 0.8720864127345083,
+      "grad_norm": 7.814578266112515,
+      "learning_rate": 2.115061596538287e-07,
+      "loss": 0.4815,
+      "step": 10738
+    },
+    {
+      "epoch": 0.8721676277105498,
+      "grad_norm": 5.583906428631362,
+      "learning_rate": 2.112415123333014e-07,
+      "loss": 0.5404,
+      "step": 10739
+    },
+    {
+      "epoch": 0.8722488426865914,
+      "grad_norm": 7.942670374026383,
+      "learning_rate": 2.1097702338274907e-07,
+      "loss": 0.4854,
+      "step": 10740
+    },
+    {
+      "epoch": 0.8723300576626329,
+      "grad_norm": 4.655699223151669,
+      "learning_rate": 2.1071269282047196e-07,
+      "loss": 0.6385,
+      "step": 10741
+    },
+    {
+      "epoch": 0.8724112726386746,
+      "grad_norm": 5.347652457276179,
+      "learning_rate": 2.1044852066476052e-07,
+      "loss": 0.5986,
+      "step": 10742
+    },
+    {
+      "epoch": 0.8724924876147162,
+      "grad_norm": 6.78105342737064,
+      "learning_rate": 2.1018450693389452e-07,
+      "loss": 0.4435,
+      "step": 10743
+    },
+    {
+      "epoch": 0.8725737025907577,
+      "grad_norm": 4.350453860498677,
+      "learning_rate": 2.099206516461419e-07,
+      "loss": 0.5274,
+      "step": 10744
+    },
+    {
+      "epoch": 0.8726549175667994,
+      "grad_norm": 4.768332070430383,
+      "learning_rate": 2.096569548197594e-07,
+      "loss": 0.551,
+      "step": 10745
+    },
+    {
+      "epoch": 0.8727361325428409,
+      "grad_norm": 6.078372849255497,
+      "learning_rate": 2.0939341647299437e-07,
+      "loss": 0.4384,
+      "step": 10746
+    },
+    {
+      "epoch": 0.8728173475188825,
+      "grad_norm": 11.608495384238454,
+      "learning_rate": 2.0913003662408254e-07,
+      "loss": 0.6075,
+      "step": 10747
+    },
+    {
+      "epoch": 0.872898562494924,
+      "grad_norm": 8.027806223001297,
+      "learning_rate": 2.0886681529124765e-07,
+      "loss": 0.4888,
+      "step": 10748
+    },
+    {
+      "epoch": 0.8729797774709657,
+      "grad_norm": 4.082351604317076,
+      "learning_rate": 2.086037524927037e-07,
+      "loss": 0.5281,
+      "step": 10749
+    },
+    {
+      "epoch": 0.8730609924470072,
+      "grad_norm": 5.82956001643256,
+      "learning_rate": 2.0834084824665314e-07,
+      "loss": 0.5854,
+      "step": 10750
+    },
+    {
+      "epoch": 0.8731422074230488,
+      "grad_norm": 9.953252455414237,
+      "learning_rate": 2.0807810257128692e-07,
+      "loss": 0.4345,
+      "step": 10751
+    },
+    {
+      "epoch": 0.8732234223990903,
+      "grad_norm": 5.897867057523133,
+      "learning_rate": 2.0781551548478607e-07,
+      "loss": 0.5252,
+      "step": 10752
+    },
+    {
+      "epoch": 0.873304637375132,
+      "grad_norm": 5.203890577661351,
+      "learning_rate": 2.0755308700532077e-07,
+      "loss": 0.3069,
+      "step": 10753
+    },
+    {
+      "epoch": 0.8733858523511736,
+      "grad_norm": 9.388507828923037,
+      "learning_rate": 2.0729081715104958e-07,
+      "loss": 0.3058,
+      "step": 10754
+    },
+    {
+      "epoch": 0.8734670673272151,
+      "grad_norm": 4.665994646948376,
+      "learning_rate": 2.070287059401191e-07,
+      "loss": 0.4658,
+      "step": 10755
+    },
+    {
+      "epoch": 0.8735482823032568,
+      "grad_norm": 4.509183704578975,
+      "learning_rate": 2.0676675339066726e-07,
+      "loss": 0.4733,
+      "step": 10756
+    },
+    {
+      "epoch": 0.8736294972792983,
+      "grad_norm": 4.982775473415814,
+      "learning_rate": 2.0650495952081935e-07,
+      "loss": 0.388,
+      "step": 10757
+    },
+    {
+      "epoch": 0.8737107122553399,
+      "grad_norm": 3.915834558563621,
+      "learning_rate": 2.062433243486897e-07,
+      "loss": 0.4357,
+      "step": 10758
+    },
+    {
+      "epoch": 0.8737919272313814,
+      "grad_norm": 5.779273705619936,
+      "learning_rate": 2.059818478923825e-07,
+      "loss": 0.4962,
+      "step": 10759
+    },
+    {
+      "epoch": 0.8738731422074231,
+      "grad_norm": 7.124975275931914,
+      "learning_rate": 2.0572053016999079e-07,
+      "loss": 0.4915,
+      "step": 10760
+    },
+    {
+      "epoch": 0.8739543571834646,
+      "grad_norm": 4.688739212653256,
+      "learning_rate": 2.0545937119959557e-07,
+      "loss": 0.5523,
+      "step": 10761
+    },
+    {
+      "epoch": 0.8740355721595062,
+      "grad_norm": 6.512142297147514,
+      "learning_rate": 2.0519837099926888e-07,
+      "loss": 0.4864,
+      "step": 10762
+    },
+    {
+      "epoch": 0.8741167871355477,
+      "grad_norm": 3.968204310846973,
+      "learning_rate": 2.0493752958706982e-07,
+      "loss": 0.4545,
+      "step": 10763
+    },
+    {
+      "epoch": 0.8741980021115894,
+      "grad_norm": 5.589051586917989,
+      "learning_rate": 2.0467684698104674e-07,
+      "loss": 0.3359,
+      "step": 10764
+    },
+    {
+      "epoch": 0.874279217087631,
+      "grad_norm": 9.371561135244747,
+      "learning_rate": 2.0441632319923798e-07,
+      "loss": 0.3767,
+      "step": 10765
+    },
+    {
+      "epoch": 0.8743604320636725,
+      "grad_norm": 6.212854338465917,
+      "learning_rate": 2.0415595825967084e-07,
+      "loss": 0.5425,
+      "step": 10766
+    },
+    {
+      "epoch": 0.8744416470397142,
+      "grad_norm": 4.1144373279605135,
+      "learning_rate": 2.0389575218036057e-07,
+      "loss": 0.6139,
+      "step": 10767
+    },
+    {
+      "epoch": 0.8745228620157557,
+      "grad_norm": 4.023569708161643,
+      "learning_rate": 2.0363570497931252e-07,
+      "loss": 0.4479,
+      "step": 10768
+    },
+    {
+      "epoch": 0.8746040769917973,
+      "grad_norm": 3.8449587292031686,
+      "learning_rate": 2.0337581667452034e-07,
+      "loss": 0.5775,
+      "step": 10769
+    },
+    {
+      "epoch": 0.8746852919678388,
+      "grad_norm": 6.240516560084869,
+      "learning_rate": 2.0311608728396658e-07,
+      "loss": 0.5565,
+      "step": 10770
+    },
+    {
+      "epoch": 0.8747665069438805,
+      "grad_norm": 7.754381694503205,
+      "learning_rate": 2.0285651682562357e-07,
+      "loss": 0.4689,
+      "step": 10771
+    },
+    {
+      "epoch": 0.874847721919922,
+      "grad_norm": 5.711700208395236,
+      "learning_rate": 2.0259710531745247e-07,
+      "loss": 0.4805,
+      "step": 10772
+    },
+    {
+      "epoch": 0.8749289368959636,
+      "grad_norm": 5.1893458798213326,
+      "learning_rate": 2.023378527774028e-07,
+      "loss": 0.4845,
+      "step": 10773
+    },
+    {
+      "epoch": 0.8750101518720051,
+      "grad_norm": 7.133320338283911,
+      "learning_rate": 2.020787592234133e-07,
+      "loss": 0.4978,
+      "step": 10774
+    },
+    {
+      "epoch": 0.8750913668480468,
+      "grad_norm": 22.519170871083364,
+      "learning_rate": 2.0181982467341238e-07,
+      "loss": 0.4116,
+      "step": 10775
+    },
+    {
+      "epoch": 0.8751725818240884,
+      "grad_norm": 9.287767664431676,
+      "learning_rate": 2.0156104914531656e-07,
+      "loss": 0.4641,
+      "step": 10776
+    },
+    {
+      "epoch": 0.8752537968001299,
+      "grad_norm": 4.893583526059918,
+      "learning_rate": 2.0130243265703148e-07,
+      "loss": 0.481,
+      "step": 10777
+    },
+    {
+      "epoch": 0.8753350117761716,
+      "grad_norm": 5.320206234300975,
+      "learning_rate": 2.010439752264523e-07,
+      "loss": 0.3727,
+      "step": 10778
+    },
+    {
+      "epoch": 0.8754162267522131,
+      "grad_norm": 4.677489372940032,
+      "learning_rate": 2.0078567687146333e-07,
+      "loss": 0.447,
+      "step": 10779
+    },
+    {
+      "epoch": 0.8754974417282547,
+      "grad_norm": 4.0249196656831465,
+      "learning_rate": 2.0052753760993693e-07,
+      "loss": 0.4569,
+      "step": 10780
+    },
+    {
+      "epoch": 0.8755786567042962,
+      "grad_norm": 6.529303245559503,
+      "learning_rate": 2.002695574597352e-07,
+      "loss": 0.3912,
+      "step": 10781
+    },
+    {
+      "epoch": 0.8756598716803379,
+      "grad_norm": 5.418332568962745,
+      "learning_rate": 2.0001173643870915e-07,
+      "loss": 0.5883,
+      "step": 10782
+    },
+    {
+      "epoch": 0.8757410866563794,
+      "grad_norm": 5.33583924657257,
+      "learning_rate": 1.9975407456469808e-07,
+      "loss": 0.4482,
+      "step": 10783
+    },
+    {
+      "epoch": 0.875822301632421,
+      "grad_norm": 5.518754450917286,
+      "learning_rate": 1.9949657185553113e-07,
+      "loss": 0.4846,
+      "step": 10784
+    },
+    {
+      "epoch": 0.8759035166084626,
+      "grad_norm": 4.466728527582831,
+      "learning_rate": 1.992392283290265e-07,
+      "loss": 0.4897,
+      "step": 10785
+    },
+    {
+      "epoch": 0.8759847315845042,
+      "grad_norm": 6.062215529147515,
+      "learning_rate": 1.9898204400299021e-07,
+      "loss": 0.4688,
+      "step": 10786
+    },
+    {
+      "epoch": 0.8760659465605458,
+      "grad_norm": 6.196645398811499,
+      "learning_rate": 1.9872501889521916e-07,
+      "loss": 0.4591,
+      "step": 10787
+    },
+    {
+      "epoch": 0.8761471615365873,
+      "grad_norm": 5.38690541555504,
+      "learning_rate": 1.984681530234972e-07,
+      "loss": 0.3928,
+      "step": 10788
+    },
+    {
+      "epoch": 0.876228376512629,
+      "grad_norm": 9.821582116837323,
+      "learning_rate": 1.9821144640559842e-07,
+      "loss": 0.4919,
+      "step": 10789
+    },
+    {
+      "epoch": 0.8763095914886705,
+      "grad_norm": 7.245646513525235,
+      "learning_rate": 1.9795489905928527e-07,
+      "loss": 0.459,
+      "step": 10790
+    },
+    {
+      "epoch": 0.8763908064647121,
+      "grad_norm": 7.6962316221213385,
+      "learning_rate": 1.976985110023094e-07,
+      "loss": 0.4777,
+      "step": 10791
+    },
+    {
+      "epoch": 0.8764720214407536,
+      "grad_norm": 4.354060604688436,
+      "learning_rate": 1.9744228225241248e-07,
+      "loss": 0.4522,
+      "step": 10792
+    },
+    {
+      "epoch": 0.8765532364167953,
+      "grad_norm": 4.660763544694212,
+      "learning_rate": 1.9718621282732302e-07,
+      "loss": 0.4305,
+      "step": 10793
+    },
+    {
+      "epoch": 0.8766344513928368,
+      "grad_norm": 6.796070237851726,
+      "learning_rate": 1.9693030274476054e-07,
+      "loss": 0.4301,
+      "step": 10794
+    },
+    {
+      "epoch": 0.8767156663688784,
+      "grad_norm": 6.8550929057120635,
+      "learning_rate": 1.9667455202243223e-07,
+      "loss": 0.4308,
+      "step": 10795
+    },
+    {
+      "epoch": 0.87679688134492,
+      "grad_norm": 10.526313953549316,
+      "learning_rate": 1.9641896067803452e-07,
+      "loss": 0.4715,
+      "step": 10796
+    },
+    {
+      "epoch": 0.8768780963209616,
+      "grad_norm": 4.6047250304925615,
+      "learning_rate": 1.9616352872925293e-07,
+      "loss": 0.529,
+      "step": 10797
+    },
+    {
+      "epoch": 0.8769593112970032,
+      "grad_norm": 6.880938367850307,
+      "learning_rate": 1.959082561937628e-07,
+      "loss": 0.5294,
+      "step": 10798
+    },
+    {
+      "epoch": 0.8770405262730447,
+      "grad_norm": 3.8531909306263272,
+      "learning_rate": 1.9565314308922666e-07,
+      "loss": 0.4086,
+      "step": 10799
+    },
+    {
+      "epoch": 0.8771217412490864,
+      "grad_norm": 5.451866169993255,
+      "learning_rate": 1.9539818943329792e-07,
+      "loss": 0.4557,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8772029562251279,
+      "grad_norm": 6.999546249788755,
+      "learning_rate": 1.9514339524361742e-07,
+      "loss": 0.4339,
+      "step": 10801
+    },
+    {
+      "epoch": 0.8772841712011695,
+      "grad_norm": 5.662909641489243,
+      "learning_rate": 1.9488876053781552e-07,
+      "loss": 0.4453,
+      "step": 10802
+    },
+    {
+      "epoch": 0.877365386177211,
+      "grad_norm": 4.898755716797307,
+      "learning_rate": 1.9463428533351202e-07,
+      "loss": 0.6062,
+      "step": 10803
+    },
+    {
+      "epoch": 0.8774466011532527,
+      "grad_norm": 4.183435356460552,
+      "learning_rate": 1.943799696483145e-07,
+      "loss": 0.4585,
+      "step": 10804
+    },
+    {
+      "epoch": 0.8775278161292942,
+      "grad_norm": 4.396794358678346,
+      "learning_rate": 1.9412581349982113e-07,
+      "loss": 0.4212,
+      "step": 10805
+    },
+    {
+      "epoch": 0.8776090311053358,
+      "grad_norm": 4.627929173347972,
+      "learning_rate": 1.938718169056175e-07,
+      "loss": 0.5645,
+      "step": 10806
+    },
+    {
+      "epoch": 0.8776902460813774,
+      "grad_norm": 4.99764782498259,
+      "learning_rate": 1.9361797988327961e-07,
+      "loss": 0.6663,
+      "step": 10807
+    },
+    {
+      "epoch": 0.877771461057419,
+      "grad_norm": 3.9890801417363937,
+      "learning_rate": 1.933643024503712e-07,
+      "loss": 0.5455,
+      "step": 10808
+    },
+    {
+      "epoch": 0.8778526760334606,
+      "grad_norm": 5.065903978646239,
+      "learning_rate": 1.9311078462444484e-07,
+      "loss": 0.633,
+      "step": 10809
+    },
+    {
+      "epoch": 0.8779338910095021,
+      "grad_norm": 7.184247053486219,
+      "learning_rate": 1.928574264230429e-07,
+      "loss": 0.3581,
+      "step": 10810
+    },
+    {
+      "epoch": 0.8780151059855438,
+      "grad_norm": 7.648845912445987,
+      "learning_rate": 1.9260422786369747e-07,
+      "loss": 0.3958,
+      "step": 10811
+    },
+    {
+      "epoch": 0.8780963209615853,
+      "grad_norm": 2.778368867161788,
+      "learning_rate": 1.9235118896392706e-07,
+      "loss": 0.3277,
+      "step": 10812
+    },
+    {
+      "epoch": 0.8781775359376269,
+      "grad_norm": 4.8097717739494685,
+      "learning_rate": 1.9209830974124183e-07,
+      "loss": 0.4247,
+      "step": 10813
+    },
+    {
+      "epoch": 0.8782587509136685,
+      "grad_norm": 10.137967528923353,
+      "learning_rate": 1.9184559021313914e-07,
+      "loss": 0.4587,
+      "step": 10814
+    },
+    {
+      "epoch": 0.8783399658897101,
+      "grad_norm": 4.796708634400599,
+      "learning_rate": 1.9159303039710558e-07,
+      "loss": 0.4826,
+      "step": 10815
+    },
+    {
+      "epoch": 0.8784211808657516,
+      "grad_norm": 9.314388410278989,
+      "learning_rate": 1.9134063031061744e-07,
+      "loss": 0.516,
+      "step": 10816
+    },
+    {
+      "epoch": 0.8785023958417932,
+      "grad_norm": 7.232628959274363,
+      "learning_rate": 1.910883899711391e-07,
+      "loss": 0.5164,
+      "step": 10817
+    },
+    {
+      "epoch": 0.8785836108178348,
+      "grad_norm": 3.9022829908795322,
+      "learning_rate": 1.9083630939612407e-07,
+      "loss": 0.5444,
+      "step": 10818
+    },
+    {
+      "epoch": 0.8786648257938764,
+      "grad_norm": 3.8942957114269627,
+      "learning_rate": 1.9058438860301621e-07,
+      "loss": 0.3569,
+      "step": 10819
+    },
+    {
+      "epoch": 0.878746040769918,
+      "grad_norm": 3.166404623353523,
+      "learning_rate": 1.9033262760924598e-07,
+      "loss": 0.4069,
+      "step": 10820
+    },
+    {
+      "epoch": 0.8788272557459595,
+      "grad_norm": 5.035265769137359,
+      "learning_rate": 1.900810264322339e-07,
+      "loss": 0.3724,
+      "step": 10821
+    },
+    {
+      "epoch": 0.8789084707220012,
+      "grad_norm": 6.137216581765903,
+      "learning_rate": 1.8982958508938998e-07,
+      "loss": 0.4983,
+      "step": 10822
+    },
+    {
+      "epoch": 0.8789896856980427,
+      "grad_norm": 6.449621491427784,
+      "learning_rate": 1.895783035981119e-07,
+      "loss": 0.5643,
+      "step": 10823
+    },
+    {
+      "epoch": 0.8790709006740843,
+      "grad_norm": 7.017513815097841,
+      "learning_rate": 1.8932718197578802e-07,
+      "loss": 0.4371,
+      "step": 10824
+    },
+    {
+      "epoch": 0.8791521156501259,
+      "grad_norm": 4.915420085875843,
+      "learning_rate": 1.890762202397936e-07,
+      "loss": 0.7439,
+      "step": 10825
+    },
+    {
+      "epoch": 0.8792333306261675,
+      "grad_norm": 5.057095171528424,
+      "learning_rate": 1.8882541840749475e-07,
+      "loss": 0.5297,
+      "step": 10826
+    },
+    {
+      "epoch": 0.879314545602209,
+      "grad_norm": 4.992472857392407,
+      "learning_rate": 1.8857477649624533e-07,
+      "loss": 0.5945,
+      "step": 10827
+    },
+    {
+      "epoch": 0.8793957605782506,
+      "grad_norm": 8.570114515507747,
+      "learning_rate": 1.883242945233879e-07,
+      "loss": 0.5583,
+      "step": 10828
+    },
+    {
+      "epoch": 0.8794769755542922,
+      "grad_norm": 4.881453778874329,
+      "learning_rate": 1.8807397250625497e-07,
+      "loss": 0.3574,
+      "step": 10829
+    },
+    {
+      "epoch": 0.8795581905303338,
+      "grad_norm": 9.774498600782247,
+      "learning_rate": 1.878238104621677e-07,
+      "loss": 0.2625,
+      "step": 10830
+    },
+    {
+      "epoch": 0.8796394055063754,
+      "grad_norm": 5.93121202645404,
+      "learning_rate": 1.8757380840843526e-07,
+      "loss": 0.4918,
+      "step": 10831
+    },
+    {
+      "epoch": 0.879720620482417,
+      "grad_norm": 3.681222565715722,
+      "learning_rate": 1.8732396636235744e-07,
+      "loss": 0.6349,
+      "step": 10832
+    },
+    {
+      "epoch": 0.8798018354584586,
+      "grad_norm": 6.108142266361864,
+      "learning_rate": 1.8707428434122155e-07,
+      "loss": 0.576,
+      "step": 10833
+    },
+    {
+      "epoch": 0.8798830504345001,
+      "grad_norm": 6.054085118997518,
+      "learning_rate": 1.8682476236230372e-07,
+      "loss": 0.4904,
+      "step": 10834
+    },
+    {
+      "epoch": 0.8799642654105417,
+      "grad_norm": 7.843292571737394,
+      "learning_rate": 1.8657540044287047e-07,
+      "loss": 0.3619,
+      "step": 10835
+    },
+    {
+      "epoch": 0.8800454803865833,
+      "grad_norm": 3.6574632924680044,
+      "learning_rate": 1.8632619860017547e-07,
+      "loss": 0.3475,
+      "step": 10836
+    },
+    {
+      "epoch": 0.8801266953626249,
+      "grad_norm": 5.74599899447566,
+      "learning_rate": 1.8607715685146244e-07,
+      "loss": 0.463,
+      "step": 10837
+    },
+    {
+      "epoch": 0.8802079103386664,
+      "grad_norm": 4.130892724817378,
+      "learning_rate": 1.8582827521396453e-07,
+      "loss": 0.5069,
+      "step": 10838
+    },
+    {
+      "epoch": 0.880289125314708,
+      "grad_norm": 6.536393003461804,
+      "learning_rate": 1.855795537049021e-07,
+      "loss": 0.4378,
+      "step": 10839
+    },
+    {
+      "epoch": 0.8803703402907496,
+      "grad_norm": 5.520938271169849,
+      "learning_rate": 1.853309923414856e-07,
+      "loss": 0.481,
+      "step": 10840
+    },
+    {
+      "epoch": 0.8804515552667912,
+      "grad_norm": 13.77624681570704,
+      "learning_rate": 1.8508259114091432e-07,
+      "loss": 0.5875,
+      "step": 10841
+    },
+    {
+      "epoch": 0.8805327702428328,
+      "grad_norm": 4.241071474476681,
+      "learning_rate": 1.8483435012037587e-07,
+      "loss": 0.3521,
+      "step": 10842
+    },
+    {
+      "epoch": 0.8806139852188744,
+      "grad_norm": 8.981810629766965,
+      "learning_rate": 1.8458626929704821e-07,
+      "loss": 0.4218,
+      "step": 10843
+    },
+    {
+      "epoch": 0.880695200194916,
+      "grad_norm": 4.1379268820071236,
+      "learning_rate": 1.843383486880959e-07,
+      "loss": 0.6531,
+      "step": 10844
+    },
+    {
+      "epoch": 0.8807764151709575,
+      "grad_norm": 4.30762221426034,
+      "learning_rate": 1.840905883106747e-07,
+      "loss": 0.6161,
+      "step": 10845
+    },
+    {
+      "epoch": 0.8808576301469991,
+      "grad_norm": 6.207428929843423,
+      "learning_rate": 1.8384298818192814e-07,
+      "loss": 0.4597,
+      "step": 10846
+    },
+    {
+      "epoch": 0.8809388451230407,
+      "grad_norm": 6.194628295310705,
+      "learning_rate": 1.835955483189883e-07,
+      "loss": 0.4932,
+      "step": 10847
+    },
+    {
+      "epoch": 0.8810200600990823,
+      "grad_norm": 4.398170364958908,
+      "learning_rate": 1.833482687389776e-07,
+      "loss": 0.4016,
+      "step": 10848
+    },
+    {
+      "epoch": 0.8811012750751238,
+      "grad_norm": 6.570347236455752,
+      "learning_rate": 1.831011494590054e-07,
+      "loss": 0.4765,
+      "step": 10849
+    },
+    {
+      "epoch": 0.8811824900511654,
+      "grad_norm": 5.039000129540002,
+      "learning_rate": 1.828541904961717e-07,
+      "loss": 0.4721,
+      "step": 10850
+    },
+    {
+      "epoch": 0.881263705027207,
+      "grad_norm": 6.470581102201411,
+      "learning_rate": 1.8260739186756527e-07,
+      "loss": 0.7508,
+      "step": 10851
+    },
+    {
+      "epoch": 0.8813449200032486,
+      "grad_norm": 4.439815787899128,
+      "learning_rate": 1.8236075359026246e-07,
+      "loss": 0.4522,
+      "step": 10852
+    },
+    {
+      "epoch": 0.8814261349792902,
+      "grad_norm": 6.875926190964903,
+      "learning_rate": 1.8211427568132932e-07,
+      "loss": 0.4571,
+      "step": 10853
+    },
+    {
+      "epoch": 0.8815073499553318,
+      "grad_norm": 5.042487513250472,
+      "learning_rate": 1.8186795815782143e-07,
+      "loss": 0.4715,
+      "step": 10854
+    },
+    {
+      "epoch": 0.8815885649313734,
+      "grad_norm": 6.319914060990893,
+      "learning_rate": 1.8162180103678177e-07,
+      "loss": 0.357,
+      "step": 10855
+    },
+    {
+      "epoch": 0.8816697799074149,
+      "grad_norm": 5.98665836166037,
+      "learning_rate": 1.813758043352437e-07,
+      "loss": 0.6732,
+      "step": 10856
+    },
+    {
+      "epoch": 0.8817509948834565,
+      "grad_norm": 4.90779966317718,
+      "learning_rate": 1.8112996807022943e-07,
+      "loss": 0.4064,
+      "step": 10857
+    },
+    {
+      "epoch": 0.8818322098594981,
+      "grad_norm": 4.954078590442837,
+      "learning_rate": 1.8088429225874865e-07,
+      "loss": 0.5962,
+      "step": 10858
+    },
+    {
+      "epoch": 0.8819134248355397,
+      "grad_norm": 7.208628926964807,
+      "learning_rate": 1.8063877691780114e-07,
+      "loss": 0.4485,
+      "step": 10859
+    },
+    {
+      "epoch": 0.8819946398115812,
+      "grad_norm": 5.399228378212703,
+      "learning_rate": 1.8039342206437494e-07,
+      "loss": 0.6773,
+      "step": 10860
+    },
+    {
+      "epoch": 0.8820758547876228,
+      "grad_norm": 6.733238765318234,
+      "learning_rate": 1.8014822771544787e-07,
+      "loss": 0.5487,
+      "step": 10861
+    },
+    {
+      "epoch": 0.8821570697636644,
+      "grad_norm": 6.204508713784901,
+      "learning_rate": 1.7990319388798527e-07,
+      "loss": 0.379,
+      "step": 10862
+    },
+    {
+      "epoch": 0.882238284739706,
+      "grad_norm": 4.9418711968437545,
+      "learning_rate": 1.79658320598943e-07,
+      "loss": 0.4973,
+      "step": 10863
+    },
+    {
+      "epoch": 0.8823194997157476,
+      "grad_norm": 7.6609348179930805,
+      "learning_rate": 1.79413607865265e-07,
+      "loss": 0.5288,
+      "step": 10864
+    },
+    {
+      "epoch": 0.8824007146917892,
+      "grad_norm": 5.302913931640557,
+      "learning_rate": 1.7916905570388387e-07,
+      "loss": 0.4702,
+      "step": 10865
+    },
+    {
+      "epoch": 0.8824819296678308,
+      "grad_norm": 5.228383412229427,
+      "learning_rate": 1.7892466413172076e-07,
+      "loss": 0.3797,
+      "step": 10866
+    },
+    {
+      "epoch": 0.8825631446438723,
+      "grad_norm": 4.948041110181531,
+      "learning_rate": 1.7868043316568718e-07,
+      "loss": 0.3931,
+      "step": 10867
+    },
+    {
+      "epoch": 0.8826443596199139,
+      "grad_norm": 8.069840917352794,
+      "learning_rate": 1.784363628226818e-07,
+      "loss": 0.4122,
+      "step": 10868
+    },
+    {
+      "epoch": 0.8827255745959555,
+      "grad_norm": 5.771520601557648,
+      "learning_rate": 1.781924531195933e-07,
+      "loss": 0.4682,
+      "step": 10869
+    },
+    {
+      "epoch": 0.8828067895719971,
+      "grad_norm": 5.964777339901852,
+      "learning_rate": 1.7794870407329968e-07,
+      "loss": 0.3881,
+      "step": 10870
+    },
+    {
+      "epoch": 0.8828880045480386,
+      "grad_norm": 17.63419248026244,
+      "learning_rate": 1.7770511570066622e-07,
+      "loss": 0.4282,
+      "step": 10871
+    },
+    {
+      "epoch": 0.8829692195240803,
+      "grad_norm": 4.937868582682188,
+      "learning_rate": 1.7746168801854786e-07,
+      "loss": 0.4074,
+      "step": 10872
+    },
+    {
+      "epoch": 0.8830504345001218,
+      "grad_norm": 7.45687522217002,
+      "learning_rate": 1.772184210437894e-07,
+      "loss": 0.3702,
+      "step": 10873
+    },
+    {
+      "epoch": 0.8831316494761634,
+      "grad_norm": 4.456345546727624,
+      "learning_rate": 1.7697531479322294e-07,
+      "loss": 0.4297,
+      "step": 10874
+    },
+    {
+      "epoch": 0.883212864452205,
+      "grad_norm": 4.869647315752035,
+      "learning_rate": 1.7673236928366976e-07,
+      "loss": 0.4171,
+      "step": 10875
+    },
+    {
+      "epoch": 0.8832940794282466,
+      "grad_norm": 4.695411262956829,
+      "learning_rate": 1.7648958453194086e-07,
+      "loss": 0.4707,
+      "step": 10876
+    },
+    {
+      "epoch": 0.8833752944042882,
+      "grad_norm": 7.882022541168915,
+      "learning_rate": 1.7624696055483643e-07,
+      "loss": 0.4886,
+      "step": 10877
+    },
+    {
+      "epoch": 0.8834565093803297,
+      "grad_norm": 7.071815658627316,
+      "learning_rate": 1.7600449736914384e-07,
+      "loss": 0.4957,
+      "step": 10878
+    },
+    {
+      "epoch": 0.8835377243563713,
+      "grad_norm": 5.21615020327953,
+      "learning_rate": 1.7576219499163995e-07,
+      "loss": 0.395,
+      "step": 10879
+    },
+    {
+      "epoch": 0.8836189393324129,
+      "grad_norm": 3.7677226715193015,
+      "learning_rate": 1.7552005343909162e-07,
+      "loss": 0.6283,
+      "step": 10880
+    },
+    {
+      "epoch": 0.8837001543084545,
+      "grad_norm": 6.221136511822739,
+      "learning_rate": 1.7527807272825326e-07,
+      "loss": 0.4107,
+      "step": 10881
+    },
+    {
+      "epoch": 0.883781369284496,
+      "grad_norm": 4.483843051687907,
+      "learning_rate": 1.7503625287586896e-07,
+      "loss": 0.4108,
+      "step": 10882
+    },
+    {
+      "epoch": 0.8838625842605377,
+      "grad_norm": 5.371900604040972,
+      "learning_rate": 1.7479459389867141e-07,
+      "loss": 0.463,
+      "step": 10883
+    },
+    {
+      "epoch": 0.8839437992365792,
+      "grad_norm": 4.729136507795324,
+      "learning_rate": 1.7455309581338204e-07,
+      "loss": 0.5657,
+      "step": 10884
+    },
+    {
+      "epoch": 0.8840250142126208,
+      "grad_norm": 5.952224993059543,
+      "learning_rate": 1.7431175863671102e-07,
+      "loss": 0.462,
+      "step": 10885
+    },
+    {
+      "epoch": 0.8841062291886624,
+      "grad_norm": 5.332245081683401,
+      "learning_rate": 1.740705823853578e-07,
+      "loss": 0.5023,
+      "step": 10886
+    },
+    {
+      "epoch": 0.884187444164704,
+      "grad_norm": 4.301376869737246,
+      "learning_rate": 1.7382956707601068e-07,
+      "loss": 0.6023,
+      "step": 10887
+    },
+    {
+      "epoch": 0.8842686591407456,
+      "grad_norm": 7.399999113034231,
+      "learning_rate": 1.7358871272534604e-07,
+      "loss": 0.383,
+      "step": 10888
+    },
+    {
+      "epoch": 0.8843498741167871,
+      "grad_norm": 6.01935069429763,
+      "learning_rate": 1.7334801935003003e-07,
+      "loss": 0.3966,
+      "step": 10889
+    },
+    {
+      "epoch": 0.8844310890928287,
+      "grad_norm": 12.267510974208244,
+      "learning_rate": 1.7310748696671791e-07,
+      "loss": 0.4748,
+      "step": 10890
+    },
+    {
+      "epoch": 0.8845123040688703,
+      "grad_norm": 4.471683226123735,
+      "learning_rate": 1.728671155920525e-07,
+      "loss": 0.557,
+      "step": 10891
+    },
+    {
+      "epoch": 0.8845935190449119,
+      "grad_norm": 4.5797052007927626,
+      "learning_rate": 1.7262690524266658e-07,
+      "loss": 0.3825,
+      "step": 10892
+    },
+    {
+      "epoch": 0.8846747340209534,
+      "grad_norm": 5.819985449696891,
+      "learning_rate": 1.7238685593518157e-07,
+      "loss": 0.4803,
+      "step": 10893
+    },
+    {
+      "epoch": 0.8847559489969951,
+      "grad_norm": 12.610582423388232,
+      "learning_rate": 1.7214696768620699e-07,
+      "loss": 0.4021,
+      "step": 10894
+    },
+    {
+      "epoch": 0.8848371639730366,
+      "grad_norm": 5.295182105134932,
+      "learning_rate": 1.719072405123423e-07,
+      "loss": 0.4256,
+      "step": 10895
+    },
+    {
+      "epoch": 0.8849183789490782,
+      "grad_norm": 4.518250999036192,
+      "learning_rate": 1.7166767443017567e-07,
+      "loss": 0.4549,
+      "step": 10896
+    },
+    {
+      "epoch": 0.8849995939251198,
+      "grad_norm": 3.165277818453232,
+      "learning_rate": 1.7142826945628353e-07,
+      "loss": 0.7135,
+      "step": 10897
+    },
+    {
+      "epoch": 0.8850808089011614,
+      "grad_norm": 14.285562335378861,
+      "learning_rate": 1.7118902560723072e-07,
+      "loss": 0.5167,
+      "step": 10898
+    },
+    {
+      "epoch": 0.885162023877203,
+      "grad_norm": 5.520164322684935,
+      "learning_rate": 1.7094994289957285e-07,
+      "loss": 0.4811,
+      "step": 10899
+    },
+    {
+      "epoch": 0.8852432388532445,
+      "grad_norm": 6.172181533567202,
+      "learning_rate": 1.7071102134985224e-07,
+      "loss": 0.4762,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8853244538292862,
+      "grad_norm": 4.846940632513523,
+      "learning_rate": 1.7047226097460123e-07,
+      "loss": 0.3914,
+      "step": 10901
+    },
+    {
+      "epoch": 0.8854056688053277,
+      "grad_norm": 3.9531646427451927,
+      "learning_rate": 1.7023366179034135e-07,
+      "loss": 0.4715,
+      "step": 10902
+    },
+    {
+      "epoch": 0.8854868837813693,
+      "grad_norm": 7.419094636905311,
+      "learning_rate": 1.6999522381358187e-07,
+      "loss": 0.3049,
+      "step": 10903
+    },
+    {
+      "epoch": 0.8855680987574108,
+      "grad_norm": 8.695140393852423,
+      "learning_rate": 1.6975694706082125e-07,
+      "loss": 0.4374,
+      "step": 10904
+    },
+    {
+      "epoch": 0.8856493137334525,
+      "grad_norm": 4.688891686843388,
+      "learning_rate": 1.6951883154854771e-07,
+      "loss": 0.3473,
+      "step": 10905
+    },
+    {
+      "epoch": 0.885730528709494,
+      "grad_norm": 9.242824312960995,
+      "learning_rate": 1.6928087729323695e-07,
+      "loss": 0.4753,
+      "step": 10906
+    },
+    {
+      "epoch": 0.8858117436855356,
+      "grad_norm": 21.34205228899008,
+      "learning_rate": 1.6904308431135414e-07,
+      "loss": 0.4638,
+      "step": 10907
+    },
+    {
+      "epoch": 0.8858929586615772,
+      "grad_norm": 7.234575319695479,
+      "learning_rate": 1.6880545261935333e-07,
+      "loss": 0.4495,
+      "step": 10908
+    },
+    {
+      "epoch": 0.8859741736376188,
+      "grad_norm": 4.584768372230779,
+      "learning_rate": 1.6856798223367777e-07,
+      "loss": 0.4903,
+      "step": 10909
+    },
+    {
+      "epoch": 0.8860553886136604,
+      "grad_norm": 3.8405465855706966,
+      "learning_rate": 1.6833067317075875e-07,
+      "loss": 0.5838,
+      "step": 10910
+    },
+    {
+      "epoch": 0.8861366035897019,
+      "grad_norm": 11.336953000753638,
+      "learning_rate": 1.680935254470173e-07,
+      "loss": 0.3647,
+      "step": 10911
+    },
+    {
+      "epoch": 0.8862178185657436,
+      "grad_norm": 3.4138677524491783,
+      "learning_rate": 1.6785653907886251e-07,
+      "loss": 0.4593,
+      "step": 10912
+    },
+    {
+      "epoch": 0.8862990335417851,
+      "grad_norm": 7.281119584083047,
+      "learning_rate": 1.6761971408269184e-07,
+      "loss": 0.4414,
+      "step": 10913
+    },
+    {
+      "epoch": 0.8863802485178267,
+      "grad_norm": 8.096587754790656,
+      "learning_rate": 1.673830504748933e-07,
+      "loss": 0.5625,
+      "step": 10914
+    },
+    {
+      "epoch": 0.8864614634938682,
+      "grad_norm": 4.3613469827373725,
+      "learning_rate": 1.6714654827184263e-07,
+      "loss": 0.6731,
+      "step": 10915
+    },
+    {
+      "epoch": 0.8865426784699099,
+      "grad_norm": 36.26226393417871,
+      "learning_rate": 1.6691020748990455e-07,
+      "loss": 0.4585,
+      "step": 10916
+    },
+    {
+      "epoch": 0.8866238934459514,
+      "grad_norm": 4.750194227102958,
+      "learning_rate": 1.6667402814543209e-07,
+      "loss": 0.5141,
+      "step": 10917
+    },
+    {
+      "epoch": 0.886705108421993,
+      "grad_norm": 7.5959663569614495,
+      "learning_rate": 1.66438010254768e-07,
+      "loss": 0.6047,
+      "step": 10918
+    },
+    {
+      "epoch": 0.8867863233980346,
+      "grad_norm": 6.985478995191811,
+      "learning_rate": 1.662021538342437e-07,
+      "loss": 0.5965,
+      "step": 10919
+    },
+    {
+      "epoch": 0.8868675383740762,
+      "grad_norm": 4.967674400774454,
+      "learning_rate": 1.6596645890017832e-07,
+      "loss": 0.4358,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8869487533501178,
+      "grad_norm": 4.562428672417342,
+      "learning_rate": 1.6573092546888132e-07,
+      "loss": 0.4619,
+      "step": 10921
+    },
+    {
+      "epoch": 0.8870299683261593,
+      "grad_norm": 4.147461101360203,
+      "learning_rate": 1.6549555355665076e-07,
+      "loss": 0.452,
+      "step": 10922
+    },
+    {
+      "epoch": 0.887111183302201,
+      "grad_norm": 3.695271801611688,
+      "learning_rate": 1.6526034317977225e-07,
+      "loss": 0.5393,
+      "step": 10923
+    },
+    {
+      "epoch": 0.8871923982782425,
+      "grad_norm": 4.837405408033375,
+      "learning_rate": 1.650252943545222e-07,
+      "loss": 0.4885,
+      "step": 10924
+    },
+    {
+      "epoch": 0.8872736132542841,
+      "grad_norm": 4.704508929048984,
+      "learning_rate": 1.647904070971637e-07,
+      "loss": 0.5001,
+      "step": 10925
+    },
+    {
+      "epoch": 0.8873548282303256,
+      "grad_norm": 3.9115802054782427,
+      "learning_rate": 1.645556814239499e-07,
+      "loss": 0.4963,
+      "step": 10926
+    },
+    {
+      "epoch": 0.8874360432063673,
+      "grad_norm": 7.938004047544957,
+      "learning_rate": 1.6432111735112277e-07,
+      "loss": 0.4367,
+      "step": 10927
+    },
+    {
+      "epoch": 0.8875172581824088,
+      "grad_norm": 12.387873573548433,
+      "learning_rate": 1.6408671489491323e-07,
+      "loss": 0.3794,
+      "step": 10928
+    },
+    {
+      "epoch": 0.8875984731584504,
+      "grad_norm": 7.148993736986469,
+      "learning_rate": 1.6385247407154025e-07,
+      "loss": 0.4554,
+      "step": 10929
+    },
+    {
+      "epoch": 0.887679688134492,
+      "grad_norm": 6.143994601377806,
+      "learning_rate": 1.6361839489721227e-07,
+      "loss": 0.46,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8877609031105336,
+      "grad_norm": 21.816971684053996,
+      "learning_rate": 1.6338447738812628e-07,
+      "loss": 0.6431,
+      "step": 10931
+    },
+    {
+      "epoch": 0.8878421180865752,
+      "grad_norm": 6.378758988903642,
+      "learning_rate": 1.631507215604683e-07,
+      "loss": 0.4062,
+      "step": 10932
+    },
+    {
+      "epoch": 0.8879233330626167,
+      "grad_norm": 8.818910417308302,
+      "learning_rate": 1.6291712743041226e-07,
+      "loss": 0.4481,
+      "step": 10933
+    },
+    {
+      "epoch": 0.8880045480386584,
+      "grad_norm": 3.3592851595714537,
+      "learning_rate": 1.6268369501412195e-07,
+      "loss": 0.6451,
+      "step": 10934
+    },
+    {
+      "epoch": 0.8880857630146999,
+      "grad_norm": 3.537250723511631,
+      "learning_rate": 1.6245042432775054e-07,
+      "loss": 0.4928,
+      "step": 10935
+    },
+    {
+      "epoch": 0.8881669779907415,
+      "grad_norm": 6.147519517852104,
+      "learning_rate": 1.622173153874379e-07,
+      "loss": 0.5083,
+      "step": 10936
+    },
+    {
+      "epoch": 0.888248192966783,
+      "grad_norm": 3.472301797973062,
+      "learning_rate": 1.61984368209315e-07,
+      "loss": 0.4876,
+      "step": 10937
+    },
+    {
+      "epoch": 0.8883294079428247,
+      "grad_norm": 4.755539554685603,
+      "learning_rate": 1.617515828095001e-07,
+      "loss": 0.3319,
+      "step": 10938
+    },
+    {
+      "epoch": 0.8884106229188662,
+      "grad_norm": 9.123690619357854,
+      "learning_rate": 1.615189592041e-07,
+      "loss": 0.5498,
+      "step": 10939
+    },
+    {
+      "epoch": 0.8884918378949078,
+      "grad_norm": 5.947150990803091,
+      "learning_rate": 1.6128649740921182e-07,
+      "loss": 0.4218,
+      "step": 10940
+    },
+    {
+      "epoch": 0.8885730528709495,
+      "grad_norm": 6.785328375370574,
+      "learning_rate": 1.6105419744092105e-07,
+      "loss": 0.5246,
+      "step": 10941
+    },
+    {
+      "epoch": 0.888654267846991,
+      "grad_norm": 7.518002211907693,
+      "learning_rate": 1.6082205931530064e-07,
+      "loss": 0.4103,
+      "step": 10942
+    },
+    {
+      "epoch": 0.8887354828230326,
+      "grad_norm": 5.241816204201098,
+      "learning_rate": 1.6059008304841417e-07,
+      "loss": 0.422,
+      "step": 10943
+    },
+    {
+      "epoch": 0.8888166977990741,
+      "grad_norm": 4.9351607203741175,
+      "learning_rate": 1.6035826865631292e-07,
+      "loss": 0.4762,
+      "step": 10944
+    },
+    {
+      "epoch": 0.8888979127751158,
+      "grad_norm": 19.003799619323694,
+      "learning_rate": 1.601266161550366e-07,
+      "loss": 0.6662,
+      "step": 10945
+    },
+    {
+      "epoch": 0.8889791277511573,
+      "grad_norm": 3.8825556090436786,
+      "learning_rate": 1.5989512556061516e-07,
+      "loss": 0.6389,
+      "step": 10946
+    },
+    {
+      "epoch": 0.8890603427271989,
+      "grad_norm": 7.947300455523199,
+      "learning_rate": 1.5966379688906576e-07,
+      "loss": 0.4424,
+      "step": 10947
+    },
+    {
+      "epoch": 0.8891415577032404,
+      "grad_norm": 5.795613017876368,
+      "learning_rate": 1.5943263015639614e-07,
+      "loss": 0.4964,
+      "step": 10948
+    },
+    {
+      "epoch": 0.8892227726792821,
+      "grad_norm": 6.8505646268664515,
+      "learning_rate": 1.592016253786008e-07,
+      "loss": 0.6119,
+      "step": 10949
+    },
+    {
+      "epoch": 0.8893039876553236,
+      "grad_norm": 8.889358889295613,
+      "learning_rate": 1.5897078257166492e-07,
+      "loss": 0.4562,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8893852026313652,
+      "grad_norm": 5.6468959215119225,
+      "learning_rate": 1.5874010175156106e-07,
+      "loss": 0.4509,
+      "step": 10951
+    },
+    {
+      "epoch": 0.8894664176074069,
+      "grad_norm": 5.671002281059167,
+      "learning_rate": 1.585095829342509e-07,
+      "loss": 0.3751,
+      "step": 10952
+    },
+    {
+      "epoch": 0.8895476325834484,
+      "grad_norm": 7.946535661783076,
+      "learning_rate": 1.5827922613568524e-07,
+      "loss": 0.4659,
+      "step": 10953
+    },
+    {
+      "epoch": 0.88962884755949,
+      "grad_norm": 5.9522004926872025,
+      "learning_rate": 1.5804903137180415e-07,
+      "loss": 0.3708,
+      "step": 10954
+    },
+    {
+      "epoch": 0.8897100625355315,
+      "grad_norm": 10.178045190705534,
+      "learning_rate": 1.5781899865853544e-07,
+      "loss": 0.5587,
+      "step": 10955
+    },
+    {
+      "epoch": 0.8897912775115732,
+      "grad_norm": 4.253031697214437,
+      "learning_rate": 1.5758912801179637e-07,
+      "loss": 0.4611,
+      "step": 10956
+    },
+    {
+      "epoch": 0.8898724924876147,
+      "grad_norm": 6.692152756992295,
+      "learning_rate": 1.5735941944749255e-07,
+      "loss": 0.4811,
+      "step": 10957
+    },
+    {
+      "epoch": 0.8899537074636563,
+      "grad_norm": 4.694878917526122,
+      "learning_rate": 1.571298729815182e-07,
+      "loss": 0.3912,
+      "step": 10958
+    },
+    {
+      "epoch": 0.8900349224396978,
+      "grad_norm": 5.740301752503594,
+      "learning_rate": 1.569004886297576e-07,
+      "loss": 0.5173,
+      "step": 10959
+    },
+    {
+      "epoch": 0.8901161374157395,
+      "grad_norm": 4.500280402580769,
+      "learning_rate": 1.5667126640808216e-07,
+      "loss": 0.5058,
+      "step": 10960
+    },
+    {
+      "epoch": 0.890197352391781,
+      "grad_norm": 9.359821537794131,
+      "learning_rate": 1.564422063323534e-07,
+      "loss": 0.4839,
+      "step": 10961
+    },
+    {
+      "epoch": 0.8902785673678226,
+      "grad_norm": 11.943572724510128,
+      "learning_rate": 1.5621330841842086e-07,
+      "loss": 0.3788,
+      "step": 10962
+    },
+    {
+      "epoch": 0.8903597823438643,
+      "grad_norm": 5.141666870810457,
+      "learning_rate": 1.5598457268212353e-07,
+      "loss": 0.4711,
+      "step": 10963
+    },
+    {
+      "epoch": 0.8904409973199058,
+      "grad_norm": 3.330373351694768,
+      "learning_rate": 1.5575599913928735e-07,
+      "loss": 0.3897,
+      "step": 10964
+    },
+    {
+      "epoch": 0.8905222122959474,
+      "grad_norm": 3.825856052673269,
+      "learning_rate": 1.5552758780572995e-07,
+      "loss": 0.4227,
+      "step": 10965
+    },
+    {
+      "epoch": 0.8906034272719889,
+      "grad_norm": 5.142764433332306,
+      "learning_rate": 1.552993386972551e-07,
+      "loss": 0.4816,
+      "step": 10966
+    },
+    {
+      "epoch": 0.8906846422480306,
+      "grad_norm": 6.3600498260391785,
+      "learning_rate": 1.5507125182965737e-07,
+      "loss": 0.684,
+      "step": 10967
+    },
+    {
+      "epoch": 0.8907658572240721,
+      "grad_norm": 8.817725414443457,
+      "learning_rate": 1.5484332721871804e-07,
+      "loss": 0.5193,
+      "step": 10968
+    },
+    {
+      "epoch": 0.8908470722001137,
+      "grad_norm": 7.292802786468984,
+      "learning_rate": 1.5461556488020945e-07,
+      "loss": 0.4265,
+      "step": 10969
+    },
+    {
+      "epoch": 0.8909282871761552,
+      "grad_norm": 18.816459578152827,
+      "learning_rate": 1.5438796482989072e-07,
+      "loss": 0.4049,
+      "step": 10970
+    },
+    {
+      "epoch": 0.8910095021521969,
+      "grad_norm": 4.741507710988063,
+      "learning_rate": 1.541605270835106e-07,
+      "loss": 0.444,
+      "step": 10971
+    },
+    {
+      "epoch": 0.8910907171282384,
+      "grad_norm": 11.630609215202053,
+      "learning_rate": 1.5393325165680707e-07,
+      "loss": 0.4207,
+      "step": 10972
+    },
+    {
+      "epoch": 0.89117193210428,
+      "grad_norm": 3.301339255185129,
+      "learning_rate": 1.5370613856550615e-07,
+      "loss": 0.5021,
+      "step": 10973
+    },
+    {
+      "epoch": 0.8912531470803217,
+      "grad_norm": 4.4372396364921105,
+      "learning_rate": 1.534791878253228e-07,
+      "loss": 0.3981,
+      "step": 10974
+    },
+    {
+      "epoch": 0.8913343620563632,
+      "grad_norm": 8.507200043869714,
+      "learning_rate": 1.5325239945196108e-07,
+      "loss": 0.4659,
+      "step": 10975
+    },
+    {
+      "epoch": 0.8914155770324048,
+      "grad_norm": 6.152072516412701,
+      "learning_rate": 1.530257734611132e-07,
+      "loss": 0.4745,
+      "step": 10976
+    },
+    {
+      "epoch": 0.8914967920084463,
+      "grad_norm": 3.5425159922458582,
+      "learning_rate": 1.5279930986846047e-07,
+      "loss": 0.6653,
+      "step": 10977
+    },
+    {
+      "epoch": 0.891578006984488,
+      "grad_norm": 5.233121096684976,
+      "learning_rate": 1.5257300868967344e-07,
+      "loss": 0.5044,
+      "step": 10978
+    },
+    {
+      "epoch": 0.8916592219605295,
+      "grad_norm": 6.291818310904449,
+      "learning_rate": 1.5234686994041016e-07,
+      "loss": 0.4987,
+      "step": 10979
+    },
+    {
+      "epoch": 0.8917404369365711,
+      "grad_norm": 5.098504129650742,
+      "learning_rate": 1.521208936363186e-07,
+      "loss": 0.3563,
+      "step": 10980
+    },
+    {
+      "epoch": 0.8918216519126126,
+      "grad_norm": 4.803655200270107,
+      "learning_rate": 1.5189507979303575e-07,
+      "loss": 0.7008,
+      "step": 10981
+    },
+    {
+      "epoch": 0.8919028668886543,
+      "grad_norm": 4.05749114440078,
+      "learning_rate": 1.5166942842618632e-07,
+      "loss": 0.3977,
+      "step": 10982
+    },
+    {
+      "epoch": 0.8919840818646958,
+      "grad_norm": 6.222213392758151,
+      "learning_rate": 1.5144393955138336e-07,
+      "loss": 0.4221,
+      "step": 10983
+    },
+    {
+      "epoch": 0.8920652968407374,
+      "grad_norm": 3.854822685859984,
+      "learning_rate": 1.512186131842308e-07,
+      "loss": 0.4855,
+      "step": 10984
+    },
+    {
+      "epoch": 0.8921465118167791,
+      "grad_norm": 7.555916028857482,
+      "learning_rate": 1.5099344934031923e-07,
+      "loss": 0.4516,
+      "step": 10985
+    },
+    {
+      "epoch": 0.8922277267928206,
+      "grad_norm": 8.539386703282604,
+      "learning_rate": 1.507684480352292e-07,
+      "loss": 0.415,
+      "step": 10986
+    },
+    {
+      "epoch": 0.8923089417688622,
+      "grad_norm": 4.964839789056717,
+      "learning_rate": 1.5054360928452915e-07,
+      "loss": 0.4186,
+      "step": 10987
+    },
+    {
+      "epoch": 0.8923901567449037,
+      "grad_norm": 6.084627536003772,
+      "learning_rate": 1.5031893310377716e-07,
+      "loss": 0.5064,
+      "step": 10988
+    },
+    {
+      "epoch": 0.8924713717209454,
+      "grad_norm": 3.940604600092638,
+      "learning_rate": 1.5009441950851965e-07,
+      "loss": 0.5235,
+      "step": 10989
+    },
+    {
+      "epoch": 0.8925525866969869,
+      "grad_norm": 5.3695438746577375,
+      "learning_rate": 1.4987006851429147e-07,
+      "loss": 0.3932,
+      "step": 10990
+    },
+    {
+      "epoch": 0.8926338016730285,
+      "grad_norm": 9.145034230965127,
+      "learning_rate": 1.4964588013661657e-07,
+      "loss": 0.3775,
+      "step": 10991
+    },
+    {
+      "epoch": 0.89271501664907,
+      "grad_norm": 7.658853077669966,
+      "learning_rate": 1.4942185439100753e-07,
+      "loss": 0.4791,
+      "step": 10992
+    },
+    {
+      "epoch": 0.8927962316251117,
+      "grad_norm": 7.168128470178995,
+      "learning_rate": 1.4919799129296615e-07,
+      "loss": 0.3937,
+      "step": 10993
+    },
+    {
+      "epoch": 0.8928774466011532,
+      "grad_norm": 5.628929909843431,
+      "learning_rate": 1.489742908579822e-07,
+      "loss": 0.4437,
+      "step": 10994
+    },
+    {
+      "epoch": 0.8929586615771948,
+      "grad_norm": 6.054942900449149,
+      "learning_rate": 1.4875075310153504e-07,
+      "loss": 0.4699,
+      "step": 10995
+    },
+    {
+      "epoch": 0.8930398765532365,
+      "grad_norm": 5.835599987062635,
+      "learning_rate": 1.4852737803909167e-07,
+      "loss": 0.6048,
+      "step": 10996
+    },
+    {
+      "epoch": 0.893121091529278,
+      "grad_norm": 7.175331718017556,
+      "learning_rate": 1.4830416568610893e-07,
+      "loss": 0.5113,
+      "step": 10997
+    },
+    {
+      "epoch": 0.8932023065053196,
+      "grad_norm": 8.62919238424243,
+      "learning_rate": 1.4808111605803117e-07,
+      "loss": 0.4906,
+      "step": 10998
+    },
+    {
+      "epoch": 0.8932835214813611,
+      "grad_norm": 11.718705876342495,
+      "learning_rate": 1.4785822917029318e-07,
+      "loss": 0.415,
+      "step": 10999
+    },
+    {
+      "epoch": 0.8933647364574028,
+      "grad_norm": 6.786219876737145,
+      "learning_rate": 1.476355050383174e-07,
+      "loss": 0.483,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8934459514334443,
+      "grad_norm": 4.549779356360936,
+      "learning_rate": 1.4741294367751484e-07,
+      "loss": 0.5062,
+      "step": 11001
+    },
+    {
+      "epoch": 0.8935271664094859,
+      "grad_norm": 4.594436506496313,
+      "learning_rate": 1.4719054510328595e-07,
+      "loss": 0.3839,
+      "step": 11002
+    },
+    {
+      "epoch": 0.8936083813855275,
+      "grad_norm": 3.87676469142472,
+      "learning_rate": 1.4696830933101868e-07,
+      "loss": 0.4757,
+      "step": 11003
+    },
+    {
+      "epoch": 0.8936895963615691,
+      "grad_norm": 4.757027256089799,
+      "learning_rate": 1.467462363760916e-07,
+      "loss": 0.5542,
+      "step": 11004
+    },
+    {
+      "epoch": 0.8937708113376106,
+      "grad_norm": 3.9669910321207587,
+      "learning_rate": 1.4652432625387013e-07,
+      "loss": 0.4679,
+      "step": 11005
+    },
+    {
+      "epoch": 0.8938520263136522,
+      "grad_norm": 5.6200804103086295,
+      "learning_rate": 1.4630257897970985e-07,
+      "loss": 0.4068,
+      "step": 11006
+    },
+    {
+      "epoch": 0.8939332412896939,
+      "grad_norm": 3.8084388168574894,
+      "learning_rate": 1.4608099456895452e-07,
+      "loss": 0.4993,
+      "step": 11007
+    },
+    {
+      "epoch": 0.8940144562657354,
+      "grad_norm": 6.108741515365576,
+      "learning_rate": 1.4585957303693664e-07,
+      "loss": 0.4431,
+      "step": 11008
+    },
+    {
+      "epoch": 0.894095671241777,
+      "grad_norm": 5.264223037773271,
+      "learning_rate": 1.4563831439897647e-07,
+      "loss": 0.5353,
+      "step": 11009
+    },
+    {
+      "epoch": 0.8941768862178185,
+      "grad_norm": 5.540285793593693,
+      "learning_rate": 1.4541721867038532e-07,
+      "loss": 0.4353,
+      "step": 11010
+    },
+    {
+      "epoch": 0.8942581011938602,
+      "grad_norm": 21.662383239919144,
+      "learning_rate": 1.4519628586646073e-07,
+      "loss": 0.4418,
+      "step": 11011
+    },
+    {
+      "epoch": 0.8943393161699017,
+      "grad_norm": 7.463645051609318,
+      "learning_rate": 1.4497551600249044e-07,
+      "loss": 0.4228,
+      "step": 11012
+    },
+    {
+      "epoch": 0.8944205311459433,
+      "grad_norm": 28.77597330752949,
+      "learning_rate": 1.447549090937511e-07,
+      "loss": 0.5227,
+      "step": 11013
+    },
+    {
+      "epoch": 0.8945017461219849,
+      "grad_norm": 3.957758324457693,
+      "learning_rate": 1.4453446515550724e-07,
+      "loss": 0.4077,
+      "step": 11014
+    },
+    {
+      "epoch": 0.8945829610980265,
+      "grad_norm": 7.878758982064296,
+      "learning_rate": 1.4431418420301157e-07,
+      "loss": 0.5104,
+      "step": 11015
+    },
+    {
+      "epoch": 0.894664176074068,
+      "grad_norm": 5.62814185101261,
+      "learning_rate": 1.440940662515075e-07,
+      "loss": 0.6012,
+      "step": 11016
+    },
+    {
+      "epoch": 0.8947453910501096,
+      "grad_norm": 5.517645951231921,
+      "learning_rate": 1.4387411131622592e-07,
+      "loss": 0.4324,
+      "step": 11017
+    },
+    {
+      "epoch": 0.8948266060261513,
+      "grad_norm": 4.575883407647924,
+      "learning_rate": 1.4365431941238544e-07,
+      "loss": 0.4455,
+      "step": 11018
+    },
+    {
+      "epoch": 0.8949078210021928,
+      "grad_norm": 4.43419935645762,
+      "learning_rate": 1.434346905551956e-07,
+      "loss": 0.3927,
+      "step": 11019
+    },
+    {
+      "epoch": 0.8949890359782344,
+      "grad_norm": 5.32989054760345,
+      "learning_rate": 1.432152247598534e-07,
+      "loss": 0.456,
+      "step": 11020
+    },
+    {
+      "epoch": 0.895070250954276,
+      "grad_norm": 4.528118555469795,
+      "learning_rate": 1.4299592204154445e-07,
+      "loss": 0.4092,
+      "step": 11021
+    },
+    {
+      "epoch": 0.8951514659303176,
+      "grad_norm": 5.024689885658865,
+      "learning_rate": 1.4277678241544328e-07,
+      "loss": 0.3977,
+      "step": 11022
+    },
+    {
+      "epoch": 0.8952326809063591,
+      "grad_norm": 6.316835150350028,
+      "learning_rate": 1.4255780589671337e-07,
+      "loss": 0.4807,
+      "step": 11023
+    },
+    {
+      "epoch": 0.8953138958824007,
+      "grad_norm": 5.3616028545471845,
+      "learning_rate": 1.423389925005067e-07,
+      "loss": 0.4474,
+      "step": 11024
+    },
+    {
+      "epoch": 0.8953951108584423,
+      "grad_norm": 3.2813886037054543,
+      "learning_rate": 1.421203422419637e-07,
+      "loss": 0.488,
+      "step": 11025
+    },
+    {
+      "epoch": 0.8954763258344839,
+      "grad_norm": 8.181117262341452,
+      "learning_rate": 1.4190185513621473e-07,
+      "loss": 0.4516,
+      "step": 11026
+    },
+    {
+      "epoch": 0.8955575408105254,
+      "grad_norm": 5.938608632171009,
+      "learning_rate": 1.416835311983772e-07,
+      "loss": 0.486,
+      "step": 11027
+    },
+    {
+      "epoch": 0.895638755786567,
+      "grad_norm": 9.269220540893665,
+      "learning_rate": 1.4146537044355785e-07,
+      "loss": 0.4106,
+      "step": 11028
+    },
+    {
+      "epoch": 0.8957199707626087,
+      "grad_norm": 18.999226973217382,
+      "learning_rate": 1.412473728868527e-07,
+      "loss": 0.4675,
+      "step": 11029
+    },
+    {
+      "epoch": 0.8958011857386502,
+      "grad_norm": 3.9421317838103103,
+      "learning_rate": 1.410295385433455e-07,
+      "loss": 0.4947,
+      "step": 11030
+    },
+    {
+      "epoch": 0.8958824007146918,
+      "grad_norm": 4.337954956064237,
+      "learning_rate": 1.4081186742810948e-07,
+      "loss": 0.4467,
+      "step": 11031
+    },
+    {
+      "epoch": 0.8959636156907334,
+      "grad_norm": 5.626043047717378,
+      "learning_rate": 1.4059435955620704e-07,
+      "loss": 0.5096,
+      "step": 11032
+    },
+    {
+      "epoch": 0.896044830666775,
+      "grad_norm": 4.67727289541783,
+      "learning_rate": 1.403770149426878e-07,
+      "loss": 0.4278,
+      "step": 11033
+    },
+    {
+      "epoch": 0.8961260456428165,
+      "grad_norm": 10.738068197611934,
+      "learning_rate": 1.4015983360259055e-07,
+      "loss": 0.5335,
+      "step": 11034
+    },
+    {
+      "epoch": 0.8962072606188581,
+      "grad_norm": 4.8693579129133235,
+      "learning_rate": 1.3994281555094386e-07,
+      "loss": 0.3716,
+      "step": 11035
+    },
+    {
+      "epoch": 0.8962884755948997,
+      "grad_norm": 2.9347782845843136,
+      "learning_rate": 1.3972596080276402e-07,
+      "loss": 0.7287,
+      "step": 11036
+    },
+    {
+      "epoch": 0.8963696905709413,
+      "grad_norm": 4.942121031320022,
+      "learning_rate": 1.395092693730557e-07,
+      "loss": 0.3883,
+      "step": 11037
+    },
+    {
+      "epoch": 0.8964509055469828,
+      "grad_norm": 5.046562171098466,
+      "learning_rate": 1.3929274127681303e-07,
+      "loss": 0.4705,
+      "step": 11038
+    },
+    {
+      "epoch": 0.8965321205230244,
+      "grad_norm": 3.9477589888855906,
+      "learning_rate": 1.3907637652901957e-07,
+      "loss": 0.5613,
+      "step": 11039
+    },
+    {
+      "epoch": 0.8966133354990661,
+      "grad_norm": 4.192994245868831,
+      "learning_rate": 1.3886017514464555e-07,
+      "loss": 0.39,
+      "step": 11040
+    },
+    {
+      "epoch": 0.8966945504751076,
+      "grad_norm": 4.637113922674406,
+      "learning_rate": 1.3864413713865098e-07,
+      "loss": 0.4266,
+      "step": 11041
+    },
+    {
+      "epoch": 0.8967757654511492,
+      "grad_norm": 5.8016593697423975,
+      "learning_rate": 1.38428262525985e-07,
+      "loss": 0.5181,
+      "step": 11042
+    },
+    {
+      "epoch": 0.8968569804271908,
+      "grad_norm": 10.961817792927185,
+      "learning_rate": 1.3821255132158456e-07,
+      "loss": 0.4214,
+      "step": 11043
+    },
+    {
+      "epoch": 0.8969381954032324,
+      "grad_norm": 6.98872156389094,
+      "learning_rate": 1.3799700354037605e-07,
+      "loss": 0.3722,
+      "step": 11044
+    },
+    {
+      "epoch": 0.8970194103792739,
+      "grad_norm": 6.014898811863191,
+      "learning_rate": 1.3778161919727472e-07,
+      "loss": 0.4123,
+      "step": 11045
+    },
+    {
+      "epoch": 0.8971006253553155,
+      "grad_norm": 4.681509306562985,
+      "learning_rate": 1.3756639830718316e-07,
+      "loss": 0.4154,
+      "step": 11046
+    },
+    {
+      "epoch": 0.8971818403313571,
+      "grad_norm": 5.1158065640859425,
+      "learning_rate": 1.373513408849936e-07,
+      "loss": 0.4988,
+      "step": 11047
+    },
+    {
+      "epoch": 0.8972630553073987,
+      "grad_norm": 5.355479142787348,
+      "learning_rate": 1.3713644694558742e-07,
+      "loss": 0.447,
+      "step": 11048
+    },
+    {
+      "epoch": 0.8973442702834402,
+      "grad_norm": 5.532841373770518,
+      "learning_rate": 1.369217165038339e-07,
+      "loss": 0.4586,
+      "step": 11049
+    },
+    {
+      "epoch": 0.8974254852594818,
+      "grad_norm": 14.185804561707538,
+      "learning_rate": 1.367071495745906e-07,
+      "loss": 0.4633,
+      "step": 11050
+    },
+    {
+      "epoch": 0.8975067002355235,
+      "grad_norm": 5.325288616133615,
+      "learning_rate": 1.3649274617270531e-07,
+      "loss": 0.539,
+      "step": 11051
+    },
+    {
+      "epoch": 0.897587915211565,
+      "grad_norm": 3.940700522997248,
+      "learning_rate": 1.3627850631301344e-07,
+      "loss": 0.4839,
+      "step": 11052
+    },
+    {
+      "epoch": 0.8976691301876066,
+      "grad_norm": 4.124491341239392,
+      "learning_rate": 1.3606443001033864e-07,
+      "loss": 0.4044,
+      "step": 11053
+    },
+    {
+      "epoch": 0.8977503451636482,
+      "grad_norm": 8.104468950407128,
+      "learning_rate": 1.3585051727949494e-07,
+      "loss": 0.4023,
+      "step": 11054
+    },
+    {
+      "epoch": 0.8978315601396898,
+      "grad_norm": 9.103539674813327,
+      "learning_rate": 1.3563676813528325e-07,
+      "loss": 0.3839,
+      "step": 11055
+    },
+    {
+      "epoch": 0.8979127751157313,
+      "grad_norm": 4.227081794199914,
+      "learning_rate": 1.354231825924937e-07,
+      "loss": 0.5269,
+      "step": 11056
+    },
+    {
+      "epoch": 0.8979939900917729,
+      "grad_norm": 4.986647735979627,
+      "learning_rate": 1.3520976066590557e-07,
+      "loss": 0.3988,
+      "step": 11057
+    },
+    {
+      "epoch": 0.8980752050678145,
+      "grad_norm": 4.88191309684984,
+      "learning_rate": 1.3499650237028677e-07,
+      "loss": 0.3746,
+      "step": 11058
+    },
+    {
+      "epoch": 0.8981564200438561,
+      "grad_norm": 8.27159257192499,
+      "learning_rate": 1.3478340772039328e-07,
+      "loss": 0.5004,
+      "step": 11059
+    },
+    {
+      "epoch": 0.8982376350198976,
+      "grad_norm": 7.866676776482577,
+      "learning_rate": 1.3457047673097024e-07,
+      "loss": 0.3978,
+      "step": 11060
+    },
+    {
+      "epoch": 0.8983188499959393,
+      "grad_norm": 6.3754067681303805,
+      "learning_rate": 1.343577094167514e-07,
+      "loss": 0.4521,
+      "step": 11061
+    },
+    {
+      "epoch": 0.8984000649719809,
+      "grad_norm": 3.9619701694967744,
+      "learning_rate": 1.341451057924592e-07,
+      "loss": 0.545,
+      "step": 11062
+    },
+    {
+      "epoch": 0.8984812799480224,
+      "grad_norm": 4.982039565280355,
+      "learning_rate": 1.3393266587280434e-07,
+      "loss": 0.8769,
+      "step": 11063
+    },
+    {
+      "epoch": 0.898562494924064,
+      "grad_norm": 4.163415192247937,
+      "learning_rate": 1.3372038967248647e-07,
+      "loss": 0.5124,
+      "step": 11064
+    },
+    {
+      "epoch": 0.8986437099001056,
+      "grad_norm": 5.048906567605772,
+      "learning_rate": 1.335082772061949e-07,
+      "loss": 0.5187,
+      "step": 11065
+    },
+    {
+      "epoch": 0.8987249248761472,
+      "grad_norm": 6.207042181936884,
+      "learning_rate": 1.3329632848860545e-07,
+      "loss": 0.6352,
+      "step": 11066
+    },
+    {
+      "epoch": 0.8988061398521887,
+      "grad_norm": 7.073497034721812,
+      "learning_rate": 1.33084543534385e-07,
+      "loss": 0.5995,
+      "step": 11067
+    },
+    {
+      "epoch": 0.8988873548282303,
+      "grad_norm": 9.491444143304246,
+      "learning_rate": 1.3287292235818732e-07,
+      "loss": 0.4552,
+      "step": 11068
+    },
+    {
+      "epoch": 0.8989685698042719,
+      "grad_norm": 3.2830020992185185,
+      "learning_rate": 1.326614649746555e-07,
+      "loss": 0.3663,
+      "step": 11069
+    },
+    {
+      "epoch": 0.8990497847803135,
+      "grad_norm": 4.692919809966446,
+      "learning_rate": 1.324501713984211e-07,
+      "loss": 0.443,
+      "step": 11070
+    },
+    {
+      "epoch": 0.899130999756355,
+      "grad_norm": 6.01529901714284,
+      "learning_rate": 1.3223904164410494e-07,
+      "loss": 0.3483,
+      "step": 11071
+    },
+    {
+      "epoch": 0.8992122147323967,
+      "grad_norm": 8.122333020006561,
+      "learning_rate": 1.3202807572631564e-07,
+      "loss": 0.5003,
+      "step": 11072
+    },
+    {
+      "epoch": 0.8992934297084383,
+      "grad_norm": 8.111867294079145,
+      "learning_rate": 1.318172736596518e-07,
+      "loss": 0.4683,
+      "step": 11073
+    },
+    {
+      "epoch": 0.8993746446844798,
+      "grad_norm": 9.069724308334372,
+      "learning_rate": 1.3160663545869896e-07,
+      "loss": 0.4858,
+      "step": 11074
+    },
+    {
+      "epoch": 0.8994558596605214,
+      "grad_norm": 7.1277675224613075,
+      "learning_rate": 1.3139616113803238e-07,
+      "loss": 0.6833,
+      "step": 11075
+    },
+    {
+      "epoch": 0.899537074636563,
+      "grad_norm": 4.1361872264740285,
+      "learning_rate": 1.3118585071221546e-07,
+      "loss": 0.6296,
+      "step": 11076
+    },
+    {
+      "epoch": 0.8996182896126046,
+      "grad_norm": 4.28970175862567,
+      "learning_rate": 1.3097570419580096e-07,
+      "loss": 0.5692,
+      "step": 11077
+    },
+    {
+      "epoch": 0.8996995045886461,
+      "grad_norm": 8.454970827458345,
+      "learning_rate": 1.3076572160333007e-07,
+      "loss": 0.431,
+      "step": 11078
+    },
+    {
+      "epoch": 0.8997807195646877,
+      "grad_norm": 9.012725656230002,
+      "learning_rate": 1.3055590294933196e-07,
+      "loss": 0.5454,
+      "step": 11079
+    },
+    {
+      "epoch": 0.8998619345407293,
+      "grad_norm": 5.94575386452439,
+      "learning_rate": 1.303462482483256e-07,
+      "loss": 0.4561,
+      "step": 11080
+    },
+    {
+      "epoch": 0.8999431495167709,
+      "grad_norm": 5.789262579104675,
+      "learning_rate": 1.301367575148177e-07,
+      "loss": 0.4424,
+      "step": 11081
+    },
+    {
+      "epoch": 0.9000243644928124,
+      "grad_norm": 38.814186598162095,
+      "learning_rate": 1.299274307633036e-07,
+      "loss": 0.5148,
+      "step": 11082
+    },
+    {
+      "epoch": 0.9001055794688541,
+      "grad_norm": 9.66072321360139,
+      "learning_rate": 1.297182680082676e-07,
+      "loss": 0.5525,
+      "step": 11083
+    },
+    {
+      "epoch": 0.9001867944448957,
+      "grad_norm": 4.4257331245909555,
+      "learning_rate": 1.2950926926418362e-07,
+      "loss": 0.5021,
+      "step": 11084
+    },
+    {
+      "epoch": 0.9002680094209372,
+      "grad_norm": 3.0817044530551954,
+      "learning_rate": 1.2930043454551178e-07,
+      "loss": 0.468,
+      "step": 11085
+    },
+    {
+      "epoch": 0.9003492243969788,
+      "grad_norm": 5.56969653241233,
+      "learning_rate": 1.2909176386670385e-07,
+      "loss": 0.3354,
+      "step": 11086
+    },
+    {
+      "epoch": 0.9004304393730204,
+      "grad_norm": 4.504308093649935,
+      "learning_rate": 1.2888325724219775e-07,
+      "loss": 0.5716,
+      "step": 11087
+    },
+    {
+      "epoch": 0.900511654349062,
+      "grad_norm": 4.663416416080067,
+      "learning_rate": 1.2867491468642106e-07,
+      "loss": 0.4541,
+      "step": 11088
+    },
+    {
+      "epoch": 0.9005928693251035,
+      "grad_norm": 4.148002537213163,
+      "learning_rate": 1.2846673621379035e-07,
+      "loss": 0.514,
+      "step": 11089
+    },
+    {
+      "epoch": 0.9006740843011452,
+      "grad_norm": 4.589279907686427,
+      "learning_rate": 1.282587218387102e-07,
+      "loss": 0.5123,
+      "step": 11090
+    },
+    {
+      "epoch": 0.9007552992771867,
+      "grad_norm": 3.974192487774201,
+      "learning_rate": 1.2805087157557434e-07,
+      "loss": 0.5089,
+      "step": 11091
+    },
+    {
+      "epoch": 0.9008365142532283,
+      "grad_norm": 5.602158178066266,
+      "learning_rate": 1.2784318543876463e-07,
+      "loss": 0.4654,
+      "step": 11092
+    },
+    {
+      "epoch": 0.9009177292292698,
+      "grad_norm": 8.541653052285914,
+      "learning_rate": 1.276356634426526e-07,
+      "loss": 0.4957,
+      "step": 11093
+    },
+    {
+      "epoch": 0.9009989442053115,
+      "grad_norm": 3.690935781414827,
+      "learning_rate": 1.274283056015968e-07,
+      "loss": 0.6248,
+      "step": 11094
+    },
+    {
+      "epoch": 0.9010801591813531,
+      "grad_norm": 4.875190589469171,
+      "learning_rate": 1.272211119299452e-07,
+      "loss": 0.4652,
+      "step": 11095
+    },
+    {
+      "epoch": 0.9011613741573946,
+      "grad_norm": 5.525144142434985,
+      "learning_rate": 1.270140824420349e-07,
+      "loss": 0.4016,
+      "step": 11096
+    },
+    {
+      "epoch": 0.9012425891334362,
+      "grad_norm": 3.8356059525271387,
+      "learning_rate": 1.2680721715219168e-07,
+      "loss": 0.4344,
+      "step": 11097
+    },
+    {
+      "epoch": 0.9013238041094778,
+      "grad_norm": 4.807969276574533,
+      "learning_rate": 1.2660051607472885e-07,
+      "loss": 0.6449,
+      "step": 11098
+    },
+    {
+      "epoch": 0.9014050190855194,
+      "grad_norm": 5.802173378442104,
+      "learning_rate": 1.2639397922394963e-07,
+      "loss": 0.5645,
+      "step": 11099
+    },
+    {
+      "epoch": 0.9014862340615609,
+      "grad_norm": 6.895498171533333,
+      "learning_rate": 1.261876066141446e-07,
+      "loss": 0.499,
+      "step": 11100
+    },
+    {
+      "epoch": 0.9015674490376026,
+      "grad_norm": 7.515564655019646,
+      "learning_rate": 1.2598139825959393e-07,
+      "loss": 0.4524,
+      "step": 11101
+    },
+    {
+      "epoch": 0.9016486640136441,
+      "grad_norm": 10.60238190605972,
+      "learning_rate": 1.2577535417456599e-07,
+      "loss": 0.3594,
+      "step": 11102
+    },
+    {
+      "epoch": 0.9017298789896857,
+      "grad_norm": 5.744098794749828,
+      "learning_rate": 1.255694743733185e-07,
+      "loss": 0.3964,
+      "step": 11103
+    },
+    {
+      "epoch": 0.9018110939657272,
+      "grad_norm": 3.9046557359930865,
+      "learning_rate": 1.253637588700965e-07,
+      "loss": 0.4082,
+      "step": 11104
+    },
+    {
+      "epoch": 0.9018923089417689,
+      "grad_norm": 5.1997775870318765,
+      "learning_rate": 1.251582076791352e-07,
+      "loss": 0.5643,
+      "step": 11105
+    },
+    {
+      "epoch": 0.9019735239178105,
+      "grad_norm": 8.139019581385137,
+      "learning_rate": 1.2495282081465747e-07,
+      "loss": 0.5973,
+      "step": 11106
+    },
+    {
+      "epoch": 0.902054738893852,
+      "grad_norm": 6.919395171153576,
+      "learning_rate": 1.2474759829087413e-07,
+      "loss": 0.5088,
+      "step": 11107
+    },
+    {
+      "epoch": 0.9021359538698936,
+      "grad_norm": 7.322279551123576,
+      "learning_rate": 1.2454254012198657e-07,
+      "loss": 0.5294,
+      "step": 11108
+    },
+    {
+      "epoch": 0.9022171688459352,
+      "grad_norm": 4.725685255377581,
+      "learning_rate": 1.2433764632218293e-07,
+      "loss": 0.4372,
+      "step": 11109
+    },
+    {
+      "epoch": 0.9022983838219768,
+      "grad_norm": 14.72704680471624,
+      "learning_rate": 1.2413291690564154e-07,
+      "loss": 0.4079,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9023795987980183,
+      "grad_norm": 5.242560339794683,
+      "learning_rate": 1.239283518865278e-07,
+      "loss": 0.5958,
+      "step": 11111
+    },
+    {
+      "epoch": 0.90246081377406,
+      "grad_norm": 3.152784855668021,
+      "learning_rate": 1.2372395127899728e-07,
+      "loss": 0.4431,
+      "step": 11112
+    },
+    {
+      "epoch": 0.9025420287501015,
+      "grad_norm": 12.381085280329925,
+      "learning_rate": 1.2351971509719312e-07,
+      "loss": 0.3836,
+      "step": 11113
+    },
+    {
+      "epoch": 0.9026232437261431,
+      "grad_norm": 6.282293943148033,
+      "learning_rate": 1.233156433552471e-07,
+      "loss": 0.4327,
+      "step": 11114
+    },
+    {
+      "epoch": 0.9027044587021846,
+      "grad_norm": 9.80917559710566,
+      "learning_rate": 1.2311173606727982e-07,
+      "loss": 0.5872,
+      "step": 11115
+    },
+    {
+      "epoch": 0.9027856736782263,
+      "grad_norm": 4.868387447251448,
+      "learning_rate": 1.2290799324740144e-07,
+      "loss": 0.604,
+      "step": 11116
+    },
+    {
+      "epoch": 0.9028668886542679,
+      "grad_norm": 4.469666914330099,
+      "learning_rate": 1.2270441490970897e-07,
+      "loss": 0.4203,
+      "step": 11117
+    },
+    {
+      "epoch": 0.9029481036303094,
+      "grad_norm": 5.0365987552230935,
+      "learning_rate": 1.2250100106828978e-07,
+      "loss": 0.4397,
+      "step": 11118
+    },
+    {
+      "epoch": 0.903029318606351,
+      "grad_norm": 4.266209336176858,
+      "learning_rate": 1.222977517372184e-07,
+      "loss": 0.8365,
+      "step": 11119
+    },
+    {
+      "epoch": 0.9031105335823926,
+      "grad_norm": 5.161680402157924,
+      "learning_rate": 1.2209466693055867e-07,
+      "loss": 0.5762,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9031917485584342,
+      "grad_norm": 4.148007256784386,
+      "learning_rate": 1.2189174666236314e-07,
+      "loss": 0.4376,
+      "step": 11121
+    },
+    {
+      "epoch": 0.9032729635344757,
+      "grad_norm": 5.331795822073706,
+      "learning_rate": 1.2168899094667257e-07,
+      "loss": 0.6871,
+      "step": 11122
+    },
+    {
+      "epoch": 0.9033541785105174,
+      "grad_norm": 4.452121593101198,
+      "learning_rate": 1.2148639979751686e-07,
+      "loss": 0.4611,
+      "step": 11123
+    },
+    {
+      "epoch": 0.9034353934865589,
+      "grad_norm": 5.092994028700953,
+      "learning_rate": 1.212839732289145e-07,
+      "loss": 0.3484,
+      "step": 11124
+    },
+    {
+      "epoch": 0.9035166084626005,
+      "grad_norm": 4.802229400487032,
+      "learning_rate": 1.2108171125487177e-07,
+      "loss": 0.3757,
+      "step": 11125
+    },
+    {
+      "epoch": 0.903597823438642,
+      "grad_norm": 5.17753286975903,
+      "learning_rate": 1.2087961388938473e-07,
+      "loss": 0.5139,
+      "step": 11126
+    },
+    {
+      "epoch": 0.9036790384146837,
+      "grad_norm": 4.289775333983445,
+      "learning_rate": 1.2067768114643635e-07,
+      "loss": 0.5869,
+      "step": 11127
+    },
+    {
+      "epoch": 0.9037602533907253,
+      "grad_norm": 5.723264204379845,
+      "learning_rate": 1.2047591304000044e-07,
+      "loss": 0.4739,
+      "step": 11128
+    },
+    {
+      "epoch": 0.9038414683667668,
+      "grad_norm": 10.15998509588337,
+      "learning_rate": 1.2027430958403808e-07,
+      "loss": 0.4689,
+      "step": 11129
+    },
+    {
+      "epoch": 0.9039226833428085,
+      "grad_norm": 4.420651137514857,
+      "learning_rate": 1.2007287079249863e-07,
+      "loss": 0.4664,
+      "step": 11130
+    },
+    {
+      "epoch": 0.90400389831885,
+      "grad_norm": 4.274612039891857,
+      "learning_rate": 1.1987159667932124e-07,
+      "loss": 0.5313,
+      "step": 11131
+    },
+    {
+      "epoch": 0.9040851132948916,
+      "grad_norm": 7.759017785750078,
+      "learning_rate": 1.1967048725843256e-07,
+      "loss": 0.4139,
+      "step": 11132
+    },
+    {
+      "epoch": 0.9041663282709331,
+      "grad_norm": 6.646277040961836,
+      "learning_rate": 1.1946954254374838e-07,
+      "loss": 0.5634,
+      "step": 11133
+    },
+    {
+      "epoch": 0.9042475432469748,
+      "grad_norm": 4.609209324854486,
+      "learning_rate": 1.1926876254917314e-07,
+      "loss": 0.4933,
+      "step": 11134
+    },
+    {
+      "epoch": 0.9043287582230163,
+      "grad_norm": 4.565133981153573,
+      "learning_rate": 1.190681472885996e-07,
+      "loss": 0.4394,
+      "step": 11135
+    },
+    {
+      "epoch": 0.9044099731990579,
+      "grad_norm": 8.18111120557696,
+      "learning_rate": 1.188676967759092e-07,
+      "loss": 0.5565,
+      "step": 11136
+    },
+    {
+      "epoch": 0.9044911881750994,
+      "grad_norm": 5.013373635726462,
+      "learning_rate": 1.1866741102497275e-07,
+      "loss": 0.5724,
+      "step": 11137
+    },
+    {
+      "epoch": 0.9045724031511411,
+      "grad_norm": 4.362396063750319,
+      "learning_rate": 1.1846729004964835e-07,
+      "loss": 0.4643,
+      "step": 11138
+    },
+    {
+      "epoch": 0.9046536181271827,
+      "grad_norm": 5.980174069654119,
+      "learning_rate": 1.1826733386378297e-07,
+      "loss": 0.4188,
+      "step": 11139
+    },
+    {
+      "epoch": 0.9047348331032242,
+      "grad_norm": 4.951897983742022,
+      "learning_rate": 1.1806754248121333e-07,
+      "loss": 0.4468,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9048160480792659,
+      "grad_norm": 4.454466679854169,
+      "learning_rate": 1.1786791591576307e-07,
+      "loss": 0.5959,
+      "step": 11141
+    },
+    {
+      "epoch": 0.9048972630553074,
+      "grad_norm": 4.311779144133112,
+      "learning_rate": 1.176684541812459e-07,
+      "loss": 0.7119,
+      "step": 11142
+    },
+    {
+      "epoch": 0.904978478031349,
+      "grad_norm": 5.216361658357243,
+      "learning_rate": 1.174691572914638e-07,
+      "loss": 0.5982,
+      "step": 11143
+    },
+    {
+      "epoch": 0.9050596930073905,
+      "grad_norm": 4.598499041200991,
+      "learning_rate": 1.1727002526020631e-07,
+      "loss": 0.5352,
+      "step": 11144
+    },
+    {
+      "epoch": 0.9051409079834322,
+      "grad_norm": 3.192274398240333,
+      "learning_rate": 1.1707105810125297e-07,
+      "loss": 0.5892,
+      "step": 11145
+    },
+    {
+      "epoch": 0.9052221229594737,
+      "grad_norm": 11.971740880860915,
+      "learning_rate": 1.1687225582837052e-07,
+      "loss": 0.3351,
+      "step": 11146
+    },
+    {
+      "epoch": 0.9053033379355153,
+      "grad_norm": 8.973707498145792,
+      "learning_rate": 1.1667361845531578e-07,
+      "loss": 0.368,
+      "step": 11147
+    },
+    {
+      "epoch": 0.9053845529115568,
+      "grad_norm": 4.2462620886968825,
+      "learning_rate": 1.164751459958327e-07,
+      "loss": 0.3972,
+      "step": 11148
+    },
+    {
+      "epoch": 0.9054657678875985,
+      "grad_norm": 4.590797394683434,
+      "learning_rate": 1.1627683846365478e-07,
+      "loss": 0.548,
+      "step": 11149
+    },
+    {
+      "epoch": 0.9055469828636401,
+      "grad_norm": 7.826708434408468,
+      "learning_rate": 1.1607869587250464e-07,
+      "loss": 0.4099,
+      "step": 11150
+    },
+    {
+      "epoch": 0.9056281978396816,
+      "grad_norm": 4.643007677952196,
+      "learning_rate": 1.1588071823609159e-07,
+      "loss": 0.6002,
+      "step": 11151
+    },
+    {
+      "epoch": 0.9057094128157233,
+      "grad_norm": 6.195712206260984,
+      "learning_rate": 1.1568290556811495e-07,
+      "loss": 0.5826,
+      "step": 11152
+    },
+    {
+      "epoch": 0.9057906277917648,
+      "grad_norm": 5.225028473648337,
+      "learning_rate": 1.1548525788226267e-07,
+      "loss": 0.5314,
+      "step": 11153
+    },
+    {
+      "epoch": 0.9058718427678064,
+      "grad_norm": 9.685757941498554,
+      "learning_rate": 1.1528777519221046e-07,
+      "loss": 0.5205,
+      "step": 11154
+    },
+    {
+      "epoch": 0.9059530577438479,
+      "grad_norm": 3.9456543583410526,
+      "learning_rate": 1.1509045751162324e-07,
+      "loss": 0.4093,
+      "step": 11155
+    },
+    {
+      "epoch": 0.9060342727198896,
+      "grad_norm": 5.960099778891394,
+      "learning_rate": 1.1489330485415479e-07,
+      "loss": 0.3512,
+      "step": 11156
+    },
+    {
+      "epoch": 0.9061154876959311,
+      "grad_norm": 5.908029100963964,
+      "learning_rate": 1.1469631723344671e-07,
+      "loss": 0.4409,
+      "step": 11157
+    },
+    {
+      "epoch": 0.9061967026719727,
+      "grad_norm": 4.606484674584257,
+      "learning_rate": 1.1449949466312893e-07,
+      "loss": 0.4521,
+      "step": 11158
+    },
+    {
+      "epoch": 0.9062779176480142,
+      "grad_norm": 5.177826704703417,
+      "learning_rate": 1.1430283715682139e-07,
+      "loss": 0.5358,
+      "step": 11159
+    },
+    {
+      "epoch": 0.9063591326240559,
+      "grad_norm": 4.966905493065145,
+      "learning_rate": 1.1410634472813098e-07,
+      "loss": 0.4514,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9064403476000975,
+      "grad_norm": 5.793696903665879,
+      "learning_rate": 1.1391001739065432e-07,
+      "loss": 0.5214,
+      "step": 11161
+    },
+    {
+      "epoch": 0.906521562576139,
+      "grad_norm": 5.420921312113236,
+      "learning_rate": 1.1371385515797695e-07,
+      "loss": 0.4365,
+      "step": 11162
+    },
+    {
+      "epoch": 0.9066027775521807,
+      "grad_norm": 3.588449960902482,
+      "learning_rate": 1.1351785804367105e-07,
+      "loss": 0.5352,
+      "step": 11163
+    },
+    {
+      "epoch": 0.9066839925282222,
+      "grad_norm": 5.796738656501365,
+      "learning_rate": 1.1332202606129938e-07,
+      "loss": 0.4415,
+      "step": 11164
+    },
+    {
+      "epoch": 0.9067652075042638,
+      "grad_norm": 5.625515444116812,
+      "learning_rate": 1.1312635922441195e-07,
+      "loss": 0.4904,
+      "step": 11165
+    },
+    {
+      "epoch": 0.9068464224803053,
+      "grad_norm": 4.830094151436432,
+      "learning_rate": 1.129308575465482e-07,
+      "loss": 0.6403,
+      "step": 11166
+    },
+    {
+      "epoch": 0.906927637456347,
+      "grad_norm": 4.562587755650938,
+      "learning_rate": 1.1273552104123564e-07,
+      "loss": 0.4543,
+      "step": 11167
+    },
+    {
+      "epoch": 0.9070088524323885,
+      "grad_norm": 7.282911765724218,
+      "learning_rate": 1.125403497219904e-07,
+      "loss": 0.4899,
+      "step": 11168
+    },
+    {
+      "epoch": 0.9070900674084301,
+      "grad_norm": 4.877190705118263,
+      "learning_rate": 1.123453436023178e-07,
+      "loss": 0.3675,
+      "step": 11169
+    },
+    {
+      "epoch": 0.9071712823844716,
+      "grad_norm": 6.299432110607247,
+      "learning_rate": 1.121505026957112e-07,
+      "loss": 0.4837,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9072524973605133,
+      "grad_norm": 5.585090356105831,
+      "learning_rate": 1.1195582701565177e-07,
+      "loss": 0.5353,
+      "step": 11171
+    },
+    {
+      "epoch": 0.9073337123365549,
+      "grad_norm": 5.1062027169899435,
+      "learning_rate": 1.1176131657561095e-07,
+      "loss": 0.5091,
+      "step": 11172
+    },
+    {
+      "epoch": 0.9074149273125964,
+      "grad_norm": 3.970272507111287,
+      "learning_rate": 1.1156697138904715e-07,
+      "loss": 0.73,
+      "step": 11173
+    },
+    {
+      "epoch": 0.9074961422886381,
+      "grad_norm": 4.9753615625859196,
+      "learning_rate": 1.1137279146940821e-07,
+      "loss": 0.3436,
+      "step": 11174
+    },
+    {
+      "epoch": 0.9075773572646796,
+      "grad_norm": 4.2196215333931,
+      "learning_rate": 1.111787768301309e-07,
+      "loss": 0.5066,
+      "step": 11175
+    },
+    {
+      "epoch": 0.9076585722407212,
+      "grad_norm": 9.64461171338353,
+      "learning_rate": 1.1098492748463945e-07,
+      "loss": 0.4379,
+      "step": 11176
+    },
+    {
+      "epoch": 0.9077397872167627,
+      "grad_norm": 7.251627691507904,
+      "learning_rate": 1.1079124344634707e-07,
+      "loss": 0.4515,
+      "step": 11177
+    },
+    {
+      "epoch": 0.9078210021928044,
+      "grad_norm": 4.400388440896803,
+      "learning_rate": 1.1059772472865632e-07,
+      "loss": 0.408,
+      "step": 11178
+    },
+    {
+      "epoch": 0.9079022171688459,
+      "grad_norm": 4.694092152410516,
+      "learning_rate": 1.1040437134495708e-07,
+      "loss": 0.4494,
+      "step": 11179
+    },
+    {
+      "epoch": 0.9079834321448875,
+      "grad_norm": 4.510297602747996,
+      "learning_rate": 1.1021118330862835e-07,
+      "loss": 0.455,
+      "step": 11180
+    },
+    {
+      "epoch": 0.908064647120929,
+      "grad_norm": 3.8454716633958546,
+      "learning_rate": 1.1001816063303805e-07,
+      "loss": 0.5094,
+      "step": 11181
+    },
+    {
+      "epoch": 0.9081458620969707,
+      "grad_norm": 7.989031039920255,
+      "learning_rate": 1.0982530333154245e-07,
+      "loss": 0.6907,
+      "step": 11182
+    },
+    {
+      "epoch": 0.9082270770730123,
+      "grad_norm": 5.9615319524726695,
+      "learning_rate": 1.0963261141748616e-07,
+      "loss": 0.3874,
+      "step": 11183
+    },
+    {
+      "epoch": 0.9083082920490538,
+      "grad_norm": 4.85080597593589,
+      "learning_rate": 1.0944008490420183e-07,
+      "loss": 0.6927,
+      "step": 11184
+    },
+    {
+      "epoch": 0.9083895070250955,
+      "grad_norm": 12.043245022297707,
+      "learning_rate": 1.0924772380501215e-07,
+      "loss": 0.4814,
+      "step": 11185
+    },
+    {
+      "epoch": 0.908470722001137,
+      "grad_norm": 3.70058521293832,
+      "learning_rate": 1.0905552813322701e-07,
+      "loss": 0.43,
+      "step": 11186
+    },
+    {
+      "epoch": 0.9085519369771786,
+      "grad_norm": 5.677961162801236,
+      "learning_rate": 1.0886349790214495e-07,
+      "loss": 0.4644,
+      "step": 11187
+    },
+    {
+      "epoch": 0.9086331519532201,
+      "grad_norm": 5.958066806913473,
+      "learning_rate": 1.0867163312505452e-07,
+      "loss": 0.5697,
+      "step": 11188
+    },
+    {
+      "epoch": 0.9087143669292618,
+      "grad_norm": 8.122121992747614,
+      "learning_rate": 1.084799338152312e-07,
+      "loss": 0.4075,
+      "step": 11189
+    },
+    {
+      "epoch": 0.9087955819053033,
+      "grad_norm": 3.68707963935213,
+      "learning_rate": 1.082883999859391e-07,
+      "loss": 0.61,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9088767968813449,
+      "grad_norm": 6.469919999736764,
+      "learning_rate": 1.0809703165043206e-07,
+      "loss": 0.4208,
+      "step": 11191
+    },
+    {
+      "epoch": 0.9089580118573864,
+      "grad_norm": 13.337941491052447,
+      "learning_rate": 1.0790582882195172e-07,
+      "loss": 0.3323,
+      "step": 11192
+    },
+    {
+      "epoch": 0.9090392268334281,
+      "grad_norm": 5.027652600451383,
+      "learning_rate": 1.0771479151372749e-07,
+      "loss": 0.3798,
+      "step": 11193
+    },
+    {
+      "epoch": 0.9091204418094697,
+      "grad_norm": 19.702276711092185,
+      "learning_rate": 1.0752391973897852e-07,
+      "loss": 0.4353,
+      "step": 11194
+    },
+    {
+      "epoch": 0.9092016567855112,
+      "grad_norm": 6.736698974070218,
+      "learning_rate": 1.0733321351091286e-07,
+      "loss": 0.4836,
+      "step": 11195
+    },
+    {
+      "epoch": 0.9092828717615529,
+      "grad_norm": 8.31372301888567,
+      "learning_rate": 1.071426728427255e-07,
+      "loss": 0.4187,
+      "step": 11196
+    },
+    {
+      "epoch": 0.9093640867375944,
+      "grad_norm": 5.818025708926807,
+      "learning_rate": 1.0695229774760147e-07,
+      "loss": 0.4106,
+      "step": 11197
+    },
+    {
+      "epoch": 0.909445301713636,
+      "grad_norm": 4.4508095887650105,
+      "learning_rate": 1.0676208823871326e-07,
+      "loss": 0.545,
+      "step": 11198
+    },
+    {
+      "epoch": 0.9095265166896775,
+      "grad_norm": 8.123684613091736,
+      "learning_rate": 1.065720443292223e-07,
+      "loss": 0.3934,
+      "step": 11199
+    },
+    {
+      "epoch": 0.9096077316657192,
+      "grad_norm": 5.968329057688115,
+      "learning_rate": 1.0638216603227892e-07,
+      "loss": 0.5218,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9096889466417607,
+      "grad_norm": 8.367605133725334,
+      "learning_rate": 1.0619245336102174e-07,
+      "loss": 0.4462,
+      "step": 11201
+    },
+    {
+      "epoch": 0.9097701616178023,
+      "grad_norm": 7.432108062190521,
+      "learning_rate": 1.060029063285778e-07,
+      "loss": 0.5753,
+      "step": 11202
+    },
+    {
+      "epoch": 0.9098513765938439,
+      "grad_norm": 5.076717044768981,
+      "learning_rate": 1.0581352494806241e-07,
+      "loss": 0.3382,
+      "step": 11203
+    },
+    {
+      "epoch": 0.9099325915698855,
+      "grad_norm": 6.177656873247412,
+      "learning_rate": 1.0562430923258037e-07,
+      "loss": 0.5211,
+      "step": 11204
+    },
+    {
+      "epoch": 0.9100138065459271,
+      "grad_norm": 6.778475917464361,
+      "learning_rate": 1.0543525919522401e-07,
+      "loss": 0.4251,
+      "step": 11205
+    },
+    {
+      "epoch": 0.9100950215219686,
+      "grad_norm": 6.121707124021081,
+      "learning_rate": 1.0524637484907424e-07,
+      "loss": 0.3896,
+      "step": 11206
+    },
+    {
+      "epoch": 0.9101762364980103,
+      "grad_norm": 6.136173762658393,
+      "learning_rate": 1.0505765620720143e-07,
+      "loss": 0.6354,
+      "step": 11207
+    },
+    {
+      "epoch": 0.9102574514740518,
+      "grad_norm": 7.9613881203228525,
+      "learning_rate": 1.0486910328266403e-07,
+      "loss": 0.3988,
+      "step": 11208
+    },
+    {
+      "epoch": 0.9103386664500934,
+      "grad_norm": 5.222762969931984,
+      "learning_rate": 1.0468071608850827e-07,
+      "loss": 0.537,
+      "step": 11209
+    },
+    {
+      "epoch": 0.910419881426135,
+      "grad_norm": 4.889505356591718,
+      "learning_rate": 1.0449249463777039e-07,
+      "loss": 0.41,
+      "step": 11210
+    },
+    {
+      "epoch": 0.9105010964021766,
+      "grad_norm": 5.393977337842649,
+      "learning_rate": 1.0430443894347358e-07,
+      "loss": 0.8052,
+      "step": 11211
+    },
+    {
+      "epoch": 0.9105823113782181,
+      "grad_norm": 4.927676914766112,
+      "learning_rate": 1.041165490186305e-07,
+      "loss": 0.4754,
+      "step": 11212
+    },
+    {
+      "epoch": 0.9106635263542597,
+      "grad_norm": 4.738799420453396,
+      "learning_rate": 1.0392882487624212e-07,
+      "loss": 0.5263,
+      "step": 11213
+    },
+    {
+      "epoch": 0.9107447413303013,
+      "grad_norm": 3.4587373103230474,
+      "learning_rate": 1.0374126652929805e-07,
+      "loss": 0.3904,
+      "step": 11214
+    },
+    {
+      "epoch": 0.9108259563063429,
+      "grad_norm": 4.320689046060023,
+      "learning_rate": 1.0355387399077627e-07,
+      "loss": 0.5351,
+      "step": 11215
+    },
+    {
+      "epoch": 0.9109071712823845,
+      "grad_norm": 6.811312905006839,
+      "learning_rate": 1.033666472736436e-07,
+      "loss": 0.4286,
+      "step": 11216
+    },
+    {
+      "epoch": 0.910988386258426,
+      "grad_norm": 3.4936475176998845,
+      "learning_rate": 1.0317958639085524e-07,
+      "loss": 0.5945,
+      "step": 11217
+    },
+    {
+      "epoch": 0.9110696012344677,
+      "grad_norm": 4.327867349353077,
+      "learning_rate": 1.0299269135535416e-07,
+      "loss": 0.3791,
+      "step": 11218
+    },
+    {
+      "epoch": 0.9111508162105092,
+      "grad_norm": 7.91773166597089,
+      "learning_rate": 1.0280596218007254e-07,
+      "loss": 0.4511,
+      "step": 11219
+    },
+    {
+      "epoch": 0.9112320311865508,
+      "grad_norm": 5.982009386221902,
+      "learning_rate": 1.0261939887793143e-07,
+      "loss": 0.4298,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9113132461625923,
+      "grad_norm": 9.681222480637057,
+      "learning_rate": 1.0243300146184048e-07,
+      "loss": 0.5457,
+      "step": 11221
+    },
+    {
+      "epoch": 0.911394461138634,
+      "grad_norm": 8.645718538389323,
+      "learning_rate": 1.0224676994469635e-07,
+      "loss": 0.4311,
+      "step": 11222
+    },
+    {
+      "epoch": 0.9114756761146755,
+      "grad_norm": 6.818193527467812,
+      "learning_rate": 1.020607043393862e-07,
+      "loss": 0.4284,
+      "step": 11223
+    },
+    {
+      "epoch": 0.9115568910907171,
+      "grad_norm": 4.876516812885544,
+      "learning_rate": 1.0187480465878418e-07,
+      "loss": 0.4452,
+      "step": 11224
+    },
+    {
+      "epoch": 0.9116381060667587,
+      "grad_norm": 4.240926353476526,
+      "learning_rate": 1.0168907091575364e-07,
+      "loss": 0.5541,
+      "step": 11225
+    },
+    {
+      "epoch": 0.9117193210428003,
+      "grad_norm": 5.868350340007438,
+      "learning_rate": 1.015035031231465e-07,
+      "loss": 0.4558,
+      "step": 11226
+    },
+    {
+      "epoch": 0.9118005360188419,
+      "grad_norm": 3.823767663089306,
+      "learning_rate": 1.0131810129380332e-07,
+      "loss": 0.6506,
+      "step": 11227
+    },
+    {
+      "epoch": 0.9118817509948834,
+      "grad_norm": 7.147187550366974,
+      "learning_rate": 1.0113286544055245e-07,
+      "loss": 0.4312,
+      "step": 11228
+    },
+    {
+      "epoch": 0.9119629659709251,
+      "grad_norm": 50.85174427248752,
+      "learning_rate": 1.0094779557621171e-07,
+      "loss": 0.4037,
+      "step": 11229
+    },
+    {
+      "epoch": 0.9120441809469666,
+      "grad_norm": 10.89692418155859,
+      "learning_rate": 1.0076289171358695e-07,
+      "loss": 0.5095,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9121253959230082,
+      "grad_norm": 6.406674309261182,
+      "learning_rate": 1.0057815386547181e-07,
+      "loss": 0.6489,
+      "step": 11231
+    },
+    {
+      "epoch": 0.9122066108990498,
+      "grad_norm": 4.4242464166233955,
+      "learning_rate": 1.0039358204464943e-07,
+      "loss": 0.5091,
+      "step": 11232
+    },
+    {
+      "epoch": 0.9122878258750914,
+      "grad_norm": 4.743762192538487,
+      "learning_rate": 1.0020917626389209e-07,
+      "loss": 0.4469,
+      "step": 11233
+    },
+    {
+      "epoch": 0.9123690408511329,
+      "grad_norm": 4.14370710399411,
+      "learning_rate": 1.0002493653595902e-07,
+      "loss": 0.39,
+      "step": 11234
+    },
+    {
+      "epoch": 0.9124502558271745,
+      "grad_norm": 14.280052275876619,
+      "learning_rate": 9.984086287359806e-08,
+      "loss": 0.5105,
+      "step": 11235
+    },
+    {
+      "epoch": 0.9125314708032161,
+      "grad_norm": 6.244441893382094,
+      "learning_rate": 9.965695528954711e-08,
+      "loss": 0.357,
+      "step": 11236
+    },
+    {
+      "epoch": 0.9126126857792577,
+      "grad_norm": 7.898334173798384,
+      "learning_rate": 9.947321379653152e-08,
+      "loss": 0.4236,
+      "step": 11237
+    },
+    {
+      "epoch": 0.9126939007552993,
+      "grad_norm": 4.788515846324651,
+      "learning_rate": 9.928963840726418e-08,
+      "loss": 0.5023,
+      "step": 11238
+    },
+    {
+      "epoch": 0.9127751157313408,
+      "grad_norm": 25.04518107661059,
+      "learning_rate": 9.910622913444856e-08,
+      "loss": 0.5852,
+      "step": 11239
+    },
+    {
+      "epoch": 0.9128563307073825,
+      "grad_norm": 4.841830459052142,
+      "learning_rate": 9.89229859907756e-08,
+      "loss": 0.4615,
+      "step": 11240
+    },
+    {
+      "epoch": 0.912937545683424,
+      "grad_norm": 7.939078029749193,
+      "learning_rate": 9.873990898892405e-08,
+      "loss": 0.4802,
+      "step": 11241
+    },
+    {
+      "epoch": 0.9130187606594656,
+      "grad_norm": 9.609761146624544,
+      "learning_rate": 9.855699814156266e-08,
+      "loss": 0.4418,
+      "step": 11242
+    },
+    {
+      "epoch": 0.9130999756355072,
+      "grad_norm": 4.6087214575599145,
+      "learning_rate": 9.837425346134771e-08,
+      "loss": 0.5425,
+      "step": 11243
+    },
+    {
+      "epoch": 0.9131811906115488,
+      "grad_norm": 5.893925469137472,
+      "learning_rate": 9.819167496092352e-08,
+      "loss": 0.4984,
+      "step": 11244
+    },
+    {
+      "epoch": 0.9132624055875903,
+      "grad_norm": 5.730428639464196,
+      "learning_rate": 9.800926265292415e-08,
+      "loss": 0.4071,
+      "step": 11245
+    },
+    {
+      "epoch": 0.9133436205636319,
+      "grad_norm": 3.5844043953410853,
+      "learning_rate": 9.782701654997145e-08,
+      "loss": 0.5314,
+      "step": 11246
+    },
+    {
+      "epoch": 0.9134248355396735,
+      "grad_norm": 4.657341875734708,
+      "learning_rate": 9.764493666467589e-08,
+      "loss": 0.5415,
+      "step": 11247
+    },
+    {
+      "epoch": 0.9135060505157151,
+      "grad_norm": 5.362707479330835,
+      "learning_rate": 9.746302300963656e-08,
+      "loss": 0.5306,
+      "step": 11248
+    },
+    {
+      "epoch": 0.9135872654917567,
+      "grad_norm": 5.144669261464033,
+      "learning_rate": 9.728127559744089e-08,
+      "loss": 0.4642,
+      "step": 11249
+    },
+    {
+      "epoch": 0.9136684804677982,
+      "grad_norm": 8.491237111181297,
+      "learning_rate": 9.709969444066436e-08,
+      "loss": 0.4926,
+      "step": 11250
+    },
+    {
+      "epoch": 0.9137496954438399,
+      "grad_norm": 4.176442395340095,
+      "learning_rate": 9.691827955187222e-08,
+      "loss": 0.6502,
+      "step": 11251
+    },
+    {
+      "epoch": 0.9138309104198814,
+      "grad_norm": 9.913453531194941,
+      "learning_rate": 9.673703094361664e-08,
+      "loss": 0.609,
+      "step": 11252
+    },
+    {
+      "epoch": 0.913912125395923,
+      "grad_norm": 4.875302448335454,
+      "learning_rate": 9.655594862843953e-08,
+      "loss": 0.3666,
+      "step": 11253
+    },
+    {
+      "epoch": 0.9139933403719646,
+      "grad_norm": 4.516782702007909,
+      "learning_rate": 9.63750326188706e-08,
+      "loss": 0.4876,
+      "step": 11254
+    },
+    {
+      "epoch": 0.9140745553480062,
+      "grad_norm": 9.6025358191127,
+      "learning_rate": 9.619428292742872e-08,
+      "loss": 0.4328,
+      "step": 11255
+    },
+    {
+      "epoch": 0.9141557703240477,
+      "grad_norm": 6.319994193854443,
+      "learning_rate": 9.601369956662054e-08,
+      "loss": 0.5117,
+      "step": 11256
+    },
+    {
+      "epoch": 0.9142369853000893,
+      "grad_norm": 5.66199645711789,
+      "learning_rate": 9.583328254894109e-08,
+      "loss": 0.657,
+      "step": 11257
+    },
+    {
+      "epoch": 0.9143182002761309,
+      "grad_norm": 3.0411540989807317,
+      "learning_rate": 9.565303188687453e-08,
+      "loss": 0.6704,
+      "step": 11258
+    },
+    {
+      "epoch": 0.9143994152521725,
+      "grad_norm": 4.937602813989695,
+      "learning_rate": 9.547294759289366e-08,
+      "loss": 0.3464,
+      "step": 11259
+    },
+    {
+      "epoch": 0.9144806302282141,
+      "grad_norm": 4.5620361618138885,
+      "learning_rate": 9.52930296794588e-08,
+      "loss": 0.4903,
+      "step": 11260
+    },
+    {
+      "epoch": 0.9145618452042557,
+      "grad_norm": 4.986561186017303,
+      "learning_rate": 9.511327815902e-08,
+      "loss": 0.5204,
+      "step": 11261
+    },
+    {
+      "epoch": 0.9146430601802973,
+      "grad_norm": 3.2596608136193366,
+      "learning_rate": 9.493369304401423e-08,
+      "loss": 0.3049,
+      "step": 11262
+    },
+    {
+      "epoch": 0.9147242751563388,
+      "grad_norm": 3.798410117128246,
+      "learning_rate": 9.475427434686824e-08,
+      "loss": 0.5748,
+      "step": 11263
+    },
+    {
+      "epoch": 0.9148054901323804,
+      "grad_norm": 6.954981893846417,
+      "learning_rate": 9.457502207999736e-08,
+      "loss": 0.5748,
+      "step": 11264
+    },
+    {
+      "epoch": 0.914886705108422,
+      "grad_norm": 3.8501120004641143,
+      "learning_rate": 9.43959362558039e-08,
+      "loss": 0.525,
+      "step": 11265
+    },
+    {
+      "epoch": 0.9149679200844636,
+      "grad_norm": 4.333034848842079,
+      "learning_rate": 9.421701688668017e-08,
+      "loss": 0.4368,
+      "step": 11266
+    },
+    {
+      "epoch": 0.9150491350605051,
+      "grad_norm": 5.609179824260918,
+      "learning_rate": 9.403826398500654e-08,
+      "loss": 0.4947,
+      "step": 11267
+    },
+    {
+      "epoch": 0.9151303500365467,
+      "grad_norm": 4.516842058122665,
+      "learning_rate": 9.385967756315201e-08,
+      "loss": 0.5003,
+      "step": 11268
+    },
+    {
+      "epoch": 0.9152115650125883,
+      "grad_norm": 5.024303496834808,
+      "learning_rate": 9.368125763347336e-08,
+      "loss": 0.502,
+      "step": 11269
+    },
+    {
+      "epoch": 0.9152927799886299,
+      "grad_norm": 3.8890995639637858,
+      "learning_rate": 9.350300420831599e-08,
+      "loss": 0.4421,
+      "step": 11270
+    },
+    {
+      "epoch": 0.9153739949646715,
+      "grad_norm": 11.469385591277504,
+      "learning_rate": 9.332491730001448e-08,
+      "loss": 0.6776,
+      "step": 11271
+    },
+    {
+      "epoch": 0.915455209940713,
+      "grad_norm": 8.781322680809653,
+      "learning_rate": 9.314699692089202e-08,
+      "loss": 0.5468,
+      "step": 11272
+    },
+    {
+      "epoch": 0.9155364249167547,
+      "grad_norm": 5.122536771418473,
+      "learning_rate": 9.296924308325905e-08,
+      "loss": 0.5422,
+      "step": 11273
+    },
+    {
+      "epoch": 0.9156176398927962,
+      "grad_norm": 5.107338953835333,
+      "learning_rate": 9.279165579941546e-08,
+      "loss": 0.5195,
+      "step": 11274
+    },
+    {
+      "epoch": 0.9156988548688378,
+      "grad_norm": 7.603688810286707,
+      "learning_rate": 9.261423508164947e-08,
+      "loss": 0.4056,
+      "step": 11275
+    },
+    {
+      "epoch": 0.9157800698448794,
+      "grad_norm": 3.231334009016635,
+      "learning_rate": 9.243698094223735e-08,
+      "loss": 0.3926,
+      "step": 11276
+    },
+    {
+      "epoch": 0.915861284820921,
+      "grad_norm": 4.19277721624568,
+      "learning_rate": 9.225989339344432e-08,
+      "loss": 0.5653,
+      "step": 11277
+    },
+    {
+      "epoch": 0.9159424997969625,
+      "grad_norm": 4.727861994059561,
+      "learning_rate": 9.208297244752362e-08,
+      "loss": 0.6066,
+      "step": 11278
+    },
+    {
+      "epoch": 0.9160237147730041,
+      "grad_norm": 4.796597984182049,
+      "learning_rate": 9.190621811671769e-08,
+      "loss": 0.4402,
+      "step": 11279
+    },
+    {
+      "epoch": 0.9161049297490457,
+      "grad_norm": 17.756697315100613,
+      "learning_rate": 9.1729630413257e-08,
+      "loss": 0.4518,
+      "step": 11280
+    },
+    {
+      "epoch": 0.9161861447250873,
+      "grad_norm": 6.095353870840864,
+      "learning_rate": 9.155320934936041e-08,
+      "loss": 0.5031,
+      "step": 11281
+    },
+    {
+      "epoch": 0.9162673597011289,
+      "grad_norm": 6.9639308227226,
+      "learning_rate": 9.137695493723481e-08,
+      "loss": 0.4176,
+      "step": 11282
+    },
+    {
+      "epoch": 0.9163485746771705,
+      "grad_norm": 4.5604405756876485,
+      "learning_rate": 9.120086718907657e-08,
+      "loss": 0.4873,
+      "step": 11283
+    },
+    {
+      "epoch": 0.9164297896532121,
+      "grad_norm": 5.467910329584174,
+      "learning_rate": 9.10249461170698e-08,
+      "loss": 0.3554,
+      "step": 11284
+    },
+    {
+      "epoch": 0.9165110046292536,
+      "grad_norm": 5.007337212383075,
+      "learning_rate": 9.084919173338758e-08,
+      "loss": 0.499,
+      "step": 11285
+    },
+    {
+      "epoch": 0.9165922196052952,
+      "grad_norm": 9.584147277340062,
+      "learning_rate": 9.067360405019099e-08,
+      "loss": 0.4328,
+      "step": 11286
+    },
+    {
+      "epoch": 0.9166734345813368,
+      "grad_norm": 4.7954029212995355,
+      "learning_rate": 9.049818307963004e-08,
+      "loss": 0.4119,
+      "step": 11287
+    },
+    {
+      "epoch": 0.9167546495573784,
+      "grad_norm": 7.375196976362717,
+      "learning_rate": 9.03229288338428e-08,
+      "loss": 0.4901,
+      "step": 11288
+    },
+    {
+      "epoch": 0.9168358645334199,
+      "grad_norm": 5.606112983529189,
+      "learning_rate": 9.014784132495542e-08,
+      "loss": 0.5983,
+      "step": 11289
+    },
+    {
+      "epoch": 0.9169170795094616,
+      "grad_norm": 5.297505673484214,
+      "learning_rate": 8.997292056508372e-08,
+      "loss": 0.3459,
+      "step": 11290
+    },
+    {
+      "epoch": 0.9169982944855031,
+      "grad_norm": 9.227347098755619,
+      "learning_rate": 8.979816656633084e-08,
+      "loss": 0.4612,
+      "step": 11291
+    },
+    {
+      "epoch": 0.9170795094615447,
+      "grad_norm": 2.966819498823606,
+      "learning_rate": 8.962357934078874e-08,
+      "loss": 0.5246,
+      "step": 11292
+    },
+    {
+      "epoch": 0.9171607244375863,
+      "grad_norm": 6.824208201533333,
+      "learning_rate": 8.944915890053891e-08,
+      "loss": 0.4562,
+      "step": 11293
+    },
+    {
+      "epoch": 0.9172419394136279,
+      "grad_norm": 5.376195914843917,
+      "learning_rate": 8.927490525764942e-08,
+      "loss": 0.6399,
+      "step": 11294
+    },
+    {
+      "epoch": 0.9173231543896695,
+      "grad_norm": 4.79660630950142,
+      "learning_rate": 8.910081842417761e-08,
+      "loss": 0.5716,
+      "step": 11295
+    },
+    {
+      "epoch": 0.917404369365711,
+      "grad_norm": 4.642397495613151,
+      "learning_rate": 8.892689841216995e-08,
+      "loss": 0.5504,
+      "step": 11296
+    },
+    {
+      "epoch": 0.9174855843417526,
+      "grad_norm": 3.4372668067850913,
+      "learning_rate": 8.875314523366014e-08,
+      "loss": 0.5241,
+      "step": 11297
+    },
+    {
+      "epoch": 0.9175667993177942,
+      "grad_norm": 5.952139631988675,
+      "learning_rate": 8.857955890067132e-08,
+      "loss": 0.4511,
+      "step": 11298
+    },
+    {
+      "epoch": 0.9176480142938358,
+      "grad_norm": 6.811625003043075,
+      "learning_rate": 8.840613942521503e-08,
+      "loss": 0.4252,
+      "step": 11299
+    },
+    {
+      "epoch": 0.9177292292698773,
+      "grad_norm": 5.451600879549154,
+      "learning_rate": 8.823288681929082e-08,
+      "loss": 0.4875,
+      "step": 11300
+    },
+    {
+      "epoch": 0.917810444245919,
+      "grad_norm": 4.68192691825503,
+      "learning_rate": 8.80598010948866e-08,
+      "loss": 0.6457,
+      "step": 11301
+    },
+    {
+      "epoch": 0.9178916592219605,
+      "grad_norm": 4.012662107572786,
+      "learning_rate": 8.788688226397917e-08,
+      "loss": 0.4899,
+      "step": 11302
+    },
+    {
+      "epoch": 0.9179728741980021,
+      "grad_norm": 6.570602578436052,
+      "learning_rate": 8.771413033853343e-08,
+      "loss": 0.5515,
+      "step": 11303
+    },
+    {
+      "epoch": 0.9180540891740437,
+      "grad_norm": 7.278974590840044,
+      "learning_rate": 8.754154533050285e-08,
+      "loss": 0.3594,
+      "step": 11304
+    },
+    {
+      "epoch": 0.9181353041500853,
+      "grad_norm": 5.819321823620186,
+      "learning_rate": 8.736912725182983e-08,
+      "loss": 0.4738,
+      "step": 11305
+    },
+    {
+      "epoch": 0.9182165191261269,
+      "grad_norm": 4.523374291477625,
+      "learning_rate": 8.719687611444483e-08,
+      "loss": 0.3854,
+      "step": 11306
+    },
+    {
+      "epoch": 0.9182977341021684,
+      "grad_norm": 3.277263217530252,
+      "learning_rate": 8.702479193026608e-08,
+      "loss": 0.4129,
+      "step": 11307
+    },
+    {
+      "epoch": 0.91837894907821,
+      "grad_norm": 4.810473743314004,
+      "learning_rate": 8.68528747112013e-08,
+      "loss": 0.3563,
+      "step": 11308
+    },
+    {
+      "epoch": 0.9184601640542516,
+      "grad_norm": 4.361251259939714,
+      "learning_rate": 8.668112446914622e-08,
+      "loss": 0.4657,
+      "step": 11309
+    },
+    {
+      "epoch": 0.9185413790302932,
+      "grad_norm": 5.725837124786675,
+      "learning_rate": 8.650954121598471e-08,
+      "loss": 0.4554,
+      "step": 11310
+    },
+    {
+      "epoch": 0.9186225940063347,
+      "grad_norm": 7.948527051871499,
+      "learning_rate": 8.633812496358973e-08,
+      "loss": 0.4677,
+      "step": 11311
+    },
+    {
+      "epoch": 0.9187038089823764,
+      "grad_norm": 8.490193251568819,
+      "learning_rate": 8.616687572382293e-08,
+      "loss": 0.4757,
+      "step": 11312
+    },
+    {
+      "epoch": 0.9187850239584179,
+      "grad_norm": 10.643140675212129,
+      "learning_rate": 8.599579350853288e-08,
+      "loss": 0.5886,
+      "step": 11313
+    },
+    {
+      "epoch": 0.9188662389344595,
+      "grad_norm": 5.12172366794015,
+      "learning_rate": 8.582487832955788e-08,
+      "loss": 0.4511,
+      "step": 11314
+    },
+    {
+      "epoch": 0.9189474539105011,
+      "grad_norm": 4.267431188352919,
+      "learning_rate": 8.565413019872488e-08,
+      "loss": 0.4457,
+      "step": 11315
+    },
+    {
+      "epoch": 0.9190286688865427,
+      "grad_norm": 6.46937742703463,
+      "learning_rate": 8.548354912784801e-08,
+      "loss": 0.5193,
+      "step": 11316
+    },
+    {
+      "epoch": 0.9191098838625843,
+      "grad_norm": 10.178148390235016,
+      "learning_rate": 8.531313512873063e-08,
+      "loss": 0.4415,
+      "step": 11317
+    },
+    {
+      "epoch": 0.9191910988386258,
+      "grad_norm": 5.783583498365241,
+      "learning_rate": 8.514288821316524e-08,
+      "loss": 0.3627,
+      "step": 11318
+    },
+    {
+      "epoch": 0.9192723138146675,
+      "grad_norm": 4.717466038627756,
+      "learning_rate": 8.497280839293159e-08,
+      "loss": 0.4085,
+      "step": 11319
+    },
+    {
+      "epoch": 0.919353528790709,
+      "grad_norm": 4.386734491499693,
+      "learning_rate": 8.480289567979776e-08,
+      "loss": 0.5006,
+      "step": 11320
+    },
+    {
+      "epoch": 0.9194347437667506,
+      "grad_norm": 4.236143189731592,
+      "learning_rate": 8.463315008552158e-08,
+      "loss": 0.5219,
+      "step": 11321
+    },
+    {
+      "epoch": 0.9195159587427921,
+      "grad_norm": 6.377920595401447,
+      "learning_rate": 8.446357162184838e-08,
+      "loss": 0.5412,
+      "step": 11322
+    },
+    {
+      "epoch": 0.9195971737188338,
+      "grad_norm": 4.393949522713649,
+      "learning_rate": 8.429416030051179e-08,
+      "loss": 0.6075,
+      "step": 11323
+    },
+    {
+      "epoch": 0.9196783886948753,
+      "grad_norm": 3.2693096226160097,
+      "learning_rate": 8.412491613323415e-08,
+      "loss": 0.4949,
+      "step": 11324
+    },
+    {
+      "epoch": 0.9197596036709169,
+      "grad_norm": 5.690081161082877,
+      "learning_rate": 8.39558391317269e-08,
+      "loss": 0.333,
+      "step": 11325
+    },
+    {
+      "epoch": 0.9198408186469585,
+      "grad_norm": 4.591799853216551,
+      "learning_rate": 8.378692930768873e-08,
+      "loss": 0.5179,
+      "step": 11326
+    },
+    {
+      "epoch": 0.9199220336230001,
+      "grad_norm": 3.4431361012922017,
+      "learning_rate": 8.361818667280724e-08,
+      "loss": 0.429,
+      "step": 11327
+    },
+    {
+      "epoch": 0.9200032485990417,
+      "grad_norm": 7.374506756234914,
+      "learning_rate": 8.344961123875895e-08,
+      "loss": 0.4514,
+      "step": 11328
+    },
+    {
+      "epoch": 0.9200844635750832,
+      "grad_norm": 3.5074173112977345,
+      "learning_rate": 8.328120301720783e-08,
+      "loss": 0.5645,
+      "step": 11329
+    },
+    {
+      "epoch": 0.9201656785511249,
+      "grad_norm": 4.89262703046508,
+      "learning_rate": 8.311296201980734e-08,
+      "loss": 0.6032,
+      "step": 11330
+    },
+    {
+      "epoch": 0.9202468935271664,
+      "grad_norm": 3.5016861133769526,
+      "learning_rate": 8.294488825819875e-08,
+      "loss": 0.9195,
+      "step": 11331
+    },
+    {
+      "epoch": 0.920328108503208,
+      "grad_norm": 5.597223977106012,
+      "learning_rate": 8.277698174401189e-08,
+      "loss": 0.644,
+      "step": 11332
+    },
+    {
+      "epoch": 0.9204093234792495,
+      "grad_norm": 4.477628857500898,
+      "learning_rate": 8.260924248886471e-08,
+      "loss": 0.5054,
+      "step": 11333
+    },
+    {
+      "epoch": 0.9204905384552912,
+      "grad_norm": 5.7562752348647885,
+      "learning_rate": 8.244167050436402e-08,
+      "loss": 0.4682,
+      "step": 11334
+    },
+    {
+      "epoch": 0.9205717534313327,
+      "grad_norm": 5.880650217377458,
+      "learning_rate": 8.22742658021053e-08,
+      "loss": 0.6894,
+      "step": 11335
+    },
+    {
+      "epoch": 0.9206529684073743,
+      "grad_norm": 5.322917229821901,
+      "learning_rate": 8.210702839367146e-08,
+      "loss": 0.4346,
+      "step": 11336
+    },
+    {
+      "epoch": 0.920734183383416,
+      "grad_norm": 4.796792390299345,
+      "learning_rate": 8.193995829063467e-08,
+      "loss": 0.4673,
+      "step": 11337
+    },
+    {
+      "epoch": 0.9208153983594575,
+      "grad_norm": 8.87203052295575,
+      "learning_rate": 8.177305550455566e-08,
+      "loss": 0.444,
+      "step": 11338
+    },
+    {
+      "epoch": 0.9208966133354991,
+      "grad_norm": 4.841733687058009,
+      "learning_rate": 8.160632004698271e-08,
+      "loss": 0.4699,
+      "step": 11339
+    },
+    {
+      "epoch": 0.9209778283115406,
+      "grad_norm": 4.075654039554347,
+      "learning_rate": 8.143975192945325e-08,
+      "loss": 0.7124,
+      "step": 11340
+    },
+    {
+      "epoch": 0.9210590432875823,
+      "grad_norm": 4.698000621909273,
+      "learning_rate": 8.127335116349305e-08,
+      "loss": 0.4646,
+      "step": 11341
+    },
+    {
+      "epoch": 0.9211402582636238,
+      "grad_norm": 4.494572377526055,
+      "learning_rate": 8.110711776061597e-08,
+      "loss": 0.4942,
+      "step": 11342
+    },
+    {
+      "epoch": 0.9212214732396654,
+      "grad_norm": 7.279761243226819,
+      "learning_rate": 8.09410517323242e-08,
+      "loss": 0.4229,
+      "step": 11343
+    },
+    {
+      "epoch": 0.9213026882157069,
+      "grad_norm": 4.863071295173902,
+      "learning_rate": 8.077515309010936e-08,
+      "loss": 0.3941,
+      "step": 11344
+    },
+    {
+      "epoch": 0.9213839031917486,
+      "grad_norm": 5.892354030279855,
+      "learning_rate": 8.060942184545034e-08,
+      "loss": 0.5919,
+      "step": 11345
+    },
+    {
+      "epoch": 0.9214651181677901,
+      "grad_norm": 4.708709314273925,
+      "learning_rate": 8.044385800981464e-08,
+      "loss": 0.3786,
+      "step": 11346
+    },
+    {
+      "epoch": 0.9215463331438317,
+      "grad_norm": 8.147418646635652,
+      "learning_rate": 8.02784615946589e-08,
+      "loss": 0.5678,
+      "step": 11347
+    },
+    {
+      "epoch": 0.9216275481198734,
+      "grad_norm": 5.376975658563418,
+      "learning_rate": 8.011323261142734e-08,
+      "loss": 0.4418,
+      "step": 11348
+    },
+    {
+      "epoch": 0.9217087630959149,
+      "grad_norm": 5.665831071686028,
+      "learning_rate": 7.994817107155301e-08,
+      "loss": 0.4363,
+      "step": 11349
+    },
+    {
+      "epoch": 0.9217899780719565,
+      "grad_norm": 4.582354055457941,
+      "learning_rate": 7.978327698645705e-08,
+      "loss": 0.4622,
+      "step": 11350
+    },
+    {
+      "epoch": 0.921871193047998,
+      "grad_norm": 7.540612277290582,
+      "learning_rate": 7.961855036754978e-08,
+      "loss": 0.3925,
+      "step": 11351
+    },
+    {
+      "epoch": 0.9219524080240397,
+      "grad_norm": 4.845886462847071,
+      "learning_rate": 7.945399122622904e-08,
+      "loss": 0.494,
+      "step": 11352
+    },
+    {
+      "epoch": 0.9220336230000812,
+      "grad_norm": 3.6709111199072013,
+      "learning_rate": 7.928959957388154e-08,
+      "loss": 0.625,
+      "step": 11353
+    },
+    {
+      "epoch": 0.9221148379761228,
+      "grad_norm": 4.457865490095833,
+      "learning_rate": 7.912537542188264e-08,
+      "loss": 0.4271,
+      "step": 11354
+    },
+    {
+      "epoch": 0.9221960529521643,
+      "grad_norm": 6.9852450558056365,
+      "learning_rate": 7.89613187815949e-08,
+      "loss": 0.5399,
+      "step": 11355
+    },
+    {
+      "epoch": 0.922277267928206,
+      "grad_norm": 5.777646011684341,
+      "learning_rate": 7.879742966437092e-08,
+      "loss": 0.4892,
+      "step": 11356
+    },
+    {
+      "epoch": 0.9223584829042475,
+      "grad_norm": 3.543496182377441,
+      "learning_rate": 7.86337080815508e-08,
+      "loss": 0.5512,
+      "step": 11357
+    },
+    {
+      "epoch": 0.9224396978802891,
+      "grad_norm": 4.258648901552076,
+      "learning_rate": 7.847015404446352e-08,
+      "loss": 0.3128,
+      "step": 11358
+    },
+    {
+      "epoch": 0.9225209128563308,
+      "grad_norm": 5.717748386729547,
+      "learning_rate": 7.830676756442529e-08,
+      "loss": 0.5077,
+      "step": 11359
+    },
+    {
+      "epoch": 0.9226021278323723,
+      "grad_norm": 5.465873996372227,
+      "learning_rate": 7.814354865274237e-08,
+      "loss": 0.4398,
+      "step": 11360
+    },
+    {
+      "epoch": 0.9226833428084139,
+      "grad_norm": 9.321969343150906,
+      "learning_rate": 7.798049732070822e-08,
+      "loss": 0.5184,
+      "step": 11361
+    },
+    {
+      "epoch": 0.9227645577844554,
+      "grad_norm": 6.551159045640905,
+      "learning_rate": 7.78176135796052e-08,
+      "loss": 0.4808,
+      "step": 11362
+    },
+    {
+      "epoch": 0.9228457727604971,
+      "grad_norm": 6.18279810098398,
+      "learning_rate": 7.765489744070459e-08,
+      "loss": 0.495,
+      "step": 11363
+    },
+    {
+      "epoch": 0.9229269877365386,
+      "grad_norm": 4.489976108293694,
+      "learning_rate": 7.749234891526486e-08,
+      "loss": 0.5977,
+      "step": 11364
+    },
+    {
+      "epoch": 0.9230082027125802,
+      "grad_norm": 4.0322329291396475,
+      "learning_rate": 7.732996801453313e-08,
+      "loss": 0.7121,
+      "step": 11365
+    },
+    {
+      "epoch": 0.9230894176886217,
+      "grad_norm": 7.911313584451032,
+      "learning_rate": 7.716775474974625e-08,
+      "loss": 0.3647,
+      "step": 11366
+    },
+    {
+      "epoch": 0.9231706326646634,
+      "grad_norm": 4.246785998863406,
+      "learning_rate": 7.70057091321283e-08,
+      "loss": 0.3889,
+      "step": 11367
+    },
+    {
+      "epoch": 0.9232518476407049,
+      "grad_norm": 21.982620189789714,
+      "learning_rate": 7.684383117289141e-08,
+      "loss": 0.5681,
+      "step": 11368
+    },
+    {
+      "epoch": 0.9233330626167465,
+      "grad_norm": 16.563097433915587,
+      "learning_rate": 7.66821208832369e-08,
+      "loss": 0.4741,
+      "step": 11369
+    },
+    {
+      "epoch": 0.9234142775927882,
+      "grad_norm": 4.738887055673006,
+      "learning_rate": 7.652057827435444e-08,
+      "loss": 0.5937,
+      "step": 11370
+    },
+    {
+      "epoch": 0.9234954925688297,
+      "grad_norm": 5.91709828726191,
+      "learning_rate": 7.635920335742203e-08,
+      "loss": 0.458,
+      "step": 11371
+    },
+    {
+      "epoch": 0.9235767075448713,
+      "grad_norm": 4.025266759061047,
+      "learning_rate": 7.619799614360573e-08,
+      "loss": 0.5648,
+      "step": 11372
+    },
+    {
+      "epoch": 0.9236579225209128,
+      "grad_norm": 5.402057661799795,
+      "learning_rate": 7.603695664406053e-08,
+      "loss": 0.4054,
+      "step": 11373
+    },
+    {
+      "epoch": 0.9237391374969545,
+      "grad_norm": 5.575279520689622,
+      "learning_rate": 7.587608486992915e-08,
+      "loss": 0.4741,
+      "step": 11374
+    },
+    {
+      "epoch": 0.923820352472996,
+      "grad_norm": 3.7480482517554115,
+      "learning_rate": 7.571538083234298e-08,
+      "loss": 0.4311,
+      "step": 11375
+    },
+    {
+      "epoch": 0.9239015674490376,
+      "grad_norm": 3.873434245465527,
+      "learning_rate": 7.555484454242229e-08,
+      "loss": 0.5094,
+      "step": 11376
+    },
+    {
+      "epoch": 0.9239827824250791,
+      "grad_norm": 4.404320153846423,
+      "learning_rate": 7.539447601127542e-08,
+      "loss": 0.4184,
+      "step": 11377
+    },
+    {
+      "epoch": 0.9240639974011208,
+      "grad_norm": 4.253052401961252,
+      "learning_rate": 7.523427524999822e-08,
+      "loss": 0.4376,
+      "step": 11378
+    },
+    {
+      "epoch": 0.9241452123771623,
+      "grad_norm": 7.182394888454542,
+      "learning_rate": 7.507424226967681e-08,
+      "loss": 0.4695,
+      "step": 11379
+    },
+    {
+      "epoch": 0.9242264273532039,
+      "grad_norm": 6.303359290206109,
+      "learning_rate": 7.491437708138372e-08,
+      "loss": 0.5214,
+      "step": 11380
+    },
+    {
+      "epoch": 0.9243076423292456,
+      "grad_norm": 4.296067574108232,
+      "learning_rate": 7.475467969618122e-08,
+      "loss": 0.4966,
+      "step": 11381
+    },
+    {
+      "epoch": 0.9243888573052871,
+      "grad_norm": 3.9154299069883187,
+      "learning_rate": 7.459515012511937e-08,
+      "loss": 0.6921,
+      "step": 11382
+    },
+    {
+      "epoch": 0.9244700722813287,
+      "grad_norm": 3.5610883326964538,
+      "learning_rate": 7.443578837923709e-08,
+      "loss": 0.4532,
+      "step": 11383
+    },
+    {
+      "epoch": 0.9245512872573702,
+      "grad_norm": 4.0069328169869625,
+      "learning_rate": 7.427659446956087e-08,
+      "loss": 0.6126,
+      "step": 11384
+    },
+    {
+      "epoch": 0.9246325022334119,
+      "grad_norm": 3.4747848677201554,
+      "learning_rate": 7.41175684071066e-08,
+      "loss": 0.5281,
+      "step": 11385
+    },
+    {
+      "epoch": 0.9247137172094534,
+      "grad_norm": 4.640308832890803,
+      "learning_rate": 7.39587102028777e-08,
+      "loss": 0.448,
+      "step": 11386
+    },
+    {
+      "epoch": 0.924794932185495,
+      "grad_norm": 4.482953706151723,
+      "learning_rate": 7.38000198678665e-08,
+      "loss": 0.5122,
+      "step": 11387
+    },
+    {
+      "epoch": 0.9248761471615365,
+      "grad_norm": 7.944047198700474,
+      "learning_rate": 7.36414974130531e-08,
+      "loss": 0.4532,
+      "step": 11388
+    },
+    {
+      "epoch": 0.9249573621375782,
+      "grad_norm": 3.2439380546377583,
+      "learning_rate": 7.348314284940706e-08,
+      "loss": 0.4904,
+      "step": 11389
+    },
+    {
+      "epoch": 0.9250385771136197,
+      "grad_norm": 4.141858638205531,
+      "learning_rate": 7.332495618788516e-08,
+      "loss": 0.5272,
+      "step": 11390
+    },
+    {
+      "epoch": 0.9251197920896613,
+      "grad_norm": 3.3315239217657773,
+      "learning_rate": 7.316693743943364e-08,
+      "loss": 0.4314,
+      "step": 11391
+    },
+    {
+      "epoch": 0.925201007065703,
+      "grad_norm": 3.9642024980963573,
+      "learning_rate": 7.300908661498602e-08,
+      "loss": 0.545,
+      "step": 11392
+    },
+    {
+      "epoch": 0.9252822220417445,
+      "grad_norm": 6.271707028230578,
+      "learning_rate": 7.28514037254649e-08,
+      "loss": 0.4979,
+      "step": 11393
+    },
+    {
+      "epoch": 0.9253634370177861,
+      "grad_norm": 6.887788740548668,
+      "learning_rate": 7.26938887817813e-08,
+      "loss": 0.5243,
+      "step": 11394
+    },
+    {
+      "epoch": 0.9254446519938276,
+      "grad_norm": 4.351861411172992,
+      "learning_rate": 7.2536541794834e-08,
+      "loss": 0.563,
+      "step": 11395
+    },
+    {
+      "epoch": 0.9255258669698693,
+      "grad_norm": 13.306179850830969,
+      "learning_rate": 7.237936277551095e-08,
+      "loss": 0.4579,
+      "step": 11396
+    },
+    {
+      "epoch": 0.9256070819459108,
+      "grad_norm": 8.741284990101086,
+      "learning_rate": 7.22223517346879e-08,
+      "loss": 0.4368,
+      "step": 11397
+    },
+    {
+      "epoch": 0.9256882969219524,
+      "grad_norm": 4.724782847856221,
+      "learning_rate": 7.206550868322947e-08,
+      "loss": 0.5499,
+      "step": 11398
+    },
+    {
+      "epoch": 0.925769511897994,
+      "grad_norm": 4.443264413967751,
+      "learning_rate": 7.190883363198815e-08,
+      "loss": 0.4174,
+      "step": 11399
+    },
+    {
+      "epoch": 0.9258507268740356,
+      "grad_norm": 10.053412856400966,
+      "learning_rate": 7.175232659180492e-08,
+      "loss": 0.6192,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9259319418500771,
+      "grad_norm": 8.327053855258349,
+      "learning_rate": 7.159598757350922e-08,
+      "loss": 0.4117,
+      "step": 11401
+    },
+    {
+      "epoch": 0.9260131568261187,
+      "grad_norm": 4.431594842283679,
+      "learning_rate": 7.143981658791933e-08,
+      "loss": 0.4699,
+      "step": 11402
+    },
+    {
+      "epoch": 0.9260943718021604,
+      "grad_norm": 3.250722005433043,
+      "learning_rate": 7.128381364584075e-08,
+      "loss": 0.5342,
+      "step": 11403
+    },
+    {
+      "epoch": 0.9261755867782019,
+      "grad_norm": 2.957886228743401,
+      "learning_rate": 7.112797875806904e-08,
+      "loss": 0.4175,
+      "step": 11404
+    },
+    {
+      "epoch": 0.9262568017542435,
+      "grad_norm": 6.59101582446327,
+      "learning_rate": 7.09723119353864e-08,
+      "loss": 0.4268,
+      "step": 11405
+    },
+    {
+      "epoch": 0.926338016730285,
+      "grad_norm": 4.053080793203663,
+      "learning_rate": 7.081681318856392e-08,
+      "loss": 0.4142,
+      "step": 11406
+    },
+    {
+      "epoch": 0.9264192317063267,
+      "grad_norm": 48.79556971471609,
+      "learning_rate": 7.066148252836219e-08,
+      "loss": 0.4385,
+      "step": 11407
+    },
+    {
+      "epoch": 0.9265004466823682,
+      "grad_norm": 4.586752487959576,
+      "learning_rate": 7.050631996552842e-08,
+      "loss": 0.4813,
+      "step": 11408
+    },
+    {
+      "epoch": 0.9265816616584098,
+      "grad_norm": 5.714365886095958,
+      "learning_rate": 7.035132551079932e-08,
+      "loss": 0.3704,
+      "step": 11409
+    },
+    {
+      "epoch": 0.9266628766344513,
+      "grad_norm": 8.579523295772237,
+      "learning_rate": 7.019649917490018e-08,
+      "loss": 0.3965,
+      "step": 11410
+    },
+    {
+      "epoch": 0.926744091610493,
+      "grad_norm": 3.0236689032912842,
+      "learning_rate": 7.004184096854356e-08,
+      "loss": 0.572,
+      "step": 11411
+    },
+    {
+      "epoch": 0.9268253065865345,
+      "grad_norm": 6.812565797682661,
+      "learning_rate": 6.988735090243142e-08,
+      "loss": 0.4516,
+      "step": 11412
+    },
+    {
+      "epoch": 0.9269065215625761,
+      "grad_norm": 7.700307129522538,
+      "learning_rate": 6.973302898725303e-08,
+      "loss": 0.4734,
+      "step": 11413
+    },
+    {
+      "epoch": 0.9269877365386178,
+      "grad_norm": 4.263009068805309,
+      "learning_rate": 6.957887523368678e-08,
+      "loss": 0.4478,
+      "step": 11414
+    },
+    {
+      "epoch": 0.9270689515146593,
+      "grad_norm": 5.759822571490719,
+      "learning_rate": 6.942488965240024e-08,
+      "loss": 0.4749,
+      "step": 11415
+    },
+    {
+      "epoch": 0.9271501664907009,
+      "grad_norm": 7.781687526540662,
+      "learning_rate": 6.92710722540471e-08,
+      "loss": 0.398,
+      "step": 11416
+    },
+    {
+      "epoch": 0.9272313814667424,
+      "grad_norm": 6.317228165360641,
+      "learning_rate": 6.911742304927166e-08,
+      "loss": 0.3865,
+      "step": 11417
+    },
+    {
+      "epoch": 0.9273125964427841,
+      "grad_norm": 6.2711857470132095,
+      "learning_rate": 6.896394204870538e-08,
+      "loss": 0.4675,
+      "step": 11418
+    },
+    {
+      "epoch": 0.9273938114188256,
+      "grad_norm": 4.60041184087192,
+      "learning_rate": 6.881062926296783e-08,
+      "loss": 0.5341,
+      "step": 11419
+    },
+    {
+      "epoch": 0.9274750263948672,
+      "grad_norm": 5.061233791327525,
+      "learning_rate": 6.865748470266803e-08,
+      "loss": 0.5314,
+      "step": 11420
+    },
+    {
+      "epoch": 0.9275562413709088,
+      "grad_norm": 3.7105712181882007,
+      "learning_rate": 6.85045083784025e-08,
+      "loss": 0.3024,
+      "step": 11421
+    },
+    {
+      "epoch": 0.9276374563469504,
+      "grad_norm": 12.81065039260799,
+      "learning_rate": 6.835170030075638e-08,
+      "loss": 0.4527,
+      "step": 11422
+    },
+    {
+      "epoch": 0.9277186713229919,
+      "grad_norm": 6.058599141728342,
+      "learning_rate": 6.819906048030345e-08,
+      "loss": 0.4631,
+      "step": 11423
+    },
+    {
+      "epoch": 0.9277998862990335,
+      "grad_norm": 6.482547478325477,
+      "learning_rate": 6.804658892760552e-08,
+      "loss": 0.4521,
+      "step": 11424
+    },
+    {
+      "epoch": 0.9278811012750752,
+      "grad_norm": 14.54413960358652,
+      "learning_rate": 6.789428565321249e-08,
+      "loss": 0.3769,
+      "step": 11425
+    },
+    {
+      "epoch": 0.9279623162511167,
+      "grad_norm": 5.183303848496903,
+      "learning_rate": 6.774215066766344e-08,
+      "loss": 0.535,
+      "step": 11426
+    },
+    {
+      "epoch": 0.9280435312271583,
+      "grad_norm": 4.383546792589708,
+      "learning_rate": 6.759018398148464e-08,
+      "loss": 0.3777,
+      "step": 11427
+    },
+    {
+      "epoch": 0.9281247462031998,
+      "grad_norm": 14.039870214824262,
+      "learning_rate": 6.743838560519189e-08,
+      "loss": 0.508,
+      "step": 11428
+    },
+    {
+      "epoch": 0.9282059611792415,
+      "grad_norm": 6.608844886900423,
+      "learning_rate": 6.728675554928898e-08,
+      "loss": 0.3648,
+      "step": 11429
+    },
+    {
+      "epoch": 0.928287176155283,
+      "grad_norm": 4.663438602760716,
+      "learning_rate": 6.713529382426726e-08,
+      "loss": 0.3972,
+      "step": 11430
+    },
+    {
+      "epoch": 0.9283683911313246,
+      "grad_norm": 3.8710006305176226,
+      "learning_rate": 6.698400044060777e-08,
+      "loss": 0.4688,
+      "step": 11431
+    },
+    {
+      "epoch": 0.9284496061073662,
+      "grad_norm": 3.7449549283487324,
+      "learning_rate": 6.683287540877853e-08,
+      "loss": 0.539,
+      "step": 11432
+    },
+    {
+      "epoch": 0.9285308210834078,
+      "grad_norm": 4.662838653734012,
+      "learning_rate": 6.668191873923701e-08,
+      "loss": 0.5233,
+      "step": 11433
+    },
+    {
+      "epoch": 0.9286120360594493,
+      "grad_norm": 4.3917574513396955,
+      "learning_rate": 6.653113044242904e-08,
+      "loss": 0.5587,
+      "step": 11434
+    },
+    {
+      "epoch": 0.9286932510354909,
+      "grad_norm": 9.052354440821007,
+      "learning_rate": 6.638051052878736e-08,
+      "loss": 0.5373,
+      "step": 11435
+    },
+    {
+      "epoch": 0.9287744660115326,
+      "grad_norm": 3.219724415933315,
+      "learning_rate": 6.623005900873474e-08,
+      "loss": 0.4253,
+      "step": 11436
+    },
+    {
+      "epoch": 0.9288556809875741,
+      "grad_norm": 3.5215886145233584,
+      "learning_rate": 6.607977589268177e-08,
+      "loss": 0.447,
+      "step": 11437
+    },
+    {
+      "epoch": 0.9289368959636157,
+      "grad_norm": 6.121975994953795,
+      "learning_rate": 6.59296611910265e-08,
+      "loss": 0.4416,
+      "step": 11438
+    },
+    {
+      "epoch": 0.9290181109396572,
+      "grad_norm": 7.404459895688475,
+      "learning_rate": 6.577971491415674e-08,
+      "loss": 0.4477,
+      "step": 11439
+    },
+    {
+      "epoch": 0.9290993259156989,
+      "grad_norm": 9.284787320280776,
+      "learning_rate": 6.56299370724478e-08,
+      "loss": 0.5049,
+      "step": 11440
+    },
+    {
+      "epoch": 0.9291805408917404,
+      "grad_norm": 7.115335886705074,
+      "learning_rate": 6.548032767626333e-08,
+      "loss": 0.4512,
+      "step": 11441
+    },
+    {
+      "epoch": 0.929261755867782,
+      "grad_norm": 6.530284682735083,
+      "learning_rate": 6.533088673595589e-08,
+      "loss": 0.3873,
+      "step": 11442
+    },
+    {
+      "epoch": 0.9293429708438236,
+      "grad_norm": 5.05350752987569,
+      "learning_rate": 6.51816142618658e-08,
+      "loss": 0.5982,
+      "step": 11443
+    },
+    {
+      "epoch": 0.9294241858198652,
+      "grad_norm": 8.402220447701428,
+      "learning_rate": 6.503251026432179e-08,
+      "loss": 0.517,
+      "step": 11444
+    },
+    {
+      "epoch": 0.9295054007959067,
+      "grad_norm": 5.712552445076794,
+      "learning_rate": 6.48835747536411e-08,
+      "loss": 0.4236,
+      "step": 11445
+    },
+    {
+      "epoch": 0.9295866157719483,
+      "grad_norm": 8.723406889128468,
+      "learning_rate": 6.473480774012941e-08,
+      "loss": 0.4614,
+      "step": 11446
+    },
+    {
+      "epoch": 0.92966783074799,
+      "grad_norm": 3.518539374576921,
+      "learning_rate": 6.458620923408044e-08,
+      "loss": 0.697,
+      "step": 11447
+    },
+    {
+      "epoch": 0.9297490457240315,
+      "grad_norm": 4.996661458005611,
+      "learning_rate": 6.443777924577676e-08,
+      "loss": 0.7767,
+      "step": 11448
+    },
+    {
+      "epoch": 0.9298302607000731,
+      "grad_norm": 6.029619825429544,
+      "learning_rate": 6.428951778548881e-08,
+      "loss": 0.5338,
+      "step": 11449
+    },
+    {
+      "epoch": 0.9299114756761147,
+      "grad_norm": 6.072175958874135,
+      "learning_rate": 6.414142486347557e-08,
+      "loss": 0.53,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9299926906521563,
+      "grad_norm": 4.106779337844846,
+      "learning_rate": 6.39935004899836e-08,
+      "loss": 0.5014,
+      "step": 11451
+    },
+    {
+      "epoch": 0.9300739056281978,
+      "grad_norm": 5.028928964661914,
+      "learning_rate": 6.38457446752494e-08,
+      "loss": 0.3896,
+      "step": 11452
+    },
+    {
+      "epoch": 0.9301551206042394,
+      "grad_norm": 3.916926797126131,
+      "learning_rate": 6.36981574294962e-08,
+      "loss": 0.4988,
+      "step": 11453
+    },
+    {
+      "epoch": 0.930236335580281,
+      "grad_norm": 11.573875343354155,
+      "learning_rate": 6.355073876293638e-08,
+      "loss": 0.4861,
+      "step": 11454
+    },
+    {
+      "epoch": 0.9303175505563226,
+      "grad_norm": 2.953512192237326,
+      "learning_rate": 6.340348868577123e-08,
+      "loss": 0.4556,
+      "step": 11455
+    },
+    {
+      "epoch": 0.9303987655323641,
+      "grad_norm": 5.593041199500861,
+      "learning_rate": 6.325640720818899e-08,
+      "loss": 0.4905,
+      "step": 11456
+    },
+    {
+      "epoch": 0.9304799805084057,
+      "grad_norm": 5.055425216937131,
+      "learning_rate": 6.310949434036707e-08,
+      "loss": 0.5173,
+      "step": 11457
+    },
+    {
+      "epoch": 0.9305611954844474,
+      "grad_norm": 5.280470361998423,
+      "learning_rate": 6.296275009247121e-08,
+      "loss": 0.4225,
+      "step": 11458
+    },
+    {
+      "epoch": 0.9306424104604889,
+      "grad_norm": 3.6832709176499785,
+      "learning_rate": 6.28161744746547e-08,
+      "loss": 0.453,
+      "step": 11459
+    },
+    {
+      "epoch": 0.9307236254365305,
+      "grad_norm": 7.653920651385478,
+      "learning_rate": 6.266976749706055e-08,
+      "loss": 0.4662,
+      "step": 11460
+    },
+    {
+      "epoch": 0.930804840412572,
+      "grad_norm": 5.873778228921937,
+      "learning_rate": 6.252352916981924e-08,
+      "loss": 0.4553,
+      "step": 11461
+    },
+    {
+      "epoch": 0.9308860553886137,
+      "grad_norm": 4.554092544444619,
+      "learning_rate": 6.237745950304963e-08,
+      "loss": 0.6166,
+      "step": 11462
+    },
+    {
+      "epoch": 0.9309672703646552,
+      "grad_norm": 5.080682795743635,
+      "learning_rate": 6.223155850685864e-08,
+      "loss": 0.5027,
+      "step": 11463
+    },
+    {
+      "epoch": 0.9310484853406968,
+      "grad_norm": 16.051986684845595,
+      "learning_rate": 6.208582619134234e-08,
+      "loss": 0.5546,
+      "step": 11464
+    },
+    {
+      "epoch": 0.9311297003167384,
+      "grad_norm": 7.21160759933066,
+      "learning_rate": 6.194026256658437e-08,
+      "loss": 0.412,
+      "step": 11465
+    },
+    {
+      "epoch": 0.93121091529278,
+      "grad_norm": 7.017201794326772,
+      "learning_rate": 6.179486764265663e-08,
+      "loss": 0.5635,
+      "step": 11466
+    },
+    {
+      "epoch": 0.9312921302688215,
+      "grad_norm": 6.919895696187786,
+      "learning_rate": 6.164964142962027e-08,
+      "loss": 0.4289,
+      "step": 11467
+    },
+    {
+      "epoch": 0.9313733452448631,
+      "grad_norm": 6.03620856801876,
+      "learning_rate": 6.15045839375239e-08,
+      "loss": 0.5504,
+      "step": 11468
+    },
+    {
+      "epoch": 0.9314545602209048,
+      "grad_norm": 6.279814421491574,
+      "learning_rate": 6.135969517640506e-08,
+      "loss": 0.3537,
+      "step": 11469
+    },
+    {
+      "epoch": 0.9315357751969463,
+      "grad_norm": 9.154160836609854,
+      "learning_rate": 6.12149751562885e-08,
+      "loss": 0.6001,
+      "step": 11470
+    },
+    {
+      "epoch": 0.9316169901729879,
+      "grad_norm": 6.53932524076683,
+      "learning_rate": 6.107042388718898e-08,
+      "loss": 0.3881,
+      "step": 11471
+    },
+    {
+      "epoch": 0.9316982051490295,
+      "grad_norm": 6.633007691995465,
+      "learning_rate": 6.092604137910768e-08,
+      "loss": 0.5065,
+      "step": 11472
+    },
+    {
+      "epoch": 0.9317794201250711,
+      "grad_norm": 5.199512015394908,
+      "learning_rate": 6.078182764203605e-08,
+      "loss": 0.5226,
+      "step": 11473
+    },
+    {
+      "epoch": 0.9318606351011126,
+      "grad_norm": 7.046681958143693,
+      "learning_rate": 6.063778268595278e-08,
+      "loss": 0.468,
+      "step": 11474
+    },
+    {
+      "epoch": 0.9319418500771542,
+      "grad_norm": 7.741942160235145,
+      "learning_rate": 6.04939065208246e-08,
+      "loss": 0.4611,
+      "step": 11475
+    },
+    {
+      "epoch": 0.9320230650531958,
+      "grad_norm": 6.7144122182681745,
+      "learning_rate": 6.035019915660717e-08,
+      "loss": 0.4898,
+      "step": 11476
+    },
+    {
+      "epoch": 0.9321042800292374,
+      "grad_norm": 4.677019846346351,
+      "learning_rate": 6.020666060324448e-08,
+      "loss": 0.4645,
+      "step": 11477
+    },
+    {
+      "epoch": 0.9321854950052789,
+      "grad_norm": 4.133353885157457,
+      "learning_rate": 6.006329087066831e-08,
+      "loss": 0.36,
+      "step": 11478
+    },
+    {
+      "epoch": 0.9322667099813206,
+      "grad_norm": 6.172109607468845,
+      "learning_rate": 5.992008996879906e-08,
+      "loss": 0.4392,
+      "step": 11479
+    },
+    {
+      "epoch": 0.9323479249573622,
+      "grad_norm": 4.757665364741502,
+      "learning_rate": 5.977705790754546e-08,
+      "loss": 0.5154,
+      "step": 11480
+    },
+    {
+      "epoch": 0.9324291399334037,
+      "grad_norm": 9.820132371604062,
+      "learning_rate": 5.963419469680543e-08,
+      "loss": 0.4637,
+      "step": 11481
+    },
+    {
+      "epoch": 0.9325103549094453,
+      "grad_norm": 4.899371695926488,
+      "learning_rate": 5.9491500346463005e-08,
+      "loss": 0.5009,
+      "step": 11482
+    },
+    {
+      "epoch": 0.9325915698854869,
+      "grad_norm": 3.7801882505750366,
+      "learning_rate": 5.934897486639307e-08,
+      "loss": 0.4272,
+      "step": 11483
+    },
+    {
+      "epoch": 0.9326727848615285,
+      "grad_norm": 7.561614245206794,
+      "learning_rate": 5.9206618266456904e-08,
+      "loss": 0.5861,
+      "step": 11484
+    },
+    {
+      "epoch": 0.93275399983757,
+      "grad_norm": 4.538003207030416,
+      "learning_rate": 5.906443055650496e-08,
+      "loss": 0.6691,
+      "step": 11485
+    },
+    {
+      "epoch": 0.9328352148136116,
+      "grad_norm": 5.474716210455667,
+      "learning_rate": 5.892241174637575e-08,
+      "loss": 0.3982,
+      "step": 11486
+    },
+    {
+      "epoch": 0.9329164297896532,
+      "grad_norm": 8.069204894922803,
+      "learning_rate": 5.8780561845896697e-08,
+      "loss": 0.52,
+      "step": 11487
+    },
+    {
+      "epoch": 0.9329976447656948,
+      "grad_norm": 5.907171852729622,
+      "learning_rate": 5.863888086488301e-08,
+      "loss": 0.5574,
+      "step": 11488
+    },
+    {
+      "epoch": 0.9330788597417363,
+      "grad_norm": 6.090121506693037,
+      "learning_rate": 5.849736881313767e-08,
+      "loss": 0.4025,
+      "step": 11489
+    },
+    {
+      "epoch": 0.933160074717778,
+      "grad_norm": 8.516034082485882,
+      "learning_rate": 5.835602570045312e-08,
+      "loss": 0.4993,
+      "step": 11490
+    },
+    {
+      "epoch": 0.9332412896938196,
+      "grad_norm": 6.277551920267633,
+      "learning_rate": 5.8214851536609326e-08,
+      "loss": 0.4205,
+      "step": 11491
+    },
+    {
+      "epoch": 0.9333225046698611,
+      "grad_norm": 8.065629651818462,
+      "learning_rate": 5.807384633137459e-08,
+      "loss": 0.4213,
+      "step": 11492
+    },
+    {
+      "epoch": 0.9334037196459027,
+      "grad_norm": 4.95647438456384,
+      "learning_rate": 5.793301009450636e-08,
+      "loss": 0.4459,
+      "step": 11493
+    },
+    {
+      "epoch": 0.9334849346219443,
+      "grad_norm": 4.983399216729912,
+      "learning_rate": 5.779234283574936e-08,
+      "loss": 0.5453,
+      "step": 11494
+    },
+    {
+      "epoch": 0.9335661495979859,
+      "grad_norm": 5.591248101931953,
+      "learning_rate": 5.765184456483664e-08,
+      "loss": 0.453,
+      "step": 11495
+    },
+    {
+      "epoch": 0.9336473645740274,
+      "grad_norm": 5.18268156667367,
+      "learning_rate": 5.7511515291490686e-08,
+      "loss": 0.4371,
+      "step": 11496
+    },
+    {
+      "epoch": 0.933728579550069,
+      "grad_norm": 6.395159139697958,
+      "learning_rate": 5.737135502542124e-08,
+      "loss": 0.5132,
+      "step": 11497
+    },
+    {
+      "epoch": 0.9338097945261106,
+      "grad_norm": 5.026018785036258,
+      "learning_rate": 5.7231363776326096e-08,
+      "loss": 0.4138,
+      "step": 11498
+    },
+    {
+      "epoch": 0.9338910095021522,
+      "grad_norm": 4.941851075904659,
+      "learning_rate": 5.709154155389279e-08,
+      "loss": 0.5284,
+      "step": 11499
+    },
+    {
+      "epoch": 0.9339722244781937,
+      "grad_norm": 5.637616770839022,
+      "learning_rate": 5.6951888367795804e-08,
+      "loss": 0.5964,
+      "step": 11500
+    },
+    {
+      "epoch": 0.9340534394542354,
+      "grad_norm": 6.004225890183753,
+      "learning_rate": 5.681240422769879e-08,
+      "loss": 0.4451,
+      "step": 11501
+    },
+    {
+      "epoch": 0.934134654430277,
+      "grad_norm": 4.723000628789615,
+      "learning_rate": 5.6673089143252646e-08,
+      "loss": 0.3078,
+      "step": 11502
+    },
+    {
+      "epoch": 0.9342158694063185,
+      "grad_norm": 6.077929375109591,
+      "learning_rate": 5.653394312409771e-08,
+      "loss": 0.6627,
+      "step": 11503
+    },
+    {
+      "epoch": 0.9342970843823601,
+      "grad_norm": 6.472316849971879,
+      "learning_rate": 5.639496617986184e-08,
+      "loss": 0.4792,
+      "step": 11504
+    },
+    {
+      "epoch": 0.9343782993584017,
+      "grad_norm": 6.293958361792907,
+      "learning_rate": 5.625615832016179e-08,
+      "loss": 0.438,
+      "step": 11505
+    },
+    {
+      "epoch": 0.9344595143344433,
+      "grad_norm": 5.09268953575727,
+      "learning_rate": 5.6117519554602375e-08,
+      "loss": 0.5255,
+      "step": 11506
+    },
+    {
+      "epoch": 0.9345407293104848,
+      "grad_norm": 5.054483428815471,
+      "learning_rate": 5.597904989277675e-08,
+      "loss": 0.437,
+      "step": 11507
+    },
+    {
+      "epoch": 0.9346219442865265,
+      "grad_norm": 3.9858596078394943,
+      "learning_rate": 5.584074934426559e-08,
+      "loss": 0.5096,
+      "step": 11508
+    },
+    {
+      "epoch": 0.934703159262568,
+      "grad_norm": 8.170161084504295,
+      "learning_rate": 5.570261791863957e-08,
+      "loss": 0.4268,
+      "step": 11509
+    },
+    {
+      "epoch": 0.9347843742386096,
+      "grad_norm": 5.199339732120409,
+      "learning_rate": 5.5564655625455766e-08,
+      "loss": 0.7232,
+      "step": 11510
+    },
+    {
+      "epoch": 0.9348655892146511,
+      "grad_norm": 4.643135079968665,
+      "learning_rate": 5.5426862474260986e-08,
+      "loss": 0.4998,
+      "step": 11511
+    },
+    {
+      "epoch": 0.9349468041906928,
+      "grad_norm": 4.163342887629166,
+      "learning_rate": 5.528923847458928e-08,
+      "loss": 0.4594,
+      "step": 11512
+    },
+    {
+      "epoch": 0.9350280191667344,
+      "grad_norm": 5.970898841399132,
+      "learning_rate": 5.5151783635964126e-08,
+      "loss": 0.46,
+      "step": 11513
+    },
+    {
+      "epoch": 0.9351092341427759,
+      "grad_norm": 5.478270806254457,
+      "learning_rate": 5.5014497967896266e-08,
+      "loss": 0.4887,
+      "step": 11514
+    },
+    {
+      "epoch": 0.9351904491188175,
+      "grad_norm": 4.460190714960654,
+      "learning_rate": 5.4877381479885307e-08,
+      "loss": 0.5598,
+      "step": 11515
+    },
+    {
+      "epoch": 0.9352716640948591,
+      "grad_norm": 4.1338050878154355,
+      "learning_rate": 5.4740434181418945e-08,
+      "loss": 0.47,
+      "step": 11516
+    },
+    {
+      "epoch": 0.9353528790709007,
+      "grad_norm": 6.435871850940438,
+      "learning_rate": 5.460365608197293e-08,
+      "loss": 0.4158,
+      "step": 11517
+    },
+    {
+      "epoch": 0.9354340940469422,
+      "grad_norm": 7.643223731934275,
+      "learning_rate": 5.4467047191011924e-08,
+      "loss": 0.3233,
+      "step": 11518
+    },
+    {
+      "epoch": 0.9355153090229839,
+      "grad_norm": 5.4260804495525825,
+      "learning_rate": 5.4330607517988635e-08,
+      "loss": 0.5388,
+      "step": 11519
+    },
+    {
+      "epoch": 0.9355965239990254,
+      "grad_norm": 4.953162854169758,
+      "learning_rate": 5.419433707234356e-08,
+      "loss": 0.4487,
+      "step": 11520
+    },
+    {
+      "epoch": 0.935677738975067,
+      "grad_norm": 3.6992673910297293,
+      "learning_rate": 5.4058235863506116e-08,
+      "loss": 0.5167,
+      "step": 11521
+    },
+    {
+      "epoch": 0.9357589539511085,
+      "grad_norm": 4.996367276417503,
+      "learning_rate": 5.392230390089404e-08,
+      "loss": 0.5104,
+      "step": 11522
+    },
+    {
+      "epoch": 0.9358401689271502,
+      "grad_norm": 4.62381128315659,
+      "learning_rate": 5.3786541193912854e-08,
+      "loss": 0.5088,
+      "step": 11523
+    },
+    {
+      "epoch": 0.9359213839031918,
+      "grad_norm": 7.587560626305793,
+      "learning_rate": 5.3650947751956174e-08,
+      "loss": 0.3284,
+      "step": 11524
+    },
+    {
+      "epoch": 0.9360025988792333,
+      "grad_norm": 5.005872247413698,
+      "learning_rate": 5.351552358440704e-08,
+      "loss": 0.4261,
+      "step": 11525
+    },
+    {
+      "epoch": 0.936083813855275,
+      "grad_norm": 5.695031699056669,
+      "learning_rate": 5.3380268700636006e-08,
+      "loss": 0.4862,
+      "step": 11526
+    },
+    {
+      "epoch": 0.9361650288313165,
+      "grad_norm": 5.066704771887348,
+      "learning_rate": 5.324518311000143e-08,
+      "loss": 0.5233,
+      "step": 11527
+    },
+    {
+      "epoch": 0.9362462438073581,
+      "grad_norm": 3.891689553812398,
+      "learning_rate": 5.311026682185139e-08,
+      "loss": 0.4053,
+      "step": 11528
+    },
+    {
+      "epoch": 0.9363274587833996,
+      "grad_norm": 4.149090685885997,
+      "learning_rate": 5.297551984552063e-08,
+      "loss": 0.5201,
+      "step": 11529
+    },
+    {
+      "epoch": 0.9364086737594413,
+      "grad_norm": 4.452851895114947,
+      "learning_rate": 5.2840942190333086e-08,
+      "loss": 0.4647,
+      "step": 11530
+    },
+    {
+      "epoch": 0.9364898887354828,
+      "grad_norm": 5.464100401282328,
+      "learning_rate": 5.270653386560104e-08,
+      "loss": 0.4444,
+      "step": 11531
+    },
+    {
+      "epoch": 0.9365711037115244,
+      "grad_norm": 7.585338070087896,
+      "learning_rate": 5.257229488062482e-08,
+      "loss": 0.4242,
+      "step": 11532
+    },
+    {
+      "epoch": 0.9366523186875659,
+      "grad_norm": 9.756341519461458,
+      "learning_rate": 5.243822524469283e-08,
+      "loss": 0.3792,
+      "step": 11533
+    },
+    {
+      "epoch": 0.9367335336636076,
+      "grad_norm": 6.0305409951018225,
+      "learning_rate": 5.23043249670821e-08,
+      "loss": 0.4986,
+      "step": 11534
+    },
+    {
+      "epoch": 0.9368147486396492,
+      "grad_norm": 8.933422659020032,
+      "learning_rate": 5.2170594057058264e-08,
+      "loss": 0.647,
+      "step": 11535
+    },
+    {
+      "epoch": 0.9368959636156907,
+      "grad_norm": 7.945666217999835,
+      "learning_rate": 5.2037032523873654e-08,
+      "loss": 0.4578,
+      "step": 11536
+    },
+    {
+      "epoch": 0.9369771785917324,
+      "grad_norm": 8.210836981044634,
+      "learning_rate": 5.190364037677142e-08,
+      "loss": 0.4933,
+      "step": 11537
+    },
+    {
+      "epoch": 0.9370583935677739,
+      "grad_norm": 4.671044337470819,
+      "learning_rate": 5.1770417624980306e-08,
+      "loss": 0.4541,
+      "step": 11538
+    },
+    {
+      "epoch": 0.9371396085438155,
+      "grad_norm": 6.043521315334802,
+      "learning_rate": 5.1637364277719595e-08,
+      "loss": 0.5181,
+      "step": 11539
+    },
+    {
+      "epoch": 0.937220823519857,
+      "grad_norm": 5.1011629271022185,
+      "learning_rate": 5.150448034419525e-08,
+      "loss": 0.5324,
+      "step": 11540
+    },
+    {
+      "epoch": 0.9373020384958987,
+      "grad_norm": 3.8201585871136734,
+      "learning_rate": 5.1371765833602703e-08,
+      "loss": 0.4803,
+      "step": 11541
+    },
+    {
+      "epoch": 0.9373832534719402,
+      "grad_norm": 5.758818308718375,
+      "learning_rate": 5.123922075512461e-08,
+      "loss": 0.3395,
+      "step": 11542
+    },
+    {
+      "epoch": 0.9374644684479818,
+      "grad_norm": 3.435971230625552,
+      "learning_rate": 5.110684511793251e-08,
+      "loss": 0.6348,
+      "step": 11543
+    },
+    {
+      "epoch": 0.9375456834240233,
+      "grad_norm": 7.13731160639292,
+      "learning_rate": 5.0974638931186036e-08,
+      "loss": 0.3566,
+      "step": 11544
+    },
+    {
+      "epoch": 0.937626898400065,
+      "grad_norm": 9.350076135939414,
+      "learning_rate": 5.084260220403342e-08,
+      "loss": 0.4923,
+      "step": 11545
+    },
+    {
+      "epoch": 0.9377081133761066,
+      "grad_norm": 4.901127986761642,
+      "learning_rate": 5.0710734945610686e-08,
+      "loss": 0.4283,
+      "step": 11546
+    },
+    {
+      "epoch": 0.9377893283521481,
+      "grad_norm": 4.685757580589546,
+      "learning_rate": 5.057903716504248e-08,
+      "loss": 0.491,
+      "step": 11547
+    },
+    {
+      "epoch": 0.9378705433281898,
+      "grad_norm": 5.048648910312251,
+      "learning_rate": 5.044750887144151e-08,
+      "loss": 0.2839,
+      "step": 11548
+    },
+    {
+      "epoch": 0.9379517583042313,
+      "grad_norm": 6.1324537629824345,
+      "learning_rate": 5.0316150073908555e-08,
+      "loss": 0.4318,
+      "step": 11549
+    },
+    {
+      "epoch": 0.9380329732802729,
+      "grad_norm": 5.300490311250102,
+      "learning_rate": 5.0184960781533844e-08,
+      "loss": 0.3291,
+      "step": 11550
+    },
+    {
+      "epoch": 0.9381141882563144,
+      "grad_norm": 4.481525918571603,
+      "learning_rate": 5.005394100339373e-08,
+      "loss": 0.6305,
+      "step": 11551
+    },
+    {
+      "epoch": 0.9381954032323561,
+      "grad_norm": 5.628772875047986,
+      "learning_rate": 4.992309074855484e-08,
+      "loss": 0.4441,
+      "step": 11552
+    },
+    {
+      "epoch": 0.9382766182083976,
+      "grad_norm": 5.631338094522123,
+      "learning_rate": 4.97924100260716e-08,
+      "loss": 0.362,
+      "step": 11553
+    },
+    {
+      "epoch": 0.9383578331844392,
+      "grad_norm": 4.450334044578378,
+      "learning_rate": 4.966189884498596e-08,
+      "loss": 0.4233,
+      "step": 11554
+    },
+    {
+      "epoch": 0.9384390481604807,
+      "grad_norm": 4.584301414669546,
+      "learning_rate": 4.953155721432873e-08,
+      "loss": 0.4751,
+      "step": 11555
+    },
+    {
+      "epoch": 0.9385202631365224,
+      "grad_norm": 4.749314594480385,
+      "learning_rate": 4.940138514311854e-08,
+      "loss": 0.5722,
+      "step": 11556
+    },
+    {
+      "epoch": 0.938601478112564,
+      "grad_norm": 4.863986365737658,
+      "learning_rate": 4.927138264036291e-08,
+      "loss": 0.4222,
+      "step": 11557
+    },
+    {
+      "epoch": 0.9386826930886055,
+      "grad_norm": 5.649739794167962,
+      "learning_rate": 4.9141549715057415e-08,
+      "loss": 0.4626,
+      "step": 11558
+    },
+    {
+      "epoch": 0.9387639080646472,
+      "grad_norm": 3.4200650534492807,
+      "learning_rate": 4.90118863761857e-08,
+      "loss": 0.412,
+      "step": 11559
+    },
+    {
+      "epoch": 0.9388451230406887,
+      "grad_norm": 8.38158690698774,
+      "learning_rate": 4.888239263271977e-08,
+      "loss": 0.5341,
+      "step": 11560
+    },
+    {
+      "epoch": 0.9389263380167303,
+      "grad_norm": 3.2373374116022893,
+      "learning_rate": 4.875306849361994e-08,
+      "loss": 0.4586,
+      "step": 11561
+    },
+    {
+      "epoch": 0.9390075529927718,
+      "grad_norm": 4.301803597627936,
+      "learning_rate": 4.862391396783461e-08,
+      "loss": 0.3589,
+      "step": 11562
+    },
+    {
+      "epoch": 0.9390887679688135,
+      "grad_norm": 5.310812898753721,
+      "learning_rate": 4.849492906430081e-08,
+      "loss": 0.4699,
+      "step": 11563
+    },
+    {
+      "epoch": 0.939169982944855,
+      "grad_norm": 5.309327329673865,
+      "learning_rate": 4.836611379194334e-08,
+      "loss": 0.5494,
+      "step": 11564
+    },
+    {
+      "epoch": 0.9392511979208966,
+      "grad_norm": 8.887102738638378,
+      "learning_rate": 4.8237468159675896e-08,
+      "loss": 0.3454,
+      "step": 11565
+    },
+    {
+      "epoch": 0.9393324128969381,
+      "grad_norm": 4.344793062039761,
+      "learning_rate": 4.810899217639997e-08,
+      "loss": 0.3838,
+      "step": 11566
+    },
+    {
+      "epoch": 0.9394136278729798,
+      "grad_norm": 3.970983027129392,
+      "learning_rate": 4.798068585100513e-08,
+      "loss": 0.463,
+      "step": 11567
+    },
+    {
+      "epoch": 0.9394948428490214,
+      "grad_norm": 6.57362417638814,
+      "learning_rate": 4.785254919236954e-08,
+      "loss": 0.4031,
+      "step": 11568
+    },
+    {
+      "epoch": 0.9395760578250629,
+      "grad_norm": 3.7805553967871552,
+      "learning_rate": 4.772458220936027e-08,
+      "loss": 0.4996,
+      "step": 11569
+    },
+    {
+      "epoch": 0.9396572728011046,
+      "grad_norm": 5.108413544198211,
+      "learning_rate": 4.7596784910830804e-08,
+      "loss": 0.4602,
+      "step": 11570
+    },
+    {
+      "epoch": 0.9397384877771461,
+      "grad_norm": 4.497648219050411,
+      "learning_rate": 4.74691573056249e-08,
+      "loss": 0.3966,
+      "step": 11571
+    },
+    {
+      "epoch": 0.9398197027531877,
+      "grad_norm": 4.1059390155620115,
+      "learning_rate": 4.7341699402573546e-08,
+      "loss": 0.3698,
+      "step": 11572
+    },
+    {
+      "epoch": 0.9399009177292292,
+      "grad_norm": 4.091197560447974,
+      "learning_rate": 4.721441121049608e-08,
+      "loss": 0.4798,
+      "step": 11573
+    },
+    {
+      "epoch": 0.9399821327052709,
+      "grad_norm": 5.006114532781862,
+      "learning_rate": 4.7087292738200454e-08,
+      "loss": 0.3846,
+      "step": 11574
+    },
+    {
+      "epoch": 0.9400633476813124,
+      "grad_norm": 7.010800653937726,
+      "learning_rate": 4.696034399448185e-08,
+      "loss": 0.3729,
+      "step": 11575
+    },
+    {
+      "epoch": 0.940144562657354,
+      "grad_norm": 5.687679281302264,
+      "learning_rate": 4.6833564988124914e-08,
+      "loss": 0.4974,
+      "step": 11576
+    },
+    {
+      "epoch": 0.9402257776333955,
+      "grad_norm": 5.740323370278037,
+      "learning_rate": 4.670695572790235e-08,
+      "loss": 0.4069,
+      "step": 11577
+    },
+    {
+      "epoch": 0.9403069926094372,
+      "grad_norm": 8.241228285955831,
+      "learning_rate": 4.658051622257437e-08,
+      "loss": 0.6067,
+      "step": 11578
+    },
+    {
+      "epoch": 0.9403882075854788,
+      "grad_norm": 4.495511178750625,
+      "learning_rate": 4.6454246480890084e-08,
+      "loss": 0.5117,
+      "step": 11579
+    },
+    {
+      "epoch": 0.9404694225615203,
+      "grad_norm": 6.551988717623826,
+      "learning_rate": 4.632814651158696e-08,
+      "loss": 0.4359,
+      "step": 11580
+    },
+    {
+      "epoch": 0.940550637537562,
+      "grad_norm": 9.683649036003146,
+      "learning_rate": 4.620221632338995e-08,
+      "loss": 0.4789,
+      "step": 11581
+    },
+    {
+      "epoch": 0.9406318525136035,
+      "grad_norm": 5.298725894305153,
+      "learning_rate": 4.607645592501347e-08,
+      "loss": 0.4102,
+      "step": 11582
+    },
+    {
+      "epoch": 0.9407130674896451,
+      "grad_norm": 5.2892762046601165,
+      "learning_rate": 4.5950865325158636e-08,
+      "loss": 0.5595,
+      "step": 11583
+    },
+    {
+      "epoch": 0.9407942824656866,
+      "grad_norm": 6.532711062097573,
+      "learning_rate": 4.582544453251597e-08,
+      "loss": 0.3798,
+      "step": 11584
+    },
+    {
+      "epoch": 0.9408754974417283,
+      "grad_norm": 7.025731896892658,
+      "learning_rate": 4.57001935557641e-08,
+      "loss": 0.444,
+      "step": 11585
+    },
+    {
+      "epoch": 0.9409567124177698,
+      "grad_norm": 4.380484134636201,
+      "learning_rate": 4.5575112403569985e-08,
+      "loss": 0.4212,
+      "step": 11586
+    },
+    {
+      "epoch": 0.9410379273938114,
+      "grad_norm": 78.3339524286403,
+      "learning_rate": 4.545020108458781e-08,
+      "loss": 0.428,
+      "step": 11587
+    },
+    {
+      "epoch": 0.941119142369853,
+      "grad_norm": 4.997755103806688,
+      "learning_rate": 4.5325459607461485e-08,
+      "loss": 0.374,
+      "step": 11588
+    },
+    {
+      "epoch": 0.9412003573458946,
+      "grad_norm": 4.4071049437250505,
+      "learning_rate": 4.5200887980821897e-08,
+      "loss": 0.6705,
+      "step": 11589
+    },
+    {
+      "epoch": 0.9412815723219362,
+      "grad_norm": 4.496642306863949,
+      "learning_rate": 4.5076486213289086e-08,
+      "loss": 0.4301,
+      "step": 11590
+    },
+    {
+      "epoch": 0.9413627872979777,
+      "grad_norm": 6.033449529408409,
+      "learning_rate": 4.495225431347089e-08,
+      "loss": 0.4866,
+      "step": 11591
+    },
+    {
+      "epoch": 0.9414440022740194,
+      "grad_norm": 8.981792164965075,
+      "learning_rate": 4.482819228996377e-08,
+      "loss": 0.6155,
+      "step": 11592
+    },
+    {
+      "epoch": 0.9415252172500609,
+      "grad_norm": 6.80461573911351,
+      "learning_rate": 4.470430015135197e-08,
+      "loss": 0.5014,
+      "step": 11593
+    },
+    {
+      "epoch": 0.9416064322261025,
+      "grad_norm": 3.2551970239032495,
+      "learning_rate": 4.458057790620779e-08,
+      "loss": 0.4495,
+      "step": 11594
+    },
+    {
+      "epoch": 0.941687647202144,
+      "grad_norm": 6.288968375149101,
+      "learning_rate": 4.4457025563092724e-08,
+      "loss": 0.4876,
+      "step": 11595
+    },
+    {
+      "epoch": 0.9417688621781857,
+      "grad_norm": 5.114911602404954,
+      "learning_rate": 4.433364313055549e-08,
+      "loss": 0.6019,
+      "step": 11596
+    },
+    {
+      "epoch": 0.9418500771542272,
+      "grad_norm": 10.285824379728702,
+      "learning_rate": 4.42104306171337e-08,
+      "loss": 0.3655,
+      "step": 11597
+    },
+    {
+      "epoch": 0.9419312921302688,
+      "grad_norm": 4.250288445214354,
+      "learning_rate": 4.4087388031353316e-08,
+      "loss": 0.5237,
+      "step": 11598
+    },
+    {
+      "epoch": 0.9420125071063103,
+      "grad_norm": 4.607829438175469,
+      "learning_rate": 4.39645153817278e-08,
+      "loss": 0.5283,
+      "step": 11599
+    },
+    {
+      "epoch": 0.942093722082352,
+      "grad_norm": 7.915466583935277,
+      "learning_rate": 4.384181267675952e-08,
+      "loss": 0.4855,
+      "step": 11600
+    },
+    {
+      "epoch": 0.9421749370583936,
+      "grad_norm": 5.46136708291508,
+      "learning_rate": 4.3719279924938626e-08,
+      "loss": 0.4598,
+      "step": 11601
+    },
+    {
+      "epoch": 0.9422561520344351,
+      "grad_norm": 9.165200607430856,
+      "learning_rate": 4.35969171347439e-08,
+      "loss": 0.6132,
+      "step": 11602
+    },
+    {
+      "epoch": 0.9423373670104768,
+      "grad_norm": 5.415018652981148,
+      "learning_rate": 4.347472431464217e-08,
+      "loss": 0.5896,
+      "step": 11603
+    },
+    {
+      "epoch": 0.9424185819865183,
+      "grad_norm": 7.296700648217176,
+      "learning_rate": 4.335270147308862e-08,
+      "loss": 0.378,
+      "step": 11604
+    },
+    {
+      "epoch": 0.9424997969625599,
+      "grad_norm": 4.175947964206813,
+      "learning_rate": 4.32308486185265e-08,
+      "loss": 0.4858,
+      "step": 11605
+    },
+    {
+      "epoch": 0.9425810119386014,
+      "grad_norm": 5.139558164610286,
+      "learning_rate": 4.3109165759387115e-08,
+      "loss": 0.3888,
+      "step": 11606
+    },
+    {
+      "epoch": 0.9426622269146431,
+      "grad_norm": 4.109603857789413,
+      "learning_rate": 4.298765290409096e-08,
+      "loss": 0.5018,
+      "step": 11607
+    },
+    {
+      "epoch": 0.9427434418906846,
+      "grad_norm": 5.23991591103504,
+      "learning_rate": 4.286631006104547e-08,
+      "loss": 0.523,
+      "step": 11608
+    },
+    {
+      "epoch": 0.9428246568667262,
+      "grad_norm": 5.661548234950453,
+      "learning_rate": 4.2745137238646984e-08,
+      "loss": 0.4094,
+      "step": 11609
+    },
+    {
+      "epoch": 0.9429058718427678,
+      "grad_norm": 5.499414318975432,
+      "learning_rate": 4.2624134445280186e-08,
+      "loss": 0.6268,
+      "step": 11610
+    },
+    {
+      "epoch": 0.9429870868188094,
+      "grad_norm": 4.495827244130177,
+      "learning_rate": 4.25033016893181e-08,
+      "loss": 0.548,
+      "step": 11611
+    },
+    {
+      "epoch": 0.943068301794851,
+      "grad_norm": 3.8103963823755094,
+      "learning_rate": 4.238263897912126e-08,
+      "loss": 0.7681,
+      "step": 11612
+    },
+    {
+      "epoch": 0.9431495167708925,
+      "grad_norm": 9.357203074745977,
+      "learning_rate": 4.22621463230391e-08,
+      "loss": 0.3692,
+      "step": 11613
+    },
+    {
+      "epoch": 0.9432307317469342,
+      "grad_norm": 3.6750136809793412,
+      "learning_rate": 4.214182372940884e-08,
+      "loss": 0.601,
+      "step": 11614
+    },
+    {
+      "epoch": 0.9433119467229757,
+      "grad_norm": 7.385444469733557,
+      "learning_rate": 4.202167120655631e-08,
+      "loss": 0.4761,
+      "step": 11615
+    },
+    {
+      "epoch": 0.9433931616990173,
+      "grad_norm": 5.303774677219012,
+      "learning_rate": 4.190168876279571e-08,
+      "loss": 0.4371,
+      "step": 11616
+    },
+    {
+      "epoch": 0.9434743766750588,
+      "grad_norm": 4.763056521648861,
+      "learning_rate": 4.1781876406428725e-08,
+      "loss": 0.4442,
+      "step": 11617
+    },
+    {
+      "epoch": 0.9435555916511005,
+      "grad_norm": 5.484993623084503,
+      "learning_rate": 4.1662234145746214e-08,
+      "loss": 0.5426,
+      "step": 11618
+    },
+    {
+      "epoch": 0.943636806627142,
+      "grad_norm": 4.303398425305994,
+      "learning_rate": 4.154276198902629e-08,
+      "loss": 0.5846,
+      "step": 11619
+    },
+    {
+      "epoch": 0.9437180216031836,
+      "grad_norm": 5.295687518325545,
+      "learning_rate": 4.1423459944536224e-08,
+      "loss": 0.6349,
+      "step": 11620
+    },
+    {
+      "epoch": 0.9437992365792252,
+      "grad_norm": 4.737076827144997,
+      "learning_rate": 4.1304328020530804e-08,
+      "loss": 0.4874,
+      "step": 11621
+    },
+    {
+      "epoch": 0.9438804515552668,
+      "grad_norm": 15.165489379351301,
+      "learning_rate": 4.118536622525315e-08,
+      "loss": 0.3641,
+      "step": 11622
+    },
+    {
+      "epoch": 0.9439616665313084,
+      "grad_norm": 3.5716624018017877,
+      "learning_rate": 4.10665745669353e-08,
+      "loss": 0.5441,
+      "step": 11623
+    },
+    {
+      "epoch": 0.9440428815073499,
+      "grad_norm": 10.121245968679512,
+      "learning_rate": 4.094795305379679e-08,
+      "loss": 0.4997,
+      "step": 11624
+    },
+    {
+      "epoch": 0.9441240964833916,
+      "grad_norm": 7.249621116373978,
+      "learning_rate": 4.082950169404548e-08,
+      "loss": 0.4358,
+      "step": 11625
+    },
+    {
+      "epoch": 0.9442053114594331,
+      "grad_norm": 7.423045226596472,
+      "learning_rate": 4.071122049587789e-08,
+      "loss": 0.4766,
+      "step": 11626
+    },
+    {
+      "epoch": 0.9442865264354747,
+      "grad_norm": 7.43889332245654,
+      "learning_rate": 4.059310946747802e-08,
+      "loss": 0.3625,
+      "step": 11627
+    },
+    {
+      "epoch": 0.9443677414115162,
+      "grad_norm": 5.999420084859953,
+      "learning_rate": 4.047516861701878e-08,
+      "loss": 0.4215,
+      "step": 11628
+    },
+    {
+      "epoch": 0.9444489563875579,
+      "grad_norm": 3.9495688216293363,
+      "learning_rate": 4.035739795266086e-08,
+      "loss": 0.5308,
+      "step": 11629
+    },
+    {
+      "epoch": 0.9445301713635994,
+      "grad_norm": 8.225837590539006,
+      "learning_rate": 4.0239797482553856e-08,
+      "loss": 0.4483,
+      "step": 11630
+    },
+    {
+      "epoch": 0.944611386339641,
+      "grad_norm": 4.507895023066098,
+      "learning_rate": 4.012236721483487e-08,
+      "loss": 0.5678,
+      "step": 11631
+    },
+    {
+      "epoch": 0.9446926013156827,
+      "grad_norm": 3.795653600393925,
+      "learning_rate": 4.0005107157628786e-08,
+      "loss": 0.5971,
+      "step": 11632
+    },
+    {
+      "epoch": 0.9447738162917242,
+      "grad_norm": 5.829598757301327,
+      "learning_rate": 3.988801731905051e-08,
+      "loss": 0.4818,
+      "step": 11633
+    },
+    {
+      "epoch": 0.9448550312677658,
+      "grad_norm": 6.893765169454947,
+      "learning_rate": 3.9771097707201056e-08,
+      "loss": 0.3942,
+      "step": 11634
+    },
+    {
+      "epoch": 0.9449362462438073,
+      "grad_norm": 7.0587960070765075,
+      "learning_rate": 3.965434833017118e-08,
+      "loss": 0.5054,
+      "step": 11635
+    },
+    {
+      "epoch": 0.945017461219849,
+      "grad_norm": 8.051189199148112,
+      "learning_rate": 3.9537769196039134e-08,
+      "loss": 0.3643,
+      "step": 11636
+    },
+    {
+      "epoch": 0.9450986761958905,
+      "grad_norm": 4.735681936699135,
+      "learning_rate": 3.9421360312871804e-08,
+      "loss": 0.6727,
+      "step": 11637
+    },
+    {
+      "epoch": 0.9451798911719321,
+      "grad_norm": 3.9449239206953925,
+      "learning_rate": 3.9305121688723855e-08,
+      "loss": 0.5668,
+      "step": 11638
+    },
+    {
+      "epoch": 0.9452611061479737,
+      "grad_norm": 6.946560704332956,
+      "learning_rate": 3.918905333163858e-08,
+      "loss": 0.6188,
+      "step": 11639
+    },
+    {
+      "epoch": 0.9453423211240153,
+      "grad_norm": 8.04163254796676,
+      "learning_rate": 3.9073155249647055e-08,
+      "loss": 0.6601,
+      "step": 11640
+    },
+    {
+      "epoch": 0.9454235361000568,
+      "grad_norm": 7.0432614921051675,
+      "learning_rate": 3.895742745076869e-08,
+      "loss": 0.4597,
+      "step": 11641
+    },
+    {
+      "epoch": 0.9455047510760984,
+      "grad_norm": 5.740385711050734,
+      "learning_rate": 3.8841869943011534e-08,
+      "loss": 0.4694,
+      "step": 11642
+    },
+    {
+      "epoch": 0.9455859660521401,
+      "grad_norm": 3.826367448009286,
+      "learning_rate": 3.872648273437168e-08,
+      "loss": 0.4556,
+      "step": 11643
+    },
+    {
+      "epoch": 0.9456671810281816,
+      "grad_norm": 3.153187096292459,
+      "learning_rate": 3.861126583283303e-08,
+      "loss": 0.3736,
+      "step": 11644
+    },
+    {
+      "epoch": 0.9457483960042232,
+      "grad_norm": 5.786888401113132,
+      "learning_rate": 3.849621924636809e-08,
+      "loss": 0.5045,
+      "step": 11645
+    },
+    {
+      "epoch": 0.9458296109802647,
+      "grad_norm": 5.70036944601648,
+      "learning_rate": 3.838134298293744e-08,
+      "loss": 0.3993,
+      "step": 11646
+    },
+    {
+      "epoch": 0.9459108259563064,
+      "grad_norm": 5.14700753120722,
+      "learning_rate": 3.8266637050489716e-08,
+      "loss": 0.6705,
+      "step": 11647
+    },
+    {
+      "epoch": 0.9459920409323479,
+      "grad_norm": 5.028518558037041,
+      "learning_rate": 3.815210145696219e-08,
+      "loss": 0.5472,
+      "step": 11648
+    },
+    {
+      "epoch": 0.9460732559083895,
+      "grad_norm": 5.08541011233092,
+      "learning_rate": 3.803773621028045e-08,
+      "loss": 0.4755,
+      "step": 11649
+    },
+    {
+      "epoch": 0.946154470884431,
+      "grad_norm": 5.674196216201576,
+      "learning_rate": 3.792354131835735e-08,
+      "loss": 0.4256,
+      "step": 11650
+    },
+    {
+      "epoch": 0.9462356858604727,
+      "grad_norm": 5.863265288069775,
+      "learning_rate": 3.780951678909489e-08,
+      "loss": 0.5258,
+      "step": 11651
+    },
+    {
+      "epoch": 0.9463169008365142,
+      "grad_norm": 3.5058013583786476,
+      "learning_rate": 3.769566263038288e-08,
+      "loss": 0.534,
+      "step": 11652
+    },
+    {
+      "epoch": 0.9463981158125558,
+      "grad_norm": 12.342326134411538,
+      "learning_rate": 3.7581978850099456e-08,
+      "loss": 0.5344,
+      "step": 11653
+    },
+    {
+      "epoch": 0.9464793307885975,
+      "grad_norm": 4.819965937928789,
+      "learning_rate": 3.7468465456110825e-08,
+      "loss": 0.4811,
+      "step": 11654
+    },
+    {
+      "epoch": 0.946560545764639,
+      "grad_norm": 7.405750445686644,
+      "learning_rate": 3.735512245627182e-08,
+      "loss": 0.3502,
+      "step": 11655
+    },
+    {
+      "epoch": 0.9466417607406806,
+      "grad_norm": 9.446771351873101,
+      "learning_rate": 3.7241949858424777e-08,
+      "loss": 0.4258,
+      "step": 11656
+    },
+    {
+      "epoch": 0.9467229757167221,
+      "grad_norm": 6.936102522174853,
+      "learning_rate": 3.712894767040093e-08,
+      "loss": 0.5589,
+      "step": 11657
+    },
+    {
+      "epoch": 0.9468041906927638,
+      "grad_norm": 4.033630186749134,
+      "learning_rate": 3.7016115900019575e-08,
+      "loss": 0.4372,
+      "step": 11658
+    },
+    {
+      "epoch": 0.9468854056688053,
+      "grad_norm": 5.941857393690612,
+      "learning_rate": 3.690345455508754e-08,
+      "loss": 0.5026,
+      "step": 11659
+    },
+    {
+      "epoch": 0.9469666206448469,
+      "grad_norm": 9.14091774177606,
+      "learning_rate": 3.679096364340079e-08,
+      "loss": 0.4604,
+      "step": 11660
+    },
+    {
+      "epoch": 0.9470478356208885,
+      "grad_norm": 4.223097671499599,
+      "learning_rate": 3.6678643172742836e-08,
+      "loss": 0.4393,
+      "step": 11661
+    },
+    {
+      "epoch": 0.9471290505969301,
+      "grad_norm": 3.9468831332746643,
+      "learning_rate": 3.656649315088606e-08,
+      "loss": 0.5762,
+      "step": 11662
+    },
+    {
+      "epoch": 0.9472102655729716,
+      "grad_norm": 4.743190373097494,
+      "learning_rate": 3.6454513585590376e-08,
+      "loss": 0.4932,
+      "step": 11663
+    },
+    {
+      "epoch": 0.9472914805490132,
+      "grad_norm": 3.896032850566264,
+      "learning_rate": 3.634270448460403e-08,
+      "loss": 0.5384,
+      "step": 11664
+    },
+    {
+      "epoch": 0.9473726955250549,
+      "grad_norm": 4.452680361174725,
+      "learning_rate": 3.623106585566388e-08,
+      "loss": 0.3909,
+      "step": 11665
+    },
+    {
+      "epoch": 0.9474539105010964,
+      "grad_norm": 6.087481905533953,
+      "learning_rate": 3.611959770649487e-08,
+      "loss": 0.4937,
+      "step": 11666
+    },
+    {
+      "epoch": 0.947535125477138,
+      "grad_norm": 4.040614567901036,
+      "learning_rate": 3.600830004480943e-08,
+      "loss": 0.5005,
+      "step": 11667
+    },
+    {
+      "epoch": 0.9476163404531796,
+      "grad_norm": 4.44420665256172,
+      "learning_rate": 3.589717287830946e-08,
+      "loss": 0.4575,
+      "step": 11668
+    },
+    {
+      "epoch": 0.9476975554292212,
+      "grad_norm": 5.616498527932605,
+      "learning_rate": 3.578621621468381e-08,
+      "loss": 0.5212,
+      "step": 11669
+    },
+    {
+      "epoch": 0.9477787704052627,
+      "grad_norm": 5.201502802582349,
+      "learning_rate": 3.567543006161051e-08,
+      "loss": 0.4217,
+      "step": 11670
+    },
+    {
+      "epoch": 0.9478599853813043,
+      "grad_norm": 16.378504345168306,
+      "learning_rate": 3.556481442675508e-08,
+      "loss": 0.4972,
+      "step": 11671
+    },
+    {
+      "epoch": 0.9479412003573459,
+      "grad_norm": 5.083869322555315,
+      "learning_rate": 3.5454369317771686e-08,
+      "loss": 0.4814,
+      "step": 11672
+    },
+    {
+      "epoch": 0.9480224153333875,
+      "grad_norm": 7.335908845803885,
+      "learning_rate": 3.534409474230255e-08,
+      "loss": 0.4003,
+      "step": 11673
+    },
+    {
+      "epoch": 0.948103630309429,
+      "grad_norm": 10.162528148573347,
+      "learning_rate": 3.523399070797795e-08,
+      "loss": 0.4699,
+      "step": 11674
+    },
+    {
+      "epoch": 0.9481848452854706,
+      "grad_norm": 4.64890275823902,
+      "learning_rate": 3.512405722241652e-08,
+      "loss": 0.4235,
+      "step": 11675
+    },
+    {
+      "epoch": 0.9482660602615123,
+      "grad_norm": 5.016364672410445,
+      "learning_rate": 3.501429429322522e-08,
+      "loss": 0.5031,
+      "step": 11676
+    },
+    {
+      "epoch": 0.9483472752375538,
+      "grad_norm": 5.858582608768708,
+      "learning_rate": 3.4904701927999385e-08,
+      "loss": 0.4388,
+      "step": 11677
+    },
+    {
+      "epoch": 0.9484284902135954,
+      "grad_norm": 4.81391584316998,
+      "learning_rate": 3.479528013432154e-08,
+      "loss": 0.4828,
+      "step": 11678
+    },
+    {
+      "epoch": 0.948509705189637,
+      "grad_norm": 8.006015757453834,
+      "learning_rate": 3.468602891976314e-08,
+      "loss": 0.5023,
+      "step": 11679
+    },
+    {
+      "epoch": 0.9485909201656786,
+      "grad_norm": 13.29141782498199,
+      "learning_rate": 3.457694829188452e-08,
+      "loss": 0.4342,
+      "step": 11680
+    },
+    {
+      "epoch": 0.9486721351417201,
+      "grad_norm": 10.20237535145513,
+      "learning_rate": 3.446803825823269e-08,
+      "loss": 0.3647,
+      "step": 11681
+    },
+    {
+      "epoch": 0.9487533501177617,
+      "grad_norm": 5.890040192499068,
+      "learning_rate": 3.435929882634415e-08,
+      "loss": 0.3275,
+      "step": 11682
+    },
+    {
+      "epoch": 0.9488345650938033,
+      "grad_norm": 10.83456434309621,
+      "learning_rate": 3.425073000374257e-08,
+      "loss": 0.4327,
+      "step": 11683
+    },
+    {
+      "epoch": 0.9489157800698449,
+      "grad_norm": 4.338209138202084,
+      "learning_rate": 3.4142331797940855e-08,
+      "loss": 0.5753,
+      "step": 11684
+    },
+    {
+      "epoch": 0.9489969950458864,
+      "grad_norm": 4.739057059307357,
+      "learning_rate": 3.4034104216439655e-08,
+      "loss": 0.5673,
+      "step": 11685
+    },
+    {
+      "epoch": 0.949078210021928,
+      "grad_norm": 7.495421522605058,
+      "learning_rate": 3.3926047266727155e-08,
+      "loss": 0.3894,
+      "step": 11686
+    },
+    {
+      "epoch": 0.9491594249979697,
+      "grad_norm": 7.2165806929599166,
+      "learning_rate": 3.381816095628071e-08,
+      "loss": 0.5159,
+      "step": 11687
+    },
+    {
+      "epoch": 0.9492406399740112,
+      "grad_norm": 3.9257600126354624,
+      "learning_rate": 3.371044529256573e-08,
+      "loss": 0.389,
+      "step": 11688
+    },
+    {
+      "epoch": 0.9493218549500528,
+      "grad_norm": 9.018221626405468,
+      "learning_rate": 3.360290028303487e-08,
+      "loss": 0.4364,
+      "step": 11689
+    },
+    {
+      "epoch": 0.9494030699260944,
+      "grad_norm": 4.246553322453479,
+      "learning_rate": 3.34955259351305e-08,
+      "loss": 0.5857,
+      "step": 11690
+    },
+    {
+      "epoch": 0.949484284902136,
+      "grad_norm": 8.521074043393677,
+      "learning_rate": 3.3388322256281694e-08,
+      "loss": 0.4385,
+      "step": 11691
+    },
+    {
+      "epoch": 0.9495654998781775,
+      "grad_norm": 8.037563989138466,
+      "learning_rate": 3.328128925390667e-08,
+      "loss": 0.5179,
+      "step": 11692
+    },
+    {
+      "epoch": 0.9496467148542191,
+      "grad_norm": 5.337969212336468,
+      "learning_rate": 3.317442693541145e-08,
+      "loss": 0.5983,
+      "step": 11693
+    },
+    {
+      "epoch": 0.9497279298302607,
+      "grad_norm": 5.550986004035013,
+      "learning_rate": 3.306773530819041e-08,
+      "loss": 0.5231,
+      "step": 11694
+    },
+    {
+      "epoch": 0.9498091448063023,
+      "grad_norm": 5.0987699618930975,
+      "learning_rate": 3.296121437962624e-08,
+      "loss": 0.6063,
+      "step": 11695
+    },
+    {
+      "epoch": 0.9498903597823438,
+      "grad_norm": 6.3006147509879105,
+      "learning_rate": 3.2854864157089164e-08,
+      "loss": 0.3196,
+      "step": 11696
+    },
+    {
+      "epoch": 0.9499715747583855,
+      "grad_norm": 4.720216946251891,
+      "learning_rate": 3.2748684647938564e-08,
+      "loss": 0.5137,
+      "step": 11697
+    },
+    {
+      "epoch": 0.9500527897344271,
+      "grad_norm": 17.425444481965904,
+      "learning_rate": 3.264267585952108e-08,
+      "loss": 0.5927,
+      "step": 11698
+    },
+    {
+      "epoch": 0.9501340047104686,
+      "grad_norm": 6.961626143865487,
+      "learning_rate": 3.253683779917194e-08,
+      "loss": 0.4547,
+      "step": 11699
+    },
+    {
+      "epoch": 0.9502152196865102,
+      "grad_norm": 7.6904422344104955,
+      "learning_rate": 3.243117047421501e-08,
+      "loss": 0.5546,
+      "step": 11700
+    },
+    {
+      "epoch": 0.9502964346625518,
+      "grad_norm": 5.910333731484439,
+      "learning_rate": 3.2325673891961394e-08,
+      "loss": 0.641,
+      "step": 11701
+    },
+    {
+      "epoch": 0.9503776496385934,
+      "grad_norm": 7.209494401473691,
+      "learning_rate": 3.222034805971136e-08,
+      "loss": 0.4257,
+      "step": 11702
+    },
+    {
+      "epoch": 0.9504588646146349,
+      "grad_norm": 5.915799539490019,
+      "learning_rate": 3.2115192984752684e-08,
+      "loss": 0.3847,
+      "step": 11703
+    },
+    {
+      "epoch": 0.9505400795906765,
+      "grad_norm": 8.657367196517342,
+      "learning_rate": 3.2010208674361774e-08,
+      "loss": 0.6126,
+      "step": 11704
+    },
+    {
+      "epoch": 0.9506212945667181,
+      "grad_norm": 5.775245158747669,
+      "learning_rate": 3.190539513580226e-08,
+      "loss": 0.4298,
+      "step": 11705
+    },
+    {
+      "epoch": 0.9507025095427597,
+      "grad_norm": 5.310693978477437,
+      "learning_rate": 3.1800752376327515e-08,
+      "loss": 0.5534,
+      "step": 11706
+    },
+    {
+      "epoch": 0.9507837245188012,
+      "grad_norm": 4.560752491636321,
+      "learning_rate": 3.169628040317785e-08,
+      "loss": 0.3459,
+      "step": 11707
+    },
+    {
+      "epoch": 0.9508649394948429,
+      "grad_norm": 3.5975956386159296,
+      "learning_rate": 3.15919792235822e-08,
+      "loss": 0.4466,
+      "step": 11708
+    },
+    {
+      "epoch": 0.9509461544708845,
+      "grad_norm": 4.190834372560731,
+      "learning_rate": 3.1487848844757865e-08,
+      "loss": 0.5566,
+      "step": 11709
+    },
+    {
+      "epoch": 0.951027369446926,
+      "grad_norm": 4.688075724324062,
+      "learning_rate": 3.138388927391017e-08,
+      "loss": 0.5138,
+      "step": 11710
+    },
+    {
+      "epoch": 0.9511085844229676,
+      "grad_norm": 6.558872976862139,
+      "learning_rate": 3.1280100518231994e-08,
+      "loss": 0.4949,
+      "step": 11711
+    },
+    {
+      "epoch": 0.9511897993990092,
+      "grad_norm": 6.1291704390663515,
+      "learning_rate": 3.1176482584905356e-08,
+      "loss": 0.6129,
+      "step": 11712
+    },
+    {
+      "epoch": 0.9512710143750508,
+      "grad_norm": 4.609364781861959,
+      "learning_rate": 3.107303548110008e-08,
+      "loss": 0.5307,
+      "step": 11713
+    },
+    {
+      "epoch": 0.9513522293510923,
+      "grad_norm": 4.543991780518672,
+      "learning_rate": 3.0969759213974324e-08,
+      "loss": 0.4039,
+      "step": 11714
+    },
+    {
+      "epoch": 0.951433444327134,
+      "grad_norm": 5.065242928777515,
+      "learning_rate": 3.086665379067405e-08,
+      "loss": 0.4611,
+      "step": 11715
+    },
+    {
+      "epoch": 0.9515146593031755,
+      "grad_norm": 7.643230789019235,
+      "learning_rate": 3.0763719218333545e-08,
+      "loss": 0.4837,
+      "step": 11716
+    },
+    {
+      "epoch": 0.9515958742792171,
+      "grad_norm": 3.926885842127398,
+      "learning_rate": 3.066095550407544e-08,
+      "loss": 0.4451,
+      "step": 11717
+    },
+    {
+      "epoch": 0.9516770892552586,
+      "grad_norm": 5.510721045832869,
+      "learning_rate": 3.0558362655010443e-08,
+      "loss": 0.5284,
+      "step": 11718
+    },
+    {
+      "epoch": 0.9517583042313003,
+      "grad_norm": 4.815199357131891,
+      "learning_rate": 3.045594067823704e-08,
+      "loss": 0.4785,
+      "step": 11719
+    },
+    {
+      "epoch": 0.9518395192073419,
+      "grad_norm": 9.394045320573086,
+      "learning_rate": 3.0353689580843174e-08,
+      "loss": 0.5275,
+      "step": 11720
+    },
+    {
+      "epoch": 0.9519207341833834,
+      "grad_norm": 7.043341703763583,
+      "learning_rate": 3.025160936990318e-08,
+      "loss": 0.4034,
+      "step": 11721
+    },
+    {
+      "epoch": 0.952001949159425,
+      "grad_norm": 5.105234133065711,
+      "learning_rate": 3.0149700052481135e-08,
+      "loss": 0.4658,
+      "step": 11722
+    },
+    {
+      "epoch": 0.9520831641354666,
+      "grad_norm": 20.843773029070668,
+      "learning_rate": 3.004796163562834e-08,
+      "loss": 0.4,
+      "step": 11723
+    },
+    {
+      "epoch": 0.9521643791115082,
+      "grad_norm": 4.834619084091482,
+      "learning_rate": 2.994639412638445e-08,
+      "loss": 0.4094,
+      "step": 11724
+    },
+    {
+      "epoch": 0.9522455940875497,
+      "grad_norm": 5.958564630244604,
+      "learning_rate": 2.984499753177772e-08,
+      "loss": 0.4347,
+      "step": 11725
+    },
+    {
+      "epoch": 0.9523268090635914,
+      "grad_norm": 7.525314088951139,
+      "learning_rate": 2.9743771858823657e-08,
+      "loss": 0.3909,
+      "step": 11726
+    },
+    {
+      "epoch": 0.9524080240396329,
+      "grad_norm": 4.660331555963379,
+      "learning_rate": 2.9642717114527208e-08,
+      "loss": 0.4202,
+      "step": 11727
+    },
+    {
+      "epoch": 0.9524892390156745,
+      "grad_norm": 4.986110534638319,
+      "learning_rate": 2.9541833305880287e-08,
+      "loss": 0.3993,
+      "step": 11728
+    },
+    {
+      "epoch": 0.952570453991716,
+      "grad_norm": 3.464614961858637,
+      "learning_rate": 2.9441120439864246e-08,
+      "loss": 0.5475,
+      "step": 11729
+    },
+    {
+      "epoch": 0.9526516689677577,
+      "grad_norm": 21.551152924357776,
+      "learning_rate": 2.9340578523447127e-08,
+      "loss": 0.4057,
+      "step": 11730
+    },
+    {
+      "epoch": 0.9527328839437993,
+      "grad_norm": 3.6579610094520283,
+      "learning_rate": 2.9240207563586142e-08,
+      "loss": 0.5294,
+      "step": 11731
+    },
+    {
+      "epoch": 0.9528140989198408,
+      "grad_norm": 5.768773438976465,
+      "learning_rate": 2.914000756722657e-08,
+      "loss": 0.4105,
+      "step": 11732
+    },
+    {
+      "epoch": 0.9528953138958824,
+      "grad_norm": 6.3584167885272285,
+      "learning_rate": 2.903997854130147e-08,
+      "loss": 0.5214,
+      "step": 11733
+    },
+    {
+      "epoch": 0.952976528871924,
+      "grad_norm": 7.692904252140693,
+      "learning_rate": 2.8940120492732537e-08,
+      "loss": 0.4954,
+      "step": 11734
+    },
+    {
+      "epoch": 0.9530577438479656,
+      "grad_norm": 5.655701924850993,
+      "learning_rate": 2.8840433428429514e-08,
+      "loss": 0.7178,
+      "step": 11735
+    },
+    {
+      "epoch": 0.9531389588240071,
+      "grad_norm": 4.970188795620924,
+      "learning_rate": 2.8740917355290222e-08,
+      "loss": 0.5701,
+      "step": 11736
+    },
+    {
+      "epoch": 0.9532201738000488,
+      "grad_norm": 5.290922267712118,
+      "learning_rate": 2.864157228019998e-08,
+      "loss": 0.4525,
+      "step": 11737
+    },
+    {
+      "epoch": 0.9533013887760903,
+      "grad_norm": 4.627627408383265,
+      "learning_rate": 2.854239821003385e-08,
+      "loss": 0.4818,
+      "step": 11738
+    },
+    {
+      "epoch": 0.9533826037521319,
+      "grad_norm": 7.8256164051009325,
+      "learning_rate": 2.8443395151653562e-08,
+      "loss": 0.5673,
+      "step": 11739
+    },
+    {
+      "epoch": 0.9534638187281734,
+      "grad_norm": 6.249505047993975,
+      "learning_rate": 2.834456311190975e-08,
+      "loss": 0.3818,
+      "step": 11740
+    },
+    {
+      "epoch": 0.9535450337042151,
+      "grad_norm": 6.005374198159657,
+      "learning_rate": 2.8245902097641388e-08,
+      "loss": 0.6104,
+      "step": 11741
+    },
+    {
+      "epoch": 0.9536262486802567,
+      "grad_norm": 5.110767892723842,
+      "learning_rate": 2.8147412115674955e-08,
+      "loss": 0.4915,
+      "step": 11742
+    },
+    {
+      "epoch": 0.9537074636562982,
+      "grad_norm": 4.5731007153331955,
+      "learning_rate": 2.8049093172825282e-08,
+      "loss": 0.4256,
+      "step": 11743
+    },
+    {
+      "epoch": 0.9537886786323398,
+      "grad_norm": 10.106582363128767,
+      "learning_rate": 2.795094527589609e-08,
+      "loss": 0.4245,
+      "step": 11744
+    },
+    {
+      "epoch": 0.9538698936083814,
+      "grad_norm": 4.273795431273181,
+      "learning_rate": 2.7852968431678064e-08,
+      "loss": 0.4379,
+      "step": 11745
+    },
+    {
+      "epoch": 0.953951108584423,
+      "grad_norm": 4.307573087745445,
+      "learning_rate": 2.7755162646950773e-08,
+      "loss": 0.6883,
+      "step": 11746
+    },
+    {
+      "epoch": 0.9540323235604645,
+      "grad_norm": 5.5704933532695105,
+      "learning_rate": 2.7657527928482418e-08,
+      "loss": 0.503,
+      "step": 11747
+    },
+    {
+      "epoch": 0.9541135385365062,
+      "grad_norm": 5.927129889565711,
+      "learning_rate": 2.756006428302843e-08,
+      "loss": 0.5112,
+      "step": 11748
+    },
+    {
+      "epoch": 0.9541947535125477,
+      "grad_norm": 7.739005076724821,
+      "learning_rate": 2.746277171733258e-08,
+      "loss": 0.4949,
+      "step": 11749
+    },
+    {
+      "epoch": 0.9542759684885893,
+      "grad_norm": 4.857345101026582,
+      "learning_rate": 2.736565023812754e-08,
+      "loss": 0.4114,
+      "step": 11750
+    },
+    {
+      "epoch": 0.9543571834646308,
+      "grad_norm": 5.671784265799197,
+      "learning_rate": 2.726869985213293e-08,
+      "loss": 0.4542,
+      "step": 11751
+    },
+    {
+      "epoch": 0.9544383984406725,
+      "grad_norm": 4.542817164476761,
+      "learning_rate": 2.717192056605783e-08,
+      "loss": 0.4281,
+      "step": 11752
+    },
+    {
+      "epoch": 0.9545196134167141,
+      "grad_norm": 4.657810531667513,
+      "learning_rate": 2.7075312386598274e-08,
+      "loss": 0.5609,
+      "step": 11753
+    },
+    {
+      "epoch": 0.9546008283927556,
+      "grad_norm": 4.663673614763122,
+      "learning_rate": 2.697887532043947e-08,
+      "loss": 0.4775,
+      "step": 11754
+    },
+    {
+      "epoch": 0.9546820433687973,
+      "grad_norm": 5.2321773718184765,
+      "learning_rate": 2.688260937425413e-08,
+      "loss": 0.5095,
+      "step": 11755
+    },
+    {
+      "epoch": 0.9547632583448388,
+      "grad_norm": 4.5112988407221835,
+      "learning_rate": 2.67865145547036e-08,
+      "loss": 0.4052,
+      "step": 11756
+    },
+    {
+      "epoch": 0.9548444733208804,
+      "grad_norm": 4.994296884071479,
+      "learning_rate": 2.6690590868436728e-08,
+      "loss": 0.5502,
+      "step": 11757
+    },
+    {
+      "epoch": 0.9549256882969219,
+      "grad_norm": 5.907644955879376,
+      "learning_rate": 2.6594838322091255e-08,
+      "loss": 0.5292,
+      "step": 11758
+    },
+    {
+      "epoch": 0.9550069032729636,
+      "grad_norm": 4.76086278575601,
+      "learning_rate": 2.6499256922292715e-08,
+      "loss": 0.532,
+      "step": 11759
+    },
+    {
+      "epoch": 0.9550881182490051,
+      "grad_norm": 10.420115891569926,
+      "learning_rate": 2.640384667565471e-08,
+      "loss": 0.5241,
+      "step": 11760
+    },
+    {
+      "epoch": 0.9551693332250467,
+      "grad_norm": 8.326204967482434,
+      "learning_rate": 2.6308607588779177e-08,
+      "loss": 0.3722,
+      "step": 11761
+    },
+    {
+      "epoch": 0.9552505482010882,
+      "grad_norm": 4.358771745047045,
+      "learning_rate": 2.6213539668256126e-08,
+      "loss": 0.4494,
+      "step": 11762
+    },
+    {
+      "epoch": 0.9553317631771299,
+      "grad_norm": 3.6696692614668676,
+      "learning_rate": 2.6118642920663906e-08,
+      "loss": 0.4823,
+      "step": 11763
+    },
+    {
+      "epoch": 0.9554129781531715,
+      "grad_norm": 3.432418349481183,
+      "learning_rate": 2.6023917352568652e-08,
+      "loss": 0.4218,
+      "step": 11764
+    },
+    {
+      "epoch": 0.955494193129213,
+      "grad_norm": 5.4984450992011515,
+      "learning_rate": 2.592936297052512e-08,
+      "loss": 0.5265,
+      "step": 11765
+    },
+    {
+      "epoch": 0.9555754081052547,
+      "grad_norm": 4.899194056779067,
+      "learning_rate": 2.5834979781075854e-08,
+      "loss": 0.3908,
+      "step": 11766
+    },
+    {
+      "epoch": 0.9556566230812962,
+      "grad_norm": 3.7039379030025317,
+      "learning_rate": 2.5740767790751463e-08,
+      "loss": 0.3868,
+      "step": 11767
+    },
+    {
+      "epoch": 0.9557378380573378,
+      "grad_norm": 7.5200178859973725,
+      "learning_rate": 2.5646727006071182e-08,
+      "loss": 0.4753,
+      "step": 11768
+    },
+    {
+      "epoch": 0.9558190530333793,
+      "grad_norm": 5.533797272476705,
+      "learning_rate": 2.55528574335423e-08,
+      "loss": 0.5694,
+      "step": 11769
+    },
+    {
+      "epoch": 0.955900268009421,
+      "grad_norm": 4.804791486887786,
+      "learning_rate": 2.5459159079659625e-08,
+      "loss": 0.4299,
+      "step": 11770
+    },
+    {
+      "epoch": 0.9559814829854625,
+      "grad_norm": 5.494381293218291,
+      "learning_rate": 2.5365631950906856e-08,
+      "loss": 0.2743,
+      "step": 11771
+    },
+    {
+      "epoch": 0.9560626979615041,
+      "grad_norm": 5.394100187529947,
+      "learning_rate": 2.5272276053755207e-08,
+      "loss": 0.5166,
+      "step": 11772
+    },
+    {
+      "epoch": 0.9561439129375456,
+      "grad_norm": 3.979565293315754,
+      "learning_rate": 2.5179091394665346e-08,
+      "loss": 0.3647,
+      "step": 11773
+    },
+    {
+      "epoch": 0.9562251279135873,
+      "grad_norm": 5.9411168688277245,
+      "learning_rate": 2.5086077980084057e-08,
+      "loss": 0.4242,
+      "step": 11774
+    },
+    {
+      "epoch": 0.9563063428896289,
+      "grad_norm": 6.4302270705636975,
+      "learning_rate": 2.4993235816448136e-08,
+      "loss": 0.4753,
+      "step": 11775
+    },
+    {
+      "epoch": 0.9563875578656704,
+      "grad_norm": 4.533132653222595,
+      "learning_rate": 2.4900564910181334e-08,
+      "loss": 0.4724,
+      "step": 11776
+    },
+    {
+      "epoch": 0.9564687728417121,
+      "grad_norm": 6.793870761327284,
+      "learning_rate": 2.4808065267696303e-08,
+      "loss": 0.3712,
+      "step": 11777
+    },
+    {
+      "epoch": 0.9565499878177536,
+      "grad_norm": 6.118222175954092,
+      "learning_rate": 2.4715736895393195e-08,
+      "loss": 0.4916,
+      "step": 11778
+    },
+    {
+      "epoch": 0.9566312027937952,
+      "grad_norm": 6.201932383609763,
+      "learning_rate": 2.462357979966107e-08,
+      "loss": 0.4406,
+      "step": 11779
+    },
+    {
+      "epoch": 0.9567124177698367,
+      "grad_norm": 5.730634163072851,
+      "learning_rate": 2.453159398687649e-08,
+      "loss": 0.47,
+      "step": 11780
+    },
+    {
+      "epoch": 0.9567936327458784,
+      "grad_norm": 5.240490008736156,
+      "learning_rate": 2.443977946340409e-08,
+      "loss": 0.5144,
+      "step": 11781
+    },
+    {
+      "epoch": 0.9568748477219199,
+      "grad_norm": 2.6161093483379783,
+      "learning_rate": 2.4348136235597398e-08,
+      "loss": 0.4603,
+      "step": 11782
+    },
+    {
+      "epoch": 0.9569560626979615,
+      "grad_norm": 4.15867662957049,
+      "learning_rate": 2.425666430979773e-08,
+      "loss": 0.5706,
+      "step": 11783
+    },
+    {
+      "epoch": 0.957037277674003,
+      "grad_norm": 9.450975853907856,
+      "learning_rate": 2.416536369233391e-08,
+      "loss": 0.477,
+      "step": 11784
+    },
+    {
+      "epoch": 0.9571184926500447,
+      "grad_norm": 7.347938995622321,
+      "learning_rate": 2.4074234389523665e-08,
+      "loss": 0.502,
+      "step": 11785
+    },
+    {
+      "epoch": 0.9571997076260863,
+      "grad_norm": 4.253055407962608,
+      "learning_rate": 2.3983276407672784e-08,
+      "loss": 0.4742,
+      "step": 11786
+    },
+    {
+      "epoch": 0.9572809226021278,
+      "grad_norm": 3.0805970942274135,
+      "learning_rate": 2.389248975307512e-08,
+      "loss": 0.6164,
+      "step": 11787
+    },
+    {
+      "epoch": 0.9573621375781695,
+      "grad_norm": 5.749655956883042,
+      "learning_rate": 2.3801874432012594e-08,
+      "loss": 0.5219,
+      "step": 11788
+    },
+    {
+      "epoch": 0.957443352554211,
+      "grad_norm": 4.8340592635662345,
+      "learning_rate": 2.371143045075519e-08,
+      "loss": 0.5919,
+      "step": 11789
+    },
+    {
+      "epoch": 0.9575245675302526,
+      "grad_norm": 3.8977820934124066,
+      "learning_rate": 2.3621157815561237e-08,
+      "loss": 0.6845,
+      "step": 11790
+    },
+    {
+      "epoch": 0.9576057825062941,
+      "grad_norm": 9.678166602329553,
+      "learning_rate": 2.3531056532677122e-08,
+      "loss": 0.5546,
+      "step": 11791
+    },
+    {
+      "epoch": 0.9576869974823358,
+      "grad_norm": 7.305637570872725,
+      "learning_rate": 2.3441126608337304e-08,
+      "loss": 0.455,
+      "step": 11792
+    },
+    {
+      "epoch": 0.9577682124583773,
+      "grad_norm": 3.8733070193102384,
+      "learning_rate": 2.335136804876459e-08,
+      "loss": 0.5323,
+      "step": 11793
+    },
+    {
+      "epoch": 0.9578494274344189,
+      "grad_norm": 3.9403704900752636,
+      "learning_rate": 2.3261780860169558e-08,
+      "loss": 0.4325,
+      "step": 11794
+    },
+    {
+      "epoch": 0.9579306424104604,
+      "grad_norm": 5.6576542673173575,
+      "learning_rate": 2.31723650487517e-08,
+      "loss": 0.4517,
+      "step": 11795
+    },
+    {
+      "epoch": 0.9580118573865021,
+      "grad_norm": 6.1828309207281755,
+      "learning_rate": 2.3083120620697453e-08,
+      "loss": 0.3581,
+      "step": 11796
+    },
+    {
+      "epoch": 0.9580930723625437,
+      "grad_norm": 4.945073828835356,
+      "learning_rate": 2.2994047582182433e-08,
+      "loss": 0.3434,
+      "step": 11797
+    },
+    {
+      "epoch": 0.9581742873385852,
+      "grad_norm": 6.8876285655205285,
+      "learning_rate": 2.2905145939369765e-08,
+      "loss": 0.4171,
+      "step": 11798
+    },
+    {
+      "epoch": 0.9582555023146269,
+      "grad_norm": 5.814280241290536,
+      "learning_rate": 2.2816415698411475e-08,
+      "loss": 0.518,
+      "step": 11799
+    },
+    {
+      "epoch": 0.9583367172906684,
+      "grad_norm": 6.165510069966149,
+      "learning_rate": 2.272785686544682e-08,
+      "loss": 0.4784,
+      "step": 11800
+    },
+    {
+      "epoch": 0.95841793226671,
+      "grad_norm": 5.755892898880744,
+      "learning_rate": 2.263946944660367e-08,
+      "loss": 0.4269,
+      "step": 11801
+    },
+    {
+      "epoch": 0.9584991472427515,
+      "grad_norm": 10.987936416240586,
+      "learning_rate": 2.2551253447997968e-08,
+      "loss": 0.3444,
+      "step": 11802
+    },
+    {
+      "epoch": 0.9585803622187932,
+      "grad_norm": 15.074930100573221,
+      "learning_rate": 2.2463208875733723e-08,
+      "loss": 0.5635,
+      "step": 11803
+    },
+    {
+      "epoch": 0.9586615771948347,
+      "grad_norm": 4.1773061107164935,
+      "learning_rate": 2.237533573590328e-08,
+      "loss": 0.5389,
+      "step": 11804
+    },
+    {
+      "epoch": 0.9587427921708763,
+      "grad_norm": 4.63238528075123,
+      "learning_rate": 2.228763403458706e-08,
+      "loss": 0.6207,
+      "step": 11805
+    },
+    {
+      "epoch": 0.9588240071469178,
+      "grad_norm": 5.9729583458173945,
+      "learning_rate": 2.2200103777853255e-08,
+      "loss": 0.5594,
+      "step": 11806
+    },
+    {
+      "epoch": 0.9589052221229595,
+      "grad_norm": 9.831624315329341,
+      "learning_rate": 2.211274497175897e-08,
+      "loss": 0.58,
+      "step": 11807
+    },
+    {
+      "epoch": 0.9589864370990011,
+      "grad_norm": 6.030561079423977,
+      "learning_rate": 2.2025557622348537e-08,
+      "loss": 0.5508,
+      "step": 11808
+    },
+    {
+      "epoch": 0.9590676520750426,
+      "grad_norm": 5.732610023291474,
+      "learning_rate": 2.1938541735655183e-08,
+      "loss": 0.5854,
+      "step": 11809
+    },
+    {
+      "epoch": 0.9591488670510843,
+      "grad_norm": 16.34752832829806,
+      "learning_rate": 2.1851697317699373e-08,
+      "loss": 0.3753,
+      "step": 11810
+    },
+    {
+      "epoch": 0.9592300820271258,
+      "grad_norm": 8.516533326876681,
+      "learning_rate": 2.1765024374491018e-08,
+      "loss": 0.4773,
+      "step": 11811
+    },
+    {
+      "epoch": 0.9593112970031674,
+      "grad_norm": 4.534008961064627,
+      "learning_rate": 2.1678522912026988e-08,
+      "loss": 0.4447,
+      "step": 11812
+    },
+    {
+      "epoch": 0.9593925119792089,
+      "grad_norm": 5.048729203766718,
+      "learning_rate": 2.1592192936292777e-08,
+      "loss": 0.4867,
+      "step": 11813
+    },
+    {
+      "epoch": 0.9594737269552506,
+      "grad_norm": 5.401071455659973,
+      "learning_rate": 2.1506034453262214e-08,
+      "loss": 0.4589,
+      "step": 11814
+    },
+    {
+      "epoch": 0.9595549419312921,
+      "grad_norm": 4.550782933767591,
+      "learning_rate": 2.142004746889692e-08,
+      "loss": 0.4399,
+      "step": 11815
+    },
+    {
+      "epoch": 0.9596361569073337,
+      "grad_norm": 7.657034015101817,
+      "learning_rate": 2.1334231989146304e-08,
+      "loss": 0.5894,
+      "step": 11816
+    },
+    {
+      "epoch": 0.9597173718833752,
+      "grad_norm": 4.68873295706115,
+      "learning_rate": 2.124858801994867e-08,
+      "loss": 0.5836,
+      "step": 11817
+    },
+    {
+      "epoch": 0.9597985868594169,
+      "grad_norm": 5.932018957670042,
+      "learning_rate": 2.1163115567230386e-08,
+      "loss": 0.6573,
+      "step": 11818
+    },
+    {
+      "epoch": 0.9598798018354585,
+      "grad_norm": 6.610287249201689,
+      "learning_rate": 2.1077814636905337e-08,
+      "loss": 0.5133,
+      "step": 11819
+    },
+    {
+      "epoch": 0.9599610168115,
+      "grad_norm": 4.576678178324454,
+      "learning_rate": 2.099268523487602e-08,
+      "loss": 0.406,
+      "step": 11820
+    },
+    {
+      "epoch": 0.9600422317875417,
+      "grad_norm": 5.480905722183299,
+      "learning_rate": 2.0907727367033005e-08,
+      "loss": 0.5465,
+      "step": 11821
+    },
+    {
+      "epoch": 0.9601234467635832,
+      "grad_norm": 10.7296131406781,
+      "learning_rate": 2.0822941039254642e-08,
+      "loss": 0.5529,
+      "step": 11822
+    },
+    {
+      "epoch": 0.9602046617396248,
+      "grad_norm": 4.499843737247379,
+      "learning_rate": 2.0738326257407904e-08,
+      "loss": 0.4012,
+      "step": 11823
+    },
+    {
+      "epoch": 0.9602858767156663,
+      "grad_norm": 5.585081499968746,
+      "learning_rate": 2.0653883027347832e-08,
+      "loss": 0.4078,
+      "step": 11824
+    },
+    {
+      "epoch": 0.960367091691708,
+      "grad_norm": 5.781778186861071,
+      "learning_rate": 2.056961135491725e-08,
+      "loss": 0.4095,
+      "step": 11825
+    },
+    {
+      "epoch": 0.9604483066677495,
+      "grad_norm": 5.2006226427597415,
+      "learning_rate": 2.048551124594733e-08,
+      "loss": 0.5062,
+      "step": 11826
+    },
+    {
+      "epoch": 0.9605295216437911,
+      "grad_norm": 9.000238161766775,
+      "learning_rate": 2.0401582706257304e-08,
+      "loss": 0.4405,
+      "step": 11827
+    },
+    {
+      "epoch": 0.9606107366198327,
+      "grad_norm": 5.406010740569412,
+      "learning_rate": 2.031782574165475e-08,
+      "loss": 0.5036,
+      "step": 11828
+    },
+    {
+      "epoch": 0.9606919515958743,
+      "grad_norm": 5.004999912991991,
+      "learning_rate": 2.0234240357935032e-08,
+      "loss": 0.4662,
+      "step": 11829
+    },
+    {
+      "epoch": 0.9607731665719159,
+      "grad_norm": 4.837946844168023,
+      "learning_rate": 2.015082656088213e-08,
+      "loss": 0.496,
+      "step": 11830
+    },
+    {
+      "epoch": 0.9608543815479574,
+      "grad_norm": 6.627370947037891,
+      "learning_rate": 2.0067584356267545e-08,
+      "loss": 0.375,
+      "step": 11831
+    },
+    {
+      "epoch": 0.9609355965239991,
+      "grad_norm": 10.674671755581318,
+      "learning_rate": 1.998451374985111e-08,
+      "loss": 0.5067,
+      "step": 11832
+    },
+    {
+      "epoch": 0.9610168115000406,
+      "grad_norm": 17.13229424192217,
+      "learning_rate": 1.9901614747381004e-08,
+      "loss": 0.3741,
+      "step": 11833
+    },
+    {
+      "epoch": 0.9610980264760822,
+      "grad_norm": 6.635134277566872,
+      "learning_rate": 1.981888735459375e-08,
+      "loss": 0.5692,
+      "step": 11834
+    },
+    {
+      "epoch": 0.9611792414521237,
+      "grad_norm": 3.8371011025366055,
+      "learning_rate": 1.973633157721283e-08,
+      "loss": 0.5985,
+      "step": 11835
+    },
+    {
+      "epoch": 0.9612604564281654,
+      "grad_norm": 5.13228830947514,
+      "learning_rate": 1.9653947420951448e-08,
+      "loss": 0.492,
+      "step": 11836
+    },
+    {
+      "epoch": 0.9613416714042069,
+      "grad_norm": 6.733892581752105,
+      "learning_rate": 1.9571734891509763e-08,
+      "loss": 0.4515,
+      "step": 11837
+    },
+    {
+      "epoch": 0.9614228863802485,
+      "grad_norm": 5.22909116303276,
+      "learning_rate": 1.9489693994576563e-08,
+      "loss": 0.6163,
+      "step": 11838
+    },
+    {
+      "epoch": 0.96150410135629,
+      "grad_norm": 4.302408834520267,
+      "learning_rate": 1.9407824735828696e-08,
+      "loss": 0.4538,
+      "step": 11839
+    },
+    {
+      "epoch": 0.9615853163323317,
+      "grad_norm": 5.043172467524965,
+      "learning_rate": 1.932612712093107e-08,
+      "loss": 0.4024,
+      "step": 11840
+    },
+    {
+      "epoch": 0.9616665313083733,
+      "grad_norm": 8.816715558441784,
+      "learning_rate": 1.9244601155536392e-08,
+      "loss": 0.4981,
+      "step": 11841
+    },
+    {
+      "epoch": 0.9617477462844148,
+      "grad_norm": 4.8079654103051475,
+      "learning_rate": 1.9163246845286253e-08,
+      "loss": 0.4698,
+      "step": 11842
+    },
+    {
+      "epoch": 0.9618289612604565,
+      "grad_norm": 6.163595049432749,
+      "learning_rate": 1.908206419580977e-08,
+      "loss": 0.5798,
+      "step": 11843
+    },
+    {
+      "epoch": 0.961910176236498,
+      "grad_norm": 4.866986542895558,
+      "learning_rate": 1.9001053212724387e-08,
+      "loss": 0.5856,
+      "step": 11844
+    },
+    {
+      "epoch": 0.9619913912125396,
+      "grad_norm": 5.782563129780745,
+      "learning_rate": 1.892021390163562e-08,
+      "loss": 0.4907,
+      "step": 11845
+    },
+    {
+      "epoch": 0.9620726061885811,
+      "grad_norm": 5.712857730829221,
+      "learning_rate": 1.8839546268137054e-08,
+      "loss": 0.4423,
+      "step": 11846
+    },
+    {
+      "epoch": 0.9621538211646228,
+      "grad_norm": 4.414806524066914,
+      "learning_rate": 1.8759050317810612e-08,
+      "loss": 0.4499,
+      "step": 11847
+    },
+    {
+      "epoch": 0.9622350361406643,
+      "grad_norm": 4.949387208955272,
+      "learning_rate": 1.8678726056226004e-08,
+      "loss": 0.4297,
+      "step": 11848
+    },
+    {
+      "epoch": 0.9623162511167059,
+      "grad_norm": 4.396280286867585,
+      "learning_rate": 1.8598573488941285e-08,
+      "loss": 0.5549,
+      "step": 11849
+    },
+    {
+      "epoch": 0.9623974660927475,
+      "grad_norm": 6.050149072377,
+      "learning_rate": 1.8518592621502852e-08,
+      "loss": 0.5692,
+      "step": 11850
+    },
+    {
+      "epoch": 0.9624786810687891,
+      "grad_norm": 5.0620158081409246,
+      "learning_rate": 1.8438783459444608e-08,
+      "loss": 0.5378,
+      "step": 11851
+    },
+    {
+      "epoch": 0.9625598960448307,
+      "grad_norm": 3.929260770758667,
+      "learning_rate": 1.8359146008289087e-08,
+      "loss": 0.5249,
+      "step": 11852
+    },
+    {
+      "epoch": 0.9626411110208722,
+      "grad_norm": 6.751642731964542,
+      "learning_rate": 1.8279680273546874e-08,
+      "loss": 0.3583,
+      "step": 11853
+    },
+    {
+      "epoch": 0.9627223259969139,
+      "grad_norm": 7.566915713105148,
+      "learning_rate": 1.8200386260716352e-08,
+      "loss": 0.5411,
+      "step": 11854
+    },
+    {
+      "epoch": 0.9628035409729554,
+      "grad_norm": 4.154405022428572,
+      "learning_rate": 1.812126397528452e-08,
+      "loss": 0.5512,
+      "step": 11855
+    },
+    {
+      "epoch": 0.962884755948997,
+      "grad_norm": 7.42761465533202,
+      "learning_rate": 1.804231342272589e-08,
+      "loss": 0.5403,
+      "step": 11856
+    },
+    {
+      "epoch": 0.9629659709250386,
+      "grad_norm": 7.694593764622218,
+      "learning_rate": 1.796353460850331e-08,
+      "loss": 0.6331,
+      "step": 11857
+    },
+    {
+      "epoch": 0.9630471859010802,
+      "grad_norm": 16.086802457999514,
+      "learning_rate": 1.7884927538068532e-08,
+      "loss": 0.4567,
+      "step": 11858
+    },
+    {
+      "epoch": 0.9631284008771217,
+      "grad_norm": 6.812124405495696,
+      "learning_rate": 1.7806492216860537e-08,
+      "loss": 0.5277,
+      "step": 11859
+    },
+    {
+      "epoch": 0.9632096158531633,
+      "grad_norm": 6.335913061374134,
+      "learning_rate": 1.77282286503061e-08,
+      "loss": 0.4104,
+      "step": 11860
+    },
+    {
+      "epoch": 0.9632908308292049,
+      "grad_norm": 5.290742984147669,
+      "learning_rate": 1.7650136843821163e-08,
+      "loss": 0.5395,
+      "step": 11861
+    },
+    {
+      "epoch": 0.9633720458052465,
+      "grad_norm": 7.397405822219676,
+      "learning_rate": 1.7572216802808907e-08,
+      "loss": 0.5077,
+      "step": 11862
+    },
+    {
+      "epoch": 0.9634532607812881,
+      "grad_norm": 5.393822284388817,
+      "learning_rate": 1.74944685326614e-08,
+      "loss": 0.447,
+      "step": 11863
+    },
+    {
+      "epoch": 0.9635344757573296,
+      "grad_norm": 4.001728592267749,
+      "learning_rate": 1.741689203875796e-08,
+      "loss": 0.5923,
+      "step": 11864
+    },
+    {
+      "epoch": 0.9636156907333713,
+      "grad_norm": 6.112171929702209,
+      "learning_rate": 1.7339487326466787e-08,
+      "loss": 0.5556,
+      "step": 11865
+    },
+    {
+      "epoch": 0.9636969057094128,
+      "grad_norm": 4.7616632757082575,
+      "learning_rate": 1.7262254401143873e-08,
+      "loss": 0.5483,
+      "step": 11866
+    },
+    {
+      "epoch": 0.9637781206854544,
+      "grad_norm": 5.833975112122491,
+      "learning_rate": 1.7185193268133282e-08,
+      "loss": 0.4426,
+      "step": 11867
+    },
+    {
+      "epoch": 0.963859335661496,
+      "grad_norm": 3.698744760296142,
+      "learning_rate": 1.7108303932767135e-08,
+      "loss": 0.4634,
+      "step": 11868
+    },
+    {
+      "epoch": 0.9639405506375376,
+      "grad_norm": 6.198330496132483,
+      "learning_rate": 1.7031586400365895e-08,
+      "loss": 0.4089,
+      "step": 11869
+    },
+    {
+      "epoch": 0.9640217656135791,
+      "grad_norm": 4.92658425213048,
+      "learning_rate": 1.695504067623782e-08,
+      "loss": 0.571,
+      "step": 11870
+    },
+    {
+      "epoch": 0.9641029805896207,
+      "grad_norm": 4.570520680219551,
+      "learning_rate": 1.6878666765679507e-08,
+      "loss": 0.4378,
+      "step": 11871
+    },
+    {
+      "epoch": 0.9641841955656623,
+      "grad_norm": 4.130327563075009,
+      "learning_rate": 1.6802464673975893e-08,
+      "loss": 0.4161,
+      "step": 11872
+    },
+    {
+      "epoch": 0.9642654105417039,
+      "grad_norm": 5.2911892010001775,
+      "learning_rate": 1.6726434406399704e-08,
+      "loss": 0.4437,
+      "step": 11873
+    },
+    {
+      "epoch": 0.9643466255177455,
+      "grad_norm": 7.92651230061654,
+      "learning_rate": 1.6650575968211458e-08,
+      "loss": 0.3474,
+      "step": 11874
+    },
+    {
+      "epoch": 0.964427840493787,
+      "grad_norm": 4.955382254039267,
+      "learning_rate": 1.6574889364660564e-08,
+      "loss": 0.4664,
+      "step": 11875
+    },
+    {
+      "epoch": 0.9645090554698287,
+      "grad_norm": 5.462298377112138,
+      "learning_rate": 1.6499374600983943e-08,
+      "loss": 0.4326,
+      "step": 11876
+    },
+    {
+      "epoch": 0.9645902704458702,
+      "grad_norm": 3.395717603402873,
+      "learning_rate": 1.642403168240686e-08,
+      "loss": 0.5458,
+      "step": 11877
+    },
+    {
+      "epoch": 0.9646714854219118,
+      "grad_norm": 6.599864821668275,
+      "learning_rate": 1.6348860614142646e-08,
+      "loss": 0.5629,
+      "step": 11878
+    },
+    {
+      "epoch": 0.9647527003979534,
+      "grad_norm": 3.790645168309616,
+      "learning_rate": 1.62738614013927e-08,
+      "loss": 0.511,
+      "step": 11879
+    },
+    {
+      "epoch": 0.964833915373995,
+      "grad_norm": 2.6841332503368918,
+      "learning_rate": 1.6199034049346474e-08,
+      "loss": 0.3197,
+      "step": 11880
+    },
+    {
+      "epoch": 0.9649151303500365,
+      "grad_norm": 8.916212568990264,
+      "learning_rate": 1.6124378563182053e-08,
+      "loss": 0.4072,
+      "step": 11881
+    },
+    {
+      "epoch": 0.9649963453260781,
+      "grad_norm": 6.5493673229576945,
+      "learning_rate": 1.6049894948064748e-08,
+      "loss": 0.5502,
+      "step": 11882
+    },
+    {
+      "epoch": 0.9650775603021197,
+      "grad_norm": 3.681237380060799,
+      "learning_rate": 1.597558320914849e-08,
+      "loss": 0.5055,
+      "step": 11883
+    },
+    {
+      "epoch": 0.9651587752781613,
+      "grad_norm": 5.868912226870429,
+      "learning_rate": 1.5901443351575563e-08,
+      "loss": 0.4587,
+      "step": 11884
+    },
+    {
+      "epoch": 0.9652399902542029,
+      "grad_norm": 3.216888608216173,
+      "learning_rate": 1.5827475380475744e-08,
+      "loss": 0.4016,
+      "step": 11885
+    },
+    {
+      "epoch": 0.9653212052302445,
+      "grad_norm": 6.658602355248441,
+      "learning_rate": 1.575367930096716e-08,
+      "loss": 0.6319,
+      "step": 11886
+    },
+    {
+      "epoch": 0.9654024202062861,
+      "grad_norm": 7.62706341624403,
+      "learning_rate": 1.5680055118156566e-08,
+      "loss": 0.4729,
+      "step": 11887
+    },
+    {
+      "epoch": 0.9654836351823276,
+      "grad_norm": 5.573341730209802,
+      "learning_rate": 1.5606602837137942e-08,
+      "loss": 0.3682,
+      "step": 11888
+    },
+    {
+      "epoch": 0.9655648501583692,
+      "grad_norm": 9.836559250337336,
+      "learning_rate": 1.5533322462993884e-08,
+      "loss": 0.3449,
+      "step": 11889
+    },
+    {
+      "epoch": 0.9656460651344108,
+      "grad_norm": 5.722705721907483,
+      "learning_rate": 1.546021400079506e-08,
+      "loss": 0.5415,
+      "step": 11890
+    },
+    {
+      "epoch": 0.9657272801104524,
+      "grad_norm": 13.038255642017731,
+      "learning_rate": 1.538727745560048e-08,
+      "loss": 0.5975,
+      "step": 11891
+    },
+    {
+      "epoch": 0.9658084950864939,
+      "grad_norm": 4.658075474008751,
+      "learning_rate": 1.5314512832456385e-08,
+      "loss": 0.3905,
+      "step": 11892
+    },
+    {
+      "epoch": 0.9658897100625355,
+      "grad_norm": 3.609827209694475,
+      "learning_rate": 1.5241920136397913e-08,
+      "loss": 0.6306,
+      "step": 11893
+    },
+    {
+      "epoch": 0.9659709250385771,
+      "grad_norm": 4.924247089415893,
+      "learning_rate": 1.516949937244827e-08,
+      "loss": 0.6758,
+      "step": 11894
+    },
+    {
+      "epoch": 0.9660521400146187,
+      "grad_norm": 7.844480475751513,
+      "learning_rate": 1.5097250545618447e-08,
+      "loss": 0.5907,
+      "step": 11895
+    },
+    {
+      "epoch": 0.9661333549906603,
+      "grad_norm": 7.5845268596408655,
+      "learning_rate": 1.5025173660907776e-08,
+      "loss": 0.3675,
+      "step": 11896
+    },
+    {
+      "epoch": 0.9662145699667019,
+      "grad_norm": 6.1316891281247425,
+      "learning_rate": 1.495326872330366e-08,
+      "loss": 0.5117,
+      "step": 11897
+    },
+    {
+      "epoch": 0.9662957849427435,
+      "grad_norm": 13.323521108000474,
+      "learning_rate": 1.4881535737781282e-08,
+      "loss": 0.2842,
+      "step": 11898
+    },
+    {
+      "epoch": 0.966376999918785,
+      "grad_norm": 6.657166105699875,
+      "learning_rate": 1.4809974709304176e-08,
+      "loss": 0.4965,
+      "step": 11899
+    },
+    {
+      "epoch": 0.9664582148948266,
+      "grad_norm": 4.572946832553962,
+      "learning_rate": 1.4738585642824488e-08,
+      "loss": 0.5393,
+      "step": 11900
+    },
+    {
+      "epoch": 0.9665394298708682,
+      "grad_norm": 5.179202728578802,
+      "learning_rate": 1.4667368543281324e-08,
+      "loss": 0.3911,
+      "step": 11901
+    },
+    {
+      "epoch": 0.9666206448469098,
+      "grad_norm": 3.397470952294324,
+      "learning_rate": 1.4596323415602965e-08,
+      "loss": 0.5457,
+      "step": 11902
+    },
+    {
+      "epoch": 0.9667018598229513,
+      "grad_norm": 5.014658354565728,
+      "learning_rate": 1.4525450264705198e-08,
+      "loss": 0.495,
+      "step": 11903
+    },
+    {
+      "epoch": 0.966783074798993,
+      "grad_norm": 4.876907477649273,
+      "learning_rate": 1.4454749095491883e-08,
+      "loss": 0.386,
+      "step": 11904
+    },
+    {
+      "epoch": 0.9668642897750345,
+      "grad_norm": 6.641652702622383,
+      "learning_rate": 1.438421991285549e-08,
+      "loss": 0.5017,
+      "step": 11905
+    },
+    {
+      "epoch": 0.9669455047510761,
+      "grad_norm": 6.852642511727459,
+      "learning_rate": 1.4313862721676285e-08,
+      "loss": 0.5132,
+      "step": 11906
+    },
+    {
+      "epoch": 0.9670267197271177,
+      "grad_norm": 4.757471303632329,
+      "learning_rate": 1.4243677526822319e-08,
+      "loss": 0.5419,
+      "step": 11907
+    },
+    {
+      "epoch": 0.9671079347031593,
+      "grad_norm": 7.349401986982847,
+      "learning_rate": 1.4173664333149983e-08,
+      "loss": 0.4691,
+      "step": 11908
+    },
+    {
+      "epoch": 0.9671891496792009,
+      "grad_norm": 4.683264212562743,
+      "learning_rate": 1.4103823145504292e-08,
+      "loss": 0.396,
+      "step": 11909
+    },
+    {
+      "epoch": 0.9672703646552424,
+      "grad_norm": 8.004000185912112,
+      "learning_rate": 1.4034153968717768e-08,
+      "loss": 0.5567,
+      "step": 11910
+    },
+    {
+      "epoch": 0.967351579631284,
+      "grad_norm": 9.131569052331146,
+      "learning_rate": 1.3964656807610721e-08,
+      "loss": 0.4341,
+      "step": 11911
+    },
+    {
+      "epoch": 0.9674327946073256,
+      "grad_norm": 4.584784368488165,
+      "learning_rate": 1.3895331666992361e-08,
+      "loss": 0.5043,
+      "step": 11912
+    },
+    {
+      "epoch": 0.9675140095833672,
+      "grad_norm": 5.0665939824517165,
+      "learning_rate": 1.3826178551659686e-08,
+      "loss": 0.4195,
+      "step": 11913
+    },
+    {
+      "epoch": 0.9675952245594087,
+      "grad_norm": 4.105167424178751,
+      "learning_rate": 1.37571974663972e-08,
+      "loss": 0.4746,
+      "step": 11914
+    },
+    {
+      "epoch": 0.9676764395354504,
+      "grad_norm": 7.389787805986678,
+      "learning_rate": 1.3688388415978581e-08,
+      "loss": 0.5362,
+      "step": 11915
+    },
+    {
+      "epoch": 0.9677576545114919,
+      "grad_norm": 5.680461317779846,
+      "learning_rate": 1.361975140516475e-08,
+      "loss": 0.5635,
+      "step": 11916
+    },
+    {
+      "epoch": 0.9678388694875335,
+      "grad_norm": 3.8294537161607476,
+      "learning_rate": 1.3551286438705513e-08,
+      "loss": 0.517,
+      "step": 11917
+    },
+    {
+      "epoch": 0.9679200844635751,
+      "grad_norm": 5.093597645059693,
+      "learning_rate": 1.3482993521337362e-08,
+      "loss": 0.4311,
+      "step": 11918
+    },
+    {
+      "epoch": 0.9680012994396167,
+      "grad_norm": 4.15007656176377,
+      "learning_rate": 1.3414872657786793e-08,
+      "loss": 0.5585,
+      "step": 11919
+    },
+    {
+      "epoch": 0.9680825144156583,
+      "grad_norm": 5.618091315100742,
+      "learning_rate": 1.3346923852766702e-08,
+      "loss": 0.4277,
+      "step": 11920
+    },
+    {
+      "epoch": 0.9681637293916998,
+      "grad_norm": 3.77740666889434,
+      "learning_rate": 1.3279147110979163e-08,
+      "loss": 0.4445,
+      "step": 11921
+    },
+    {
+      "epoch": 0.9682449443677414,
+      "grad_norm": 14.059774837034583,
+      "learning_rate": 1.3211542437113755e-08,
+      "loss": 0.4987,
+      "step": 11922
+    },
+    {
+      "epoch": 0.968326159343783,
+      "grad_norm": 5.451347013670637,
+      "learning_rate": 1.3144109835848685e-08,
+      "loss": 0.4612,
+      "step": 11923
+    },
+    {
+      "epoch": 0.9684073743198246,
+      "grad_norm": 9.206132420636997,
+      "learning_rate": 1.3076849311849382e-08,
+      "loss": 0.5038,
+      "step": 11924
+    },
+    {
+      "epoch": 0.9684885892958661,
+      "grad_norm": 11.483754762302455,
+      "learning_rate": 1.300976086977046e-08,
+      "loss": 0.4655,
+      "step": 11925
+    },
+    {
+      "epoch": 0.9685698042719078,
+      "grad_norm": 6.743288245848484,
+      "learning_rate": 1.2942844514254038e-08,
+      "loss": 0.5264,
+      "step": 11926
+    },
+    {
+      "epoch": 0.9686510192479493,
+      "grad_norm": 6.053479065590689,
+      "learning_rate": 1.2876100249930024e-08,
+      "loss": 0.4499,
+      "step": 11927
+    },
+    {
+      "epoch": 0.9687322342239909,
+      "grad_norm": 3.6207902961878853,
+      "learning_rate": 1.2809528081416667e-08,
+      "loss": 0.536,
+      "step": 11928
+    },
+    {
+      "epoch": 0.9688134492000325,
+      "grad_norm": 6.866156397584753,
+      "learning_rate": 1.2743128013321115e-08,
+      "loss": 0.3062,
+      "step": 11929
+    },
+    {
+      "epoch": 0.9688946641760741,
+      "grad_norm": 5.924869562979613,
+      "learning_rate": 1.2676900050237472e-08,
+      "loss": 0.4935,
+      "step": 11930
+    },
+    {
+      "epoch": 0.9689758791521157,
+      "grad_norm": 5.999357098508331,
+      "learning_rate": 1.2610844196748184e-08,
+      "loss": 0.3896,
+      "step": 11931
+    },
+    {
+      "epoch": 0.9690570941281572,
+      "grad_norm": 24.091678245998175,
+      "learning_rate": 1.2544960457424316e-08,
+      "loss": 0.4804,
+      "step": 11932
+    },
+    {
+      "epoch": 0.9691383091041988,
+      "grad_norm": 4.839750186365689,
+      "learning_rate": 1.2479248836824165e-08,
+      "loss": 0.4291,
+      "step": 11933
+    },
+    {
+      "epoch": 0.9692195240802404,
+      "grad_norm": 5.821546388018869,
+      "learning_rate": 1.2413709339495205e-08,
+      "loss": 0.3908,
+      "step": 11934
+    },
+    {
+      "epoch": 0.969300739056282,
+      "grad_norm": 4.807004678084896,
+      "learning_rate": 1.2348341969972143e-08,
+      "loss": 0.461,
+      "step": 11935
+    },
+    {
+      "epoch": 0.9693819540323235,
+      "grad_norm": 11.290614300170317,
+      "learning_rate": 1.2283146732778306e-08,
+      "loss": 0.498,
+      "step": 11936
+    },
+    {
+      "epoch": 0.9694631690083652,
+      "grad_norm": 4.900177210666115,
+      "learning_rate": 1.2218123632424527e-08,
+      "loss": 0.5445,
+      "step": 11937
+    },
+    {
+      "epoch": 0.9695443839844067,
+      "grad_norm": 4.9843232783971665,
+      "learning_rate": 1.2153272673409989e-08,
+      "loss": 0.4344,
+      "step": 11938
+    },
+    {
+      "epoch": 0.9696255989604483,
+      "grad_norm": 9.329861274666198,
+      "learning_rate": 1.2088593860222487e-08,
+      "loss": 0.523,
+      "step": 11939
+    },
+    {
+      "epoch": 0.9697068139364899,
+      "grad_norm": 5.949707334731747,
+      "learning_rate": 1.2024087197337053e-08,
+      "loss": 0.4087,
+      "step": 11940
+    },
+    {
+      "epoch": 0.9697880289125315,
+      "grad_norm": 3.2312999306605934,
+      "learning_rate": 1.1959752689217342e-08,
+      "loss": 0.4933,
+      "step": 11941
+    },
+    {
+      "epoch": 0.9698692438885731,
+      "grad_norm": 4.579568524123334,
+      "learning_rate": 1.1895590340315343e-08,
+      "loss": 0.5973,
+      "step": 11942
+    },
+    {
+      "epoch": 0.9699504588646146,
+      "grad_norm": 4.04721463907932,
+      "learning_rate": 1.183160015507001e-08,
+      "loss": 0.5191,
+      "step": 11943
+    },
+    {
+      "epoch": 0.9700316738406563,
+      "grad_norm": 3.4507628169869564,
+      "learning_rate": 1.1767782137909467e-08,
+      "loss": 0.4671,
+      "step": 11944
+    },
+    {
+      "epoch": 0.9701128888166978,
+      "grad_norm": 4.199682308907727,
+      "learning_rate": 1.17041362932499e-08,
+      "loss": 0.5631,
+      "step": 11945
+    },
+    {
+      "epoch": 0.9701941037927394,
+      "grad_norm": 4.406904013569409,
+      "learning_rate": 1.1640662625494737e-08,
+      "loss": 0.5799,
+      "step": 11946
+    },
+    {
+      "epoch": 0.9702753187687809,
+      "grad_norm": 6.977075217955481,
+      "learning_rate": 1.1577361139036292e-08,
+      "loss": 0.5053,
+      "step": 11947
+    },
+    {
+      "epoch": 0.9703565337448226,
+      "grad_norm": 5.4502036515579295,
+      "learning_rate": 1.1514231838254674e-08,
+      "loss": 0.4786,
+      "step": 11948
+    },
+    {
+      "epoch": 0.9704377487208641,
+      "grad_norm": 11.48385762088185,
+      "learning_rate": 1.1451274727518058e-08,
+      "loss": 0.4975,
+      "step": 11949
+    },
+    {
+      "epoch": 0.9705189636969057,
+      "grad_norm": 4.167108685353742,
+      "learning_rate": 1.1388489811182957e-08,
+      "loss": 0.5192,
+      "step": 11950
+    },
+    {
+      "epoch": 0.9706001786729473,
+      "grad_norm": 5.2318458897159115,
+      "learning_rate": 1.1325877093593396e-08,
+      "loss": 0.5431,
+      "step": 11951
+    },
+    {
+      "epoch": 0.9706813936489889,
+      "grad_norm": 6.6175704623771505,
+      "learning_rate": 1.1263436579082022e-08,
+      "loss": 0.6257,
+      "step": 11952
+    },
+    {
+      "epoch": 0.9707626086250305,
+      "grad_norm": 3.4493770443599465,
+      "learning_rate": 1.1201168271969266e-08,
+      "loss": 0.5008,
+      "step": 11953
+    },
+    {
+      "epoch": 0.970843823601072,
+      "grad_norm": 9.823049761761046,
+      "learning_rate": 1.1139072176564181e-08,
+      "loss": 0.6227,
+      "step": 11954
+    },
+    {
+      "epoch": 0.9709250385771137,
+      "grad_norm": 12.037056297654933,
+      "learning_rate": 1.1077148297163053e-08,
+      "loss": 0.4436,
+      "step": 11955
+    },
+    {
+      "epoch": 0.9710062535531552,
+      "grad_norm": 47.76636023695995,
+      "learning_rate": 1.101539663805079e-08,
+      "loss": 0.44,
+      "step": 11956
+    },
+    {
+      "epoch": 0.9710874685291968,
+      "grad_norm": 5.334949326289944,
+      "learning_rate": 1.0953817203500084e-08,
+      "loss": 0.4165,
+      "step": 11957
+    },
+    {
+      "epoch": 0.9711686835052383,
+      "grad_norm": 9.47687334833281,
+      "learning_rate": 1.0892409997772524e-08,
+      "loss": 0.4979,
+      "step": 11958
+    },
+    {
+      "epoch": 0.97124989848128,
+      "grad_norm": 5.061927562596902,
+      "learning_rate": 1.0831175025116658e-08,
+      "loss": 0.4791,
+      "step": 11959
+    },
+    {
+      "epoch": 0.9713311134573215,
+      "grad_norm": 8.5596022285608,
+      "learning_rate": 1.0770112289769653e-08,
+      "loss": 0.5829,
+      "step": 11960
+    },
+    {
+      "epoch": 0.9714123284333631,
+      "grad_norm": 4.508772275279344,
+      "learning_rate": 1.0709221795956738e-08,
+      "loss": 0.5874,
+      "step": 11961
+    },
+    {
+      "epoch": 0.9714935434094047,
+      "grad_norm": 4.701424170250187,
+      "learning_rate": 1.0648503547891487e-08,
+      "loss": 0.4767,
+      "step": 11962
+    },
+    {
+      "epoch": 0.9715747583854463,
+      "grad_norm": 3.965969498170774,
+      "learning_rate": 1.0587957549774986e-08,
+      "loss": 0.5583,
+      "step": 11963
+    },
+    {
+      "epoch": 0.9716559733614879,
+      "grad_norm": 7.964781868944689,
+      "learning_rate": 1.052758380579666e-08,
+      "loss": 0.5368,
+      "step": 11964
+    },
+    {
+      "epoch": 0.9717371883375294,
+      "grad_norm": 5.249845918760695,
+      "learning_rate": 1.0467382320134279e-08,
+      "loss": 0.4113,
+      "step": 11965
+    },
+    {
+      "epoch": 0.971818403313571,
+      "grad_norm": 5.825530015816844,
+      "learning_rate": 1.0407353096953398e-08,
+      "loss": 0.3504,
+      "step": 11966
+    },
+    {
+      "epoch": 0.9718996182896126,
+      "grad_norm": 9.06281122515128,
+      "learning_rate": 1.034749614040792e-08,
+      "loss": 0.3699,
+      "step": 11967
+    },
+    {
+      "epoch": 0.9719808332656542,
+      "grad_norm": 7.498622251328044,
+      "learning_rate": 1.0287811454639252e-08,
+      "loss": 0.4268,
+      "step": 11968
+    },
+    {
+      "epoch": 0.9720620482416957,
+      "grad_norm": 6.263774598272872,
+      "learning_rate": 1.0228299043777146e-08,
+      "loss": 0.4959,
+      "step": 11969
+    },
+    {
+      "epoch": 0.9721432632177374,
+      "grad_norm": 5.061985215032839,
+      "learning_rate": 1.0168958911939975e-08,
+      "loss": 0.5099,
+      "step": 11970
+    },
+    {
+      "epoch": 0.9722244781937789,
+      "grad_norm": 8.537598436187121,
+      "learning_rate": 1.0109791063233898e-08,
+      "loss": 0.3745,
+      "step": 11971
+    },
+    {
+      "epoch": 0.9723056931698205,
+      "grad_norm": 3.710316240876283,
+      "learning_rate": 1.0050795501752309e-08,
+      "loss": 0.462,
+      "step": 11972
+    },
+    {
+      "epoch": 0.9723869081458621,
+      "grad_norm": 4.819345845034754,
+      "learning_rate": 9.991972231577774e-09,
+      "loss": 0.568,
+      "step": 11973
+    },
+    {
+      "epoch": 0.9724681231219037,
+      "grad_norm": 7.126395810841205,
+      "learning_rate": 9.933321256780925e-09,
+      "loss": 0.4745,
+      "step": 11974
+    },
+    {
+      "epoch": 0.9725493380979453,
+      "grad_norm": 4.57445118125883,
+      "learning_rate": 9.874842581419631e-09,
+      "loss": 0.3644,
+      "step": 11975
+    },
+    {
+      "epoch": 0.9726305530739868,
+      "grad_norm": 8.751285985895777,
+      "learning_rate": 9.816536209540373e-09,
+      "loss": 0.3933,
+      "step": 11976
+    },
+    {
+      "epoch": 0.9727117680500285,
+      "grad_norm": 6.666232456269325,
+      "learning_rate": 9.758402145177703e-09,
+      "loss": 0.499,
+      "step": 11977
+    },
+    {
+      "epoch": 0.97279298302607,
+      "grad_norm": 7.576271556752231,
+      "learning_rate": 9.70044039235396e-09,
+      "loss": 0.4075,
+      "step": 11978
+    },
+    {
+      "epoch": 0.9728741980021116,
+      "grad_norm": 4.487672059560348,
+      "learning_rate": 9.642650955080379e-09,
+      "loss": 0.3907,
+      "step": 11979
+    },
+    {
+      "epoch": 0.9729554129781531,
+      "grad_norm": 3.6526471218089864,
+      "learning_rate": 9.585033837355151e-09,
+      "loss": 0.5327,
+      "step": 11980
+    },
+    {
+      "epoch": 0.9730366279541948,
+      "grad_norm": 4.83213034119808,
+      "learning_rate": 9.527589043165086e-09,
+      "loss": 0.5155,
+      "step": 11981
+    },
+    {
+      "epoch": 0.9731178429302363,
+      "grad_norm": 4.370525558131413,
+      "learning_rate": 9.470316576485616e-09,
+      "loss": 0.4436,
+      "step": 11982
+    },
+    {
+      "epoch": 0.9731990579062779,
+      "grad_norm": 6.8962180265204065,
+      "learning_rate": 9.41321644127885e-09,
+      "loss": 0.4835,
+      "step": 11983
+    },
+    {
+      "epoch": 0.9732802728823196,
+      "grad_norm": 5.928084396311675,
+      "learning_rate": 9.356288641496624e-09,
+      "loss": 0.3822,
+      "step": 11984
+    },
+    {
+      "epoch": 0.9733614878583611,
+      "grad_norm": 5.61471404924446,
+      "learning_rate": 9.299533181077458e-09,
+      "loss": 0.36,
+      "step": 11985
+    },
+    {
+      "epoch": 0.9734427028344027,
+      "grad_norm": 4.928900628567097,
+      "learning_rate": 9.242950063948763e-09,
+      "loss": 0.4844,
+      "step": 11986
+    },
+    {
+      "epoch": 0.9735239178104442,
+      "grad_norm": 6.227465912824045,
+      "learning_rate": 9.18653929402602e-09,
+      "loss": 0.6315,
+      "step": 11987
+    },
+    {
+      "epoch": 0.9736051327864859,
+      "grad_norm": 5.256061365992016,
+      "learning_rate": 9.13030087521194e-09,
+      "loss": 0.5286,
+      "step": 11988
+    },
+    {
+      "epoch": 0.9736863477625274,
+      "grad_norm": 5.4504099987714465,
+      "learning_rate": 9.074234811398408e-09,
+      "loss": 0.4854,
+      "step": 11989
+    },
+    {
+      "epoch": 0.973767562738569,
+      "grad_norm": 6.539725065060604,
+      "learning_rate": 9.018341106464823e-09,
+      "loss": 0.4206,
+      "step": 11990
+    },
+    {
+      "epoch": 0.9738487777146105,
+      "grad_norm": 4.947085140988941,
+      "learning_rate": 8.962619764278923e-09,
+      "loss": 0.4808,
+      "step": 11991
+    },
+    {
+      "epoch": 0.9739299926906522,
+      "grad_norm": 6.855737833780214,
+      "learning_rate": 8.907070788695681e-09,
+      "loss": 0.3113,
+      "step": 11992
+    },
+    {
+      "epoch": 0.9740112076666937,
+      "grad_norm": 5.986387589137889,
+      "learning_rate": 8.851694183559523e-09,
+      "loss": 0.3221,
+      "step": 11993
+    },
+    {
+      "epoch": 0.9740924226427353,
+      "grad_norm": 6.396980239435106,
+      "learning_rate": 8.796489952701825e-09,
+      "loss": 0.5371,
+      "step": 11994
+    },
+    {
+      "epoch": 0.974173637618777,
+      "grad_norm": 4.469906653866564,
+      "learning_rate": 8.741458099942313e-09,
+      "loss": 0.5594,
+      "step": 11995
+    },
+    {
+      "epoch": 0.9742548525948185,
+      "grad_norm": 6.102062048541683,
+      "learning_rate": 8.686598629089326e-09,
+      "loss": 0.382,
+      "step": 11996
+    },
+    {
+      "epoch": 0.9743360675708601,
+      "grad_norm": 4.275407491240544,
+      "learning_rate": 8.63191154393872e-09,
+      "loss": 0.4869,
+      "step": 11997
+    },
+    {
+      "epoch": 0.9744172825469016,
+      "grad_norm": 6.27848767223886,
+      "learning_rate": 8.577396848274134e-09,
+      "loss": 0.4735,
+      "step": 11998
+    },
+    {
+      "epoch": 0.9744984975229433,
+      "grad_norm": 6.216106726624828,
+      "learning_rate": 8.523054545868381e-09,
+      "loss": 0.5145,
+      "step": 11999
+    },
+    {
+      "epoch": 0.9745797124989848,
+      "grad_norm": 6.83477643803915,
+      "learning_rate": 8.468884640480956e-09,
+      "loss": 0.4235,
+      "step": 12000
+    },
+    {
+      "epoch": 0.9746609274750264,
+      "grad_norm": 12.60083591448807,
+      "learning_rate": 8.414887135860528e-09,
+      "loss": 0.3258,
+      "step": 12001
+    },
+    {
+      "epoch": 0.9747421424510679,
+      "grad_norm": 4.645778183923307,
+      "learning_rate": 8.36106203574355e-09,
+      "loss": 0.5243,
+      "step": 12002
+    },
+    {
+      "epoch": 0.9748233574271096,
+      "grad_norm": 6.124524623697457,
+      "learning_rate": 8.307409343854267e-09,
+      "loss": 0.599,
+      "step": 12003
+    },
+    {
+      "epoch": 0.9749045724031511,
+      "grad_norm": 21.473022616832324,
+      "learning_rate": 8.253929063904986e-09,
+      "loss": 0.382,
+      "step": 12004
+    },
+    {
+      "epoch": 0.9749857873791927,
+      "grad_norm": 4.661852786954021,
+      "learning_rate": 8.200621199596359e-09,
+      "loss": 0.6825,
+      "step": 12005
+    },
+    {
+      "epoch": 0.9750670023552344,
+      "grad_norm": 3.538632513896874,
+      "learning_rate": 8.147485754617379e-09,
+      "loss": 0.4224,
+      "step": 12006
+    },
+    {
+      "epoch": 0.9751482173312759,
+      "grad_norm": 4.889222428461826,
+      "learning_rate": 8.094522732644272e-09,
+      "loss": 0.5328,
+      "step": 12007
+    },
+    {
+      "epoch": 0.9752294323073175,
+      "grad_norm": 8.835696874530043,
+      "learning_rate": 8.041732137341885e-09,
+      "loss": 0.3872,
+      "step": 12008
+    },
+    {
+      "epoch": 0.975310647283359,
+      "grad_norm": 5.49476829680145,
+      "learning_rate": 7.989113972363406e-09,
+      "loss": 0.6145,
+      "step": 12009
+    },
+    {
+      "epoch": 0.9753918622594007,
+      "grad_norm": 5.131999174481017,
+      "learning_rate": 7.936668241349255e-09,
+      "loss": 0.5043,
+      "step": 12010
+    },
+    {
+      "epoch": 0.9754730772354422,
+      "grad_norm": 4.572903995548805,
+      "learning_rate": 7.884394947928476e-09,
+      "loss": 0.5749,
+      "step": 12011
+    },
+    {
+      "epoch": 0.9755542922114838,
+      "grad_norm": 4.3477540547148825,
+      "learning_rate": 7.832294095718452e-09,
+      "loss": 0.4026,
+      "step": 12012
+    },
+    {
+      "epoch": 0.9756355071875253,
+      "grad_norm": 7.454961091377597,
+      "learning_rate": 7.780365688323798e-09,
+      "loss": 0.3078,
+      "step": 12013
+    },
+    {
+      "epoch": 0.975716722163567,
+      "grad_norm": 4.170795190072248,
+      "learning_rate": 7.72860972933831e-09,
+      "loss": 0.5571,
+      "step": 12014
+    },
+    {
+      "epoch": 0.9757979371396085,
+      "grad_norm": 7.707305765402371,
+      "learning_rate": 7.677026222342454e-09,
+      "loss": 0.3938,
+      "step": 12015
+    },
+    {
+      "epoch": 0.9758791521156501,
+      "grad_norm": 4.686956759781069,
+      "learning_rate": 7.625615170906153e-09,
+      "loss": 0.6371,
+      "step": 12016
+    },
+    {
+      "epoch": 0.9759603670916918,
+      "grad_norm": 6.526544319588936,
+      "learning_rate": 7.57437657858684e-09,
+      "loss": 0.4409,
+      "step": 12017
+    },
+    {
+      "epoch": 0.9760415820677333,
+      "grad_norm": 5.749185856694462,
+      "learning_rate": 7.523310448929178e-09,
+      "loss": 0.2862,
+      "step": 12018
+    },
+    {
+      "epoch": 0.9761227970437749,
+      "grad_norm": 6.413191599955744,
+      "learning_rate": 7.472416785467563e-09,
+      "loss": 0.5205,
+      "step": 12019
+    },
+    {
+      "epoch": 0.9762040120198164,
+      "grad_norm": 4.702110106071757,
+      "learning_rate": 7.421695591723066e-09,
+      "loss": 0.4963,
+      "step": 12020
+    },
+    {
+      "epoch": 0.9762852269958581,
+      "grad_norm": 4.552204118507779,
+      "learning_rate": 7.371146871205381e-09,
+      "loss": 0.5284,
+      "step": 12021
+    },
+    {
+      "epoch": 0.9763664419718996,
+      "grad_norm": 5.639485771969086,
+      "learning_rate": 7.320770627412543e-09,
+      "loss": 0.5295,
+      "step": 12022
+    },
+    {
+      "epoch": 0.9764476569479412,
+      "grad_norm": 6.8637890745131145,
+      "learning_rate": 7.27056686382982e-09,
+      "loss": 0.4764,
+      "step": 12023
+    },
+    {
+      "epoch": 0.9765288719239827,
+      "grad_norm": 5.248727371974205,
+      "learning_rate": 7.220535583931099e-09,
+      "loss": 0.4443,
+      "step": 12024
+    },
+    {
+      "epoch": 0.9766100869000244,
+      "grad_norm": 5.140443422767862,
+      "learning_rate": 7.17067679117861e-09,
+      "loss": 0.5761,
+      "step": 12025
+    },
+    {
+      "epoch": 0.9766913018760659,
+      "grad_norm": 4.603187702423827,
+      "learning_rate": 7.120990489022373e-09,
+      "loss": 0.4652,
+      "step": 12026
+    },
+    {
+      "epoch": 0.9767725168521075,
+      "grad_norm": 5.4771501562832245,
+      "learning_rate": 7.071476680900191e-09,
+      "loss": 0.3577,
+      "step": 12027
+    },
+    {
+      "epoch": 0.9768537318281492,
+      "grad_norm": 9.457901018585092,
+      "learning_rate": 7.022135370237937e-09,
+      "loss": 0.5396,
+      "step": 12028
+    },
+    {
+      "epoch": 0.9769349468041907,
+      "grad_norm": 6.7496814308010995,
+      "learning_rate": 6.972966560450101e-09,
+      "loss": 0.4375,
+      "step": 12029
+    },
+    {
+      "epoch": 0.9770161617802323,
+      "grad_norm": 9.618733980000252,
+      "learning_rate": 6.923970254938961e-09,
+      "loss": 0.463,
+      "step": 12030
+    },
+    {
+      "epoch": 0.9770973767562738,
+      "grad_norm": 3.6437791716015155,
+      "learning_rate": 6.875146457094583e-09,
+      "loss": 0.466,
+      "step": 12031
+    },
+    {
+      "epoch": 0.9771785917323155,
+      "grad_norm": 4.087276926389463,
+      "learning_rate": 6.8264951702951e-09,
+      "loss": 0.3838,
+      "step": 12032
+    },
+    {
+      "epoch": 0.977259806708357,
+      "grad_norm": 7.648508375870743,
+      "learning_rate": 6.778016397907539e-09,
+      "loss": 0.3517,
+      "step": 12033
+    },
+    {
+      "epoch": 0.9773410216843986,
+      "grad_norm": 6.478817726316786,
+      "learning_rate": 6.729710143286161e-09,
+      "loss": 0.3296,
+      "step": 12034
+    },
+    {
+      "epoch": 0.9774222366604401,
+      "grad_norm": 4.785457296804816,
+      "learning_rate": 6.681576409773016e-09,
+      "loss": 0.5188,
+      "step": 12035
+    },
+    {
+      "epoch": 0.9775034516364818,
+      "grad_norm": 7.296274375209841,
+      "learning_rate": 6.633615200699328e-09,
+      "loss": 0.6136,
+      "step": 12036
+    },
+    {
+      "epoch": 0.9775846666125233,
+      "grad_norm": 4.160049221454699,
+      "learning_rate": 6.5858265193835536e-09,
+      "loss": 0.514,
+      "step": 12037
+    },
+    {
+      "epoch": 0.9776658815885649,
+      "grad_norm": 5.830756185483621,
+      "learning_rate": 6.538210369132214e-09,
+      "loss": 0.4866,
+      "step": 12038
+    },
+    {
+      "epoch": 0.9777470965646066,
+      "grad_norm": 5.407452911485265,
+      "learning_rate": 6.490766753240174e-09,
+      "loss": 0.6008,
+      "step": 12039
+    },
+    {
+      "epoch": 0.9778283115406481,
+      "grad_norm": 5.873928827822523,
+      "learning_rate": 6.443495674990641e-09,
+      "loss": 0.5819,
+      "step": 12040
+    },
+    {
+      "epoch": 0.9779095265166897,
+      "grad_norm": 6.00288359521902,
+      "learning_rate": 6.396397137654054e-09,
+      "loss": 0.3921,
+      "step": 12041
+    },
+    {
+      "epoch": 0.9779907414927312,
+      "grad_norm": 7.297473039045752,
+      "learning_rate": 6.3494711444897495e-09,
+      "loss": 0.4012,
+      "step": 12042
+    },
+    {
+      "epoch": 0.9780719564687729,
+      "grad_norm": 4.082939335655915,
+      "learning_rate": 6.302717698744298e-09,
+      "loss": 0.4805,
+      "step": 12043
+    },
+    {
+      "epoch": 0.9781531714448144,
+      "grad_norm": 9.09294805829235,
+      "learning_rate": 6.2561368036531676e-09,
+      "loss": 0.5639,
+      "step": 12044
+    },
+    {
+      "epoch": 0.978234386420856,
+      "grad_norm": 6.290896132053336,
+      "learning_rate": 6.209728462439613e-09,
+      "loss": 0.5002,
+      "step": 12045
+    },
+    {
+      "epoch": 0.9783156013968975,
+      "grad_norm": 8.382100561674394,
+      "learning_rate": 6.1634926783143975e-09,
+      "loss": 0.4165,
+      "step": 12046
+    },
+    {
+      "epoch": 0.9783968163729392,
+      "grad_norm": 5.891844640379527,
+      "learning_rate": 6.117429454477186e-09,
+      "loss": 0.3735,
+      "step": 12047
+    },
+    {
+      "epoch": 0.9784780313489807,
+      "grad_norm": 5.2181840686707766,
+      "learning_rate": 6.071538794115151e-09,
+      "loss": 0.5358,
+      "step": 12048
+    },
+    {
+      "epoch": 0.9785592463250223,
+      "grad_norm": 5.170788538583498,
+      "learning_rate": 6.025820700403529e-09,
+      "loss": 0.5949,
+      "step": 12049
+    },
+    {
+      "epoch": 0.978640461301064,
+      "grad_norm": 9.342590625760383,
+      "learning_rate": 5.9802751765061785e-09,
+      "loss": 0.553,
+      "step": 12050
+    },
+    {
+      "epoch": 0.9787216762771055,
+      "grad_norm": 6.856775571985887,
+      "learning_rate": 5.9349022255741905e-09,
+      "loss": 0.5531,
+      "step": 12051
+    },
+    {
+      "epoch": 0.9788028912531471,
+      "grad_norm": 4.222463014742318,
+      "learning_rate": 5.889701850747276e-09,
+      "loss": 0.5136,
+      "step": 12052
+    },
+    {
+      "epoch": 0.9788841062291886,
+      "grad_norm": 6.316498496538767,
+      "learning_rate": 5.844674055153487e-09,
+      "loss": 0.4999,
+      "step": 12053
+    },
+    {
+      "epoch": 0.9789653212052303,
+      "grad_norm": 5.951917290095282,
+      "learning_rate": 5.799818841907556e-09,
+      "loss": 0.4719,
+      "step": 12054
+    },
+    {
+      "epoch": 0.9790465361812718,
+      "grad_norm": 10.835272741624479,
+      "learning_rate": 5.7551362141142205e-09,
+      "loss": 0.3861,
+      "step": 12055
+    },
+    {
+      "epoch": 0.9791277511573134,
+      "grad_norm": 6.226477036302456,
+      "learning_rate": 5.71062617486462e-09,
+      "loss": 0.5043,
+      "step": 12056
+    },
+    {
+      "epoch": 0.979208966133355,
+      "grad_norm": 9.360240176470688,
+      "learning_rate": 5.666288727239066e-09,
+      "loss": 0.4041,
+      "step": 12057
+    },
+    {
+      "epoch": 0.9792901811093966,
+      "grad_norm": 8.608220193471967,
+      "learning_rate": 5.622123874305108e-09,
+      "loss": 0.3628,
+      "step": 12058
+    },
+    {
+      "epoch": 0.9793713960854381,
+      "grad_norm": 5.828438095145518,
+      "learning_rate": 5.578131619118909e-09,
+      "loss": 0.3776,
+      "step": 12059
+    },
+    {
+      "epoch": 0.9794526110614797,
+      "grad_norm": 4.52241346057629,
+      "learning_rate": 5.534311964724426e-09,
+      "loss": 0.4952,
+      "step": 12060
+    },
+    {
+      "epoch": 0.9795338260375214,
+      "grad_norm": 3.8196975497917736,
+      "learning_rate": 5.490664914153676e-09,
+      "loss": 0.4629,
+      "step": 12061
+    },
+    {
+      "epoch": 0.9796150410135629,
+      "grad_norm": 4.37711626897485,
+      "learning_rate": 5.447190470427022e-09,
+      "loss": 0.4722,
+      "step": 12062
+    },
+    {
+      "epoch": 0.9796962559896045,
+      "grad_norm": 3.977099684985591,
+      "learning_rate": 5.4038886365523346e-09,
+      "loss": 0.5534,
+      "step": 12063
+    },
+    {
+      "epoch": 0.979777470965646,
+      "grad_norm": 3.64657798618061,
+      "learning_rate": 5.360759415526385e-09,
+      "loss": 0.4216,
+      "step": 12064
+    },
+    {
+      "epoch": 0.9798586859416877,
+      "grad_norm": 5.625694324125192,
+      "learning_rate": 5.3178028103331725e-09,
+      "loss": 0.4299,
+      "step": 12065
+    },
+    {
+      "epoch": 0.9799399009177292,
+      "grad_norm": 3.7515462457345823,
+      "learning_rate": 5.275018823945044e-09,
+      "loss": 0.4106,
+      "step": 12066
+    },
+    {
+      "epoch": 0.9800211158937708,
+      "grad_norm": 4.423453741648708,
+      "learning_rate": 5.232407459322408e-09,
+      "loss": 0.3944,
+      "step": 12067
+    },
+    {
+      "epoch": 0.9801023308698124,
+      "grad_norm": 5.1224553290352635,
+      "learning_rate": 5.189968719413741e-09,
+      "loss": 0.379,
+      "step": 12068
+    },
+    {
+      "epoch": 0.980183545845854,
+      "grad_norm": 7.181016487036675,
+      "learning_rate": 5.14770260715558e-09,
+      "loss": 0.4096,
+      "step": 12069
+    },
+    {
+      "epoch": 0.9802647608218955,
+      "grad_norm": 5.486473028790214,
+      "learning_rate": 5.10560912547281e-09,
+      "loss": 0.5059,
+      "step": 12070
+    },
+    {
+      "epoch": 0.9803459757979371,
+      "grad_norm": 4.843785469577583,
+      "learning_rate": 5.063688277277545e-09,
+      "loss": 0.444,
+      "step": 12071
+    },
+    {
+      "epoch": 0.9804271907739788,
+      "grad_norm": 12.938046238864795,
+      "learning_rate": 5.021940065471076e-09,
+      "loss": 0.4866,
+      "step": 12072
+    },
+    {
+      "epoch": 0.9805084057500203,
+      "grad_norm": 10.093029172739994,
+      "learning_rate": 4.980364492941924e-09,
+      "loss": 0.4679,
+      "step": 12073
+    },
+    {
+      "epoch": 0.9805896207260619,
+      "grad_norm": 7.273329994281363,
+      "learning_rate": 4.938961562566402e-09,
+      "loss": 0.3914,
+      "step": 12074
+    },
+    {
+      "epoch": 0.9806708357021034,
+      "grad_norm": 3.3725698256307437,
+      "learning_rate": 4.8977312772102715e-09,
+      "loss": 0.5183,
+      "step": 12075
+    },
+    {
+      "epoch": 0.9807520506781451,
+      "grad_norm": 5.1217541737470835,
+      "learning_rate": 4.856673639725695e-09,
+      "loss": 0.498,
+      "step": 12076
+    },
+    {
+      "epoch": 0.9808332656541866,
+      "grad_norm": 3.8730107181355535,
+      "learning_rate": 4.815788652954012e-09,
+      "loss": 0.4615,
+      "step": 12077
+    },
+    {
+      "epoch": 0.9809144806302282,
+      "grad_norm": 8.294253938971737,
+      "learning_rate": 4.775076319724348e-09,
+      "loss": 0.5011,
+      "step": 12078
+    },
+    {
+      "epoch": 0.9809956956062698,
+      "grad_norm": 3.478676476091108,
+      "learning_rate": 4.734536642853338e-09,
+      "loss": 0.5997,
+      "step": 12079
+    },
+    {
+      "epoch": 0.9810769105823114,
+      "grad_norm": 4.181120233195193,
+      "learning_rate": 4.6941696251465165e-09,
+      "loss": 0.6451,
+      "step": 12080
+    },
+    {
+      "epoch": 0.9811581255583529,
+      "grad_norm": 3.6134990519276466,
+      "learning_rate": 4.6539752693969265e-09,
+      "loss": 0.4854,
+      "step": 12081
+    },
+    {
+      "epoch": 0.9812393405343945,
+      "grad_norm": 11.700469427119463,
+      "learning_rate": 4.613953578385954e-09,
+      "loss": 0.3663,
+      "step": 12082
+    },
+    {
+      "epoch": 0.9813205555104362,
+      "grad_norm": 4.369290046988292,
+      "learning_rate": 4.574104554882497e-09,
+      "loss": 0.7481,
+      "step": 12083
+    },
+    {
+      "epoch": 0.9814017704864777,
+      "grad_norm": 5.401987698332693,
+      "learning_rate": 4.534428201644348e-09,
+      "loss": 0.423,
+      "step": 12084
+    },
+    {
+      "epoch": 0.9814829854625193,
+      "grad_norm": 9.430350487972829,
+      "learning_rate": 4.494924521416533e-09,
+      "loss": 0.4895,
+      "step": 12085
+    },
+    {
+      "epoch": 0.9815642004385609,
+      "grad_norm": 3.7707234936613565,
+      "learning_rate": 4.455593516932699e-09,
+      "loss": 0.5178,
+      "step": 12086
+    },
+    {
+      "epoch": 0.9816454154146025,
+      "grad_norm": 6.985985247371312,
+      "learning_rate": 4.4164351909142815e-09,
+      "loss": 0.3872,
+      "step": 12087
+    },
+    {
+      "epoch": 0.981726630390644,
+      "grad_norm": 5.1330117460195215,
+      "learning_rate": 4.377449546071055e-09,
+      "loss": 0.4994,
+      "step": 12088
+    },
+    {
+      "epoch": 0.9818078453666856,
+      "grad_norm": 6.234609568273921,
+      "learning_rate": 4.338636585100309e-09,
+      "loss": 0.5062,
+      "step": 12089
+    },
+    {
+      "epoch": 0.9818890603427272,
+      "grad_norm": 7.581218288677617,
+      "learning_rate": 4.299996310687671e-09,
+      "loss": 0.5828,
+      "step": 12090
+    },
+    {
+      "epoch": 0.9819702753187688,
+      "grad_norm": 5.845752308546046,
+      "learning_rate": 4.261528725507113e-09,
+      "loss": 0.472,
+      "step": 12091
+    },
+    {
+      "epoch": 0.9820514902948103,
+      "grad_norm": 7.146820683688464,
+      "learning_rate": 4.223233832220397e-09,
+      "loss": 0.4632,
+      "step": 12092
+    },
+    {
+      "epoch": 0.982132705270852,
+      "grad_norm": 3.2381775735767935,
+      "learning_rate": 4.18511163347679e-09,
+      "loss": 0.6537,
+      "step": 12093
+    },
+    {
+      "epoch": 0.9822139202468936,
+      "grad_norm": 5.730027453232988,
+      "learning_rate": 4.147162131914739e-09,
+      "loss": 0.3759,
+      "step": 12094
+    },
+    {
+      "epoch": 0.9822951352229351,
+      "grad_norm": 3.5984616089033827,
+      "learning_rate": 4.109385330159921e-09,
+      "loss": 0.5452,
+      "step": 12095
+    },
+    {
+      "epoch": 0.9823763501989767,
+      "grad_norm": 6.769530777534317,
+      "learning_rate": 4.071781230826355e-09,
+      "loss": 0.4365,
+      "step": 12096
+    },
+    {
+      "epoch": 0.9824575651750183,
+      "grad_norm": 5.819933789150477,
+      "learning_rate": 4.034349836516127e-09,
+      "loss": 0.4541,
+      "step": 12097
+    },
+    {
+      "epoch": 0.9825387801510599,
+      "grad_norm": 5.491397739648694,
+      "learning_rate": 3.99709114981911e-09,
+      "loss": 0.3554,
+      "step": 12098
+    },
+    {
+      "epoch": 0.9826199951271014,
+      "grad_norm": 5.29270938282257,
+      "learning_rate": 3.960005173313519e-09,
+      "loss": 0.3631,
+      "step": 12099
+    },
+    {
+      "epoch": 0.982701210103143,
+      "grad_norm": 5.1898426034045615,
+      "learning_rate": 3.923091909565357e-09,
+      "loss": 0.4277,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9827824250791846,
+      "grad_norm": 4.856901331420813,
+      "learning_rate": 3.88635136112897e-09,
+      "loss": 0.4089,
+      "step": 12101
+    },
+    {
+      "epoch": 0.9828636400552262,
+      "grad_norm": 3.4872568138932762,
+      "learning_rate": 3.8497835305464915e-09,
+      "loss": 0.6127,
+      "step": 12102
+    },
+    {
+      "epoch": 0.9829448550312677,
+      "grad_norm": 13.086059636580629,
+      "learning_rate": 3.813388420348396e-09,
+      "loss": 0.4843,
+      "step": 12103
+    },
+    {
+      "epoch": 0.9830260700073093,
+      "grad_norm": 5.395590802101605,
+      "learning_rate": 3.777166033052948e-09,
+      "loss": 0.4887,
+      "step": 12104
+    },
+    {
+      "epoch": 0.983107284983351,
+      "grad_norm": 4.809183849119693,
+      "learning_rate": 3.741116371166476e-09,
+      "loss": 0.5797,
+      "step": 12105
+    },
+    {
+      "epoch": 0.9831884999593925,
+      "grad_norm": 7.000518609996093,
+      "learning_rate": 3.705239437183372e-09,
+      "loss": 0.6074,
+      "step": 12106
+    },
+    {
+      "epoch": 0.9832697149354341,
+      "grad_norm": 7.749862840244079,
+      "learning_rate": 3.6695352335863745e-09,
+      "loss": 0.3449,
+      "step": 12107
+    },
+    {
+      "epoch": 0.9833509299114757,
+      "grad_norm": 5.2466883835337494,
+      "learning_rate": 3.6340037628460057e-09,
+      "loss": 0.4941,
+      "step": 12108
+    },
+    {
+      "epoch": 0.9834321448875173,
+      "grad_norm": 20.385619121552477,
+      "learning_rate": 3.5986450274205776e-09,
+      "loss": 0.4583,
+      "step": 12109
+    },
+    {
+      "epoch": 0.9835133598635588,
+      "grad_norm": 4.0593617565984035,
+      "learning_rate": 3.5634590297570215e-09,
+      "loss": 0.4225,
+      "step": 12110
+    },
+    {
+      "epoch": 0.9835945748396004,
+      "grad_norm": 6.465413464659027,
+      "learning_rate": 3.528445772289779e-09,
+      "loss": 0.5616,
+      "step": 12111
+    },
+    {
+      "epoch": 0.983675789815642,
+      "grad_norm": 8.991522570276532,
+      "learning_rate": 3.4936052574416345e-09,
+      "loss": 0.3596,
+      "step": 12112
+    },
+    {
+      "epoch": 0.9837570047916836,
+      "grad_norm": 4.818729279084131,
+      "learning_rate": 3.458937487623437e-09,
+      "loss": 0.4936,
+      "step": 12113
+    },
+    {
+      "epoch": 0.9838382197677251,
+      "grad_norm": 6.112039073938186,
+      "learning_rate": 3.424442465234101e-09,
+      "loss": 0.4451,
+      "step": 12114
+    },
+    {
+      "epoch": 0.9839194347437668,
+      "grad_norm": 6.296027842842914,
+      "learning_rate": 3.3901201926606063e-09,
+      "loss": 0.5395,
+      "step": 12115
+    },
+    {
+      "epoch": 0.9840006497198084,
+      "grad_norm": 5.253597703880921,
+      "learning_rate": 3.3559706722774423e-09,
+      "loss": 0.6561,
+      "step": 12116
+    },
+    {
+      "epoch": 0.9840818646958499,
+      "grad_norm": 5.551682606805065,
+      "learning_rate": 3.3219939064477182e-09,
+      "loss": 0.3641,
+      "step": 12117
+    },
+    {
+      "epoch": 0.9841630796718915,
+      "grad_norm": 7.465132500098328,
+      "learning_rate": 3.288189897522609e-09,
+      "loss": 0.4635,
+      "step": 12118
+    },
+    {
+      "epoch": 0.9842442946479331,
+      "grad_norm": 8.173920927315894,
+      "learning_rate": 3.254558647841077e-09,
+      "loss": 0.5209,
+      "step": 12119
+    },
+    {
+      "epoch": 0.9843255096239747,
+      "grad_norm": 4.788329870369596,
+      "learning_rate": 3.2211001597304283e-09,
+      "loss": 0.5823,
+      "step": 12120
+    },
+    {
+      "epoch": 0.9844067246000162,
+      "grad_norm": 6.8587617655903355,
+      "learning_rate": 3.187814435505199e-09,
+      "loss": 0.4951,
+      "step": 12121
+    },
+    {
+      "epoch": 0.9844879395760578,
+      "grad_norm": 5.481814887439901,
+      "learning_rate": 3.1547014774693797e-09,
+      "loss": 0.3174,
+      "step": 12122
+    },
+    {
+      "epoch": 0.9845691545520994,
+      "grad_norm": 8.885860015107669,
+      "learning_rate": 3.1217612879139158e-09,
+      "loss": 0.4645,
+      "step": 12123
+    },
+    {
+      "epoch": 0.984650369528141,
+      "grad_norm": 4.284808548816115,
+      "learning_rate": 3.088993869117818e-09,
+      "loss": 0.4884,
+      "step": 12124
+    },
+    {
+      "epoch": 0.9847315845041825,
+      "grad_norm": 4.896143196735318,
+      "learning_rate": 3.056399223348716e-09,
+      "loss": 0.3905,
+      "step": 12125
+    },
+    {
+      "epoch": 0.9848127994802242,
+      "grad_norm": 4.786981373349127,
+      "learning_rate": 3.023977352861751e-09,
+      "loss": 0.6363,
+      "step": 12126
+    },
+    {
+      "epoch": 0.9848940144562658,
+      "grad_norm": 5.440565139296447,
+      "learning_rate": 2.991728259900684e-09,
+      "loss": 0.4777,
+      "step": 12127
+    },
+    {
+      "epoch": 0.9849752294323073,
+      "grad_norm": 4.432486814000843,
+      "learning_rate": 2.959651946696507e-09,
+      "loss": 0.4696,
+      "step": 12128
+    },
+    {
+      "epoch": 0.9850564444083489,
+      "grad_norm": 4.593612284410864,
+      "learning_rate": 2.927748415469389e-09,
+      "loss": 0.4412,
+      "step": 12129
+    },
+    {
+      "epoch": 0.9851376593843905,
+      "grad_norm": 11.349530667232669,
+      "learning_rate": 2.8960176684261767e-09,
+      "loss": 0.5133,
+      "step": 12130
+    },
+    {
+      "epoch": 0.9852188743604321,
+      "grad_norm": 7.977641323262256,
+      "learning_rate": 2.86445970776289e-09,
+      "loss": 0.4928,
+      "step": 12131
+    },
+    {
+      "epoch": 0.9853000893364736,
+      "grad_norm": 4.0771155412598805,
+      "learning_rate": 2.833074535663338e-09,
+      "loss": 0.4539,
+      "step": 12132
+    },
+    {
+      "epoch": 0.9853813043125152,
+      "grad_norm": 4.941422589211022,
+      "learning_rate": 2.8018621542988402e-09,
+      "loss": 0.4797,
+      "step": 12133
+    },
+    {
+      "epoch": 0.9854625192885568,
+      "grad_norm": 8.421332670522345,
+      "learning_rate": 2.7708225658290566e-09,
+      "loss": 0.5164,
+      "step": 12134
+    },
+    {
+      "epoch": 0.9855437342645984,
+      "grad_norm": 7.591462565867755,
+      "learning_rate": 2.739955772401992e-09,
+      "loss": 0.621,
+      "step": 12135
+    },
+    {
+      "epoch": 0.9856249492406399,
+      "grad_norm": 5.772643166679314,
+      "learning_rate": 2.709261776153438e-09,
+      "loss": 0.4788,
+      "step": 12136
+    },
+    {
+      "epoch": 0.9857061642166816,
+      "grad_norm": 6.132821992629694,
+      "learning_rate": 2.6787405792072507e-09,
+      "loss": 0.4194,
+      "step": 12137
+    },
+    {
+      "epoch": 0.9857873791927232,
+      "grad_norm": 5.663712178693497,
+      "learning_rate": 2.6483921836753525e-09,
+      "loss": 0.4067,
+      "step": 12138
+    },
+    {
+      "epoch": 0.9858685941687647,
+      "grad_norm": 5.534184126963757,
+      "learning_rate": 2.6182165916577295e-09,
+      "loss": 0.547,
+      "step": 12139
+    },
+    {
+      "epoch": 0.9859498091448063,
+      "grad_norm": 4.047838425287923,
+      "learning_rate": 2.5882138052421567e-09,
+      "loss": 0.4595,
+      "step": 12140
+    },
+    {
+      "epoch": 0.9860310241208479,
+      "grad_norm": 3.5712541301580667,
+      "learning_rate": 2.5583838265050286e-09,
+      "loss": 0.4067,
+      "step": 12141
+    },
+    {
+      "epoch": 0.9861122390968895,
+      "grad_norm": 5.1536702438204065,
+      "learning_rate": 2.52872665751025e-09,
+      "loss": 0.5706,
+      "step": 12142
+    },
+    {
+      "epoch": 0.986193454072931,
+      "grad_norm": 16.616810366998326,
+      "learning_rate": 2.4992423003095124e-09,
+      "loss": 0.5131,
+      "step": 12143
+    },
+    {
+      "epoch": 0.9862746690489727,
+      "grad_norm": 5.2196354867906685,
+      "learning_rate": 2.4699307569436835e-09,
+      "loss": 0.4475,
+      "step": 12144
+    },
+    {
+      "epoch": 0.9863558840250142,
+      "grad_norm": 4.94188443326249,
+      "learning_rate": 2.4407920294405864e-09,
+      "loss": 0.381,
+      "step": 12145
+    },
+    {
+      "epoch": 0.9864370990010558,
+      "grad_norm": 6.982635794914591,
+      "learning_rate": 2.4118261198166625e-09,
+      "loss": 0.4683,
+      "step": 12146
+    },
+    {
+      "epoch": 0.9865183139770973,
+      "grad_norm": 8.68074477826504,
+      "learning_rate": 2.383033030075865e-09,
+      "loss": 0.4486,
+      "step": 12147
+    },
+    {
+      "epoch": 0.986599528953139,
+      "grad_norm": 4.488613369870494,
+      "learning_rate": 2.354412762210767e-09,
+      "loss": 0.5145,
+      "step": 12148
+    },
+    {
+      "epoch": 0.9866807439291806,
+      "grad_norm": 8.342369544568935,
+      "learning_rate": 2.325965318201728e-09,
+      "loss": 0.4742,
+      "step": 12149
+    },
+    {
+      "epoch": 0.9867619589052221,
+      "grad_norm": 4.473725413880338,
+      "learning_rate": 2.2976907000171743e-09,
+      "loss": 0.4517,
+      "step": 12150
+    },
+    {
+      "epoch": 0.9868431738812637,
+      "grad_norm": 5.732059155908079,
+      "learning_rate": 2.2695889096133184e-09,
+      "loss": 0.5598,
+      "step": 12151
+    },
+    {
+      "epoch": 0.9869243888573053,
+      "grad_norm": 6.627352082908765,
+      "learning_rate": 2.2416599489349933e-09,
+      "loss": 0.4432,
+      "step": 12152
+    },
+    {
+      "epoch": 0.9870056038333469,
+      "grad_norm": 6.4462903139711525,
+      "learning_rate": 2.2139038199145424e-09,
+      "loss": 0.641,
+      "step": 12153
+    },
+    {
+      "epoch": 0.9870868188093884,
+      "grad_norm": 4.235134074339075,
+      "learning_rate": 2.1863205244726514e-09,
+      "loss": 0.3059,
+      "step": 12154
+    },
+    {
+      "epoch": 0.98716803378543,
+      "grad_norm": 10.34898874466498,
+      "learning_rate": 2.1589100645180715e-09,
+      "loss": 0.4217,
+      "step": 12155
+    },
+    {
+      "epoch": 0.9872492487614716,
+      "grad_norm": 7.083106400400486,
+      "learning_rate": 2.1316724419470637e-09,
+      "loss": 0.4262,
+      "step": 12156
+    },
+    {
+      "epoch": 0.9873304637375132,
+      "grad_norm": 4.880842030469988,
+      "learning_rate": 2.1046076586445084e-09,
+      "loss": 0.4115,
+      "step": 12157
+    },
+    {
+      "epoch": 0.9874116787135547,
+      "grad_norm": 5.428807688681806,
+      "learning_rate": 2.077715716483353e-09,
+      "loss": 0.4031,
+      "step": 12158
+    },
+    {
+      "epoch": 0.9874928936895964,
+      "grad_norm": 3.5615581342610594,
+      "learning_rate": 2.0509966173240524e-09,
+      "loss": 0.4478,
+      "step": 12159
+    },
+    {
+      "epoch": 0.987574108665638,
+      "grad_norm": 4.9802028956802245,
+      "learning_rate": 2.0244503630154066e-09,
+      "loss": 0.566,
+      "step": 12160
+    },
+    {
+      "epoch": 0.9876553236416795,
+      "grad_norm": 3.88910810121843,
+      "learning_rate": 1.9980769553948344e-09,
+      "loss": 0.3933,
+      "step": 12161
+    },
+    {
+      "epoch": 0.9877365386177211,
+      "grad_norm": 4.551525974114428,
+      "learning_rate": 1.9718763962867094e-09,
+      "loss": 0.4622,
+      "step": 12162
+    },
+    {
+      "epoch": 0.9878177535937627,
+      "grad_norm": 6.650520927884947,
+      "learning_rate": 1.945848687504026e-09,
+      "loss": 0.5204,
+      "step": 12163
+    },
+    {
+      "epoch": 0.9878989685698043,
+      "grad_norm": 5.106322645005586,
+      "learning_rate": 1.919993830847844e-09,
+      "loss": 0.3685,
+      "step": 12164
+    },
+    {
+      "epoch": 0.9879801835458458,
+      "grad_norm": 5.293912542968066,
+      "learning_rate": 1.8943118281070095e-09,
+      "loss": 0.4464,
+      "step": 12165
+    },
+    {
+      "epoch": 0.9880613985218875,
+      "grad_norm": 4.885311151399439,
+      "learning_rate": 1.86880268105899e-09,
+      "loss": 0.5068,
+      "step": 12166
+    },
+    {
+      "epoch": 0.988142613497929,
+      "grad_norm": 5.1380448467955295,
+      "learning_rate": 1.8434663914687623e-09,
+      "loss": 0.4485,
+      "step": 12167
+    },
+    {
+      "epoch": 0.9882238284739706,
+      "grad_norm": 4.9192525339798925,
+      "learning_rate": 1.8183029610890912e-09,
+      "loss": 0.5346,
+      "step": 12168
+    },
+    {
+      "epoch": 0.9883050434500121,
+      "grad_norm": 4.9277595849293565,
+      "learning_rate": 1.7933123916613614e-09,
+      "loss": 0.4065,
+      "step": 12169
+    },
+    {
+      "epoch": 0.9883862584260538,
+      "grad_norm": 5.8993559652023215,
+      "learning_rate": 1.7684946849150232e-09,
+      "loss": 0.459,
+      "step": 12170
+    },
+    {
+      "epoch": 0.9884674734020954,
+      "grad_norm": 4.936973153328386,
+      "learning_rate": 1.7438498425673135e-09,
+      "loss": 0.5248,
+      "step": 12171
+    },
+    {
+      "epoch": 0.9885486883781369,
+      "grad_norm": 4.4227762528236125,
+      "learning_rate": 1.7193778663229799e-09,
+      "loss": 0.5324,
+      "step": 12172
+    },
+    {
+      "epoch": 0.9886299033541786,
+      "grad_norm": 4.957741580907209,
+      "learning_rate": 1.6950787578759453e-09,
+      "loss": 0.5095,
+      "step": 12173
+    },
+    {
+      "epoch": 0.9887111183302201,
+      "grad_norm": 6.64145740938836,
+      "learning_rate": 1.6709525189073649e-09,
+      "loss": 0.4332,
+      "step": 12174
+    },
+    {
+      "epoch": 0.9887923333062617,
+      "grad_norm": 6.145921162498026,
+      "learning_rate": 1.646999151086459e-09,
+      "loss": 0.3923,
+      "step": 12175
+    },
+    {
+      "epoch": 0.9888735482823032,
+      "grad_norm": 5.254183282644911,
+      "learning_rate": 1.6232186560710684e-09,
+      "loss": 0.4596,
+      "step": 12176
+    },
+    {
+      "epoch": 0.9889547632583449,
+      "grad_norm": 17.99187941281074,
+      "learning_rate": 1.599611035506543e-09,
+      "loss": 0.6023,
+      "step": 12177
+    },
+    {
+      "epoch": 0.9890359782343864,
+      "grad_norm": 10.290258462437267,
+      "learning_rate": 1.5761762910260214e-09,
+      "loss": 0.5383,
+      "step": 12178
+    },
+    {
+      "epoch": 0.989117193210428,
+      "grad_norm": 6.646644372837433,
+      "learning_rate": 1.5529144242518167e-09,
+      "loss": 0.5216,
+      "step": 12179
+    },
+    {
+      "epoch": 0.9891984081864695,
+      "grad_norm": 4.4766333642428675,
+      "learning_rate": 1.5298254367926424e-09,
+      "loss": 0.4035,
+      "step": 12180
+    },
+    {
+      "epoch": 0.9892796231625112,
+      "grad_norm": 9.914986537758677,
+      "learning_rate": 1.5069093302469418e-09,
+      "loss": 0.4324,
+      "step": 12181
+    },
+    {
+      "epoch": 0.9893608381385528,
+      "grad_norm": 4.176667675043893,
+      "learning_rate": 1.4841661061998358e-09,
+      "loss": 0.4502,
+      "step": 12182
+    },
+    {
+      "epoch": 0.9894420531145943,
+      "grad_norm": 3.795368320920534,
+      "learning_rate": 1.4615957662250657e-09,
+      "loss": 0.4701,
+      "step": 12183
+    },
+    {
+      "epoch": 0.989523268090636,
+      "grad_norm": 5.32972800464418,
+      "learning_rate": 1.4391983118847152e-09,
+      "loss": 0.605,
+      "step": 12184
+    },
+    {
+      "epoch": 0.9896044830666775,
+      "grad_norm": 6.008043856391167,
+      "learning_rate": 1.4169737447283782e-09,
+      "loss": 0.3262,
+      "step": 12185
+    },
+    {
+      "epoch": 0.9896856980427191,
+      "grad_norm": 7.84003745927447,
+      "learning_rate": 1.394922066293991e-09,
+      "loss": 0.5486,
+      "step": 12186
+    },
+    {
+      "epoch": 0.9897669130187606,
+      "grad_norm": 6.066959827307744,
+      "learning_rate": 1.3730432781070002e-09,
+      "loss": 0.4331,
+      "step": 12187
+    },
+    {
+      "epoch": 0.9898481279948023,
+      "grad_norm": 17.157061321750632,
+      "learning_rate": 1.3513373816820274e-09,
+      "loss": 0.642,
+      "step": 12188
+    },
+    {
+      "epoch": 0.9899293429708438,
+      "grad_norm": 3.932173196085788,
+      "learning_rate": 1.3298043785203718e-09,
+      "loss": 0.4469,
+      "step": 12189
+    },
+    {
+      "epoch": 0.9900105579468854,
+      "grad_norm": 8.805868997589002,
+      "learning_rate": 1.30844427011223e-09,
+      "loss": 0.485,
+      "step": 12190
+    },
+    {
+      "epoch": 0.9900917729229269,
+      "grad_norm": 4.027618081932802,
+      "learning_rate": 1.287257057935587e-09,
+      "loss": 0.4275,
+      "step": 12191
+    },
+    {
+      "epoch": 0.9901729878989686,
+      "grad_norm": 8.085780436510987,
+      "learning_rate": 1.2662427434564916e-09,
+      "loss": 0.5487,
+      "step": 12192
+    },
+    {
+      "epoch": 0.9902542028750102,
+      "grad_norm": 6.6055275665447875,
+      "learning_rate": 1.2454013281290589e-09,
+      "loss": 0.4155,
+      "step": 12193
+    },
+    {
+      "epoch": 0.9903354178510517,
+      "grad_norm": 5.7796904505260365,
+      "learning_rate": 1.2247328133954683e-09,
+      "loss": 0.4356,
+      "step": 12194
+    },
+    {
+      "epoch": 0.9904166328270934,
+      "grad_norm": 4.4243001082018445,
+      "learning_rate": 1.2042372006856873e-09,
+      "loss": 0.4525,
+      "step": 12195
+    },
+    {
+      "epoch": 0.9904978478031349,
+      "grad_norm": 5.706811651819559,
+      "learning_rate": 1.1839144914180256e-09,
+      "loss": 0.3391,
+      "step": 12196
+    },
+    {
+      "epoch": 0.9905790627791765,
+      "grad_norm": 5.337953154660965,
+      "learning_rate": 1.1637646869985809e-09,
+      "loss": 0.4905,
+      "step": 12197
+    },
+    {
+      "epoch": 0.990660277755218,
+      "grad_norm": 7.132197947930379,
+      "learning_rate": 1.143787788821793e-09,
+      "loss": 0.4736,
+      "step": 12198
+    },
+    {
+      "epoch": 0.9907414927312597,
+      "grad_norm": 8.488570779233747,
+      "learning_rate": 1.1239837982698898e-09,
+      "loss": 0.4538,
+      "step": 12199
+    },
+    {
+      "epoch": 0.9908227077073012,
+      "grad_norm": 8.566390662400726,
+      "learning_rate": 1.104352716713164e-09,
+      "loss": 0.5289,
+      "step": 12200
+    },
+    {
+      "epoch": 0.9909039226833428,
+      "grad_norm": 7.713592201697891,
+      "learning_rate": 1.0848945455099734e-09,
+      "loss": 0.4059,
+      "step": 12201
+    },
+    {
+      "epoch": 0.9909851376593843,
+      "grad_norm": 8.262715346834725,
+      "learning_rate": 1.0656092860067413e-09,
+      "loss": 0.367,
+      "step": 12202
+    },
+    {
+      "epoch": 0.991066352635426,
+      "grad_norm": 5.432388233712127,
+      "learning_rate": 1.046496939538233e-09,
+      "loss": 0.3928,
+      "step": 12203
+    },
+    {
+      "epoch": 0.9911475676114676,
+      "grad_norm": 5.147447230251187,
+      "learning_rate": 1.027557507426169e-09,
+      "loss": 0.5214,
+      "step": 12204
+    },
+    {
+      "epoch": 0.9912287825875091,
+      "grad_norm": 5.0721034321084515,
+      "learning_rate": 1.0087909909817228e-09,
+      "loss": 0.3809,
+      "step": 12205
+    },
+    {
+      "epoch": 0.9913099975635508,
+      "grad_norm": 3.1443588440586963,
+      "learning_rate": 9.901973915033004e-10,
+      "loss": 0.5828,
+      "step": 12206
+    },
+    {
+      "epoch": 0.9913912125395923,
+      "grad_norm": 7.70458840955126,
+      "learning_rate": 9.717767102770947e-10,
+      "loss": 0.3964,
+      "step": 12207
+    },
+    {
+      "epoch": 0.9914724275156339,
+      "grad_norm": 5.013950694674269,
+      "learning_rate": 9.535289485781973e-10,
+      "loss": 0.5808,
+      "step": 12208
+    },
+    {
+      "epoch": 0.9915536424916754,
+      "grad_norm": 5.467259982799362,
+      "learning_rate": 9.354541076692092e-10,
+      "loss": 0.3382,
+      "step": 12209
+    },
+    {
+      "epoch": 0.9916348574677171,
+      "grad_norm": 3.502450749974113,
+      "learning_rate": 9.17552188800519e-10,
+      "loss": 0.4828,
+      "step": 12210
+    },
+    {
+      "epoch": 0.9917160724437586,
+      "grad_norm": 4.810040707477097,
+      "learning_rate": 8.998231932108581e-10,
+      "loss": 0.4266,
+      "step": 12211
+    },
+    {
+      "epoch": 0.9917972874198002,
+      "grad_norm": 4.375638079472969,
+      "learning_rate": 8.822671221273005e-10,
+      "loss": 0.5797,
+      "step": 12212
+    },
+    {
+      "epoch": 0.9918785023958417,
+      "grad_norm": 9.293670755342289,
+      "learning_rate": 8.648839767644302e-10,
+      "loss": 0.621,
+      "step": 12213
+    },
+    {
+      "epoch": 0.9919597173718834,
+      "grad_norm": 6.6261486011288335,
+      "learning_rate": 8.476737583251737e-10,
+      "loss": 0.4257,
+      "step": 12214
+    },
+    {
+      "epoch": 0.992040932347925,
+      "grad_norm": 4.4991871759075375,
+      "learning_rate": 8.306364680002454e-10,
+      "loss": 0.522,
+      "step": 12215
+    },
+    {
+      "epoch": 0.9921221473239665,
+      "grad_norm": 4.881228997280039,
+      "learning_rate": 8.137721069687021e-10,
+      "loss": 0.526,
+      "step": 12216
+    },
+    {
+      "epoch": 0.9922033623000082,
+      "grad_norm": 3.6297898118275964,
+      "learning_rate": 7.970806763973882e-10,
+      "loss": 0.4575,
+      "step": 12217
+    },
+    {
+      "epoch": 0.9922845772760497,
+      "grad_norm": 6.246461103137779,
+      "learning_rate": 7.805621774409356e-10,
+      "loss": 0.3662,
+      "step": 12218
+    },
+    {
+      "epoch": 0.9923657922520913,
+      "grad_norm": 5.111388107129699,
+      "learning_rate": 7.642166112428739e-10,
+      "loss": 0.5867,
+      "step": 12219
+    },
+    {
+      "epoch": 0.9924470072281328,
+      "grad_norm": 6.103417103656217,
+      "learning_rate": 7.480439789339655e-10,
+      "loss": 0.489,
+      "step": 12220
+    },
+    {
+      "epoch": 0.9925282222041745,
+      "grad_norm": 5.872345642365211,
+      "learning_rate": 7.320442816333151e-10,
+      "loss": 0.4852,
+      "step": 12221
+    },
+    {
+      "epoch": 0.992609437180216,
+      "grad_norm": 4.896384077479555,
+      "learning_rate": 7.162175204480926e-10,
+      "loss": 0.415,
+      "step": 12222
+    },
+    {
+      "epoch": 0.9926906521562576,
+      "grad_norm": 11.24619397959855,
+      "learning_rate": 7.005636964732554e-10,
+      "loss": 0.5892,
+      "step": 12223
+    },
+    {
+      "epoch": 0.9927718671322991,
+      "grad_norm": 6.417486574394487,
+      "learning_rate": 6.850828107921037e-10,
+      "loss": 0.4542,
+      "step": 12224
+    },
+    {
+      "epoch": 0.9928530821083408,
+      "grad_norm": 5.669079347074385,
+      "learning_rate": 6.697748644757252e-10,
+      "loss": 0.4736,
+      "step": 12225
+    },
+    {
+      "epoch": 0.9929342970843824,
+      "grad_norm": 5.1508249817000475,
+      "learning_rate": 6.546398585832725e-10,
+      "loss": 0.3881,
+      "step": 12226
+    },
+    {
+      "epoch": 0.9930155120604239,
+      "grad_norm": 6.284781187211057,
+      "learning_rate": 6.396777941622412e-10,
+      "loss": 0.5204,
+      "step": 12227
+    },
+    {
+      "epoch": 0.9930967270364656,
+      "grad_norm": 5.867023818166328,
+      "learning_rate": 6.248886722479142e-10,
+      "loss": 0.3868,
+      "step": 12228
+    },
+    {
+      "epoch": 0.9931779420125071,
+      "grad_norm": 7.865759003313058,
+      "learning_rate": 6.10272493863362e-10,
+      "loss": 0.4652,
+      "step": 12229
+    },
+    {
+      "epoch": 0.9932591569885487,
+      "grad_norm": 5.771846652668555,
+      "learning_rate": 5.958292600202753e-10,
+      "loss": 0.4368,
+      "step": 12230
+    },
+    {
+      "epoch": 0.9933403719645902,
+      "grad_norm": 5.555877566179887,
+      "learning_rate": 5.81558971717855e-10,
+      "loss": 0.4511,
+      "step": 12231
+    },
+    {
+      "epoch": 0.9934215869406319,
+      "grad_norm": 7.602271947993169,
+      "learning_rate": 5.674616299436441e-10,
+      "loss": 0.3559,
+      "step": 12232
+    },
+    {
+      "epoch": 0.9935028019166734,
+      "grad_norm": 3.168714257439046,
+      "learning_rate": 5.53537235672974e-10,
+      "loss": 0.6782,
+      "step": 12233
+    },
+    {
+      "epoch": 0.993584016892715,
+      "grad_norm": 8.0565315209248,
+      "learning_rate": 5.397857898692404e-10,
+      "loss": 0.4928,
+      "step": 12234
+    },
+    {
+      "epoch": 0.9936652318687565,
+      "grad_norm": 6.7281592050291525,
+      "learning_rate": 5.262072934841822e-10,
+      "loss": 0.4097,
+      "step": 12235
+    },
+    {
+      "epoch": 0.9937464468447982,
+      "grad_norm": 5.03523479370938,
+      "learning_rate": 5.128017474573254e-10,
+      "loss": 0.5035,
+      "step": 12236
+    },
+    {
+      "epoch": 0.9938276618208398,
+      "grad_norm": 5.581578649372743,
+      "learning_rate": 4.995691527162616e-10,
+      "loss": 0.6748,
+      "step": 12237
+    },
+    {
+      "epoch": 0.9939088767968813,
+      "grad_norm": 5.1959863608312755,
+      "learning_rate": 4.86509510176647e-10,
+      "loss": 0.325,
+      "step": 12238
+    },
+    {
+      "epoch": 0.993990091772923,
+      "grad_norm": 5.218818979766321,
+      "learning_rate": 4.736228207419258e-10,
+      "loss": 0.5546,
+      "step": 12239
+    },
+    {
+      "epoch": 0.9940713067489645,
+      "grad_norm": 6.565937064015396,
+      "learning_rate": 4.60909085304162e-10,
+      "loss": 0.4902,
+      "step": 12240
+    },
+    {
+      "epoch": 0.9941525217250061,
+      "grad_norm": 4.111169296382956,
+      "learning_rate": 4.4836830474265235e-10,
+      "loss": 0.4279,
+      "step": 12241
+    },
+    {
+      "epoch": 0.9942337367010476,
+      "grad_norm": 4.28308564761494,
+      "learning_rate": 4.3600047992559124e-10,
+      "loss": 0.5878,
+      "step": 12242
+    },
+    {
+      "epoch": 0.9943149516770893,
+      "grad_norm": 7.57733637051035,
+      "learning_rate": 4.2380561170840553e-10,
+      "loss": 0.7041,
+      "step": 12243
+    },
+    {
+      "epoch": 0.9943961666531308,
+      "grad_norm": 6.372067043265081,
+      "learning_rate": 4.1178370093486463e-10,
+      "loss": 0.4418,
+      "step": 12244
+    },
+    {
+      "epoch": 0.9944773816291724,
+      "grad_norm": 3.884746049499681,
+      "learning_rate": 3.9993474843735837e-10,
+      "loss": 0.4731,
+      "step": 12245
+    },
+    {
+      "epoch": 0.994558596605214,
+      "grad_norm": 8.66268673867393,
+      "learning_rate": 3.882587550349537e-10,
+      "loss": 0.4845,
+      "step": 12246
+    },
+    {
+      "epoch": 0.9946398115812556,
+      "grad_norm": 4.113892528273869,
+      "learning_rate": 3.7675572153644814e-10,
+      "loss": 0.4564,
+      "step": 12247
+    },
+    {
+      "epoch": 0.9947210265572972,
+      "grad_norm": 4.958163424799454,
+      "learning_rate": 3.6542564873731645e-10,
+      "loss": 0.4016,
+      "step": 12248
+    },
+    {
+      "epoch": 0.9948022415333387,
+      "grad_norm": 3.6688498786954162,
+      "learning_rate": 3.5426853742137613e-10,
+      "loss": 0.6589,
+      "step": 12249
+    },
+    {
+      "epoch": 0.9948834565093804,
+      "grad_norm": 10.717003613916313,
+      "learning_rate": 3.432843883610648e-10,
+      "loss": 0.4469,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9949646714854219,
+      "grad_norm": 4.462573986139782,
+      "learning_rate": 3.3247320231605265e-10,
+      "loss": 0.5045,
+      "step": 12251
+    },
+    {
+      "epoch": 0.9950458864614635,
+      "grad_norm": 5.007934033669896,
+      "learning_rate": 3.218349800346299e-10,
+      "loss": 0.4665,
+      "step": 12252
+    },
+    {
+      "epoch": 0.995127101437505,
+      "grad_norm": 3.873104868996485,
+      "learning_rate": 3.1136972225315197e-10,
+      "loss": 0.5173,
+      "step": 12253
+    },
+    {
+      "epoch": 0.9952083164135467,
+      "grad_norm": 3.3876902753690796,
+      "learning_rate": 3.0107742969520683e-10,
+      "loss": 0.5485,
+      "step": 12254
+    },
+    {
+      "epoch": 0.9952895313895882,
+      "grad_norm": 6.699091681394484,
+      "learning_rate": 2.9095810307328e-10,
+      "loss": 0.367,
+      "step": 12255
+    },
+    {
+      "epoch": 0.9953707463656298,
+      "grad_norm": 6.171401454308493,
+      "learning_rate": 2.810117430873671e-10,
+      "loss": 0.3962,
+      "step": 12256
+    },
+    {
+      "epoch": 0.9954519613416714,
+      "grad_norm": 4.923317857251655,
+      "learning_rate": 2.71238350426084e-10,
+      "loss": 0.4938,
+      "step": 12257
+    },
+    {
+      "epoch": 0.995533176317713,
+      "grad_norm": 5.816170040992201,
+      "learning_rate": 2.61637925765279e-10,
+      "loss": 0.3092,
+      "step": 12258
+    },
+    {
+      "epoch": 0.9956143912937546,
+      "grad_norm": 4.783682438235011,
+      "learning_rate": 2.522104697696981e-10,
+      "loss": 0.4861,
+      "step": 12259
+    },
+    {
+      "epoch": 0.9956956062697961,
+      "grad_norm": 6.542965672565677,
+      "learning_rate": 2.4295598309131973e-10,
+      "loss": 0.4059,
+      "step": 12260
+    },
+    {
+      "epoch": 0.9957768212458378,
+      "grad_norm": 5.290685230558423,
+      "learning_rate": 2.3387446637046506e-10,
+      "loss": 0.433,
+      "step": 12261
+    },
+    {
+      "epoch": 0.9958580362218793,
+      "grad_norm": 4.569512260893426,
+      "learning_rate": 2.2496592023579789e-10,
+      "loss": 0.5698,
+      "step": 12262
+    },
+    {
+      "epoch": 0.9959392511979209,
+      "grad_norm": 13.1705343773864,
+      "learning_rate": 2.1623034530349197e-10,
+      "loss": 0.4349,
+      "step": 12263
+    },
+    {
+      "epoch": 0.9960204661739624,
+      "grad_norm": 11.027548037308877,
+      "learning_rate": 2.076677421783413e-10,
+      "loss": 0.3677,
+      "step": 12264
+    },
+    {
+      "epoch": 0.9961016811500041,
+      "grad_norm": 7.761548829661015,
+      "learning_rate": 1.992781114523723e-10,
+      "loss": 0.4086,
+      "step": 12265
+    },
+    {
+      "epoch": 0.9961828961260456,
+      "grad_norm": 4.496137975811519,
+      "learning_rate": 1.910614537065092e-10,
+      "loss": 0.523,
+      "step": 12266
+    },
+    {
+      "epoch": 0.9962641111020872,
+      "grad_norm": 4.01367922274939,
+      "learning_rate": 1.8301776950918615e-10,
+      "loss": 0.4518,
+      "step": 12267
+    },
+    {
+      "epoch": 0.9963453260781288,
+      "grad_norm": 11.271093585704282,
+      "learning_rate": 1.7514705941690247e-10,
+      "loss": 0.4174,
+      "step": 12268
+    },
+    {
+      "epoch": 0.9964265410541704,
+      "grad_norm": 5.748127242285962,
+      "learning_rate": 1.6744932397422254e-10,
+      "loss": 0.5133,
+      "step": 12269
+    },
+    {
+      "epoch": 0.996507756030212,
+      "grad_norm": 5.1019116955334916,
+      "learning_rate": 1.5992456371377584e-10,
+      "loss": 0.3742,
+      "step": 12270
+    },
+    {
+      "epoch": 0.9965889710062535,
+      "grad_norm": 6.408098676761947,
+      "learning_rate": 1.5257277915653458e-10,
+      "loss": 0.4672,
+      "step": 12271
+    },
+    {
+      "epoch": 0.9966701859822952,
+      "grad_norm": 6.484656082022969,
+      "learning_rate": 1.4539397081070328e-10,
+      "loss": 0.3491,
+      "step": 12272
+    },
+    {
+      "epoch": 0.9967514009583367,
+      "grad_norm": 3.999338555705399,
+      "learning_rate": 1.3838813917366188e-10,
+      "loss": 0.5493,
+      "step": 12273
+    },
+    {
+      "epoch": 0.9968326159343783,
+      "grad_norm": 5.4475266405766565,
+      "learning_rate": 1.3155528472974523e-10,
+      "loss": 0.3952,
+      "step": 12274
+    },
+    {
+      "epoch": 0.9969138309104199,
+      "grad_norm": 4.872599475892686,
+      "learning_rate": 1.2489540795163068e-10,
+      "loss": 0.5474,
+      "step": 12275
+    },
+    {
+      "epoch": 0.9969950458864615,
+      "grad_norm": 3.670731259776475,
+      "learning_rate": 1.18408509300616e-10,
+      "loss": 0.5644,
+      "step": 12276
+    },
+    {
+      "epoch": 0.997076260862503,
+      "grad_norm": 6.2602727752630045,
+      "learning_rate": 1.1209458922495365e-10,
+      "loss": 0.5351,
+      "step": 12277
+    },
+    {
+      "epoch": 0.9971574758385446,
+      "grad_norm": 4.875410969826889,
+      "learning_rate": 1.0595364816207155e-10,
+      "loss": 0.5048,
+      "step": 12278
+    },
+    {
+      "epoch": 0.9972386908145862,
+      "grad_norm": 4.9678315360142395,
+      "learning_rate": 9.998568653690754e-11,
+      "loss": 0.5401,
+      "step": 12279
+    },
+    {
+      "epoch": 0.9973199057906278,
+      "grad_norm": 4.51671927225658,
+      "learning_rate": 9.41907047619095e-11,
+      "loss": 0.516,
+      "step": 12280
+    },
+    {
+      "epoch": 0.9974011207666694,
+      "grad_norm": 4.930864987720356,
+      "learning_rate": 8.856870323842304e-11,
+      "loss": 0.5275,
+      "step": 12281
+    },
+    {
+      "epoch": 0.997482335742711,
+      "grad_norm": 6.047707965039309,
+      "learning_rate": 8.311968235530376e-11,
+      "loss": 0.4833,
+      "step": 12282
+    },
+    {
+      "epoch": 0.9975635507187526,
+      "grad_norm": 7.174138199102842,
+      "learning_rate": 7.784364248974996e-11,
+      "loss": 0.5413,
+      "step": 12283
+    },
+    {
+      "epoch": 0.9976447656947941,
+      "grad_norm": 6.751919593589131,
+      "learning_rate": 7.274058400674744e-11,
+      "loss": 0.4525,
+      "step": 12284
+    },
+    {
+      "epoch": 0.9977259806708357,
+      "grad_norm": 5.318816576138242,
+      "learning_rate": 6.781050725962468e-11,
+      "loss": 0.4845,
+      "step": 12285
+    },
+    {
+      "epoch": 0.9978071956468773,
+      "grad_norm": 4.612324143619252,
+      "learning_rate": 6.30534125889426e-11,
+      "loss": 0.5208,
+      "step": 12286
+    },
+    {
+      "epoch": 0.9978884106229189,
+      "grad_norm": 6.7018444490349065,
+      "learning_rate": 5.846930032443743e-11,
+      "loss": 0.4859,
+      "step": 12287
+    },
+    {
+      "epoch": 0.9979696255989604,
+      "grad_norm": 7.423300865438458,
+      "learning_rate": 5.4058170783077845e-11,
+      "loss": 0.3708,
+      "step": 12288
+    },
+    {
+      "epoch": 0.998050840575002,
+      "grad_norm": 6.668113758835644,
+      "learning_rate": 4.982002427017518e-11,
+      "loss": 0.5712,
+      "step": 12289
+    },
+    {
+      "epoch": 0.9981320555510436,
+      "grad_norm": 7.202885642026081,
+      "learning_rate": 4.5754861078828314e-11,
+      "loss": 0.4483,
+      "step": 12290
+    },
+    {
+      "epoch": 0.9982132705270852,
+      "grad_norm": 8.215415290405625,
+      "learning_rate": 4.186268149047879e-11,
+      "loss": 0.4036,
+      "step": 12291
+    },
+    {
+      "epoch": 0.9982944855031268,
+      "grad_norm": 3.4403988676406008,
+      "learning_rate": 3.814348577435567e-11,
+      "loss": 0.4738,
+      "step": 12292
+    },
+    {
+      "epoch": 0.9983757004791683,
+      "grad_norm": 4.536469686726864,
+      "learning_rate": 3.4597274187753163e-11,
+      "loss": 0.5479,
+      "step": 12293
+    },
+    {
+      "epoch": 0.99845691545521,
+      "grad_norm": 4.166942290927568,
+      "learning_rate": 3.122404697603054e-11,
+      "loss": 0.4988,
+      "step": 12294
+    },
+    {
+      "epoch": 0.9985381304312515,
+      "grad_norm": 3.821490393417459,
+      "learning_rate": 2.8023804372889762e-11,
+      "loss": 0.6298,
+      "step": 12295
+    },
+    {
+      "epoch": 0.9986193454072931,
+      "grad_norm": 4.579385987949685,
+      "learning_rate": 2.499654659954276e-11,
+      "loss": 0.4335,
+      "step": 12296
+    },
+    {
+      "epoch": 0.9987005603833347,
+      "grad_norm": 3.5660288116414067,
+      "learning_rate": 2.214227386554413e-11,
+      "loss": 0.4252,
+      "step": 12297
+    },
+    {
+      "epoch": 0.9987817753593763,
+      "grad_norm": 3.8887534314797567,
+      "learning_rate": 1.9460986368513568e-11,
+      "loss": 0.5141,
+      "step": 12298
+    },
+    {
+      "epoch": 0.9988629903354178,
+      "grad_norm": 9.174267597596703,
+      "learning_rate": 1.6952684293580767e-11,
+      "loss": 0.4097,
+      "step": 12299
+    },
+    {
+      "epoch": 0.9989442053114594,
+      "grad_norm": 4.342933188190943,
+      "learning_rate": 1.4617367814495632e-11,
+      "loss": 0.498,
+      "step": 12300
+    },
+    {
+      "epoch": 0.999025420287501,
+      "grad_norm": 5.079424217042842,
+      "learning_rate": 1.2455037093073163e-11,
+      "loss": 0.5183,
+      "step": 12301
+    },
+    {
+      "epoch": 0.9991066352635426,
+      "grad_norm": 8.135085893159895,
+      "learning_rate": 1.0465692278638361e-11,
+      "loss": 0.5642,
+      "step": 12302
+    },
+    {
+      "epoch": 0.9991878502395842,
+      "grad_norm": 30.1448448129456,
+      "learning_rate": 8.649333509136438e-12,
+      "loss": 0.4073,
+      "step": 12303
+    },
+    {
+      "epoch": 0.9992690652156258,
+      "grad_norm": 3.5750066680144217,
+      "learning_rate": 7.005960910022591e-12,
+      "loss": 0.4412,
+      "step": 12304
+    },
+    {
+      "epoch": 0.9993502801916674,
+      "grad_norm": 13.396727140828338,
+      "learning_rate": 5.535574594817128e-12,
+      "loss": 0.3997,
+      "step": 12305
+    },
+    {
+      "epoch": 0.9994314951677089,
+      "grad_norm": 6.163825951561884,
+      "learning_rate": 4.238174665938122e-12,
+      "loss": 0.3945,
+      "step": 12306
+    },
+    {
+      "epoch": 0.9995127101437505,
+      "grad_norm": 5.7285348007196975,
+      "learning_rate": 3.11376121248097e-12,
+      "loss": 0.4673,
+      "step": 12307
+    },
+    {
+      "epoch": 0.9995939251197921,
+      "grad_norm": 4.103396802719272,
+      "learning_rate": 2.1623343124388405e-12,
+      "loss": 0.4399,
+      "step": 12308
+    },
+    {
+      "epoch": 0.9996751400958337,
+      "grad_norm": 4.630608713223542,
+      "learning_rate": 1.3838940318700034e-12,
+      "loss": 0.4208,
+      "step": 12309
+    },
+    {
+      "epoch": 0.9997563550718752,
+      "grad_norm": 5.594155269906072,
+      "learning_rate": 7.784404243427191e-13,
+      "loss": 0.5584,
+      "step": 12310
+    },
+    {
+      "epoch": 0.9998375700479168,
+      "grad_norm": 7.544818372955951,
+      "learning_rate": 3.4597353176790696e-13,
+      "loss": 0.4048,
+      "step": 12311
+    },
+    {
+      "epoch": 0.9999187850239584,
+      "grad_norm": 4.9742349717120895,
+      "learning_rate": 8.649338439914445e-14,
+      "loss": 0.5781,
+      "step": 12312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.419415245809011,
+      "learning_rate": 0.0,
+      "loss": 0.4966,
+      "step": 12313
+    },
+    {
+      "epoch": 1.0,
+      "step": 12313,
+      "total_flos": 582090100039680.0,
+      "train_loss": 0.5386334688529102,
+      "train_runtime": 33230.6995,
+      "train_samples_per_second": 11.857,
+      "train_steps_per_second": 0.371
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12313,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 4101,
+  "total_flos": 582090100039680.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}