diff --git "a/checkpoint-2000/trainer_state.json" "b/checkpoint-2000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2000/trainer_state.json"
@@ -0,0 +1,14033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.48297512678097076,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002414875633904854,
+      "grad_norm": 0.49120378494262695,
+      "learning_rate": 1.2048192771084338e-06,
+      "loss": 2.1193,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004829751267809708,
+      "grad_norm": 0.4886572062969208,
+      "learning_rate": 2.4096385542168676e-06,
+      "loss": 1.9434,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007244626901714562,
+      "grad_norm": 2.381781578063965,
+      "learning_rate": 3.614457831325301e-06,
+      "loss": 2.3871,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009659502535619416,
+      "grad_norm": 0.5716086030006409,
+      "learning_rate": 4.819277108433735e-06,
+      "loss": 2.0503,
+      "step": 4
+    },
+    {
+      "epoch": 0.001207437816952427,
+      "grad_norm": 1.3921763896942139,
+      "learning_rate": 6.024096385542169e-06,
+      "loss": 1.8545,
+      "step": 5
+    },
+    {
+      "epoch": 0.0014489253803429123,
+      "grad_norm": 0.6292646527290344,
+      "learning_rate": 7.228915662650602e-06,
+      "loss": 2.0255,
+      "step": 6
+    },
+    {
+      "epoch": 0.0016904129437333977,
+      "grad_norm": 0.49030834436416626,
+      "learning_rate": 8.433734939759036e-06,
+      "loss": 1.994,
+      "step": 7
+    },
+    {
+      "epoch": 0.001931900507123883,
+      "grad_norm": 0.47718971967697144,
+      "learning_rate": 9.63855421686747e-06,
+      "loss": 2.0012,
+      "step": 8
+    },
+    {
+      "epoch": 0.0021733880705143687,
+      "grad_norm": 0.6809853315353394,
+      "learning_rate": 1.0843373493975904e-05,
+      "loss": 1.9578,
+      "step": 9
+    },
+    {
+      "epoch": 0.002414875633904854,
+      "grad_norm": 0.4382535517215729,
+      "learning_rate": 1.2048192771084338e-05,
+      "loss": 2.0007,
+      "step": 10
+    },
+    {
+      "epoch": 0.0026563631972953395,
+      "grad_norm": 0.40765514969825745,
+      "learning_rate": 1.3253012048192772e-05,
+      "loss": 1.8652,
+      "step": 11
+    },
+    {
+      "epoch": 0.0028978507606858247,
+      "grad_norm": 0.47924724221229553,
+      "learning_rate": 1.4457831325301205e-05,
+      "loss": 1.9401,
+      "step": 12
+    },
+    {
+      "epoch": 0.0031393383240763103,
+      "grad_norm": 0.40365689992904663,
+      "learning_rate": 1.566265060240964e-05,
+      "loss": 2.016,
+      "step": 13
+    },
+    {
+      "epoch": 0.0033808258874667954,
+      "grad_norm": 0.46215543150901794,
+      "learning_rate": 1.6867469879518073e-05,
+      "loss": 1.9878,
+      "step": 14
+    },
+    {
+      "epoch": 0.003622313450857281,
+      "grad_norm": 0.41287553310394287,
+      "learning_rate": 1.8072289156626505e-05,
+      "loss": 1.8019,
+      "step": 15
+    },
+    {
+      "epoch": 0.003863801014247766,
+      "grad_norm": 0.4323561191558838,
+      "learning_rate": 1.927710843373494e-05,
+      "loss": 1.9787,
+      "step": 16
+    },
+    {
+      "epoch": 0.004105288577638252,
+      "grad_norm": 0.4473198652267456,
+      "learning_rate": 2.0481927710843373e-05,
+      "loss": 1.9847,
+      "step": 17
+    },
+    {
+      "epoch": 0.004346776141028737,
+      "grad_norm": 0.6705676317214966,
+      "learning_rate": 2.168674698795181e-05,
+      "loss": 2.2759,
+      "step": 18
+    },
+    {
+      "epoch": 0.004588263704419222,
+      "grad_norm": 0.4056943655014038,
+      "learning_rate": 2.289156626506024e-05,
+      "loss": 2.0165,
+      "step": 19
+    },
+    {
+      "epoch": 0.004829751267809708,
+      "grad_norm": 0.36379876732826233,
+      "learning_rate": 2.4096385542168677e-05,
+      "loss": 1.896,
+      "step": 20
+    },
+    {
+      "epoch": 0.005071238831200193,
+      "grad_norm": 0.35651546716690063,
+      "learning_rate": 2.530120481927711e-05,
+      "loss": 1.6728,
+      "step": 21
+    },
+    {
+      "epoch": 0.005312726394590679,
+      "grad_norm": 0.3706895411014557,
+      "learning_rate": 2.6506024096385545e-05,
+      "loss": 1.9103,
+      "step": 22
+    },
+    {
+      "epoch": 0.005554213957981164,
+      "grad_norm": 0.355390727519989,
+      "learning_rate": 2.7710843373493977e-05,
+      "loss": 1.7729,
+      "step": 23
+    },
+    {
+      "epoch": 0.005795701521371649,
+      "grad_norm": 0.4054727256298065,
+      "learning_rate": 2.891566265060241e-05,
+      "loss": 1.999,
+      "step": 24
+    },
+    {
+      "epoch": 0.006037189084762135,
+      "grad_norm": 0.3950248062610626,
+      "learning_rate": 3.012048192771085e-05,
+      "loss": 1.8431,
+      "step": 25
+    },
+    {
+      "epoch": 0.0062786766481526205,
+      "grad_norm": 0.35835927724838257,
+      "learning_rate": 3.132530120481928e-05,
+      "loss": 1.713,
+      "step": 26
+    },
+    {
+      "epoch": 0.006520164211543105,
+      "grad_norm": 0.42847758531570435,
+      "learning_rate": 3.253012048192771e-05,
+      "loss": 1.859,
+      "step": 27
+    },
+    {
+      "epoch": 0.006761651774933591,
+      "grad_norm": 0.3585743010044098,
+      "learning_rate": 3.3734939759036146e-05,
+      "loss": 1.7553,
+      "step": 28
+    },
+    {
+      "epoch": 0.0070031393383240765,
+      "grad_norm": 0.35875406861305237,
+      "learning_rate": 3.4939759036144585e-05,
+      "loss": 1.9552,
+      "step": 29
+    },
+    {
+      "epoch": 0.007244626901714562,
+      "grad_norm": 0.3377876281738281,
+      "learning_rate": 3.614457831325301e-05,
+      "loss": 1.7706,
+      "step": 30
+    },
+    {
+      "epoch": 0.007486114465105047,
+      "grad_norm": 0.3570369780063629,
+      "learning_rate": 3.734939759036144e-05,
+      "loss": 1.8494,
+      "step": 31
+    },
+    {
+      "epoch": 0.007727602028495532,
+      "grad_norm": 0.3713253438472748,
+      "learning_rate": 3.855421686746988e-05,
+      "loss": 2.0651,
+      "step": 32
+    },
+    {
+      "epoch": 0.007969089591886018,
+      "grad_norm": 0.3895401656627655,
+      "learning_rate": 3.9759036144578314e-05,
+      "loss": 1.9954,
+      "step": 33
+    },
+    {
+      "epoch": 0.008210577155276504,
+      "grad_norm": 0.3511655032634735,
+      "learning_rate": 4.0963855421686746e-05,
+      "loss": 1.7921,
+      "step": 34
+    },
+    {
+      "epoch": 0.00845206471866699,
+      "grad_norm": 0.32599174976348877,
+      "learning_rate": 4.2168674698795186e-05,
+      "loss": 1.7539,
+      "step": 35
+    },
+    {
+      "epoch": 0.008693552282057475,
+      "grad_norm": 0.32991236448287964,
+      "learning_rate": 4.337349397590362e-05,
+      "loss": 1.7223,
+      "step": 36
+    },
+    {
+      "epoch": 0.008935039845447959,
+      "grad_norm": 0.3855891227722168,
+      "learning_rate": 4.457831325301205e-05,
+      "loss": 1.7663,
+      "step": 37
+    },
+    {
+      "epoch": 0.009176527408838444,
+      "grad_norm": 0.36162635684013367,
+      "learning_rate": 4.578313253012048e-05,
+      "loss": 1.9458,
+      "step": 38
+    },
+    {
+      "epoch": 0.00941801497222893,
+      "grad_norm": 0.33777403831481934,
+      "learning_rate": 4.698795180722892e-05,
+      "loss": 1.854,
+      "step": 39
+    },
+    {
+      "epoch": 0.009659502535619416,
+      "grad_norm": 0.41380664706230164,
+      "learning_rate": 4.8192771084337354e-05,
+      "loss": 1.9908,
+      "step": 40
+    },
+    {
+      "epoch": 0.009900990099009901,
+      "grad_norm": 0.34837591648101807,
+      "learning_rate": 4.9397590361445786e-05,
+      "loss": 1.9103,
+      "step": 41
+    },
+    {
+      "epoch": 0.010142477662400387,
+      "grad_norm": 0.3593843877315521,
+      "learning_rate": 5.060240963855422e-05,
+      "loss": 2.0174,
+      "step": 42
+    },
+    {
+      "epoch": 0.010383965225790872,
+      "grad_norm": 0.3512469530105591,
+      "learning_rate": 5.180722891566265e-05,
+      "loss": 1.8854,
+      "step": 43
+    },
+    {
+      "epoch": 0.010625452789181358,
+      "grad_norm": 0.3344199061393738,
+      "learning_rate": 5.301204819277109e-05,
+      "loss": 1.771,
+      "step": 44
+    },
+    {
+      "epoch": 0.010866940352571842,
+      "grad_norm": 0.31484276056289673,
+      "learning_rate": 5.4216867469879516e-05,
+      "loss": 1.7191,
+      "step": 45
+    },
+    {
+      "epoch": 0.011108427915962327,
+      "grad_norm": 0.33984580636024475,
+      "learning_rate": 5.5421686746987955e-05,
+      "loss": 1.8153,
+      "step": 46
+    },
+    {
+      "epoch": 0.011349915479352813,
+      "grad_norm": 0.37721672654151917,
+      "learning_rate": 5.6626506024096394e-05,
+      "loss": 1.932,
+      "step": 47
+    },
+    {
+      "epoch": 0.011591403042743299,
+      "grad_norm": 0.32222846150398254,
+      "learning_rate": 5.783132530120482e-05,
+      "loss": 1.7607,
+      "step": 48
+    },
+    {
+      "epoch": 0.011832890606133784,
+      "grad_norm": 0.35319098830223083,
+      "learning_rate": 5.903614457831326e-05,
+      "loss": 1.9254,
+      "step": 49
+    },
+    {
+      "epoch": 0.01207437816952427,
+      "grad_norm": 0.37980917096138,
+      "learning_rate": 6.02409638554217e-05,
+      "loss": 1.9552,
+      "step": 50
+    },
+    {
+      "epoch": 0.012315865732914755,
+      "grad_norm": 0.3551584780216217,
+      "learning_rate": 6.144578313253012e-05,
+      "loss": 2.0627,
+      "step": 51
+    },
+    {
+      "epoch": 0.012557353296305241,
+      "grad_norm": 0.3481822609901428,
+      "learning_rate": 6.265060240963856e-05,
+      "loss": 1.9323,
+      "step": 52
+    },
+    {
+      "epoch": 0.012798840859695725,
+      "grad_norm": 0.34489256143569946,
+      "learning_rate": 6.385542168674698e-05,
+      "loss": 1.863,
+      "step": 53
+    },
+    {
+      "epoch": 0.01304032842308621,
+      "grad_norm": 0.34144216775894165,
+      "learning_rate": 6.506024096385543e-05,
+      "loss": 1.9177,
+      "step": 54
+    },
+    {
+      "epoch": 0.013281815986476696,
+      "grad_norm": 0.3356941044330597,
+      "learning_rate": 6.626506024096386e-05,
+      "loss": 1.8424,
+      "step": 55
+    },
+    {
+      "epoch": 0.013523303549867182,
+      "grad_norm": 0.3361498713493347,
+      "learning_rate": 6.746987951807229e-05,
+      "loss": 1.7362,
+      "step": 56
+    },
+    {
+      "epoch": 0.013764791113257667,
+      "grad_norm": 0.37799155712127686,
+      "learning_rate": 6.867469879518072e-05,
+      "loss": 2.1137,
+      "step": 57
+    },
+    {
+      "epoch": 0.014006278676648153,
+      "grad_norm": 0.3507687747478485,
+      "learning_rate": 6.987951807228917e-05,
+      "loss": 1.9008,
+      "step": 58
+    },
+    {
+      "epoch": 0.014247766240038639,
+      "grad_norm": 0.3450543284416199,
+      "learning_rate": 7.108433734939759e-05,
+      "loss": 1.868,
+      "step": 59
+    },
+    {
+      "epoch": 0.014489253803429124,
+      "grad_norm": 0.33962032198905945,
+      "learning_rate": 7.228915662650602e-05,
+      "loss": 1.7856,
+      "step": 60
+    },
+    {
+      "epoch": 0.014730741366819608,
+      "grad_norm": 0.4293012022972107,
+      "learning_rate": 7.349397590361447e-05,
+      "loss": 2.1798,
+      "step": 61
+    },
+    {
+      "epoch": 0.014972228930210094,
+      "grad_norm": 0.35071665048599243,
+      "learning_rate": 7.469879518072289e-05,
+      "loss": 1.8269,
+      "step": 62
+    },
+    {
+      "epoch": 0.01521371649360058,
+      "grad_norm": 0.3420826494693756,
+      "learning_rate": 7.590361445783133e-05,
+      "loss": 1.8103,
+      "step": 63
+    },
+    {
+      "epoch": 0.015455204056991065,
+      "grad_norm": 0.31659749150276184,
+      "learning_rate": 7.710843373493976e-05,
+      "loss": 1.7773,
+      "step": 64
+    },
+    {
+      "epoch": 0.01569669162038155,
+      "grad_norm": 0.3348800837993622,
+      "learning_rate": 7.83132530120482e-05,
+      "loss": 1.8194,
+      "step": 65
+    },
+    {
+      "epoch": 0.015938179183772036,
+      "grad_norm": 0.3300272524356842,
+      "learning_rate": 7.951807228915663e-05,
+      "loss": 1.8126,
+      "step": 66
+    },
+    {
+      "epoch": 0.01617966674716252,
+      "grad_norm": 0.3315858542919159,
+      "learning_rate": 8.072289156626507e-05,
+      "loss": 1.8791,
+      "step": 67
+    },
+    {
+      "epoch": 0.016421154310553007,
+      "grad_norm": 0.34572306275367737,
+      "learning_rate": 8.192771084337349e-05,
+      "loss": 1.8424,
+      "step": 68
+    },
+    {
+      "epoch": 0.016662641873943493,
+      "grad_norm": 0.31780093908309937,
+      "learning_rate": 8.313253012048194e-05,
+      "loss": 1.7617,
+      "step": 69
+    },
+    {
+      "epoch": 0.01690412943733398,
+      "grad_norm": 0.35894426703453064,
+      "learning_rate": 8.433734939759037e-05,
+      "loss": 1.9024,
+      "step": 70
+    },
+    {
+      "epoch": 0.017145617000724464,
+      "grad_norm": 0.32306596636772156,
+      "learning_rate": 8.55421686746988e-05,
+      "loss": 1.8297,
+      "step": 71
+    },
+    {
+      "epoch": 0.01738710456411495,
+      "grad_norm": 0.3225422203540802,
+      "learning_rate": 8.674698795180724e-05,
+      "loss": 1.8403,
+      "step": 72
+    },
+    {
+      "epoch": 0.017628592127505432,
+      "grad_norm": 0.31123486161231995,
+      "learning_rate": 8.795180722891567e-05,
+      "loss": 1.7744,
+      "step": 73
+    },
+    {
+      "epoch": 0.017870079690895917,
+      "grad_norm": 0.32267558574676514,
+      "learning_rate": 8.91566265060241e-05,
+      "loss": 1.8111,
+      "step": 74
+    },
+    {
+      "epoch": 0.018111567254286403,
+      "grad_norm": 0.3544601798057556,
+      "learning_rate": 9.036144578313253e-05,
+      "loss": 2.1014,
+      "step": 75
+    },
+    {
+      "epoch": 0.01835305481767689,
+      "grad_norm": 0.30508580803871155,
+      "learning_rate": 9.156626506024096e-05,
+      "loss": 1.7758,
+      "step": 76
+    },
+    {
+      "epoch": 0.018594542381067374,
+      "grad_norm": 0.321807324886322,
+      "learning_rate": 9.27710843373494e-05,
+      "loss": 1.7328,
+      "step": 77
+    },
+    {
+      "epoch": 0.01883602994445786,
+      "grad_norm": 0.3235253393650055,
+      "learning_rate": 9.397590361445784e-05,
+      "loss": 1.8749,
+      "step": 78
+    },
+    {
+      "epoch": 0.019077517507848345,
+      "grad_norm": 0.29553085565567017,
+      "learning_rate": 9.518072289156626e-05,
+      "loss": 1.6523,
+      "step": 79
+    },
+    {
+      "epoch": 0.01931900507123883,
+      "grad_norm": 0.33573049306869507,
+      "learning_rate": 9.638554216867471e-05,
+      "loss": 1.9299,
+      "step": 80
+    },
+    {
+      "epoch": 0.019560492634629317,
+      "grad_norm": 0.32041892409324646,
+      "learning_rate": 9.759036144578314e-05,
+      "loss": 1.8036,
+      "step": 81
+    },
+    {
+      "epoch": 0.019801980198019802,
+      "grad_norm": 0.3136705458164215,
+      "learning_rate": 9.879518072289157e-05,
+      "loss": 1.7463,
+      "step": 82
+    },
+    {
+      "epoch": 0.020043467761410288,
+      "grad_norm": 0.31231194734573364,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 83
+    },
+    {
+      "epoch": 0.020284955324800773,
+      "grad_norm": 0.3328460454940796,
+      "learning_rate": 9.999998501641807e-05,
+      "loss": 1.9319,
+      "step": 84
+    },
+    {
+      "epoch": 0.02052644288819126,
+      "grad_norm": 0.31596639752388,
+      "learning_rate": 9.999994006568125e-05,
+      "loss": 1.8929,
+      "step": 85
+    },
+    {
+      "epoch": 0.020767930451581745,
+      "grad_norm": 0.42943811416625977,
+      "learning_rate": 9.99998651478165e-05,
+      "loss": 2.0316,
+      "step": 86
+    },
+    {
+      "epoch": 0.02100941801497223,
+      "grad_norm": 0.3127952516078949,
+      "learning_rate": 9.99997602628687e-05,
+      "loss": 1.7111,
+      "step": 87
+    },
+    {
+      "epoch": 0.021250905578362716,
+      "grad_norm": 0.33090245723724365,
+      "learning_rate": 9.999962541090071e-05,
+      "loss": 1.7233,
+      "step": 88
+    },
+    {
+      "epoch": 0.021492393141753198,
+      "grad_norm": 0.3207198679447174,
+      "learning_rate": 9.999946059199337e-05,
+      "loss": 1.8777,
+      "step": 89
+    },
+    {
+      "epoch": 0.021733880705143684,
+      "grad_norm": 0.3291267156600952,
+      "learning_rate": 9.999926580624545e-05,
+      "loss": 1.883,
+      "step": 90
+    },
+    {
+      "epoch": 0.02197536826853417,
+      "grad_norm": 0.3264913856983185,
+      "learning_rate": 9.999904105377372e-05,
+      "loss": 1.8955,
+      "step": 91
+    },
+    {
+      "epoch": 0.022216855831924655,
+      "grad_norm": 0.33100128173828125,
+      "learning_rate": 9.999878633471285e-05,
+      "loss": 2.0645,
+      "step": 92
+    },
+    {
+      "epoch": 0.02245834339531514,
+      "grad_norm": 0.3285035789012909,
+      "learning_rate": 9.999850164921551e-05,
+      "loss": 1.9464,
+      "step": 93
+    },
+    {
+      "epoch": 0.022699830958705626,
+      "grad_norm": 0.31835636496543884,
+      "learning_rate": 9.999818699745234e-05,
+      "loss": 1.834,
+      "step": 94
+    },
+    {
+      "epoch": 0.02294131852209611,
+      "grad_norm": 0.2967548966407776,
+      "learning_rate": 9.999784237961192e-05,
+      "loss": 1.7508,
+      "step": 95
+    },
+    {
+      "epoch": 0.023182806085486597,
+      "grad_norm": 0.37682053446769714,
+      "learning_rate": 9.999746779590078e-05,
+      "loss": 1.9776,
+      "step": 96
+    },
+    {
+      "epoch": 0.023424293648877083,
+      "grad_norm": 0.29876354336738586,
+      "learning_rate": 9.999706324654343e-05,
+      "loss": 1.5886,
+      "step": 97
+    },
+    {
+      "epoch": 0.02366578121226757,
+      "grad_norm": 0.32981058955192566,
+      "learning_rate": 9.999662873178236e-05,
+      "loss": 1.9233,
+      "step": 98
+    },
+    {
+      "epoch": 0.023907268775658054,
+      "grad_norm": 0.33086058497428894,
+      "learning_rate": 9.999616425187796e-05,
+      "loss": 1.9578,
+      "step": 99
+    },
+    {
+      "epoch": 0.02414875633904854,
+      "grad_norm": 0.307336688041687,
+      "learning_rate": 9.999566980710863e-05,
+      "loss": 1.783,
+      "step": 100
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 0.3072247803211212,
+      "learning_rate": 9.999514539777071e-05,
+      "loss": 1.8958,
+      "step": 101
+    },
+    {
+      "epoch": 0.02463173146582951,
+      "grad_norm": 0.2904624044895172,
+      "learning_rate": 9.999459102417847e-05,
+      "loss": 1.7192,
+      "step": 102
+    },
+    {
+      "epoch": 0.024873219029219996,
+      "grad_norm": 0.31059005856513977,
+      "learning_rate": 9.999400668666423e-05,
+      "loss": 1.8329,
+      "step": 103
+    },
+    {
+      "epoch": 0.025114706592610482,
+      "grad_norm": 0.30321741104125977,
+      "learning_rate": 9.999339238557816e-05,
+      "loss": 2.0038,
+      "step": 104
+    },
+    {
+      "epoch": 0.025356194156000968,
+      "grad_norm": 0.3106807768344879,
+      "learning_rate": 9.999274812128847e-05,
+      "loss": 1.8293,
+      "step": 105
+    },
+    {
+      "epoch": 0.02559768171939145,
+      "grad_norm": 0.304987370967865,
+      "learning_rate": 9.999207389418129e-05,
+      "loss": 1.8311,
+      "step": 106
+    },
+    {
+      "epoch": 0.025839169282781935,
+      "grad_norm": 0.31797996163368225,
+      "learning_rate": 9.999136970466067e-05,
+      "loss": 1.9911,
+      "step": 107
+    },
+    {
+      "epoch": 0.02608065684617242,
+      "grad_norm": 0.2915087044239044,
+      "learning_rate": 9.999063555314872e-05,
+      "loss": 1.7453,
+      "step": 108
+    },
+    {
+      "epoch": 0.026322144409562907,
+      "grad_norm": 0.31751200556755066,
+      "learning_rate": 9.99898714400854e-05,
+      "loss": 1.7372,
+      "step": 109
+    },
+    {
+      "epoch": 0.026563631972953392,
+      "grad_norm": 0.2950197458267212,
+      "learning_rate": 9.998907736592873e-05,
+      "loss": 1.7576,
+      "step": 110
+    },
+    {
+      "epoch": 0.026805119536343878,
+      "grad_norm": 0.3286873996257782,
+      "learning_rate": 9.998825333115458e-05,
+      "loss": 2.0046,
+      "step": 111
+    },
+    {
+      "epoch": 0.027046607099734363,
+      "grad_norm": 0.3000810146331787,
+      "learning_rate": 9.998739933625686e-05,
+      "loss": 1.6593,
+      "step": 112
+    },
+    {
+      "epoch": 0.02728809466312485,
+      "grad_norm": 0.33321690559387207,
+      "learning_rate": 9.99865153817474e-05,
+      "loss": 2.0058,
+      "step": 113
+    },
+    {
+      "epoch": 0.027529582226515335,
+      "grad_norm": 0.3196820020675659,
+      "learning_rate": 9.998560146815599e-05,
+      "loss": 2.0243,
+      "step": 114
+    },
+    {
+      "epoch": 0.02777106978990582,
+      "grad_norm": 0.31422409415245056,
+      "learning_rate": 9.998465759603037e-05,
+      "loss": 1.8517,
+      "step": 115
+    },
+    {
+      "epoch": 0.028012557353296306,
+      "grad_norm": 0.29709357023239136,
+      "learning_rate": 9.998368376593627e-05,
+      "loss": 1.7622,
+      "step": 116
+    },
+    {
+      "epoch": 0.02825404491668679,
+      "grad_norm": 0.29982519149780273,
+      "learning_rate": 9.998267997845732e-05,
+      "loss": 1.6767,
+      "step": 117
+    },
+    {
+      "epoch": 0.028495532480077277,
+      "grad_norm": 0.2890397906303406,
+      "learning_rate": 9.998164623419513e-05,
+      "loss": 1.8896,
+      "step": 118
+    },
+    {
+      "epoch": 0.028737020043467763,
+      "grad_norm": 0.28895923495292664,
+      "learning_rate": 9.99805825337693e-05,
+      "loss": 1.8441,
+      "step": 119
+    },
+    {
+      "epoch": 0.028978507606858248,
+      "grad_norm": 0.3054564893245697,
+      "learning_rate": 9.997948887781732e-05,
+      "loss": 1.8496,
+      "step": 120
+    },
+    {
+      "epoch": 0.029219995170248734,
+      "grad_norm": 0.3080004155635834,
+      "learning_rate": 9.99783652669947e-05,
+      "loss": 1.7408,
+      "step": 121
+    },
+    {
+      "epoch": 0.029461482733639216,
+      "grad_norm": 0.2757675349712372,
+      "learning_rate": 9.997721170197482e-05,
+      "loss": 1.7709,
+      "step": 122
+    },
+    {
+      "epoch": 0.0297029702970297,
+      "grad_norm": 0.32389089465141296,
+      "learning_rate": 9.997602818344911e-05,
+      "loss": 1.8518,
+      "step": 123
+    },
+    {
+      "epoch": 0.029944457860420187,
+      "grad_norm": 0.3012002408504486,
+      "learning_rate": 9.997481471212688e-05,
+      "loss": 1.7495,
+      "step": 124
+    },
+    {
+      "epoch": 0.030185945423810673,
+      "grad_norm": 0.3027832508087158,
+      "learning_rate": 9.997357128873541e-05,
+      "loss": 1.8043,
+      "step": 125
+    },
+    {
+      "epoch": 0.03042743298720116,
+      "grad_norm": 0.28534793853759766,
+      "learning_rate": 9.997229791401993e-05,
+      "loss": 1.7346,
+      "step": 126
+    },
+    {
+      "epoch": 0.030668920550591644,
+      "grad_norm": 0.31898003816604614,
+      "learning_rate": 9.997099458874369e-05,
+      "loss": 1.8055,
+      "step": 127
+    },
+    {
+      "epoch": 0.03091040811398213,
+      "grad_norm": 0.31042003631591797,
+      "learning_rate": 9.996966131368775e-05,
+      "loss": 1.7288,
+      "step": 128
+    },
+    {
+      "epoch": 0.031151895677372615,
+      "grad_norm": 0.30200737714767456,
+      "learning_rate": 9.996829808965122e-05,
+      "loss": 1.814,
+      "step": 129
+    },
+    {
+      "epoch": 0.0313933832407631,
+      "grad_norm": 0.3008061647415161,
+      "learning_rate": 9.996690491745118e-05,
+      "loss": 1.8479,
+      "step": 130
+    },
+    {
+      "epoch": 0.031634870804153586,
+      "grad_norm": 0.3122195601463318,
+      "learning_rate": 9.996548179792257e-05,
+      "loss": 1.7699,
+      "step": 131
+    },
+    {
+      "epoch": 0.03187635836754407,
+      "grad_norm": 0.3094089925289154,
+      "learning_rate": 9.996402873191837e-05,
+      "loss": 1.8328,
+      "step": 132
+    },
+    {
+      "epoch": 0.03211784593093456,
+      "grad_norm": 0.3003721237182617,
+      "learning_rate": 9.996254572030943e-05,
+      "loss": 1.7759,
+      "step": 133
+    },
+    {
+      "epoch": 0.03235933349432504,
+      "grad_norm": 0.32589980959892273,
+      "learning_rate": 9.996103276398461e-05,
+      "loss": 1.8535,
+      "step": 134
+    },
+    {
+      "epoch": 0.03260082105771553,
+      "grad_norm": 0.3078492283821106,
+      "learning_rate": 9.995948986385067e-05,
+      "loss": 1.8138,
+      "step": 135
+    },
+    {
+      "epoch": 0.032842308621106014,
+      "grad_norm": 0.28524380922317505,
+      "learning_rate": 9.995791702083232e-05,
+      "loss": 1.7354,
+      "step": 136
+    },
+    {
+      "epoch": 0.0330837961844965,
+      "grad_norm": 0.3418903648853302,
+      "learning_rate": 9.995631423587228e-05,
+      "loss": 1.9039,
+      "step": 137
+    },
+    {
+      "epoch": 0.033325283747886986,
+      "grad_norm": 0.29714182019233704,
+      "learning_rate": 9.995468150993114e-05,
+      "loss": 1.8988,
+      "step": 138
+    },
+    {
+      "epoch": 0.03356677131127747,
+      "grad_norm": 0.34489384293556213,
+      "learning_rate": 9.995301884398747e-05,
+      "loss": 1.9888,
+      "step": 139
+    },
+    {
+      "epoch": 0.03380825887466796,
+      "grad_norm": 0.2920478284358978,
+      "learning_rate": 9.995132623903777e-05,
+      "loss": 1.6168,
+      "step": 140
+    },
+    {
+      "epoch": 0.03404974643805844,
+      "grad_norm": 0.2992519438266754,
+      "learning_rate": 9.994960369609649e-05,
+      "loss": 1.8919,
+      "step": 141
+    },
+    {
+      "epoch": 0.03429123400144893,
+      "grad_norm": 0.29365047812461853,
+      "learning_rate": 9.994785121619605e-05,
+      "loss": 1.7853,
+      "step": 142
+    },
+    {
+      "epoch": 0.034532721564839414,
+      "grad_norm": 0.30724966526031494,
+      "learning_rate": 9.994606880038676e-05,
+      "loss": 1.9399,
+      "step": 143
+    },
+    {
+      "epoch": 0.0347742091282299,
+      "grad_norm": 0.41282230615615845,
+      "learning_rate": 9.994425644973691e-05,
+      "loss": 2.0693,
+      "step": 144
+    },
+    {
+      "epoch": 0.035015696691620385,
+      "grad_norm": 0.30199870467185974,
+      "learning_rate": 9.99424141653327e-05,
+      "loss": 1.8273,
+      "step": 145
+    },
+    {
+      "epoch": 0.035257184255010864,
+      "grad_norm": 0.28794577717781067,
+      "learning_rate": 9.994054194827832e-05,
+      "loss": 1.8246,
+      "step": 146
+    },
+    {
+      "epoch": 0.03549867181840135,
+      "grad_norm": 0.2896375358104706,
+      "learning_rate": 9.993863979969587e-05,
+      "loss": 1.8404,
+      "step": 147
+    },
+    {
+      "epoch": 0.035740159381791835,
+      "grad_norm": 0.32256078720092773,
+      "learning_rate": 9.993670772072537e-05,
+      "loss": 1.8969,
+      "step": 148
+    },
+    {
+      "epoch": 0.03598164694518232,
+      "grad_norm": 0.27402758598327637,
+      "learning_rate": 9.993474571252479e-05,
+      "loss": 1.7122,
+      "step": 149
+    },
+    {
+      "epoch": 0.036223134508572806,
+      "grad_norm": 0.32481542229652405,
+      "learning_rate": 9.993275377627008e-05,
+      "loss": 2.0565,
+      "step": 150
+    },
+    {
+      "epoch": 0.03646462207196329,
+      "grad_norm": 0.26531848311424255,
+      "learning_rate": 9.993073191315507e-05,
+      "loss": 1.5844,
+      "step": 151
+    },
+    {
+      "epoch": 0.03670610963535378,
+      "grad_norm": 0.2869581878185272,
+      "learning_rate": 9.992868012439159e-05,
+      "loss": 1.7028,
+      "step": 152
+    },
+    {
+      "epoch": 0.03694759719874426,
+      "grad_norm": 0.2975616753101349,
+      "learning_rate": 9.992659841120929e-05,
+      "loss": 1.9539,
+      "step": 153
+    },
+    {
+      "epoch": 0.03718908476213475,
+      "grad_norm": 0.322748064994812,
+      "learning_rate": 9.99244867748559e-05,
+      "loss": 2.0498,
+      "step": 154
+    },
+    {
+      "epoch": 0.037430572325525234,
+      "grad_norm": 0.2980214059352875,
+      "learning_rate": 9.992234521659698e-05,
+      "loss": 1.9035,
+      "step": 155
+    },
+    {
+      "epoch": 0.03767205988891572,
+      "grad_norm": 0.3023289144039154,
+      "learning_rate": 9.992017373771607e-05,
+      "loss": 1.8374,
+      "step": 156
+    },
+    {
+      "epoch": 0.037913547452306205,
+      "grad_norm": 0.2790420949459076,
+      "learning_rate": 9.991797233951462e-05,
+      "loss": 1.4934,
+      "step": 157
+    },
+    {
+      "epoch": 0.03815503501569669,
+      "grad_norm": 0.2871580719947815,
+      "learning_rate": 9.991574102331204e-05,
+      "loss": 1.7839,
+      "step": 158
+    },
+    {
+      "epoch": 0.038396522579087176,
+      "grad_norm": 0.2993938624858856,
+      "learning_rate": 9.991347979044566e-05,
+      "loss": 1.8198,
+      "step": 159
+    },
+    {
+      "epoch": 0.03863801014247766,
+      "grad_norm": 0.30797529220581055,
+      "learning_rate": 9.99111886422707e-05,
+      "loss": 1.8831,
+      "step": 160
+    },
+    {
+      "epoch": 0.03887949770586815,
+      "grad_norm": 0.30790555477142334,
+      "learning_rate": 9.990886758016038e-05,
+      "loss": 1.9753,
+      "step": 161
+    },
+    {
+      "epoch": 0.03912098526925863,
+      "grad_norm": 0.2984273135662079,
+      "learning_rate": 9.990651660550579e-05,
+      "loss": 1.656,
+      "step": 162
+    },
+    {
+      "epoch": 0.03936247283264912,
+      "grad_norm": 0.30045798420906067,
+      "learning_rate": 9.990413571971598e-05,
+      "loss": 1.7617,
+      "step": 163
+    },
+    {
+      "epoch": 0.039603960396039604,
+      "grad_norm": 0.2841191291809082,
+      "learning_rate": 9.990172492421791e-05,
+      "loss": 1.7359,
+      "step": 164
+    },
+    {
+      "epoch": 0.03984544795943009,
+      "grad_norm": 0.363344669342041,
+      "learning_rate": 9.98992842204565e-05,
+      "loss": 2.0574,
+      "step": 165
+    },
+    {
+      "epoch": 0.040086935522820576,
+      "grad_norm": 0.2936279773712158,
+      "learning_rate": 9.989681360989454e-05,
+      "loss": 1.8597,
+      "step": 166
+    },
+    {
+      "epoch": 0.04032842308621106,
+      "grad_norm": 0.31181731820106506,
+      "learning_rate": 9.989431309401277e-05,
+      "loss": 2.0107,
+      "step": 167
+    },
+    {
+      "epoch": 0.04056991064960155,
+      "grad_norm": 0.28021594882011414,
+      "learning_rate": 9.989178267430988e-05,
+      "loss": 1.5916,
+      "step": 168
+    },
+    {
+      "epoch": 0.04081139821299203,
+      "grad_norm": 0.3215208053588867,
+      "learning_rate": 9.988922235230247e-05,
+      "loss": 1.9517,
+      "step": 169
+    },
+    {
+      "epoch": 0.04105288577638252,
+      "grad_norm": 0.2947755753993988,
+      "learning_rate": 9.988663212952502e-05,
+      "loss": 1.8771,
+      "step": 170
+    },
+    {
+      "epoch": 0.041294373339773004,
+      "grad_norm": 0.28413596749305725,
+      "learning_rate": 9.988401200752997e-05,
+      "loss": 1.8199,
+      "step": 171
+    },
+    {
+      "epoch": 0.04153586090316349,
+      "grad_norm": 0.3089015781879425,
+      "learning_rate": 9.988136198788769e-05,
+      "loss": 1.7263,
+      "step": 172
+    },
+    {
+      "epoch": 0.041777348466553975,
+      "grad_norm": 0.30298158526420593,
+      "learning_rate": 9.987868207218645e-05,
+      "loss": 1.7411,
+      "step": 173
+    },
+    {
+      "epoch": 0.04201883602994446,
+      "grad_norm": 0.2908902168273926,
+      "learning_rate": 9.98759722620324e-05,
+      "loss": 1.7757,
+      "step": 174
+    },
+    {
+      "epoch": 0.042260323593334946,
+      "grad_norm": 0.3005327880382538,
+      "learning_rate": 9.98732325590497e-05,
+      "loss": 1.7623,
+      "step": 175
+    },
+    {
+      "epoch": 0.04250181115672543,
+      "grad_norm": 0.31084272265434265,
+      "learning_rate": 9.987046296488034e-05,
+      "loss": 1.8229,
+      "step": 176
+    },
+    {
+      "epoch": 0.04274329872011592,
+      "grad_norm": 0.3010302186012268,
+      "learning_rate": 9.986766348118428e-05,
+      "loss": 1.847,
+      "step": 177
+    },
+    {
+      "epoch": 0.042984786283506396,
+      "grad_norm": 0.5154542922973633,
+      "learning_rate": 9.986483410963934e-05,
+      "loss": 1.9374,
+      "step": 178
+    },
+    {
+      "epoch": 0.04322627384689688,
+      "grad_norm": 0.2857230603694916,
+      "learning_rate": 9.98619748519413e-05,
+      "loss": 1.6965,
+      "step": 179
+    },
+    {
+      "epoch": 0.04346776141028737,
+      "grad_norm": 0.29431548714637756,
+      "learning_rate": 9.985908570980387e-05,
+      "loss": 1.8925,
+      "step": 180
+    },
+    {
+      "epoch": 0.04370924897367785,
+      "grad_norm": 0.2822973430156708,
+      "learning_rate": 9.985616668495859e-05,
+      "loss": 1.7492,
+      "step": 181
+    },
+    {
+      "epoch": 0.04395073653706834,
+      "grad_norm": 0.28814923763275146,
+      "learning_rate": 9.985321777915497e-05,
+      "loss": 1.669,
+      "step": 182
+    },
+    {
+      "epoch": 0.044192224100458824,
+      "grad_norm": 0.30659806728363037,
+      "learning_rate": 9.985023899416041e-05,
+      "loss": 1.7437,
+      "step": 183
+    },
+    {
+      "epoch": 0.04443371166384931,
+      "grad_norm": 0.4332549273967743,
+      "learning_rate": 9.984723033176026e-05,
+      "loss": 2.4215,
+      "step": 184
+    },
+    {
+      "epoch": 0.044675199227239795,
+      "grad_norm": 0.29047444462776184,
+      "learning_rate": 9.984419179375773e-05,
+      "loss": 1.8027,
+      "step": 185
+    },
+    {
+      "epoch": 0.04491668679063028,
+      "grad_norm": 0.3238014876842499,
+      "learning_rate": 9.984112338197392e-05,
+      "loss": 1.9597,
+      "step": 186
+    },
+    {
+      "epoch": 0.045158174354020766,
+      "grad_norm": 0.29162389039993286,
+      "learning_rate": 9.983802509824786e-05,
+      "loss": 1.8871,
+      "step": 187
+    },
+    {
+      "epoch": 0.04539966191741125,
+      "grad_norm": 0.3024207651615143,
+      "learning_rate": 9.983489694443652e-05,
+      "loss": 1.7083,
+      "step": 188
+    },
+    {
+      "epoch": 0.04564114948080174,
+      "grad_norm": 0.31812480092048645,
+      "learning_rate": 9.983173892241472e-05,
+      "loss": 1.8127,
+      "step": 189
+    },
+    {
+      "epoch": 0.04588263704419222,
+      "grad_norm": 0.2975609004497528,
+      "learning_rate": 9.982855103407521e-05,
+      "loss": 1.8012,
+      "step": 190
+    },
+    {
+      "epoch": 0.04612412460758271,
+      "grad_norm": 0.29555127024650574,
+      "learning_rate": 9.982533328132862e-05,
+      "loss": 1.8123,
+      "step": 191
+    },
+    {
+      "epoch": 0.046365612170973194,
+      "grad_norm": 0.35191309452056885,
+      "learning_rate": 9.982208566610348e-05,
+      "loss": 1.8852,
+      "step": 192
+    },
+    {
+      "epoch": 0.04660709973436368,
+      "grad_norm": 0.2966957688331604,
+      "learning_rate": 9.981880819034624e-05,
+      "loss": 1.807,
+      "step": 193
+    },
+    {
+      "epoch": 0.046848587297754166,
+      "grad_norm": 0.3028441071510315,
+      "learning_rate": 9.981550085602123e-05,
+      "loss": 1.8438,
+      "step": 194
+    },
+    {
+      "epoch": 0.04709007486114465,
+      "grad_norm": 0.2984451353549957,
+      "learning_rate": 9.981216366511067e-05,
+      "loss": 1.8438,
+      "step": 195
+    },
+    {
+      "epoch": 0.04733156242453514,
+      "grad_norm": 0.30961814522743225,
+      "learning_rate": 9.980879661961469e-05,
+      "loss": 2.0188,
+      "step": 196
+    },
+    {
+      "epoch": 0.04757304998792562,
+      "grad_norm": 0.2839982807636261,
+      "learning_rate": 9.98053997215513e-05,
+      "loss": 1.6599,
+      "step": 197
+    },
+    {
+      "epoch": 0.04781453755131611,
+      "grad_norm": 0.3275313377380371,
+      "learning_rate": 9.980197297295644e-05,
+      "loss": 2.0287,
+      "step": 198
+    },
+    {
+      "epoch": 0.048056025114706594,
+      "grad_norm": 0.27704620361328125,
+      "learning_rate": 9.979851637588386e-05,
+      "loss": 1.67,
+      "step": 199
+    },
+    {
+      "epoch": 0.04829751267809708,
+      "grad_norm": 0.3120492696762085,
+      "learning_rate": 9.979502993240528e-05,
+      "loss": 2.0147,
+      "step": 200
+    },
+    {
+      "epoch": 0.048539000241487565,
+      "grad_norm": 1.1097334623336792,
+      "learning_rate": 9.979151364461025e-05,
+      "loss": 1.5357,
+      "step": 201
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 0.3085632622241974,
+      "learning_rate": 9.978796751460627e-05,
+      "loss": 1.8097,
+      "step": 202
+    },
+    {
+      "epoch": 0.049021975368268536,
+      "grad_norm": 0.2954683303833008,
+      "learning_rate": 9.978439154451867e-05,
+      "loss": 1.7476,
+      "step": 203
+    },
+    {
+      "epoch": 0.04926346293165902,
+      "grad_norm": 0.36696991324424744,
+      "learning_rate": 9.978078573649068e-05,
+      "loss": 1.8692,
+      "step": 204
+    },
+    {
+      "epoch": 0.04950495049504951,
+      "grad_norm": 0.2759391963481903,
+      "learning_rate": 9.977715009268343e-05,
+      "loss": 1.8649,
+      "step": 205
+    },
+    {
+      "epoch": 0.04974643805843999,
+      "grad_norm": 0.28329887986183167,
+      "learning_rate": 9.97734846152759e-05,
+      "loss": 1.7726,
+      "step": 206
+    },
+    {
+      "epoch": 0.04998792562183048,
+      "grad_norm": 0.2912246286869049,
+      "learning_rate": 9.976978930646497e-05,
+      "loss": 1.7696,
+      "step": 207
+    },
+    {
+      "epoch": 0.050229413185220964,
+      "grad_norm": 0.2736220061779022,
+      "learning_rate": 9.976606416846542e-05,
+      "loss": 1.6264,
+      "step": 208
+    },
+    {
+      "epoch": 0.05047090074861145,
+      "grad_norm": 0.3122333586215973,
+      "learning_rate": 9.976230920350985e-05,
+      "loss": 1.7356,
+      "step": 209
+    },
+    {
+      "epoch": 0.050712388312001935,
+      "grad_norm": 0.3272218406200409,
+      "learning_rate": 9.975852441384882e-05,
+      "loss": 1.8347,
+      "step": 210
+    },
+    {
+      "epoch": 0.050953875875392414,
+      "grad_norm": 0.28549280762672424,
+      "learning_rate": 9.975470980175068e-05,
+      "loss": 1.8455,
+      "step": 211
+    },
+    {
+      "epoch": 0.0511953634387829,
+      "grad_norm": 0.3388885259628296,
+      "learning_rate": 9.975086536950171e-05,
+      "loss": 1.7048,
+      "step": 212
+    },
+    {
+      "epoch": 0.051436851002173385,
+      "grad_norm": 0.2744766175746918,
+      "learning_rate": 9.974699111940605e-05,
+      "loss": 1.7144,
+      "step": 213
+    },
+    {
+      "epoch": 0.05167833856556387,
+      "grad_norm": 0.28307104110717773,
+      "learning_rate": 9.97430870537857e-05,
+      "loss": 1.6624,
+      "step": 214
+    },
+    {
+      "epoch": 0.051919826128954356,
+      "grad_norm": 0.31305596232414246,
+      "learning_rate": 9.973915317498053e-05,
+      "loss": 1.9547,
+      "step": 215
+    },
+    {
+      "epoch": 0.05216131369234484,
+      "grad_norm": 0.28607913851737976,
+      "learning_rate": 9.973518948534826e-05,
+      "loss": 1.7163,
+      "step": 216
+    },
+    {
+      "epoch": 0.05240280125573533,
+      "grad_norm": 0.28641268610954285,
+      "learning_rate": 9.973119598726456e-05,
+      "loss": 1.7557,
+      "step": 217
+    },
+    {
+      "epoch": 0.05264428881912581,
+      "grad_norm": 0.2846461534500122,
+      "learning_rate": 9.972717268312286e-05,
+      "loss": 1.835,
+      "step": 218
+    },
+    {
+      "epoch": 0.0528857763825163,
+      "grad_norm": 0.30998507142066956,
+      "learning_rate": 9.972311957533452e-05,
+      "loss": 1.8987,
+      "step": 219
+    },
+    {
+      "epoch": 0.053127263945906784,
+      "grad_norm": 0.28558000922203064,
+      "learning_rate": 9.971903666632873e-05,
+      "loss": 1.8831,
+      "step": 220
+    },
+    {
+      "epoch": 0.05336875150929727,
+      "grad_norm": 0.3230333924293518,
+      "learning_rate": 9.971492395855256e-05,
+      "loss": 1.9183,
+      "step": 221
+    },
+    {
+      "epoch": 0.053610239072687756,
+      "grad_norm": 0.2764125466346741,
+      "learning_rate": 9.971078145447092e-05,
+      "loss": 1.6978,
+      "step": 222
+    },
+    {
+      "epoch": 0.05385172663607824,
+      "grad_norm": 0.2793998122215271,
+      "learning_rate": 9.970660915656664e-05,
+      "loss": 1.8111,
+      "step": 223
+    },
+    {
+      "epoch": 0.05409321419946873,
+      "grad_norm": 0.30916768312454224,
+      "learning_rate": 9.970240706734031e-05,
+      "loss": 1.8777,
+      "step": 224
+    },
+    {
+      "epoch": 0.05433470176285921,
+      "grad_norm": 0.28356650471687317,
+      "learning_rate": 9.969817518931043e-05,
+      "loss": 1.8503,
+      "step": 225
+    },
+    {
+      "epoch": 0.0545761893262497,
+      "grad_norm": 0.28640586137771606,
+      "learning_rate": 9.969391352501336e-05,
+      "loss": 1.9375,
+      "step": 226
+    },
+    {
+      "epoch": 0.054817676889640184,
+      "grad_norm": 0.2708923816680908,
+      "learning_rate": 9.96896220770033e-05,
+      "loss": 1.693,
+      "step": 227
+    },
+    {
+      "epoch": 0.05505916445303067,
+      "grad_norm": 0.3123991787433624,
+      "learning_rate": 9.96853008478523e-05,
+      "loss": 1.9336,
+      "step": 228
+    },
+    {
+      "epoch": 0.055300652016421155,
+      "grad_norm": 0.2919258773326874,
+      "learning_rate": 9.968094984015025e-05,
+      "loss": 1.8187,
+      "step": 229
+    },
+    {
+      "epoch": 0.05554213957981164,
+      "grad_norm": 0.2894388735294342,
+      "learning_rate": 9.96765690565049e-05,
+      "loss": 1.7333,
+      "step": 230
+    },
+    {
+      "epoch": 0.055783627143202126,
+      "grad_norm": 0.30706238746643066,
+      "learning_rate": 9.967215849954185e-05,
+      "loss": 2.0766,
+      "step": 231
+    },
+    {
+      "epoch": 0.05602511470659261,
+      "grad_norm": 0.2915318012237549,
+      "learning_rate": 9.966771817190455e-05,
+      "loss": 1.8268,
+      "step": 232
+    },
+    {
+      "epoch": 0.0562666022699831,
+      "grad_norm": 0.28697696328163147,
+      "learning_rate": 9.966324807625424e-05,
+      "loss": 1.8361,
+      "step": 233
+    },
+    {
+      "epoch": 0.05650808983337358,
+      "grad_norm": 0.3011184334754944,
+      "learning_rate": 9.965874821527008e-05,
+      "loss": 1.9076,
+      "step": 234
+    },
+    {
+      "epoch": 0.05674957739676407,
+      "grad_norm": 0.2749825119972229,
+      "learning_rate": 9.965421859164903e-05,
+      "loss": 1.6789,
+      "step": 235
+    },
+    {
+      "epoch": 0.056991064960154554,
+      "grad_norm": 0.281316876411438,
+      "learning_rate": 9.964965920810586e-05,
+      "loss": 1.735,
+      "step": 236
+    },
+    {
+      "epoch": 0.05723255252354504,
+      "grad_norm": 0.2849574387073517,
+      "learning_rate": 9.964507006737322e-05,
+      "loss": 1.8415,
+      "step": 237
+    },
+    {
+      "epoch": 0.057474040086935525,
+      "grad_norm": 0.26775339245796204,
+      "learning_rate": 9.964045117220161e-05,
+      "loss": 1.7411,
+      "step": 238
+    },
+    {
+      "epoch": 0.05771552765032601,
+      "grad_norm": 0.3151552379131317,
+      "learning_rate": 9.96358025253593e-05,
+      "loss": 1.7063,
+      "step": 239
+    },
+    {
+      "epoch": 0.057957015213716497,
+      "grad_norm": 0.2982829213142395,
+      "learning_rate": 9.963112412963242e-05,
+      "loss": 1.8773,
+      "step": 240
+    },
+    {
+      "epoch": 0.05819850277710698,
+      "grad_norm": 0.2909030020236969,
+      "learning_rate": 9.962641598782495e-05,
+      "loss": 1.891,
+      "step": 241
+    },
+    {
+      "epoch": 0.05843999034049747,
+      "grad_norm": 0.3036896288394928,
+      "learning_rate": 9.962167810275869e-05,
+      "loss": 1.9357,
+      "step": 242
+    },
+    {
+      "epoch": 0.058681477903887946,
+      "grad_norm": 0.2985946536064148,
+      "learning_rate": 9.961691047727324e-05,
+      "loss": 1.837,
+      "step": 243
+    },
+    {
+      "epoch": 0.05892296546727843,
+      "grad_norm": 0.278512567281723,
+      "learning_rate": 9.961211311422607e-05,
+      "loss": 1.7386,
+      "step": 244
+    },
+    {
+      "epoch": 0.05916445303066892,
+      "grad_norm": 0.2755807340145111,
+      "learning_rate": 9.960728601649242e-05,
+      "loss": 1.6836,
+      "step": 245
+    },
+    {
+      "epoch": 0.0594059405940594,
+      "grad_norm": 0.293917715549469,
+      "learning_rate": 9.960242918696539e-05,
+      "loss": 1.8625,
+      "step": 246
+    },
+    {
+      "epoch": 0.05964742815744989,
+      "grad_norm": 0.28448382019996643,
+      "learning_rate": 9.95975426285559e-05,
+      "loss": 1.7909,
+      "step": 247
+    },
+    {
+      "epoch": 0.059888915720840374,
+      "grad_norm": 0.2795000970363617,
+      "learning_rate": 9.959262634419266e-05,
+      "loss": 1.6246,
+      "step": 248
+    },
+    {
+      "epoch": 0.06013040328423086,
+      "grad_norm": 0.29517507553100586,
+      "learning_rate": 9.958768033682222e-05,
+      "loss": 1.7155,
+      "step": 249
+    },
+    {
+      "epoch": 0.060371890847621346,
+      "grad_norm": 0.27212411165237427,
+      "learning_rate": 9.958270460940893e-05,
+      "loss": 1.7492,
+      "step": 250
+    },
+    {
+      "epoch": 0.06061337841101183,
+      "grad_norm": 0.29364678263664246,
+      "learning_rate": 9.957769916493496e-05,
+      "loss": 1.7754,
+      "step": 251
+    },
+    {
+      "epoch": 0.06085486597440232,
+      "grad_norm": 0.29167085886001587,
+      "learning_rate": 9.95726640064003e-05,
+      "loss": 1.5348,
+      "step": 252
+    },
+    {
+      "epoch": 0.0610963535377928,
+      "grad_norm": 0.2989068925380707,
+      "learning_rate": 9.956759913682272e-05,
+      "loss": 1.7296,
+      "step": 253
+    },
+    {
+      "epoch": 0.06133784110118329,
+      "grad_norm": 0.2934739887714386,
+      "learning_rate": 9.956250455923784e-05,
+      "loss": 1.8387,
+      "step": 254
+    },
+    {
+      "epoch": 0.061579328664573774,
+      "grad_norm": 0.2777479588985443,
+      "learning_rate": 9.955738027669903e-05,
+      "loss": 1.7859,
+      "step": 255
+    },
+    {
+      "epoch": 0.06182081622796426,
+      "grad_norm": 0.3208652436733246,
+      "learning_rate": 9.955222629227752e-05,
+      "loss": 1.8396,
+      "step": 256
+    },
+    {
+      "epoch": 0.062062303791354745,
+      "grad_norm": 0.2889414131641388,
+      "learning_rate": 9.95470426090623e-05,
+      "loss": 1.7689,
+      "step": 257
+    },
+    {
+      "epoch": 0.06230379135474523,
+      "grad_norm": 0.29929113388061523,
+      "learning_rate": 9.95418292301602e-05,
+      "loss": 1.7447,
+      "step": 258
+    },
+    {
+      "epoch": 0.06254527891813572,
+      "grad_norm": 0.304668664932251,
+      "learning_rate": 9.953658615869579e-05,
+      "loss": 1.8408,
+      "step": 259
+    },
+    {
+      "epoch": 0.0627867664815262,
+      "grad_norm": 0.26577287912368774,
+      "learning_rate": 9.953131339781149e-05,
+      "loss": 1.7391,
+      "step": 260
+    },
+    {
+      "epoch": 0.06302825404491669,
+      "grad_norm": 0.26456665992736816,
+      "learning_rate": 9.952601095066749e-05,
+      "loss": 1.588,
+      "step": 261
+    },
+    {
+      "epoch": 0.06326974160830717,
+      "grad_norm": 0.2887117862701416,
+      "learning_rate": 9.952067882044177e-05,
+      "loss": 1.7707,
+      "step": 262
+    },
+    {
+      "epoch": 0.06351122917169766,
+      "grad_norm": 0.2779730260372162,
+      "learning_rate": 9.951531701033012e-05,
+      "loss": 1.7844,
+      "step": 263
+    },
+    {
+      "epoch": 0.06375271673508814,
+      "grad_norm": 0.27247482538223267,
+      "learning_rate": 9.95099255235461e-05,
+      "loss": 1.6796,
+      "step": 264
+    },
+    {
+      "epoch": 0.06399420429847863,
+      "grad_norm": 0.2854093611240387,
+      "learning_rate": 9.950450436332103e-05,
+      "loss": 1.7844,
+      "step": 265
+    },
+    {
+      "epoch": 0.06423569186186912,
+      "grad_norm": 0.3516107201576233,
+      "learning_rate": 9.94990535329041e-05,
+      "loss": 2.0682,
+      "step": 266
+    },
+    {
+      "epoch": 0.0644771794252596,
+      "grad_norm": 0.2657569944858551,
+      "learning_rate": 9.94935730355622e-05,
+      "loss": 1.796,
+      "step": 267
+    },
+    {
+      "epoch": 0.06471866698865009,
+      "grad_norm": 0.29638686776161194,
+      "learning_rate": 9.948806287458003e-05,
+      "loss": 1.7465,
+      "step": 268
+    },
+    {
+      "epoch": 0.06496015455204057,
+      "grad_norm": 0.28134340047836304,
+      "learning_rate": 9.948252305326008e-05,
+      "loss": 1.5347,
+      "step": 269
+    },
+    {
+      "epoch": 0.06520164211543106,
+      "grad_norm": 0.2832060158252716,
+      "learning_rate": 9.947695357492258e-05,
+      "loss": 1.8296,
+      "step": 270
+    },
+    {
+      "epoch": 0.06544312967882154,
+      "grad_norm": 0.3167952597141266,
+      "learning_rate": 9.947135444290558e-05,
+      "loss": 2.1621,
+      "step": 271
+    },
+    {
+      "epoch": 0.06568461724221203,
+      "grad_norm": 0.28933095932006836,
+      "learning_rate": 9.946572566056487e-05,
+      "loss": 1.764,
+      "step": 272
+    },
+    {
+      "epoch": 0.06592610480560251,
+      "grad_norm": 0.30958911776542664,
+      "learning_rate": 9.946006723127403e-05,
+      "loss": 1.9259,
+      "step": 273
+    },
+    {
+      "epoch": 0.066167592368993,
+      "grad_norm": 0.27871811389923096,
+      "learning_rate": 9.945437915842441e-05,
+      "loss": 1.8828,
+      "step": 274
+    },
+    {
+      "epoch": 0.06640907993238349,
+      "grad_norm": 0.2978667914867401,
+      "learning_rate": 9.944866144542511e-05,
+      "loss": 1.8764,
+      "step": 275
+    },
+    {
+      "epoch": 0.06665056749577397,
+      "grad_norm": 0.27303117513656616,
+      "learning_rate": 9.9442914095703e-05,
+      "loss": 1.7793,
+      "step": 276
+    },
+    {
+      "epoch": 0.06689205505916446,
+      "grad_norm": 0.28109732270240784,
+      "learning_rate": 9.94371371127027e-05,
+      "loss": 1.7233,
+      "step": 277
+    },
+    {
+      "epoch": 0.06713354262255494,
+      "grad_norm": 0.29150810837745667,
+      "learning_rate": 9.943133049988663e-05,
+      "loss": 1.8243,
+      "step": 278
+    },
+    {
+      "epoch": 0.06737503018594543,
+      "grad_norm": 0.3005611300468445,
+      "learning_rate": 9.942549426073494e-05,
+      "loss": 1.8877,
+      "step": 279
+    },
+    {
+      "epoch": 0.06761651774933591,
+      "grad_norm": 0.2712879478931427,
+      "learning_rate": 9.941962839874555e-05,
+      "loss": 1.8297,
+      "step": 280
+    },
+    {
+      "epoch": 0.0678580053127264,
+      "grad_norm": 0.29506200551986694,
+      "learning_rate": 9.94137329174341e-05,
+      "loss": 1.746,
+      "step": 281
+    },
+    {
+      "epoch": 0.06809949287611688,
+      "grad_norm": 0.2840847671031952,
+      "learning_rate": 9.940780782033401e-05,
+      "loss": 1.7606,
+      "step": 282
+    },
+    {
+      "epoch": 0.06834098043950737,
+      "grad_norm": 0.2767079472541809,
+      "learning_rate": 9.940185311099648e-05,
+      "loss": 1.7573,
+      "step": 283
+    },
+    {
+      "epoch": 0.06858246800289786,
+      "grad_norm": 0.28070494532585144,
+      "learning_rate": 9.939586879299038e-05,
+      "loss": 1.7798,
+      "step": 284
+    },
+    {
+      "epoch": 0.06882395556628834,
+      "grad_norm": 0.2791160047054291,
+      "learning_rate": 9.938985486990241e-05,
+      "loss": 1.8631,
+      "step": 285
+    },
+    {
+      "epoch": 0.06906544312967883,
+      "grad_norm": 0.27911651134490967,
+      "learning_rate": 9.938381134533695e-05,
+      "loss": 1.7894,
+      "step": 286
+    },
+    {
+      "epoch": 0.06930693069306931,
+      "grad_norm": 0.2858154773712158,
+      "learning_rate": 9.937773822291617e-05,
+      "loss": 1.8625,
+      "step": 287
+    },
+    {
+      "epoch": 0.0695484182564598,
+      "grad_norm": 0.2822221517562866,
+      "learning_rate": 9.93716355062799e-05,
+      "loss": 1.8163,
+      "step": 288
+    },
+    {
+      "epoch": 0.06978990581985028,
+      "grad_norm": 0.28583386540412903,
+      "learning_rate": 9.936550319908583e-05,
+      "loss": 1.7072,
+      "step": 289
+    },
+    {
+      "epoch": 0.07003139338324077,
+      "grad_norm": 0.27519020438194275,
+      "learning_rate": 9.935934130500927e-05,
+      "loss": 1.7768,
+      "step": 290
+    },
+    {
+      "epoch": 0.07027288094663126,
+      "grad_norm": 0.2802446186542511,
+      "learning_rate": 9.935314982774335e-05,
+      "loss": 1.8043,
+      "step": 291
+    },
+    {
+      "epoch": 0.07051436851002173,
+      "grad_norm": 0.2888818085193634,
+      "learning_rate": 9.934692877099885e-05,
+      "loss": 1.8006,
+      "step": 292
+    },
+    {
+      "epoch": 0.07075585607341221,
+      "grad_norm": 0.2915632128715515,
+      "learning_rate": 9.934067813850434e-05,
+      "loss": 1.8006,
+      "step": 293
+    },
+    {
+      "epoch": 0.0709973436368027,
+      "grad_norm": 0.281309574842453,
+      "learning_rate": 9.93343979340061e-05,
+      "loss": 1.8207,
+      "step": 294
+    },
+    {
+      "epoch": 0.07123883120019318,
+      "grad_norm": 0.2658507525920868,
+      "learning_rate": 9.93280881612681e-05,
+      "loss": 1.8466,
+      "step": 295
+    },
+    {
+      "epoch": 0.07148031876358367,
+      "grad_norm": 0.2695015072822571,
+      "learning_rate": 9.932174882407209e-05,
+      "loss": 1.7938,
+      "step": 296
+    },
+    {
+      "epoch": 0.07172180632697416,
+      "grad_norm": 0.28267964720726013,
+      "learning_rate": 9.931537992621749e-05,
+      "loss": 1.859,
+      "step": 297
+    },
+    {
+      "epoch": 0.07196329389036464,
+      "grad_norm": 0.3001064360141754,
+      "learning_rate": 9.930898147152148e-05,
+      "loss": 1.8308,
+      "step": 298
+    },
+    {
+      "epoch": 0.07220478145375513,
+      "grad_norm": 0.2865334153175354,
+      "learning_rate": 9.93025534638189e-05,
+      "loss": 1.8846,
+      "step": 299
+    },
+    {
+      "epoch": 0.07244626901714561,
+      "grad_norm": 0.29224735498428345,
+      "learning_rate": 9.929609590696236e-05,
+      "loss": 1.974,
+      "step": 300
+    },
+    {
+      "epoch": 0.0726877565805361,
+      "grad_norm": 0.29061272740364075,
+      "learning_rate": 9.928960880482214e-05,
+      "loss": 1.891,
+      "step": 301
+    },
+    {
+      "epoch": 0.07292924414392658,
+      "grad_norm": 0.27877870202064514,
+      "learning_rate": 9.928309216128623e-05,
+      "loss": 1.6629,
+      "step": 302
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 0.27829432487487793,
+      "learning_rate": 9.927654598026035e-05,
+      "loss": 1.9232,
+      "step": 303
+    },
+    {
+      "epoch": 0.07341221927070755,
+      "grad_norm": 0.28853073716163635,
+      "learning_rate": 9.926997026566793e-05,
+      "loss": 1.6872,
+      "step": 304
+    },
+    {
+      "epoch": 0.07365370683409804,
+      "grad_norm": 0.3017938435077667,
+      "learning_rate": 9.926336502145004e-05,
+      "loss": 1.834,
+      "step": 305
+    },
+    {
+      "epoch": 0.07389519439748853,
+      "grad_norm": 0.3079606294631958,
+      "learning_rate": 9.92567302515655e-05,
+      "loss": 1.9919,
+      "step": 306
+    },
+    {
+      "epoch": 0.07413668196087901,
+      "grad_norm": 0.284600168466568,
+      "learning_rate": 9.925006595999083e-05,
+      "loss": 1.8107,
+      "step": 307
+    },
+    {
+      "epoch": 0.0743781695242695,
+      "grad_norm": 0.3207036852836609,
+      "learning_rate": 9.924337215072023e-05,
+      "loss": 1.8985,
+      "step": 308
+    },
+    {
+      "epoch": 0.07461965708765998,
+      "grad_norm": 0.27426770329475403,
+      "learning_rate": 9.923664882776557e-05,
+      "loss": 1.7316,
+      "step": 309
+    },
+    {
+      "epoch": 0.07486114465105047,
+      "grad_norm": 0.30607980489730835,
+      "learning_rate": 9.922989599515644e-05,
+      "loss": 1.9892,
+      "step": 310
+    },
+    {
+      "epoch": 0.07510263221444095,
+      "grad_norm": 0.2927166819572449,
+      "learning_rate": 9.922311365694011e-05,
+      "loss": 1.8746,
+      "step": 311
+    },
+    {
+      "epoch": 0.07534411977783144,
+      "grad_norm": 0.2725332975387573,
+      "learning_rate": 9.921630181718152e-05,
+      "loss": 1.7773,
+      "step": 312
+    },
+    {
+      "epoch": 0.07558560734122192,
+      "grad_norm": 0.2871837317943573,
+      "learning_rate": 9.92094604799633e-05,
+      "loss": 1.9641,
+      "step": 313
+    },
+    {
+      "epoch": 0.07582709490461241,
+      "grad_norm": 0.284435510635376,
+      "learning_rate": 9.920258964938576e-05,
+      "loss": 1.8081,
+      "step": 314
+    },
+    {
+      "epoch": 0.0760685824680029,
+      "grad_norm": 0.27971702814102173,
+      "learning_rate": 9.91956893295669e-05,
+      "loss": 1.8659,
+      "step": 315
+    },
+    {
+      "epoch": 0.07631007003139338,
+      "grad_norm": 0.29805755615234375,
+      "learning_rate": 9.918875952464234e-05,
+      "loss": 2.0772,
+      "step": 316
+    },
+    {
+      "epoch": 0.07655155759478387,
+      "grad_norm": 0.2688223421573639,
+      "learning_rate": 9.918180023876548e-05,
+      "loss": 1.7697,
+      "step": 317
+    },
+    {
+      "epoch": 0.07679304515817435,
+      "grad_norm": 0.2922739088535309,
+      "learning_rate": 9.917481147610725e-05,
+      "loss": 1.8815,
+      "step": 318
+    },
+    {
+      "epoch": 0.07703453272156484,
+      "grad_norm": 0.2920701801776886,
+      "learning_rate": 9.916779324085636e-05,
+      "loss": 1.6683,
+      "step": 319
+    },
+    {
+      "epoch": 0.07727602028495532,
+      "grad_norm": 0.2858518958091736,
+      "learning_rate": 9.916074553721915e-05,
+      "loss": 1.7205,
+      "step": 320
+    },
+    {
+      "epoch": 0.07751750784834581,
+      "grad_norm": 0.2882474660873413,
+      "learning_rate": 9.915366836941957e-05,
+      "loss": 1.8518,
+      "step": 321
+    },
+    {
+      "epoch": 0.0777589954117363,
+      "grad_norm": 0.2800799310207367,
+      "learning_rate": 9.91465617416993e-05,
+      "loss": 1.7858,
+      "step": 322
+    },
+    {
+      "epoch": 0.07800048297512678,
+      "grad_norm": 0.30860286951065063,
+      "learning_rate": 9.913942565831766e-05,
+      "loss": 2.0491,
+      "step": 323
+    },
+    {
+      "epoch": 0.07824197053851727,
+      "grad_norm": 0.2933506965637207,
+      "learning_rate": 9.91322601235516e-05,
+      "loss": 1.8936,
+      "step": 324
+    },
+    {
+      "epoch": 0.07848345810190775,
+      "grad_norm": 0.2754722237586975,
+      "learning_rate": 9.912506514169573e-05,
+      "loss": 1.7845,
+      "step": 325
+    },
+    {
+      "epoch": 0.07872494566529824,
+      "grad_norm": 0.2775117754936218,
+      "learning_rate": 9.911784071706233e-05,
+      "loss": 1.6883,
+      "step": 326
+    },
+    {
+      "epoch": 0.07896643322868872,
+      "grad_norm": 0.27762269973754883,
+      "learning_rate": 9.91105868539813e-05,
+      "loss": 1.7333,
+      "step": 327
+    },
+    {
+      "epoch": 0.07920792079207921,
+      "grad_norm": 0.27511125802993774,
+      "learning_rate": 9.910330355680018e-05,
+      "loss": 1.7933,
+      "step": 328
+    },
+    {
+      "epoch": 0.0794494083554697,
+      "grad_norm": 0.27266359329223633,
+      "learning_rate": 9.909599082988419e-05,
+      "loss": 1.8126,
+      "step": 329
+    },
+    {
+      "epoch": 0.07969089591886018,
+      "grad_norm": 0.28638163208961487,
+      "learning_rate": 9.908864867761616e-05,
+      "loss": 1.7719,
+      "step": 330
+    },
+    {
+      "epoch": 0.07993238348225067,
+      "grad_norm": 0.287117063999176,
+      "learning_rate": 9.908127710439654e-05,
+      "loss": 1.7776,
+      "step": 331
+    },
+    {
+      "epoch": 0.08017387104564115,
+      "grad_norm": 0.2886422574520111,
+      "learning_rate": 9.907387611464347e-05,
+      "loss": 1.7783,
+      "step": 332
+    },
+    {
+      "epoch": 0.08041535860903164,
+      "grad_norm": 0.28264063596725464,
+      "learning_rate": 9.906644571279265e-05,
+      "loss": 1.7754,
+      "step": 333
+    },
+    {
+      "epoch": 0.08065684617242212,
+      "grad_norm": 0.2666427493095398,
+      "learning_rate": 9.905898590329744e-05,
+      "loss": 1.5789,
+      "step": 334
+    },
+    {
+      "epoch": 0.08089833373581261,
+      "grad_norm": 0.3525300621986389,
+      "learning_rate": 9.905149669062884e-05,
+      "loss": 2.2147,
+      "step": 335
+    },
+    {
+      "epoch": 0.0811398212992031,
+      "grad_norm": 0.28527653217315674,
+      "learning_rate": 9.904397807927546e-05,
+      "loss": 1.6824,
+      "step": 336
+    },
+    {
+      "epoch": 0.08138130886259358,
+      "grad_norm": 0.3258638381958008,
+      "learning_rate": 9.903643007374352e-05,
+      "loss": 1.9276,
+      "step": 337
+    },
+    {
+      "epoch": 0.08162279642598406,
+      "grad_norm": 0.2898681163787842,
+      "learning_rate": 9.902885267855689e-05,
+      "loss": 1.8823,
+      "step": 338
+    },
+    {
+      "epoch": 0.08186428398937455,
+      "grad_norm": 0.28534162044525146,
+      "learning_rate": 9.902124589825701e-05,
+      "loss": 1.7955,
+      "step": 339
+    },
+    {
+      "epoch": 0.08210577155276504,
+      "grad_norm": 0.29654279351234436,
+      "learning_rate": 9.901360973740295e-05,
+      "loss": 1.8217,
+      "step": 340
+    },
+    {
+      "epoch": 0.08234725911615552,
+      "grad_norm": 0.27160125970840454,
+      "learning_rate": 9.900594420057139e-05,
+      "loss": 1.7292,
+      "step": 341
+    },
+    {
+      "epoch": 0.08258874667954601,
+      "grad_norm": 0.28140437602996826,
+      "learning_rate": 9.899824929235663e-05,
+      "loss": 1.883,
+      "step": 342
+    },
+    {
+      "epoch": 0.08283023424293649,
+      "grad_norm": 0.2921160161495209,
+      "learning_rate": 9.899052501737056e-05,
+      "loss": 1.9536,
+      "step": 343
+    },
+    {
+      "epoch": 0.08307172180632698,
+      "grad_norm": 0.2749026417732239,
+      "learning_rate": 9.898277138024268e-05,
+      "loss": 1.7172,
+      "step": 344
+    },
+    {
+      "epoch": 0.08331320936971746,
+      "grad_norm": 0.2894373834133148,
+      "learning_rate": 9.897498838562006e-05,
+      "loss": 1.9665,
+      "step": 345
+    },
+    {
+      "epoch": 0.08355469693310795,
+      "grad_norm": 0.312502920627594,
+      "learning_rate": 9.89671760381674e-05,
+      "loss": 1.8277,
+      "step": 346
+    },
+    {
+      "epoch": 0.08379618449649844,
+      "grad_norm": 0.281454861164093,
+      "learning_rate": 9.895933434256696e-05,
+      "loss": 1.6576,
+      "step": 347
+    },
+    {
+      "epoch": 0.08403767205988892,
+      "grad_norm": 0.280256062746048,
+      "learning_rate": 9.895146330351865e-05,
+      "loss": 1.9035,
+      "step": 348
+    },
+    {
+      "epoch": 0.0842791596232794,
+      "grad_norm": 0.2698526382446289,
+      "learning_rate": 9.894356292573988e-05,
+      "loss": 1.7014,
+      "step": 349
+    },
+    {
+      "epoch": 0.08452064718666989,
+      "grad_norm": 0.2966681718826294,
+      "learning_rate": 9.893563321396568e-05,
+      "loss": 1.9057,
+      "step": 350
+    },
+    {
+      "epoch": 0.08476213475006038,
+      "grad_norm": 0.2751750946044922,
+      "learning_rate": 9.892767417294872e-05,
+      "loss": 1.9109,
+      "step": 351
+    },
+    {
+      "epoch": 0.08500362231345086,
+      "grad_norm": 0.27698180079460144,
+      "learning_rate": 9.891968580745917e-05,
+      "loss": 1.7143,
+      "step": 352
+    },
+    {
+      "epoch": 0.08524510987684135,
+      "grad_norm": 0.30382975935935974,
+      "learning_rate": 9.89116681222848e-05,
+      "loss": 1.8529,
+      "step": 353
+    },
+    {
+      "epoch": 0.08548659744023183,
+      "grad_norm": 0.28126025199890137,
+      "learning_rate": 9.890362112223094e-05,
+      "loss": 1.8243,
+      "step": 354
+    },
+    {
+      "epoch": 0.08572808500362232,
+      "grad_norm": 0.26443248987197876,
+      "learning_rate": 9.889554481212056e-05,
+      "loss": 1.5424,
+      "step": 355
+    },
+    {
+      "epoch": 0.08596957256701279,
+      "grad_norm": 0.302097886800766,
+      "learning_rate": 9.888743919679409e-05,
+      "loss": 1.7993,
+      "step": 356
+    },
+    {
+      "epoch": 0.08621106013040328,
+      "grad_norm": 0.2876918911933899,
+      "learning_rate": 9.887930428110959e-05,
+      "loss": 1.8625,
+      "step": 357
+    },
+    {
+      "epoch": 0.08645254769379376,
+      "grad_norm": 0.2955901324748993,
+      "learning_rate": 9.887114006994267e-05,
+      "loss": 1.7146,
+      "step": 358
+    },
+    {
+      "epoch": 0.08669403525718425,
+      "grad_norm": 0.2705610394477844,
+      "learning_rate": 9.886294656818651e-05,
+      "loss": 1.8299,
+      "step": 359
+    },
+    {
+      "epoch": 0.08693552282057473,
+      "grad_norm": 0.2755456566810608,
+      "learning_rate": 9.88547237807518e-05,
+      "loss": 1.8367,
+      "step": 360
+    },
+    {
+      "epoch": 0.08717701038396522,
+      "grad_norm": 0.29839128255844116,
+      "learning_rate": 9.884647171256683e-05,
+      "loss": 1.7689,
+      "step": 361
+    },
+    {
+      "epoch": 0.0874184979473557,
+      "grad_norm": 0.27627283334732056,
+      "learning_rate": 9.883819036857743e-05,
+      "loss": 1.7107,
+      "step": 362
+    },
+    {
+      "epoch": 0.08765998551074619,
+      "grad_norm": 0.2801618278026581,
+      "learning_rate": 9.882987975374696e-05,
+      "loss": 1.8248,
+      "step": 363
+    },
+    {
+      "epoch": 0.08790147307413668,
+      "grad_norm": 0.2711143493652344,
+      "learning_rate": 9.882153987305632e-05,
+      "loss": 1.6291,
+      "step": 364
+    },
+    {
+      "epoch": 0.08814296063752716,
+      "grad_norm": 0.2989214062690735,
+      "learning_rate": 9.881317073150396e-05,
+      "loss": 1.8169,
+      "step": 365
+    },
+    {
+      "epoch": 0.08838444820091765,
+      "grad_norm": 0.27399736642837524,
+      "learning_rate": 9.880477233410587e-05,
+      "loss": 1.7848,
+      "step": 366
+    },
+    {
+      "epoch": 0.08862593576430813,
+      "grad_norm": 0.29605332016944885,
+      "learning_rate": 9.879634468589561e-05,
+      "loss": 1.8587,
+      "step": 367
+    },
+    {
+      "epoch": 0.08886742332769862,
+      "grad_norm": 0.3036874234676361,
+      "learning_rate": 9.87878877919242e-05,
+      "loss": 1.8376,
+      "step": 368
+    },
+    {
+      "epoch": 0.0891089108910891,
+      "grad_norm": 0.27524060010910034,
+      "learning_rate": 9.877940165726022e-05,
+      "loss": 1.7785,
+      "step": 369
+    },
+    {
+      "epoch": 0.08935039845447959,
+      "grad_norm": 0.29827842116355896,
+      "learning_rate": 9.87708862869898e-05,
+      "loss": 1.8899,
+      "step": 370
+    },
+    {
+      "epoch": 0.08959188601787008,
+      "grad_norm": 0.28207167983055115,
+      "learning_rate": 9.876234168621654e-05,
+      "loss": 1.818,
+      "step": 371
+    },
+    {
+      "epoch": 0.08983337358126056,
+      "grad_norm": 0.2751006782054901,
+      "learning_rate": 9.875376786006161e-05,
+      "loss": 1.8081,
+      "step": 372
+    },
+    {
+      "epoch": 0.09007486114465105,
+      "grad_norm": 0.2890884280204773,
+      "learning_rate": 9.874516481366367e-05,
+      "loss": 1.719,
+      "step": 373
+    },
+    {
+      "epoch": 0.09031634870804153,
+      "grad_norm": 0.29088371992111206,
+      "learning_rate": 9.87365325521789e-05,
+      "loss": 1.7928,
+      "step": 374
+    },
+    {
+      "epoch": 0.09055783627143202,
+      "grad_norm": 0.287952721118927,
+      "learning_rate": 9.872787108078098e-05,
+      "loss": 1.8332,
+      "step": 375
+    },
+    {
+      "epoch": 0.0907993238348225,
+      "grad_norm": 0.2880597710609436,
+      "learning_rate": 9.871918040466113e-05,
+      "loss": 1.8148,
+      "step": 376
+    },
+    {
+      "epoch": 0.09104081139821299,
+      "grad_norm": 0.2922658622264862,
+      "learning_rate": 9.871046052902802e-05,
+      "loss": 1.8887,
+      "step": 377
+    },
+    {
+      "epoch": 0.09128229896160348,
+      "grad_norm": 0.3039756119251251,
+      "learning_rate": 9.870171145910784e-05,
+      "loss": 1.7143,
+      "step": 378
+    },
+    {
+      "epoch": 0.09152378652499396,
+      "grad_norm": 0.27396541833877563,
+      "learning_rate": 9.869293320014433e-05,
+      "loss": 1.7626,
+      "step": 379
+    },
+    {
+      "epoch": 0.09176527408838445,
+      "grad_norm": 0.30368801951408386,
+      "learning_rate": 9.868412575739864e-05,
+      "loss": 1.9307,
+      "step": 380
+    },
+    {
+      "epoch": 0.09200676165177493,
+      "grad_norm": 0.2743094563484192,
+      "learning_rate": 9.867528913614947e-05,
+      "loss": 1.6689,
+      "step": 381
+    },
+    {
+      "epoch": 0.09224824921516542,
+      "grad_norm": 0.2829776406288147,
+      "learning_rate": 9.866642334169298e-05,
+      "loss": 1.7646,
+      "step": 382
+    },
+    {
+      "epoch": 0.0924897367785559,
+      "grad_norm": 0.27822762727737427,
+      "learning_rate": 9.865752837934283e-05,
+      "loss": 1.8207,
+      "step": 383
+    },
+    {
+      "epoch": 0.09273122434194639,
+      "grad_norm": 0.2803753614425659,
+      "learning_rate": 9.864860425443016e-05,
+      "loss": 1.7833,
+      "step": 384
+    },
+    {
+      "epoch": 0.09297271190533687,
+      "grad_norm": 0.28696754574775696,
+      "learning_rate": 9.86396509723036e-05,
+      "loss": 1.8914,
+      "step": 385
+    },
+    {
+      "epoch": 0.09321419946872736,
+      "grad_norm": 0.27866771817207336,
+      "learning_rate": 9.863066853832921e-05,
+      "loss": 1.8337,
+      "step": 386
+    },
+    {
+      "epoch": 0.09345568703211785,
+      "grad_norm": 0.2905522584915161,
+      "learning_rate": 9.862165695789054e-05,
+      "loss": 1.8121,
+      "step": 387
+    },
+    {
+      "epoch": 0.09369717459550833,
+      "grad_norm": 0.2901919186115265,
+      "learning_rate": 9.861261623638867e-05,
+      "loss": 1.6512,
+      "step": 388
+    },
+    {
+      "epoch": 0.09393866215889882,
+      "grad_norm": 0.29584378004074097,
+      "learning_rate": 9.860354637924204e-05,
+      "loss": 1.9358,
+      "step": 389
+    },
+    {
+      "epoch": 0.0941801497222893,
+      "grad_norm": 0.27355754375457764,
+      "learning_rate": 9.859444739188665e-05,
+      "loss": 1.7889,
+      "step": 390
+    },
+    {
+      "epoch": 0.09442163728567979,
+      "grad_norm": 0.28625258803367615,
+      "learning_rate": 9.85853192797759e-05,
+      "loss": 1.7989,
+      "step": 391
+    },
+    {
+      "epoch": 0.09466312484907027,
+      "grad_norm": 0.2943178415298462,
+      "learning_rate": 9.857616204838065e-05,
+      "loss": 1.8327,
+      "step": 392
+    },
+    {
+      "epoch": 0.09490461241246076,
+      "grad_norm": 0.2840021550655365,
+      "learning_rate": 9.856697570318927e-05,
+      "loss": 1.8274,
+      "step": 393
+    },
+    {
+      "epoch": 0.09514609997585124,
+      "grad_norm": 0.28078603744506836,
+      "learning_rate": 9.855776024970749e-05,
+      "loss": 1.7316,
+      "step": 394
+    },
+    {
+      "epoch": 0.09538758753924173,
+      "grad_norm": 0.27110886573791504,
+      "learning_rate": 9.854851569345854e-05,
+      "loss": 1.6548,
+      "step": 395
+    },
+    {
+      "epoch": 0.09562907510263222,
+      "grad_norm": 0.2627575695514679,
+      "learning_rate": 9.853924203998308e-05,
+      "loss": 1.7053,
+      "step": 396
+    },
+    {
+      "epoch": 0.0958705626660227,
+      "grad_norm": 0.28886744379997253,
+      "learning_rate": 9.852993929483922e-05,
+      "loss": 1.6716,
+      "step": 397
+    },
+    {
+      "epoch": 0.09611205022941319,
+      "grad_norm": 0.2770794928073883,
+      "learning_rate": 9.85206074636025e-05,
+      "loss": 1.7156,
+      "step": 398
+    },
+    {
+      "epoch": 0.09635353779280367,
+      "grad_norm": 0.2742690443992615,
+      "learning_rate": 9.85112465518659e-05,
+      "loss": 1.6733,
+      "step": 399
+    },
+    {
+      "epoch": 0.09659502535619416,
+      "grad_norm": 0.28999531269073486,
+      "learning_rate": 9.850185656523978e-05,
+      "loss": 1.8832,
+      "step": 400
+    },
+    {
+      "epoch": 0.09683651291958464,
+      "grad_norm": 0.2888575494289398,
+      "learning_rate": 9.849243750935202e-05,
+      "loss": 1.8296,
+      "step": 401
+    },
+    {
+      "epoch": 0.09707800048297513,
+      "grad_norm": 0.2710040807723999,
+      "learning_rate": 9.848298938984781e-05,
+      "loss": 1.7656,
+      "step": 402
+    },
+    {
+      "epoch": 0.09731948804636562,
+      "grad_norm": 0.2809413969516754,
+      "learning_rate": 9.847351221238987e-05,
+      "loss": 1.7886,
+      "step": 403
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 0.2801872789859772,
+      "learning_rate": 9.846400598265824e-05,
+      "loss": 1.7092,
+      "step": 404
+    },
+    {
+      "epoch": 0.09780246317314659,
+      "grad_norm": 0.2998405396938324,
+      "learning_rate": 9.845447070635046e-05,
+      "loss": 1.8165,
+      "step": 405
+    },
+    {
+      "epoch": 0.09804395073653707,
+      "grad_norm": 0.33238905668258667,
+      "learning_rate": 9.844490638918138e-05,
+      "loss": 2.0468,
+      "step": 406
+    },
+    {
+      "epoch": 0.09828543829992756,
+      "grad_norm": 0.2929340600967407,
+      "learning_rate": 9.843531303688335e-05,
+      "loss": 1.8788,
+      "step": 407
+    },
+    {
+      "epoch": 0.09852692586331804,
+      "grad_norm": 0.2869876027107239,
+      "learning_rate": 9.842569065520606e-05,
+      "loss": 1.8086,
+      "step": 408
+    },
+    {
+      "epoch": 0.09876841342670853,
+      "grad_norm": 0.28498587012290955,
+      "learning_rate": 9.841603924991663e-05,
+      "loss": 1.6979,
+      "step": 409
+    },
+    {
+      "epoch": 0.09900990099009901,
+      "grad_norm": 0.2809644341468811,
+      "learning_rate": 9.840635882679956e-05,
+      "loss": 1.7499,
+      "step": 410
+    },
+    {
+      "epoch": 0.0992513885534895,
+      "grad_norm": 0.2747531533241272,
+      "learning_rate": 9.839664939165675e-05,
+      "loss": 1.7549,
+      "step": 411
+    },
+    {
+      "epoch": 0.09949287611687999,
+      "grad_norm": 0.2902829349040985,
+      "learning_rate": 9.83869109503075e-05,
+      "loss": 1.8995,
+      "step": 412
+    },
+    {
+      "epoch": 0.09973436368027047,
+      "grad_norm": 0.2756800949573517,
+      "learning_rate": 9.837714350858846e-05,
+      "loss": 1.7057,
+      "step": 413
+    },
+    {
+      "epoch": 0.09997585124366096,
+      "grad_norm": 0.289368599653244,
+      "learning_rate": 9.836734707235366e-05,
+      "loss": 1.933,
+      "step": 414
+    },
+    {
+      "epoch": 0.10021733880705144,
+      "grad_norm": 0.29384687542915344,
+      "learning_rate": 9.835752164747457e-05,
+      "loss": 1.9249,
+      "step": 415
+    },
+    {
+      "epoch": 0.10045882637044193,
+      "grad_norm": 0.2748492360115051,
+      "learning_rate": 9.834766723983997e-05,
+      "loss": 1.8197,
+      "step": 416
+    },
+    {
+      "epoch": 0.10070031393383241,
+      "grad_norm": 0.29578012228012085,
+      "learning_rate": 9.833778385535603e-05,
+      "loss": 1.9807,
+      "step": 417
+    },
+    {
+      "epoch": 0.1009418014972229,
+      "grad_norm": 0.2878894507884979,
+      "learning_rate": 9.832787149994631e-05,
+      "loss": 1.6623,
+      "step": 418
+    },
+    {
+      "epoch": 0.10118328906061339,
+      "grad_norm": 0.28690582513809204,
+      "learning_rate": 9.831793017955168e-05,
+      "loss": 1.8893,
+      "step": 419
+    },
+    {
+      "epoch": 0.10142477662400387,
+      "grad_norm": 0.27932021021842957,
+      "learning_rate": 9.830795990013046e-05,
+      "loss": 1.7858,
+      "step": 420
+    },
+    {
+      "epoch": 0.10166626418739434,
+      "grad_norm": 0.27387359738349915,
+      "learning_rate": 9.82979606676582e-05,
+      "loss": 1.8473,
+      "step": 421
+    },
+    {
+      "epoch": 0.10190775175078483,
+      "grad_norm": 0.26410290598869324,
+      "learning_rate": 9.828793248812792e-05,
+      "loss": 1.6531,
+      "step": 422
+    },
+    {
+      "epoch": 0.10214923931417531,
+      "grad_norm": 0.2906973361968994,
+      "learning_rate": 9.827787536754994e-05,
+      "loss": 1.8561,
+      "step": 423
+    },
+    {
+      "epoch": 0.1023907268775658,
+      "grad_norm": 0.28027960658073425,
+      "learning_rate": 9.82677893119519e-05,
+      "loss": 1.7624,
+      "step": 424
+    },
+    {
+      "epoch": 0.10263221444095628,
+      "grad_norm": 0.26871955394744873,
+      "learning_rate": 9.825767432737883e-05,
+      "loss": 1.6901,
+      "step": 425
+    },
+    {
+      "epoch": 0.10287370200434677,
+      "grad_norm": 0.28086093068122864,
+      "learning_rate": 9.824753041989308e-05,
+      "loss": 1.8209,
+      "step": 426
+    },
+    {
+      "epoch": 0.10311518956773726,
+      "grad_norm": 0.2760232388973236,
+      "learning_rate": 9.823735759557431e-05,
+      "loss": 1.7899,
+      "step": 427
+    },
+    {
+      "epoch": 0.10335667713112774,
+      "grad_norm": 0.27731311321258545,
+      "learning_rate": 9.822715586051957e-05,
+      "loss": 1.7834,
+      "step": 428
+    },
+    {
+      "epoch": 0.10359816469451823,
+      "grad_norm": 0.26586011052131653,
+      "learning_rate": 9.821692522084316e-05,
+      "loss": 1.7721,
+      "step": 429
+    },
+    {
+      "epoch": 0.10383965225790871,
+      "grad_norm": 0.2568017542362213,
+      "learning_rate": 9.820666568267678e-05,
+      "loss": 1.5376,
+      "step": 430
+    },
+    {
+      "epoch": 0.1040811398212992,
+      "grad_norm": 0.29189375042915344,
+      "learning_rate": 9.819637725216941e-05,
+      "loss": 1.8376,
+      "step": 431
+    },
+    {
+      "epoch": 0.10432262738468968,
+      "grad_norm": 0.28995007276535034,
+      "learning_rate": 9.818605993548733e-05,
+      "loss": 1.8012,
+      "step": 432
+    },
+    {
+      "epoch": 0.10456411494808017,
+      "grad_norm": 0.2857837378978729,
+      "learning_rate": 9.817571373881417e-05,
+      "loss": 1.9442,
+      "step": 433
+    },
+    {
+      "epoch": 0.10480560251147066,
+      "grad_norm": 0.28825604915618896,
+      "learning_rate": 9.816533866835087e-05,
+      "loss": 1.8442,
+      "step": 434
+    },
+    {
+      "epoch": 0.10504709007486114,
+      "grad_norm": 0.28988251090049744,
+      "learning_rate": 9.815493473031563e-05,
+      "loss": 1.7784,
+      "step": 435
+    },
+    {
+      "epoch": 0.10528857763825163,
+      "grad_norm": 0.2709440290927887,
+      "learning_rate": 9.814450193094398e-05,
+      "loss": 1.7199,
+      "step": 436
+    },
+    {
+      "epoch": 0.10553006520164211,
+      "grad_norm": 0.2710571587085724,
+      "learning_rate": 9.813404027648877e-05,
+      "loss": 1.6485,
+      "step": 437
+    },
+    {
+      "epoch": 0.1057715527650326,
+      "grad_norm": 0.29289767146110535,
+      "learning_rate": 9.812354977322011e-05,
+      "loss": 1.6867,
+      "step": 438
+    },
+    {
+      "epoch": 0.10601304032842308,
+      "grad_norm": 0.27464166283607483,
+      "learning_rate": 9.811303042742542e-05,
+      "loss": 1.54,
+      "step": 439
+    },
+    {
+      "epoch": 0.10625452789181357,
+      "grad_norm": 0.27407106757164,
+      "learning_rate": 9.81024822454094e-05,
+      "loss": 1.6428,
+      "step": 440
+    },
+    {
+      "epoch": 0.10649601545520405,
+      "grad_norm": 0.29014259576797485,
+      "learning_rate": 9.809190523349399e-05,
+      "loss": 1.7654,
+      "step": 441
+    },
+    {
+      "epoch": 0.10673750301859454,
+      "grad_norm": 0.2837926149368286,
+      "learning_rate": 9.808129939801851e-05,
+      "loss": 1.8607,
+      "step": 442
+    },
+    {
+      "epoch": 0.10697899058198503,
+      "grad_norm": 0.3138989508152008,
+      "learning_rate": 9.807066474533946e-05,
+      "loss": 1.9194,
+      "step": 443
+    },
+    {
+      "epoch": 0.10722047814537551,
+      "grad_norm": 0.2894023060798645,
+      "learning_rate": 9.806000128183069e-05,
+      "loss": 1.8462,
+      "step": 444
+    },
+    {
+      "epoch": 0.107461965708766,
+      "grad_norm": 0.27192845940589905,
+      "learning_rate": 9.80493090138832e-05,
+      "loss": 1.8293,
+      "step": 445
+    },
+    {
+      "epoch": 0.10770345327215648,
+      "grad_norm": 0.2810100018978119,
+      "learning_rate": 9.803858794790541e-05,
+      "loss": 1.8501,
+      "step": 446
+    },
+    {
+      "epoch": 0.10794494083554697,
+      "grad_norm": 0.2710225284099579,
+      "learning_rate": 9.802783809032288e-05,
+      "loss": 1.7248,
+      "step": 447
+    },
+    {
+      "epoch": 0.10818642839893745,
+      "grad_norm": 0.28958362340927124,
+      "learning_rate": 9.801705944757846e-05,
+      "loss": 1.7555,
+      "step": 448
+    },
+    {
+      "epoch": 0.10842791596232794,
+      "grad_norm": 0.27755028009414673,
+      "learning_rate": 9.800625202613225e-05,
+      "loss": 1.8206,
+      "step": 449
+    },
+    {
+      "epoch": 0.10866940352571842,
+      "grad_norm": 0.2801584005355835,
+      "learning_rate": 9.799541583246163e-05,
+      "loss": 1.764,
+      "step": 450
+    },
+    {
+      "epoch": 0.10891089108910891,
+      "grad_norm": 0.2989168167114258,
+      "learning_rate": 9.79845508730612e-05,
+      "loss": 1.8531,
+      "step": 451
+    },
+    {
+      "epoch": 0.1091523786524994,
+      "grad_norm": 0.28735461831092834,
+      "learning_rate": 9.797365715444278e-05,
+      "loss": 1.9467,
+      "step": 452
+    },
+    {
+      "epoch": 0.10939386621588988,
+      "grad_norm": 0.2826710641384125,
+      "learning_rate": 9.796273468313546e-05,
+      "loss": 1.6993,
+      "step": 453
+    },
+    {
+      "epoch": 0.10963535377928037,
+      "grad_norm": 0.3037896156311035,
+      "learning_rate": 9.795178346568555e-05,
+      "loss": 1.7925,
+      "step": 454
+    },
+    {
+      "epoch": 0.10987684134267085,
+      "grad_norm": 0.27401891350746155,
+      "learning_rate": 9.794080350865658e-05,
+      "loss": 1.8314,
+      "step": 455
+    },
+    {
+      "epoch": 0.11011832890606134,
+      "grad_norm": 0.2809910774230957,
+      "learning_rate": 9.792979481862933e-05,
+      "loss": 1.8314,
+      "step": 456
+    },
+    {
+      "epoch": 0.11035981646945182,
+      "grad_norm": 0.27638763189315796,
+      "learning_rate": 9.791875740220175e-05,
+      "loss": 1.702,
+      "step": 457
+    },
+    {
+      "epoch": 0.11060130403284231,
+      "grad_norm": 0.2893375754356384,
+      "learning_rate": 9.79076912659891e-05,
+      "loss": 1.8864,
+      "step": 458
+    },
+    {
+      "epoch": 0.1108427915962328,
+      "grad_norm": 0.2902871072292328,
+      "learning_rate": 9.789659641662373e-05,
+      "loss": 1.8047,
+      "step": 459
+    },
+    {
+      "epoch": 0.11108427915962328,
+      "grad_norm": 0.2804037928581238,
+      "learning_rate": 9.788547286075531e-05,
+      "loss": 1.7456,
+      "step": 460
+    },
+    {
+      "epoch": 0.11132576672301377,
+      "grad_norm": 0.2759838402271271,
+      "learning_rate": 9.787432060505064e-05,
+      "loss": 1.7112,
+      "step": 461
+    },
+    {
+      "epoch": 0.11156725428640425,
+      "grad_norm": 0.28496286273002625,
+      "learning_rate": 9.786313965619376e-05,
+      "loss": 1.7838,
+      "step": 462
+    },
+    {
+      "epoch": 0.11180874184979474,
+      "grad_norm": 0.28257259726524353,
+      "learning_rate": 9.78519300208859e-05,
+      "loss": 1.8171,
+      "step": 463
+    },
+    {
+      "epoch": 0.11205022941318522,
+      "grad_norm": 0.2747897505760193,
+      "learning_rate": 9.784069170584547e-05,
+      "loss": 1.7337,
+      "step": 464
+    },
+    {
+      "epoch": 0.11229171697657571,
+      "grad_norm": 0.29796814918518066,
+      "learning_rate": 9.78294247178081e-05,
+      "loss": 1.977,
+      "step": 465
+    },
+    {
+      "epoch": 0.1125332045399662,
+      "grad_norm": 0.2776499390602112,
+      "learning_rate": 9.781812906352657e-05,
+      "loss": 1.78,
+      "step": 466
+    },
+    {
+      "epoch": 0.11277469210335668,
+      "grad_norm": 0.3614672124385834,
+      "learning_rate": 9.780680474977083e-05,
+      "loss": 1.7773,
+      "step": 467
+    },
+    {
+      "epoch": 0.11301617966674717,
+      "grad_norm": 0.28927668929100037,
+      "learning_rate": 9.779545178332807e-05,
+      "loss": 1.8433,
+      "step": 468
+    },
+    {
+      "epoch": 0.11325766723013765,
+      "grad_norm": 0.2913784086704254,
+      "learning_rate": 9.778407017100259e-05,
+      "loss": 1.9236,
+      "step": 469
+    },
+    {
+      "epoch": 0.11349915479352814,
+      "grad_norm": 0.30389007925987244,
+      "learning_rate": 9.77726599196159e-05,
+      "loss": 1.7944,
+      "step": 470
+    },
+    {
+      "epoch": 0.11374064235691862,
+      "grad_norm": 0.2783800959587097,
+      "learning_rate": 9.776122103600665e-05,
+      "loss": 1.7486,
+      "step": 471
+    },
+    {
+      "epoch": 0.11398212992030911,
+      "grad_norm": 0.2779453992843628,
+      "learning_rate": 9.774975352703065e-05,
+      "loss": 1.6498,
+      "step": 472
+    },
+    {
+      "epoch": 0.1142236174836996,
+      "grad_norm": 0.2728698253631592,
+      "learning_rate": 9.773825739956089e-05,
+      "loss": 1.7032,
+      "step": 473
+    },
+    {
+      "epoch": 0.11446510504709008,
+      "grad_norm": 0.4645857512950897,
+      "learning_rate": 9.772673266048747e-05,
+      "loss": 1.9575,
+      "step": 474
+    },
+    {
+      "epoch": 0.11470659261048056,
+      "grad_norm": 0.2809775471687317,
+      "learning_rate": 9.771517931671769e-05,
+      "loss": 1.7906,
+      "step": 475
+    },
+    {
+      "epoch": 0.11494808017387105,
+      "grad_norm": 0.2997235953807831,
+      "learning_rate": 9.770359737517599e-05,
+      "loss": 1.8539,
+      "step": 476
+    },
+    {
+      "epoch": 0.11518956773726154,
+      "grad_norm": 0.2767902910709381,
+      "learning_rate": 9.769198684280387e-05,
+      "loss": 1.7607,
+      "step": 477
+    },
+    {
+      "epoch": 0.11543105530065202,
+      "grad_norm": 0.312592089176178,
+      "learning_rate": 9.768034772656006e-05,
+      "loss": 1.9156,
+      "step": 478
+    },
+    {
+      "epoch": 0.11567254286404251,
+      "grad_norm": 0.2808191180229187,
+      "learning_rate": 9.766868003342038e-05,
+      "loss": 1.7381,
+      "step": 479
+    },
+    {
+      "epoch": 0.11591403042743299,
+      "grad_norm": 0.28857922554016113,
+      "learning_rate": 9.765698377037778e-05,
+      "loss": 1.8443,
+      "step": 480
+    },
+    {
+      "epoch": 0.11615551799082348,
+      "grad_norm": 0.28472453355789185,
+      "learning_rate": 9.764525894444235e-05,
+      "loss": 1.816,
+      "step": 481
+    },
+    {
+      "epoch": 0.11639700555421396,
+      "grad_norm": 0.2768532335758209,
+      "learning_rate": 9.763350556264128e-05,
+      "loss": 1.7091,
+      "step": 482
+    },
+    {
+      "epoch": 0.11663849311760445,
+      "grad_norm": 0.3824092447757721,
+      "learning_rate": 9.762172363201887e-05,
+      "loss": 1.6476,
+      "step": 483
+    },
+    {
+      "epoch": 0.11687998068099494,
+      "grad_norm": 0.28541287779808044,
+      "learning_rate": 9.760991315963655e-05,
+      "loss": 1.7566,
+      "step": 484
+    },
+    {
+      "epoch": 0.11712146824438542,
+      "grad_norm": 0.2787253260612488,
+      "learning_rate": 9.759807415257285e-05,
+      "loss": 1.8362,
+      "step": 485
+    },
+    {
+      "epoch": 0.11736295580777589,
+      "grad_norm": 0.2788980305194855,
+      "learning_rate": 9.75862066179234e-05,
+      "loss": 1.7361,
+      "step": 486
+    },
+    {
+      "epoch": 0.11760444337116638,
+      "grad_norm": 0.30717000365257263,
+      "learning_rate": 9.75743105628009e-05,
+      "loss": 2.0405,
+      "step": 487
+    },
+    {
+      "epoch": 0.11784593093455686,
+      "grad_norm": 0.28835999965667725,
+      "learning_rate": 9.756238599433521e-05,
+      "loss": 1.8115,
+      "step": 488
+    },
+    {
+      "epoch": 0.11808741849794735,
+      "grad_norm": 0.26419007778167725,
+      "learning_rate": 9.755043291967321e-05,
+      "loss": 1.5545,
+      "step": 489
+    },
+    {
+      "epoch": 0.11832890606133784,
+      "grad_norm": 0.2989584505558014,
+      "learning_rate": 9.753845134597892e-05,
+      "loss": 1.9747,
+      "step": 490
+    },
+    {
+      "epoch": 0.11857039362472832,
+      "grad_norm": 0.27482473850250244,
+      "learning_rate": 9.752644128043338e-05,
+      "loss": 1.7562,
+      "step": 491
+    },
+    {
+      "epoch": 0.1188118811881188,
+      "grad_norm": 0.26998311281204224,
+      "learning_rate": 9.751440273023478e-05,
+      "loss": 1.6722,
+      "step": 492
+    },
+    {
+      "epoch": 0.11905336875150929,
+      "grad_norm": 0.2935127019882202,
+      "learning_rate": 9.750233570259833e-05,
+      "loss": 1.8002,
+      "step": 493
+    },
+    {
+      "epoch": 0.11929485631489978,
+      "grad_norm": 0.2886837422847748,
+      "learning_rate": 9.749024020475632e-05,
+      "loss": 1.8251,
+      "step": 494
+    },
+    {
+      "epoch": 0.11953634387829026,
+      "grad_norm": 0.27557656168937683,
+      "learning_rate": 9.747811624395811e-05,
+      "loss": 1.6533,
+      "step": 495
+    },
+    {
+      "epoch": 0.11977783144168075,
+      "grad_norm": 0.2803681194782257,
+      "learning_rate": 9.746596382747011e-05,
+      "loss": 1.7325,
+      "step": 496
+    },
+    {
+      "epoch": 0.12001931900507123,
+      "grad_norm": 0.3031095862388611,
+      "learning_rate": 9.745378296257579e-05,
+      "loss": 1.8844,
+      "step": 497
+    },
+    {
+      "epoch": 0.12026080656846172,
+      "grad_norm": 0.28464269638061523,
+      "learning_rate": 9.744157365657567e-05,
+      "loss": 1.8316,
+      "step": 498
+    },
+    {
+      "epoch": 0.1205022941318522,
+      "grad_norm": 0.29071709513664246,
+      "learning_rate": 9.742933591678731e-05,
+      "loss": 1.7188,
+      "step": 499
+    },
+    {
+      "epoch": 0.12074378169524269,
+      "grad_norm": 0.29018208384513855,
+      "learning_rate": 9.741706975054532e-05,
+      "loss": 1.6691,
+      "step": 500
+    },
+    {
+      "epoch": 0.12098526925863318,
+      "grad_norm": 0.3093045949935913,
+      "learning_rate": 9.740477516520135e-05,
+      "loss": 1.9989,
+      "step": 501
+    },
+    {
+      "epoch": 0.12122675682202366,
+      "grad_norm": 0.3274978697299957,
+      "learning_rate": 9.739245216812407e-05,
+      "loss": 1.9223,
+      "step": 502
+    },
+    {
+      "epoch": 0.12146824438541415,
+      "grad_norm": 0.2763381004333496,
+      "learning_rate": 9.738010076669919e-05,
+      "loss": 1.8004,
+      "step": 503
+    },
+    {
+      "epoch": 0.12170973194880463,
+      "grad_norm": 0.27695074677467346,
+      "learning_rate": 9.736772096832945e-05,
+      "loss": 1.8704,
+      "step": 504
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 0.28391584753990173,
+      "learning_rate": 9.735531278043457e-05,
+      "loss": 1.8722,
+      "step": 505
+    },
+    {
+      "epoch": 0.1221927070755856,
+      "grad_norm": 0.2790353000164032,
+      "learning_rate": 9.734287621045133e-05,
+      "loss": 1.7409,
+      "step": 506
+    },
+    {
+      "epoch": 0.12243419463897609,
+      "grad_norm": 0.31203344464302063,
+      "learning_rate": 9.733041126583351e-05,
+      "loss": 1.8611,
+      "step": 507
+    },
+    {
+      "epoch": 0.12267568220236658,
+      "grad_norm": 0.28117474913597107,
+      "learning_rate": 9.73179179540519e-05,
+      "loss": 1.7635,
+      "step": 508
+    },
+    {
+      "epoch": 0.12291716976575706,
+      "grad_norm": 0.2965591251850128,
+      "learning_rate": 9.730539628259425e-05,
+      "loss": 1.7152,
+      "step": 509
+    },
+    {
+      "epoch": 0.12315865732914755,
+      "grad_norm": 0.2742213010787964,
+      "learning_rate": 9.729284625896536e-05,
+      "loss": 1.7753,
+      "step": 510
+    },
+    {
+      "epoch": 0.12340014489253803,
+      "grad_norm": 0.2835814952850342,
+      "learning_rate": 9.7280267890687e-05,
+      "loss": 1.8058,
+      "step": 511
+    },
+    {
+      "epoch": 0.12364163245592852,
+      "grad_norm": 0.27745455503463745,
+      "learning_rate": 9.726766118529795e-05,
+      "loss": 1.614,
+      "step": 512
+    },
+    {
+      "epoch": 0.123883120019319,
+      "grad_norm": 0.2921353876590729,
+      "learning_rate": 9.725502615035393e-05,
+      "loss": 2.0323,
+      "step": 513
+    },
+    {
+      "epoch": 0.12412460758270949,
+      "grad_norm": 0.278221994638443,
+      "learning_rate": 9.724236279342764e-05,
+      "loss": 1.6944,
+      "step": 514
+    },
+    {
+      "epoch": 0.12436609514609998,
+      "grad_norm": 0.2910671532154083,
+      "learning_rate": 9.722967112210883e-05,
+      "loss": 1.8207,
+      "step": 515
+    },
+    {
+      "epoch": 0.12460758270949046,
+      "grad_norm": 0.28481143712997437,
+      "learning_rate": 9.721695114400416e-05,
+      "loss": 1.5918,
+      "step": 516
+    },
+    {
+      "epoch": 0.12484907027288095,
+      "grad_norm": 0.2787121832370758,
+      "learning_rate": 9.720420286673722e-05,
+      "loss": 1.6884,
+      "step": 517
+    },
+    {
+      "epoch": 0.12509055783627143,
+      "grad_norm": 0.2825336158275604,
+      "learning_rate": 9.719142629794864e-05,
+      "loss": 1.7568,
+      "step": 518
+    },
+    {
+      "epoch": 0.12533204539966192,
+      "grad_norm": 0.30035796761512756,
+      "learning_rate": 9.717862144529596e-05,
+      "loss": 1.7803,
+      "step": 519
+    },
+    {
+      "epoch": 0.1255735329630524,
+      "grad_norm": 0.2755514085292816,
+      "learning_rate": 9.716578831645368e-05,
+      "loss": 1.8127,
+      "step": 520
+    },
+    {
+      "epoch": 0.1258150205264429,
+      "grad_norm": 0.27801671624183655,
+      "learning_rate": 9.715292691911326e-05,
+      "loss": 1.8203,
+      "step": 521
+    },
+    {
+      "epoch": 0.12605650808983337,
+      "grad_norm": 0.27820059657096863,
+      "learning_rate": 9.71400372609831e-05,
+      "loss": 1.8601,
+      "step": 522
+    },
+    {
+      "epoch": 0.12629799565322386,
+      "grad_norm": 0.3146127760410309,
+      "learning_rate": 9.712711934978851e-05,
+      "loss": 1.896,
+      "step": 523
+    },
+    {
+      "epoch": 0.12653948321661435,
+      "grad_norm": 0.2702305018901825,
+      "learning_rate": 9.711417319327174e-05,
+      "loss": 1.7463,
+      "step": 524
+    },
+    {
+      "epoch": 0.12678097078000483,
+      "grad_norm": 0.27754876017570496,
+      "learning_rate": 9.710119879919201e-05,
+      "loss": 1.7289,
+      "step": 525
+    },
+    {
+      "epoch": 0.12702245834339532,
+      "grad_norm": 0.27290308475494385,
+      "learning_rate": 9.708819617532546e-05,
+      "loss": 1.715,
+      "step": 526
+    },
+    {
+      "epoch": 0.1272639459067858,
+      "grad_norm": 0.2878723740577698,
+      "learning_rate": 9.707516532946506e-05,
+      "loss": 1.782,
+      "step": 527
+    },
+    {
+      "epoch": 0.1275054334701763,
+      "grad_norm": 0.29204249382019043,
+      "learning_rate": 9.706210626942078e-05,
+      "loss": 1.7625,
+      "step": 528
+    },
+    {
+      "epoch": 0.12774692103356677,
+      "grad_norm": 0.28486737608909607,
+      "learning_rate": 9.704901900301952e-05,
+      "loss": 1.8964,
+      "step": 529
+    },
+    {
+      "epoch": 0.12798840859695726,
+      "grad_norm": 0.28245386481285095,
+      "learning_rate": 9.703590353810501e-05,
+      "loss": 1.7631,
+      "step": 530
+    },
+    {
+      "epoch": 0.12822989616034774,
+      "grad_norm": 0.28806790709495544,
+      "learning_rate": 9.702275988253791e-05,
+      "loss": 1.7705,
+      "step": 531
+    },
+    {
+      "epoch": 0.12847138372373823,
+      "grad_norm": 0.2874543070793152,
+      "learning_rate": 9.700958804419581e-05,
+      "loss": 1.6925,
+      "step": 532
+    },
+    {
+      "epoch": 0.12871287128712872,
+      "grad_norm": 0.2873690128326416,
+      "learning_rate": 9.699638803097313e-05,
+      "loss": 1.7817,
+      "step": 533
+    },
+    {
+      "epoch": 0.1289543588505192,
+      "grad_norm": 0.2811433970928192,
+      "learning_rate": 9.698315985078124e-05,
+      "loss": 1.6764,
+      "step": 534
+    },
+    {
+      "epoch": 0.1291958464139097,
+      "grad_norm": 0.2692316174507141,
+      "learning_rate": 9.696990351154836e-05,
+      "loss": 1.654,
+      "step": 535
+    },
+    {
+      "epoch": 0.12943733397730017,
+      "grad_norm": 0.27523520588874817,
+      "learning_rate": 9.695661902121955e-05,
+      "loss": 1.7807,
+      "step": 536
+    },
+    {
+      "epoch": 0.12967882154069066,
+      "grad_norm": 0.27212703227996826,
+      "learning_rate": 9.694330638775681e-05,
+      "loss": 1.588,
+      "step": 537
+    },
+    {
+      "epoch": 0.12992030910408114,
+      "grad_norm": 0.2714000940322876,
+      "learning_rate": 9.692996561913899e-05,
+      "loss": 1.7409,
+      "step": 538
+    },
+    {
+      "epoch": 0.13016179666747163,
+      "grad_norm": 0.2825371026992798,
+      "learning_rate": 9.691659672336175e-05,
+      "loss": 1.6836,
+      "step": 539
+    },
+    {
+      "epoch": 0.13040328423086212,
+      "grad_norm": 0.27138465642929077,
+      "learning_rate": 9.690319970843768e-05,
+      "loss": 1.6183,
+      "step": 540
+    },
+    {
+      "epoch": 0.1306447717942526,
+      "grad_norm": 0.2688143253326416,
+      "learning_rate": 9.688977458239619e-05,
+      "loss": 1.7082,
+      "step": 541
+    },
+    {
+      "epoch": 0.1308862593576431,
+      "grad_norm": 0.27580389380455017,
+      "learning_rate": 9.687632135328353e-05,
+      "loss": 1.804,
+      "step": 542
+    },
+    {
+      "epoch": 0.13112774692103357,
+      "grad_norm": 0.2860567569732666,
+      "learning_rate": 9.686284002916279e-05,
+      "loss": 1.7159,
+      "step": 543
+    },
+    {
+      "epoch": 0.13136923448442406,
+      "grad_norm": 0.28194060921669006,
+      "learning_rate": 9.684933061811393e-05,
+      "loss": 1.8439,
+      "step": 544
+    },
+    {
+      "epoch": 0.13161072204781454,
+      "grad_norm": 0.28102830052375793,
+      "learning_rate": 9.683579312823372e-05,
+      "loss": 1.8587,
+      "step": 545
+    },
+    {
+      "epoch": 0.13185220961120503,
+      "grad_norm": 0.28536900877952576,
+      "learning_rate": 9.682222756763576e-05,
+      "loss": 1.5734,
+      "step": 546
+    },
+    {
+      "epoch": 0.13209369717459551,
+      "grad_norm": 0.28035515546798706,
+      "learning_rate": 9.680863394445048e-05,
+      "loss": 1.6984,
+      "step": 547
+    },
+    {
+      "epoch": 0.132335184737986,
+      "grad_norm": 0.2830459773540497,
+      "learning_rate": 9.679501226682513e-05,
+      "loss": 1.7716,
+      "step": 548
+    },
+    {
+      "epoch": 0.13257667230137649,
+      "grad_norm": 0.312172532081604,
+      "learning_rate": 9.678136254292376e-05,
+      "loss": 1.9346,
+      "step": 549
+    },
+    {
+      "epoch": 0.13281815986476697,
+      "grad_norm": 0.2721990644931793,
+      "learning_rate": 9.676768478092724e-05,
+      "loss": 1.6447,
+      "step": 550
+    },
+    {
+      "epoch": 0.13305964742815746,
+      "grad_norm": 0.31954988837242126,
+      "learning_rate": 9.675397898903327e-05,
+      "loss": 2.0035,
+      "step": 551
+    },
+    {
+      "epoch": 0.13330113499154794,
+      "grad_norm": 0.27436164021492004,
+      "learning_rate": 9.674024517545628e-05,
+      "loss": 1.7928,
+      "step": 552
+    },
+    {
+      "epoch": 0.13354262255493843,
+      "grad_norm": 0.2752690017223358,
+      "learning_rate": 9.67264833484276e-05,
+      "loss": 1.7733,
+      "step": 553
+    },
+    {
+      "epoch": 0.1337841101183289,
+      "grad_norm": 0.28749632835388184,
+      "learning_rate": 9.671269351619523e-05,
+      "loss": 1.8311,
+      "step": 554
+    },
+    {
+      "epoch": 0.1340255976817194,
+      "grad_norm": 0.28053197264671326,
+      "learning_rate": 9.669887568702405e-05,
+      "loss": 1.778,
+      "step": 555
+    },
+    {
+      "epoch": 0.13426708524510989,
+      "grad_norm": 0.2810681462287903,
+      "learning_rate": 9.668502986919566e-05,
+      "loss": 1.7463,
+      "step": 556
+    },
+    {
+      "epoch": 0.13450857280850037,
+      "grad_norm": 0.2764918804168701,
+      "learning_rate": 9.667115607100846e-05,
+      "loss": 1.7359,
+      "step": 557
+    },
+    {
+      "epoch": 0.13475006037189086,
+      "grad_norm": 0.28768905997276306,
+      "learning_rate": 9.665725430077764e-05,
+      "loss": 1.6103,
+      "step": 558
+    },
+    {
+      "epoch": 0.13499154793528134,
+      "grad_norm": 0.29200369119644165,
+      "learning_rate": 9.664332456683511e-05,
+      "loss": 1.6831,
+      "step": 559
+    },
+    {
+      "epoch": 0.13523303549867183,
+      "grad_norm": 0.31363463401794434,
+      "learning_rate": 9.662936687752958e-05,
+      "loss": 2.0143,
+      "step": 560
+    },
+    {
+      "epoch": 0.1354745230620623,
+      "grad_norm": 0.2834382653236389,
+      "learning_rate": 9.661538124122648e-05,
+      "loss": 1.7506,
+      "step": 561
+    },
+    {
+      "epoch": 0.1357160106254528,
+      "grad_norm": 0.29083454608917236,
+      "learning_rate": 9.660136766630801e-05,
+      "loss": 1.8281,
+      "step": 562
+    },
+    {
+      "epoch": 0.13595749818884328,
+      "grad_norm": 0.27659663558006287,
+      "learning_rate": 9.658732616117312e-05,
+      "loss": 1.6351,
+      "step": 563
+    },
+    {
+      "epoch": 0.13619898575223377,
+      "grad_norm": 0.28885459899902344,
+      "learning_rate": 9.657325673423749e-05,
+      "loss": 1.8258,
+      "step": 564
+    },
+    {
+      "epoch": 0.13644047331562426,
+      "grad_norm": 0.2921139895915985,
+      "learning_rate": 9.655915939393354e-05,
+      "loss": 2.0592,
+      "step": 565
+    },
+    {
+      "epoch": 0.13668196087901474,
+      "grad_norm": 0.28540095686912537,
+      "learning_rate": 9.654503414871041e-05,
+      "loss": 1.8457,
+      "step": 566
+    },
+    {
+      "epoch": 0.13692344844240523,
+      "grad_norm": 0.291238933801651,
+      "learning_rate": 9.653088100703395e-05,
+      "loss": 1.7909,
+      "step": 567
+    },
+    {
+      "epoch": 0.1371649360057957,
+      "grad_norm": 0.28494998812675476,
+      "learning_rate": 9.65166999773868e-05,
+      "loss": 1.7552,
+      "step": 568
+    },
+    {
+      "epoch": 0.1374064235691862,
+      "grad_norm": 0.2778211832046509,
+      "learning_rate": 9.650249106826823e-05,
+      "loss": 1.7522,
+      "step": 569
+    },
+    {
+      "epoch": 0.13764791113257668,
+      "grad_norm": 0.2810254395008087,
+      "learning_rate": 9.648825428819426e-05,
+      "loss": 1.616,
+      "step": 570
+    },
+    {
+      "epoch": 0.13788939869596717,
+      "grad_norm": 0.2753067910671234,
+      "learning_rate": 9.647398964569761e-05,
+      "loss": 1.6586,
+      "step": 571
+    },
+    {
+      "epoch": 0.13813088625935765,
+      "grad_norm": 0.28924110531806946,
+      "learning_rate": 9.645969714932768e-05,
+      "loss": 1.848,
+      "step": 572
+    },
+    {
+      "epoch": 0.13837237382274814,
+      "grad_norm": 0.29547610878944397,
+      "learning_rate": 9.644537680765061e-05,
+      "loss": 1.8143,
+      "step": 573
+    },
+    {
+      "epoch": 0.13861386138613863,
+      "grad_norm": 0.2900772988796234,
+      "learning_rate": 9.64310286292492e-05,
+      "loss": 1.9223,
+      "step": 574
+    },
+    {
+      "epoch": 0.1388553489495291,
+      "grad_norm": 0.30921927094459534,
+      "learning_rate": 9.64166526227229e-05,
+      "loss": 1.6421,
+      "step": 575
+    },
+    {
+      "epoch": 0.1390968365129196,
+      "grad_norm": 0.2801254987716675,
+      "learning_rate": 9.640224879668791e-05,
+      "loss": 1.6986,
+      "step": 576
+    },
+    {
+      "epoch": 0.13933832407631008,
+      "grad_norm": 0.3085472583770752,
+      "learning_rate": 9.638781715977704e-05,
+      "loss": 1.961,
+      "step": 577
+    },
+    {
+      "epoch": 0.13957981163970057,
+      "grad_norm": 0.29188886284828186,
+      "learning_rate": 9.637335772063983e-05,
+      "loss": 1.6771,
+      "step": 578
+    },
+    {
+      "epoch": 0.13982129920309105,
+      "grad_norm": 0.30945637822151184,
+      "learning_rate": 9.635887048794242e-05,
+      "loss": 2.1181,
+      "step": 579
+    },
+    {
+      "epoch": 0.14006278676648154,
+      "grad_norm": 0.2721393406391144,
+      "learning_rate": 9.634435547036763e-05,
+      "loss": 1.7944,
+      "step": 580
+    },
+    {
+      "epoch": 0.14030427432987203,
+      "grad_norm": 0.2815290093421936,
+      "learning_rate": 9.632981267661496e-05,
+      "loss": 1.5639,
+      "step": 581
+    },
+    {
+      "epoch": 0.1405457618932625,
+      "grad_norm": 0.2875358760356903,
+      "learning_rate": 9.631524211540051e-05,
+      "loss": 1.8578,
+      "step": 582
+    },
+    {
+      "epoch": 0.14078724945665297,
+      "grad_norm": 0.3072429597377777,
+      "learning_rate": 9.630064379545707e-05,
+      "loss": 1.6873,
+      "step": 583
+    },
+    {
+      "epoch": 0.14102873702004345,
+      "grad_norm": 0.3023555278778076,
+      "learning_rate": 9.628601772553403e-05,
+      "loss": 1.7457,
+      "step": 584
+    },
+    {
+      "epoch": 0.14127022458343394,
+      "grad_norm": 0.3081570267677307,
+      "learning_rate": 9.627136391439744e-05,
+      "loss": 1.7089,
+      "step": 585
+    },
+    {
+      "epoch": 0.14151171214682443,
+      "grad_norm": 0.2926265299320221,
+      "learning_rate": 9.625668237082995e-05,
+      "loss": 1.8465,
+      "step": 586
+    },
+    {
+      "epoch": 0.1417531997102149,
+      "grad_norm": 0.30102863907814026,
+      "learning_rate": 9.624197310363085e-05,
+      "loss": 1.779,
+      "step": 587
+    },
+    {
+      "epoch": 0.1419946872736054,
+      "grad_norm": 0.3015720844268799,
+      "learning_rate": 9.622723612161606e-05,
+      "loss": 1.8684,
+      "step": 588
+    },
+    {
+      "epoch": 0.14223617483699588,
+      "grad_norm": 0.2746000587940216,
+      "learning_rate": 9.621247143361806e-05,
+      "loss": 1.734,
+      "step": 589
+    },
+    {
+      "epoch": 0.14247766240038637,
+      "grad_norm": 0.2910473048686981,
+      "learning_rate": 9.619767904848598e-05,
+      "loss": 1.9269,
+      "step": 590
+    },
+    {
+      "epoch": 0.14271914996377685,
+      "grad_norm": 0.27721109986305237,
+      "learning_rate": 9.618285897508553e-05,
+      "loss": 1.6957,
+      "step": 591
+    },
+    {
+      "epoch": 0.14296063752716734,
+      "grad_norm": 0.2783820629119873,
+      "learning_rate": 9.616801122229904e-05,
+      "loss": 1.7603,
+      "step": 592
+    },
+    {
+      "epoch": 0.14320212509055782,
+      "grad_norm": 0.2656041979789734,
+      "learning_rate": 9.615313579902538e-05,
+      "loss": 1.6831,
+      "step": 593
+    },
+    {
+      "epoch": 0.1434436126539483,
+      "grad_norm": 0.2902359068393707,
+      "learning_rate": 9.613823271418006e-05,
+      "loss": 1.7025,
+      "step": 594
+    },
+    {
+      "epoch": 0.1436851002173388,
+      "grad_norm": 0.2851390838623047,
+      "learning_rate": 9.612330197669514e-05,
+      "loss": 1.6937,
+      "step": 595
+    },
+    {
+      "epoch": 0.14392658778072928,
+      "grad_norm": 0.2772185206413269,
+      "learning_rate": 9.610834359551926e-05,
+      "loss": 1.6978,
+      "step": 596
+    },
+    {
+      "epoch": 0.14416807534411977,
+      "grad_norm": 0.2872218191623688,
+      "learning_rate": 9.609335757961761e-05,
+      "loss": 1.8004,
+      "step": 597
+    },
+    {
+      "epoch": 0.14440956290751025,
+      "grad_norm": 0.27167755365371704,
+      "learning_rate": 9.607834393797197e-05,
+      "loss": 1.6695,
+      "step": 598
+    },
+    {
+      "epoch": 0.14465105047090074,
+      "grad_norm": 0.275107741355896,
+      "learning_rate": 9.606330267958066e-05,
+      "loss": 1.8051,
+      "step": 599
+    },
+    {
+      "epoch": 0.14489253803429122,
+      "grad_norm": 0.2851438522338867,
+      "learning_rate": 9.604823381345856e-05,
+      "loss": 1.863,
+      "step": 600
+    },
+    {
+      "epoch": 0.1451340255976817,
+      "grad_norm": 0.2878473103046417,
+      "learning_rate": 9.603313734863708e-05,
+      "loss": 1.8116,
+      "step": 601
+    },
+    {
+      "epoch": 0.1453755131610722,
+      "grad_norm": 0.2840920388698578,
+      "learning_rate": 9.601801329416422e-05,
+      "loss": 1.6956,
+      "step": 602
+    },
+    {
+      "epoch": 0.14561700072446268,
+      "grad_norm": 0.27009525895118713,
+      "learning_rate": 9.600286165910445e-05,
+      "loss": 1.6875,
+      "step": 603
+    },
+    {
+      "epoch": 0.14585848828785317,
+      "grad_norm": 0.31799811124801636,
+      "learning_rate": 9.598768245253879e-05,
+      "loss": 1.8778,
+      "step": 604
+    },
+    {
+      "epoch": 0.14609997585124365,
+      "grad_norm": 0.2943333685398102,
+      "learning_rate": 9.597247568356484e-05,
+      "loss": 1.6921,
+      "step": 605
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 0.2869589030742645,
+      "learning_rate": 9.595724136129665e-05,
+      "loss": 1.8085,
+      "step": 606
+    },
+    {
+      "epoch": 0.14658295097802462,
+      "grad_norm": 0.2950487434864044,
+      "learning_rate": 9.594197949486478e-05,
+      "loss": 1.8381,
+      "step": 607
+    },
+    {
+      "epoch": 0.1468244385414151,
+      "grad_norm": 0.2995087206363678,
+      "learning_rate": 9.592669009341637e-05,
+      "loss": 1.7566,
+      "step": 608
+    },
+    {
+      "epoch": 0.1470659261048056,
+      "grad_norm": 0.302502304315567,
+      "learning_rate": 9.5911373166115e-05,
+      "loss": 1.8809,
+      "step": 609
+    },
+    {
+      "epoch": 0.14730741366819608,
+      "grad_norm": 0.30947795510292053,
+      "learning_rate": 9.589602872214079e-05,
+      "loss": 1.9043,
+      "step": 610
+    },
+    {
+      "epoch": 0.14754890123158657,
+      "grad_norm": 0.29188233613967896,
+      "learning_rate": 9.588065677069029e-05,
+      "loss": 1.7712,
+      "step": 611
+    },
+    {
+      "epoch": 0.14779038879497705,
+      "grad_norm": 0.29185009002685547,
+      "learning_rate": 9.58652573209766e-05,
+      "loss": 1.907,
+      "step": 612
+    },
+    {
+      "epoch": 0.14803187635836754,
+      "grad_norm": 0.2943671941757202,
+      "learning_rate": 9.584983038222928e-05,
+      "loss": 2.0036,
+      "step": 613
+    },
+    {
+      "epoch": 0.14827336392175802,
+      "grad_norm": 0.2962663471698761,
+      "learning_rate": 9.583437596369434e-05,
+      "loss": 1.8042,
+      "step": 614
+    },
+    {
+      "epoch": 0.1485148514851485,
+      "grad_norm": 0.28282076120376587,
+      "learning_rate": 9.581889407463431e-05,
+      "loss": 1.8427,
+      "step": 615
+    },
+    {
+      "epoch": 0.148756339048539,
+      "grad_norm": 0.27998435497283936,
+      "learning_rate": 9.580338472432815e-05,
+      "loss": 1.8312,
+      "step": 616
+    },
+    {
+      "epoch": 0.14899782661192948,
+      "grad_norm": 0.285490483045578,
+      "learning_rate": 9.578784792207127e-05,
+      "loss": 1.745,
+      "step": 617
+    },
+    {
+      "epoch": 0.14923931417531996,
+      "grad_norm": 0.2885023057460785,
+      "learning_rate": 9.577228367717554e-05,
+      "loss": 1.6725,
+      "step": 618
+    },
+    {
+      "epoch": 0.14948080173871045,
+      "grad_norm": 0.3022843897342682,
+      "learning_rate": 9.575669199896933e-05,
+      "loss": 1.7719,
+      "step": 619
+    },
+    {
+      "epoch": 0.14972228930210094,
+      "grad_norm": 0.27724575996398926,
+      "learning_rate": 9.574107289679737e-05,
+      "loss": 1.6429,
+      "step": 620
+    },
+    {
+      "epoch": 0.14996377686549142,
+      "grad_norm": 0.3123320937156677,
+      "learning_rate": 9.572542638002087e-05,
+      "loss": 1.9326,
+      "step": 621
+    },
+    {
+      "epoch": 0.1502052644288819,
+      "grad_norm": 0.33049526810646057,
+      "learning_rate": 9.570975245801746e-05,
+      "loss": 2.183,
+      "step": 622
+    },
+    {
+      "epoch": 0.1504467519922724,
+      "grad_norm": 0.2905597686767578,
+      "learning_rate": 9.569405114018122e-05,
+      "loss": 1.838,
+      "step": 623
+    },
+    {
+      "epoch": 0.15068823955566288,
+      "grad_norm": 0.2684404253959656,
+      "learning_rate": 9.567832243592261e-05,
+      "loss": 1.6479,
+      "step": 624
+    },
+    {
+      "epoch": 0.15092972711905336,
+      "grad_norm": 0.2934153378009796,
+      "learning_rate": 9.566256635466852e-05,
+      "loss": 1.8059,
+      "step": 625
+    },
+    {
+      "epoch": 0.15117121468244385,
+      "grad_norm": 0.28340545296669006,
+      "learning_rate": 9.564678290586229e-05,
+      "loss": 1.9517,
+      "step": 626
+    },
+    {
+      "epoch": 0.15141270224583434,
+      "grad_norm": 0.3298986852169037,
+      "learning_rate": 9.563097209896357e-05,
+      "loss": 1.6387,
+      "step": 627
+    },
+    {
+      "epoch": 0.15165418980922482,
+      "grad_norm": 0.2907513380050659,
+      "learning_rate": 9.56151339434485e-05,
+      "loss": 1.8626,
+      "step": 628
+    },
+    {
+      "epoch": 0.1518956773726153,
+      "grad_norm": 0.29203519225120544,
+      "learning_rate": 9.559926844880956e-05,
+      "loss": 1.8376,
+      "step": 629
+    },
+    {
+      "epoch": 0.1521371649360058,
+      "grad_norm": 0.29703041911125183,
+      "learning_rate": 9.558337562455561e-05,
+      "loss": 1.7033,
+      "step": 630
+    },
+    {
+      "epoch": 0.15237865249939628,
+      "grad_norm": 0.28163012862205505,
+      "learning_rate": 9.556745548021193e-05,
+      "loss": 1.7325,
+      "step": 631
+    },
+    {
+      "epoch": 0.15262014006278676,
+      "grad_norm": 0.3157091736793518,
+      "learning_rate": 9.555150802532016e-05,
+      "loss": 1.77,
+      "step": 632
+    },
+    {
+      "epoch": 0.15286162762617725,
+      "grad_norm": 0.295043021440506,
+      "learning_rate": 9.553553326943826e-05,
+      "loss": 1.8613,
+      "step": 633
+    },
+    {
+      "epoch": 0.15310311518956773,
+      "grad_norm": 0.29055267572402954,
+      "learning_rate": 9.551953122214064e-05,
+      "loss": 1.8194,
+      "step": 634
+    },
+    {
+      "epoch": 0.15334460275295822,
+      "grad_norm": 0.31686437129974365,
+      "learning_rate": 9.550350189301797e-05,
+      "loss": 2.034,
+      "step": 635
+    },
+    {
+      "epoch": 0.1535860903163487,
+      "grad_norm": 0.30802491307258606,
+      "learning_rate": 9.548744529167737e-05,
+      "loss": 1.8503,
+      "step": 636
+    },
+    {
+      "epoch": 0.1538275778797392,
+      "grad_norm": 0.2771475613117218,
+      "learning_rate": 9.547136142774222e-05,
+      "loss": 1.6846,
+      "step": 637
+    },
+    {
+      "epoch": 0.15406906544312968,
+      "grad_norm": 0.3178752362728119,
+      "learning_rate": 9.54552503108523e-05,
+      "loss": 1.8496,
+      "step": 638
+    },
+    {
+      "epoch": 0.15431055300652016,
+      "grad_norm": 0.29422762989997864,
+      "learning_rate": 9.543911195066367e-05,
+      "loss": 1.7731,
+      "step": 639
+    },
+    {
+      "epoch": 0.15455204056991065,
+      "grad_norm": 0.27632033824920654,
+      "learning_rate": 9.542294635684878e-05,
+      "loss": 1.6732,
+      "step": 640
+    },
+    {
+      "epoch": 0.15479352813330113,
+      "grad_norm": 0.2904300093650818,
+      "learning_rate": 9.540675353909636e-05,
+      "loss": 1.6868,
+      "step": 641
+    },
+    {
+      "epoch": 0.15503501569669162,
+      "grad_norm": 0.30907413363456726,
+      "learning_rate": 9.539053350711145e-05,
+      "loss": 2.053,
+      "step": 642
+    },
+    {
+      "epoch": 0.1552765032600821,
+      "grad_norm": 0.3101508319377899,
+      "learning_rate": 9.537428627061545e-05,
+      "loss": 1.9812,
+      "step": 643
+    },
+    {
+      "epoch": 0.1555179908234726,
+      "grad_norm": 0.3084372878074646,
+      "learning_rate": 9.535801183934598e-05,
+      "loss": 1.9835,
+      "step": 644
+    },
+    {
+      "epoch": 0.15575947838686308,
+      "grad_norm": 0.2710745632648468,
+      "learning_rate": 9.534171022305706e-05,
+      "loss": 1.6607,
+      "step": 645
+    },
+    {
+      "epoch": 0.15600096595025356,
+      "grad_norm": 0.2810233235359192,
+      "learning_rate": 9.532538143151895e-05,
+      "loss": 1.7959,
+      "step": 646
+    },
+    {
+      "epoch": 0.15624245351364405,
+      "grad_norm": 0.2879350483417511,
+      "learning_rate": 9.530902547451817e-05,
+      "loss": 1.7773,
+      "step": 647
+    },
+    {
+      "epoch": 0.15648394107703453,
+      "grad_norm": 0.2893047332763672,
+      "learning_rate": 9.529264236185758e-05,
+      "loss": 1.6512,
+      "step": 648
+    },
+    {
+      "epoch": 0.15672542864042502,
+      "grad_norm": 0.29465562105178833,
+      "learning_rate": 9.527623210335628e-05,
+      "loss": 1.9087,
+      "step": 649
+    },
+    {
+      "epoch": 0.1569669162038155,
+      "grad_norm": 0.2800419330596924,
+      "learning_rate": 9.525979470884963e-05,
+      "loss": 1.8204,
+      "step": 650
+    },
+    {
+      "epoch": 0.157208403767206,
+      "grad_norm": 0.27463045716285706,
+      "learning_rate": 9.524333018818931e-05,
+      "loss": 1.6238,
+      "step": 651
+    },
+    {
+      "epoch": 0.15744989133059648,
+      "grad_norm": 0.2779013216495514,
+      "learning_rate": 9.522683855124319e-05,
+      "loss": 1.7749,
+      "step": 652
+    },
+    {
+      "epoch": 0.15769137889398696,
+      "grad_norm": 0.28004294633865356,
+      "learning_rate": 9.521031980789543e-05,
+      "loss": 1.7464,
+      "step": 653
+    },
+    {
+      "epoch": 0.15793286645737745,
+      "grad_norm": 0.2746169865131378,
+      "learning_rate": 9.519377396804644e-05,
+      "loss": 1.7046,
+      "step": 654
+    },
+    {
+      "epoch": 0.15817435402076793,
+      "grad_norm": 0.2713209092617035,
+      "learning_rate": 9.517720104161284e-05,
+      "loss": 1.6901,
+      "step": 655
+    },
+    {
+      "epoch": 0.15841584158415842,
+      "grad_norm": 0.2889479696750641,
+      "learning_rate": 9.51606010385275e-05,
+      "loss": 1.8167,
+      "step": 656
+    },
+    {
+      "epoch": 0.1586573291475489,
+      "grad_norm": 0.4551263451576233,
+      "learning_rate": 9.514397396873956e-05,
+      "loss": 2.2537,
+      "step": 657
+    },
+    {
+      "epoch": 0.1588988167109394,
+      "grad_norm": 0.2781124413013458,
+      "learning_rate": 9.512731984221428e-05,
+      "loss": 1.6174,
+      "step": 658
+    },
+    {
+      "epoch": 0.15914030427432987,
+      "grad_norm": 0.30449262261390686,
+      "learning_rate": 9.511063866893324e-05,
+      "loss": 1.8033,
+      "step": 659
+    },
+    {
+      "epoch": 0.15938179183772036,
+      "grad_norm": 0.2778739035129547,
+      "learning_rate": 9.509393045889417e-05,
+      "loss": 1.7579,
+      "step": 660
+    },
+    {
+      "epoch": 0.15962327940111085,
+      "grad_norm": 0.31164950132369995,
+      "learning_rate": 9.507719522211104e-05,
+      "loss": 1.8609,
+      "step": 661
+    },
+    {
+      "epoch": 0.15986476696450133,
+      "grad_norm": 0.28629255294799805,
+      "learning_rate": 9.506043296861399e-05,
+      "loss": 1.8606,
+      "step": 662
+    },
+    {
+      "epoch": 0.16010625452789182,
+      "grad_norm": 0.29869675636291504,
+      "learning_rate": 9.504364370844938e-05,
+      "loss": 1.9861,
+      "step": 663
+    },
+    {
+      "epoch": 0.1603477420912823,
+      "grad_norm": 0.2828952968120575,
+      "learning_rate": 9.502682745167972e-05,
+      "loss": 1.6205,
+      "step": 664
+    },
+    {
+      "epoch": 0.1605892296546728,
+      "grad_norm": 0.30284184217453003,
+      "learning_rate": 9.500998420838374e-05,
+      "loss": 1.8012,
+      "step": 665
+    },
+    {
+      "epoch": 0.16083071721806327,
+      "grad_norm": 0.2756812274456024,
+      "learning_rate": 9.49931139886563e-05,
+      "loss": 1.7166,
+      "step": 666
+    },
+    {
+      "epoch": 0.16107220478145376,
+      "grad_norm": 0.2983384132385254,
+      "learning_rate": 9.497621680260848e-05,
+      "loss": 1.7543,
+      "step": 667
+    },
+    {
+      "epoch": 0.16131369234484425,
+      "grad_norm": 0.32105186581611633,
+      "learning_rate": 9.495929266036747e-05,
+      "loss": 1.6919,
+      "step": 668
+    },
+    {
+      "epoch": 0.16155517990823473,
+      "grad_norm": 0.2885781526565552,
+      "learning_rate": 9.494234157207665e-05,
+      "loss": 1.723,
+      "step": 669
+    },
+    {
+      "epoch": 0.16179666747162522,
+      "grad_norm": 0.28944310545921326,
+      "learning_rate": 9.492536354789554e-05,
+      "loss": 1.7258,
+      "step": 670
+    },
+    {
+      "epoch": 0.1620381550350157,
+      "grad_norm": 0.31029340624809265,
+      "learning_rate": 9.490835859799983e-05,
+      "loss": 1.7345,
+      "step": 671
+    },
+    {
+      "epoch": 0.1622796425984062,
+      "grad_norm": 0.2986032962799072,
+      "learning_rate": 9.489132673258128e-05,
+      "loss": 1.6178,
+      "step": 672
+    },
+    {
+      "epoch": 0.16252113016179667,
+      "grad_norm": 0.2751445770263672,
+      "learning_rate": 9.487426796184785e-05,
+      "loss": 1.7689,
+      "step": 673
+    },
+    {
+      "epoch": 0.16276261772518716,
+      "grad_norm": 0.34798896312713623,
+      "learning_rate": 9.485718229602359e-05,
+      "loss": 2.016,
+      "step": 674
+    },
+    {
+      "epoch": 0.16300410528857764,
+      "grad_norm": 0.31324881315231323,
+      "learning_rate": 9.484006974534869e-05,
+      "loss": 1.831,
+      "step": 675
+    },
+    {
+      "epoch": 0.16324559285196813,
+      "grad_norm": 0.28822532296180725,
+      "learning_rate": 9.482293032007943e-05,
+      "loss": 1.8226,
+      "step": 676
+    },
+    {
+      "epoch": 0.16348708041535862,
+      "grad_norm": 0.2884620726108551,
+      "learning_rate": 9.480576403048822e-05,
+      "loss": 1.6349,
+      "step": 677
+    },
+    {
+      "epoch": 0.1637285679787491,
+      "grad_norm": 0.29445016384124756,
+      "learning_rate": 9.478857088686355e-05,
+      "loss": 1.6545,
+      "step": 678
+    },
+    {
+      "epoch": 0.1639700555421396,
+      "grad_norm": 0.28695762157440186,
+      "learning_rate": 9.477135089951001e-05,
+      "loss": 1.6071,
+      "step": 679
+    },
+    {
+      "epoch": 0.16421154310553007,
+      "grad_norm": 0.2973484694957733,
+      "learning_rate": 9.47541040787483e-05,
+      "loss": 1.9545,
+      "step": 680
+    },
+    {
+      "epoch": 0.16445303066892056,
+      "grad_norm": 0.3042725920677185,
+      "learning_rate": 9.473683043491518e-05,
+      "loss": 1.8942,
+      "step": 681
+    },
+    {
+      "epoch": 0.16469451823231104,
+      "grad_norm": 0.2901724874973297,
+      "learning_rate": 9.471952997836348e-05,
+      "loss": 1.6541,
+      "step": 682
+    },
+    {
+      "epoch": 0.16493600579570153,
+      "grad_norm": 0.27054041624069214,
+      "learning_rate": 9.470220271946215e-05,
+      "loss": 1.6211,
+      "step": 683
+    },
+    {
+      "epoch": 0.16517749335909201,
+      "grad_norm": 0.28445035219192505,
+      "learning_rate": 9.468484866859612e-05,
+      "loss": 1.7979,
+      "step": 684
+    },
+    {
+      "epoch": 0.1654189809224825,
+      "grad_norm": 0.29308000206947327,
+      "learning_rate": 9.466746783616645e-05,
+      "loss": 1.7483,
+      "step": 685
+    },
+    {
+      "epoch": 0.16566046848587299,
+      "grad_norm": 0.2914656698703766,
+      "learning_rate": 9.465006023259022e-05,
+      "loss": 1.7282,
+      "step": 686
+    },
+    {
+      "epoch": 0.16590195604926347,
+      "grad_norm": 0.2674042284488678,
+      "learning_rate": 9.463262586830056e-05,
+      "loss": 1.7118,
+      "step": 687
+    },
+    {
+      "epoch": 0.16614344361265396,
+      "grad_norm": 0.2843051552772522,
+      "learning_rate": 9.461516475374663e-05,
+      "loss": 1.8195,
+      "step": 688
+    },
+    {
+      "epoch": 0.16638493117604444,
+      "grad_norm": 0.29515916109085083,
+      "learning_rate": 9.459767689939365e-05,
+      "loss": 1.7215,
+      "step": 689
+    },
+    {
+      "epoch": 0.16662641873943493,
+      "grad_norm": 0.2934631109237671,
+      "learning_rate": 9.458016231572283e-05,
+      "loss": 1.7785,
+      "step": 690
+    },
+    {
+      "epoch": 0.16686790630282541,
+      "grad_norm": 0.28389841318130493,
+      "learning_rate": 9.45626210132314e-05,
+      "loss": 1.6819,
+      "step": 691
+    },
+    {
+      "epoch": 0.1671093938662159,
+      "grad_norm": 0.30072394013404846,
+      "learning_rate": 9.45450530024327e-05,
+      "loss": 1.7743,
+      "step": 692
+    },
+    {
+      "epoch": 0.16735088142960639,
+      "grad_norm": 0.28151774406433105,
+      "learning_rate": 9.452745829385592e-05,
+      "loss": 1.7631,
+      "step": 693
+    },
+    {
+      "epoch": 0.16759236899299687,
+      "grad_norm": 0.2804087698459625,
+      "learning_rate": 9.450983689804636e-05,
+      "loss": 1.82,
+      "step": 694
+    },
+    {
+      "epoch": 0.16783385655638736,
+      "grad_norm": 0.2899933457374573,
+      "learning_rate": 9.449218882556528e-05,
+      "loss": 1.7561,
+      "step": 695
+    },
+    {
+      "epoch": 0.16807534411977784,
+      "grad_norm": 0.3071437180042267,
+      "learning_rate": 9.447451408698993e-05,
+      "loss": 1.8007,
+      "step": 696
+    },
+    {
+      "epoch": 0.16831683168316833,
+      "grad_norm": 0.2796712815761566,
+      "learning_rate": 9.445681269291354e-05,
+      "loss": 1.6631,
+      "step": 697
+    },
+    {
+      "epoch": 0.1685583192465588,
+      "grad_norm": 0.3034926652908325,
+      "learning_rate": 9.443908465394535e-05,
+      "loss": 1.879,
+      "step": 698
+    },
+    {
+      "epoch": 0.1687998068099493,
+      "grad_norm": 0.2936783730983734,
+      "learning_rate": 9.442132998071052e-05,
+      "loss": 1.7325,
+      "step": 699
+    },
+    {
+      "epoch": 0.16904129437333978,
+      "grad_norm": 0.2850625813007355,
+      "learning_rate": 9.44035486838502e-05,
+      "loss": 1.7287,
+      "step": 700
+    },
+    {
+      "epoch": 0.16928278193673027,
+      "grad_norm": 0.28181543946266174,
+      "learning_rate": 9.43857407740215e-05,
+      "loss": 1.7932,
+      "step": 701
+    },
+    {
+      "epoch": 0.16952426950012076,
+      "grad_norm": 0.2871532142162323,
+      "learning_rate": 9.436790626189745e-05,
+      "loss": 1.7514,
+      "step": 702
+    },
+    {
+      "epoch": 0.16976575706351124,
+      "grad_norm": 0.28033384680747986,
+      "learning_rate": 9.435004515816703e-05,
+      "loss": 1.7618,
+      "step": 703
+    },
+    {
+      "epoch": 0.17000724462690173,
+      "grad_norm": 0.28677472472190857,
+      "learning_rate": 9.433215747353522e-05,
+      "loss": 1.7955,
+      "step": 704
+    },
+    {
+      "epoch": 0.1702487321902922,
+      "grad_norm": 0.28708505630493164,
+      "learning_rate": 9.431424321872286e-05,
+      "loss": 1.7104,
+      "step": 705
+    },
+    {
+      "epoch": 0.1704902197536827,
+      "grad_norm": 0.2901146113872528,
+      "learning_rate": 9.429630240446674e-05,
+      "loss": 1.7043,
+      "step": 706
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 0.3041292130947113,
+      "learning_rate": 9.427833504151955e-05,
+      "loss": 1.645,
+      "step": 707
+    },
+    {
+      "epoch": 0.17097319488046367,
+      "grad_norm": 0.2819676995277405,
+      "learning_rate": 9.426034114064994e-05,
+      "loss": 1.7287,
+      "step": 708
+    },
+    {
+      "epoch": 0.17121468244385415,
+      "grad_norm": 0.2727168798446655,
+      "learning_rate": 9.424232071264241e-05,
+      "loss": 1.6778,
+      "step": 709
+    },
+    {
+      "epoch": 0.17145617000724464,
+      "grad_norm": 0.28442737460136414,
+      "learning_rate": 9.422427376829739e-05,
+      "loss": 1.7041,
+      "step": 710
+    },
+    {
+      "epoch": 0.17169765757063513,
+      "grad_norm": 0.49621260166168213,
+      "learning_rate": 9.42062003184312e-05,
+      "loss": 1.558,
+      "step": 711
+    },
+    {
+      "epoch": 0.17193914513402558,
+      "grad_norm": 0.2837514877319336,
+      "learning_rate": 9.418810037387603e-05,
+      "loss": 1.7101,
+      "step": 712
+    },
+    {
+      "epoch": 0.17218063269741607,
+      "grad_norm": 0.27921706438064575,
+      "learning_rate": 9.416997394547996e-05,
+      "loss": 1.8319,
+      "step": 713
+    },
+    {
+      "epoch": 0.17242212026080656,
+      "grad_norm": 0.2778397798538208,
+      "learning_rate": 9.415182104410693e-05,
+      "loss": 1.724,
+      "step": 714
+    },
+    {
+      "epoch": 0.17266360782419704,
+      "grad_norm": 0.28968319296836853,
+      "learning_rate": 9.413364168063681e-05,
+      "loss": 1.6833,
+      "step": 715
+    },
+    {
+      "epoch": 0.17290509538758753,
+      "grad_norm": 0.2843336760997772,
+      "learning_rate": 9.411543586596523e-05,
+      "loss": 1.7913,
+      "step": 716
+    },
+    {
+      "epoch": 0.173146582950978,
+      "grad_norm": 0.2921014428138733,
+      "learning_rate": 9.409720361100374e-05,
+      "loss": 1.7845,
+      "step": 717
+    },
+    {
+      "epoch": 0.1733880705143685,
+      "grad_norm": 0.3008785545825958,
+      "learning_rate": 9.407894492667972e-05,
+      "loss": 1.7576,
+      "step": 718
+    },
+    {
+      "epoch": 0.17362955807775898,
+      "grad_norm": 0.2838538587093353,
+      "learning_rate": 9.406065982393639e-05,
+      "loss": 1.7193,
+      "step": 719
+    },
+    {
+      "epoch": 0.17387104564114947,
+      "grad_norm": 0.3217141628265381,
+      "learning_rate": 9.40423483137328e-05,
+      "loss": 1.87,
+      "step": 720
+    },
+    {
+      "epoch": 0.17411253320453995,
+      "grad_norm": 0.29201528429985046,
+      "learning_rate": 9.402401040704383e-05,
+      "loss": 1.7704,
+      "step": 721
+    },
+    {
+      "epoch": 0.17435402076793044,
+      "grad_norm": 0.2874560058116913,
+      "learning_rate": 9.400564611486017e-05,
+      "loss": 1.7274,
+      "step": 722
+    },
+    {
+      "epoch": 0.17459550833132093,
+      "grad_norm": 0.27963292598724365,
+      "learning_rate": 9.398725544818838e-05,
+      "loss": 1.6752,
+      "step": 723
+    },
+    {
+      "epoch": 0.1748369958947114,
+      "grad_norm": 0.2946629524230957,
+      "learning_rate": 9.396883841805073e-05,
+      "loss": 1.9184,
+      "step": 724
+    },
+    {
+      "epoch": 0.1750784834581019,
+      "grad_norm": 0.2995729446411133,
+      "learning_rate": 9.395039503548537e-05,
+      "loss": 1.7893,
+      "step": 725
+    },
+    {
+      "epoch": 0.17531997102149238,
+      "grad_norm": 0.292082816362381,
+      "learning_rate": 9.393192531154621e-05,
+      "loss": 1.8141,
+      "step": 726
+    },
+    {
+      "epoch": 0.17556145858488287,
+      "grad_norm": 0.27504876255989075,
+      "learning_rate": 9.391342925730295e-05,
+      "loss": 1.6477,
+      "step": 727
+    },
+    {
+      "epoch": 0.17580294614827335,
+      "grad_norm": 0.2799321413040161,
+      "learning_rate": 9.389490688384109e-05,
+      "loss": 1.6797,
+      "step": 728
+    },
+    {
+      "epoch": 0.17604443371166384,
+      "grad_norm": 0.2817777991294861,
+      "learning_rate": 9.387635820226189e-05,
+      "loss": 1.7789,
+      "step": 729
+    },
+    {
+      "epoch": 0.17628592127505432,
+      "grad_norm": 0.266647607088089,
+      "learning_rate": 9.385778322368235e-05,
+      "loss": 1.6869,
+      "step": 730
+    },
+    {
+      "epoch": 0.1765274088384448,
+      "grad_norm": 0.2966921925544739,
+      "learning_rate": 9.383918195923527e-05,
+      "loss": 1.8977,
+      "step": 731
+    },
+    {
+      "epoch": 0.1767688964018353,
+      "grad_norm": 0.2740517258644104,
+      "learning_rate": 9.382055442006923e-05,
+      "loss": 1.7156,
+      "step": 732
+    },
+    {
+      "epoch": 0.17701038396522578,
+      "grad_norm": 0.27741682529449463,
+      "learning_rate": 9.380190061734847e-05,
+      "loss": 1.6714,
+      "step": 733
+    },
+    {
+      "epoch": 0.17725187152861627,
+      "grad_norm": 0.2760551869869232,
+      "learning_rate": 9.378322056225305e-05,
+      "loss": 1.7512,
+      "step": 734
+    },
+    {
+      "epoch": 0.17749335909200675,
+      "grad_norm": 0.2780584394931793,
+      "learning_rate": 9.376451426597872e-05,
+      "loss": 1.7036,
+      "step": 735
+    },
+    {
+      "epoch": 0.17773484665539724,
+      "grad_norm": 0.2876451909542084,
+      "learning_rate": 9.374578173973698e-05,
+      "loss": 1.6687,
+      "step": 736
+    },
+    {
+      "epoch": 0.17797633421878772,
+      "grad_norm": 0.29878392815589905,
+      "learning_rate": 9.372702299475506e-05,
+      "loss": 1.772,
+      "step": 737
+    },
+    {
+      "epoch": 0.1782178217821782,
+      "grad_norm": 0.28005820512771606,
+      "learning_rate": 9.370823804227586e-05,
+      "loss": 1.7543,
+      "step": 738
+    },
+    {
+      "epoch": 0.1784593093455687,
+      "grad_norm": 0.26662278175354004,
+      "learning_rate": 9.368942689355804e-05,
+      "loss": 1.6537,
+      "step": 739
+    },
+    {
+      "epoch": 0.17870079690895918,
+      "grad_norm": 0.2788659632205963,
+      "learning_rate": 9.367058955987591e-05,
+      "loss": 1.7021,
+      "step": 740
+    },
+    {
+      "epoch": 0.17894228447234967,
+      "grad_norm": 0.2913663685321808,
+      "learning_rate": 9.36517260525195e-05,
+      "loss": 1.8001,
+      "step": 741
+    },
+    {
+      "epoch": 0.17918377203574015,
+      "grad_norm": 0.2656884789466858,
+      "learning_rate": 9.363283638279457e-05,
+      "loss": 1.6377,
+      "step": 742
+    },
+    {
+      "epoch": 0.17942525959913064,
+      "grad_norm": 0.3025050759315491,
+      "learning_rate": 9.361392056202246e-05,
+      "loss": 1.8939,
+      "step": 743
+    },
+    {
+      "epoch": 0.17966674716252112,
+      "grad_norm": 0.30079180002212524,
+      "learning_rate": 9.359497860154027e-05,
+      "loss": 1.8976,
+      "step": 744
+    },
+    {
+      "epoch": 0.1799082347259116,
+      "grad_norm": 0.28714820742607117,
+      "learning_rate": 9.357601051270074e-05,
+      "loss": 1.7784,
+      "step": 745
+    },
+    {
+      "epoch": 0.1801497222893021,
+      "grad_norm": 0.29758569598197937,
+      "learning_rate": 9.355701630687225e-05,
+      "loss": 1.7182,
+      "step": 746
+    },
+    {
+      "epoch": 0.18039120985269258,
+      "grad_norm": 0.26696479320526123,
+      "learning_rate": 9.353799599543887e-05,
+      "loss": 1.6108,
+      "step": 747
+    },
+    {
+      "epoch": 0.18063269741608307,
+      "grad_norm": 0.2863609790802002,
+      "learning_rate": 9.351894958980026e-05,
+      "loss": 1.8426,
+      "step": 748
+    },
+    {
+      "epoch": 0.18087418497947355,
+      "grad_norm": 0.2715589702129364,
+      "learning_rate": 9.349987710137179e-05,
+      "loss": 1.7442,
+      "step": 749
+    },
+    {
+      "epoch": 0.18111567254286404,
+      "grad_norm": 0.2864156663417816,
+      "learning_rate": 9.348077854158441e-05,
+      "loss": 1.7882,
+      "step": 750
+    },
+    {
+      "epoch": 0.18135716010625452,
+      "grad_norm": 0.30364513397216797,
+      "learning_rate": 9.346165392188473e-05,
+      "loss": 1.64,
+      "step": 751
+    },
+    {
+      "epoch": 0.181598647669645,
+      "grad_norm": 0.3006398379802704,
+      "learning_rate": 9.344250325373495e-05,
+      "loss": 1.901,
+      "step": 752
+    },
+    {
+      "epoch": 0.1818401352330355,
+      "grad_norm": 0.2953626811504364,
+      "learning_rate": 9.34233265486129e-05,
+      "loss": 1.797,
+      "step": 753
+    },
+    {
+      "epoch": 0.18208162279642598,
+      "grad_norm": 0.2777319550514221,
+      "learning_rate": 9.3404123818012e-05,
+      "loss": 1.7733,
+      "step": 754
+    },
+    {
+      "epoch": 0.18232311035981646,
+      "grad_norm": 0.2769542932510376,
+      "learning_rate": 9.338489507344128e-05,
+      "loss": 1.7931,
+      "step": 755
+    },
+    {
+      "epoch": 0.18256459792320695,
+      "grad_norm": 0.28098264336586,
+      "learning_rate": 9.336564032642536e-05,
+      "loss": 1.7173,
+      "step": 756
+    },
+    {
+      "epoch": 0.18280608548659744,
+      "grad_norm": 0.278130441904068,
+      "learning_rate": 9.334635958850446e-05,
+      "loss": 1.7329,
+      "step": 757
+    },
+    {
+      "epoch": 0.18304757304998792,
+      "grad_norm": 0.28257855772972107,
+      "learning_rate": 9.332705287123434e-05,
+      "loss": 1.7744,
+      "step": 758
+    },
+    {
+      "epoch": 0.1832890606133784,
+      "grad_norm": 0.28167739510536194,
+      "learning_rate": 9.330772018618634e-05,
+      "loss": 1.821,
+      "step": 759
+    },
+    {
+      "epoch": 0.1835305481767689,
+      "grad_norm": 0.29732733964920044,
+      "learning_rate": 9.32883615449474e-05,
+      "loss": 1.7651,
+      "step": 760
+    },
+    {
+      "epoch": 0.18377203574015938,
+      "grad_norm": 0.27904653549194336,
+      "learning_rate": 9.326897695912e-05,
+      "loss": 1.884,
+      "step": 761
+    },
+    {
+      "epoch": 0.18401352330354986,
+      "grad_norm": 0.29639023542404175,
+      "learning_rate": 9.324956644032213e-05,
+      "loss": 1.9033,
+      "step": 762
+    },
+    {
+      "epoch": 0.18425501086694035,
+      "grad_norm": 0.28502506017684937,
+      "learning_rate": 9.323013000018736e-05,
+      "loss": 1.8383,
+      "step": 763
+    },
+    {
+      "epoch": 0.18449649843033084,
+      "grad_norm": 0.2888893187046051,
+      "learning_rate": 9.32106676503648e-05,
+      "loss": 1.7931,
+      "step": 764
+    },
+    {
+      "epoch": 0.18473798599372132,
+      "grad_norm": 0.29819416999816895,
+      "learning_rate": 9.31911794025191e-05,
+      "loss": 1.8012,
+      "step": 765
+    },
+    {
+      "epoch": 0.1849794735571118,
+      "grad_norm": 0.2912543714046478,
+      "learning_rate": 9.317166526833035e-05,
+      "loss": 1.8254,
+      "step": 766
+    },
+    {
+      "epoch": 0.1852209611205023,
+      "grad_norm": 0.2873326539993286,
+      "learning_rate": 9.315212525949428e-05,
+      "loss": 1.8009,
+      "step": 767
+    },
+    {
+      "epoch": 0.18546244868389278,
+      "grad_norm": 0.27630600333213806,
+      "learning_rate": 9.313255938772202e-05,
+      "loss": 1.7497,
+      "step": 768
+    },
+    {
+      "epoch": 0.18570393624728326,
+      "grad_norm": 0.28622114658355713,
+      "learning_rate": 9.311296766474027e-05,
+      "loss": 1.6059,
+      "step": 769
+    },
+    {
+      "epoch": 0.18594542381067375,
+      "grad_norm": 0.2943037748336792,
+      "learning_rate": 9.309335010229118e-05,
+      "loss": 1.728,
+      "step": 770
+    },
+    {
+      "epoch": 0.18618691137406423,
+      "grad_norm": 0.29202359914779663,
+      "learning_rate": 9.30737067121324e-05,
+      "loss": 1.7766,
+      "step": 771
+    },
+    {
+      "epoch": 0.18642839893745472,
+      "grad_norm": 0.2959134578704834,
+      "learning_rate": 9.30540375060371e-05,
+      "loss": 1.8195,
+      "step": 772
+    },
+    {
+      "epoch": 0.1866698865008452,
+      "grad_norm": 0.28923317790031433,
+      "learning_rate": 9.303434249579384e-05,
+      "loss": 1.6646,
+      "step": 773
+    },
+    {
+      "epoch": 0.1869113740642357,
+      "grad_norm": 0.2839503288269043,
+      "learning_rate": 9.301462169320672e-05,
+      "loss": 1.7667,
+      "step": 774
+    },
+    {
+      "epoch": 0.18715286162762618,
+      "grad_norm": 0.28493398427963257,
+      "learning_rate": 9.299487511009525e-05,
+      "loss": 1.8927,
+      "step": 775
+    },
+    {
+      "epoch": 0.18739434919101666,
+      "grad_norm": 0.3119465112686157,
+      "learning_rate": 9.297510275829444e-05,
+      "loss": 1.8846,
+      "step": 776
+    },
+    {
+      "epoch": 0.18763583675440715,
+      "grad_norm": 0.29498162865638733,
+      "learning_rate": 9.29553046496547e-05,
+      "loss": 1.7002,
+      "step": 777
+    },
+    {
+      "epoch": 0.18787732431779763,
+      "grad_norm": 0.28077542781829834,
+      "learning_rate": 9.293548079604189e-05,
+      "loss": 1.7079,
+      "step": 778
+    },
+    {
+      "epoch": 0.18811881188118812,
+      "grad_norm": 0.2838478982448578,
+      "learning_rate": 9.291563120933731e-05,
+      "loss": 1.6531,
+      "step": 779
+    },
+    {
+      "epoch": 0.1883602994445786,
+      "grad_norm": 0.29093652963638306,
+      "learning_rate": 9.289575590143766e-05,
+      "loss": 1.6198,
+      "step": 780
+    },
+    {
+      "epoch": 0.1886017870079691,
+      "grad_norm": 0.28098905086517334,
+      "learning_rate": 9.28758548842551e-05,
+      "loss": 1.7282,
+      "step": 781
+    },
+    {
+      "epoch": 0.18884327457135958,
+      "grad_norm": 0.31896719336509705,
+      "learning_rate": 9.285592816971716e-05,
+      "loss": 2.0032,
+      "step": 782
+    },
+    {
+      "epoch": 0.18908476213475006,
+      "grad_norm": 0.2947934865951538,
+      "learning_rate": 9.283597576976677e-05,
+      "loss": 1.8173,
+      "step": 783
+    },
+    {
+      "epoch": 0.18932624969814055,
+      "grad_norm": 0.2865394949913025,
+      "learning_rate": 9.281599769636226e-05,
+      "loss": 1.7014,
+      "step": 784
+    },
+    {
+      "epoch": 0.18956773726153103,
+      "grad_norm": 0.2883725166320801,
+      "learning_rate": 9.27959939614774e-05,
+      "loss": 1.9053,
+      "step": 785
+    },
+    {
+      "epoch": 0.18980922482492152,
+      "grad_norm": 0.3068401515483856,
+      "learning_rate": 9.277596457710125e-05,
+      "loss": 1.7265,
+      "step": 786
+    },
+    {
+      "epoch": 0.190050712388312,
+      "grad_norm": 0.29906126856803894,
+      "learning_rate": 9.275590955523831e-05,
+      "loss": 1.7202,
+      "step": 787
+    },
+    {
+      "epoch": 0.1902921999517025,
+      "grad_norm": 0.2799694240093231,
+      "learning_rate": 9.273582890790842e-05,
+      "loss": 1.7297,
+      "step": 788
+    },
+    {
+      "epoch": 0.19053368751509298,
+      "grad_norm": 0.29376327991485596,
+      "learning_rate": 9.271572264714675e-05,
+      "loss": 1.7631,
+      "step": 789
+    },
+    {
+      "epoch": 0.19077517507848346,
+      "grad_norm": 0.29661813378334045,
+      "learning_rate": 9.269559078500389e-05,
+      "loss": 1.7929,
+      "step": 790
+    },
+    {
+      "epoch": 0.19101666264187395,
+      "grad_norm": 0.28860270977020264,
+      "learning_rate": 9.267543333354571e-05,
+      "loss": 1.8858,
+      "step": 791
+    },
+    {
+      "epoch": 0.19125815020526443,
+      "grad_norm": 0.29066187143325806,
+      "learning_rate": 9.265525030485348e-05,
+      "loss": 1.7584,
+      "step": 792
+    },
+    {
+      "epoch": 0.19149963776865492,
+      "grad_norm": 0.29545891284942627,
+      "learning_rate": 9.263504171102372e-05,
+      "loss": 1.6921,
+      "step": 793
+    },
+    {
+      "epoch": 0.1917411253320454,
+      "grad_norm": 0.3049885034561157,
+      "learning_rate": 9.261480756416835e-05,
+      "loss": 1.778,
+      "step": 794
+    },
+    {
+      "epoch": 0.1919826128954359,
+      "grad_norm": 0.31124168634414673,
+      "learning_rate": 9.259454787641454e-05,
+      "loss": 1.9121,
+      "step": 795
+    },
+    {
+      "epoch": 0.19222410045882637,
+      "grad_norm": 0.3002880811691284,
+      "learning_rate": 9.257426265990481e-05,
+      "loss": 1.8388,
+      "step": 796
+    },
+    {
+      "epoch": 0.19246558802221686,
+      "grad_norm": 0.32877475023269653,
+      "learning_rate": 9.255395192679697e-05,
+      "loss": 2.0557,
+      "step": 797
+    },
+    {
+      "epoch": 0.19270707558560735,
+      "grad_norm": 0.28469905257225037,
+      "learning_rate": 9.253361568926413e-05,
+      "loss": 1.8209,
+      "step": 798
+    },
+    {
+      "epoch": 0.19294856314899783,
+      "grad_norm": 0.2947460114955902,
+      "learning_rate": 9.251325395949466e-05,
+      "loss": 1.7037,
+      "step": 799
+    },
+    {
+      "epoch": 0.19319005071238832,
+      "grad_norm": 0.2917262017726898,
+      "learning_rate": 9.249286674969224e-05,
+      "loss": 1.7924,
+      "step": 800
+    },
+    {
+      "epoch": 0.1934315382757788,
+      "grad_norm": 0.2889077961444855,
+      "learning_rate": 9.247245407207579e-05,
+      "loss": 1.7546,
+      "step": 801
+    },
+    {
+      "epoch": 0.1936730258391693,
+      "grad_norm": 0.3094068169593811,
+      "learning_rate": 9.245201593887952e-05,
+      "loss": 1.8725,
+      "step": 802
+    },
+    {
+      "epoch": 0.19391451340255977,
+      "grad_norm": 0.29328641295433044,
+      "learning_rate": 9.24315523623529e-05,
+      "loss": 1.8439,
+      "step": 803
+    },
+    {
+      "epoch": 0.19415600096595026,
+      "grad_norm": 0.2785046696662903,
+      "learning_rate": 9.241106335476061e-05,
+      "loss": 1.6542,
+      "step": 804
+    },
+    {
+      "epoch": 0.19439748852934075,
+      "grad_norm": 0.2787037193775177,
+      "learning_rate": 9.239054892838264e-05,
+      "loss": 1.675,
+      "step": 805
+    },
+    {
+      "epoch": 0.19463897609273123,
+      "grad_norm": 0.2860700488090515,
+      "learning_rate": 9.237000909551415e-05,
+      "loss": 1.7113,
+      "step": 806
+    },
+    {
+      "epoch": 0.19488046365612172,
+      "grad_norm": 0.2860807776451111,
+      "learning_rate": 9.234944386846552e-05,
+      "loss": 1.7551,
+      "step": 807
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 0.2986462414264679,
+      "learning_rate": 9.232885325956243e-05,
+      "loss": 1.7758,
+      "step": 808
+    },
+    {
+      "epoch": 0.1953634387829027,
+      "grad_norm": 0.9163442254066467,
+      "learning_rate": 9.23082372811457e-05,
+      "loss": 2.1011,
+      "step": 809
+    },
+    {
+      "epoch": 0.19560492634629317,
+      "grad_norm": 0.3197281062602997,
+      "learning_rate": 9.22875959455714e-05,
+      "loss": 1.9184,
+      "step": 810
+    },
+    {
+      "epoch": 0.19584641390968366,
+      "grad_norm": 0.288152813911438,
+      "learning_rate": 9.226692926521072e-05,
+      "loss": 1.5923,
+      "step": 811
+    },
+    {
+      "epoch": 0.19608790147307414,
+      "grad_norm": 0.3113833963871002,
+      "learning_rate": 9.224623725245016e-05,
+      "loss": 1.7094,
+      "step": 812
+    },
+    {
+      "epoch": 0.19632938903646463,
+      "grad_norm": 0.28525310754776,
+      "learning_rate": 9.22255199196913e-05,
+      "loss": 1.7011,
+      "step": 813
+    },
+    {
+      "epoch": 0.19657087659985512,
+      "grad_norm": 0.28587856888771057,
+      "learning_rate": 9.220477727935095e-05,
+      "loss": 1.5106,
+      "step": 814
+    },
+    {
+      "epoch": 0.1968123641632456,
+      "grad_norm": 0.324881911277771,
+      "learning_rate": 9.218400934386106e-05,
+      "loss": 1.934,
+      "step": 815
+    },
+    {
+      "epoch": 0.1970538517266361,
+      "grad_norm": 0.28722596168518066,
+      "learning_rate": 9.216321612566879e-05,
+      "loss": 1.8429,
+      "step": 816
+    },
+    {
+      "epoch": 0.19729533929002657,
+      "grad_norm": 0.298000693321228,
+      "learning_rate": 9.214239763723636e-05,
+      "loss": 1.6637,
+      "step": 817
+    },
+    {
+      "epoch": 0.19753682685341706,
+      "grad_norm": 0.29165783524513245,
+      "learning_rate": 9.21215538910412e-05,
+      "loss": 1.7989,
+      "step": 818
+    },
+    {
+      "epoch": 0.19777831441680754,
+      "grad_norm": 0.28170880675315857,
+      "learning_rate": 9.210068489957591e-05,
+      "loss": 1.6985,
+      "step": 819
+    },
+    {
+      "epoch": 0.19801980198019803,
+      "grad_norm": 0.29324185848236084,
+      "learning_rate": 9.207979067534815e-05,
+      "loss": 1.7398,
+      "step": 820
+    },
+    {
+      "epoch": 0.19826128954358851,
+      "grad_norm": 0.29249295592308044,
+      "learning_rate": 9.205887123088075e-05,
+      "loss": 1.7616,
+      "step": 821
+    },
+    {
+      "epoch": 0.198502777106979,
+      "grad_norm": 0.2885819375514984,
+      "learning_rate": 9.203792657871163e-05,
+      "loss": 1.6781,
+      "step": 822
+    },
+    {
+      "epoch": 0.19874426467036949,
+      "grad_norm": 0.2888461947441101,
+      "learning_rate": 9.201695673139381e-05,
+      "loss": 1.7416,
+      "step": 823
+    },
+    {
+      "epoch": 0.19898575223375997,
+      "grad_norm": 0.27995526790618896,
+      "learning_rate": 9.199596170149543e-05,
+      "loss": 1.7762,
+      "step": 824
+    },
+    {
+      "epoch": 0.19922723979715046,
+      "grad_norm": 0.30026763677597046,
+      "learning_rate": 9.197494150159976e-05,
+      "loss": 1.8803,
+      "step": 825
+    },
+    {
+      "epoch": 0.19946872736054094,
+      "grad_norm": 0.2981962561607361,
+      "learning_rate": 9.195389614430508e-05,
+      "loss": 1.6778,
+      "step": 826
+    },
+    {
+      "epoch": 0.19971021492393143,
+      "grad_norm": 0.29537999629974365,
+      "learning_rate": 9.193282564222477e-05,
+      "loss": 1.6503,
+      "step": 827
+    },
+    {
+      "epoch": 0.19995170248732191,
+      "grad_norm": 0.3164536654949188,
+      "learning_rate": 9.191173000798732e-05,
+      "loss": 1.9035,
+      "step": 828
+    },
+    {
+      "epoch": 0.2001931900507124,
+      "grad_norm": 0.36678019165992737,
+      "learning_rate": 9.189060925423624e-05,
+      "loss": 2.0678,
+      "step": 829
+    },
+    {
+      "epoch": 0.20043467761410289,
+      "grad_norm": 0.29730290174484253,
+      "learning_rate": 9.186946339363013e-05,
+      "loss": 1.9599,
+      "step": 830
+    },
+    {
+      "epoch": 0.20067616517749337,
+      "grad_norm": 0.2871508002281189,
+      "learning_rate": 9.184829243884264e-05,
+      "loss": 1.748,
+      "step": 831
+    },
+    {
+      "epoch": 0.20091765274088386,
+      "grad_norm": 0.2855934500694275,
+      "learning_rate": 9.182709640256238e-05,
+      "loss": 1.7618,
+      "step": 832
+    },
+    {
+      "epoch": 0.20115914030427434,
+      "grad_norm": 0.2966662049293518,
+      "learning_rate": 9.180587529749309e-05,
+      "loss": 1.7395,
+      "step": 833
+    },
+    {
+      "epoch": 0.20140062786766483,
+      "grad_norm": 0.29292529821395874,
+      "learning_rate": 9.178462913635349e-05,
+      "loss": 1.7003,
+      "step": 834
+    },
+    {
+      "epoch": 0.2016421154310553,
+      "grad_norm": 0.29136964678764343,
+      "learning_rate": 9.176335793187732e-05,
+      "loss": 1.8498,
+      "step": 835
+    },
+    {
+      "epoch": 0.2018836029944458,
+      "grad_norm": 0.3045142590999603,
+      "learning_rate": 9.174206169681334e-05,
+      "loss": 1.6792,
+      "step": 836
+    },
+    {
+      "epoch": 0.20212509055783628,
+      "grad_norm": 0.3062690198421478,
+      "learning_rate": 9.172074044392529e-05,
+      "loss": 1.8017,
+      "step": 837
+    },
+    {
+      "epoch": 0.20236657812122677,
+      "grad_norm": 0.27704912424087524,
+      "learning_rate": 9.169939418599194e-05,
+      "loss": 1.6993,
+      "step": 838
+    },
+    {
+      "epoch": 0.20260806568461726,
+      "grad_norm": 0.29494112730026245,
+      "learning_rate": 9.167802293580703e-05,
+      "loss": 1.6878,
+      "step": 839
+    },
+    {
+      "epoch": 0.20284955324800774,
+      "grad_norm": 0.3410157561302185,
+      "learning_rate": 9.165662670617925e-05,
+      "loss": 2.0581,
+      "step": 840
+    },
+    {
+      "epoch": 0.20309104081139823,
+      "grad_norm": 0.28066882491111755,
+      "learning_rate": 9.16352055099323e-05,
+      "loss": 1.7619,
+      "step": 841
+    },
+    {
+      "epoch": 0.20333252837478868,
+      "grad_norm": 0.27249637246131897,
+      "learning_rate": 9.161375935990482e-05,
+      "loss": 1.5283,
+      "step": 842
+    },
+    {
+      "epoch": 0.20357401593817917,
+      "grad_norm": 0.3004692494869232,
+      "learning_rate": 9.159228826895043e-05,
+      "loss": 1.8467,
+      "step": 843
+    },
+    {
+      "epoch": 0.20381550350156966,
+      "grad_norm": 0.2757605016231537,
+      "learning_rate": 9.15707922499377e-05,
+      "loss": 1.6736,
+      "step": 844
+    },
+    {
+      "epoch": 0.20405699106496014,
+      "grad_norm": 0.281103253364563,
+      "learning_rate": 9.154927131575009e-05,
+      "loss": 1.7212,
+      "step": 845
+    },
+    {
+      "epoch": 0.20429847862835063,
+      "grad_norm": 0.28760358691215515,
+      "learning_rate": 9.152772547928601e-05,
+      "loss": 1.6699,
+      "step": 846
+    },
+    {
+      "epoch": 0.2045399661917411,
+      "grad_norm": 0.2798704504966736,
+      "learning_rate": 9.150615475345888e-05,
+      "loss": 1.6742,
+      "step": 847
+    },
+    {
+      "epoch": 0.2047814537551316,
+      "grad_norm": 0.29464927315711975,
+      "learning_rate": 9.148455915119691e-05,
+      "loss": 1.8128,
+      "step": 848
+    },
+    {
+      "epoch": 0.20502294131852208,
+      "grad_norm": 0.30707064270973206,
+      "learning_rate": 9.146293868544333e-05,
+      "loss": 1.916,
+      "step": 849
+    },
+    {
+      "epoch": 0.20526442888191257,
+      "grad_norm": 0.2912966012954712,
+      "learning_rate": 9.144129336915616e-05,
+      "loss": 1.7689,
+      "step": 850
+    },
+    {
+      "epoch": 0.20550591644530306,
+      "grad_norm": 0.28413230180740356,
+      "learning_rate": 9.141962321530842e-05,
+      "loss": 1.7297,
+      "step": 851
+    },
+    {
+      "epoch": 0.20574740400869354,
+      "grad_norm": 0.2876090109348297,
+      "learning_rate": 9.139792823688797e-05,
+      "loss": 1.6406,
+      "step": 852
+    },
+    {
+      "epoch": 0.20598889157208403,
+      "grad_norm": 0.3216966390609741,
+      "learning_rate": 9.137620844689752e-05,
+      "loss": 2.0118,
+      "step": 853
+    },
+    {
+      "epoch": 0.2062303791354745,
+      "grad_norm": 0.28935253620147705,
+      "learning_rate": 9.135446385835471e-05,
+      "loss": 1.9161,
+      "step": 854
+    },
+    {
+      "epoch": 0.206471866698865,
+      "grad_norm": 0.2942180931568146,
+      "learning_rate": 9.133269448429199e-05,
+      "loss": 1.7412,
+      "step": 855
+    },
+    {
+      "epoch": 0.20671335426225548,
+      "grad_norm": 0.3060719072818756,
+      "learning_rate": 9.13109003377567e-05,
+      "loss": 1.9438,
+      "step": 856
+    },
+    {
+      "epoch": 0.20695484182564597,
+      "grad_norm": 0.2826623320579529,
+      "learning_rate": 9.128908143181101e-05,
+      "loss": 1.7717,
+      "step": 857
+    },
+    {
+      "epoch": 0.20719632938903645,
+      "grad_norm": 0.30230000615119934,
+      "learning_rate": 9.126723777953195e-05,
+      "loss": 1.8519,
+      "step": 858
+    },
+    {
+      "epoch": 0.20743781695242694,
+      "grad_norm": 0.305400013923645,
+      "learning_rate": 9.124536939401135e-05,
+      "loss": 1.7981,
+      "step": 859
+    },
+    {
+      "epoch": 0.20767930451581743,
+      "grad_norm": 0.2971997559070587,
+      "learning_rate": 9.122347628835589e-05,
+      "loss": 1.9272,
+      "step": 860
+    },
+    {
+      "epoch": 0.2079207920792079,
+      "grad_norm": 0.2872750461101532,
+      "learning_rate": 9.120155847568704e-05,
+      "loss": 1.5787,
+      "step": 861
+    },
+    {
+      "epoch": 0.2081622796425984,
+      "grad_norm": 0.2825159728527069,
+      "learning_rate": 9.117961596914109e-05,
+      "loss": 1.729,
+      "step": 862
+    },
+    {
+      "epoch": 0.20840376720598888,
+      "grad_norm": 0.2875017821788788,
+      "learning_rate": 9.115764878186917e-05,
+      "loss": 1.7238,
+      "step": 863
+    },
+    {
+      "epoch": 0.20864525476937937,
+      "grad_norm": 0.2908998727798462,
+      "learning_rate": 9.113565692703714e-05,
+      "loss": 1.6799,
+      "step": 864
+    },
+    {
+      "epoch": 0.20888674233276985,
+      "grad_norm": 0.28806325793266296,
+      "learning_rate": 9.111364041782566e-05,
+      "loss": 1.7153,
+      "step": 865
+    },
+    {
+      "epoch": 0.20912822989616034,
+      "grad_norm": 0.29157885909080505,
+      "learning_rate": 9.10915992674302e-05,
+      "loss": 1.6363,
+      "step": 866
+    },
+    {
+      "epoch": 0.20936971745955082,
+      "grad_norm": 0.29479745030403137,
+      "learning_rate": 9.106953348906097e-05,
+      "loss": 1.9421,
+      "step": 867
+    },
+    {
+      "epoch": 0.2096112050229413,
+      "grad_norm": 0.3266473412513733,
+      "learning_rate": 9.104744309594294e-05,
+      "loss": 1.9736,
+      "step": 868
+    },
+    {
+      "epoch": 0.2098526925863318,
+      "grad_norm": 0.2955428659915924,
+      "learning_rate": 9.102532810131584e-05,
+      "loss": 1.945,
+      "step": 869
+    },
+    {
+      "epoch": 0.21009418014972228,
+      "grad_norm": 0.29601460695266724,
+      "learning_rate": 9.100318851843415e-05,
+      "loss": 1.8434,
+      "step": 870
+    },
+    {
+      "epoch": 0.21033566771311277,
+      "grad_norm": 0.2987689673900604,
+      "learning_rate": 9.098102436056706e-05,
+      "loss": 1.9485,
+      "step": 871
+    },
+    {
+      "epoch": 0.21057715527650325,
+      "grad_norm": 0.27123886346817017,
+      "learning_rate": 9.095883564099854e-05,
+      "loss": 1.6151,
+      "step": 872
+    },
+    {
+      "epoch": 0.21081864283989374,
+      "grad_norm": 0.28310680389404297,
+      "learning_rate": 9.093662237302722e-05,
+      "loss": 1.731,
+      "step": 873
+    },
+    {
+      "epoch": 0.21106013040328422,
+      "grad_norm": 0.281590074300766,
+      "learning_rate": 9.091438456996648e-05,
+      "loss": 1.7374,
+      "step": 874
+    },
+    {
+      "epoch": 0.2113016179666747,
+      "grad_norm": 0.2950079143047333,
+      "learning_rate": 9.089212224514441e-05,
+      "loss": 1.7526,
+      "step": 875
+    },
+    {
+      "epoch": 0.2115431055300652,
+      "grad_norm": 0.2879007160663605,
+      "learning_rate": 9.086983541190379e-05,
+      "loss": 1.7355,
+      "step": 876
+    },
+    {
+      "epoch": 0.21178459309345568,
+      "grad_norm": 0.2673324644565582,
+      "learning_rate": 9.084752408360206e-05,
+      "loss": 1.6847,
+      "step": 877
+    },
+    {
+      "epoch": 0.21202608065684617,
+      "grad_norm": 0.2840292751789093,
+      "learning_rate": 9.082518827361137e-05,
+      "loss": 1.6642,
+      "step": 878
+    },
+    {
+      "epoch": 0.21226756822023665,
+      "grad_norm": 0.2851640582084656,
+      "learning_rate": 9.080282799531855e-05,
+      "loss": 1.6895,
+      "step": 879
+    },
+    {
+      "epoch": 0.21250905578362714,
+      "grad_norm": 0.30625519156455994,
+      "learning_rate": 9.078044326212507e-05,
+      "loss": 1.8807,
+      "step": 880
+    },
+    {
+      "epoch": 0.21275054334701762,
+      "grad_norm": 0.2955930233001709,
+      "learning_rate": 9.07580340874471e-05,
+      "loss": 1.8051,
+      "step": 881
+    },
+    {
+      "epoch": 0.2129920309104081,
+      "grad_norm": 0.3044041395187378,
+      "learning_rate": 9.073560048471538e-05,
+      "loss": 1.8127,
+      "step": 882
+    },
+    {
+      "epoch": 0.2132335184737986,
+      "grad_norm": 0.26778867840766907,
+      "learning_rate": 9.071314246737536e-05,
+      "loss": 1.6614,
+      "step": 883
+    },
+    {
+      "epoch": 0.21347500603718908,
+      "grad_norm": 0.28641968965530396,
+      "learning_rate": 9.069066004888712e-05,
+      "loss": 1.8822,
+      "step": 884
+    },
+    {
+      "epoch": 0.21371649360057957,
+      "grad_norm": 0.2820203900337219,
+      "learning_rate": 9.066815324272533e-05,
+      "loss": 1.625,
+      "step": 885
+    },
+    {
+      "epoch": 0.21395798116397005,
+      "grad_norm": 0.2761995792388916,
+      "learning_rate": 9.064562206237929e-05,
+      "loss": 1.6245,
+      "step": 886
+    },
+    {
+      "epoch": 0.21419946872736054,
+      "grad_norm": 0.2768464684486389,
+      "learning_rate": 9.062306652135292e-05,
+      "loss": 1.7488,
+      "step": 887
+    },
+    {
+      "epoch": 0.21444095629075102,
+      "grad_norm": 0.2869665026664734,
+      "learning_rate": 9.060048663316472e-05,
+      "loss": 1.7725,
+      "step": 888
+    },
+    {
+      "epoch": 0.2146824438541415,
+      "grad_norm": 0.28295812010765076,
+      "learning_rate": 9.05778824113478e-05,
+      "loss": 1.724,
+      "step": 889
+    },
+    {
+      "epoch": 0.214923931417532,
+      "grad_norm": 0.2879597246646881,
+      "learning_rate": 9.055525386944985e-05,
+      "loss": 1.7923,
+      "step": 890
+    },
+    {
+      "epoch": 0.21516541898092248,
+      "grad_norm": 0.2905023396015167,
+      "learning_rate": 9.053260102103315e-05,
+      "loss": 1.6549,
+      "step": 891
+    },
+    {
+      "epoch": 0.21540690654431296,
+      "grad_norm": 0.2903376817703247,
+      "learning_rate": 9.05099238796745e-05,
+      "loss": 1.7865,
+      "step": 892
+    },
+    {
+      "epoch": 0.21564839410770345,
+      "grad_norm": 0.2778032422065735,
+      "learning_rate": 9.048722245896531e-05,
+      "loss": 1.6382,
+      "step": 893
+    },
+    {
+      "epoch": 0.21588988167109394,
+      "grad_norm": 0.3031550943851471,
+      "learning_rate": 9.046449677251155e-05,
+      "loss": 1.7798,
+      "step": 894
+    },
+    {
+      "epoch": 0.21613136923448442,
+      "grad_norm": 0.2920134365558624,
+      "learning_rate": 9.044174683393365e-05,
+      "loss": 1.7833,
+      "step": 895
+    },
+    {
+      "epoch": 0.2163728567978749,
+      "grad_norm": 0.27324342727661133,
+      "learning_rate": 9.041897265686666e-05,
+      "loss": 1.6678,
+      "step": 896
+    },
+    {
+      "epoch": 0.2166143443612654,
+      "grad_norm": 0.289196640253067,
+      "learning_rate": 9.039617425496015e-05,
+      "loss": 1.7339,
+      "step": 897
+    },
+    {
+      "epoch": 0.21685583192465588,
+      "grad_norm": 0.26956117153167725,
+      "learning_rate": 9.037335164187817e-05,
+      "loss": 1.7743,
+      "step": 898
+    },
+    {
+      "epoch": 0.21709731948804636,
+      "grad_norm": 0.2876354455947876,
+      "learning_rate": 9.035050483129928e-05,
+      "loss": 1.7449,
+      "step": 899
+    },
+    {
+      "epoch": 0.21733880705143685,
+      "grad_norm": 0.29619812965393066,
+      "learning_rate": 9.032763383691662e-05,
+      "loss": 1.8701,
+      "step": 900
+    },
+    {
+      "epoch": 0.21758029461482734,
+      "grad_norm": 0.27593353390693665,
+      "learning_rate": 9.03047386724377e-05,
+      "loss": 1.6191,
+      "step": 901
+    },
+    {
+      "epoch": 0.21782178217821782,
+      "grad_norm": 0.2756110727787018,
+      "learning_rate": 9.028181935158463e-05,
+      "loss": 1.7214,
+      "step": 902
+    },
+    {
+      "epoch": 0.2180632697416083,
+      "grad_norm": 0.29085442423820496,
+      "learning_rate": 9.025887588809394e-05,
+      "loss": 1.8983,
+      "step": 903
+    },
+    {
+      "epoch": 0.2183047573049988,
+      "grad_norm": 0.30053767561912537,
+      "learning_rate": 9.023590829571663e-05,
+      "loss": 2.027,
+      "step": 904
+    },
+    {
+      "epoch": 0.21854624486838928,
+      "grad_norm": 0.27619075775146484,
+      "learning_rate": 9.021291658821815e-05,
+      "loss": 1.6083,
+      "step": 905
+    },
+    {
+      "epoch": 0.21878773243177976,
+      "grad_norm": 0.2857961356639862,
+      "learning_rate": 9.018990077937848e-05,
+      "loss": 1.7759,
+      "step": 906
+    },
+    {
+      "epoch": 0.21902921999517025,
+      "grad_norm": 0.2999686598777771,
+      "learning_rate": 9.016686088299196e-05,
+      "loss": 1.7725,
+      "step": 907
+    },
+    {
+      "epoch": 0.21927070755856073,
+      "grad_norm": 0.2941741943359375,
+      "learning_rate": 9.01437969128674e-05,
+      "loss": 1.7253,
+      "step": 908
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 0.2836509644985199,
+      "learning_rate": 9.012070888282803e-05,
+      "loss": 1.7763,
+      "step": 909
+    },
+    {
+      "epoch": 0.2197536826853417,
+      "grad_norm": 0.2917303740978241,
+      "learning_rate": 9.009759680671151e-05,
+      "loss": 1.6533,
+      "step": 910
+    },
+    {
+      "epoch": 0.2199951702487322,
+      "grad_norm": 0.27767467498779297,
+      "learning_rate": 9.007446069836991e-05,
+      "loss": 1.7,
+      "step": 911
+    },
+    {
+      "epoch": 0.22023665781212268,
+      "grad_norm": 0.27494391798973083,
+      "learning_rate": 9.005130057166971e-05,
+      "loss": 1.6589,
+      "step": 912
+    },
+    {
+      "epoch": 0.22047814537551316,
+      "grad_norm": 0.3042466342449188,
+      "learning_rate": 9.002811644049176e-05,
+      "loss": 1.8579,
+      "step": 913
+    },
+    {
+      "epoch": 0.22071963293890365,
+      "grad_norm": 0.308646023273468,
+      "learning_rate": 9.00049083187313e-05,
+      "loss": 1.911,
+      "step": 914
+    },
+    {
+      "epoch": 0.22096112050229413,
+      "grad_norm": 0.2836478054523468,
+      "learning_rate": 8.998167622029801e-05,
+      "loss": 1.8678,
+      "step": 915
+    },
+    {
+      "epoch": 0.22120260806568462,
+      "grad_norm": 0.3099924325942993,
+      "learning_rate": 8.995842015911584e-05,
+      "loss": 1.8014,
+      "step": 916
+    },
+    {
+      "epoch": 0.2214440956290751,
+      "grad_norm": 0.2966037094593048,
+      "learning_rate": 8.993514014912318e-05,
+      "loss": 1.8265,
+      "step": 917
+    },
+    {
+      "epoch": 0.2216855831924656,
+      "grad_norm": 0.2903251051902771,
+      "learning_rate": 8.991183620427275e-05,
+      "loss": 1.7539,
+      "step": 918
+    },
+    {
+      "epoch": 0.22192707075585608,
+      "grad_norm": 0.28033512830734253,
+      "learning_rate": 8.98885083385316e-05,
+      "loss": 1.5729,
+      "step": 919
+    },
+    {
+      "epoch": 0.22216855831924656,
+      "grad_norm": 0.29812338948249817,
+      "learning_rate": 8.986515656588114e-05,
+      "loss": 1.6498,
+      "step": 920
+    },
+    {
+      "epoch": 0.22241004588263705,
+      "grad_norm": 0.29040244221687317,
+      "learning_rate": 8.98417809003171e-05,
+      "loss": 1.8032,
+      "step": 921
+    },
+    {
+      "epoch": 0.22265153344602753,
+      "grad_norm": 0.2998603880405426,
+      "learning_rate": 8.981838135584953e-05,
+      "loss": 1.7509,
+      "step": 922
+    },
+    {
+      "epoch": 0.22289302100941802,
+      "grad_norm": 0.2951640188694,
+      "learning_rate": 8.979495794650276e-05,
+      "loss": 1.8133,
+      "step": 923
+    },
+    {
+      "epoch": 0.2231345085728085,
+      "grad_norm": 0.2719232439994812,
+      "learning_rate": 8.977151068631548e-05,
+      "loss": 1.6761,
+      "step": 924
+    },
+    {
+      "epoch": 0.223375996136199,
+      "grad_norm": 0.29052817821502686,
+      "learning_rate": 8.974803958934064e-05,
+      "loss": 1.7221,
+      "step": 925
+    },
+    {
+      "epoch": 0.22361748369958948,
+      "grad_norm": 0.2881939113140106,
+      "learning_rate": 8.972454466964551e-05,
+      "loss": 1.6036,
+      "step": 926
+    },
+    {
+      "epoch": 0.22385897126297996,
+      "grad_norm": 0.2801560163497925,
+      "learning_rate": 8.970102594131156e-05,
+      "loss": 1.59,
+      "step": 927
+    },
+    {
+      "epoch": 0.22410045882637045,
+      "grad_norm": 0.289957195520401,
+      "learning_rate": 8.967748341843462e-05,
+      "loss": 1.9019,
+      "step": 928
+    },
+    {
+      "epoch": 0.22434194638976093,
+      "grad_norm": 0.29187458753585815,
+      "learning_rate": 8.965391711512475e-05,
+      "loss": 1.6741,
+      "step": 929
+    },
+    {
+      "epoch": 0.22458343395315142,
+      "grad_norm": 0.2863827049732208,
+      "learning_rate": 8.96303270455062e-05,
+      "loss": 1.8284,
+      "step": 930
+    },
+    {
+      "epoch": 0.2248249215165419,
+      "grad_norm": 0.29024264216423035,
+      "learning_rate": 8.960671322371759e-05,
+      "loss": 1.7697,
+      "step": 931
+    },
+    {
+      "epoch": 0.2250664090799324,
+      "grad_norm": 0.273242712020874,
+      "learning_rate": 8.958307566391166e-05,
+      "loss": 1.7714,
+      "step": 932
+    },
+    {
+      "epoch": 0.22530789664332287,
+      "grad_norm": 0.298308789730072,
+      "learning_rate": 8.955941438025545e-05,
+      "loss": 1.7871,
+      "step": 933
+    },
+    {
+      "epoch": 0.22554938420671336,
+      "grad_norm": 0.28436291217803955,
+      "learning_rate": 8.953572938693016e-05,
+      "loss": 1.7225,
+      "step": 934
+    },
+    {
+      "epoch": 0.22579087177010385,
+      "grad_norm": 0.2844787836074829,
+      "learning_rate": 8.951202069813126e-05,
+      "loss": 1.6326,
+      "step": 935
+    },
+    {
+      "epoch": 0.22603235933349433,
+      "grad_norm": 0.28223082423210144,
+      "learning_rate": 8.948828832806838e-05,
+      "loss": 1.8086,
+      "step": 936
+    },
+    {
+      "epoch": 0.22627384689688482,
+      "grad_norm": 0.3171670734882355,
+      "learning_rate": 8.946453229096537e-05,
+      "loss": 1.7749,
+      "step": 937
+    },
+    {
+      "epoch": 0.2265153344602753,
+      "grad_norm": 0.27827945351600647,
+      "learning_rate": 8.944075260106022e-05,
+      "loss": 1.7798,
+      "step": 938
+    },
+    {
+      "epoch": 0.2267568220236658,
+      "grad_norm": 0.3321087062358856,
+      "learning_rate": 8.941694927260516e-05,
+      "loss": 2.0057,
+      "step": 939
+    },
+    {
+      "epoch": 0.22699830958705627,
+      "grad_norm": 0.3078181743621826,
+      "learning_rate": 8.939312231986655e-05,
+      "loss": 1.7606,
+      "step": 940
+    },
+    {
+      "epoch": 0.22723979715044676,
+      "grad_norm": 0.3000055253505707,
+      "learning_rate": 8.936927175712488e-05,
+      "loss": 1.8077,
+      "step": 941
+    },
+    {
+      "epoch": 0.22748128471383725,
+      "grad_norm": 0.3055233955383301,
+      "learning_rate": 8.934539759867487e-05,
+      "loss": 1.8196,
+      "step": 942
+    },
+    {
+      "epoch": 0.22772277227722773,
+      "grad_norm": 0.30723288655281067,
+      "learning_rate": 8.932149985882529e-05,
+      "loss": 1.7099,
+      "step": 943
+    },
+    {
+      "epoch": 0.22796425984061822,
+      "grad_norm": 0.2921900153160095,
+      "learning_rate": 8.929757855189915e-05,
+      "loss": 1.7364,
+      "step": 944
+    },
+    {
+      "epoch": 0.2282057474040087,
+      "grad_norm": 0.3061923384666443,
+      "learning_rate": 8.927363369223348e-05,
+      "loss": 1.6376,
+      "step": 945
+    },
+    {
+      "epoch": 0.2284472349673992,
+      "grad_norm": 0.2868843078613281,
+      "learning_rate": 8.924966529417949e-05,
+      "loss": 1.7675,
+      "step": 946
+    },
+    {
+      "epoch": 0.22868872253078967,
+      "grad_norm": 0.2890067994594574,
+      "learning_rate": 8.922567337210245e-05,
+      "loss": 1.7781,
+      "step": 947
+    },
+    {
+      "epoch": 0.22893021009418016,
+      "grad_norm": 0.30553463101387024,
+      "learning_rate": 8.920165794038177e-05,
+      "loss": 1.8075,
+      "step": 948
+    },
+    {
+      "epoch": 0.22917169765757064,
+      "grad_norm": 0.27772581577301025,
+      "learning_rate": 8.917761901341097e-05,
+      "loss": 1.6267,
+      "step": 949
+    },
+    {
+      "epoch": 0.22941318522096113,
+      "grad_norm": 0.3060512840747833,
+      "learning_rate": 8.915355660559758e-05,
+      "loss": 1.8912,
+      "step": 950
+    },
+    {
+      "epoch": 0.22965467278435162,
+      "grad_norm": 0.30589228868484497,
+      "learning_rate": 8.912947073136327e-05,
+      "loss": 1.8268,
+      "step": 951
+    },
+    {
+      "epoch": 0.2298961603477421,
+      "grad_norm": 0.29312169551849365,
+      "learning_rate": 8.910536140514372e-05,
+      "loss": 1.6905,
+      "step": 952
+    },
+    {
+      "epoch": 0.2301376479111326,
+      "grad_norm": 0.29430586099624634,
+      "learning_rate": 8.908122864138872e-05,
+      "loss": 1.7801,
+      "step": 953
+    },
+    {
+      "epoch": 0.23037913547452307,
+      "grad_norm": 0.28983962535858154,
+      "learning_rate": 8.905707245456206e-05,
+      "loss": 1.8132,
+      "step": 954
+    },
+    {
+      "epoch": 0.23062062303791356,
+      "grad_norm": 0.3049819767475128,
+      "learning_rate": 8.903289285914159e-05,
+      "loss": 1.8283,
+      "step": 955
+    },
+    {
+      "epoch": 0.23086211060130404,
+      "grad_norm": 0.2923794090747833,
+      "learning_rate": 8.900868986961919e-05,
+      "loss": 1.8282,
+      "step": 956
+    },
+    {
+      "epoch": 0.23110359816469453,
+      "grad_norm": 0.2887613773345947,
+      "learning_rate": 8.898446350050076e-05,
+      "loss": 1.7958,
+      "step": 957
+    },
+    {
+      "epoch": 0.23134508572808501,
+      "grad_norm": 0.2974589169025421,
+      "learning_rate": 8.896021376630621e-05,
+      "loss": 1.8724,
+      "step": 958
+    },
+    {
+      "epoch": 0.2315865732914755,
+      "grad_norm": 0.2794170081615448,
+      "learning_rate": 8.893594068156947e-05,
+      "loss": 1.7259,
+      "step": 959
+    },
+    {
+      "epoch": 0.23182806085486599,
+      "grad_norm": 0.2978573441505432,
+      "learning_rate": 8.891164426083842e-05,
+      "loss": 1.7693,
+      "step": 960
+    },
+    {
+      "epoch": 0.23206954841825647,
+      "grad_norm": 0.30293652415275574,
+      "learning_rate": 8.888732451867498e-05,
+      "loss": 1.7213,
+      "step": 961
+    },
+    {
+      "epoch": 0.23231103598164696,
+      "grad_norm": 0.2683996558189392,
+      "learning_rate": 8.886298146965501e-05,
+      "loss": 1.7209,
+      "step": 962
+    },
+    {
+      "epoch": 0.23255252354503744,
+      "grad_norm": 0.2920992970466614,
+      "learning_rate": 8.883861512836839e-05,
+      "loss": 1.5465,
+      "step": 963
+    },
+    {
+      "epoch": 0.23279401110842793,
+      "grad_norm": 0.2909085154533386,
+      "learning_rate": 8.881422550941887e-05,
+      "loss": 1.723,
+      "step": 964
+    },
+    {
+      "epoch": 0.23303549867181841,
+      "grad_norm": 0.28865692019462585,
+      "learning_rate": 8.878981262742423e-05,
+      "loss": 1.8612,
+      "step": 965
+    },
+    {
+      "epoch": 0.2332769862352089,
+      "grad_norm": 0.27150455117225647,
+      "learning_rate": 8.876537649701618e-05,
+      "loss": 1.5894,
+      "step": 966
+    },
+    {
+      "epoch": 0.23351847379859939,
+      "grad_norm": 0.2986278533935547,
+      "learning_rate": 8.87409171328403e-05,
+      "loss": 1.8431,
+      "step": 967
+    },
+    {
+      "epoch": 0.23375996136198987,
+      "grad_norm": 0.29470667243003845,
+      "learning_rate": 8.87164345495562e-05,
+      "loss": 1.9004,
+      "step": 968
+    },
+    {
+      "epoch": 0.23400144892538036,
+      "grad_norm": 0.2897234559059143,
+      "learning_rate": 8.869192876183734e-05,
+      "loss": 1.78,
+      "step": 969
+    },
+    {
+      "epoch": 0.23424293648877084,
+      "grad_norm": 0.29372528195381165,
+      "learning_rate": 8.866739978437109e-05,
+      "loss": 1.8018,
+      "step": 970
+    },
+    {
+      "epoch": 0.2344844240521613,
+      "grad_norm": 0.27670425176620483,
+      "learning_rate": 8.864284763185873e-05,
+      "loss": 1.6661,
+      "step": 971
+    },
+    {
+      "epoch": 0.23472591161555179,
+      "grad_norm": 0.2911158800125122,
+      "learning_rate": 8.861827231901541e-05,
+      "loss": 1.9302,
+      "step": 972
+    },
+    {
+      "epoch": 0.23496739917894227,
+      "grad_norm": 0.27287617325782776,
+      "learning_rate": 8.85936738605702e-05,
+      "loss": 1.6371,
+      "step": 973
+    },
+    {
+      "epoch": 0.23520888674233276,
+      "grad_norm": 0.2705206871032715,
+      "learning_rate": 8.856905227126602e-05,
+      "loss": 1.6918,
+      "step": 974
+    },
+    {
+      "epoch": 0.23545037430572324,
+      "grad_norm": 0.28230130672454834,
+      "learning_rate": 8.854440756585964e-05,
+      "loss": 1.7005,
+      "step": 975
+    },
+    {
+      "epoch": 0.23569186186911373,
+      "grad_norm": 0.2819898724555969,
+      "learning_rate": 8.851973975912173e-05,
+      "loss": 1.7495,
+      "step": 976
+    },
+    {
+      "epoch": 0.2359333494325042,
+      "grad_norm": 0.2814916670322418,
+      "learning_rate": 8.849504886583674e-05,
+      "loss": 1.6623,
+      "step": 977
+    },
+    {
+      "epoch": 0.2361748369958947,
+      "grad_norm": 0.2934419512748718,
+      "learning_rate": 8.847033490080301e-05,
+      "loss": 1.7376,
+      "step": 978
+    },
+    {
+      "epoch": 0.23641632455928518,
+      "grad_norm": 0.3402636647224426,
+      "learning_rate": 8.844559787883266e-05,
+      "loss": 2.0732,
+      "step": 979
+    },
+    {
+      "epoch": 0.23665781212267567,
+      "grad_norm": 0.2801571786403656,
+      "learning_rate": 8.84208378147517e-05,
+      "loss": 1.6666,
+      "step": 980
+    },
+    {
+      "epoch": 0.23689929968606616,
+      "grad_norm": 0.31147944927215576,
+      "learning_rate": 8.839605472339988e-05,
+      "loss": 1.7559,
+      "step": 981
+    },
+    {
+      "epoch": 0.23714078724945664,
+      "grad_norm": 0.2805061936378479,
+      "learning_rate": 8.83712486196308e-05,
+      "loss": 1.7002,
+      "step": 982
+    },
+    {
+      "epoch": 0.23738227481284713,
+      "grad_norm": 0.2966560423374176,
+      "learning_rate": 8.83464195183118e-05,
+      "loss": 1.7956,
+      "step": 983
+    },
+    {
+      "epoch": 0.2376237623762376,
+      "grad_norm": 0.29622846841812134,
+      "learning_rate": 8.832156743432405e-05,
+      "loss": 1.7916,
+      "step": 984
+    },
+    {
+      "epoch": 0.2378652499396281,
+      "grad_norm": 0.29022935032844543,
+      "learning_rate": 8.829669238256251e-05,
+      "loss": 1.7386,
+      "step": 985
+    },
+    {
+      "epoch": 0.23810673750301858,
+      "grad_norm": 0.2820836007595062,
+      "learning_rate": 8.827179437793582e-05,
+      "loss": 1.7125,
+      "step": 986
+    },
+    {
+      "epoch": 0.23834822506640907,
+      "grad_norm": 0.2786775529384613,
+      "learning_rate": 8.824687343536648e-05,
+      "loss": 1.9048,
+      "step": 987
+    },
+    {
+      "epoch": 0.23858971262979956,
+      "grad_norm": 0.273759663105011,
+      "learning_rate": 8.822192956979066e-05,
+      "loss": 1.6711,
+      "step": 988
+    },
+    {
+      "epoch": 0.23883120019319004,
+      "grad_norm": 0.2763570249080658,
+      "learning_rate": 8.81969627961583e-05,
+      "loss": 1.6444,
+      "step": 989
+    },
+    {
+      "epoch": 0.23907268775658053,
+      "grad_norm": 0.2795482277870178,
+      "learning_rate": 8.817197312943308e-05,
+      "loss": 1.7041,
+      "step": 990
+    },
+    {
+      "epoch": 0.239314175319971,
+      "grad_norm": 0.28513190150260925,
+      "learning_rate": 8.81469605845924e-05,
+      "loss": 1.74,
+      "step": 991
+    },
+    {
+      "epoch": 0.2395556628833615,
+      "grad_norm": 0.3679279088973999,
+      "learning_rate": 8.812192517662732e-05,
+      "loss": 1.6976,
+      "step": 992
+    },
+    {
+      "epoch": 0.23979715044675198,
+      "grad_norm": 0.3091084957122803,
+      "learning_rate": 8.809686692054268e-05,
+      "loss": 1.9301,
+      "step": 993
+    },
+    {
+      "epoch": 0.24003863801014247,
+      "grad_norm": 0.3088282644748688,
+      "learning_rate": 8.807178583135697e-05,
+      "loss": 1.8976,
+      "step": 994
+    },
+    {
+      "epoch": 0.24028012557353295,
+      "grad_norm": 0.275089293718338,
+      "learning_rate": 8.804668192410235e-05,
+      "loss": 1.7477,
+      "step": 995
+    },
+    {
+      "epoch": 0.24052161313692344,
+      "grad_norm": 0.2925816476345062,
+      "learning_rate": 8.802155521382471e-05,
+      "loss": 1.9009,
+      "step": 996
+    },
+    {
+      "epoch": 0.24076310070031393,
+      "grad_norm": 0.2990473806858063,
+      "learning_rate": 8.799640571558357e-05,
+      "loss": 1.796,
+      "step": 997
+    },
+    {
+      "epoch": 0.2410045882637044,
+      "grad_norm": 0.2928846776485443,
+      "learning_rate": 8.797123344445208e-05,
+      "loss": 1.7707,
+      "step": 998
+    },
+    {
+      "epoch": 0.2412460758270949,
+      "grad_norm": 0.28296223282814026,
+      "learning_rate": 8.794603841551712e-05,
+      "loss": 1.7324,
+      "step": 999
+    },
+    {
+      "epoch": 0.24148756339048538,
+      "grad_norm": 0.291424036026001,
+      "learning_rate": 8.79208206438791e-05,
+      "loss": 1.9164,
+      "step": 1000
+    },
+    {
+      "epoch": 0.24172905095387587,
+      "grad_norm": 0.2974405884742737,
+      "learning_rate": 8.789558014465218e-05,
+      "loss": 1.8217,
+      "step": 1001
+    },
+    {
+      "epoch": 0.24197053851726635,
+      "grad_norm": 0.2871314287185669,
+      "learning_rate": 8.787031693296407e-05,
+      "loss": 1.7372,
+      "step": 1002
+    },
+    {
+      "epoch": 0.24221202608065684,
+      "grad_norm": 0.28517138957977295,
+      "learning_rate": 8.784503102395609e-05,
+      "loss": 1.7268,
+      "step": 1003
+    },
+    {
+      "epoch": 0.24245351364404732,
+      "grad_norm": 0.2855314612388611,
+      "learning_rate": 8.781972243278319e-05,
+      "loss": 1.7284,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2426950012074378,
+      "grad_norm": 0.30058592557907104,
+      "learning_rate": 8.779439117461388e-05,
+      "loss": 1.676,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2429364887708283,
+      "grad_norm": 0.2805761694908142,
+      "learning_rate": 8.776903726463031e-05,
+      "loss": 1.6762,
+      "step": 1006
+    },
+    {
+      "epoch": 0.24317797633421878,
+      "grad_norm": 0.2890090346336365,
+      "learning_rate": 8.774366071802817e-05,
+      "loss": 1.772,
+      "step": 1007
+    },
+    {
+      "epoch": 0.24341946389760927,
+      "grad_norm": 0.287696897983551,
+      "learning_rate": 8.771826155001673e-05,
+      "loss": 1.8927,
+      "step": 1008
+    },
+    {
+      "epoch": 0.24366095146099975,
+      "grad_norm": 0.28548726439476013,
+      "learning_rate": 8.769283977581878e-05,
+      "loss": 1.7648,
+      "step": 1009
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 0.28890261054039,
+      "learning_rate": 8.766739541067072e-05,
+      "loss": 1.8778,
+      "step": 1010
+    },
+    {
+      "epoch": 0.24414392658778072,
+      "grad_norm": 0.28177410364151,
+      "learning_rate": 8.764192846982244e-05,
+      "loss": 1.7047,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2443854141511712,
+      "grad_norm": 0.2893199026584625,
+      "learning_rate": 8.761643896853739e-05,
+      "loss": 1.6994,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2446269017145617,
+      "grad_norm": 0.2756044566631317,
+      "learning_rate": 8.759092692209251e-05,
+      "loss": 1.7141,
+      "step": 1013
+    },
+    {
+      "epoch": 0.24486838927795218,
+      "grad_norm": 0.29576975107192993,
+      "learning_rate": 8.756539234577834e-05,
+      "loss": 1.7438,
+      "step": 1014
+    },
+    {
+      "epoch": 0.24510987684134267,
+      "grad_norm": 0.31498026847839355,
+      "learning_rate": 8.753983525489878e-05,
+      "loss": 1.9633,
+      "step": 1015
+    },
+    {
+      "epoch": 0.24535136440473315,
+      "grad_norm": 0.28949326276779175,
+      "learning_rate": 8.751425566477131e-05,
+      "loss": 1.7926,
+      "step": 1016
+    },
+    {
+      "epoch": 0.24559285196812364,
+      "grad_norm": 0.2963774502277374,
+      "learning_rate": 8.748865359072694e-05,
+      "loss": 1.7185,
+      "step": 1017
+    },
+    {
+      "epoch": 0.24583433953151412,
+      "grad_norm": 0.27066588401794434,
+      "learning_rate": 8.746302904811004e-05,
+      "loss": 1.7726,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2460758270949046,
+      "grad_norm": 0.29991111159324646,
+      "learning_rate": 8.743738205227854e-05,
+      "loss": 1.9028,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2463173146582951,
+      "grad_norm": 0.3058822453022003,
+      "learning_rate": 8.74117126186038e-05,
+      "loss": 1.7786,
+      "step": 1020
+    },
+    {
+      "epoch": 0.24655880222168558,
+      "grad_norm": 0.28362077474594116,
+      "learning_rate": 8.73860207624706e-05,
+      "loss": 1.6691,
+      "step": 1021
+    },
+    {
+      "epoch": 0.24680028978507607,
+      "grad_norm": 0.29726213216781616,
+      "learning_rate": 8.736030649927717e-05,
+      "loss": 1.7436,
+      "step": 1022
+    },
+    {
+      "epoch": 0.24704177734846655,
+      "grad_norm": 0.278328001499176,
+      "learning_rate": 8.733456984443522e-05,
+      "loss": 1.6813,
+      "step": 1023
+    },
+    {
+      "epoch": 0.24728326491185704,
+      "grad_norm": 0.30932289361953735,
+      "learning_rate": 8.730881081336984e-05,
+      "loss": 1.9074,
+      "step": 1024
+    },
+    {
+      "epoch": 0.24752475247524752,
+      "grad_norm": 0.28863757848739624,
+      "learning_rate": 8.728302942151949e-05,
+      "loss": 1.7997,
+      "step": 1025
+    },
+    {
+      "epoch": 0.247766240038638,
+      "grad_norm": 0.28880390524864197,
+      "learning_rate": 8.725722568433612e-05,
+      "loss": 1.7873,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2480077276020285,
+      "grad_norm": 0.27976545691490173,
+      "learning_rate": 8.723139961728499e-05,
+      "loss": 1.6941,
+      "step": 1027
+    },
+    {
+      "epoch": 0.24824921516541898,
+      "grad_norm": 0.2951839864253998,
+      "learning_rate": 8.72055512358448e-05,
+      "loss": 1.8255,
+      "step": 1028
+    },
+    {
+      "epoch": 0.24849070272880946,
+      "grad_norm": 0.29703572392463684,
+      "learning_rate": 8.717968055550759e-05,
+      "loss": 1.6139,
+      "step": 1029
+    },
+    {
+      "epoch": 0.24873219029219995,
+      "grad_norm": 0.3209615647792816,
+      "learning_rate": 8.71537875917788e-05,
+      "loss": 1.8684,
+      "step": 1030
+    },
+    {
+      "epoch": 0.24897367785559044,
+      "grad_norm": 0.32092544436454773,
+      "learning_rate": 8.712787236017719e-05,
+      "loss": 1.8249,
+      "step": 1031
+    },
+    {
+      "epoch": 0.24921516541898092,
+      "grad_norm": 0.28598418831825256,
+      "learning_rate": 8.710193487623488e-05,
+      "loss": 1.7439,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2494566529823714,
+      "grad_norm": 0.30598631501197815,
+      "learning_rate": 8.707597515549733e-05,
+      "loss": 1.7287,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2496981405457619,
+      "grad_norm": 0.31710535287857056,
+      "learning_rate": 8.704999321352331e-05,
+      "loss": 1.9099,
+      "step": 1034
+    },
+    {
+      "epoch": 0.24993962810915238,
+      "grad_norm": 0.3359716236591339,
+      "learning_rate": 8.702398906588494e-05,
+      "loss": 1.9935,
+      "step": 1035
+    },
+    {
+      "epoch": 0.25018111567254286,
+      "grad_norm": 0.3071852922439575,
+      "learning_rate": 8.699796272816762e-05,
+      "loss": 1.7408,
+      "step": 1036
+    },
+    {
+      "epoch": 0.25042260323593335,
+      "grad_norm": 0.3323792815208435,
+      "learning_rate": 8.697191421597008e-05,
+      "loss": 1.8794,
+      "step": 1037
+    },
+    {
+      "epoch": 0.25066409079932384,
+      "grad_norm": 0.29584386944770813,
+      "learning_rate": 8.69458435449043e-05,
+      "loss": 1.7439,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2509055783627143,
+      "grad_norm": 0.28571659326553345,
+      "learning_rate": 8.691975073059555e-05,
+      "loss": 1.6339,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2511470659261048,
+      "grad_norm": 0.2837628722190857,
+      "learning_rate": 8.689363578868241e-05,
+      "loss": 1.717,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2513885534894953,
+      "grad_norm": 0.2912003993988037,
+      "learning_rate": 8.68674987348167e-05,
+      "loss": 1.648,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2516300410528858,
+      "grad_norm": 0.30284371972084045,
+      "learning_rate": 8.684133958466347e-05,
+      "loss": 1.8681,
+      "step": 1042
+    },
+    {
+      "epoch": 0.25187152861627626,
+      "grad_norm": 0.27904531359672546,
+      "learning_rate": 8.681515835390101e-05,
+      "loss": 1.6447,
+      "step": 1043
+    },
+    {
+      "epoch": 0.25211301617966675,
+      "grad_norm": 0.300796240568161,
+      "learning_rate": 8.678895505822091e-05,
+      "loss": 1.6384,
+      "step": 1044
+    },
+    {
+      "epoch": 0.25235450374305723,
+      "grad_norm": 0.2925683557987213,
+      "learning_rate": 8.676272971332791e-05,
+      "loss": 1.7102,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2525959913064477,
+      "grad_norm": 0.29522907733917236,
+      "learning_rate": 8.673648233494e-05,
+      "loss": 1.7875,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2528374788698382,
+      "grad_norm": 0.2745014429092407,
+      "learning_rate": 8.671021293878838e-05,
+      "loss": 1.7905,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2530789664332287,
+      "grad_norm": 0.28517085313796997,
+      "learning_rate": 8.668392154061741e-05,
+      "loss": 1.7684,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2533204539966192,
+      "grad_norm": 0.29019320011138916,
+      "learning_rate": 8.66576081561847e-05,
+      "loss": 1.8825,
+      "step": 1049
+    },
+    {
+      "epoch": 0.25356194156000966,
+      "grad_norm": 0.2811944782733917,
+      "learning_rate": 8.663127280126096e-05,
+      "loss": 1.7097,
+      "step": 1050
+    },
+    {
+      "epoch": 0.25380342912340015,
+      "grad_norm": 0.2872609496116638,
+      "learning_rate": 8.660491549163014e-05,
+      "loss": 1.7745,
+      "step": 1051
+    },
+    {
+      "epoch": 0.25404491668679063,
+      "grad_norm": 0.27062755823135376,
+      "learning_rate": 8.65785362430893e-05,
+      "loss": 1.536,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2542864042501811,
+      "grad_norm": 0.28564590215682983,
+      "learning_rate": 8.655213507144868e-05,
+      "loss": 1.6888,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2545278918135716,
+      "grad_norm": 0.295981228351593,
+      "learning_rate": 8.652571199253161e-05,
+      "loss": 1.8429,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2547693793769621,
+      "grad_norm": 0.27825748920440674,
+      "learning_rate": 8.649926702217463e-05,
+      "loss": 1.7498,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2550108669403526,
+      "grad_norm": 0.26681530475616455,
+      "learning_rate": 8.647280017622733e-05,
+      "loss": 1.7221,
+      "step": 1056
+    },
+    {
+      "epoch": 0.25525235450374306,
+      "grad_norm": 0.2864789664745331,
+      "learning_rate": 8.644631147055244e-05,
+      "loss": 1.7624,
+      "step": 1057
+    },
+    {
+      "epoch": 0.25549384206713355,
+      "grad_norm": 0.28378716111183167,
+      "learning_rate": 8.641980092102579e-05,
+      "loss": 1.7936,
+      "step": 1058
+    },
+    {
+      "epoch": 0.25573532963052403,
+      "grad_norm": 0.2861391007900238,
+      "learning_rate": 8.639326854353628e-05,
+      "loss": 1.7771,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2559768171939145,
+      "grad_norm": 0.29720035195350647,
+      "learning_rate": 8.636671435398595e-05,
+      "loss": 1.7993,
+      "step": 1060
+    },
+    {
+      "epoch": 0.256218304757305,
+      "grad_norm": 0.28660377860069275,
+      "learning_rate": 8.634013836828985e-05,
+      "loss": 1.6742,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2564597923206955,
+      "grad_norm": 0.2879665195941925,
+      "learning_rate": 8.631354060237614e-05,
+      "loss": 1.7474,
+      "step": 1062
+    },
+    {
+      "epoch": 0.256701279884086,
+      "grad_norm": 0.30121028423309326,
+      "learning_rate": 8.628692107218596e-05,
+      "loss": 1.7211,
+      "step": 1063
+    },
+    {
+      "epoch": 0.25694276744747646,
+      "grad_norm": 0.2766074538230896,
+      "learning_rate": 8.62602797936736e-05,
+      "loss": 1.6623,
+      "step": 1064
+    },
+    {
+      "epoch": 0.25718425501086695,
+      "grad_norm": 0.2935093939304352,
+      "learning_rate": 8.62336167828063e-05,
+      "loss": 1.6995,
+      "step": 1065
+    },
+    {
+      "epoch": 0.25742574257425743,
+      "grad_norm": 0.2767849862575531,
+      "learning_rate": 8.62069320555644e-05,
+      "loss": 1.794,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2576672301376479,
+      "grad_norm": 0.29953351616859436,
+      "learning_rate": 8.618022562794114e-05,
+      "loss": 1.8499,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2579087177010384,
+      "grad_norm": 0.4503518342971802,
+      "learning_rate": 8.615349751594291e-05,
+      "loss": 1.7237,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2581502052644289,
+      "grad_norm": 0.278839647769928,
+      "learning_rate": 8.612674773558899e-05,
+      "loss": 1.6459,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2583916928278194,
+      "grad_norm": 0.2816259264945984,
+      "learning_rate": 8.609997630291167e-05,
+      "loss": 1.7863,
+      "step": 1070
+    },
+    {
+      "epoch": 0.25863318039120986,
+      "grad_norm": 0.30857524275779724,
+      "learning_rate": 8.607318323395626e-05,
+      "loss": 1.8289,
+      "step": 1071
+    },
+    {
+      "epoch": 0.25887466795460035,
+      "grad_norm": 0.288631796836853,
+      "learning_rate": 8.604636854478098e-05,
+      "loss": 1.7618,
+      "step": 1072
+    },
+    {
+      "epoch": 0.25911615551799083,
+      "grad_norm": 0.29358530044555664,
+      "learning_rate": 8.601953225145704e-05,
+      "loss": 1.9032,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2593576430813813,
+      "grad_norm": 0.29718780517578125,
+      "learning_rate": 8.599267437006859e-05,
+      "loss": 1.8647,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2595991306447718,
+      "grad_norm": 0.2939820885658264,
+      "learning_rate": 8.596579491671273e-05,
+      "loss": 1.7947,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2598406182081623,
+      "grad_norm": 0.29942068457603455,
+      "learning_rate": 8.593889390749947e-05,
+      "loss": 1.7313,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2600821057715528,
+      "grad_norm": 0.28265923261642456,
+      "learning_rate": 8.591197135855175e-05,
+      "loss": 1.8637,
+      "step": 1077
+    },
+    {
+      "epoch": 0.26032359333494326,
+      "grad_norm": 0.28485462069511414,
+      "learning_rate": 8.588502728600544e-05,
+      "loss": 1.7327,
+      "step": 1078
+    },
+    {
+      "epoch": 0.26056508089833375,
+      "grad_norm": 0.2807444632053375,
+      "learning_rate": 8.585806170600926e-05,
+      "loss": 1.784,
+      "step": 1079
+    },
+    {
+      "epoch": 0.26080656846172423,
+      "grad_norm": 0.3553418219089508,
+      "learning_rate": 8.583107463472484e-05,
+      "loss": 1.6733,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2610480560251147,
+      "grad_norm": 0.27911999821662903,
+      "learning_rate": 8.580406608832675e-05,
+      "loss": 1.7077,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2612895435885052,
+      "grad_norm": 0.28452926874160767,
+      "learning_rate": 8.577703608300234e-05,
+      "loss": 1.7534,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2615310311518957,
+      "grad_norm": 0.3004327118396759,
+      "learning_rate": 8.574998463495187e-05,
+      "loss": 1.7699,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2617725187152862,
+      "grad_norm": 0.28714221715927124,
+      "learning_rate": 8.572291176038845e-05,
+      "loss": 1.7046,
+      "step": 1084
+    },
+    {
+      "epoch": 0.26201400627867666,
+      "grad_norm": 0.2888624668121338,
+      "learning_rate": 8.569581747553801e-05,
+      "loss": 1.764,
+      "step": 1085
+    },
+    {
+      "epoch": 0.26225549384206714,
+      "grad_norm": 0.2808159291744232,
+      "learning_rate": 8.566870179663935e-05,
+      "loss": 1.6601,
+      "step": 1086
+    },
+    {
+      "epoch": 0.26249698140545763,
+      "grad_norm": 0.2881624102592468,
+      "learning_rate": 8.564156473994405e-05,
+      "loss": 1.8495,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2627384689688481,
+      "grad_norm": 0.28251782059669495,
+      "learning_rate": 8.561440632171653e-05,
+      "loss": 1.6594,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2629799565322386,
+      "grad_norm": 0.3031906187534332,
+      "learning_rate": 8.558722655823401e-05,
+      "loss": 1.8254,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2632214440956291,
+      "grad_norm": 0.31372591853141785,
+      "learning_rate": 8.55600254657865e-05,
+      "loss": 1.7989,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2634629316590196,
+      "grad_norm": 0.2720772922039032,
+      "learning_rate": 8.553280306067678e-05,
+      "loss": 1.6419,
+      "step": 1091
+    },
+    {
+      "epoch": 0.26370441922241006,
+      "grad_norm": 0.26789841055870056,
+      "learning_rate": 8.550555935922042e-05,
+      "loss": 1.6165,
+      "step": 1092
+    },
+    {
+      "epoch": 0.26394590678580054,
+      "grad_norm": 0.2883191406726837,
+      "learning_rate": 8.547829437774577e-05,
+      "loss": 1.6455,
+      "step": 1093
+    },
+    {
+      "epoch": 0.26418739434919103,
+      "grad_norm": 0.29379168152809143,
+      "learning_rate": 8.545100813259387e-05,
+      "loss": 1.785,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2644288819125815,
+      "grad_norm": 0.2879287004470825,
+      "learning_rate": 8.542370064011858e-05,
+      "loss": 1.8095,
+      "step": 1095
+    },
+    {
+      "epoch": 0.264670369475972,
+      "grad_norm": 0.29632505774497986,
+      "learning_rate": 8.539637191668646e-05,
+      "loss": 1.7733,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2649118570393625,
+      "grad_norm": 0.28169459104537964,
+      "learning_rate": 8.53690219786768e-05,
+      "loss": 1.6289,
+      "step": 1097
+    },
+    {
+      "epoch": 0.26515334460275297,
+      "grad_norm": 0.31244707107543945,
+      "learning_rate": 8.534165084248157e-05,
+      "loss": 1.9603,
+      "step": 1098
+    },
+    {
+      "epoch": 0.26539483216614346,
+      "grad_norm": 0.29472532868385315,
+      "learning_rate": 8.531425852450552e-05,
+      "loss": 1.8188,
+      "step": 1099
+    },
+    {
+      "epoch": 0.26563631972953394,
+      "grad_norm": 0.27472302317619324,
+      "learning_rate": 8.528684504116601e-05,
+      "loss": 1.5912,
+      "step": 1100
+    },
+    {
+      "epoch": 0.26587780729292443,
+      "grad_norm": 0.2806059420108795,
+      "learning_rate": 8.525941040889315e-05,
+      "loss": 1.5864,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2661192948563149,
+      "grad_norm": 0.3072028160095215,
+      "learning_rate": 8.523195464412972e-05,
+      "loss": 1.7789,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2663607824197054,
+      "grad_norm": 0.2739306390285492,
+      "learning_rate": 8.520447776333113e-05,
+      "loss": 1.631,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2666022699830959,
+      "grad_norm": 0.3055983781814575,
+      "learning_rate": 8.517697978296544e-05,
+      "loss": 1.8486,
+      "step": 1104
+    },
+    {
+      "epoch": 0.26684375754648637,
+      "grad_norm": 0.2936011552810669,
+      "learning_rate": 8.514946071951342e-05,
+      "loss": 1.6725,
+      "step": 1105
+    },
+    {
+      "epoch": 0.26708524510987686,
+      "grad_norm": 0.28947994112968445,
+      "learning_rate": 8.51219205894684e-05,
+      "loss": 1.7471,
+      "step": 1106
+    },
+    {
+      "epoch": 0.26732673267326734,
+      "grad_norm": 0.3159331679344177,
+      "learning_rate": 8.50943594093364e-05,
+      "loss": 1.8664,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2675682202366578,
+      "grad_norm": 0.2945443093776703,
+      "learning_rate": 8.5066777195636e-05,
+      "loss": 1.7574,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2678097078000483,
+      "grad_norm": 0.29109689593315125,
+      "learning_rate": 8.503917396489847e-05,
+      "loss": 1.6845,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2680511953634388,
+      "grad_norm": 0.2907101809978485,
+      "learning_rate": 8.501154973366754e-05,
+      "loss": 1.7214,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 0.27306222915649414,
+      "learning_rate": 8.498390451849967e-05,
+      "loss": 1.6801,
+      "step": 1111
+    },
+    {
+      "epoch": 0.26853417049021977,
+      "grad_norm": 0.28223463892936707,
+      "learning_rate": 8.495623833596382e-05,
+      "loss": 1.7371,
+      "step": 1112
+    },
+    {
+      "epoch": 0.26877565805361026,
+      "grad_norm": 0.29571643471717834,
+      "learning_rate": 8.492855120264151e-05,
+      "loss": 1.6157,
+      "step": 1113
+    },
+    {
+      "epoch": 0.26901714561700074,
+      "grad_norm": 0.27164462208747864,
+      "learning_rate": 8.490084313512685e-05,
+      "loss": 1.6855,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2692586331803912,
+      "grad_norm": 0.28996163606643677,
+      "learning_rate": 8.48731141500265e-05,
+      "loss": 1.7985,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2695001207437817,
+      "grad_norm": 0.2976994812488556,
+      "learning_rate": 8.484536426395962e-05,
+      "loss": 1.7614,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2697416083071722,
+      "grad_norm": 0.2953455448150635,
+      "learning_rate": 8.481759349355791e-05,
+      "loss": 1.8129,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2699830958705627,
+      "grad_norm": 0.30153965950012207,
+      "learning_rate": 8.478980185546562e-05,
+      "loss": 1.9054,
+      "step": 1118
+    },
+    {
+      "epoch": 0.27022458343395317,
+      "grad_norm": 0.29038190841674805,
+      "learning_rate": 8.476198936633946e-05,
+      "loss": 1.7608,
+      "step": 1119
+    },
+    {
+      "epoch": 0.27046607099734365,
+      "grad_norm": 0.295154869556427,
+      "learning_rate": 8.473415604284869e-05,
+      "loss": 1.8574,
+      "step": 1120
+    },
+    {
+      "epoch": 0.27070755856073414,
+      "grad_norm": 0.30491289496421814,
+      "learning_rate": 8.470630190167499e-05,
+      "loss": 1.5952,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2709490461241246,
+      "grad_norm": 0.28542447090148926,
+      "learning_rate": 8.467842695951256e-05,
+      "loss": 1.6682,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2711905336875151,
+      "grad_norm": 0.2852567434310913,
+      "learning_rate": 8.465053123306806e-05,
+      "loss": 1.7543,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2714320212509056,
+      "grad_norm": 0.2947034537792206,
+      "learning_rate": 8.462261473906063e-05,
+      "loss": 1.778,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2716735088142961,
+      "grad_norm": 0.3023829758167267,
+      "learning_rate": 8.45946774942218e-05,
+      "loss": 1.6253,
+      "step": 1125
+    },
+    {
+      "epoch": 0.27191499637768657,
+      "grad_norm": 0.2906095087528229,
+      "learning_rate": 8.456671951529559e-05,
+      "loss": 1.6695,
+      "step": 1126
+    },
+    {
+      "epoch": 0.27215648394107705,
+      "grad_norm": 0.28810620307922363,
+      "learning_rate": 8.453874081903841e-05,
+      "loss": 1.8087,
+      "step": 1127
+    },
+    {
+      "epoch": 0.27239797150446754,
+      "grad_norm": 0.2939632833003998,
+      "learning_rate": 8.451074142221913e-05,
+      "loss": 1.8691,
+      "step": 1128
+    },
+    {
+      "epoch": 0.272639459067858,
+      "grad_norm": 0.47220274806022644,
+      "learning_rate": 8.448272134161896e-05,
+      "loss": 1.8504,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2728809466312485,
+      "grad_norm": 0.26881143450737,
+      "learning_rate": 8.445468059403159e-05,
+      "loss": 1.6685,
+      "step": 1130
+    },
+    {
+      "epoch": 0.273122434194639,
+      "grad_norm": 0.29267653822898865,
+      "learning_rate": 8.442661919626305e-05,
+      "loss": 1.814,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2733639217580295,
+      "grad_norm": 0.29204314947128296,
+      "learning_rate": 8.439853716513171e-05,
+      "loss": 1.7541,
+      "step": 1132
+    },
+    {
+      "epoch": 0.27360540932141997,
+      "grad_norm": 0.28699198365211487,
+      "learning_rate": 8.437043451746837e-05,
+      "loss": 1.7636,
+      "step": 1133
+    },
+    {
+      "epoch": 0.27384689688481045,
+      "grad_norm": 0.2884604334831238,
+      "learning_rate": 8.434231127011617e-05,
+      "loss": 1.7746,
+      "step": 1134
+    },
+    {
+      "epoch": 0.27408838444820094,
+      "grad_norm": 0.275660902261734,
+      "learning_rate": 8.431416743993059e-05,
+      "loss": 1.5836,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2743298720115914,
+      "grad_norm": 0.2811957597732544,
+      "learning_rate": 8.428600304377942e-05,
+      "loss": 1.7617,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2745713595749819,
+      "grad_norm": 0.28643205761909485,
+      "learning_rate": 8.425781809854285e-05,
+      "loss": 1.6944,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2748128471383724,
+      "grad_norm": 0.286495566368103,
+      "learning_rate": 8.42296126211133e-05,
+      "loss": 1.7086,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2750543347017629,
+      "grad_norm": 0.28585320711135864,
+      "learning_rate": 8.420138662839552e-05,
+      "loss": 1.6291,
+      "step": 1139
+    },
+    {
+      "epoch": 0.27529582226515337,
+      "grad_norm": 0.2841216027736664,
+      "learning_rate": 8.417314013730662e-05,
+      "loss": 1.6028,
+      "step": 1140
+    },
+    {
+      "epoch": 0.27553730982854385,
+      "grad_norm": 0.2992970645427704,
+      "learning_rate": 8.414487316477589e-05,
+      "loss": 1.6879,
+      "step": 1141
+    },
+    {
+      "epoch": 0.27577879739193434,
+      "grad_norm": 0.2880629897117615,
+      "learning_rate": 8.411658572774498e-05,
+      "loss": 1.799,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2760202849553248,
+      "grad_norm": 0.286931574344635,
+      "learning_rate": 8.408827784316777e-05,
+      "loss": 1.8065,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2762617725187153,
+      "grad_norm": 0.29781198501586914,
+      "learning_rate": 8.405994952801042e-05,
+      "loss": 1.7046,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2765032600821058,
+      "grad_norm": 0.2892191708087921,
+      "learning_rate": 8.403160079925127e-05,
+      "loss": 1.8779,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2767447476454963,
+      "grad_norm": 0.2801336646080017,
+      "learning_rate": 8.400323167388098e-05,
+      "loss": 1.6652,
+      "step": 1146
+    },
+    {
+      "epoch": 0.27698623520888677,
+      "grad_norm": 0.28385961055755615,
+      "learning_rate": 8.397484216890237e-05,
+      "loss": 1.5174,
+      "step": 1147
+    },
+    {
+      "epoch": 0.27722772277227725,
+      "grad_norm": 0.2902655005455017,
+      "learning_rate": 8.39464323013305e-05,
+      "loss": 1.7628,
+      "step": 1148
+    },
+    {
+      "epoch": 0.27746921033566774,
+      "grad_norm": 0.31906628608703613,
+      "learning_rate": 8.391800208819267e-05,
+      "loss": 1.9554,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2777106978990582,
+      "grad_norm": 0.2920599579811096,
+      "learning_rate": 8.38895515465283e-05,
+      "loss": 1.8078,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2779521854624487,
+      "grad_norm": 0.2755699157714844,
+      "learning_rate": 8.386108069338903e-05,
+      "loss": 1.6479,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2781936730258392,
+      "grad_norm": 0.31938937306404114,
+      "learning_rate": 8.383258954583868e-05,
+      "loss": 1.7485,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2784351605892297,
+      "grad_norm": 0.3132666349411011,
+      "learning_rate": 8.380407812095325e-05,
+      "loss": 1.7489,
+      "step": 1153
+    },
+    {
+      "epoch": 0.27867664815262017,
+      "grad_norm": 0.2854362428188324,
+      "learning_rate": 8.377554643582084e-05,
+      "loss": 1.8807,
+      "step": 1154
+    },
+    {
+      "epoch": 0.27891813571601065,
+      "grad_norm": 0.3173444867134094,
+      "learning_rate": 8.374699450754174e-05,
+      "loss": 1.8614,
+      "step": 1155
+    },
+    {
+      "epoch": 0.27915962327940114,
+      "grad_norm": 0.27949345111846924,
+      "learning_rate": 8.371842235322836e-05,
+      "loss": 1.731,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2794011108427916,
+      "grad_norm": 0.27879253029823303,
+      "learning_rate": 8.368982999000521e-05,
+      "loss": 1.6843,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2796425984061821,
+      "grad_norm": 0.2896178364753723,
+      "learning_rate": 8.366121743500895e-05,
+      "loss": 1.589,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2798840859695726,
+      "grad_norm": 0.32527682185173035,
+      "learning_rate": 8.363258470538832e-05,
+      "loss": 1.8235,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2801255735329631,
+      "grad_norm": 0.3003365695476532,
+      "learning_rate": 8.360393181830414e-05,
+      "loss": 1.9417,
+      "step": 1160
+    },
+    {
+      "epoch": 0.28036706109635356,
+      "grad_norm": 0.30865755677223206,
+      "learning_rate": 8.357525879092933e-05,
+      "loss": 1.759,
+      "step": 1161
+    },
+    {
+      "epoch": 0.28060854865974405,
+      "grad_norm": 0.29030731320381165,
+      "learning_rate": 8.35465656404489e-05,
+      "loss": 1.7726,
+      "step": 1162
+    },
+    {
+      "epoch": 0.28085003622313454,
+      "grad_norm": 0.2997666895389557,
+      "learning_rate": 8.351785238405985e-05,
+      "loss": 1.6101,
+      "step": 1163
+    },
+    {
+      "epoch": 0.281091523786525,
+      "grad_norm": 0.2943105399608612,
+      "learning_rate": 8.348911903897132e-05,
+      "loss": 1.7153,
+      "step": 1164
+    },
+    {
+      "epoch": 0.28133301134991545,
+      "grad_norm": 0.2931126058101654,
+      "learning_rate": 8.346036562240444e-05,
+      "loss": 1.6673,
+      "step": 1165
+    },
+    {
+      "epoch": 0.28157449891330594,
+      "grad_norm": 0.31237348914146423,
+      "learning_rate": 8.343159215159235e-05,
+      "loss": 1.965,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2818159864766964,
+      "grad_norm": 0.30502235889434814,
+      "learning_rate": 8.340279864378026e-05,
+      "loss": 1.6968,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2820574740400869,
+      "grad_norm": 0.28124526143074036,
+      "learning_rate": 8.337398511622536e-05,
+      "loss": 1.7541,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2822989616034774,
+      "grad_norm": 0.2825506031513214,
+      "learning_rate": 8.334515158619685e-05,
+      "loss": 1.625,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2825404491668679,
+      "grad_norm": 0.29996126890182495,
+      "learning_rate": 8.33162980709759e-05,
+      "loss": 1.8162,
+      "step": 1170
+    },
+    {
+      "epoch": 0.28278193673025837,
+      "grad_norm": 0.28735971450805664,
+      "learning_rate": 8.328742458785568e-05,
+      "loss": 1.6568,
+      "step": 1171
+    },
+    {
+      "epoch": 0.28302342429364885,
+      "grad_norm": 0.30295073986053467,
+      "learning_rate": 8.325853115414132e-05,
+      "loss": 1.768,
+      "step": 1172
+    },
+    {
+      "epoch": 0.28326491185703934,
+      "grad_norm": 0.29709091782569885,
+      "learning_rate": 8.322961778714989e-05,
+      "loss": 1.7637,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2835063994204298,
+      "grad_norm": 0.29945021867752075,
+      "learning_rate": 8.320068450421044e-05,
+      "loss": 1.7821,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2837478869838203,
+      "grad_norm": 0.29444846510887146,
+      "learning_rate": 8.317173132266392e-05,
+      "loss": 1.8555,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2839893745472108,
+      "grad_norm": 0.27773210406303406,
+      "learning_rate": 8.314275825986325e-05,
+      "loss": 1.6424,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2842308621106013,
+      "grad_norm": 0.31503719091415405,
+      "learning_rate": 8.311376533317321e-05,
+      "loss": 1.985,
+      "step": 1177
+    },
+    {
+      "epoch": 0.28447234967399176,
+      "grad_norm": 0.3152500092983246,
+      "learning_rate": 8.308475255997055e-05,
+      "loss": 1.8576,
+      "step": 1178
+    },
+    {
+      "epoch": 0.28471383723738225,
+      "grad_norm": 0.28431928157806396,
+      "learning_rate": 8.305571995764385e-05,
+      "loss": 1.6745,
+      "step": 1179
+    },
+    {
+      "epoch": 0.28495532480077274,
+      "grad_norm": 0.29776814579963684,
+      "learning_rate": 8.302666754359363e-05,
+      "loss": 1.6738,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2851968123641632,
+      "grad_norm": 0.29094064235687256,
+      "learning_rate": 8.299759533523222e-05,
+      "loss": 1.7761,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2854382999275537,
+      "grad_norm": 0.28861159086227417,
+      "learning_rate": 8.296850334998391e-05,
+      "loss": 1.7584,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2856797874909442,
+      "grad_norm": 0.32065683603286743,
+      "learning_rate": 8.293939160528474e-05,
+      "loss": 1.9342,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2859212750543347,
+      "grad_norm": 0.3114961087703705,
+      "learning_rate": 8.291026011858266e-05,
+      "loss": 1.9161,
+      "step": 1184
+    },
+    {
+      "epoch": 0.28616276261772516,
+      "grad_norm": 0.2836478650569916,
+      "learning_rate": 8.288110890733741e-05,
+      "loss": 1.6908,
+      "step": 1185
+    },
+    {
+      "epoch": 0.28640425018111565,
+      "grad_norm": 0.289570689201355,
+      "learning_rate": 8.28519379890206e-05,
+      "loss": 1.7737,
+      "step": 1186
+    },
+    {
+      "epoch": 0.28664573774450613,
+      "grad_norm": 0.2810715138912201,
+      "learning_rate": 8.28227473811156e-05,
+      "loss": 1.7923,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2868872253078966,
+      "grad_norm": 0.2763075828552246,
+      "learning_rate": 8.279353710111761e-05,
+      "loss": 1.6779,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2871287128712871,
+      "grad_norm": 0.28934016823768616,
+      "learning_rate": 8.276430716653363e-05,
+      "loss": 1.7764,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2873702004346776,
+      "grad_norm": 0.3152405619621277,
+      "learning_rate": 8.273505759488241e-05,
+      "loss": 1.9608,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2876116879980681,
+      "grad_norm": 0.3540990948677063,
+      "learning_rate": 8.270578840369449e-05,
+      "loss": 1.817,
+      "step": 1191
+    },
+    {
+      "epoch": 0.28785317556145856,
+      "grad_norm": 0.27829819917678833,
+      "learning_rate": 8.267649961051219e-05,
+      "loss": 1.632,
+      "step": 1192
+    },
+    {
+      "epoch": 0.28809466312484905,
+      "grad_norm": 0.3032492995262146,
+      "learning_rate": 8.264719123288949e-05,
+      "loss": 1.853,
+      "step": 1193
+    },
+    {
+      "epoch": 0.28833615068823953,
+      "grad_norm": 0.2742372453212738,
+      "learning_rate": 8.26178632883922e-05,
+      "loss": 1.6385,
+      "step": 1194
+    },
+    {
+      "epoch": 0.28857763825163,
+      "grad_norm": 0.27677586674690247,
+      "learning_rate": 8.258851579459783e-05,
+      "loss": 1.6423,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2888191258150205,
+      "grad_norm": 0.2932434380054474,
+      "learning_rate": 8.25591487690956e-05,
+      "loss": 1.8024,
+      "step": 1196
+    },
+    {
+      "epoch": 0.289060613378411,
+      "grad_norm": 0.2936798632144928,
+      "learning_rate": 8.252976222948647e-05,
+      "loss": 1.8136,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2893021009418015,
+      "grad_norm": 0.3236442804336548,
+      "learning_rate": 8.250035619338302e-05,
+      "loss": 1.7561,
+      "step": 1198
+    },
+    {
+      "epoch": 0.28954358850519196,
+      "grad_norm": 0.3159981966018677,
+      "learning_rate": 8.247093067840956e-05,
+      "loss": 1.9135,
+      "step": 1199
+    },
+    {
+      "epoch": 0.28978507606858245,
+      "grad_norm": 0.2588784098625183,
+      "learning_rate": 8.244148570220211e-05,
+      "loss": 1.5257,
+      "step": 1200
+    },
+    {
+      "epoch": 0.29002656363197293,
+      "grad_norm": 0.29745063185691833,
+      "learning_rate": 8.241202128240829e-05,
+      "loss": 1.833,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2902680511953634,
+      "grad_norm": 0.28102126717567444,
+      "learning_rate": 8.23825374366874e-05,
+      "loss": 1.673,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2905095387587539,
+      "grad_norm": 0.27232810854911804,
+      "learning_rate": 8.23530341827104e-05,
+      "loss": 1.6953,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2907510263221444,
+      "grad_norm": 0.28626495599746704,
+      "learning_rate": 8.232351153815988e-05,
+      "loss": 1.695,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2909925138855349,
+      "grad_norm": 0.30019521713256836,
+      "learning_rate": 8.229396952073001e-05,
+      "loss": 1.6836,
+      "step": 1205
+    },
+    {
+      "epoch": 0.29123400144892536,
+      "grad_norm": 0.2773890197277069,
+      "learning_rate": 8.226440814812662e-05,
+      "loss": 1.7156,
+      "step": 1206
+    },
+    {
+      "epoch": 0.29147548901231585,
+      "grad_norm": 0.3097532093524933,
+      "learning_rate": 8.223482743806709e-05,
+      "loss": 1.8985,
+      "step": 1207
+    },
+    {
+      "epoch": 0.29171697657570633,
+      "grad_norm": 0.3096452057361603,
+      "learning_rate": 8.220522740828046e-05,
+      "loss": 1.9616,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2919584641390968,
+      "grad_norm": 0.2948162853717804,
+      "learning_rate": 8.217560807650728e-05,
+      "loss": 1.7219,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2921999517024873,
+      "grad_norm": 0.2883644700050354,
+      "learning_rate": 8.21459694604997e-05,
+      "loss": 1.7977,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2924414392658778,
+      "grad_norm": 0.3046281635761261,
+      "learning_rate": 8.211631157802144e-05,
+      "loss": 2.0905,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 0.31257206201553345,
+      "learning_rate": 8.208663444684776e-05,
+      "loss": 1.9364,
+      "step": 1212
+    },
+    {
+      "epoch": 0.29292441439265876,
+      "grad_norm": 0.29145514965057373,
+      "learning_rate": 8.20569380847654e-05,
+      "loss": 1.7626,
+      "step": 1213
+    },
+    {
+      "epoch": 0.29316590195604925,
+      "grad_norm": 0.2752838730812073,
+      "learning_rate": 8.202722250957273e-05,
+      "loss": 1.5806,
+      "step": 1214
+    },
+    {
+      "epoch": 0.29340738951943973,
+      "grad_norm": 0.2736091911792755,
+      "learning_rate": 8.199748773907956e-05,
+      "loss": 1.6547,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2936488770828302,
+      "grad_norm": 0.3060460090637207,
+      "learning_rate": 8.196773379110722e-05,
+      "loss": 1.8482,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2938903646462207,
+      "grad_norm": 0.2833315134048462,
+      "learning_rate": 8.193796068348853e-05,
+      "loss": 1.6437,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2941318522096112,
+      "grad_norm": 0.2926206588745117,
+      "learning_rate": 8.190816843406783e-05,
+      "loss": 1.8218,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2943733397730017,
+      "grad_norm": 0.29806745052337646,
+      "learning_rate": 8.187835706070089e-05,
+      "loss": 1.8769,
+      "step": 1219
+    },
+    {
+      "epoch": 0.29461482733639216,
+      "grad_norm": 0.2936302125453949,
+      "learning_rate": 8.184852658125494e-05,
+      "loss": 1.7622,
+      "step": 1220
+    },
+    {
+      "epoch": 0.29485631489978265,
+      "grad_norm": 0.30911576747894287,
+      "learning_rate": 8.18186770136087e-05,
+      "loss": 1.9,
+      "step": 1221
+    },
+    {
+      "epoch": 0.29509780246317313,
+      "grad_norm": 0.2942495048046112,
+      "learning_rate": 8.178880837565228e-05,
+      "loss": 1.7852,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2953392900265636,
+      "grad_norm": 0.2813624441623688,
+      "learning_rate": 8.17589206852873e-05,
+      "loss": 1.738,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2955807775899541,
+      "grad_norm": 0.2759395241737366,
+      "learning_rate": 8.172901396042669e-05,
+      "loss": 1.6633,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2958222651533446,
+      "grad_norm": 0.28530681133270264,
+      "learning_rate": 8.169908821899489e-05,
+      "loss": 1.6496,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2960637527167351,
+      "grad_norm": 0.28899961709976196,
+      "learning_rate": 8.166914347892764e-05,
+      "loss": 1.7069,
+      "step": 1226
+    },
+    {
+      "epoch": 0.29630524028012556,
+      "grad_norm": 0.3118133842945099,
+      "learning_rate": 8.163917975817217e-05,
+      "loss": 1.7776,
+      "step": 1227
+    },
+    {
+      "epoch": 0.29654672784351604,
+      "grad_norm": 0.288650244474411,
+      "learning_rate": 8.1609197074687e-05,
+      "loss": 1.6574,
+      "step": 1228
+    },
+    {
+      "epoch": 0.29678821540690653,
+      "grad_norm": 0.28079408407211304,
+      "learning_rate": 8.157919544644206e-05,
+      "loss": 1.6564,
+      "step": 1229
+    },
+    {
+      "epoch": 0.297029702970297,
+      "grad_norm": 0.28103119134902954,
+      "learning_rate": 8.154917489141865e-05,
+      "loss": 1.6368,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2972711905336875,
+      "grad_norm": 0.2799832820892334,
+      "learning_rate": 8.151913542760934e-05,
+      "loss": 1.6049,
+      "step": 1231
+    },
+    {
+      "epoch": 0.297512678097078,
+      "grad_norm": 0.2864444851875305,
+      "learning_rate": 8.148907707301811e-05,
+      "loss": 1.6147,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2977541656604685,
+      "grad_norm": 0.28280696272850037,
+      "learning_rate": 8.145899984566024e-05,
+      "loss": 1.8078,
+      "step": 1233
+    },
+    {
+      "epoch": 0.29799565322385896,
+      "grad_norm": 0.3180987238883972,
+      "learning_rate": 8.142890376356229e-05,
+      "loss": 1.8723,
+      "step": 1234
+    },
+    {
+      "epoch": 0.29823714078724944,
+      "grad_norm": 0.28342029452323914,
+      "learning_rate": 8.139878884476216e-05,
+      "loss": 1.6105,
+      "step": 1235
+    },
+    {
+      "epoch": 0.29847862835063993,
+      "grad_norm": 0.2965908944606781,
+      "learning_rate": 8.136865510730903e-05,
+      "loss": 1.7005,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2987201159140304,
+      "grad_norm": 0.28203895688056946,
+      "learning_rate": 8.133850256926335e-05,
+      "loss": 1.7559,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2989616034774209,
+      "grad_norm": 0.29620274901390076,
+      "learning_rate": 8.130833124869681e-05,
+      "loss": 1.8065,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2992030910408114,
+      "grad_norm": 0.2921755909919739,
+      "learning_rate": 8.127814116369242e-05,
+      "loss": 1.7224,
+      "step": 1239
+    },
+    {
+      "epoch": 0.29944457860420187,
+      "grad_norm": 0.3076702356338501,
+      "learning_rate": 8.12479323323444e-05,
+      "loss": 1.9597,
+      "step": 1240
+    },
+    {
+      "epoch": 0.29968606616759236,
+      "grad_norm": 0.3073546290397644,
+      "learning_rate": 8.121770477275821e-05,
+      "loss": 1.7661,
+      "step": 1241
+    },
+    {
+      "epoch": 0.29992755373098284,
+      "grad_norm": 0.27899089455604553,
+      "learning_rate": 8.118745850305054e-05,
+      "loss": 1.6729,
+      "step": 1242
+    },
+    {
+      "epoch": 0.30016904129437333,
+      "grad_norm": 0.28602051734924316,
+      "learning_rate": 8.115719354134926e-05,
+      "loss": 1.6565,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3004105288577638,
+      "grad_norm": 0.3022647798061371,
+      "learning_rate": 8.11269099057935e-05,
+      "loss": 1.7554,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3006520164211543,
+      "grad_norm": 0.2899514436721802,
+      "learning_rate": 8.109660761453355e-05,
+      "loss": 1.7334,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3008935039845448,
+      "grad_norm": 0.2962121367454529,
+      "learning_rate": 8.106628668573087e-05,
+      "loss": 1.769,
+      "step": 1246
+    },
+    {
+      "epoch": 0.30113499154793527,
+      "grad_norm": 0.29414042830467224,
+      "learning_rate": 8.103594713755813e-05,
+      "loss": 1.8094,
+      "step": 1247
+    },
+    {
+      "epoch": 0.30137647911132576,
+      "grad_norm": 0.33522823452949524,
+      "learning_rate": 8.100558898819912e-05,
+      "loss": 1.9885,
+      "step": 1248
+    },
+    {
+      "epoch": 0.30161796667471624,
+      "grad_norm": 0.32740721106529236,
+      "learning_rate": 8.097521225584876e-05,
+      "loss": 1.8678,
+      "step": 1249
+    },
+    {
+      "epoch": 0.30185945423810673,
+      "grad_norm": 0.2977084815502167,
+      "learning_rate": 8.094481695871319e-05,
+      "loss": 1.6237,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3021009418014972,
+      "grad_norm": 0.29849734902381897,
+      "learning_rate": 8.091440311500963e-05,
+      "loss": 1.8449,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3023424293648877,
+      "grad_norm": 0.2919251620769501,
+      "learning_rate": 8.088397074296636e-05,
+      "loss": 1.7264,
+      "step": 1252
+    },
+    {
+      "epoch": 0.3025839169282782,
+      "grad_norm": 0.28110143542289734,
+      "learning_rate": 8.085351986082287e-05,
+      "loss": 1.6551,
+      "step": 1253
+    },
+    {
+      "epoch": 0.30282540449166867,
+      "grad_norm": 0.28460079431533813,
+      "learning_rate": 8.082305048682966e-05,
+      "loss": 1.7723,
+      "step": 1254
+    },
+    {
+      "epoch": 0.30306689205505916,
+      "grad_norm": 0.2959478795528412,
+      "learning_rate": 8.079256263924836e-05,
+      "loss": 1.7073,
+      "step": 1255
+    },
+    {
+      "epoch": 0.30330837961844964,
+      "grad_norm": 0.30130937695503235,
+      "learning_rate": 8.076205633635165e-05,
+      "loss": 1.8099,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3035498671818401,
+      "grad_norm": 0.2832283675670624,
+      "learning_rate": 8.073153159642328e-05,
+      "loss": 1.7325,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3037913547452306,
+      "grad_norm": 0.27280157804489136,
+      "learning_rate": 8.070098843775804e-05,
+      "loss": 1.7815,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3040328423086211,
+      "grad_norm": 0.3622336983680725,
+      "learning_rate": 8.067042687866178e-05,
+      "loss": 2.2295,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3042743298720116,
+      "grad_norm": 0.29381340742111206,
+      "learning_rate": 8.063984693745136e-05,
+      "loss": 1.8125,
+      "step": 1260
+    },
+    {
+      "epoch": 0.30451581743540207,
+      "grad_norm": 0.28132760524749756,
+      "learning_rate": 8.060924863245466e-05,
+      "loss": 1.4983,
+      "step": 1261
+    },
+    {
+      "epoch": 0.30475730499879256,
+      "grad_norm": 0.30654898285865784,
+      "learning_rate": 8.057863198201056e-05,
+      "loss": 1.7907,
+      "step": 1262
+    },
+    {
+      "epoch": 0.30499879256218304,
+      "grad_norm": 0.2809448838233948,
+      "learning_rate": 8.054799700446896e-05,
+      "loss": 1.7044,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3052402801255735,
+      "grad_norm": 0.2905556559562683,
+      "learning_rate": 8.051734371819072e-05,
+      "loss": 1.7571,
+      "step": 1264
+    },
+    {
+      "epoch": 0.305481767688964,
+      "grad_norm": 0.30511733889579773,
+      "learning_rate": 8.048667214154769e-05,
+      "loss": 1.8195,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3057232552523545,
+      "grad_norm": 0.2864232659339905,
+      "learning_rate": 8.045598229292265e-05,
+      "loss": 1.5109,
+      "step": 1266
+    },
+    {
+      "epoch": 0.305964742815745,
+      "grad_norm": 0.2991337776184082,
+      "learning_rate": 8.042527419070938e-05,
+      "loss": 1.568,
+      "step": 1267
+    },
+    {
+      "epoch": 0.30620623037913547,
+      "grad_norm": 0.327675998210907,
+      "learning_rate": 8.039454785331256e-05,
+      "loss": 2.1194,
+      "step": 1268
+    },
+    {
+      "epoch": 0.30644771794252595,
+      "grad_norm": 0.28206297755241394,
+      "learning_rate": 8.036380329914781e-05,
+      "loss": 1.6651,
+      "step": 1269
+    },
+    {
+      "epoch": 0.30668920550591644,
+      "grad_norm": 0.286390483379364,
+      "learning_rate": 8.03330405466417e-05,
+      "loss": 1.695,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3069306930693069,
+      "grad_norm": 0.3134172260761261,
+      "learning_rate": 8.030225961423165e-05,
+      "loss": 1.8522,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3071721806326974,
+      "grad_norm": 0.30529677867889404,
+      "learning_rate": 8.0271460520366e-05,
+      "loss": 1.7846,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3074136681960879,
+      "grad_norm": 0.2850496768951416,
+      "learning_rate": 8.0240643283504e-05,
+      "loss": 1.8515,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3076551557594784,
+      "grad_norm": 0.31650790572166443,
+      "learning_rate": 8.020980792211576e-05,
+      "loss": 1.7635,
+      "step": 1274
+    },
+    {
+      "epoch": 0.30789664332286887,
+      "grad_norm": 0.3554689288139343,
+      "learning_rate": 8.017895445468222e-05,
+      "loss": 1.6921,
+      "step": 1275
+    },
+    {
+      "epoch": 0.30813813088625935,
+      "grad_norm": 0.28742724657058716,
+      "learning_rate": 8.014808289969523e-05,
+      "loss": 1.8524,
+      "step": 1276
+    },
+    {
+      "epoch": 0.30837961844964984,
+      "grad_norm": 0.27519968152046204,
+      "learning_rate": 8.01171932756574e-05,
+      "loss": 1.7443,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3086211060130403,
+      "grad_norm": 0.288105845451355,
+      "learning_rate": 8.008628560108227e-05,
+      "loss": 1.6894,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3088625935764308,
+      "grad_norm": 0.29678773880004883,
+      "learning_rate": 8.005535989449411e-05,
+      "loss": 1.6458,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3091040811398213,
+      "grad_norm": 0.2868390679359436,
+      "learning_rate": 8.002441617442807e-05,
+      "loss": 1.5688,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3093455687032118,
+      "grad_norm": 0.302274614572525,
+      "learning_rate": 7.999345445943003e-05,
+      "loss": 1.6713,
+      "step": 1281
+    },
+    {
+      "epoch": 0.30958705626660227,
+      "grad_norm": 0.30040210485458374,
+      "learning_rate": 7.99624747680567e-05,
+      "loss": 1.7323,
+      "step": 1282
+    },
+    {
+      "epoch": 0.30982854382999275,
+      "grad_norm": 0.2940504550933838,
+      "learning_rate": 7.993147711887554e-05,
+      "loss": 1.776,
+      "step": 1283
+    },
+    {
+      "epoch": 0.31007003139338324,
+      "grad_norm": 0.31216347217559814,
+      "learning_rate": 7.99004615304648e-05,
+      "loss": 1.6848,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3103115189567737,
+      "grad_norm": 0.3009435534477234,
+      "learning_rate": 7.986942802141346e-05,
+      "loss": 1.8094,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3105530065201642,
+      "grad_norm": 0.31435760855674744,
+      "learning_rate": 7.983837661032123e-05,
+      "loss": 1.6859,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3107944940835547,
+      "grad_norm": 0.2916475236415863,
+      "learning_rate": 7.980730731579856e-05,
+      "loss": 1.7225,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3110359816469452,
+      "grad_norm": 0.29502925276756287,
+      "learning_rate": 7.977622015646667e-05,
+      "loss": 1.7175,
+      "step": 1288
+    },
+    {
+      "epoch": 0.31127746921033567,
+      "grad_norm": 0.3051729202270508,
+      "learning_rate": 7.974511515095738e-05,
+      "loss": 1.6778,
+      "step": 1289
+    },
+    {
+      "epoch": 0.31151895677372615,
+      "grad_norm": 0.29863834381103516,
+      "learning_rate": 7.971399231791328e-05,
+      "loss": 1.6819,
+      "step": 1290
+    },
+    {
+      "epoch": 0.31176044433711664,
+      "grad_norm": 0.27716687321662903,
+      "learning_rate": 7.968285167598766e-05,
+      "loss": 1.6873,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3120019319005071,
+      "grad_norm": 0.30108094215393066,
+      "learning_rate": 7.965169324384445e-05,
+      "loss": 1.6907,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3122434194638976,
+      "grad_norm": 0.3081267774105072,
+      "learning_rate": 7.96205170401582e-05,
+      "loss": 1.6812,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3124849070272881,
+      "grad_norm": 0.2833951711654663,
+      "learning_rate": 7.958932308361422e-05,
+      "loss": 1.6024,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3127263945906786,
+      "grad_norm": 0.3181832730770111,
+      "learning_rate": 7.955811139290837e-05,
+      "loss": 1.896,
+      "step": 1295
+    },
+    {
+      "epoch": 0.31296788215406907,
+      "grad_norm": 0.29816505312919617,
+      "learning_rate": 7.952688198674714e-05,
+      "loss": 1.8003,
+      "step": 1296
+    },
+    {
+      "epoch": 0.31320936971745955,
+      "grad_norm": 0.33011528849601746,
+      "learning_rate": 7.949563488384772e-05,
+      "loss": 1.9278,
+      "step": 1297
+    },
+    {
+      "epoch": 0.31345085728085004,
+      "grad_norm": 0.30737966299057007,
+      "learning_rate": 7.946437010293781e-05,
+      "loss": 1.8645,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3136923448442405,
+      "grad_norm": 0.26630595326423645,
+      "learning_rate": 7.943308766275577e-05,
+      "loss": 1.494,
+      "step": 1299
+    },
+    {
+      "epoch": 0.313933832407631,
+      "grad_norm": 0.298896849155426,
+      "learning_rate": 7.94017875820505e-05,
+      "loss": 1.8095,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3141753199710215,
+      "grad_norm": 0.2950008809566498,
+      "learning_rate": 7.93704698795815e-05,
+      "loss": 1.8302,
+      "step": 1301
+    },
+    {
+      "epoch": 0.314416807534412,
+      "grad_norm": 0.29186734557151794,
+      "learning_rate": 7.933913457411884e-05,
+      "loss": 1.7945,
+      "step": 1302
+    },
+    {
+      "epoch": 0.31465829509780247,
+      "grad_norm": 0.29058417677879333,
+      "learning_rate": 7.93077816844431e-05,
+      "loss": 1.6974,
+      "step": 1303
+    },
+    {
+      "epoch": 0.31489978266119295,
+      "grad_norm": 0.3064737021923065,
+      "learning_rate": 7.927641122934543e-05,
+      "loss": 1.8599,
+      "step": 1304
+    },
+    {
+      "epoch": 0.31514127022458344,
+      "grad_norm": 0.3019355535507202,
+      "learning_rate": 7.924502322762752e-05,
+      "loss": 1.8489,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3153827577879739,
+      "grad_norm": 0.27845993638038635,
+      "learning_rate": 7.921361769810153e-05,
+      "loss": 1.6851,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3156242453513644,
+      "grad_norm": 0.28601062297821045,
+      "learning_rate": 7.918219465959018e-05,
+      "loss": 1.7533,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3158657329147549,
+      "grad_norm": 0.28513553738594055,
+      "learning_rate": 7.915075413092664e-05,
+      "loss": 1.6407,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3161072204781454,
+      "grad_norm": 0.29098600149154663,
+      "learning_rate": 7.91192961309546e-05,
+      "loss": 1.6635,
+      "step": 1309
+    },
+    {
+      "epoch": 0.31634870804153586,
+      "grad_norm": 0.3085253834724426,
+      "learning_rate": 7.908782067852816e-05,
+      "loss": 1.7706,
+      "step": 1310
+    },
+    {
+      "epoch": 0.31659019560492635,
+      "grad_norm": 0.29994043707847595,
+      "learning_rate": 7.905632779251195e-05,
+      "loss": 1.7929,
+      "step": 1311
+    },
+    {
+      "epoch": 0.31683168316831684,
+      "grad_norm": 0.3095788061618805,
+      "learning_rate": 7.902481749178101e-05,
+      "loss": 1.8123,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 0.30506932735443115,
+      "learning_rate": 7.899328979522085e-05,
+      "loss": 1.6084,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3173146582950978,
+      "grad_norm": 0.28293487429618835,
+      "learning_rate": 7.896174472172735e-05,
+      "loss": 1.7082,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3175561458584883,
+      "grad_norm": 0.3032687306404114,
+      "learning_rate": 7.893018229020686e-05,
+      "loss": 1.7877,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3177976334218788,
+      "grad_norm": 0.3086721897125244,
+      "learning_rate": 7.889860251957609e-05,
+      "loss": 1.8977,
+      "step": 1316
+    },
+    {
+      "epoch": 0.31803912098526926,
+      "grad_norm": 0.3076464831829071,
+      "learning_rate": 7.886700542876218e-05,
+      "loss": 1.783,
+      "step": 1317
+    },
+    {
+      "epoch": 0.31828060854865975,
+      "grad_norm": 0.30208855867385864,
+      "learning_rate": 7.883539103670264e-05,
+      "loss": 1.7573,
+      "step": 1318
+    },
+    {
+      "epoch": 0.31852209611205023,
+      "grad_norm": 0.3009052574634552,
+      "learning_rate": 7.880375936234534e-05,
+      "loss": 1.7234,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3187635836754407,
+      "grad_norm": 0.3098020553588867,
+      "learning_rate": 7.87721104246485e-05,
+      "loss": 1.8425,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3190050712388312,
+      "grad_norm": 0.2842791974544525,
+      "learning_rate": 7.874044424258069e-05,
+      "loss": 1.7381,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3192465588022217,
+      "grad_norm": 0.3024597764015198,
+      "learning_rate": 7.870876083512084e-05,
+      "loss": 1.6135,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3194880463656122,
+      "grad_norm": 0.2864610254764557,
+      "learning_rate": 7.867706022125819e-05,
+      "loss": 1.6423,
+      "step": 1323
+    },
+    {
+      "epoch": 0.31972953392900266,
+      "grad_norm": 0.3033413887023926,
+      "learning_rate": 7.864534241999228e-05,
+      "loss": 1.8562,
+      "step": 1324
+    },
+    {
+      "epoch": 0.31997102149239315,
+      "grad_norm": 0.2795904874801636,
+      "learning_rate": 7.861360745033297e-05,
+      "loss": 1.6943,
+      "step": 1325
+    },
+    {
+      "epoch": 0.32021250905578363,
+      "grad_norm": 0.2918721139431,
+      "learning_rate": 7.85818553313004e-05,
+      "loss": 1.7442,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3204539966191741,
+      "grad_norm": 0.2953207194805145,
+      "learning_rate": 7.855008608192498e-05,
+      "loss": 1.7553,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3206954841825646,
+      "grad_norm": 0.29291972517967224,
+      "learning_rate": 7.85182997212474e-05,
+      "loss": 1.738,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3209369717459551,
+      "grad_norm": 0.28523069620132446,
+      "learning_rate": 7.848649626831862e-05,
+      "loss": 1.8182,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3211784593093456,
+      "grad_norm": 0.2963908910751343,
+      "learning_rate": 7.845467574219978e-05,
+      "loss": 1.7844,
+      "step": 1330
+    },
+    {
+      "epoch": 0.32141994687273606,
+      "grad_norm": 0.293582022190094,
+      "learning_rate": 7.842283816196232e-05,
+      "loss": 1.7315,
+      "step": 1331
+    },
+    {
+      "epoch": 0.32166143443612655,
+      "grad_norm": 0.287688672542572,
+      "learning_rate": 7.839098354668791e-05,
+      "loss": 1.5484,
+      "step": 1332
+    },
+    {
+      "epoch": 0.32190292199951703,
+      "grad_norm": 0.3031911253929138,
+      "learning_rate": 7.835911191546836e-05,
+      "loss": 1.8595,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3221444095629075,
+      "grad_norm": 0.3043416738510132,
+      "learning_rate": 7.832722328740575e-05,
+      "loss": 1.8035,
+      "step": 1334
+    },
+    {
+      "epoch": 0.322385897126298,
+      "grad_norm": 0.284494549036026,
+      "learning_rate": 7.82953176816123e-05,
+      "loss": 1.6982,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3226273846896885,
+      "grad_norm": 0.28933319449424744,
+      "learning_rate": 7.82633951172104e-05,
+      "loss": 1.7081,
+      "step": 1336
+    },
+    {
+      "epoch": 0.322868872253079,
+      "grad_norm": 0.2961520850658417,
+      "learning_rate": 7.823145561333266e-05,
+      "loss": 1.8237,
+      "step": 1337
+    },
+    {
+      "epoch": 0.32311035981646946,
+      "grad_norm": 0.33161845803260803,
+      "learning_rate": 7.819949918912178e-05,
+      "loss": 1.9155,
+      "step": 1338
+    },
+    {
+      "epoch": 0.32335184737985995,
+      "grad_norm": 0.28777819871902466,
+      "learning_rate": 7.816752586373064e-05,
+      "loss": 1.7327,
+      "step": 1339
+    },
+    {
+      "epoch": 0.32359333494325043,
+      "grad_norm": 0.28206440806388855,
+      "learning_rate": 7.813553565632222e-05,
+      "loss": 1.7053,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3238348225066409,
+      "grad_norm": 0.2921142578125,
+      "learning_rate": 7.810352858606966e-05,
+      "loss": 1.7221,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3240763100700314,
+      "grad_norm": 0.30265289545059204,
+      "learning_rate": 7.807150467215619e-05,
+      "loss": 1.8833,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3243177976334219,
+      "grad_norm": 0.3193051218986511,
+      "learning_rate": 7.803946393377511e-05,
+      "loss": 1.9729,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3245592851968124,
+      "grad_norm": 0.2971660792827606,
+      "learning_rate": 7.80074063901298e-05,
+      "loss": 1.8479,
+      "step": 1344
+    },
+    {
+      "epoch": 0.32480077276020286,
+      "grad_norm": 0.28120097517967224,
+      "learning_rate": 7.797533206043377e-05,
+      "loss": 1.6295,
+      "step": 1345
+    },
+    {
+      "epoch": 0.32504226032359335,
+      "grad_norm": 0.2861529588699341,
+      "learning_rate": 7.794324096391055e-05,
+      "loss": 1.571,
+      "step": 1346
+    },
+    {
+      "epoch": 0.32528374788698383,
+      "grad_norm": 0.28205791115760803,
+      "learning_rate": 7.79111331197937e-05,
+      "loss": 1.6045,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3255252354503743,
+      "grad_norm": 0.3102561831474304,
+      "learning_rate": 7.787900854732686e-05,
+      "loss": 1.9812,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3257667230137648,
+      "grad_norm": 0.29614007472991943,
+      "learning_rate": 7.784686726576364e-05,
+      "loss": 1.7906,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3260082105771553,
+      "grad_norm": 0.27276650071144104,
+      "learning_rate": 7.781470929436776e-05,
+      "loss": 1.5792,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3262496981405458,
+      "grad_norm": 0.30475497245788574,
+      "learning_rate": 7.778253465241286e-05,
+      "loss": 1.7098,
+      "step": 1351
+    },
+    {
+      "epoch": 0.32649118570393626,
+      "grad_norm": 0.28294622898101807,
+      "learning_rate": 7.775034335918256e-05,
+      "loss": 1.711,
+      "step": 1352
+    },
+    {
+      "epoch": 0.32673267326732675,
+      "grad_norm": 0.30347758531570435,
+      "learning_rate": 7.771813543397055e-05,
+      "loss": 1.8165,
+      "step": 1353
+    },
+    {
+      "epoch": 0.32697416083071723,
+      "grad_norm": 0.2811405062675476,
+      "learning_rate": 7.768591089608042e-05,
+      "loss": 1.7231,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3272156483941077,
+      "grad_norm": 0.2834404408931732,
+      "learning_rate": 7.765366976482568e-05,
+      "loss": 1.6664,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3274571359574982,
+      "grad_norm": 0.30318307876586914,
+      "learning_rate": 7.762141205952991e-05,
+      "loss": 1.8148,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3276986235208887,
+      "grad_norm": 0.30411288142204285,
+      "learning_rate": 7.75891377995265e-05,
+      "loss": 1.7338,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3279401110842792,
+      "grad_norm": 0.28736481070518494,
+      "learning_rate": 7.755684700415881e-05,
+      "loss": 1.6882,
+      "step": 1358
+    },
+    {
+      "epoch": 0.32818159864766966,
+      "grad_norm": 0.2920001149177551,
+      "learning_rate": 7.752453969278014e-05,
+      "loss": 1.7787,
+      "step": 1359
+    },
+    {
+      "epoch": 0.32842308621106014,
+      "grad_norm": 0.30056503415107727,
+      "learning_rate": 7.749221588475363e-05,
+      "loss": 1.6943,
+      "step": 1360
+    },
+    {
+      "epoch": 0.32866457377445063,
+      "grad_norm": 0.2995055913925171,
+      "learning_rate": 7.745987559945236e-05,
+      "loss": 1.7648,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3289060613378411,
+      "grad_norm": 0.29763472080230713,
+      "learning_rate": 7.742751885625926e-05,
+      "loss": 1.8365,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3291475489012316,
+      "grad_norm": 0.2970298230648041,
+      "learning_rate": 7.739514567456712e-05,
+      "loss": 1.8233,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3293890364646221,
+      "grad_norm": 0.28750601410865784,
+      "learning_rate": 7.736275607377859e-05,
+      "loss": 1.7402,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3296305240280126,
+      "grad_norm": 0.29735836386680603,
+      "learning_rate": 7.733035007330615e-05,
+      "loss": 1.5324,
+      "step": 1365
+    },
+    {
+      "epoch": 0.32987201159140306,
+      "grad_norm": 0.2967546284198761,
+      "learning_rate": 7.729792769257214e-05,
+      "loss": 1.7626,
+      "step": 1366
+    },
+    {
+      "epoch": 0.33011349915479354,
+      "grad_norm": 0.29205527901649475,
+      "learning_rate": 7.726548895100867e-05,
+      "loss": 1.7793,
+      "step": 1367
+    },
+    {
+      "epoch": 0.33035498671818403,
+      "grad_norm": 0.2876708507537842,
+      "learning_rate": 7.72330338680577e-05,
+      "loss": 1.8089,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3305964742815745,
+      "grad_norm": 0.3068663477897644,
+      "learning_rate": 7.720056246317096e-05,
+      "loss": 1.918,
+      "step": 1369
+    },
+    {
+      "epoch": 0.330837961844965,
+      "grad_norm": 0.29179996252059937,
+      "learning_rate": 7.716807475580997e-05,
+      "loss": 1.6605,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3310794494083555,
+      "grad_norm": 0.30181172490119934,
+      "learning_rate": 7.713557076544601e-05,
+      "loss": 1.6363,
+      "step": 1371
+    },
+    {
+      "epoch": 0.33132093697174597,
+      "grad_norm": 0.29619401693344116,
+      "learning_rate": 7.710305051156015e-05,
+      "loss": 1.7194,
+      "step": 1372
+    },
+    {
+      "epoch": 0.33156242453513646,
+      "grad_norm": 0.2949178218841553,
+      "learning_rate": 7.707051401364318e-05,
+      "loss": 1.803,
+      "step": 1373
+    },
+    {
+      "epoch": 0.33180391209852694,
+      "grad_norm": 0.2937702536582947,
+      "learning_rate": 7.70379612911956e-05,
+      "loss": 1.7161,
+      "step": 1374
+    },
+    {
+      "epoch": 0.33204539966191743,
+      "grad_norm": 0.30445921421051025,
+      "learning_rate": 7.70053923637277e-05,
+      "loss": 1.8072,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3322868872253079,
+      "grad_norm": 0.3168744146823883,
+      "learning_rate": 7.697280725075944e-05,
+      "loss": 1.7361,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3325283747886984,
+      "grad_norm": 0.31621459126472473,
+      "learning_rate": 7.694020597182048e-05,
+      "loss": 1.763,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3327698623520889,
+      "grad_norm": 0.2877269685268402,
+      "learning_rate": 7.690758854645018e-05,
+      "loss": 1.7322,
+      "step": 1378
+    },
+    {
+      "epoch": 0.33301134991547937,
+      "grad_norm": 0.2892703115940094,
+      "learning_rate": 7.687495499419757e-05,
+      "loss": 1.6159,
+      "step": 1379
+    },
+    {
+      "epoch": 0.33325283747886986,
+      "grad_norm": 0.29529622197151184,
+      "learning_rate": 7.684230533462138e-05,
+      "loss": 1.7719,
+      "step": 1380
+    },
+    {
+      "epoch": 0.33349432504226034,
+      "grad_norm": 0.2938997745513916,
+      "learning_rate": 7.680963958728993e-05,
+      "loss": 1.8697,
+      "step": 1381
+    },
+    {
+      "epoch": 0.33373581260565083,
+      "grad_norm": 0.2826845645904541,
+      "learning_rate": 7.67769577717812e-05,
+      "loss": 1.7505,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3339773001690413,
+      "grad_norm": 0.29490166902542114,
+      "learning_rate": 7.674425990768288e-05,
+      "loss": 1.8577,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3342187877324318,
+      "grad_norm": 0.29277679324150085,
+      "learning_rate": 7.671154601459215e-05,
+      "loss": 1.6305,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3344602752958223,
+      "grad_norm": 0.28395023941993713,
+      "learning_rate": 7.667881611211592e-05,
+      "loss": 1.7382,
+      "step": 1385
+    },
+    {
+      "epoch": 0.33470176285921277,
+      "grad_norm": 0.30212682485580444,
+      "learning_rate": 7.664607021987058e-05,
+      "loss": 1.8324,
+      "step": 1386
+    },
+    {
+      "epoch": 0.33494325042260326,
+      "grad_norm": 0.28818273544311523,
+      "learning_rate": 7.661330835748219e-05,
+      "loss": 1.8119,
+      "step": 1387
+    },
+    {
+      "epoch": 0.33518473798599374,
+      "grad_norm": 0.30034732818603516,
+      "learning_rate": 7.658053054458636e-05,
+      "loss": 1.7632,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3354262255493842,
+      "grad_norm": 0.2828637659549713,
+      "learning_rate": 7.654773680082823e-05,
+      "loss": 1.734,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3356677131127747,
+      "grad_norm": 0.2823232114315033,
+      "learning_rate": 7.651492714586253e-05,
+      "loss": 1.7447,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3359092006761652,
+      "grad_norm": 0.3083249032497406,
+      "learning_rate": 7.64821015993535e-05,
+      "loss": 1.855,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3361506882395557,
+      "grad_norm": 0.30150431394577026,
+      "learning_rate": 7.644926018097491e-05,
+      "loss": 1.8758,
+      "step": 1392
+    },
+    {
+      "epoch": 0.33639217580294617,
+      "grad_norm": 0.3692246079444885,
+      "learning_rate": 7.641640291041004e-05,
+      "loss": 1.7376,
+      "step": 1393
+    },
+    {
+      "epoch": 0.33663366336633666,
+      "grad_norm": 0.28748929500579834,
+      "learning_rate": 7.638352980735167e-05,
+      "loss": 1.6032,
+      "step": 1394
+    },
+    {
+      "epoch": 0.33687515092972714,
+      "grad_norm": 0.30165570974349976,
+      "learning_rate": 7.635064089150209e-05,
+      "loss": 1.8655,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3371166384931176,
+      "grad_norm": 0.280122309923172,
+      "learning_rate": 7.631773618257301e-05,
+      "loss": 1.5955,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3373581260565081,
+      "grad_norm": 0.30419206619262695,
+      "learning_rate": 7.62848157002857e-05,
+      "loss": 1.8109,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3375996136198986,
+      "grad_norm": 0.3030405342578888,
+      "learning_rate": 7.62518794643708e-05,
+      "loss": 1.6872,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3378411011832891,
+      "grad_norm": 0.29341575503349304,
+      "learning_rate": 7.62189274945684e-05,
+      "loss": 1.7672,
+      "step": 1399
+    },
+    {
+      "epoch": 0.33808258874667957,
+      "grad_norm": 0.28583914041519165,
+      "learning_rate": 7.61859598106281e-05,
+      "loss": 1.6706,
+      "step": 1400
+    },
+    {
+      "epoch": 0.33832407631007005,
+      "grad_norm": 0.2980149984359741,
+      "learning_rate": 7.615297643230883e-05,
+      "loss": 1.7647,
+      "step": 1401
+    },
+    {
+      "epoch": 0.33856556387346054,
+      "grad_norm": 0.2835332453250885,
+      "learning_rate": 7.611997737937895e-05,
+      "loss": 1.6417,
+      "step": 1402
+    },
+    {
+      "epoch": 0.338807051436851,
+      "grad_norm": 0.29867133498191833,
+      "learning_rate": 7.608696267161621e-05,
+      "loss": 1.7539,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3390485390002415,
+      "grad_norm": 0.2891360819339752,
+      "learning_rate": 7.60539323288078e-05,
+      "loss": 1.6868,
+      "step": 1404
+    },
+    {
+      "epoch": 0.339290026563632,
+      "grad_norm": 0.2839174270629883,
+      "learning_rate": 7.602088637075016e-05,
+      "loss": 1.7187,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3395315141270225,
+      "grad_norm": 0.28129449486732483,
+      "learning_rate": 7.598782481724923e-05,
+      "loss": 1.5241,
+      "step": 1406
+    },
+    {
+      "epoch": 0.33977300169041297,
+      "grad_norm": 0.30046090483665466,
+      "learning_rate": 7.595474768812021e-05,
+      "loss": 1.7827,
+      "step": 1407
+    },
+    {
+      "epoch": 0.34001448925380345,
+      "grad_norm": 0.3015134632587433,
+      "learning_rate": 7.592165500318761e-05,
+      "loss": 1.656,
+      "step": 1408
+    },
+    {
+      "epoch": 0.34025597681719394,
+      "grad_norm": 0.28390491008758545,
+      "learning_rate": 7.588854678228539e-05,
+      "loss": 1.5548,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3404974643805844,
+      "grad_norm": 0.29424551129341125,
+      "learning_rate": 7.585542304525667e-05,
+      "loss": 1.5973,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3407389519439749,
+      "grad_norm": 0.3038700819015503,
+      "learning_rate": 7.5822283811954e-05,
+      "loss": 1.7428,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3409804395073654,
+      "grad_norm": 0.2812209129333496,
+      "learning_rate": 7.57891291022391e-05,
+      "loss": 1.603,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3412219270707559,
+      "grad_norm": 0.28015992045402527,
+      "learning_rate": 7.575595893598304e-05,
+      "loss": 1.6477,
+      "step": 1413
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 0.2998206913471222,
+      "learning_rate": 7.572277333306614e-05,
+      "loss": 1.7594,
+      "step": 1414
+    },
+    {
+      "epoch": 0.34170490219753685,
+      "grad_norm": 0.3038783073425293,
+      "learning_rate": 7.568957231337799e-05,
+      "loss": 1.9082,
+      "step": 1415
+    },
+    {
+      "epoch": 0.34194638976092734,
+      "grad_norm": 0.2797698378562927,
+      "learning_rate": 7.565635589681737e-05,
+      "loss": 1.6403,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3421878773243178,
+      "grad_norm": 0.2895340025424957,
+      "learning_rate": 7.56231241032923e-05,
+      "loss": 1.7442,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3424293648877083,
+      "grad_norm": 0.28286561369895935,
+      "learning_rate": 7.558987695272009e-05,
+      "loss": 1.8109,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3426708524510988,
+      "grad_norm": 0.2754989564418793,
+      "learning_rate": 7.555661446502714e-05,
+      "loss": 1.6567,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3429123400144893,
+      "grad_norm": 0.2939462959766388,
+      "learning_rate": 7.552333666014913e-05,
+      "loss": 1.7232,
+      "step": 1420
+    },
+    {
+      "epoch": 0.34315382757787977,
+      "grad_norm": 0.3267726004123688,
+      "learning_rate": 7.549004355803086e-05,
+      "loss": 1.9031,
+      "step": 1421
+    },
+    {
+      "epoch": 0.34339531514127025,
+      "grad_norm": 0.3062046766281128,
+      "learning_rate": 7.545673517862637e-05,
+      "loss": 1.8201,
+      "step": 1422
+    },
+    {
+      "epoch": 0.34363680270466074,
+      "grad_norm": 0.2913348376750946,
+      "learning_rate": 7.542341154189878e-05,
+      "loss": 1.6796,
+      "step": 1423
+    },
+    {
+      "epoch": 0.34387829026805117,
+      "grad_norm": 0.29292842745780945,
+      "learning_rate": 7.539007266782039e-05,
+      "loss": 1.8692,
+      "step": 1424
+    },
+    {
+      "epoch": 0.34411977783144165,
+      "grad_norm": 0.2910190224647522,
+      "learning_rate": 7.535671857637265e-05,
+      "loss": 1.7212,
+      "step": 1425
+    },
+    {
+      "epoch": 0.34436126539483214,
+      "grad_norm": 0.30508074164390564,
+      "learning_rate": 7.532334928754608e-05,
+      "loss": 1.6877,
+      "step": 1426
+    },
+    {
+      "epoch": 0.3446027529582226,
+      "grad_norm": 0.30790939927101135,
+      "learning_rate": 7.528996482134038e-05,
+      "loss": 1.8494,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3448442405216131,
+      "grad_norm": 0.30237263441085815,
+      "learning_rate": 7.525656519776427e-05,
+      "loss": 1.7703,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3450857280850036,
+      "grad_norm": 0.2987865209579468,
+      "learning_rate": 7.522315043683559e-05,
+      "loss": 1.7415,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3453272156483941,
+      "grad_norm": 0.2867589294910431,
+      "learning_rate": 7.518972055858127e-05,
+      "loss": 1.7095,
+      "step": 1430
+    },
+    {
+      "epoch": 0.34556870321178457,
+      "grad_norm": 0.2851780652999878,
+      "learning_rate": 7.515627558303728e-05,
+      "loss": 1.6923,
+      "step": 1431
+    },
+    {
+      "epoch": 0.34581019077517505,
+      "grad_norm": 0.3218798339366913,
+      "learning_rate": 7.512281553024863e-05,
+      "loss": 1.9602,
+      "step": 1432
+    },
+    {
+      "epoch": 0.34605167833856554,
+      "grad_norm": 0.30771544575691223,
+      "learning_rate": 7.508934042026938e-05,
+      "loss": 1.968,
+      "step": 1433
+    },
+    {
+      "epoch": 0.346293165901956,
+      "grad_norm": 0.28604716062545776,
+      "learning_rate": 7.505585027316265e-05,
+      "loss": 1.7667,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3465346534653465,
+      "grad_norm": 0.29297733306884766,
+      "learning_rate": 7.502234510900046e-05,
+      "loss": 1.7588,
+      "step": 1435
+    },
+    {
+      "epoch": 0.346776141028737,
+      "grad_norm": 0.29484325647354126,
+      "learning_rate": 7.498882494786396e-05,
+      "loss": 1.8345,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3470176285921275,
+      "grad_norm": 0.30629557371139526,
+      "learning_rate": 7.495528980984324e-05,
+      "loss": 1.8114,
+      "step": 1437
+    },
+    {
+      "epoch": 0.34725911615551797,
+      "grad_norm": 0.2843991219997406,
+      "learning_rate": 7.492173971503732e-05,
+      "loss": 1.7903,
+      "step": 1438
+    },
+    {
+      "epoch": 0.34750060371890845,
+      "grad_norm": 0.29854124784469604,
+      "learning_rate": 7.488817468355426e-05,
+      "loss": 1.7366,
+      "step": 1439
+    },
+    {
+      "epoch": 0.34774209128229894,
+      "grad_norm": 0.2846215069293976,
+      "learning_rate": 7.485459473551101e-05,
+      "loss": 1.6711,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3479835788456894,
+      "grad_norm": 0.2946016192436218,
+      "learning_rate": 7.482099989103349e-05,
+      "loss": 1.7971,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3482250664090799,
+      "grad_norm": 0.2970103621482849,
+      "learning_rate": 7.478739017025654e-05,
+      "loss": 1.8869,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3484665539724704,
+      "grad_norm": 0.30067914724349976,
+      "learning_rate": 7.475376559332396e-05,
+      "loss": 1.6867,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3487080415358609,
+      "grad_norm": 0.2954392731189728,
+      "learning_rate": 7.472012618038835e-05,
+      "loss": 1.8272,
+      "step": 1444
+    },
+    {
+      "epoch": 0.34894952909925137,
+      "grad_norm": 0.29047641158103943,
+      "learning_rate": 7.468647195161132e-05,
+      "loss": 1.6909,
+      "step": 1445
+    },
+    {
+      "epoch": 0.34919101666264185,
+      "grad_norm": 0.2894633710384369,
+      "learning_rate": 7.465280292716329e-05,
+      "loss": 1.747,
+      "step": 1446
+    },
+    {
+      "epoch": 0.34943250422603234,
+      "grad_norm": 0.28781044483184814,
+      "learning_rate": 7.461911912722355e-05,
+      "loss": 1.5743,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3496739917894228,
+      "grad_norm": 0.30114755034446716,
+      "learning_rate": 7.458542057198027e-05,
+      "loss": 1.8015,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3499154793528133,
+      "grad_norm": 0.2965746819972992,
+      "learning_rate": 7.455170728163045e-05,
+      "loss": 1.7565,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3501569669162038,
+      "grad_norm": 0.32243311405181885,
+      "learning_rate": 7.451797927637992e-05,
+      "loss": 1.8916,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3503984544795943,
+      "grad_norm": 0.3025962710380554,
+      "learning_rate": 7.448423657644336e-05,
+      "loss": 1.6573,
+      "step": 1451
+    },
+    {
+      "epoch": 0.35063994204298476,
+      "grad_norm": 0.2864728271961212,
+      "learning_rate": 7.445047920204418e-05,
+      "loss": 1.6866,
+      "step": 1452
+    },
+    {
+      "epoch": 0.35088142960637525,
+      "grad_norm": 0.3004949986934662,
+      "learning_rate": 7.441670717341466e-05,
+      "loss": 1.7636,
+      "step": 1453
+    },
+    {
+      "epoch": 0.35112291716976574,
+      "grad_norm": 0.2901725471019745,
+      "learning_rate": 7.438292051079588e-05,
+      "loss": 1.6909,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3513644047331562,
+      "grad_norm": 0.3008805513381958,
+      "learning_rate": 7.434911923443757e-05,
+      "loss": 1.8379,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3516058922965467,
+      "grad_norm": 0.2823113799095154,
+      "learning_rate": 7.431530336459837e-05,
+      "loss": 1.7156,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3518473798599372,
+      "grad_norm": 0.2956325113773346,
+      "learning_rate": 7.428147292154554e-05,
+      "loss": 1.819,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3520888674233277,
+      "grad_norm": 0.2923109233379364,
+      "learning_rate": 7.424762792555516e-05,
+      "loss": 1.7036,
+      "step": 1458
+    },
+    {
+      "epoch": 0.35233035498671816,
+      "grad_norm": 0.28335970640182495,
+      "learning_rate": 7.4213768396912e-05,
+      "loss": 1.6259,
+      "step": 1459
+    },
+    {
+      "epoch": 0.35257184255010865,
+      "grad_norm": 0.29262620210647583,
+      "learning_rate": 7.417989435590953e-05,
+      "loss": 1.6768,
+      "step": 1460
+    },
+    {
+      "epoch": 0.35281333011349914,
+      "grad_norm": 0.3107213079929352,
+      "learning_rate": 7.414600582284992e-05,
+      "loss": 1.9349,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3530548176768896,
+      "grad_norm": 0.2939004600048065,
+      "learning_rate": 7.411210281804407e-05,
+      "loss": 1.7732,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3532963052402801,
+      "grad_norm": 0.28268906474113464,
+      "learning_rate": 7.407818536181148e-05,
+      "loss": 1.6477,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3535377928036706,
+      "grad_norm": 0.2945854365825653,
+      "learning_rate": 7.404425347448036e-05,
+      "loss": 1.756,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3537792803670611,
+      "grad_norm": 0.29590821266174316,
+      "learning_rate": 7.401030717638758e-05,
+      "loss": 1.716,
+      "step": 1465
+    },
+    {
+      "epoch": 0.35402076793045156,
+      "grad_norm": 0.278075248003006,
+      "learning_rate": 7.397634648787859e-05,
+      "loss": 1.6162,
+      "step": 1466
+    },
+    {
+      "epoch": 0.35426225549384205,
+      "grad_norm": 0.3146173655986786,
+      "learning_rate": 7.394237142930751e-05,
+      "loss": 1.794,
+      "step": 1467
+    },
+    {
+      "epoch": 0.35450374305723253,
+      "grad_norm": 0.28644752502441406,
+      "learning_rate": 7.390838202103709e-05,
+      "loss": 1.7197,
+      "step": 1468
+    },
+    {
+      "epoch": 0.354745230620623,
+      "grad_norm": 0.3296287953853607,
+      "learning_rate": 7.387437828343864e-05,
+      "loss": 1.6563,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3549867181840135,
+      "grad_norm": 0.2927229106426239,
+      "learning_rate": 7.384036023689204e-05,
+      "loss": 1.783,
+      "step": 1470
+    },
+    {
+      "epoch": 0.355228205747404,
+      "grad_norm": 0.3050139546394348,
+      "learning_rate": 7.380632790178583e-05,
+      "loss": 1.6854,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3554696933107945,
+      "grad_norm": 0.3071853816509247,
+      "learning_rate": 7.377228129851703e-05,
+      "loss": 1.8396,
+      "step": 1472
+    },
+    {
+      "epoch": 0.35571118087418496,
+      "grad_norm": 0.3828865587711334,
+      "learning_rate": 7.373822044749124e-05,
+      "loss": 1.6513,
+      "step": 1473
+    },
+    {
+      "epoch": 0.35595266843757545,
+      "grad_norm": 0.31001341342926025,
+      "learning_rate": 7.370414536912261e-05,
+      "loss": 1.9578,
+      "step": 1474
+    },
+    {
+      "epoch": 0.35619415600096593,
+      "grad_norm": 0.3114062249660492,
+      "learning_rate": 7.367005608383382e-05,
+      "loss": 1.8608,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3564356435643564,
+      "grad_norm": 0.29749569296836853,
+      "learning_rate": 7.363595261205602e-05,
+      "loss": 1.8161,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3566771311277469,
+      "grad_norm": 0.3024919629096985,
+      "learning_rate": 7.360183497422895e-05,
+      "loss": 1.8777,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3569186186911374,
+      "grad_norm": 0.2933084964752197,
+      "learning_rate": 7.356770319080074e-05,
+      "loss": 1.685,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3571601062545279,
+      "grad_norm": 0.30307063460350037,
+      "learning_rate": 7.353355728222804e-05,
+      "loss": 1.8085,
+      "step": 1479
+    },
+    {
+      "epoch": 0.35740159381791836,
+      "grad_norm": 0.3066220283508301,
+      "learning_rate": 7.349939726897604e-05,
+      "loss": 1.7865,
+      "step": 1480
+    },
+    {
+      "epoch": 0.35764308138130885,
+      "grad_norm": 0.30603402853012085,
+      "learning_rate": 7.346522317151823e-05,
+      "loss": 1.8973,
+      "step": 1481
+    },
+    {
+      "epoch": 0.35788456894469933,
+      "grad_norm": 0.3032185137271881,
+      "learning_rate": 7.343103501033667e-05,
+      "loss": 1.7836,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3581260565080898,
+      "grad_norm": 0.2852492332458496,
+      "learning_rate": 7.33968328059218e-05,
+      "loss": 1.7213,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3583675440714803,
+      "grad_norm": 0.2975196838378906,
+      "learning_rate": 7.336261657877247e-05,
+      "loss": 1.8799,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3586090316348708,
+      "grad_norm": 0.2893730103969574,
+      "learning_rate": 7.332838634939597e-05,
+      "loss": 1.7396,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3588505191982613,
+      "grad_norm": 0.28530189394950867,
+      "learning_rate": 7.329414213830793e-05,
+      "loss": 1.7311,
+      "step": 1486
+    },
+    {
+      "epoch": 0.35909200676165176,
+      "grad_norm": 0.3011293411254883,
+      "learning_rate": 7.325988396603241e-05,
+      "loss": 1.8305,
+      "step": 1487
+    },
+    {
+      "epoch": 0.35933349432504225,
+      "grad_norm": 0.30045387148857117,
+      "learning_rate": 7.322561185310178e-05,
+      "loss": 1.7941,
+      "step": 1488
+    },
+    {
+      "epoch": 0.35957498188843273,
+      "grad_norm": 0.30552494525909424,
+      "learning_rate": 7.319132582005685e-05,
+      "loss": 1.624,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3598164694518232,
+      "grad_norm": 0.3145065903663635,
+      "learning_rate": 7.315702588744668e-05,
+      "loss": 1.8659,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3600579570152137,
+      "grad_norm": 0.28973329067230225,
+      "learning_rate": 7.312271207582873e-05,
+      "loss": 1.8122,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3602994445786042,
+      "grad_norm": 0.2974574863910675,
+      "learning_rate": 7.308838440576875e-05,
+      "loss": 1.9315,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3605409321419947,
+      "grad_norm": 0.310698539018631,
+      "learning_rate": 7.305404289784079e-05,
+      "loss": 2.0308,
+      "step": 1493
+    },
+    {
+      "epoch": 0.36078241970538516,
+      "grad_norm": 0.27670934796333313,
+      "learning_rate": 7.301968757262721e-05,
+      "loss": 1.6081,
+      "step": 1494
+    },
+    {
+      "epoch": 0.36102390726877565,
+      "grad_norm": 0.27521172165870667,
+      "learning_rate": 7.298531845071861e-05,
+      "loss": 1.6248,
+      "step": 1495
+    },
+    {
+      "epoch": 0.36126539483216613,
+      "grad_norm": 0.2828007638454437,
+      "learning_rate": 7.295093555271395e-05,
+      "loss": 1.7688,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3615068823955566,
+      "grad_norm": 0.29078638553619385,
+      "learning_rate": 7.291653889922035e-05,
+      "loss": 1.7569,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3617483699589471,
+      "grad_norm": 0.2831771969795227,
+      "learning_rate": 7.288212851085322e-05,
+      "loss": 1.6375,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3619898575223376,
+      "grad_norm": 0.307668000459671,
+      "learning_rate": 7.28477044082362e-05,
+      "loss": 1.9078,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3622313450857281,
+      "grad_norm": 0.2992483675479889,
+      "learning_rate": 7.281326661200114e-05,
+      "loss": 1.8321,
+      "step": 1500
+    },
+    {
+      "epoch": 0.36247283264911856,
+      "grad_norm": 0.3095793128013611,
+      "learning_rate": 7.277881514278811e-05,
+      "loss": 1.8003,
+      "step": 1501
+    },
+    {
+      "epoch": 0.36271432021250904,
+      "grad_norm": 0.29581761360168457,
+      "learning_rate": 7.274435002124534e-05,
+      "loss": 1.8205,
+      "step": 1502
+    },
+    {
+      "epoch": 0.36295580777589953,
+      "grad_norm": 0.2918574810028076,
+      "learning_rate": 7.27098712680293e-05,
+      "loss": 1.7362,
+      "step": 1503
+    },
+    {
+      "epoch": 0.36319729533929,
+      "grad_norm": 0.29628437757492065,
+      "learning_rate": 7.267537890380459e-05,
+      "loss": 1.6339,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3634387829026805,
+      "grad_norm": 0.2891917824745178,
+      "learning_rate": 7.264087294924397e-05,
+      "loss": 1.7657,
+      "step": 1505
+    },
+    {
+      "epoch": 0.363680270466071,
+      "grad_norm": 0.28467097878456116,
+      "learning_rate": 7.260635342502837e-05,
+      "loss": 1.6824,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3639217580294615,
+      "grad_norm": 0.2941914200782776,
+      "learning_rate": 7.25718203518468e-05,
+      "loss": 1.6893,
+      "step": 1507
+    },
+    {
+      "epoch": 0.36416324559285196,
+      "grad_norm": 0.28581947088241577,
+      "learning_rate": 7.253727375039646e-05,
+      "loss": 1.7599,
+      "step": 1508
+    },
+    {
+      "epoch": 0.36440473315624244,
+      "grad_norm": 0.29284581542015076,
+      "learning_rate": 7.250271364138259e-05,
+      "loss": 1.8434,
+      "step": 1509
+    },
+    {
+      "epoch": 0.36464622071963293,
+      "grad_norm": 0.3031042516231537,
+      "learning_rate": 7.24681400455186e-05,
+      "loss": 1.7305,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3648877082830234,
+      "grad_norm": 0.29574331641197205,
+      "learning_rate": 7.24335529835259e-05,
+      "loss": 1.6791,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3651291958464139,
+      "grad_norm": 0.3356575667858124,
+      "learning_rate": 7.239895247613404e-05,
+      "loss": 1.99,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3653706834098044,
+      "grad_norm": 0.28749579191207886,
+      "learning_rate": 7.236433854408061e-05,
+      "loss": 1.564,
+      "step": 1513
+    },
+    {
+      "epoch": 0.36561217097319487,
+      "grad_norm": 0.2926652729511261,
+      "learning_rate": 7.232971120811121e-05,
+      "loss": 1.7391,
+      "step": 1514
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 0.28827694058418274,
+      "learning_rate": 7.229507048897952e-05,
+      "loss": 1.7495,
+      "step": 1515
+    },
+    {
+      "epoch": 0.36609514609997584,
+      "grad_norm": 0.29912760853767395,
+      "learning_rate": 7.226041640744721e-05,
+      "loss": 1.6743,
+      "step": 1516
+    },
+    {
+      "epoch": 0.36633663366336633,
+      "grad_norm": 0.29889270663261414,
+      "learning_rate": 7.222574898428398e-05,
+      "loss": 1.8151,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3665781212267568,
+      "grad_norm": 0.30631643533706665,
+      "learning_rate": 7.219106824026751e-05,
+      "loss": 1.879,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3668196087901473,
+      "grad_norm": 0.303406298160553,
+      "learning_rate": 7.215637419618347e-05,
+      "loss": 1.6758,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3670610963535378,
+      "grad_norm": 0.2735936939716339,
+      "learning_rate": 7.212166687282551e-05,
+      "loss": 1.6312,
+      "step": 1520
+    },
+    {
+      "epoch": 0.36730258391692827,
+      "grad_norm": 0.2837960720062256,
+      "learning_rate": 7.208694629099523e-05,
+      "loss": 1.6888,
+      "step": 1521
+    },
+    {
+      "epoch": 0.36754407148031876,
+      "grad_norm": 0.2823384404182434,
+      "learning_rate": 7.205221247150218e-05,
+      "loss": 1.7177,
+      "step": 1522
+    },
+    {
+      "epoch": 0.36778555904370924,
+      "grad_norm": 0.2750958204269409,
+      "learning_rate": 7.201746543516382e-05,
+      "loss": 1.5654,
+      "step": 1523
+    },
+    {
+      "epoch": 0.36802704660709973,
+      "grad_norm": 0.29475295543670654,
+      "learning_rate": 7.198270520280559e-05,
+      "loss": 1.7343,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3682685341704902,
+      "grad_norm": 0.26865559816360474,
+      "learning_rate": 7.194793179526078e-05,
+      "loss": 1.5128,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3685100217338807,
+      "grad_norm": 0.28918734192848206,
+      "learning_rate": 7.191314523337058e-05,
+      "loss": 1.8486,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3687515092972712,
+      "grad_norm": 0.279786616563797,
+      "learning_rate": 7.18783455379841e-05,
+      "loss": 1.7636,
+      "step": 1527
+    },
+    {
+      "epoch": 0.36899299686066167,
+      "grad_norm": 0.289742648601532,
+      "learning_rate": 7.184353272995831e-05,
+      "loss": 1.7157,
+      "step": 1528
+    },
+    {
+      "epoch": 0.36923448442405216,
+      "grad_norm": 0.2943384647369385,
+      "learning_rate": 7.180870683015802e-05,
+      "loss": 1.8054,
+      "step": 1529
+    },
+    {
+      "epoch": 0.36947597198744264,
+      "grad_norm": 0.2821480929851532,
+      "learning_rate": 7.177386785945591e-05,
+      "loss": 1.7366,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3697174595508331,
+      "grad_norm": 0.289713978767395,
+      "learning_rate": 7.173901583873248e-05,
+      "loss": 1.8236,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3699589471142236,
+      "grad_norm": 0.29103440046310425,
+      "learning_rate": 7.170415078887604e-05,
+      "loss": 1.6997,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3702004346776141,
+      "grad_norm": 0.28201615810394287,
+      "learning_rate": 7.166927273078275e-05,
+      "loss": 1.6628,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3704419222410046,
+      "grad_norm": 0.290673166513443,
+      "learning_rate": 7.163438168535652e-05,
+      "loss": 1.6033,
+      "step": 1534
+    },
+    {
+      "epoch": 0.37068340980439507,
+      "grad_norm": 0.28538063168525696,
+      "learning_rate": 7.159947767350907e-05,
+      "loss": 1.6938,
+      "step": 1535
+    },
+    {
+      "epoch": 0.37092489736778556,
+      "grad_norm": 0.2745797634124756,
+      "learning_rate": 7.156456071615989e-05,
+      "loss": 1.6851,
+      "step": 1536
+    },
+    {
+      "epoch": 0.37116638493117604,
+      "grad_norm": 0.2976955473423004,
+      "learning_rate": 7.15296308342362e-05,
+      "loss": 1.8967,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3714078724945665,
+      "grad_norm": 0.27819645404815674,
+      "learning_rate": 7.149468804867303e-05,
+      "loss": 1.6837,
+      "step": 1538
+    },
+    {
+      "epoch": 0.371649360057957,
+      "grad_norm": 0.2899627387523651,
+      "learning_rate": 7.145973238041306e-05,
+      "loss": 1.7605,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3718908476213475,
+      "grad_norm": 0.2979569733142853,
+      "learning_rate": 7.142476385040677e-05,
+      "loss": 1.7937,
+      "step": 1540
+    },
+    {
+      "epoch": 0.372132335184738,
+      "grad_norm": 0.28263163566589355,
+      "learning_rate": 7.138978247961228e-05,
+      "loss": 1.7896,
+      "step": 1541
+    },
+    {
+      "epoch": 0.37237382274812847,
+      "grad_norm": 0.28612080216407776,
+      "learning_rate": 7.135478828899546e-05,
+      "loss": 1.6894,
+      "step": 1542
+    },
+    {
+      "epoch": 0.37261531031151895,
+      "grad_norm": 0.29869383573532104,
+      "learning_rate": 7.131978129952985e-05,
+      "loss": 1.658,
+      "step": 1543
+    },
+    {
+      "epoch": 0.37285679787490944,
+      "grad_norm": 0.29021137952804565,
+      "learning_rate": 7.128476153219665e-05,
+      "loss": 1.8043,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3730982854382999,
+      "grad_norm": 0.29576247930526733,
+      "learning_rate": 7.124972900798471e-05,
+      "loss": 1.8886,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3733397730016904,
+      "grad_norm": 0.30245786905288696,
+      "learning_rate": 7.121468374789055e-05,
+      "loss": 1.6748,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3735812605650809,
+      "grad_norm": 0.2910307049751282,
+      "learning_rate": 7.11796257729183e-05,
+      "loss": 1.6754,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3738227481284714,
+      "grad_norm": 0.3084256052970886,
+      "learning_rate": 7.114455510407974e-05,
+      "loss": 1.7797,
+      "step": 1548
+    },
+    {
+      "epoch": 0.37406423569186187,
+      "grad_norm": 0.29425761103630066,
+      "learning_rate": 7.11094717623942e-05,
+      "loss": 1.657,
+      "step": 1549
+    },
+    {
+      "epoch": 0.37430572325525235,
+      "grad_norm": 0.29802206158638,
+      "learning_rate": 7.107437576888869e-05,
+      "loss": 1.7229,
+      "step": 1550
+    },
+    {
+      "epoch": 0.37454721081864284,
+      "grad_norm": 0.3134438693523407,
+      "learning_rate": 7.103926714459774e-05,
+      "loss": 1.8207,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3747886983820333,
+      "grad_norm": 0.2775018513202667,
+      "learning_rate": 7.100414591056347e-05,
+      "loss": 1.6175,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3750301859454238,
+      "grad_norm": 0.2902098000049591,
+      "learning_rate": 7.096901208783554e-05,
+      "loss": 1.7434,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3752716735088143,
+      "grad_norm": 0.28874725103378296,
+      "learning_rate": 7.093386569747121e-05,
+      "loss": 1.6254,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3755131610722048,
+      "grad_norm": 0.2993185520172119,
+      "learning_rate": 7.089870676053519e-05,
+      "loss": 1.738,
+      "step": 1555
+    },
+    {
+      "epoch": 0.37575464863559527,
+      "grad_norm": 0.29154321551322937,
+      "learning_rate": 7.086353529809975e-05,
+      "loss": 1.8475,
+      "step": 1556
+    },
+    {
+      "epoch": 0.37599613619898575,
+      "grad_norm": 0.2712036073207855,
+      "learning_rate": 7.082835133124472e-05,
+      "loss": 1.6923,
+      "step": 1557
+    },
+    {
+      "epoch": 0.37623762376237624,
+      "grad_norm": 0.2821873128414154,
+      "learning_rate": 7.079315488105732e-05,
+      "loss": 1.652,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3764791113257667,
+      "grad_norm": 0.29745790362358093,
+      "learning_rate": 7.075794596863234e-05,
+      "loss": 1.7248,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3767205988891572,
+      "grad_norm": 0.2806411385536194,
+      "learning_rate": 7.072272461507198e-05,
+      "loss": 1.6761,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3769620864525477,
+      "grad_norm": 0.2909418046474457,
+      "learning_rate": 7.068749084148596e-05,
+      "loss": 1.6984,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3772035740159382,
+      "grad_norm": 0.29692786931991577,
+      "learning_rate": 7.065224466899137e-05,
+      "loss": 1.7558,
+      "step": 1562
+    },
+    {
+      "epoch": 0.37744506157932867,
+      "grad_norm": 0.3164198100566864,
+      "learning_rate": 7.061698611871276e-05,
+      "loss": 1.7037,
+      "step": 1563
+    },
+    {
+      "epoch": 0.37768654914271915,
+      "grad_norm": 0.28125908970832825,
+      "learning_rate": 7.058171521178214e-05,
+      "loss": 1.6706,
+      "step": 1564
+    },
+    {
+      "epoch": 0.37792803670610964,
+      "grad_norm": 0.30474957823753357,
+      "learning_rate": 7.054643196933885e-05,
+      "loss": 1.8193,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3781695242695001,
+      "grad_norm": 0.30764004588127136,
+      "learning_rate": 7.051113641252971e-05,
+      "loss": 1.8221,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3784110118328906,
+      "grad_norm": 0.27815955877304077,
+      "learning_rate": 7.047582856250885e-05,
+      "loss": 1.6797,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3786524993962811,
+      "grad_norm": 0.3051069974899292,
+      "learning_rate": 7.044050844043778e-05,
+      "loss": 1.8329,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3788939869596716,
+      "grad_norm": 0.29128777980804443,
+      "learning_rate": 7.040517606748541e-05,
+      "loss": 1.5598,
+      "step": 1569
+    },
+    {
+      "epoch": 0.37913547452306207,
+      "grad_norm": 0.3010561466217041,
+      "learning_rate": 7.036983146482794e-05,
+      "loss": 1.7792,
+      "step": 1570
+    },
+    {
+      "epoch": 0.37937696208645255,
+      "grad_norm": 0.32075226306915283,
+      "learning_rate": 7.033447465364891e-05,
+      "loss": 1.8827,
+      "step": 1571
+    },
+    {
+      "epoch": 0.37961844964984304,
+      "grad_norm": 0.29643532633781433,
+      "learning_rate": 7.029910565513919e-05,
+      "loss": 1.8178,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3798599372132335,
+      "grad_norm": 0.312639445066452,
+      "learning_rate": 7.026372449049699e-05,
+      "loss": 1.7738,
+      "step": 1573
+    },
+    {
+      "epoch": 0.380101424776624,
+      "grad_norm": 0.3031156659126282,
+      "learning_rate": 7.022833118092773e-05,
+      "loss": 1.7509,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3803429123400145,
+      "grad_norm": 0.28569579124450684,
+      "learning_rate": 7.019292574764415e-05,
+      "loss": 1.6701,
+      "step": 1575
+    },
+    {
+      "epoch": 0.380584399903405,
+      "grad_norm": 0.29795435070991516,
+      "learning_rate": 7.015750821186628e-05,
+      "loss": 1.7447,
+      "step": 1576
+    },
+    {
+      "epoch": 0.38082588746679547,
+      "grad_norm": 0.3067462742328644,
+      "learning_rate": 7.012207859482138e-05,
+      "loss": 1.8176,
+      "step": 1577
+    },
+    {
+      "epoch": 0.38106737503018595,
+      "grad_norm": 0.2866966426372528,
+      "learning_rate": 7.008663691774396e-05,
+      "loss": 1.64,
+      "step": 1578
+    },
+    {
+      "epoch": 0.38130886259357644,
+      "grad_norm": 0.28150928020477295,
+      "learning_rate": 7.005118320187573e-05,
+      "loss": 1.6103,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3815503501569669,
+      "grad_norm": 0.28614646196365356,
+      "learning_rate": 7.001571746846565e-05,
+      "loss": 1.7165,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3817918377203574,
+      "grad_norm": 0.29418009519577026,
+      "learning_rate": 6.998023973876987e-05,
+      "loss": 1.7223,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3820333252837479,
+      "grad_norm": 0.2841942310333252,
+      "learning_rate": 6.99447500340517e-05,
+      "loss": 1.5126,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3822748128471384,
+      "grad_norm": 0.30715563893318176,
+      "learning_rate": 6.990924837558171e-05,
+      "loss": 1.8462,
+      "step": 1583
+    },
+    {
+      "epoch": 0.38251630041052886,
+      "grad_norm": 0.2975330948829651,
+      "learning_rate": 6.987373478463754e-05,
+      "loss": 1.83,
+      "step": 1584
+    },
+    {
+      "epoch": 0.38275778797391935,
+      "grad_norm": 0.28202196955680847,
+      "learning_rate": 6.983820928250402e-05,
+      "loss": 1.6507,
+      "step": 1585
+    },
+    {
+      "epoch": 0.38299927553730984,
+      "grad_norm": 0.28683874011039734,
+      "learning_rate": 6.980267189047314e-05,
+      "loss": 1.6067,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3832407631007003,
+      "grad_norm": 0.2874414622783661,
+      "learning_rate": 6.976712262984401e-05,
+      "loss": 1.5705,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3834822506640908,
+      "grad_norm": 0.29029643535614014,
+      "learning_rate": 6.973156152192279e-05,
+      "loss": 1.6031,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3837237382274813,
+      "grad_norm": 0.2800596356391907,
+      "learning_rate": 6.969598858802285e-05,
+      "loss": 1.7132,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3839652257908718,
+      "grad_norm": 0.28613579273223877,
+      "learning_rate": 6.966040384946454e-05,
+      "loss": 1.7108,
+      "step": 1590
+    },
+    {
+      "epoch": 0.38420671335426226,
+      "grad_norm": 0.30830395221710205,
+      "learning_rate": 6.962480732757535e-05,
+      "loss": 1.6721,
+      "step": 1591
+    },
+    {
+      "epoch": 0.38444820091765275,
+      "grad_norm": 0.29725363850593567,
+      "learning_rate": 6.958919904368983e-05,
+      "loss": 1.8466,
+      "step": 1592
+    },
+    {
+      "epoch": 0.38468968848104323,
+      "grad_norm": 0.37713706493377686,
+      "learning_rate": 6.955357901914957e-05,
+      "loss": 1.9359,
+      "step": 1593
+    },
+    {
+      "epoch": 0.3849311760444337,
+      "grad_norm": 0.29880592226982117,
+      "learning_rate": 6.951794727530315e-05,
+      "loss": 1.6493,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3851726636078242,
+      "grad_norm": 0.29536154866218567,
+      "learning_rate": 6.948230383350626e-05,
+      "loss": 1.714,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3854141511712147,
+      "grad_norm": 0.2920708954334259,
+      "learning_rate": 6.944664871512154e-05,
+      "loss": 1.7116,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3856556387346052,
+      "grad_norm": 0.29399019479751587,
+      "learning_rate": 6.941098194151864e-05,
+      "loss": 1.708,
+      "step": 1597
+    },
+    {
+      "epoch": 0.38589712629799566,
+      "grad_norm": 0.29361873865127563,
+      "learning_rate": 6.93753035340742e-05,
+      "loss": 1.6959,
+      "step": 1598
+    },
+    {
+      "epoch": 0.38613861386138615,
+      "grad_norm": 0.28866901993751526,
+      "learning_rate": 6.933961351417186e-05,
+      "loss": 1.6782,
+      "step": 1599
+    },
+    {
+      "epoch": 0.38638010142477663,
+      "grad_norm": 0.2916489243507385,
+      "learning_rate": 6.930391190320217e-05,
+      "loss": 1.6748,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3866215889881671,
+      "grad_norm": 0.2900312840938568,
+      "learning_rate": 6.926819872256264e-05,
+      "loss": 1.7323,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3868630765515576,
+      "grad_norm": 0.2876428961753845,
+      "learning_rate": 6.923247399365776e-05,
+      "loss": 1.5876,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3871045641149481,
+      "grad_norm": 0.3108595609664917,
+      "learning_rate": 6.919673773789888e-05,
+      "loss": 1.7827,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3873460516783386,
+      "grad_norm": 0.30719828605651855,
+      "learning_rate": 6.91609899767043e-05,
+      "loss": 1.7054,
+      "step": 1604
+    },
+    {
+      "epoch": 0.38758753924172906,
+      "grad_norm": 0.2866917848587036,
+      "learning_rate": 6.912523073149917e-05,
+      "loss": 1.7833,
+      "step": 1605
+    },
+    {
+      "epoch": 0.38782902680511955,
+      "grad_norm": 0.28398698568344116,
+      "learning_rate": 6.908946002371559e-05,
+      "loss": 1.7662,
+      "step": 1606
+    },
+    {
+      "epoch": 0.38807051436851003,
+      "grad_norm": 0.30139845609664917,
+      "learning_rate": 6.905367787479248e-05,
+      "loss": 1.7251,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3883120019319005,
+      "grad_norm": 0.30310872197151184,
+      "learning_rate": 6.901788430617562e-05,
+      "loss": 1.6172,
+      "step": 1608
+    },
+    {
+      "epoch": 0.388553489495291,
+      "grad_norm": 0.30176272988319397,
+      "learning_rate": 6.898207933931764e-05,
+      "loss": 1.7503,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3887949770586815,
+      "grad_norm": 0.2845175266265869,
+      "learning_rate": 6.894626299567804e-05,
+      "loss": 1.7345,
+      "step": 1610
+    },
+    {
+      "epoch": 0.389036464622072,
+      "grad_norm": 0.29985424876213074,
+      "learning_rate": 6.891043529672308e-05,
+      "loss": 1.7745,
+      "step": 1611
+    },
+    {
+      "epoch": 0.38927795218546246,
+      "grad_norm": 0.2928183674812317,
+      "learning_rate": 6.887459626392585e-05,
+      "loss": 1.739,
+      "step": 1612
+    },
+    {
+      "epoch": 0.38951943974885295,
+      "grad_norm": 0.3028491735458374,
+      "learning_rate": 6.883874591876624e-05,
+      "loss": 1.7981,
+      "step": 1613
+    },
+    {
+      "epoch": 0.38976092731224343,
+      "grad_norm": 0.2912960350513458,
+      "learning_rate": 6.880288428273091e-05,
+      "loss": 1.7531,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3900024148756339,
+      "grad_norm": 0.29823577404022217,
+      "learning_rate": 6.876701137731329e-05,
+      "loss": 1.7662,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 0.283291220664978,
+      "learning_rate": 6.873112722401357e-05,
+      "loss": 1.6241,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3904853900024149,
+      "grad_norm": 0.2959177792072296,
+      "learning_rate": 6.869523184433867e-05,
+      "loss": 1.8359,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3907268775658054,
+      "grad_norm": 0.28464576601982117,
+      "learning_rate": 6.865932525980224e-05,
+      "loss": 1.8036,
+      "step": 1618
+    },
+    {
+      "epoch": 0.39096836512919586,
+      "grad_norm": 0.3036990463733673,
+      "learning_rate": 6.862340749192467e-05,
+      "loss": 1.9036,
+      "step": 1619
+    },
+    {
+      "epoch": 0.39120985269258635,
+      "grad_norm": 0.29203951358795166,
+      "learning_rate": 6.8587478562233e-05,
+      "loss": 1.8115,
+      "step": 1620
+    },
+    {
+      "epoch": 0.39145134025597683,
+      "grad_norm": 0.3313468098640442,
+      "learning_rate": 6.855153849226102e-05,
+      "loss": 1.7969,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3916928278193673,
+      "grad_norm": 0.27904316782951355,
+      "learning_rate": 6.851558730354917e-05,
+      "loss": 1.6888,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3919343153827578,
+      "grad_norm": 0.2998868227005005,
+      "learning_rate": 6.847962501764454e-05,
+      "loss": 1.7044,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3921758029461483,
+      "grad_norm": 0.29342302680015564,
+      "learning_rate": 6.844365165610089e-05,
+      "loss": 1.8111,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3924172905095388,
+      "grad_norm": 0.30202576518058777,
+      "learning_rate": 6.840766724047862e-05,
+      "loss": 1.7587,
+      "step": 1625
+    },
+    {
+      "epoch": 0.39265877807292926,
+      "grad_norm": 0.2904617488384247,
+      "learning_rate": 6.837167179234472e-05,
+      "loss": 1.7509,
+      "step": 1626
+    },
+    {
+      "epoch": 0.39290026563631975,
+      "grad_norm": 0.29565051198005676,
+      "learning_rate": 6.833566533327285e-05,
+      "loss": 1.7288,
+      "step": 1627
+    },
+    {
+      "epoch": 0.39314175319971023,
+      "grad_norm": 0.28997743129730225,
+      "learning_rate": 6.829964788484322e-05,
+      "loss": 1.6617,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3933832407631007,
+      "grad_norm": 0.3013928532600403,
+      "learning_rate": 6.826361946864266e-05,
+      "loss": 1.6874,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3936247283264912,
+      "grad_norm": 0.30889344215393066,
+      "learning_rate": 6.822758010626456e-05,
+      "loss": 1.7683,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3938662158898817,
+      "grad_norm": 0.30394911766052246,
+      "learning_rate": 6.819152981930883e-05,
+      "loss": 1.6567,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3941077034532722,
+      "grad_norm": 0.3007825016975403,
+      "learning_rate": 6.815546862938202e-05,
+      "loss": 1.8263,
+      "step": 1632
+    },
+    {
+      "epoch": 0.39434919101666266,
+      "grad_norm": 0.2870945632457733,
+      "learning_rate": 6.811939655809715e-05,
+      "loss": 1.7453,
+      "step": 1633
+    },
+    {
+      "epoch": 0.39459067858005314,
+      "grad_norm": 0.30396586656570435,
+      "learning_rate": 6.808331362707377e-05,
+      "loss": 1.9653,
+      "step": 1634
+    },
+    {
+      "epoch": 0.39483216614344363,
+      "grad_norm": 0.29016467928886414,
+      "learning_rate": 6.804721985793792e-05,
+      "loss": 1.7476,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3950736537068341,
+      "grad_norm": 0.28695061802864075,
+      "learning_rate": 6.801111527232217e-05,
+      "loss": 1.6255,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3953151412702246,
+      "grad_norm": 0.29611697793006897,
+      "learning_rate": 6.797499989186558e-05,
+      "loss": 1.7653,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3955566288336151,
+      "grad_norm": 0.29037386178970337,
+      "learning_rate": 6.793887373821363e-05,
+      "loss": 1.7265,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3957981163970056,
+      "grad_norm": 0.30254122614860535,
+      "learning_rate": 6.790273683301831e-05,
+      "loss": 1.6677,
+      "step": 1639
+    },
+    {
+      "epoch": 0.39603960396039606,
+      "grad_norm": 0.30332258343696594,
+      "learning_rate": 6.786658919793803e-05,
+      "loss": 1.6915,
+      "step": 1640
+    },
+    {
+      "epoch": 0.39628109152378654,
+      "grad_norm": 0.30312207341194153,
+      "learning_rate": 6.783043085463762e-05,
+      "loss": 1.8231,
+      "step": 1641
+    },
+    {
+      "epoch": 0.39652257908717703,
+      "grad_norm": 0.29442471265792847,
+      "learning_rate": 6.779426182478833e-05,
+      "loss": 1.772,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3967640666505675,
+      "grad_norm": 0.2834378182888031,
+      "learning_rate": 6.775808213006787e-05,
+      "loss": 1.5351,
+      "step": 1643
+    },
+    {
+      "epoch": 0.397005554213958,
+      "grad_norm": 0.2885272204875946,
+      "learning_rate": 6.772189179216024e-05,
+      "loss": 1.7361,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3972470417773485,
+      "grad_norm": 0.29504212737083435,
+      "learning_rate": 6.768569083275594e-05,
+      "loss": 1.5897,
+      "step": 1645
+    },
+    {
+      "epoch": 0.39748852934073897,
+      "grad_norm": 0.28233128786087036,
+      "learning_rate": 6.76494792735517e-05,
+      "loss": 1.682,
+      "step": 1646
+    },
+    {
+      "epoch": 0.39773001690412946,
+      "grad_norm": 0.2948426902294159,
+      "learning_rate": 6.761325713625072e-05,
+      "loss": 1.8264,
+      "step": 1647
+    },
+    {
+      "epoch": 0.39797150446751994,
+      "grad_norm": 0.3149384558200836,
+      "learning_rate": 6.75770244425625e-05,
+      "loss": 1.8042,
+      "step": 1648
+    },
+    {
+      "epoch": 0.39821299203091043,
+      "grad_norm": 0.5162968039512634,
+      "learning_rate": 6.754078121420283e-05,
+      "loss": 1.7922,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3984544795943009,
+      "grad_norm": 0.3127591013908386,
+      "learning_rate": 6.750452747289386e-05,
+      "loss": 1.7433,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3986959671576914,
+      "grad_norm": 0.3073781430721283,
+      "learning_rate": 6.746826324036403e-05,
+      "loss": 1.9478,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3989374547210819,
+      "grad_norm": 0.29865652322769165,
+      "learning_rate": 6.743198853834806e-05,
+      "loss": 1.6781,
+      "step": 1652
+    },
+    {
+      "epoch": 0.39917894228447237,
+      "grad_norm": 0.29269179701805115,
+      "learning_rate": 6.739570338858693e-05,
+      "loss": 1.6779,
+      "step": 1653
+    },
+    {
+      "epoch": 0.39942042984786286,
+      "grad_norm": 0.29973065853118896,
+      "learning_rate": 6.735940781282796e-05,
+      "loss": 1.6762,
+      "step": 1654
+    },
+    {
+      "epoch": 0.39966191741125334,
+      "grad_norm": 0.29493921995162964,
+      "learning_rate": 6.732310183282461e-05,
+      "loss": 1.7821,
+      "step": 1655
+    },
+    {
+      "epoch": 0.39990340497464383,
+      "grad_norm": 0.2976763844490051,
+      "learning_rate": 6.728678547033662e-05,
+      "loss": 1.7797,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4001448925380343,
+      "grad_norm": 0.3011676073074341,
+      "learning_rate": 6.725045874712999e-05,
+      "loss": 1.8517,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4003863801014248,
+      "grad_norm": 0.2780913710594177,
+      "learning_rate": 6.721412168497686e-05,
+      "loss": 1.6398,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4006278676648153,
+      "grad_norm": 0.3012704849243164,
+      "learning_rate": 6.717777430565563e-05,
+      "loss": 1.7828,
+      "step": 1659
+    },
+    {
+      "epoch": 0.40086935522820577,
+      "grad_norm": 0.2912366986274719,
+      "learning_rate": 6.714141663095085e-05,
+      "loss": 1.6365,
+      "step": 1660
+    },
+    {
+      "epoch": 0.40111084279159626,
+      "grad_norm": 0.29877522587776184,
+      "learning_rate": 6.710504868265326e-05,
+      "loss": 1.7992,
+      "step": 1661
+    },
+    {
+      "epoch": 0.40135233035498674,
+      "grad_norm": 0.2767972946166992,
+      "learning_rate": 6.706867048255972e-05,
+      "loss": 1.7815,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4015938179183772,
+      "grad_norm": 0.3042682707309723,
+      "learning_rate": 6.703228205247329e-05,
+      "loss": 1.7805,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4018353054817677,
+      "grad_norm": 0.30128151178359985,
+      "learning_rate": 6.699588341420308e-05,
+      "loss": 1.5885,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4020767930451582,
+      "grad_norm": 0.3044366240501404,
+      "learning_rate": 6.695947458956443e-05,
+      "loss": 1.7927,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4023182806085487,
+      "grad_norm": 0.31940144300460815,
+      "learning_rate": 6.692305560037871e-05,
+      "loss": 1.7555,
+      "step": 1666
+    },
+    {
+      "epoch": 0.40255976817193917,
+      "grad_norm": 0.29257991909980774,
+      "learning_rate": 6.688662646847337e-05,
+      "loss": 1.7296,
+      "step": 1667
+    },
+    {
+      "epoch": 0.40280125573532966,
+      "grad_norm": 0.2922995090484619,
+      "learning_rate": 6.685018721568197e-05,
+      "loss": 1.859,
+      "step": 1668
+    },
+    {
+      "epoch": 0.40304274329872014,
+      "grad_norm": 0.29321998357772827,
+      "learning_rate": 6.681373786384416e-05,
+      "loss": 1.8717,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4032842308621106,
+      "grad_norm": 0.2926195561885834,
+      "learning_rate": 6.67772784348056e-05,
+      "loss": 1.744,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4035257184255011,
+      "grad_norm": 0.2928321659564972,
+      "learning_rate": 6.674080895041798e-05,
+      "loss": 1.6524,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4037672059888916,
+      "grad_norm": 0.27669477462768555,
+      "learning_rate": 6.670432943253906e-05,
+      "loss": 1.6994,
+      "step": 1672
+    },
+    {
+      "epoch": 0.4040086935522821,
+      "grad_norm": 0.2832506597042084,
+      "learning_rate": 6.66678399030326e-05,
+      "loss": 1.656,
+      "step": 1673
+    },
+    {
+      "epoch": 0.40425018111567257,
+      "grad_norm": 0.2894239127635956,
+      "learning_rate": 6.663134038376835e-05,
+      "loss": 1.7739,
+      "step": 1674
+    },
+    {
+      "epoch": 0.40449166867906305,
+      "grad_norm": 0.3202664852142334,
+      "learning_rate": 6.659483089662203e-05,
+      "loss": 1.8945,
+      "step": 1675
+    },
+    {
+      "epoch": 0.40473315624245354,
+      "grad_norm": 0.30392253398895264,
+      "learning_rate": 6.65583114634754e-05,
+      "loss": 1.8153,
+      "step": 1676
+    },
+    {
+      "epoch": 0.404974643805844,
+      "grad_norm": 0.2862154543399811,
+      "learning_rate": 6.652178210621609e-05,
+      "loss": 1.5981,
+      "step": 1677
+    },
+    {
+      "epoch": 0.4052161313692345,
+      "grad_norm": 0.2755371630191803,
+      "learning_rate": 6.648524284673777e-05,
+      "loss": 1.5838,
+      "step": 1678
+    },
+    {
+      "epoch": 0.405457618932625,
+      "grad_norm": 0.2834921181201935,
+      "learning_rate": 6.644869370693995e-05,
+      "loss": 1.7238,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4056991064960155,
+      "grad_norm": 0.3005118668079376,
+      "learning_rate": 6.641213470872814e-05,
+      "loss": 1.8139,
+      "step": 1680
+    },
+    {
+      "epoch": 0.40594059405940597,
+      "grad_norm": 0.2832629084587097,
+      "learning_rate": 6.637556587401374e-05,
+      "loss": 1.6167,
+      "step": 1681
+    },
+    {
+      "epoch": 0.40618208162279645,
+      "grad_norm": 0.28953537344932556,
+      "learning_rate": 6.633898722471402e-05,
+      "loss": 1.7748,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4064235691861869,
+      "grad_norm": 0.3091205358505249,
+      "learning_rate": 6.630239878275214e-05,
+      "loss": 1.8577,
+      "step": 1683
+    },
+    {
+      "epoch": 0.40666505674957737,
+      "grad_norm": 0.29937854409217834,
+      "learning_rate": 6.626580057005716e-05,
+      "loss": 1.7232,
+      "step": 1684
+    },
+    {
+      "epoch": 0.40690654431296785,
+      "grad_norm": 0.3058737516403198,
+      "learning_rate": 6.622919260856392e-05,
+      "loss": 1.7831,
+      "step": 1685
+    },
+    {
+      "epoch": 0.40714803187635834,
+      "grad_norm": 0.3024592399597168,
+      "learning_rate": 6.619257492021324e-05,
+      "loss": 1.7563,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4073895194397488,
+      "grad_norm": 0.28710269927978516,
+      "learning_rate": 6.615594752695163e-05,
+      "loss": 1.5895,
+      "step": 1687
+    },
+    {
+      "epoch": 0.4076310070031393,
+      "grad_norm": 0.2993939220905304,
+      "learning_rate": 6.611931045073147e-05,
+      "loss": 1.7654,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4078724945665298,
+      "grad_norm": 0.2898264527320862,
+      "learning_rate": 6.608266371351094e-05,
+      "loss": 1.6705,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4081139821299203,
+      "grad_norm": 0.3162345886230469,
+      "learning_rate": 6.604600733725407e-05,
+      "loss": 1.9323,
+      "step": 1690
+    },
+    {
+      "epoch": 0.40835546969331077,
+      "grad_norm": 0.29158738255500793,
+      "learning_rate": 6.600934134393055e-05,
+      "loss": 1.6548,
+      "step": 1691
+    },
+    {
+      "epoch": 0.40859695725670125,
+      "grad_norm": 0.3384915888309479,
+      "learning_rate": 6.597266575551593e-05,
+      "loss": 1.9476,
+      "step": 1692
+    },
+    {
+      "epoch": 0.40883844482009174,
+      "grad_norm": 0.3284933269023895,
+      "learning_rate": 6.593598059399148e-05,
+      "loss": 2.0154,
+      "step": 1693
+    },
+    {
+      "epoch": 0.4090799323834822,
+      "grad_norm": 0.2912117838859558,
+      "learning_rate": 6.589928588134418e-05,
+      "loss": 1.5946,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4093214199468727,
+      "grad_norm": 0.29039883613586426,
+      "learning_rate": 6.586258163956677e-05,
+      "loss": 1.6569,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4095629075102632,
+      "grad_norm": 0.29182976484298706,
+      "learning_rate": 6.582586789065768e-05,
+      "loss": 1.7651,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4098043950736537,
+      "grad_norm": 0.287898987531662,
+      "learning_rate": 6.578914465662109e-05,
+      "loss": 1.8019,
+      "step": 1697
+    },
+    {
+      "epoch": 0.41004588263704417,
+      "grad_norm": 0.28599241375923157,
+      "learning_rate": 6.575241195946679e-05,
+      "loss": 1.7098,
+      "step": 1698
+    },
+    {
+      "epoch": 0.41028737020043465,
+      "grad_norm": 0.29934966564178467,
+      "learning_rate": 6.571566982121027e-05,
+      "loss": 1.772,
+      "step": 1699
+    },
+    {
+      "epoch": 0.41052885776382514,
+      "grad_norm": 0.28729212284088135,
+      "learning_rate": 6.56789182638727e-05,
+      "loss": 1.7225,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4107703453272156,
+      "grad_norm": 0.29589328169822693,
+      "learning_rate": 6.564215730948089e-05,
+      "loss": 1.7541,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4110118328906061,
+      "grad_norm": 0.30237460136413574,
+      "learning_rate": 6.560538698006725e-05,
+      "loss": 1.7909,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4112533204539966,
+      "grad_norm": 0.2996138036251068,
+      "learning_rate": 6.55686072976698e-05,
+      "loss": 1.7098,
+      "step": 1703
+    },
+    {
+      "epoch": 0.4114948080173871,
+      "grad_norm": 0.31643104553222656,
+      "learning_rate": 6.553181828433225e-05,
+      "loss": 1.7339,
+      "step": 1704
+    },
+    {
+      "epoch": 0.41173629558077757,
+      "grad_norm": 0.26532241702079773,
+      "learning_rate": 6.549501996210385e-05,
+      "loss": 1.5175,
+      "step": 1705
+    },
+    {
+      "epoch": 0.41197778314416805,
+      "grad_norm": 0.2829447090625763,
+      "learning_rate": 6.545821235303938e-05,
+      "loss": 1.7096,
+      "step": 1706
+    },
+    {
+      "epoch": 0.41221927070755854,
+      "grad_norm": 0.2905785143375397,
+      "learning_rate": 6.542139547919927e-05,
+      "loss": 1.6579,
+      "step": 1707
+    },
+    {
+      "epoch": 0.412460758270949,
+      "grad_norm": 0.28984224796295166,
+      "learning_rate": 6.538456936264943e-05,
+      "loss": 1.6448,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4127022458343395,
+      "grad_norm": 0.27520114183425903,
+      "learning_rate": 6.534773402546138e-05,
+      "loss": 1.5512,
+      "step": 1709
+    },
+    {
+      "epoch": 0.41294373339773,
+      "grad_norm": 0.3095702528953552,
+      "learning_rate": 6.531088948971212e-05,
+      "loss": 1.7525,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4131852209611205,
+      "grad_norm": 0.2894667387008667,
+      "learning_rate": 6.527403577748419e-05,
+      "loss": 1.63,
+      "step": 1711
+    },
+    {
+      "epoch": 0.41342670852451097,
+      "grad_norm": 0.28379741311073303,
+      "learning_rate": 6.523717291086558e-05,
+      "loss": 1.6679,
+      "step": 1712
+    },
+    {
+      "epoch": 0.41366819608790145,
+      "grad_norm": 0.2986467480659485,
+      "learning_rate": 6.520030091194984e-05,
+      "loss": 1.7661,
+      "step": 1713
+    },
+    {
+      "epoch": 0.41390968365129194,
+      "grad_norm": 0.3017704486846924,
+      "learning_rate": 6.516341980283593e-05,
+      "loss": 1.7938,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4141511712146824,
+      "grad_norm": 0.30169060826301575,
+      "learning_rate": 6.512652960562829e-05,
+      "loss": 1.7313,
+      "step": 1715
+    },
+    {
+      "epoch": 0.4143926587780729,
+      "grad_norm": 0.30517810583114624,
+      "learning_rate": 6.508963034243683e-05,
+      "loss": 1.8192,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 0.2784494161605835,
+      "learning_rate": 6.505272203537687e-05,
+      "loss": 1.6988,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4148756339048539,
+      "grad_norm": 0.6281806826591492,
+      "learning_rate": 6.501580470656913e-05,
+      "loss": 1.6437,
+      "step": 1718
+    },
+    {
+      "epoch": 0.41511712146824437,
+      "grad_norm": 0.28869011998176575,
+      "learning_rate": 6.497887837813981e-05,
+      "loss": 1.7,
+      "step": 1719
+    },
+    {
+      "epoch": 0.41535860903163485,
+      "grad_norm": 0.30811411142349243,
+      "learning_rate": 6.494194307222043e-05,
+      "loss": 1.8591,
+      "step": 1720
+    },
+    {
+      "epoch": 0.41560009659502534,
+      "grad_norm": 0.2911246716976166,
+      "learning_rate": 6.49049988109479e-05,
+      "loss": 1.7103,
+      "step": 1721
+    },
+    {
+      "epoch": 0.4158415841584158,
+      "grad_norm": 0.27908068895339966,
+      "learning_rate": 6.486804561646456e-05,
+      "loss": 1.6012,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4160830717218063,
+      "grad_norm": 0.2703852951526642,
+      "learning_rate": 6.483108351091802e-05,
+      "loss": 1.3896,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4163245592851968,
+      "grad_norm": 0.2969891130924225,
+      "learning_rate": 6.479411251646128e-05,
+      "loss": 1.8627,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4165660468485873,
+      "grad_norm": 0.31090447306632996,
+      "learning_rate": 6.475713265525267e-05,
+      "loss": 1.8897,
+      "step": 1725
+    },
+    {
+      "epoch": 0.41680753441197776,
+      "grad_norm": 0.293481707572937,
+      "learning_rate": 6.472014394945582e-05,
+      "loss": 1.7154,
+      "step": 1726
+    },
+    {
+      "epoch": 0.41704902197536825,
+      "grad_norm": 0.28805553913116455,
+      "learning_rate": 6.468314642123963e-05,
+      "loss": 1.8052,
+      "step": 1727
+    },
+    {
+      "epoch": 0.41729050953875874,
+      "grad_norm": 0.3230077624320984,
+      "learning_rate": 6.464614009277837e-05,
+      "loss": 1.8545,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4175319971021492,
+      "grad_norm": 0.29030686616897583,
+      "learning_rate": 6.460912498625147e-05,
+      "loss": 1.7401,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4177734846655397,
+      "grad_norm": 0.30003416538238525,
+      "learning_rate": 6.457210112384376e-05,
+      "loss": 1.7905,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4180149722289302,
+      "grad_norm": 0.2969563901424408,
+      "learning_rate": 6.453506852774518e-05,
+      "loss": 1.6655,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4182564597923207,
+      "grad_norm": 0.29960548877716064,
+      "learning_rate": 6.4498027220151e-05,
+      "loss": 1.7215,
+      "step": 1732
+    },
+    {
+      "epoch": 0.41849794735571116,
+      "grad_norm": 0.2943073511123657,
+      "learning_rate": 6.446097722326166e-05,
+      "loss": 1.8138,
+      "step": 1733
+    },
+    {
+      "epoch": 0.41873943491910165,
+      "grad_norm": 0.30021798610687256,
+      "learning_rate": 6.442391855928285e-05,
+      "loss": 1.6481,
+      "step": 1734
+    },
+    {
+      "epoch": 0.41898092248249214,
+      "grad_norm": 0.30978235602378845,
+      "learning_rate": 6.438685125042543e-05,
+      "loss": 1.7445,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4192224100458826,
+      "grad_norm": 0.2761182487010956,
+      "learning_rate": 6.434977531890542e-05,
+      "loss": 1.6566,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4194638976092731,
+      "grad_norm": 0.3045799434185028,
+      "learning_rate": 6.431269078694405e-05,
+      "loss": 1.7833,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4197053851726636,
+      "grad_norm": 0.32096797227859497,
+      "learning_rate": 6.427559767676767e-05,
+      "loss": 1.6964,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4199468727360541,
+      "grad_norm": 0.31089526414871216,
+      "learning_rate": 6.423849601060778e-05,
+      "loss": 1.7477,
+      "step": 1739
+    },
+    {
+      "epoch": 0.42018836029944456,
+      "grad_norm": 0.2792001962661743,
+      "learning_rate": 6.420138581070106e-05,
+      "loss": 1.5561,
+      "step": 1740
+    },
+    {
+      "epoch": 0.42042984786283505,
+      "grad_norm": 0.3000398874282837,
+      "learning_rate": 6.41642670992892e-05,
+      "loss": 1.8557,
+      "step": 1741
+    },
+    {
+      "epoch": 0.42067133542622553,
+      "grad_norm": 0.2896421253681183,
+      "learning_rate": 6.412713989861907e-05,
+      "loss": 1.6449,
+      "step": 1742
+    },
+    {
+      "epoch": 0.420912822989616,
+      "grad_norm": 0.29571399092674255,
+      "learning_rate": 6.409000423094264e-05,
+      "loss": 1.9293,
+      "step": 1743
+    },
+    {
+      "epoch": 0.4211543105530065,
+      "grad_norm": 0.3130645155906677,
+      "learning_rate": 6.405286011851689e-05,
+      "loss": 1.926,
+      "step": 1744
+    },
+    {
+      "epoch": 0.421395798116397,
+      "grad_norm": 0.31119149923324585,
+      "learning_rate": 6.40157075836039e-05,
+      "loss": 1.7332,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4216372856797875,
+      "grad_norm": 0.2839961051940918,
+      "learning_rate": 6.397854664847078e-05,
+      "loss": 1.7123,
+      "step": 1746
+    },
+    {
+      "epoch": 0.42187877324317796,
+      "grad_norm": 0.29415860772132874,
+      "learning_rate": 6.394137733538972e-05,
+      "loss": 1.7624,
+      "step": 1747
+    },
+    {
+      "epoch": 0.42212026080656845,
+      "grad_norm": 0.2900754511356354,
+      "learning_rate": 6.390419966663785e-05,
+      "loss": 1.7625,
+      "step": 1748
+    },
+    {
+      "epoch": 0.42236174836995893,
+      "grad_norm": 0.2831285297870636,
+      "learning_rate": 6.386701366449741e-05,
+      "loss": 1.8242,
+      "step": 1749
+    },
+    {
+      "epoch": 0.4226032359333494,
+      "grad_norm": 0.29872801899909973,
+      "learning_rate": 6.382981935125555e-05,
+      "loss": 1.8123,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4228447234967399,
+      "grad_norm": 0.2907567620277405,
+      "learning_rate": 6.379261674920443e-05,
+      "loss": 1.8015,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4230862110601304,
+      "grad_norm": 0.28774112462997437,
+      "learning_rate": 6.375540588064118e-05,
+      "loss": 1.6881,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4233276986235209,
+      "grad_norm": 0.2798502445220947,
+      "learning_rate": 6.37181867678679e-05,
+      "loss": 1.6737,
+      "step": 1753
+    },
+    {
+      "epoch": 0.42356918618691136,
+      "grad_norm": 0.3050326704978943,
+      "learning_rate": 6.368095943319157e-05,
+      "loss": 1.9083,
+      "step": 1754
+    },
+    {
+      "epoch": 0.42381067375030185,
+      "grad_norm": 0.27888479828834534,
+      "learning_rate": 6.364372389892423e-05,
+      "loss": 1.6069,
+      "step": 1755
+    },
+    {
+      "epoch": 0.42405216131369233,
+      "grad_norm": 0.2976183295249939,
+      "learning_rate": 6.360648018738267e-05,
+      "loss": 1.6508,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4242936488770828,
+      "grad_norm": 0.2891102433204651,
+      "learning_rate": 6.356922832088867e-05,
+      "loss": 1.6793,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4245351364404733,
+      "grad_norm": 0.29795247316360474,
+      "learning_rate": 6.35319683217689e-05,
+      "loss": 1.846,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4247766240038638,
+      "grad_norm": 0.3060801029205322,
+      "learning_rate": 6.349470021235491e-05,
+      "loss": 1.7774,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4250181115672543,
+      "grad_norm": 0.2822737395763397,
+      "learning_rate": 6.345742401498305e-05,
+      "loss": 1.6143,
+      "step": 1760
+    },
+    {
+      "epoch": 0.42525959913064476,
+      "grad_norm": 0.30945831537246704,
+      "learning_rate": 6.342013975199459e-05,
+      "loss": 1.8053,
+      "step": 1761
+    },
+    {
+      "epoch": 0.42550108669403525,
+      "grad_norm": 0.32073312997817993,
+      "learning_rate": 6.338284744573557e-05,
+      "loss": 1.8592,
+      "step": 1762
+    },
+    {
+      "epoch": 0.42574257425742573,
+      "grad_norm": 0.28697794675827026,
+      "learning_rate": 6.334554711855692e-05,
+      "loss": 1.7174,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4259840618208162,
+      "grad_norm": 0.27200600504875183,
+      "learning_rate": 6.330823879281431e-05,
+      "loss": 1.5327,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4262255493842067,
+      "grad_norm": 0.31434571743011475,
+      "learning_rate": 6.327092249086825e-05,
+      "loss": 2.0499,
+      "step": 1765
+    },
+    {
+      "epoch": 0.4264670369475972,
+      "grad_norm": 0.3031150996685028,
+      "learning_rate": 6.323359823508401e-05,
+      "loss": 1.8122,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4267085245109877,
+      "grad_norm": 0.31416165828704834,
+      "learning_rate": 6.319626604783166e-05,
+      "loss": 1.6704,
+      "step": 1767
+    },
+    {
+      "epoch": 0.42695001207437816,
+      "grad_norm": 0.2872917354106903,
+      "learning_rate": 6.315892595148593e-05,
+      "loss": 1.6125,
+      "step": 1768
+    },
+    {
+      "epoch": 0.42719149963776865,
+      "grad_norm": 0.27335771918296814,
+      "learning_rate": 6.312157796842641e-05,
+      "loss": 1.7027,
+      "step": 1769
+    },
+    {
+      "epoch": 0.42743298720115913,
+      "grad_norm": 0.29706230759620667,
+      "learning_rate": 6.308422212103736e-05,
+      "loss": 1.9199,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4276744747645496,
+      "grad_norm": 0.29793912172317505,
+      "learning_rate": 6.304685843170772e-05,
+      "loss": 1.7714,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4279159623279401,
+      "grad_norm": 0.2788400650024414,
+      "learning_rate": 6.300948692283122e-05,
+      "loss": 1.6532,
+      "step": 1772
+    },
+    {
+      "epoch": 0.4281574498913306,
+      "grad_norm": 0.2901703715324402,
+      "learning_rate": 6.297210761680618e-05,
+      "loss": 1.6379,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4283989374547211,
+      "grad_norm": 0.3157692849636078,
+      "learning_rate": 6.293472053603563e-05,
+      "loss": 1.9983,
+      "step": 1774
+    },
+    {
+      "epoch": 0.42864042501811156,
+      "grad_norm": 0.2885459363460541,
+      "learning_rate": 6.289732570292728e-05,
+      "loss": 1.7734,
+      "step": 1775
+    },
+    {
+      "epoch": 0.42888191258150204,
+      "grad_norm": 0.2981279492378235,
+      "learning_rate": 6.28599231398935e-05,
+      "loss": 1.75,
+      "step": 1776
+    },
+    {
+      "epoch": 0.42912340014489253,
+      "grad_norm": 0.2864707410335541,
+      "learning_rate": 6.282251286935121e-05,
+      "loss": 1.6159,
+      "step": 1777
+    },
+    {
+      "epoch": 0.429364887708283,
+      "grad_norm": 0.28859418630599976,
+      "learning_rate": 6.278509491372207e-05,
+      "loss": 1.6682,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4296063752716735,
+      "grad_norm": 0.3026987612247467,
+      "learning_rate": 6.274766929543221e-05,
+      "loss": 1.7819,
+      "step": 1779
+    },
+    {
+      "epoch": 0.429847862835064,
+      "grad_norm": 0.29130101203918457,
+      "learning_rate": 6.271023603691248e-05,
+      "loss": 1.7521,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4300893503984545,
+      "grad_norm": 0.28929030895233154,
+      "learning_rate": 6.267279516059823e-05,
+      "loss": 1.8133,
+      "step": 1781
+    },
+    {
+      "epoch": 0.43033083796184496,
+      "grad_norm": 0.2978893220424652,
+      "learning_rate": 6.263534668892938e-05,
+      "loss": 1.7392,
+      "step": 1782
+    },
+    {
+      "epoch": 0.43057232552523544,
+      "grad_norm": 0.2781008183956146,
+      "learning_rate": 6.259789064435044e-05,
+      "loss": 1.7491,
+      "step": 1783
+    },
+    {
+      "epoch": 0.43081381308862593,
+      "grad_norm": 0.2996642291545868,
+      "learning_rate": 6.256042704931045e-05,
+      "loss": 1.7415,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4310553006520164,
+      "grad_norm": 0.2978329360485077,
+      "learning_rate": 6.252295592626293e-05,
+      "loss": 1.783,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4312967882154069,
+      "grad_norm": 0.29910266399383545,
+      "learning_rate": 6.248547729766596e-05,
+      "loss": 1.8312,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4315382757787974,
+      "grad_norm": 0.2915440499782562,
+      "learning_rate": 6.244799118598211e-05,
+      "loss": 1.8281,
+      "step": 1787
+    },
+    {
+      "epoch": 0.43177976334218787,
+      "grad_norm": 0.2931154668331146,
+      "learning_rate": 6.241049761367845e-05,
+      "loss": 1.8002,
+      "step": 1788
+    },
+    {
+      "epoch": 0.43202125090557836,
+      "grad_norm": 0.2912518382072449,
+      "learning_rate": 6.237299660322645e-05,
+      "loss": 1.8242,
+      "step": 1789
+    },
+    {
+      "epoch": 0.43226273846896884,
+      "grad_norm": 0.2783883810043335,
+      "learning_rate": 6.233548817710213e-05,
+      "loss": 1.6564,
+      "step": 1790
+    },
+    {
+      "epoch": 0.43250422603235933,
+      "grad_norm": 0.29445287585258484,
+      "learning_rate": 6.22979723577859e-05,
+      "loss": 1.7016,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4327457135957498,
+      "grad_norm": 0.27962037920951843,
+      "learning_rate": 6.22604491677626e-05,
+      "loss": 1.5158,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4329872011591403,
+      "grad_norm": 0.28880858421325684,
+      "learning_rate": 6.222291862952153e-05,
+      "loss": 1.6425,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4332286887225308,
+      "grad_norm": 0.2967377007007599,
+      "learning_rate": 6.218538076555635e-05,
+      "loss": 1.8272,
+      "step": 1794
+    },
+    {
+      "epoch": 0.43347017628592127,
+      "grad_norm": 0.29760682582855225,
+      "learning_rate": 6.214783559836512e-05,
+      "loss": 1.7123,
+      "step": 1795
+    },
+    {
+      "epoch": 0.43371166384931176,
+      "grad_norm": 0.26291459798812866,
+      "learning_rate": 6.211028315045027e-05,
+      "loss": 1.4946,
+      "step": 1796
+    },
+    {
+      "epoch": 0.43395315141270224,
+      "grad_norm": 0.3057827651500702,
+      "learning_rate": 6.207272344431866e-05,
+      "loss": 1.7002,
+      "step": 1797
+    },
+    {
+      "epoch": 0.43419463897609273,
+      "grad_norm": 0.3144783079624176,
+      "learning_rate": 6.20351565024814e-05,
+      "loss": 1.8826,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4344361265394832,
+      "grad_norm": 0.29561561346054077,
+      "learning_rate": 6.199758234745401e-05,
+      "loss": 1.7635,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4346776141028737,
+      "grad_norm": 0.2834794521331787,
+      "learning_rate": 6.196000100175629e-05,
+      "loss": 1.7412,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4349191016662642,
+      "grad_norm": 0.29855379462242126,
+      "learning_rate": 6.192241248791236e-05,
+      "loss": 1.8944,
+      "step": 1801
+    },
+    {
+      "epoch": 0.43516058922965467,
+      "grad_norm": 0.30267342925071716,
+      "learning_rate": 6.188481682845068e-05,
+      "loss": 1.7823,
+      "step": 1802
+    },
+    {
+      "epoch": 0.43540207679304516,
+      "grad_norm": 0.30372753739356995,
+      "learning_rate": 6.184721404590392e-05,
+      "loss": 1.871,
+      "step": 1803
+    },
+    {
+      "epoch": 0.43564356435643564,
+      "grad_norm": 0.2923888862133026,
+      "learning_rate": 6.180960416280906e-05,
+      "loss": 1.7681,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4358850519198261,
+      "grad_norm": 0.3047378361225128,
+      "learning_rate": 6.177198720170735e-05,
+      "loss": 1.8215,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4361265394832166,
+      "grad_norm": 0.30522942543029785,
+      "learning_rate": 6.173436318514425e-05,
+      "loss": 1.8184,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4363680270466071,
+      "grad_norm": 0.29536402225494385,
+      "learning_rate": 6.169673213566946e-05,
+      "loss": 1.7936,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4366095146099976,
+      "grad_norm": 0.29366007447242737,
+      "learning_rate": 6.165909407583687e-05,
+      "loss": 1.8101,
+      "step": 1808
+    },
+    {
+      "epoch": 0.43685100217338807,
+      "grad_norm": 0.3027951419353485,
+      "learning_rate": 6.162144902820467e-05,
+      "loss": 1.8094,
+      "step": 1809
+    },
+    {
+      "epoch": 0.43709248973677856,
+      "grad_norm": 0.30503374338150024,
+      "learning_rate": 6.15837970153351e-05,
+      "loss": 1.5617,
+      "step": 1810
+    },
+    {
+      "epoch": 0.43733397730016904,
+      "grad_norm": 0.279544472694397,
+      "learning_rate": 6.154613805979467e-05,
+      "loss": 1.6252,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4375754648635595,
+      "grad_norm": 0.300307959318161,
+      "learning_rate": 6.150847218415401e-05,
+      "loss": 1.8711,
+      "step": 1812
+    },
+    {
+      "epoch": 0.43781695242695,
+      "grad_norm": 0.2886538505554199,
+      "learning_rate": 6.14707994109879e-05,
+      "loss": 1.5643,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4380584399903405,
+      "grad_norm": 0.2958506643772125,
+      "learning_rate": 6.143311976287531e-05,
+      "loss": 1.7646,
+      "step": 1814
+    },
+    {
+      "epoch": 0.438299927553731,
+      "grad_norm": 0.29687565565109253,
+      "learning_rate": 6.139543326239926e-05,
+      "loss": 1.642,
+      "step": 1815
+    },
+    {
+      "epoch": 0.43854141511712147,
+      "grad_norm": 0.2901788353919983,
+      "learning_rate": 6.135773993214685e-05,
+      "loss": 1.5288,
+      "step": 1816
+    },
+    {
+      "epoch": 0.43878290268051195,
+      "grad_norm": 0.29023823142051697,
+      "learning_rate": 6.132003979470939e-05,
+      "loss": 1.6135,
+      "step": 1817
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 0.2979198098182678,
+      "learning_rate": 6.128233287268216e-05,
+      "loss": 1.7553,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4392658778072929,
+      "grad_norm": 0.28564679622650146,
+      "learning_rate": 6.124461918866459e-05,
+      "loss": 1.6581,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4395073653706834,
+      "grad_norm": 0.2789301872253418,
+      "learning_rate": 6.12068987652601e-05,
+      "loss": 1.6782,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4397488529340739,
+      "grad_norm": 0.2890610992908478,
+      "learning_rate": 6.116917162507618e-05,
+      "loss": 1.6198,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4399903404974644,
+      "grad_norm": 0.29789113998413086,
+      "learning_rate": 6.113143779072432e-05,
+      "loss": 1.8246,
+      "step": 1822
+    },
+    {
+      "epoch": 0.44023182806085487,
+      "grad_norm": 0.2954878807067871,
+      "learning_rate": 6.109369728482006e-05,
+      "loss": 1.8309,
+      "step": 1823
+    },
+    {
+      "epoch": 0.44047331562424535,
+      "grad_norm": 0.29515141248703003,
+      "learning_rate": 6.105595012998293e-05,
+      "loss": 1.8473,
+      "step": 1824
+    },
+    {
+      "epoch": 0.44071480318763584,
+      "grad_norm": 0.28120049834251404,
+      "learning_rate": 6.1018196348836384e-05,
+      "loss": 1.6964,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4409562907510263,
+      "grad_norm": 0.2976786196231842,
+      "learning_rate": 6.098043596400795e-05,
+      "loss": 1.7344,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4411977783144168,
+      "grad_norm": 0.27497339248657227,
+      "learning_rate": 6.094266899812904e-05,
+      "loss": 1.5725,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4414392658778073,
+      "grad_norm": 0.29393255710601807,
+      "learning_rate": 6.090489547383502e-05,
+      "loss": 1.7235,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4416807534411978,
+      "grad_norm": 0.28918004035949707,
+      "learning_rate": 6.0867115413765206e-05,
+      "loss": 1.6561,
+      "step": 1829
+    },
+    {
+      "epoch": 0.44192224100458827,
+      "grad_norm": 0.2973797917366028,
+      "learning_rate": 6.0829328840562836e-05,
+      "loss": 1.8733,
+      "step": 1830
+    },
+    {
+      "epoch": 0.44216372856797875,
+      "grad_norm": 0.2833927571773529,
+      "learning_rate": 6.079153577687502e-05,
+      "loss": 1.6147,
+      "step": 1831
+    },
+    {
+      "epoch": 0.44240521613136924,
+      "grad_norm": 0.2931411862373352,
+      "learning_rate": 6.075373624535279e-05,
+      "loss": 1.7791,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4426467036947597,
+      "grad_norm": 0.30425596237182617,
+      "learning_rate": 6.071593026865102e-05,
+      "loss": 1.8605,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4428881912581502,
+      "grad_norm": 0.2918083369731903,
+      "learning_rate": 6.067811786942851e-05,
+      "loss": 1.814,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4431296788215407,
+      "grad_norm": 0.2972649931907654,
+      "learning_rate": 6.064029907034783e-05,
+      "loss": 1.6886,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4433711663849312,
+      "grad_norm": 0.2869977653026581,
+      "learning_rate": 6.0602473894075416e-05,
+      "loss": 1.6742,
+      "step": 1836
+    },
+    {
+      "epoch": 0.44361265394832167,
+      "grad_norm": 0.2964032292366028,
+      "learning_rate": 6.0564642363281577e-05,
+      "loss": 1.6807,
+      "step": 1837
+    },
+    {
+      "epoch": 0.44385414151171215,
+      "grad_norm": 0.4216827154159546,
+      "learning_rate": 6.052680450064036e-05,
+      "loss": 1.8796,
+      "step": 1838
+    },
+    {
+      "epoch": 0.44409562907510264,
+      "grad_norm": 0.2836448848247528,
+      "learning_rate": 6.0488960328829615e-05,
+      "loss": 1.8391,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4443371166384931,
+      "grad_norm": 0.2807202637195587,
+      "learning_rate": 6.045110987053102e-05,
+      "loss": 1.6688,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4445786042018836,
+      "grad_norm": 0.2864443361759186,
+      "learning_rate": 6.0413253148429986e-05,
+      "loss": 1.7344,
+      "step": 1841
+    },
+    {
+      "epoch": 0.4448200917652741,
+      "grad_norm": 0.31129157543182373,
+      "learning_rate": 6.0375390185215664e-05,
+      "loss": 1.8526,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4450615793286646,
+      "grad_norm": 0.3081095218658447,
+      "learning_rate": 6.033752100358101e-05,
+      "loss": 1.8352,
+      "step": 1843
+    },
+    {
+      "epoch": 0.44530306689205507,
+      "grad_norm": 0.28839313983917236,
+      "learning_rate": 6.0299645626222636e-05,
+      "loss": 1.7259,
+      "step": 1844
+    },
+    {
+      "epoch": 0.44554455445544555,
+      "grad_norm": 0.28718453645706177,
+      "learning_rate": 6.026176407584088e-05,
+      "loss": 1.8188,
+      "step": 1845
+    },
+    {
+      "epoch": 0.44578604201883604,
+      "grad_norm": 0.29002997279167175,
+      "learning_rate": 6.022387637513983e-05,
+      "loss": 1.6686,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4460275295822265,
+      "grad_norm": 0.29619693756103516,
+      "learning_rate": 6.0185982546827205e-05,
+      "loss": 1.8598,
+      "step": 1847
+    },
+    {
+      "epoch": 0.446269017145617,
+      "grad_norm": 0.29213395714759827,
+      "learning_rate": 6.014808261361441e-05,
+      "loss": 1.6915,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4465105047090075,
+      "grad_norm": 0.2871073782444,
+      "learning_rate": 6.0110176598216525e-05,
+      "loss": 1.6989,
+      "step": 1849
+    },
+    {
+      "epoch": 0.446751992272398,
+      "grad_norm": 0.2969827353954315,
+      "learning_rate": 6.0072264523352275e-05,
+      "loss": 1.6611,
+      "step": 1850
+    },
+    {
+      "epoch": 0.44699347983578847,
+      "grad_norm": 0.28079450130462646,
+      "learning_rate": 6.003434641174398e-05,
+      "loss": 1.6835,
+      "step": 1851
+    },
+    {
+      "epoch": 0.44723496739917895,
+      "grad_norm": 0.3196025490760803,
+      "learning_rate": 5.999642228611761e-05,
+      "loss": 1.9651,
+      "step": 1852
+    },
+    {
+      "epoch": 0.44747645496256944,
+      "grad_norm": 0.3039729595184326,
+      "learning_rate": 5.9958492169202776e-05,
+      "loss": 1.6505,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4477179425259599,
+      "grad_norm": 0.2899879217147827,
+      "learning_rate": 5.9920556083732585e-05,
+      "loss": 1.6848,
+      "step": 1854
+    },
+    {
+      "epoch": 0.4479594300893504,
+      "grad_norm": 0.2984606921672821,
+      "learning_rate": 5.988261405244382e-05,
+      "loss": 1.833,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4482009176527409,
+      "grad_norm": 0.3010087311267853,
+      "learning_rate": 5.984466609807675e-05,
+      "loss": 1.7083,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4484424052161314,
+      "grad_norm": 0.2915293872356415,
+      "learning_rate": 5.980671224337523e-05,
+      "loss": 1.75,
+      "step": 1857
+    },
+    {
+      "epoch": 0.44868389277952186,
+      "grad_norm": 0.2936859428882599,
+      "learning_rate": 5.976875251108667e-05,
+      "loss": 1.7717,
+      "step": 1858
+    },
+    {
+      "epoch": 0.44892538034291235,
+      "grad_norm": 0.2996618151664734,
+      "learning_rate": 5.9730786923961946e-05,
+      "loss": 1.7103,
+      "step": 1859
+    },
+    {
+      "epoch": 0.44916686790630284,
+      "grad_norm": 0.2864699363708496,
+      "learning_rate": 5.9692815504755505e-05,
+      "loss": 1.686,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4494083554696933,
+      "grad_norm": 0.2949967086315155,
+      "learning_rate": 5.965483827622526e-05,
+      "loss": 1.6495,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4496498430330838,
+      "grad_norm": 0.29744744300842285,
+      "learning_rate": 5.9616855261132586e-05,
+      "loss": 1.7475,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4498913305964743,
+      "grad_norm": 0.2873002290725708,
+      "learning_rate": 5.957886648224238e-05,
+      "loss": 1.6152,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4501328181598648,
+      "grad_norm": 0.3030773401260376,
+      "learning_rate": 5.954087196232292e-05,
+      "loss": 1.9727,
+      "step": 1864
+    },
+    {
+      "epoch": 0.45037430572325526,
+      "grad_norm": 0.31282663345336914,
+      "learning_rate": 5.9502871724146014e-05,
+      "loss": 1.7839,
+      "step": 1865
+    },
+    {
+      "epoch": 0.45061579328664575,
+      "grad_norm": 0.2903457283973694,
+      "learning_rate": 5.9464865790486815e-05,
+      "loss": 1.7714,
+      "step": 1866
+    },
+    {
+      "epoch": 0.45085728085003623,
+      "grad_norm": 0.2956843376159668,
+      "learning_rate": 5.9426854184123935e-05,
+      "loss": 1.7269,
+      "step": 1867
+    },
+    {
+      "epoch": 0.4510987684134267,
+      "grad_norm": 0.30317896604537964,
+      "learning_rate": 5.938883692783938e-05,
+      "loss": 1.8399,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4513402559768172,
+      "grad_norm": 0.3119915723800659,
+      "learning_rate": 5.9350814044418515e-05,
+      "loss": 1.7026,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4515817435402077,
+      "grad_norm": 0.29706525802612305,
+      "learning_rate": 5.931278555665013e-05,
+      "loss": 1.7608,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4518232311035982,
+      "grad_norm": 0.30105337500572205,
+      "learning_rate": 5.927475148732633e-05,
+      "loss": 1.9063,
+      "step": 1871
+    },
+    {
+      "epoch": 0.45206471866698866,
+      "grad_norm": 0.28921782970428467,
+      "learning_rate": 5.923671185924255e-05,
+      "loss": 1.5752,
+      "step": 1872
+    },
+    {
+      "epoch": 0.45230620623037915,
+      "grad_norm": 0.29720595479011536,
+      "learning_rate": 5.919866669519765e-05,
+      "loss": 1.7688,
+      "step": 1873
+    },
+    {
+      "epoch": 0.45254769379376963,
+      "grad_norm": 0.30140766501426697,
+      "learning_rate": 5.9160616017993685e-05,
+      "loss": 1.7024,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4527891813571601,
+      "grad_norm": 0.29524683952331543,
+      "learning_rate": 5.9122559850436085e-05,
+      "loss": 1.9465,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4530306689205506,
+      "grad_norm": 0.30466195940971375,
+      "learning_rate": 5.90844982153336e-05,
+      "loss": 1.905,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4532721564839411,
+      "grad_norm": 0.3002755641937256,
+      "learning_rate": 5.904643113549816e-05,
+      "loss": 1.7018,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4535136440473316,
+      "grad_norm": 0.2979101836681366,
+      "learning_rate": 5.900835863374503e-05,
+      "loss": 1.709,
+      "step": 1878
+    },
+    {
+      "epoch": 0.45375513161072206,
+      "grad_norm": 0.2791266441345215,
+      "learning_rate": 5.897028073289272e-05,
+      "loss": 1.6399,
+      "step": 1879
+    },
+    {
+      "epoch": 0.45399661917411255,
+      "grad_norm": 0.3100603222846985,
+      "learning_rate": 5.8932197455762974e-05,
+      "loss": 1.7315,
+      "step": 1880
+    },
+    {
+      "epoch": 0.45423810673750303,
+      "grad_norm": 0.29227718710899353,
+      "learning_rate": 5.889410882518071e-05,
+      "loss": 1.8037,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4544795943008935,
+      "grad_norm": 0.2878359258174896,
+      "learning_rate": 5.885601486397412e-05,
+      "loss": 1.6143,
+      "step": 1882
+    },
+    {
+      "epoch": 0.454721081864284,
+      "grad_norm": 0.30222880840301514,
+      "learning_rate": 5.881791559497456e-05,
+      "loss": 1.6901,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4549625694276745,
+      "grad_norm": 0.32026219367980957,
+      "learning_rate": 5.877981104101657e-05,
+      "loss": 1.8235,
+      "step": 1884
+    },
+    {
+      "epoch": 0.455204056991065,
+      "grad_norm": 0.3001136779785156,
+      "learning_rate": 5.874170122493784e-05,
+      "loss": 1.9317,
+      "step": 1885
+    },
+    {
+      "epoch": 0.45544554455445546,
+      "grad_norm": 0.27894434332847595,
+      "learning_rate": 5.870358616957926e-05,
+      "loss": 1.5573,
+      "step": 1886
+    },
+    {
+      "epoch": 0.45568703211784595,
+      "grad_norm": 0.2884087860584259,
+      "learning_rate": 5.8665465897784824e-05,
+      "loss": 1.6473,
+      "step": 1887
+    },
+    {
+      "epoch": 0.45592851968123643,
+      "grad_norm": 0.2925865054130554,
+      "learning_rate": 5.8627340432401656e-05,
+      "loss": 1.7021,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4561700072446269,
+      "grad_norm": 0.30170294642448425,
+      "learning_rate": 5.8589209796279996e-05,
+      "loss": 1.8159,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4564114948080174,
+      "grad_norm": 0.3007247745990753,
+      "learning_rate": 5.855107401227319e-05,
+      "loss": 1.8116,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4566529823714079,
+      "grad_norm": 0.28724128007888794,
+      "learning_rate": 5.851293310323767e-05,
+      "loss": 1.7403,
+      "step": 1891
+    },
+    {
+      "epoch": 0.4568944699347984,
+      "grad_norm": 0.28229671716690063,
+      "learning_rate": 5.847478709203291e-05,
+      "loss": 1.7085,
+      "step": 1892
+    },
+    {
+      "epoch": 0.45713595749818886,
+      "grad_norm": 0.3043849766254425,
+      "learning_rate": 5.843663600152148e-05,
+      "loss": 1.7451,
+      "step": 1893
+    },
+    {
+      "epoch": 0.45737744506157935,
+      "grad_norm": 0.2898251712322235,
+      "learning_rate": 5.8398479854569e-05,
+      "loss": 1.82,
+      "step": 1894
+    },
+    {
+      "epoch": 0.45761893262496983,
+      "grad_norm": 0.3063147962093353,
+      "learning_rate": 5.836031867404406e-05,
+      "loss": 1.9651,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4578604201883603,
+      "grad_norm": 0.2852920889854431,
+      "learning_rate": 5.832215248281834e-05,
+      "loss": 1.6368,
+      "step": 1896
+    },
+    {
+      "epoch": 0.4581019077517508,
+      "grad_norm": 0.30048584938049316,
+      "learning_rate": 5.828398130376645e-05,
+      "loss": 1.738,
+      "step": 1897
+    },
+    {
+      "epoch": 0.4583433953151413,
+      "grad_norm": 0.2849442660808563,
+      "learning_rate": 5.824580515976608e-05,
+      "loss": 1.6016,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4585848828785318,
+      "grad_norm": 0.28128960728645325,
+      "learning_rate": 5.8207624073697816e-05,
+      "loss": 1.6994,
+      "step": 1899
+    },
+    {
+      "epoch": 0.45882637044192226,
+      "grad_norm": 0.29790449142456055,
+      "learning_rate": 5.8169438068445246e-05,
+      "loss": 1.7874,
+      "step": 1900
+    },
+    {
+      "epoch": 0.45906785800531275,
+      "grad_norm": 0.28888222575187683,
+      "learning_rate": 5.813124716689488e-05,
+      "loss": 1.7332,
+      "step": 1901
+    },
+    {
+      "epoch": 0.45930934556870323,
+      "grad_norm": 0.2996605932712555,
+      "learning_rate": 5.809305139193619e-05,
+      "loss": 1.8353,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4595508331320937,
+      "grad_norm": 0.29596322774887085,
+      "learning_rate": 5.805485076646156e-05,
+      "loss": 1.682,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4597923206954842,
+      "grad_norm": 0.3038746416568756,
+      "learning_rate": 5.801664531336628e-05,
+      "loss": 1.6489,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4600338082588747,
+      "grad_norm": 0.29957544803619385,
+      "learning_rate": 5.797843505554852e-05,
+      "loss": 1.7067,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4602752958222652,
+      "grad_norm": 0.29159390926361084,
+      "learning_rate": 5.794022001590935e-05,
+      "loss": 1.7833,
+      "step": 1906
+    },
+    {
+      "epoch": 0.46051678338565566,
+      "grad_norm": 0.2858039140701294,
+      "learning_rate": 5.790200021735268e-05,
+      "loss": 1.6577,
+      "step": 1907
+    },
+    {
+      "epoch": 0.46075827094904614,
+      "grad_norm": 0.3013875484466553,
+      "learning_rate": 5.78637756827853e-05,
+      "loss": 1.771,
+      "step": 1908
+    },
+    {
+      "epoch": 0.46099975851243663,
+      "grad_norm": 0.30382275581359863,
+      "learning_rate": 5.782554643511685e-05,
+      "loss": 1.7432,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4612412460758271,
+      "grad_norm": 0.28172066807746887,
+      "learning_rate": 5.778731249725974e-05,
+      "loss": 1.6688,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4614827336392176,
+      "grad_norm": 0.28853094577789307,
+      "learning_rate": 5.7749073892129244e-05,
+      "loss": 1.6464,
+      "step": 1911
+    },
+    {
+      "epoch": 0.4617242212026081,
+      "grad_norm": 0.29917535185813904,
+      "learning_rate": 5.77108306426434e-05,
+      "loss": 1.7856,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4619657087659986,
+      "grad_norm": 0.2910739779472351,
+      "learning_rate": 5.767258277172305e-05,
+      "loss": 1.7508,
+      "step": 1913
+    },
+    {
+      "epoch": 0.46220719632938906,
+      "grad_norm": 0.29367324709892273,
+      "learning_rate": 5.7634330302291795e-05,
+      "loss": 1.6982,
+      "step": 1914
+    },
+    {
+      "epoch": 0.46244868389277954,
+      "grad_norm": 0.297710657119751,
+      "learning_rate": 5.759607325727601e-05,
+      "loss": 1.679,
+      "step": 1915
+    },
+    {
+      "epoch": 0.46269017145617003,
+      "grad_norm": 0.316152423620224,
+      "learning_rate": 5.7557811659604766e-05,
+      "loss": 2.0318,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4629316590195605,
+      "grad_norm": 0.28985244035720825,
+      "learning_rate": 5.751954553220992e-05,
+      "loss": 1.767,
+      "step": 1917
+    },
+    {
+      "epoch": 0.463173146582951,
+      "grad_norm": 0.28868818283081055,
+      "learning_rate": 5.748127489802601e-05,
+      "loss": 1.7489,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 0.2903616428375244,
+      "learning_rate": 5.744299977999027e-05,
+      "loss": 1.663,
+      "step": 1919
+    },
+    {
+      "epoch": 0.46365612170973197,
+      "grad_norm": 0.289998859167099,
+      "learning_rate": 5.7404720201042664e-05,
+      "loss": 1.796,
+      "step": 1920
+    },
+    {
+      "epoch": 0.46389760927312246,
+      "grad_norm": 0.3104515075683594,
+      "learning_rate": 5.736643618412579e-05,
+      "loss": 1.8674,
+      "step": 1921
+    },
+    {
+      "epoch": 0.46413909683651294,
+      "grad_norm": 0.30258458852767944,
+      "learning_rate": 5.73281477521849e-05,
+      "loss": 1.7722,
+      "step": 1922
+    },
+    {
+      "epoch": 0.46438058439990343,
+      "grad_norm": 0.2958606779575348,
+      "learning_rate": 5.728985492816793e-05,
+      "loss": 1.7653,
+      "step": 1923
+    },
+    {
+      "epoch": 0.4646220719632939,
+      "grad_norm": 0.2986416220664978,
+      "learning_rate": 5.725155773502542e-05,
+      "loss": 1.7348,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4648635595266844,
+      "grad_norm": 0.28947514295578003,
+      "learning_rate": 5.7213256195710516e-05,
+      "loss": 1.7064,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4651050470900749,
+      "grad_norm": 0.29144924879074097,
+      "learning_rate": 5.717495033317901e-05,
+      "loss": 1.7251,
+      "step": 1926
+    },
+    {
+      "epoch": 0.46534653465346537,
+      "grad_norm": 0.2858518660068512,
+      "learning_rate": 5.7136640170389263e-05,
+      "loss": 1.5974,
+      "step": 1927
+    },
+    {
+      "epoch": 0.46558802221685586,
+      "grad_norm": 0.3050241768360138,
+      "learning_rate": 5.7098325730302195e-05,
+      "loss": 1.6366,
+      "step": 1928
+    },
+    {
+      "epoch": 0.46582950978024634,
+      "grad_norm": 0.3163412809371948,
+      "learning_rate": 5.7060007035881324e-05,
+      "loss": 1.7627,
+      "step": 1929
+    },
+    {
+      "epoch": 0.46607099734363683,
+      "grad_norm": 0.31081393361091614,
+      "learning_rate": 5.702168411009271e-05,
+      "loss": 1.9454,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4663124849070273,
+      "grad_norm": 0.2910256087779999,
+      "learning_rate": 5.698335697590491e-05,
+      "loss": 1.665,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4665539724704178,
+      "grad_norm": 0.3363916277885437,
+      "learning_rate": 5.6945025656289076e-05,
+      "loss": 1.8586,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4667954600338083,
+      "grad_norm": 0.29299837350845337,
+      "learning_rate": 5.690669017421879e-05,
+      "loss": 1.714,
+      "step": 1933
+    },
+    {
+      "epoch": 0.46703694759719877,
+      "grad_norm": 0.2883269488811493,
+      "learning_rate": 5.686835055267018e-05,
+      "loss": 1.7377,
+      "step": 1934
+    },
+    {
+      "epoch": 0.46727843516058926,
+      "grad_norm": 0.2988598644733429,
+      "learning_rate": 5.683000681462185e-05,
+      "loss": 1.8745,
+      "step": 1935
+    },
+    {
+      "epoch": 0.46751992272397974,
+      "grad_norm": 0.28724566102027893,
+      "learning_rate": 5.679165898305485e-05,
+      "loss": 1.6266,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4677614102873702,
+      "grad_norm": 0.3093075752258301,
+      "learning_rate": 5.675330708095268e-05,
+      "loss": 1.7257,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4680028978507607,
+      "grad_norm": 0.2962707579135895,
+      "learning_rate": 5.671495113130133e-05,
+      "loss": 1.7822,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4682443854141512,
+      "grad_norm": 0.29928648471832275,
+      "learning_rate": 5.6676591157089166e-05,
+      "loss": 1.8773,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4684858729775417,
+      "grad_norm": 0.29388922452926636,
+      "learning_rate": 5.663822718130696e-05,
+      "loss": 1.812,
+      "step": 1940
+    },
+    {
+      "epoch": 0.46872736054093217,
+      "grad_norm": 0.30245307087898254,
+      "learning_rate": 5.659985922694792e-05,
+      "loss": 1.8397,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4689688481043226,
+      "grad_norm": 0.2745733857154846,
+      "learning_rate": 5.656148731700763e-05,
+      "loss": 1.537,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4692103356677131,
+      "grad_norm": 0.3115253150463104,
+      "learning_rate": 5.6523111474484016e-05,
+      "loss": 1.7199,
+      "step": 1943
+    },
+    {
+      "epoch": 0.46945182323110357,
+      "grad_norm": 0.29406628012657166,
+      "learning_rate": 5.64847317223774e-05,
+      "loss": 1.5904,
+      "step": 1944
+    },
+    {
+      "epoch": 0.46969331079449406,
+      "grad_norm": 0.28329282999038696,
+      "learning_rate": 5.644634808369044e-05,
+      "loss": 1.5586,
+      "step": 1945
+    },
+    {
+      "epoch": 0.46993479835788454,
+      "grad_norm": 0.2810218334197998,
+      "learning_rate": 5.6407960581428064e-05,
+      "loss": 1.6041,
+      "step": 1946
+    },
+    {
+      "epoch": 0.47017628592127503,
+      "grad_norm": 0.29527994990348816,
+      "learning_rate": 5.6369569238597606e-05,
+      "loss": 1.7334,
+      "step": 1947
+    },
+    {
+      "epoch": 0.4704177734846655,
+      "grad_norm": 0.2843940556049347,
+      "learning_rate": 5.633117407820865e-05,
+      "loss": 1.7189,
+      "step": 1948
+    },
+    {
+      "epoch": 0.470659261048056,
+      "grad_norm": 0.2926733195781708,
+      "learning_rate": 5.629277512327306e-05,
+      "loss": 1.6232,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4709007486114465,
+      "grad_norm": 0.2885076403617859,
+      "learning_rate": 5.625437239680501e-05,
+      "loss": 1.613,
+      "step": 1950
+    },
+    {
+      "epoch": 0.47114223617483697,
+      "grad_norm": 0.2991282045841217,
+      "learning_rate": 5.621596592182091e-05,
+      "loss": 1.8069,
+      "step": 1951
+    },
+    {
+      "epoch": 0.47138372373822746,
+      "grad_norm": 0.28910577297210693,
+      "learning_rate": 5.617755572133942e-05,
+      "loss": 1.6859,
+      "step": 1952
+    },
+    {
+      "epoch": 0.47162521130161794,
+      "grad_norm": 0.30354511737823486,
+      "learning_rate": 5.6139141818381446e-05,
+      "loss": 1.8275,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4718666988650084,
+      "grad_norm": 0.31015917658805847,
+      "learning_rate": 5.610072423597008e-05,
+      "loss": 1.7884,
+      "step": 1954
+    },
+    {
+      "epoch": 0.4721081864283989,
+      "grad_norm": 0.29766714572906494,
+      "learning_rate": 5.606230299713068e-05,
+      "loss": 1.7359,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4723496739917894,
+      "grad_norm": 0.28761327266693115,
+      "learning_rate": 5.6023878124890716e-05,
+      "loss": 1.6276,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4725911615551799,
+      "grad_norm": 0.2790350317955017,
+      "learning_rate": 5.59854496422799e-05,
+      "loss": 1.668,
+      "step": 1957
+    },
+    {
+      "epoch": 0.47283264911857037,
+      "grad_norm": 0.2867283225059509,
+      "learning_rate": 5.5947017572330084e-05,
+      "loss": 1.6826,
+      "step": 1958
+    },
+    {
+      "epoch": 0.47307413668196086,
+      "grad_norm": 0.295149028301239,
+      "learning_rate": 5.5908581938075264e-05,
+      "loss": 1.7293,
+      "step": 1959
+    },
+    {
+      "epoch": 0.47331562424535134,
+      "grad_norm": 0.28681740164756775,
+      "learning_rate": 5.587014276255157e-05,
+      "loss": 1.6374,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4735571118087418,
+      "grad_norm": 0.30145061016082764,
+      "learning_rate": 5.5831700068797286e-05,
+      "loss": 1.7667,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4737985993721323,
+      "grad_norm": 0.28270459175109863,
+      "learning_rate": 5.579325387985277e-05,
+      "loss": 1.5991,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4740400869355228,
+      "grad_norm": 0.3064315915107727,
+      "learning_rate": 5.575480421876048e-05,
+      "loss": 1.6987,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4742815744989133,
+      "grad_norm": 0.3125905692577362,
+      "learning_rate": 5.5716351108564965e-05,
+      "loss": 1.829,
+      "step": 1964
+    },
+    {
+      "epoch": 0.47452306206230377,
+      "grad_norm": 0.27573850750923157,
+      "learning_rate": 5.567789457231286e-05,
+      "loss": 1.5916,
+      "step": 1965
+    },
+    {
+      "epoch": 0.47476454962569425,
+      "grad_norm": 0.31593382358551025,
+      "learning_rate": 5.56394346330528e-05,
+      "loss": 1.7972,
+      "step": 1966
+    },
+    {
+      "epoch": 0.47500603718908474,
+      "grad_norm": 0.31528183817863464,
+      "learning_rate": 5.560097131383552e-05,
+      "loss": 1.7862,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4752475247524752,
+      "grad_norm": 0.29852205514907837,
+      "learning_rate": 5.556250463771373e-05,
+      "loss": 1.7726,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4754890123158657,
+      "grad_norm": 0.2947467863559723,
+      "learning_rate": 5.5524034627742174e-05,
+      "loss": 1.5666,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4757304998792562,
+      "grad_norm": 0.2866770923137665,
+      "learning_rate": 5.5485561306977604e-05,
+      "loss": 1.683,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4759719874426467,
+      "grad_norm": 0.29122698307037354,
+      "learning_rate": 5.5447084698478746e-05,
+      "loss": 1.5568,
+      "step": 1971
+    },
+    {
+      "epoch": 0.47621347500603717,
+      "grad_norm": 0.29016897082328796,
+      "learning_rate": 5.540860482530629e-05,
+      "loss": 1.8156,
+      "step": 1972
+    },
+    {
+      "epoch": 0.47645496256942765,
+      "grad_norm": 0.30687373876571655,
+      "learning_rate": 5.537012171052289e-05,
+      "loss": 1.9504,
+      "step": 1973
+    },
+    {
+      "epoch": 0.47669645013281814,
+      "grad_norm": 0.2921820282936096,
+      "learning_rate": 5.533163537719315e-05,
+      "loss": 1.7727,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4769379376962086,
+      "grad_norm": 0.28375446796417236,
+      "learning_rate": 5.5293145848383576e-05,
+      "loss": 1.6884,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4771794252595991,
+      "grad_norm": 0.2858161926269531,
+      "learning_rate": 5.5254653147162626e-05,
+      "loss": 1.6366,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4774209128229896,
+      "grad_norm": 0.2857281565666199,
+      "learning_rate": 5.521615729660066e-05,
+      "loss": 1.6584,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4776624003863801,
+      "grad_norm": 0.37559181451797485,
+      "learning_rate": 5.5177658319769856e-05,
+      "loss": 1.7549,
+      "step": 1978
+    },
+    {
+      "epoch": 0.47790388794977057,
+      "grad_norm": 0.2909361720085144,
+      "learning_rate": 5.513915623974436e-05,
+      "loss": 1.7672,
+      "step": 1979
+    },
+    {
+      "epoch": 0.47814537551316105,
+      "grad_norm": 0.28941604495048523,
+      "learning_rate": 5.510065107960013e-05,
+      "loss": 1.7656,
+      "step": 1980
+    },
+    {
+      "epoch": 0.47838686307655154,
+      "grad_norm": 0.31622543931007385,
+      "learning_rate": 5.506214286241495e-05,
+      "loss": 1.9301,
+      "step": 1981
+    },
+    {
+      "epoch": 0.478628350639942,
+      "grad_norm": 0.3030341565608978,
+      "learning_rate": 5.5023631611268486e-05,
+      "loss": 1.7637,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4788698382033325,
+      "grad_norm": 0.31502172350883484,
+      "learning_rate": 5.498511734924219e-05,
+      "loss": 1.735,
+      "step": 1983
+    },
+    {
+      "epoch": 0.479111325766723,
+      "grad_norm": 0.2744358479976654,
+      "learning_rate": 5.4946600099419306e-05,
+      "loss": 1.7905,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4793528133301135,
+      "grad_norm": 0.2697215974330902,
+      "learning_rate": 5.4908079884884925e-05,
+      "loss": 1.5816,
+      "step": 1985
+    },
+    {
+      "epoch": 0.47959430089350397,
+      "grad_norm": 0.27491477131843567,
+      "learning_rate": 5.486955672872584e-05,
+      "loss": 1.6728,
+      "step": 1986
+    },
+    {
+      "epoch": 0.47983578845689445,
+      "grad_norm": 0.2729800343513489,
+      "learning_rate": 5.483103065403069e-05,
+      "loss": 1.4615,
+      "step": 1987
+    },
+    {
+      "epoch": 0.48007727602028494,
+      "grad_norm": 0.2833001911640167,
+      "learning_rate": 5.4792501683889784e-05,
+      "loss": 1.6446,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4803187635836754,
+      "grad_norm": 0.2760225832462311,
+      "learning_rate": 5.4753969841395224e-05,
+      "loss": 1.6635,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4805602511470659,
+      "grad_norm": 0.3052677512168884,
+      "learning_rate": 5.471543514964078e-05,
+      "loss": 1.6142,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4808017387104564,
+      "grad_norm": 0.2885996997356415,
+      "learning_rate": 5.4676897631721993e-05,
+      "loss": 1.6738,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4810432262738469,
+      "grad_norm": 0.28482338786125183,
+      "learning_rate": 5.4638357310736054e-05,
+      "loss": 1.71,
+      "step": 1992
+    },
+    {
+      "epoch": 0.48128471383723737,
+      "grad_norm": 0.2904854118824005,
+      "learning_rate": 5.459981420978183e-05,
+      "loss": 1.6673,
+      "step": 1993
+    },
+    {
+      "epoch": 0.48152620140062785,
+      "grad_norm": 0.3114590346813202,
+      "learning_rate": 5.456126835195989e-05,
+      "loss": 1.9513,
+      "step": 1994
+    },
+    {
+      "epoch": 0.48176768896401834,
+      "grad_norm": 0.30344778299331665,
+      "learning_rate": 5.4522719760372435e-05,
+      "loss": 1.779,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4820091765274088,
+      "grad_norm": 0.29543033242225647,
+      "learning_rate": 5.448416845812329e-05,
+      "loss": 1.6585,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4822506640907993,
+      "grad_norm": 0.29839324951171875,
+      "learning_rate": 5.444561446831792e-05,
+      "loss": 1.7803,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4824921516541898,
+      "grad_norm": 0.2827812731266022,
+      "learning_rate": 5.440705781406341e-05,
+      "loss": 1.7993,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4827336392175803,
+      "grad_norm": 0.292828768491745,
+      "learning_rate": 5.4368498518468414e-05,
+      "loss": 1.6683,
+      "step": 1999
+    },
+    {
+      "epoch": 0.48297512678097076,
+      "grad_norm": 0.29965028166770935,
+      "learning_rate": 5.4329936604643216e-05,
+      "loss": 1.7802,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4141,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.082997855158272e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}