diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,42978 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 15.0,
+  "eval_steps": 500,
+  "global_step": 6135,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0024449877750611247,
+      "grad_norm": 0.056970333624040005,
+      "learning_rate": 4.9999996722213865e-06,
+      "loss": 0.6502,
+      "step": 1
+    },
+    {
+      "epoch": 0.004889975550122249,
+      "grad_norm": 0.0642923240948625,
+      "learning_rate": 4.99999868888563e-06,
+      "loss": 0.7052,
+      "step": 2
+    },
+    {
+      "epoch": 0.007334963325183374,
+      "grad_norm": 0.07295774500273575,
+      "learning_rate": 4.999997049992989e-06,
+      "loss": 1.163,
+      "step": 3
+    },
+    {
+      "epoch": 0.009779951100244499,
+      "grad_norm": 0.08454639435282074,
+      "learning_rate": 4.999994755543892e-06,
+      "loss": 0.7364,
+      "step": 4
+    },
+    {
+      "epoch": 0.012224938875305624,
+      "grad_norm": 0.06591401969976246,
+      "learning_rate": 4.999991805538943e-06,
+      "loss": 0.8502,
+      "step": 5
+    },
+    {
+      "epoch": 0.014669926650366748,
+      "grad_norm": 0.056687605494465804,
+      "learning_rate": 4.999988199978913e-06,
+      "loss": 0.5687,
+      "step": 6
+    },
+    {
+      "epoch": 0.017114914425427872,
+      "grad_norm": 0.07161055534855067,
+      "learning_rate": 4.999983938864751e-06,
+      "loss": 0.6998,
+      "step": 7
+    },
+    {
+      "epoch": 0.019559902200488997,
+      "grad_norm": 0.058300557445260116,
+      "learning_rate": 4.999979022197571e-06,
+      "loss": 0.6303,
+      "step": 8
+    },
+    {
+      "epoch": 0.022004889975550123,
+      "grad_norm": 0.05437433502479284,
+      "learning_rate": 4.999973449978664e-06,
+      "loss": 0.5753,
+      "step": 9
+    },
+    {
+      "epoch": 0.02444987775061125,
+      "grad_norm": 0.06723667478202587,
+      "learning_rate": 4.999967222209489e-06,
+      "loss": 0.9733,
+      "step": 10
+    },
+    {
+      "epoch": 0.02689486552567237,
+      "grad_norm": 0.08315652473020697,
+      "learning_rate": 4.999960338891682e-06,
+      "loss": 1.1614,
+      "step": 11
+    },
+    {
+      "epoch": 0.029339853300733496,
+      "grad_norm": 0.07785854113341294,
+      "learning_rate": 4.999952800027046e-06,
+      "loss": 0.7975,
+      "step": 12
+    },
+    {
+      "epoch": 0.03178484107579462,
+      "grad_norm": 0.07307268247146252,
+      "learning_rate": 4.999944605617559e-06,
+      "loss": 1.0792,
+      "step": 13
+    },
+    {
+      "epoch": 0.034229828850855744,
+      "grad_norm": 0.06504667944919901,
+      "learning_rate": 4.999935755665369e-06,
+      "loss": 0.6103,
+      "step": 14
+    },
+    {
+      "epoch": 0.03667481662591687,
+      "grad_norm": 0.06531060627142818,
+      "learning_rate": 4.999926250172798e-06,
+      "loss": 0.6697,
+      "step": 15
+    },
+    {
+      "epoch": 0.039119804400977995,
+      "grad_norm": 0.057727976859811216,
+      "learning_rate": 4.999916089142336e-06,
+      "loss": 0.6186,
+      "step": 16
+    },
+    {
+      "epoch": 0.04156479217603912,
+      "grad_norm": 0.07073003216152718,
+      "learning_rate": 4.99990527257665e-06,
+      "loss": 1.0184,
+      "step": 17
+    },
+    {
+      "epoch": 0.044009779951100246,
+      "grad_norm": 0.06607794689704095,
+      "learning_rate": 4.999893800478573e-06,
+      "loss": 0.6661,
+      "step": 18
+    },
+    {
+      "epoch": 0.04645476772616137,
+      "grad_norm": 0.09361726958951792,
+      "learning_rate": 4.999881672851118e-06,
+      "loss": 0.9106,
+      "step": 19
+    },
+    {
+      "epoch": 0.0488997555012225,
+      "grad_norm": 0.07718233312983057,
+      "learning_rate": 4.9998688896974625e-06,
+      "loss": 0.7479,
+      "step": 20
+    },
+    {
+      "epoch": 0.05134474327628362,
+      "grad_norm": 0.06509269497362077,
+      "learning_rate": 4.999855451020958e-06,
+      "loss": 0.6082,
+      "step": 21
+    },
+    {
+      "epoch": 0.05378973105134474,
+      "grad_norm": 0.07152985737396737,
+      "learning_rate": 4.999841356825129e-06,
+      "loss": 0.6822,
+      "step": 22
+    },
+    {
+      "epoch": 0.05623471882640587,
+      "grad_norm": 0.07769426391009598,
+      "learning_rate": 4.9998266071136714e-06,
+      "loss": 0.7239,
+      "step": 23
+    },
+    {
+      "epoch": 0.05867970660146699,
+      "grad_norm": 0.08276748551801542,
+      "learning_rate": 4.999811201890453e-06,
+      "loss": 0.6978,
+      "step": 24
+    },
+    {
+      "epoch": 0.061124694376528114,
+      "grad_norm": 0.09628639792438537,
+      "learning_rate": 4.999795141159514e-06,
+      "loss": 0.8996,
+      "step": 25
+    },
+    {
+      "epoch": 0.06356968215158924,
+      "grad_norm": 0.06258737253229976,
+      "learning_rate": 4.999778424925065e-06,
+      "loss": 0.5211,
+      "step": 26
+    },
+    {
+      "epoch": 0.06601466992665037,
+      "grad_norm": 0.08811261681853647,
+      "learning_rate": 4.99976105319149e-06,
+      "loss": 0.7886,
+      "step": 27
+    },
+    {
+      "epoch": 0.06845965770171149,
+      "grad_norm": 0.09319511948769583,
+      "learning_rate": 4.999743025963343e-06,
+      "loss": 0.9864,
+      "step": 28
+    },
+    {
+      "epoch": 0.07090464547677261,
+      "grad_norm": 0.082190275231665,
+      "learning_rate": 4.999724343245352e-06,
+      "loss": 0.8088,
+      "step": 29
+    },
+    {
+      "epoch": 0.07334963325183375,
+      "grad_norm": 0.10764497651069097,
+      "learning_rate": 4.999705005042417e-06,
+      "loss": 0.6992,
+      "step": 30
+    },
+    {
+      "epoch": 0.07579462102689487,
+      "grad_norm": 0.09511750481525305,
+      "learning_rate": 4.9996850113596075e-06,
+      "loss": 0.8331,
+      "step": 31
+    },
+    {
+      "epoch": 0.07823960880195599,
+      "grad_norm": 0.09242231072962526,
+      "learning_rate": 4.999664362202167e-06,
+      "loss": 0.708,
+      "step": 32
+    },
+    {
+      "epoch": 0.08068459657701711,
+      "grad_norm": 0.08341700201579721,
+      "learning_rate": 4.99964305757551e-06,
+      "loss": 1.0805,
+      "step": 33
+    },
+    {
+      "epoch": 0.08312958435207823,
+      "grad_norm": 0.06155034900954256,
+      "learning_rate": 4.999621097485223e-06,
+      "loss": 0.4807,
+      "step": 34
+    },
+    {
+      "epoch": 0.08557457212713937,
+      "grad_norm": 0.08420997143915426,
+      "learning_rate": 4.999598481937064e-06,
+      "loss": 0.7115,
+      "step": 35
+    },
+    {
+      "epoch": 0.08801955990220049,
+      "grad_norm": 0.08645529689872837,
+      "learning_rate": 4.999575210936965e-06,
+      "loss": 0.6044,
+      "step": 36
+    },
+    {
+      "epoch": 0.09046454767726161,
+      "grad_norm": 0.07531317485402605,
+      "learning_rate": 4.999551284491027e-06,
+      "loss": 0.8446,
+      "step": 37
+    },
+    {
+      "epoch": 0.09290953545232274,
+      "grad_norm": 0.09254647443966509,
+      "learning_rate": 4.9995267026055225e-06,
+      "loss": 1.0609,
+      "step": 38
+    },
+    {
+      "epoch": 0.09535452322738386,
+      "grad_norm": 0.06902120774533693,
+      "learning_rate": 4.9995014652869e-06,
+      "loss": 0.5849,
+      "step": 39
+    },
+    {
+      "epoch": 0.097799511002445,
+      "grad_norm": 0.11450312213133851,
+      "learning_rate": 4.999475572541776e-06,
+      "loss": 1.2489,
+      "step": 40
+    },
+    {
+      "epoch": 0.10024449877750612,
+      "grad_norm": 0.12143513035395985,
+      "learning_rate": 4.9994490243769406e-06,
+      "loss": 0.8074,
+      "step": 41
+    },
+    {
+      "epoch": 0.10268948655256724,
+      "grad_norm": 0.10095751319167787,
+      "learning_rate": 4.999421820799355e-06,
+      "loss": 0.8203,
+      "step": 42
+    },
+    {
+      "epoch": 0.10513447432762836,
+      "grad_norm": 0.08911137880429622,
+      "learning_rate": 4.999393961816154e-06,
+      "loss": 0.7197,
+      "step": 43
+    },
+    {
+      "epoch": 0.10757946210268948,
+      "grad_norm": 0.129400332443115,
+      "learning_rate": 4.99936544743464e-06,
+      "loss": 1.2678,
+      "step": 44
+    },
+    {
+      "epoch": 0.1100244498777506,
+      "grad_norm": 0.10031345707812848,
+      "learning_rate": 4.9993362776622924e-06,
+      "loss": 1.5299,
+      "step": 45
+    },
+    {
+      "epoch": 0.11246943765281174,
+      "grad_norm": 0.07753346019013206,
+      "learning_rate": 4.999306452506759e-06,
+      "loss": 0.6727,
+      "step": 46
+    },
+    {
+      "epoch": 0.11491442542787286,
+      "grad_norm": 0.08482223258078687,
+      "learning_rate": 4.999275971975862e-06,
+      "loss": 0.4733,
+      "step": 47
+    },
+    {
+      "epoch": 0.11735941320293398,
+      "grad_norm": 0.12129946290133806,
+      "learning_rate": 4.999244836077593e-06,
+      "loss": 0.8762,
+      "step": 48
+    },
+    {
+      "epoch": 0.1198044009779951,
+      "grad_norm": 0.10928653456458395,
+      "learning_rate": 4.9992130448201174e-06,
+      "loss": 0.5832,
+      "step": 49
+    },
+    {
+      "epoch": 0.12224938875305623,
+      "grad_norm": 0.10988768445643009,
+      "learning_rate": 4.999180598211771e-06,
+      "loss": 0.7233,
+      "step": 50
+    },
+    {
+      "epoch": 0.12469437652811736,
+      "grad_norm": 0.1053130155345633,
+      "learning_rate": 4.999147496261062e-06,
+      "loss": 0.8271,
+      "step": 51
+    },
+    {
+      "epoch": 0.1271393643031785,
+      "grad_norm": 0.10759620901058406,
+      "learning_rate": 4.999113738976669e-06,
+      "loss": 0.6885,
+      "step": 52
+    },
+    {
+      "epoch": 0.1295843520782396,
+      "grad_norm": 0.11577751614456157,
+      "learning_rate": 4.999079326367448e-06,
+      "loss": 0.701,
+      "step": 53
+    },
+    {
+      "epoch": 0.13202933985330073,
+      "grad_norm": 0.09535428814755058,
+      "learning_rate": 4.999044258442419e-06,
+      "loss": 0.6498,
+      "step": 54
+    },
+    {
+      "epoch": 0.13447432762836187,
+      "grad_norm": 0.16209145805899675,
+      "learning_rate": 4.999008535210778e-06,
+      "loss": 0.691,
+      "step": 55
+    },
+    {
+      "epoch": 0.13691931540342298,
+      "grad_norm": 0.12612499277133693,
+      "learning_rate": 4.998972156681895e-06,
+      "loss": 0.7276,
+      "step": 56
+    },
+    {
+      "epoch": 0.1393643031784841,
+      "grad_norm": 0.10406117297804278,
+      "learning_rate": 4.9989351228653075e-06,
+      "loss": 0.7196,
+      "step": 57
+    },
+    {
+      "epoch": 0.14180929095354522,
+      "grad_norm": 0.11912931166857814,
+      "learning_rate": 4.998897433770726e-06,
+      "loss": 0.9538,
+      "step": 58
+    },
+    {
+      "epoch": 0.14425427872860636,
+      "grad_norm": 0.09672628819333685,
+      "learning_rate": 4.998859089408034e-06,
+      "loss": 1.1227,
+      "step": 59
+    },
+    {
+      "epoch": 0.1466992665036675,
+      "grad_norm": 0.10215295318189853,
+      "learning_rate": 4.998820089787287e-06,
+      "loss": 0.6743,
+      "step": 60
+    },
+    {
+      "epoch": 0.1491442542787286,
+      "grad_norm": 0.11280790975393021,
+      "learning_rate": 4.998780434918712e-06,
+      "loss": 0.5333,
+      "step": 61
+    },
+    {
+      "epoch": 0.15158924205378974,
+      "grad_norm": 0.1774356389549198,
+      "learning_rate": 4.998740124812704e-06,
+      "loss": 0.9371,
+      "step": 62
+    },
+    {
+      "epoch": 0.15403422982885084,
+      "grad_norm": 0.10426167157172311,
+      "learning_rate": 4.9986991594798375e-06,
+      "loss": 0.5998,
+      "step": 63
+    },
+    {
+      "epoch": 0.15647921760391198,
+      "grad_norm": 0.11819600790108543,
+      "learning_rate": 4.998657538930852e-06,
+      "loss": 0.8305,
+      "step": 64
+    },
+    {
+      "epoch": 0.15892420537897312,
+      "grad_norm": 0.10182772481798326,
+      "learning_rate": 4.998615263176663e-06,
+      "loss": 0.6427,
+      "step": 65
+    },
+    {
+      "epoch": 0.16136919315403422,
+      "grad_norm": 0.10620039619674311,
+      "learning_rate": 4.998572332228355e-06,
+      "loss": 0.5889,
+      "step": 66
+    },
+    {
+      "epoch": 0.16381418092909536,
+      "grad_norm": 0.12021868850600481,
+      "learning_rate": 4.9985287460971856e-06,
+      "loss": 0.6706,
+      "step": 67
+    },
+    {
+      "epoch": 0.16625916870415647,
+      "grad_norm": 0.10893869913926743,
+      "learning_rate": 4.998484504794584e-06,
+      "loss": 0.6363,
+      "step": 68
+    },
+    {
+      "epoch": 0.1687041564792176,
+      "grad_norm": 0.13194236620987754,
+      "learning_rate": 4.998439608332152e-06,
+      "loss": 1.3301,
+      "step": 69
+    },
+    {
+      "epoch": 0.17114914425427874,
+      "grad_norm": 0.17068657483797256,
+      "learning_rate": 4.998394056721661e-06,
+      "loss": 1.0071,
+      "step": 70
+    },
+    {
+      "epoch": 0.17359413202933985,
+      "grad_norm": 0.10829113413634824,
+      "learning_rate": 4.998347849975057e-06,
+      "loss": 0.7784,
+      "step": 71
+    },
+    {
+      "epoch": 0.17603911980440098,
+      "grad_norm": 0.1291929918135214,
+      "learning_rate": 4.998300988104457e-06,
+      "loss": 0.7124,
+      "step": 72
+    },
+    {
+      "epoch": 0.1784841075794621,
+      "grad_norm": 0.1235367727071986,
+      "learning_rate": 4.998253471122148e-06,
+      "loss": 0.6257,
+      "step": 73
+    },
+    {
+      "epoch": 0.18092909535452323,
+      "grad_norm": 0.12967411869926201,
+      "learning_rate": 4.9982052990405895e-06,
+      "loss": 0.6845,
+      "step": 74
+    },
+    {
+      "epoch": 0.18337408312958436,
+      "grad_norm": 0.11283010140159752,
+      "learning_rate": 4.998156471872415e-06,
+      "loss": 0.7685,
+      "step": 75
+    },
+    {
+      "epoch": 0.18581907090464547,
+      "grad_norm": 0.14859626261811418,
+      "learning_rate": 4.998106989630428e-06,
+      "loss": 0.9241,
+      "step": 76
+    },
+    {
+      "epoch": 0.1882640586797066,
+      "grad_norm": 0.10453236434422224,
+      "learning_rate": 4.998056852327602e-06,
+      "loss": 0.8729,
+      "step": 77
+    },
+    {
+      "epoch": 0.19070904645476772,
+      "grad_norm": 0.16878089089242582,
+      "learning_rate": 4.998006059977087e-06,
+      "loss": 0.6309,
+      "step": 78
+    },
+    {
+      "epoch": 0.19315403422982885,
+      "grad_norm": 0.11545542520278311,
+      "learning_rate": 4.997954612592198e-06,
+      "loss": 0.7567,
+      "step": 79
+    },
+    {
+      "epoch": 0.19559902200489,
+      "grad_norm": 0.1375865837735131,
+      "learning_rate": 4.9979025101864294e-06,
+      "loss": 1.0424,
+      "step": 80
+    },
+    {
+      "epoch": 0.1980440097799511,
+      "grad_norm": 0.12810477596360154,
+      "learning_rate": 4.997849752773443e-06,
+      "loss": 0.7293,
+      "step": 81
+    },
+    {
+      "epoch": 0.20048899755501223,
+      "grad_norm": 0.13802070118221518,
+      "learning_rate": 4.997796340367071e-06,
+      "loss": 0.9855,
+      "step": 82
+    },
+    {
+      "epoch": 0.20293398533007334,
+      "grad_norm": 0.12223213324491261,
+      "learning_rate": 4.99774227298132e-06,
+      "loss": 0.6349,
+      "step": 83
+    },
+    {
+      "epoch": 0.20537897310513448,
+      "grad_norm": 0.12684251294283894,
+      "learning_rate": 4.99768755063037e-06,
+      "loss": 0.8088,
+      "step": 84
+    },
+    {
+      "epoch": 0.2078239608801956,
+      "grad_norm": 0.13824509671408847,
+      "learning_rate": 4.997632173328568e-06,
+      "loss": 0.6177,
+      "step": 85
+    },
+    {
+      "epoch": 0.21026894865525672,
+      "grad_norm": 0.15409914738654823,
+      "learning_rate": 4.997576141090436e-06,
+      "loss": 0.9274,
+      "step": 86
+    },
+    {
+      "epoch": 0.21271393643031786,
+      "grad_norm": 0.1465455864569219,
+      "learning_rate": 4.9975194539306675e-06,
+      "loss": 1.0025,
+      "step": 87
+    },
+    {
+      "epoch": 0.21515892420537897,
+      "grad_norm": 0.08760359472319448,
+      "learning_rate": 4.9974621118641255e-06,
+      "loss": 0.6117,
+      "step": 88
+    },
+    {
+      "epoch": 0.2176039119804401,
+      "grad_norm": 0.10955988313080253,
+      "learning_rate": 4.997404114905847e-06,
+      "loss": 0.6335,
+      "step": 89
+    },
+    {
+      "epoch": 0.2200488997555012,
+      "grad_norm": 0.13032374035452762,
+      "learning_rate": 4.997345463071041e-06,
+      "loss": 0.9161,
+      "step": 90
+    },
+    {
+      "epoch": 0.22249388753056235,
+      "grad_norm": 0.10333150887648376,
+      "learning_rate": 4.997286156375088e-06,
+      "loss": 0.5768,
+      "step": 91
+    },
+    {
+      "epoch": 0.22493887530562348,
+      "grad_norm": 0.11095398870073611,
+      "learning_rate": 4.997226194833537e-06,
+      "loss": 0.6353,
+      "step": 92
+    },
+    {
+      "epoch": 0.2273838630806846,
+      "grad_norm": 0.11422289268164039,
+      "learning_rate": 4.9971655784621146e-06,
+      "loss": 1.0907,
+      "step": 93
+    },
+    {
+      "epoch": 0.22982885085574573,
+      "grad_norm": 0.10393048195804086,
+      "learning_rate": 4.9971043072767126e-06,
+      "loss": 0.7662,
+      "step": 94
+    },
+    {
+      "epoch": 0.23227383863080683,
+      "grad_norm": 0.17083560410434256,
+      "learning_rate": 4.9970423812934e-06,
+      "loss": 1.3592,
+      "step": 95
+    },
+    {
+      "epoch": 0.23471882640586797,
+      "grad_norm": 0.12888083529675434,
+      "learning_rate": 4.996979800528413e-06,
+      "loss": 0.6691,
+      "step": 96
+    },
+    {
+      "epoch": 0.2371638141809291,
+      "grad_norm": 0.09896919452550294,
+      "learning_rate": 4.996916564998165e-06,
+      "loss": 0.646,
+      "step": 97
+    },
+    {
+      "epoch": 0.2396088019559902,
+      "grad_norm": 0.13022303525623774,
+      "learning_rate": 4.996852674719234e-06,
+      "loss": 0.962,
+      "step": 98
+    },
+    {
+      "epoch": 0.24205378973105135,
+      "grad_norm": 0.13590235028802428,
+      "learning_rate": 4.996788129708375e-06,
+      "loss": 1.3377,
+      "step": 99
+    },
+    {
+      "epoch": 0.24449877750611246,
+      "grad_norm": 0.10680948537478709,
+      "learning_rate": 4.996722929982515e-06,
+      "loss": 0.6796,
+      "step": 100
+    },
+    {
+      "epoch": 0.2469437652811736,
+      "grad_norm": 0.13975673207873412,
+      "learning_rate": 4.996657075558749e-06,
+      "loss": 0.7666,
+      "step": 101
+    },
+    {
+      "epoch": 0.24938875305623473,
+      "grad_norm": 0.11324922241216796,
+      "learning_rate": 4.9965905664543455e-06,
+      "loss": 0.8291,
+      "step": 102
+    },
+    {
+      "epoch": 0.25183374083129584,
+      "grad_norm": 0.11304832716241121,
+      "learning_rate": 4.996523402686745e-06,
+      "loss": 0.9367,
+      "step": 103
+    },
+    {
+      "epoch": 0.254278728606357,
+      "grad_norm": 0.12834753980055527,
+      "learning_rate": 4.9964555842735605e-06,
+      "loss": 0.6685,
+      "step": 104
+    },
+    {
+      "epoch": 0.2567237163814181,
+      "grad_norm": 0.08988623728847779,
+      "learning_rate": 4.996387111232574e-06,
+      "loss": 0.6556,
+      "step": 105
+    },
+    {
+      "epoch": 0.2591687041564792,
+      "grad_norm": 0.12673117114876908,
+      "learning_rate": 4.99631798358174e-06,
+      "loss": 1.2048,
+      "step": 106
+    },
+    {
+      "epoch": 0.2616136919315403,
+      "grad_norm": 0.12374158370008943,
+      "learning_rate": 4.996248201339188e-06,
+      "loss": 0.6128,
+      "step": 107
+    },
+    {
+      "epoch": 0.26405867970660146,
+      "grad_norm": 0.09576497724521663,
+      "learning_rate": 4.996177764523215e-06,
+      "loss": 0.9087,
+      "step": 108
+    },
+    {
+      "epoch": 0.2665036674816626,
+      "grad_norm": 0.1299861945815397,
+      "learning_rate": 4.996106673152291e-06,
+      "loss": 0.5431,
+      "step": 109
+    },
+    {
+      "epoch": 0.26894865525672373,
+      "grad_norm": 0.11488409073340845,
+      "learning_rate": 4.996034927245057e-06,
+      "loss": 0.832,
+      "step": 110
+    },
+    {
+      "epoch": 0.2713936430317848,
+      "grad_norm": 0.16739301953270402,
+      "learning_rate": 4.9959625268203284e-06,
+      "loss": 0.7255,
+      "step": 111
+    },
+    {
+      "epoch": 0.27383863080684595,
+      "grad_norm": 0.0911522743248785,
+      "learning_rate": 4.9958894718970895e-06,
+      "loss": 0.5702,
+      "step": 112
+    },
+    {
+      "epoch": 0.2762836185819071,
+      "grad_norm": 0.13020888607160633,
+      "learning_rate": 4.995815762494496e-06,
+      "loss": 0.7724,
+      "step": 113
+    },
+    {
+      "epoch": 0.2787286063569682,
+      "grad_norm": 0.07606783563632581,
+      "learning_rate": 4.995741398631878e-06,
+      "loss": 0.5307,
+      "step": 114
+    },
+    {
+      "epoch": 0.28117359413202936,
+      "grad_norm": 0.09308595601431244,
+      "learning_rate": 4.995666380328733e-06,
+      "loss": 0.9479,
+      "step": 115
+    },
+    {
+      "epoch": 0.28361858190709044,
+      "grad_norm": 0.07887678996394007,
+      "learning_rate": 4.995590707604735e-06,
+      "loss": 0.6134,
+      "step": 116
+    },
+    {
+      "epoch": 0.2860635696821516,
+      "grad_norm": 0.07836620825794727,
+      "learning_rate": 4.9955143804797255e-06,
+      "loss": 0.6522,
+      "step": 117
+    },
+    {
+      "epoch": 0.2885085574572127,
+      "grad_norm": 0.10758103214466581,
+      "learning_rate": 4.995437398973719e-06,
+      "loss": 0.7114,
+      "step": 118
+    },
+    {
+      "epoch": 0.29095354523227385,
+      "grad_norm": 0.1028771022908388,
+      "learning_rate": 4.995359763106904e-06,
+      "loss": 0.6825,
+      "step": 119
+    },
+    {
+      "epoch": 0.293398533007335,
+      "grad_norm": 0.08658710914184022,
+      "learning_rate": 4.995281472899636e-06,
+      "loss": 0.5607,
+      "step": 120
+    },
+    {
+      "epoch": 0.29584352078239606,
+      "grad_norm": 0.10877855073612133,
+      "learning_rate": 4.995202528372446e-06,
+      "loss": 0.8574,
+      "step": 121
+    },
+    {
+      "epoch": 0.2982885085574572,
+      "grad_norm": 0.08475753927946016,
+      "learning_rate": 4.995122929546034e-06,
+      "loss": 0.5852,
+      "step": 122
+    },
+    {
+      "epoch": 0.30073349633251834,
+      "grad_norm": 0.07951060549243796,
+      "learning_rate": 4.995042676441273e-06,
+      "loss": 0.6117,
+      "step": 123
+    },
+    {
+      "epoch": 0.30317848410757947,
+      "grad_norm": 0.09217102200073693,
+      "learning_rate": 4.994961769079207e-06,
+      "loss": 0.5907,
+      "step": 124
+    },
+    {
+      "epoch": 0.3056234718826406,
+      "grad_norm": 0.08913805490526741,
+      "learning_rate": 4.994880207481053e-06,
+      "loss": 0.8031,
+      "step": 125
+    },
+    {
+      "epoch": 0.3080684596577017,
+      "grad_norm": 0.11884806379374013,
+      "learning_rate": 4.994797991668197e-06,
+      "loss": 1.0872,
+      "step": 126
+    },
+    {
+      "epoch": 0.3105134474327628,
+      "grad_norm": 0.11868854741395989,
+      "learning_rate": 4.994715121662198e-06,
+      "loss": 0.5876,
+      "step": 127
+    },
+    {
+      "epoch": 0.31295843520782396,
+      "grad_norm": 0.08173481608418573,
+      "learning_rate": 4.994631597484787e-06,
+      "loss": 0.7205,
+      "step": 128
+    },
+    {
+      "epoch": 0.3154034229828851,
+      "grad_norm": 0.08210949719491274,
+      "learning_rate": 4.994547419157866e-06,
+      "loss": 0.5876,
+      "step": 129
+    },
+    {
+      "epoch": 0.31784841075794623,
+      "grad_norm": 0.15524404469106484,
+      "learning_rate": 4.994462586703506e-06,
+      "loss": 0.6807,
+      "step": 130
+    },
+    {
+      "epoch": 0.3202933985330073,
+      "grad_norm": 0.06519568859034326,
+      "learning_rate": 4.994377100143957e-06,
+      "loss": 0.8641,
+      "step": 131
+    },
+    {
+      "epoch": 0.32273838630806845,
+      "grad_norm": 0.08912855892712686,
+      "learning_rate": 4.994290959501631e-06,
+      "loss": 0.782,
+      "step": 132
+    },
+    {
+      "epoch": 0.3251833740831296,
+      "grad_norm": 0.08772647938292119,
+      "learning_rate": 4.994204164799118e-06,
+      "loss": 0.6825,
+      "step": 133
+    },
+    {
+      "epoch": 0.3276283618581907,
+      "grad_norm": 0.08539072589695366,
+      "learning_rate": 4.994116716059177e-06,
+      "loss": 0.6012,
+      "step": 134
+    },
+    {
+      "epoch": 0.33007334963325186,
+      "grad_norm": 0.1137745075536628,
+      "learning_rate": 4.99402861330474e-06,
+      "loss": 1.1046,
+      "step": 135
+    },
+    {
+      "epoch": 0.33251833740831294,
+      "grad_norm": 0.11184409656681345,
+      "learning_rate": 4.993939856558908e-06,
+      "loss": 0.8484,
+      "step": 136
+    },
+    {
+      "epoch": 0.33496332518337407,
+      "grad_norm": 0.07911652357186882,
+      "learning_rate": 4.993850445844956e-06,
+      "loss": 0.6034,
+      "step": 137
+    },
+    {
+      "epoch": 0.3374083129584352,
+      "grad_norm": 0.12022480170363364,
+      "learning_rate": 4.99376038118633e-06,
+      "loss": 0.673,
+      "step": 138
+    },
+    {
+      "epoch": 0.33985330073349634,
+      "grad_norm": 0.12788232296499707,
+      "learning_rate": 4.993669662606646e-06,
+      "loss": 0.8366,
+      "step": 139
+    },
+    {
+      "epoch": 0.3422982885085575,
+      "grad_norm": 0.08129035332772191,
+      "learning_rate": 4.993578290129694e-06,
+      "loss": 0.6278,
+      "step": 140
+    },
+    {
+      "epoch": 0.34474327628361856,
+      "grad_norm": 0.12858988487215975,
+      "learning_rate": 4.993486263779432e-06,
+      "loss": 1.0111,
+      "step": 141
+    },
+    {
+      "epoch": 0.3471882640586797,
+      "grad_norm": 0.08500035990172021,
+      "learning_rate": 4.993393583579992e-06,
+      "loss": 0.732,
+      "step": 142
+    },
+    {
+      "epoch": 0.34963325183374083,
+      "grad_norm": 0.09539096134569047,
+      "learning_rate": 4.993300249555678e-06,
+      "loss": 0.5754,
+      "step": 143
+    },
+    {
+      "epoch": 0.35207823960880197,
+      "grad_norm": 0.0651542161701372,
+      "learning_rate": 4.993206261730963e-06,
+      "loss": 0.5034,
+      "step": 144
+    },
+    {
+      "epoch": 0.3545232273838631,
+      "grad_norm": 0.08111164431686171,
+      "learning_rate": 4.993111620130493e-06,
+      "loss": 0.7779,
+      "step": 145
+    },
+    {
+      "epoch": 0.3569682151589242,
+      "grad_norm": 0.08996874840635453,
+      "learning_rate": 4.993016324779086e-06,
+      "loss": 1.1616,
+      "step": 146
+    },
+    {
+      "epoch": 0.3594132029339853,
+      "grad_norm": 0.0879884319379973,
+      "learning_rate": 4.9929203757017295e-06,
+      "loss": 0.6133,
+      "step": 147
+    },
+    {
+      "epoch": 0.36185819070904646,
+      "grad_norm": 0.10527125061550155,
+      "learning_rate": 4.992823772923585e-06,
+      "loss": 0.8257,
+      "step": 148
+    },
+    {
+      "epoch": 0.3643031784841076,
+      "grad_norm": 0.06923364001443395,
+      "learning_rate": 4.992726516469983e-06,
+      "loss": 0.9113,
+      "step": 149
+    },
+    {
+      "epoch": 0.36674816625916873,
+      "grad_norm": 0.0911528537182838,
+      "learning_rate": 4.9926286063664265e-06,
+      "loss": 1.377,
+      "step": 150
+    },
+    {
+      "epoch": 0.3691931540342298,
+      "grad_norm": 0.06336974023058425,
+      "learning_rate": 4.9925300426385896e-06,
+      "loss": 0.5948,
+      "step": 151
+    },
+    {
+      "epoch": 0.37163814180929094,
+      "grad_norm": 0.09224131658729902,
+      "learning_rate": 4.992430825312317e-06,
+      "loss": 0.5692,
+      "step": 152
+    },
+    {
+      "epoch": 0.3740831295843521,
+      "grad_norm": 0.0678876532088674,
+      "learning_rate": 4.992330954413628e-06,
+      "loss": 0.9431,
+      "step": 153
+    },
+    {
+      "epoch": 0.3765281173594132,
+      "grad_norm": 0.0674633387529636,
+      "learning_rate": 4.99223042996871e-06,
+      "loss": 0.5367,
+      "step": 154
+    },
+    {
+      "epoch": 0.37897310513447435,
+      "grad_norm": 0.1611234007061063,
+      "learning_rate": 4.9921292520039224e-06,
+      "loss": 0.657,
+      "step": 155
+    },
+    {
+      "epoch": 0.38141809290953543,
+      "grad_norm": 0.09120305182979466,
+      "learning_rate": 4.9920274205457964e-06,
+      "loss": 0.6761,
+      "step": 156
+    },
+    {
+      "epoch": 0.38386308068459657,
+      "grad_norm": 0.10383633337877886,
+      "learning_rate": 4.991924935621035e-06,
+      "loss": 1.3519,
+      "step": 157
+    },
+    {
+      "epoch": 0.3863080684596577,
+      "grad_norm": 0.09009436772559609,
+      "learning_rate": 4.991821797256512e-06,
+      "loss": 1.1781,
+      "step": 158
+    },
+    {
+      "epoch": 0.38875305623471884,
+      "grad_norm": 0.19129170972605575,
+      "learning_rate": 4.991718005479272e-06,
+      "loss": 0.6583,
+      "step": 159
+    },
+    {
+      "epoch": 0.39119804400978,
+      "grad_norm": 0.07005021371016842,
+      "learning_rate": 4.991613560316533e-06,
+      "loss": 0.7996,
+      "step": 160
+    },
+    {
+      "epoch": 0.39364303178484106,
+      "grad_norm": 0.0996636908815799,
+      "learning_rate": 4.991508461795682e-06,
+      "loss": 0.6251,
+      "step": 161
+    },
+    {
+      "epoch": 0.3960880195599022,
+      "grad_norm": 0.1549911052603041,
+      "learning_rate": 4.991402709944277e-06,
+      "loss": 0.7505,
+      "step": 162
+    },
+    {
+      "epoch": 0.39853300733496333,
+      "grad_norm": 0.06689787692146222,
+      "learning_rate": 4.9912963047900515e-06,
+      "loss": 0.5236,
+      "step": 163
+    },
+    {
+      "epoch": 0.40097799511002447,
+      "grad_norm": 0.08118830506393378,
+      "learning_rate": 4.991189246360905e-06,
+      "loss": 1.0172,
+      "step": 164
+    },
+    {
+      "epoch": 0.4034229828850856,
+      "grad_norm": 0.06949004489354703,
+      "learning_rate": 4.991081534684911e-06,
+      "loss": 0.6692,
+      "step": 165
+    },
+    {
+      "epoch": 0.4058679706601467,
+      "grad_norm": 0.09296671391886693,
+      "learning_rate": 4.990973169790315e-06,
+      "loss": 0.6944,
+      "step": 166
+    },
+    {
+      "epoch": 0.4083129584352078,
+      "grad_norm": 0.07484237489228208,
+      "learning_rate": 4.990864151705532e-06,
+      "loss": 0.5256,
+      "step": 167
+    },
+    {
+      "epoch": 0.41075794621026895,
+      "grad_norm": 0.10433853476942256,
+      "learning_rate": 4.990754480459149e-06,
+      "loss": 0.9094,
+      "step": 168
+    },
+    {
+      "epoch": 0.4132029339853301,
+      "grad_norm": 0.12247812230834339,
+      "learning_rate": 4.990644156079925e-06,
+      "loss": 0.7113,
+      "step": 169
+    },
+    {
+      "epoch": 0.4156479217603912,
+      "grad_norm": 0.05804892945521931,
+      "learning_rate": 4.990533178596788e-06,
+      "loss": 0.7857,
+      "step": 170
+    },
+    {
+      "epoch": 0.4180929095354523,
+      "grad_norm": 0.07237411770989431,
+      "learning_rate": 4.99042154803884e-06,
+      "loss": 0.961,
+      "step": 171
+    },
+    {
+      "epoch": 0.42053789731051344,
+      "grad_norm": 0.07228973210856648,
+      "learning_rate": 4.990309264435354e-06,
+      "loss": 0.6686,
+      "step": 172
+    },
+    {
+      "epoch": 0.4229828850855746,
+      "grad_norm": 0.06901133513710181,
+      "learning_rate": 4.9901963278157715e-06,
+      "loss": 0.816,
+      "step": 173
+    },
+    {
+      "epoch": 0.4254278728606357,
+      "grad_norm": 0.07174856426018103,
+      "learning_rate": 4.990082738209707e-06,
+      "loss": 0.595,
+      "step": 174
+    },
+    {
+      "epoch": 0.4278728606356968,
+      "grad_norm": 0.06384119191550144,
+      "learning_rate": 4.989968495646949e-06,
+      "loss": 0.8963,
+      "step": 175
+    },
+    {
+      "epoch": 0.43031784841075793,
+      "grad_norm": 0.060398178651819484,
+      "learning_rate": 4.989853600157451e-06,
+      "loss": 0.5991,
+      "step": 176
+    },
+    {
+      "epoch": 0.43276283618581907,
+      "grad_norm": 0.0713760600484018,
+      "learning_rate": 4.989738051771344e-06,
+      "loss": 0.7191,
+      "step": 177
+    },
+    {
+      "epoch": 0.4352078239608802,
+      "grad_norm": 0.08843620230170797,
+      "learning_rate": 4.989621850518925e-06,
+      "loss": 0.6627,
+      "step": 178
+    },
+    {
+      "epoch": 0.43765281173594134,
+      "grad_norm": 0.051954527341017354,
+      "learning_rate": 4.989504996430667e-06,
+      "loss": 0.4925,
+      "step": 179
+    },
+    {
+      "epoch": 0.4400977995110024,
+      "grad_norm": 0.0708987237498778,
+      "learning_rate": 4.989387489537211e-06,
+      "loss": 0.706,
+      "step": 180
+    },
+    {
+      "epoch": 0.44254278728606355,
+      "grad_norm": 0.0801266850649354,
+      "learning_rate": 4.98926932986937e-06,
+      "loss": 0.6977,
+      "step": 181
+    },
+    {
+      "epoch": 0.4449877750611247,
+      "grad_norm": 0.10512480393573406,
+      "learning_rate": 4.989150517458127e-06,
+      "loss": 0.521,
+      "step": 182
+    },
+    {
+      "epoch": 0.4474327628361858,
+      "grad_norm": 0.06320368064017733,
+      "learning_rate": 4.9890310523346386e-06,
+      "loss": 0.7557,
+      "step": 183
+    },
+    {
+      "epoch": 0.44987775061124696,
+      "grad_norm": 0.056159744095997874,
+      "learning_rate": 4.988910934530231e-06,
+      "loss": 0.5743,
+      "step": 184
+    },
+    {
+      "epoch": 0.45232273838630804,
+      "grad_norm": 0.12744617399117472,
+      "learning_rate": 4.988790164076401e-06,
+      "loss": 0.6362,
+      "step": 185
+    },
+    {
+      "epoch": 0.4547677261613692,
+      "grad_norm": 0.08930527880819754,
+      "learning_rate": 4.98866874100482e-06,
+      "loss": 0.5671,
+      "step": 186
+    },
+    {
+      "epoch": 0.4572127139364303,
+      "grad_norm": 0.07149671636535629,
+      "learning_rate": 4.988546665347325e-06,
+      "loss": 0.8099,
+      "step": 187
+    },
+    {
+      "epoch": 0.45965770171149145,
+      "grad_norm": 0.07794984955470813,
+      "learning_rate": 4.988423937135927e-06,
+      "loss": 0.5913,
+      "step": 188
+    },
+    {
+      "epoch": 0.4621026894865526,
+      "grad_norm": 0.10918496095120744,
+      "learning_rate": 4.988300556402811e-06,
+      "loss": 1.0037,
+      "step": 189
+    },
+    {
+      "epoch": 0.46454767726161367,
+      "grad_norm": 0.11961317163600049,
+      "learning_rate": 4.988176523180328e-06,
+      "loss": 0.5171,
+      "step": 190
+    },
+    {
+      "epoch": 0.4669926650366748,
+      "grad_norm": 0.06838542971570936,
+      "learning_rate": 4.9880518375010024e-06,
+      "loss": 0.615,
+      "step": 191
+    },
+    {
+      "epoch": 0.46943765281173594,
+      "grad_norm": 0.07856151694766761,
+      "learning_rate": 4.98792649939753e-06,
+      "loss": 0.8231,
+      "step": 192
+    },
+    {
+      "epoch": 0.4718826405867971,
+      "grad_norm": 0.12075114188625571,
+      "learning_rate": 4.987800508902778e-06,
+      "loss": 0.7494,
+      "step": 193
+    },
+    {
+      "epoch": 0.4743276283618582,
+      "grad_norm": 0.06830223665328845,
+      "learning_rate": 4.987673866049784e-06,
+      "loss": 0.6304,
+      "step": 194
+    },
+    {
+      "epoch": 0.4767726161369193,
+      "grad_norm": 0.10675312491364994,
+      "learning_rate": 4.987546570871755e-06,
+      "loss": 0.7213,
+      "step": 195
+    },
+    {
+      "epoch": 0.4792176039119804,
+      "grad_norm": 0.08088271092943289,
+      "learning_rate": 4.9874186234020715e-06,
+      "loss": 0.501,
+      "step": 196
+    },
+    {
+      "epoch": 0.48166259168704156,
+      "grad_norm": 0.06901160602750281,
+      "learning_rate": 4.987290023674285e-06,
+      "loss": 0.6102,
+      "step": 197
+    },
+    {
+      "epoch": 0.4841075794621027,
+      "grad_norm": 0.08330753598800536,
+      "learning_rate": 4.987160771722117e-06,
+      "loss": 0.9907,
+      "step": 198
+    },
+    {
+      "epoch": 0.48655256723716384,
+      "grad_norm": 0.0642050181277398,
+      "learning_rate": 4.9870308675794604e-06,
+      "loss": 0.6304,
+      "step": 199
+    },
+    {
+      "epoch": 0.4889975550122249,
+      "grad_norm": 0.05503037907417415,
+      "learning_rate": 4.9869003112803785e-06,
+      "loss": 0.803,
+      "step": 200
+    },
+    {
+      "epoch": 0.49144254278728605,
+      "grad_norm": 0.05343579542611549,
+      "learning_rate": 4.986769102859106e-06,
+      "loss": 0.4103,
+      "step": 201
+    },
+    {
+      "epoch": 0.4938875305623472,
+      "grad_norm": 0.08640774665689151,
+      "learning_rate": 4.98663724235005e-06,
+      "loss": 0.555,
+      "step": 202
+    },
+    {
+      "epoch": 0.4963325183374083,
+      "grad_norm": 0.08602755799563155,
+      "learning_rate": 4.986504729787786e-06,
+      "loss": 0.7291,
+      "step": 203
+    },
+    {
+      "epoch": 0.49877750611246946,
+      "grad_norm": 0.10140175293983117,
+      "learning_rate": 4.986371565207062e-06,
+      "loss": 0.9228,
+      "step": 204
+    },
+    {
+      "epoch": 0.5012224938875306,
+      "grad_norm": 0.07644994488427348,
+      "learning_rate": 4.986237748642798e-06,
+      "loss": 0.7743,
+      "step": 205
+    },
+    {
+      "epoch": 0.5036674816625917,
+      "grad_norm": 0.07819041605586327,
+      "learning_rate": 4.986103280130083e-06,
+      "loss": 0.6905,
+      "step": 206
+    },
+    {
+      "epoch": 0.5061124694376528,
+      "grad_norm": 0.0791217176414588,
+      "learning_rate": 4.985968159704176e-06,
+      "loss": 0.7729,
+      "step": 207
+    },
+    {
+      "epoch": 0.508557457212714,
+      "grad_norm": 0.06863020871294506,
+      "learning_rate": 4.985832387400512e-06,
+      "loss": 0.5923,
+      "step": 208
+    },
+    {
+      "epoch": 0.511002444987775,
+      "grad_norm": 0.08929132493963589,
+      "learning_rate": 4.985695963254692e-06,
+      "loss": 0.6418,
+      "step": 209
+    },
+    {
+      "epoch": 0.5134474327628362,
+      "grad_norm": 0.09899762276269189,
+      "learning_rate": 4.985558887302489e-06,
+      "loss": 0.9342,
+      "step": 210
+    },
+    {
+      "epoch": 0.5158924205378973,
+      "grad_norm": 0.08193665403820839,
+      "learning_rate": 4.985421159579847e-06,
+      "loss": 1.0324,
+      "step": 211
+    },
+    {
+      "epoch": 0.5183374083129584,
+      "grad_norm": 0.09959839296825133,
+      "learning_rate": 4.9852827801228835e-06,
+      "loss": 0.9113,
+      "step": 212
+    },
+    {
+      "epoch": 0.5207823960880196,
+      "grad_norm": 0.1315051002524026,
+      "learning_rate": 4.985143748967883e-06,
+      "loss": 0.5565,
+      "step": 213
+    },
+    {
+      "epoch": 0.5232273838630807,
+      "grad_norm": 0.06915030902244963,
+      "learning_rate": 4.985004066151304e-06,
+      "loss": 0.4822,
+      "step": 214
+    },
+    {
+      "epoch": 0.5256723716381418,
+      "grad_norm": 0.10091503632779024,
+      "learning_rate": 4.984863731709772e-06,
+      "loss": 0.5046,
+      "step": 215
+    },
+    {
+      "epoch": 0.5281173594132029,
+      "grad_norm": 0.06507643097510693,
+      "learning_rate": 4.984722745680089e-06,
+      "loss": 0.9793,
+      "step": 216
+    },
+    {
+      "epoch": 0.530562347188264,
+      "grad_norm": 0.0879113975750124,
+      "learning_rate": 4.984581108099223e-06,
+      "loss": 0.6283,
+      "step": 217
+    },
+    {
+      "epoch": 0.5330073349633252,
+      "grad_norm": 0.09582350126030938,
+      "learning_rate": 4.984438819004315e-06,
+      "loss": 0.5287,
+      "step": 218
+    },
+    {
+      "epoch": 0.5354523227383863,
+      "grad_norm": 0.06701152242509115,
+      "learning_rate": 4.984295878432677e-06,
+      "loss": 0.7057,
+      "step": 219
+    },
+    {
+      "epoch": 0.5378973105134475,
+      "grad_norm": 0.0718490412776238,
+      "learning_rate": 4.98415228642179e-06,
+      "loss": 0.6648,
+      "step": 220
+    },
+    {
+      "epoch": 0.5403422982885085,
+      "grad_norm": 0.09307015191606556,
+      "learning_rate": 4.984008043009308e-06,
+      "loss": 0.7452,
+      "step": 221
+    },
+    {
+      "epoch": 0.5427872860635696,
+      "grad_norm": 0.10303736607634002,
+      "learning_rate": 4.983863148233055e-06,
+      "loss": 0.7592,
+      "step": 222
+    },
+    {
+      "epoch": 0.5452322738386308,
+      "grad_norm": 0.08702846146876625,
+      "learning_rate": 4.983717602131025e-06,
+      "loss": 0.5841,
+      "step": 223
+    },
+    {
+      "epoch": 0.5476772616136919,
+      "grad_norm": 0.05850403452669591,
+      "learning_rate": 4.983571404741384e-06,
+      "loss": 0.5318,
+      "step": 224
+    },
+    {
+      "epoch": 0.5501222493887531,
+      "grad_norm": 0.08068398717142776,
+      "learning_rate": 4.983424556102469e-06,
+      "loss": 0.717,
+      "step": 225
+    },
+    {
+      "epoch": 0.5525672371638142,
+      "grad_norm": 0.07123259556915239,
+      "learning_rate": 4.983277056252787e-06,
+      "loss": 0.8039,
+      "step": 226
+    },
+    {
+      "epoch": 0.5550122249388753,
+      "grad_norm": 0.09188810320332673,
+      "learning_rate": 4.983128905231014e-06,
+      "loss": 0.8955,
+      "step": 227
+    },
+    {
+      "epoch": 0.5574572127139364,
+      "grad_norm": 0.07481109001543708,
+      "learning_rate": 4.982980103076e-06,
+      "loss": 0.6147,
+      "step": 228
+    },
+    {
+      "epoch": 0.5599022004889975,
+      "grad_norm": 0.12774035377023493,
+      "learning_rate": 4.982830649826764e-06,
+      "loss": 0.5272,
+      "step": 229
+    },
+    {
+      "epoch": 0.5623471882640587,
+      "grad_norm": 0.08134594677844227,
+      "learning_rate": 4.982680545522497e-06,
+      "loss": 0.5825,
+      "step": 230
+    },
+    {
+      "epoch": 0.5647921760391198,
+      "grad_norm": 0.06308804044499798,
+      "learning_rate": 4.982529790202559e-06,
+      "loss": 0.7986,
+      "step": 231
+    },
+    {
+      "epoch": 0.5672371638141809,
+      "grad_norm": 0.07309317085146093,
+      "learning_rate": 4.982378383906481e-06,
+      "loss": 0.8898,
+      "step": 232
+    },
+    {
+      "epoch": 0.5696821515892421,
+      "grad_norm": 0.06734889017367603,
+      "learning_rate": 4.982226326673966e-06,
+      "loss": 0.6332,
+      "step": 233
+    },
+    {
+      "epoch": 0.5721271393643031,
+      "grad_norm": 0.06932860707953248,
+      "learning_rate": 4.982073618544886e-06,
+      "loss": 1.1766,
+      "step": 234
+    },
+    {
+      "epoch": 0.5745721271393643,
+      "grad_norm": 0.06996838757752463,
+      "learning_rate": 4.981920259559285e-06,
+      "loss": 0.7283,
+      "step": 235
+    },
+    {
+      "epoch": 0.5770171149144254,
+      "grad_norm": 0.05357288768842464,
+      "learning_rate": 4.981766249757378e-06,
+      "loss": 0.5487,
+      "step": 236
+    },
+    {
+      "epoch": 0.5794621026894865,
+      "grad_norm": 0.08579442726514055,
+      "learning_rate": 4.981611589179549e-06,
+      "loss": 0.9645,
+      "step": 237
+    },
+    {
+      "epoch": 0.5819070904645477,
+      "grad_norm": 0.08745993239648804,
+      "learning_rate": 4.981456277866355e-06,
+      "loss": 0.9741,
+      "step": 238
+    },
+    {
+      "epoch": 0.5843520782396088,
+      "grad_norm": 0.07579885675925718,
+      "learning_rate": 4.98130031585852e-06,
+      "loss": 0.5609,
+      "step": 239
+    },
+    {
+      "epoch": 0.58679706601467,
+      "grad_norm": 0.09232934876594046,
+      "learning_rate": 4.981143703196941e-06,
+      "loss": 0.5401,
+      "step": 240
+    },
+    {
+      "epoch": 0.589242053789731,
+      "grad_norm": 0.08730183070991888,
+      "learning_rate": 4.980986439922688e-06,
+      "loss": 0.7162,
+      "step": 241
+    },
+    {
+      "epoch": 0.5916870415647921,
+      "grad_norm": 0.0874818677863648,
+      "learning_rate": 4.980828526076996e-06,
+      "loss": 0.6562,
+      "step": 242
+    },
+    {
+      "epoch": 0.5941320293398533,
+      "grad_norm": 0.06531945462857051,
+      "learning_rate": 4.980669961701276e-06,
+      "loss": 0.6015,
+      "step": 243
+    },
+    {
+      "epoch": 0.5965770171149144,
+      "grad_norm": 0.11538804730109553,
+      "learning_rate": 4.9805107468371055e-06,
+      "loss": 0.6362,
+      "step": 244
+    },
+    {
+      "epoch": 0.5990220048899756,
+      "grad_norm": 0.08947315493090728,
+      "learning_rate": 4.980350881526235e-06,
+      "loss": 0.8897,
+      "step": 245
+    },
+    {
+      "epoch": 0.6014669926650367,
+      "grad_norm": 0.07891063165167597,
+      "learning_rate": 4.980190365810584e-06,
+      "loss": 0.7953,
+      "step": 246
+    },
+    {
+      "epoch": 0.6039119804400978,
+      "grad_norm": 0.09066564849285777,
+      "learning_rate": 4.980029199732245e-06,
+      "loss": 1.0367,
+      "step": 247
+    },
+    {
+      "epoch": 0.6063569682151589,
+      "grad_norm": 0.09298654353803167,
+      "learning_rate": 4.979867383333479e-06,
+      "loss": 0.6467,
+      "step": 248
+    },
+    {
+      "epoch": 0.60880195599022,
+      "grad_norm": 0.07442509371951192,
+      "learning_rate": 4.979704916656717e-06,
+      "loss": 0.6237,
+      "step": 249
+    },
+    {
+      "epoch": 0.6112469437652812,
+      "grad_norm": 0.08031377511460987,
+      "learning_rate": 4.9795417997445615e-06,
+      "loss": 1.4601,
+      "step": 250
+    },
+    {
+      "epoch": 0.6136919315403423,
+      "grad_norm": 0.07068617682262945,
+      "learning_rate": 4.979378032639787e-06,
+      "loss": 0.8154,
+      "step": 251
+    },
+    {
+      "epoch": 0.6161369193154034,
+      "grad_norm": 0.08803651290720371,
+      "learning_rate": 4.979213615385334e-06,
+      "loss": 0.7019,
+      "step": 252
+    },
+    {
+      "epoch": 0.6185819070904646,
+      "grad_norm": 0.09405352195827332,
+      "learning_rate": 4.9790485480243205e-06,
+      "loss": 0.6672,
+      "step": 253
+    },
+    {
+      "epoch": 0.6210268948655256,
+      "grad_norm": 0.07611275328318817,
+      "learning_rate": 4.9788828306000285e-06,
+      "loss": 0.7606,
+      "step": 254
+    },
+    {
+      "epoch": 0.6234718826405868,
+      "grad_norm": 0.054319948235132615,
+      "learning_rate": 4.978716463155913e-06,
+      "loss": 0.431,
+      "step": 255
+    },
+    {
+      "epoch": 0.6259168704156479,
+      "grad_norm": 0.10166914464066547,
+      "learning_rate": 4.9785494457356e-06,
+      "loss": 0.9267,
+      "step": 256
+    },
+    {
+      "epoch": 0.628361858190709,
+      "grad_norm": 0.07316616910070232,
+      "learning_rate": 4.9783817783828845e-06,
+      "loss": 1.1379,
+      "step": 257
+    },
+    {
+      "epoch": 0.6308068459657702,
+      "grad_norm": 0.07284557902814078,
+      "learning_rate": 4.978213461141733e-06,
+      "loss": 0.8238,
+      "step": 258
+    },
+    {
+      "epoch": 0.6332518337408313,
+      "grad_norm": 0.07729127701923229,
+      "learning_rate": 4.978044494056282e-06,
+      "loss": 0.6282,
+      "step": 259
+    },
+    {
+      "epoch": 0.6356968215158925,
+      "grad_norm": 0.10965476253473438,
+      "learning_rate": 4.97787487717084e-06,
+      "loss": 0.7746,
+      "step": 260
+    },
+    {
+      "epoch": 0.6381418092909535,
+      "grad_norm": 0.08328320785681154,
+      "learning_rate": 4.977704610529883e-06,
+      "loss": 0.7129,
+      "step": 261
+    },
+    {
+      "epoch": 0.6405867970660146,
+      "grad_norm": 0.07096269237068077,
+      "learning_rate": 4.977533694178058e-06,
+      "loss": 0.8062,
+      "step": 262
+    },
+    {
+      "epoch": 0.6430317848410758,
+      "grad_norm": 0.11117598057147879,
+      "learning_rate": 4.977362128160186e-06,
+      "loss": 0.6799,
+      "step": 263
+    },
+    {
+      "epoch": 0.6454767726161369,
+      "grad_norm": 0.09735342376530616,
+      "learning_rate": 4.9771899125212524e-06,
+      "loss": 0.8234,
+      "step": 264
+    },
+    {
+      "epoch": 0.6479217603911981,
+      "grad_norm": 0.09144966869987771,
+      "learning_rate": 4.977017047306418e-06,
+      "loss": 0.7613,
+      "step": 265
+    },
+    {
+      "epoch": 0.6503667481662592,
+      "grad_norm": 0.09346114112765455,
+      "learning_rate": 4.9768435325610106e-06,
+      "loss": 0.7318,
+      "step": 266
+    },
+    {
+      "epoch": 0.6528117359413202,
+      "grad_norm": 0.06779016560747965,
+      "learning_rate": 4.9766693683305326e-06,
+      "loss": 0.6908,
+      "step": 267
+    },
+    {
+      "epoch": 0.6552567237163814,
+      "grad_norm": 0.13383732323227207,
+      "learning_rate": 4.97649455466065e-06,
+      "loss": 1.3706,
+      "step": 268
+    },
+    {
+      "epoch": 0.6577017114914425,
+      "grad_norm": 0.0788712267195195,
+      "learning_rate": 4.976319091597206e-06,
+      "loss": 0.6005,
+      "step": 269
+    },
+    {
+      "epoch": 0.6601466992665037,
+      "grad_norm": 0.06570350427696932,
+      "learning_rate": 4.976142979186209e-06,
+      "loss": 0.8911,
+      "step": 270
+    },
+    {
+      "epoch": 0.6625916870415648,
+      "grad_norm": 0.08063591302532015,
+      "learning_rate": 4.975966217473841e-06,
+      "loss": 0.8505,
+      "step": 271
+    },
+    {
+      "epoch": 0.6650366748166259,
+      "grad_norm": 0.0826057148632207,
+      "learning_rate": 4.975788806506452e-06,
+      "loss": 0.7124,
+      "step": 272
+    },
+    {
+      "epoch": 0.6674816625916871,
+      "grad_norm": 0.07363801056784766,
+      "learning_rate": 4.975610746330565e-06,
+      "loss": 0.9228,
+      "step": 273
+    },
+    {
+      "epoch": 0.6699266503667481,
+      "grad_norm": 0.08839135450743345,
+      "learning_rate": 4.975432036992869e-06,
+      "loss": 0.8497,
+      "step": 274
+    },
+    {
+      "epoch": 0.6723716381418093,
+      "grad_norm": 0.06362767692976444,
+      "learning_rate": 4.9752526785402276e-06,
+      "loss": 0.5299,
+      "step": 275
+    },
+    {
+      "epoch": 0.6748166259168704,
+      "grad_norm": 0.09506102741641534,
+      "learning_rate": 4.975072671019671e-06,
+      "loss": 0.6699,
+      "step": 276
+    },
+    {
+      "epoch": 0.6772616136919315,
+      "grad_norm": 0.0732489846524302,
+      "learning_rate": 4.974892014478403e-06,
+      "loss": 0.5213,
+      "step": 277
+    },
+    {
+      "epoch": 0.6797066014669927,
+      "grad_norm": 0.07792011711234956,
+      "learning_rate": 4.974710708963795e-06,
+      "loss": 0.7079,
+      "step": 278
+    },
+    {
+      "epoch": 0.6821515892420538,
+      "grad_norm": 0.06443317774634497,
+      "learning_rate": 4.9745287545233904e-06,
+      "loss": 0.784,
+      "step": 279
+    },
+    {
+      "epoch": 0.684596577017115,
+      "grad_norm": 0.08024122985918529,
+      "learning_rate": 4.9743461512049e-06,
+      "loss": 0.6112,
+      "step": 280
+    },
+    {
+      "epoch": 0.687041564792176,
+      "grad_norm": 0.09010298542382467,
+      "learning_rate": 4.974162899056207e-06,
+      "loss": 1.0372,
+      "step": 281
+    },
+    {
+      "epoch": 0.6894865525672371,
+      "grad_norm": 0.07100616876826107,
+      "learning_rate": 4.973978998125366e-06,
+      "loss": 1.023,
+      "step": 282
+    },
+    {
+      "epoch": 0.6919315403422983,
+      "grad_norm": 0.09164189430137386,
+      "learning_rate": 4.9737944484605995e-06,
+      "loss": 1.0753,
+      "step": 283
+    },
+    {
+      "epoch": 0.6943765281173594,
+      "grad_norm": 0.0718132356345249,
+      "learning_rate": 4.9736092501102985e-06,
+      "loss": 0.6648,
+      "step": 284
+    },
+    {
+      "epoch": 0.6968215158924206,
+      "grad_norm": 0.07884549975722853,
+      "learning_rate": 4.973423403123029e-06,
+      "loss": 0.7576,
+      "step": 285
+    },
+    {
+      "epoch": 0.6992665036674817,
+      "grad_norm": 0.10110247892652786,
+      "learning_rate": 4.973236907547523e-06,
+      "loss": 0.7716,
+      "step": 286
+    },
+    {
+      "epoch": 0.7017114914425427,
+      "grad_norm": 0.07723692555839858,
+      "learning_rate": 4.973049763432684e-06,
+      "loss": 0.7622,
+      "step": 287
+    },
+    {
+      "epoch": 0.7041564792176039,
+      "grad_norm": 0.07134818457945936,
+      "learning_rate": 4.972861970827585e-06,
+      "loss": 0.5653,
+      "step": 288
+    },
+    {
+      "epoch": 0.706601466992665,
+      "grad_norm": 0.07619769017028144,
+      "learning_rate": 4.97267352978147e-06,
+      "loss": 0.9603,
+      "step": 289
+    },
+    {
+      "epoch": 0.7090464547677262,
+      "grad_norm": 0.0839175935654035,
+      "learning_rate": 4.972484440343754e-06,
+      "loss": 0.8661,
+      "step": 290
+    },
+    {
+      "epoch": 0.7114914425427873,
+      "grad_norm": 0.08203933567175961,
+      "learning_rate": 4.9722947025640175e-06,
+      "loss": 0.6861,
+      "step": 291
+    },
+    {
+      "epoch": 0.7139364303178484,
+      "grad_norm": 0.08518667181270598,
+      "learning_rate": 4.972104316492016e-06,
+      "loss": 1.0165,
+      "step": 292
+    },
+    {
+      "epoch": 0.7163814180929096,
+      "grad_norm": 0.10970590133781132,
+      "learning_rate": 4.971913282177674e-06,
+      "loss": 0.6495,
+      "step": 293
+    },
+    {
+      "epoch": 0.7188264058679706,
+      "grad_norm": 0.06977398950464615,
+      "learning_rate": 4.971721599671083e-06,
+      "loss": 0.4463,
+      "step": 294
+    },
+    {
+      "epoch": 0.7212713936430318,
+      "grad_norm": 0.0953089382370969,
+      "learning_rate": 4.971529269022508e-06,
+      "loss": 1.1599,
+      "step": 295
+    },
+    {
+      "epoch": 0.7237163814180929,
+      "grad_norm": 0.07231474734901186,
+      "learning_rate": 4.971336290282382e-06,
+      "loss": 0.5773,
+      "step": 296
+    },
+    {
+      "epoch": 0.726161369193154,
+      "grad_norm": 0.10325309775054034,
+      "learning_rate": 4.971142663501308e-06,
+      "loss": 0.5606,
+      "step": 297
+    },
+    {
+      "epoch": 0.7286063569682152,
+      "grad_norm": 0.06493355868743536,
+      "learning_rate": 4.970948388730059e-06,
+      "loss": 0.5219,
+      "step": 298
+    },
+    {
+      "epoch": 0.7310513447432763,
+      "grad_norm": 0.05275171607060085,
+      "learning_rate": 4.970753466019581e-06,
+      "loss": 0.4623,
+      "step": 299
+    },
+    {
+      "epoch": 0.7334963325183375,
+      "grad_norm": 0.062003179052265364,
+      "learning_rate": 4.970557895420984e-06,
+      "loss": 0.8119,
+      "step": 300
+    },
+    {
+      "epoch": 0.7359413202933985,
+      "grad_norm": 0.08765537692196579,
+      "learning_rate": 4.970361676985553e-06,
+      "loss": 1.2166,
+      "step": 301
+    },
+    {
+      "epoch": 0.7383863080684596,
+      "grad_norm": 0.1057985963435635,
+      "learning_rate": 4.97016481076474e-06,
+      "loss": 0.6923,
+      "step": 302
+    },
+    {
+      "epoch": 0.7408312958435208,
+      "grad_norm": 0.0715389858662257,
+      "learning_rate": 4.969967296810168e-06,
+      "loss": 0.535,
+      "step": 303
+    },
+    {
+      "epoch": 0.7432762836185819,
+      "grad_norm": 0.09288946983927428,
+      "learning_rate": 4.96976913517363e-06,
+      "loss": 0.6887,
+      "step": 304
+    },
+    {
+      "epoch": 0.7457212713936431,
+      "grad_norm": 0.06614264379965222,
+      "learning_rate": 4.969570325907088e-06,
+      "loss": 0.5541,
+      "step": 305
+    },
+    {
+      "epoch": 0.7481662591687042,
+      "grad_norm": 0.07789423963719386,
+      "learning_rate": 4.969370869062676e-06,
+      "loss": 0.905,
+      "step": 306
+    },
+    {
+      "epoch": 0.7506112469437652,
+      "grad_norm": 0.09710120230405622,
+      "learning_rate": 4.969170764692693e-06,
+      "loss": 0.573,
+      "step": 307
+    },
+    {
+      "epoch": 0.7530562347188264,
+      "grad_norm": 0.0645825950181619,
+      "learning_rate": 4.968970012849614e-06,
+      "loss": 0.7413,
+      "step": 308
+    },
+    {
+      "epoch": 0.7555012224938875,
+      "grad_norm": 0.08158264825180002,
+      "learning_rate": 4.96876861358608e-06,
+      "loss": 0.9351,
+      "step": 309
+    },
+    {
+      "epoch": 0.7579462102689487,
+      "grad_norm": 0.0642139113205452,
+      "learning_rate": 4.968566566954901e-06,
+      "loss": 0.5706,
+      "step": 310
+    },
+    {
+      "epoch": 0.7603911980440098,
+      "grad_norm": 0.08775438735010123,
+      "learning_rate": 4.96836387300906e-06,
+      "loss": 0.6897,
+      "step": 311
+    },
+    {
+      "epoch": 0.7628361858190709,
+      "grad_norm": 0.06561298612751973,
+      "learning_rate": 4.968160531801707e-06,
+      "loss": 0.6126,
+      "step": 312
+    },
+    {
+      "epoch": 0.7652811735941321,
+      "grad_norm": 0.15019058027061738,
+      "learning_rate": 4.967956543386163e-06,
+      "loss": 1.1087,
+      "step": 313
+    },
+    {
+      "epoch": 0.7677261613691931,
+      "grad_norm": 0.10563230650201169,
+      "learning_rate": 4.967751907815918e-06,
+      "loss": 0.6726,
+      "step": 314
+    },
+    {
+      "epoch": 0.7701711491442543,
+      "grad_norm": 0.13651356976880127,
+      "learning_rate": 4.967546625144633e-06,
+      "loss": 0.7918,
+      "step": 315
+    },
+    {
+      "epoch": 0.7726161369193154,
+      "grad_norm": 0.08055315644947807,
+      "learning_rate": 4.967340695426137e-06,
+      "loss": 0.7152,
+      "step": 316
+    },
+    {
+      "epoch": 0.7750611246943765,
+      "grad_norm": 0.10646718030189253,
+      "learning_rate": 4.9671341187144305e-06,
+      "loss": 1.0296,
+      "step": 317
+    },
+    {
+      "epoch": 0.7775061124694377,
+      "grad_norm": 0.08035804242125767,
+      "learning_rate": 4.966926895063682e-06,
+      "loss": 0.4675,
+      "step": 318
+    },
+    {
+      "epoch": 0.7799511002444988,
+      "grad_norm": 0.09080078591258836,
+      "learning_rate": 4.966719024528229e-06,
+      "loss": 0.5322,
+      "step": 319
+    },
+    {
+      "epoch": 0.78239608801956,
+      "grad_norm": 0.07755784595559201,
+      "learning_rate": 4.966510507162583e-06,
+      "loss": 0.5181,
+      "step": 320
+    },
+    {
+      "epoch": 0.784841075794621,
+      "grad_norm": 0.08574576983190237,
+      "learning_rate": 4.9663013430214194e-06,
+      "loss": 0.8412,
+      "step": 321
+    },
+    {
+      "epoch": 0.7872860635696821,
+      "grad_norm": 0.07292180204112599,
+      "learning_rate": 4.966091532159586e-06,
+      "loss": 0.966,
+      "step": 322
+    },
+    {
+      "epoch": 0.7897310513447433,
+      "grad_norm": 0.10982141476577932,
+      "learning_rate": 4.965881074632102e-06,
+      "loss": 1.0051,
+      "step": 323
+    },
+    {
+      "epoch": 0.7921760391198044,
+      "grad_norm": 0.09431756479791206,
+      "learning_rate": 4.965669970494151e-06,
+      "loss": 0.8688,
+      "step": 324
+    },
+    {
+      "epoch": 0.7946210268948656,
+      "grad_norm": 0.06966225643153073,
+      "learning_rate": 4.965458219801093e-06,
+      "loss": 0.543,
+      "step": 325
+    },
+    {
+      "epoch": 0.7970660146699267,
+      "grad_norm": 0.09866879177917375,
+      "learning_rate": 4.965245822608451e-06,
+      "loss": 1.0296,
+      "step": 326
+    },
+    {
+      "epoch": 0.7995110024449877,
+      "grad_norm": 0.07347140347158336,
+      "learning_rate": 4.965032778971922e-06,
+      "loss": 0.5063,
+      "step": 327
+    },
+    {
+      "epoch": 0.8019559902200489,
+      "grad_norm": 0.19054475281388922,
+      "learning_rate": 4.96481908894737e-06,
+      "loss": 0.6894,
+      "step": 328
+    },
+    {
+      "epoch": 0.80440097799511,
+      "grad_norm": 0.07284733096473339,
+      "learning_rate": 4.96460475259083e-06,
+      "loss": 0.5695,
+      "step": 329
+    },
+    {
+      "epoch": 0.8068459657701712,
+      "grad_norm": 0.09008535141694302,
+      "learning_rate": 4.964389769958506e-06,
+      "loss": 1.3029,
+      "step": 330
+    },
+    {
+      "epoch": 0.8092909535452323,
+      "grad_norm": 0.089770829238767,
+      "learning_rate": 4.964174141106771e-06,
+      "loss": 0.8726,
+      "step": 331
+    },
+    {
+      "epoch": 0.8117359413202934,
+      "grad_norm": 0.06873500003410662,
+      "learning_rate": 4.963957866092167e-06,
+      "loss": 0.6982,
+      "step": 332
+    },
+    {
+      "epoch": 0.8141809290953546,
+      "grad_norm": 0.07904686905540631,
+      "learning_rate": 4.963740944971408e-06,
+      "loss": 0.7142,
+      "step": 333
+    },
+    {
+      "epoch": 0.8166259168704156,
+      "grad_norm": 0.08736362318728487,
+      "learning_rate": 4.963523377801375e-06,
+      "loss": 0.6935,
+      "step": 334
+    },
+    {
+      "epoch": 0.8190709046454768,
+      "grad_norm": 0.0685205181370088,
+      "learning_rate": 4.963305164639119e-06,
+      "loss": 0.8552,
+      "step": 335
+    },
+    {
+      "epoch": 0.8215158924205379,
+      "grad_norm": 0.07608190464825812,
+      "learning_rate": 4.9630863055418586e-06,
+      "loss": 1.0129,
+      "step": 336
+    },
+    {
+      "epoch": 0.823960880195599,
+      "grad_norm": 0.08989916733561043,
+      "learning_rate": 4.962866800566987e-06,
+      "loss": 0.6522,
+      "step": 337
+    },
+    {
+      "epoch": 0.8264058679706602,
+      "grad_norm": 0.08925436348801553,
+      "learning_rate": 4.962646649772062e-06,
+      "loss": 0.6081,
+      "step": 338
+    },
+    {
+      "epoch": 0.8288508557457213,
+      "grad_norm": 0.08726652834221434,
+      "learning_rate": 4.9624258532148116e-06,
+      "loss": 0.4843,
+      "step": 339
+    },
+    {
+      "epoch": 0.8312958435207825,
+      "grad_norm": 0.0719249289345701,
+      "learning_rate": 4.962204410953135e-06,
+      "loss": 0.8632,
+      "step": 340
+    },
+    {
+      "epoch": 0.8337408312958435,
+      "grad_norm": 0.11864905600234564,
+      "learning_rate": 4.961982323045097e-06,
+      "loss": 0.478,
+      "step": 341
+    },
+    {
+      "epoch": 0.8361858190709046,
+      "grad_norm": 0.09827018736440911,
+      "learning_rate": 4.961759589548937e-06,
+      "loss": 0.6793,
+      "step": 342
+    },
+    {
+      "epoch": 0.8386308068459658,
+      "grad_norm": 0.09566327812901528,
+      "learning_rate": 4.9615362105230595e-06,
+      "loss": 0.6049,
+      "step": 343
+    },
+    {
+      "epoch": 0.8410757946210269,
+      "grad_norm": 0.07070144260042295,
+      "learning_rate": 4.96131218602604e-06,
+      "loss": 0.6632,
+      "step": 344
+    },
+    {
+      "epoch": 0.843520782396088,
+      "grad_norm": 0.0830820616569169,
+      "learning_rate": 4.961087516116622e-06,
+      "loss": 0.5394,
+      "step": 345
+    },
+    {
+      "epoch": 0.8459657701711492,
+      "grad_norm": 0.11057233351152325,
+      "learning_rate": 4.96086220085372e-06,
+      "loss": 0.5714,
+      "step": 346
+    },
+    {
+      "epoch": 0.8484107579462102,
+      "grad_norm": 0.06940174870366157,
+      "learning_rate": 4.9606362402964155e-06,
+      "loss": 0.5592,
+      "step": 347
+    },
+    {
+      "epoch": 0.8508557457212714,
+      "grad_norm": 0.1000557755141698,
+      "learning_rate": 4.960409634503962e-06,
+      "loss": 0.8992,
+      "step": 348
+    },
+    {
+      "epoch": 0.8533007334963325,
+      "grad_norm": 0.059909846258367434,
+      "learning_rate": 4.96018238353578e-06,
+      "loss": 0.7048,
+      "step": 349
+    },
+    {
+      "epoch": 0.8557457212713936,
+      "grad_norm": 0.09166004278216504,
+      "learning_rate": 4.959954487451461e-06,
+      "loss": 0.762,
+      "step": 350
+    },
+    {
+      "epoch": 0.8581907090464548,
+      "grad_norm": 0.07974581790684761,
+      "learning_rate": 4.959725946310763e-06,
+      "loss": 0.8647,
+      "step": 351
+    },
+    {
+      "epoch": 0.8606356968215159,
+      "grad_norm": 0.07236511919690002,
+      "learning_rate": 4.959496760173615e-06,
+      "loss": 0.8792,
+      "step": 352
+    },
+    {
+      "epoch": 0.863080684596577,
+      "grad_norm": 0.09899137329821,
+      "learning_rate": 4.959266929100116e-06,
+      "loss": 1.0709,
+      "step": 353
+    },
+    {
+      "epoch": 0.8655256723716381,
+      "grad_norm": 0.11413955816663791,
+      "learning_rate": 4.959036453150532e-06,
+      "loss": 0.5481,
+      "step": 354
+    },
+    {
+      "epoch": 0.8679706601466992,
+      "grad_norm": 0.07755742570827054,
+      "learning_rate": 4.958805332385299e-06,
+      "loss": 0.634,
+      "step": 355
+    },
+    {
+      "epoch": 0.8704156479217604,
+      "grad_norm": 0.07744569811774757,
+      "learning_rate": 4.9585735668650225e-06,
+      "loss": 0.682,
+      "step": 356
+    },
+    {
+      "epoch": 0.8728606356968215,
+      "grad_norm": 0.06647153060123005,
+      "learning_rate": 4.958341156650477e-06,
+      "loss": 0.4184,
+      "step": 357
+    },
+    {
+      "epoch": 0.8753056234718827,
+      "grad_norm": 0.06996065107298906,
+      "learning_rate": 4.958108101802605e-06,
+      "loss": 0.5719,
+      "step": 358
+    },
+    {
+      "epoch": 0.8777506112469438,
+      "grad_norm": 0.06901613460968989,
+      "learning_rate": 4.95787440238252e-06,
+      "loss": 0.601,
+      "step": 359
+    },
+    {
+      "epoch": 0.8801955990220048,
+      "grad_norm": 0.10908189815492843,
+      "learning_rate": 4.957640058451502e-06,
+      "loss": 0.6224,
+      "step": 360
+    },
+    {
+      "epoch": 0.882640586797066,
+      "grad_norm": 0.07519971673926353,
+      "learning_rate": 4.957405070071001e-06,
+      "loss": 0.4477,
+      "step": 361
+    },
+    {
+      "epoch": 0.8850855745721271,
+      "grad_norm": 0.0941555591491303,
+      "learning_rate": 4.957169437302639e-06,
+      "loss": 0.7558,
+      "step": 362
+    },
+    {
+      "epoch": 0.8875305623471883,
+      "grad_norm": 0.07441197948763678,
+      "learning_rate": 4.9569331602082015e-06,
+      "loss": 0.6555,
+      "step": 363
+    },
+    {
+      "epoch": 0.8899755501222494,
+      "grad_norm": 0.10066309657301664,
+      "learning_rate": 4.956696238849647e-06,
+      "loss": 0.8644,
+      "step": 364
+    },
+    {
+      "epoch": 0.8924205378973105,
+      "grad_norm": 0.060322858349314744,
+      "learning_rate": 4.956458673289102e-06,
+      "loss": 0.6688,
+      "step": 365
+    },
+    {
+      "epoch": 0.8948655256723717,
+      "grad_norm": 0.07459723173599125,
+      "learning_rate": 4.95622046358886e-06,
+      "loss": 0.8178,
+      "step": 366
+    },
+    {
+      "epoch": 0.8973105134474327,
+      "grad_norm": 0.07249063362436482,
+      "learning_rate": 4.9559816098113874e-06,
+      "loss": 0.7303,
+      "step": 367
+    },
+    {
+      "epoch": 0.8997555012224939,
+      "grad_norm": 0.08828997974047631,
+      "learning_rate": 4.955742112019316e-06,
+      "loss": 0.5852,
+      "step": 368
+    },
+    {
+      "epoch": 0.902200488997555,
+      "grad_norm": 0.06782498962008572,
+      "learning_rate": 4.955501970275447e-06,
+      "loss": 0.7972,
+      "step": 369
+    },
+    {
+      "epoch": 0.9046454767726161,
+      "grad_norm": 0.09193553943494716,
+      "learning_rate": 4.955261184642752e-06,
+      "loss": 0.8673,
+      "step": 370
+    },
+    {
+      "epoch": 0.9070904645476773,
+      "grad_norm": 0.0815775893964367,
+      "learning_rate": 4.955019755184369e-06,
+      "loss": 0.4362,
+      "step": 371
+    },
+    {
+      "epoch": 0.9095354523227384,
+      "grad_norm": 0.0735163465256261,
+      "learning_rate": 4.954777681963609e-06,
+      "loss": 0.8682,
+      "step": 372
+    },
+    {
+      "epoch": 0.9119804400977995,
+      "grad_norm": 0.0607622140169798,
+      "learning_rate": 4.954534965043947e-06,
+      "loss": 0.6241,
+      "step": 373
+    },
+    {
+      "epoch": 0.9144254278728606,
+      "grad_norm": 0.09097140987050431,
+      "learning_rate": 4.954291604489029e-06,
+      "loss": 0.4739,
+      "step": 374
+    },
+    {
+      "epoch": 0.9168704156479217,
+      "grad_norm": 0.09009782142347368,
+      "learning_rate": 4.95404760036267e-06,
+      "loss": 0.5854,
+      "step": 375
+    },
+    {
+      "epoch": 0.9193154034229829,
+      "grad_norm": 0.10026963855738673,
+      "learning_rate": 4.9538029527288545e-06,
+      "loss": 0.5275,
+      "step": 376
+    },
+    {
+      "epoch": 0.921760391198044,
+      "grad_norm": 0.06555658920279356,
+      "learning_rate": 4.953557661651733e-06,
+      "loss": 0.503,
+      "step": 377
+    },
+    {
+      "epoch": 0.9242053789731052,
+      "grad_norm": 0.089032835475639,
+      "learning_rate": 4.9533117271956286e-06,
+      "loss": 1.0431,
+      "step": 378
+    },
+    {
+      "epoch": 0.9266503667481663,
+      "grad_norm": 0.08650107379464002,
+      "learning_rate": 4.953065149425029e-06,
+      "loss": 0.9569,
+      "step": 379
+    },
+    {
+      "epoch": 0.9290953545232273,
+      "grad_norm": 0.11733557999855201,
+      "learning_rate": 4.952817928404593e-06,
+      "loss": 0.5429,
+      "step": 380
+    },
+    {
+      "epoch": 0.9315403422982885,
+      "grad_norm": 0.09000604566338713,
+      "learning_rate": 4.952570064199148e-06,
+      "loss": 0.8854,
+      "step": 381
+    },
+    {
+      "epoch": 0.9339853300733496,
+      "grad_norm": 0.07898219178600274,
+      "learning_rate": 4.952321556873689e-06,
+      "loss": 0.4985,
+      "step": 382
+    },
+    {
+      "epoch": 0.9364303178484108,
+      "grad_norm": 0.08131984444295022,
+      "learning_rate": 4.952072406493382e-06,
+      "loss": 0.7474,
+      "step": 383
+    },
+    {
+      "epoch": 0.9388753056234719,
+      "grad_norm": 0.07659531144361392,
+      "learning_rate": 4.951822613123558e-06,
+      "loss": 0.5776,
+      "step": 384
+    },
+    {
+      "epoch": 0.941320293398533,
+      "grad_norm": 0.10311268616339668,
+      "learning_rate": 4.95157217682972e-06,
+      "loss": 0.644,
+      "step": 385
+    },
+    {
+      "epoch": 0.9437652811735942,
+      "grad_norm": 0.11823311556952831,
+      "learning_rate": 4.951321097677536e-06,
+      "loss": 0.7785,
+      "step": 386
+    },
+    {
+      "epoch": 0.9462102689486552,
+      "grad_norm": 0.09248650345575686,
+      "learning_rate": 4.9510693757328464e-06,
+      "loss": 0.9739,
+      "step": 387
+    },
+    {
+      "epoch": 0.9486552567237164,
+      "grad_norm": 0.05170086674421056,
+      "learning_rate": 4.950817011061659e-06,
+      "loss": 0.3465,
+      "step": 388
+    },
+    {
+      "epoch": 0.9511002444987775,
+      "grad_norm": 0.07553393219105838,
+      "learning_rate": 4.950564003730148e-06,
+      "loss": 0.9992,
+      "step": 389
+    },
+    {
+      "epoch": 0.9535452322738386,
+      "grad_norm": 0.1098893091484965,
+      "learning_rate": 4.950310353804659e-06,
+      "loss": 0.5216,
+      "step": 390
+    },
+    {
+      "epoch": 0.9559902200488998,
+      "grad_norm": 0.08676183287703625,
+      "learning_rate": 4.950056061351704e-06,
+      "loss": 0.7807,
+      "step": 391
+    },
+    {
+      "epoch": 0.9584352078239609,
+      "grad_norm": 0.0969873752285447,
+      "learning_rate": 4.949801126437965e-06,
+      "loss": 0.9221,
+      "step": 392
+    },
+    {
+      "epoch": 0.960880195599022,
+      "grad_norm": 0.0854835249192429,
+      "learning_rate": 4.94954554913029e-06,
+      "loss": 0.8477,
+      "step": 393
+    },
+    {
+      "epoch": 0.9633251833740831,
+      "grad_norm": 0.1207802459263373,
+      "learning_rate": 4.9492893294957e-06,
+      "loss": 0.61,
+      "step": 394
+    },
+    {
+      "epoch": 0.9657701711491442,
+      "grad_norm": 0.09380341945695975,
+      "learning_rate": 4.94903246760138e-06,
+      "loss": 0.4529,
+      "step": 395
+    },
+    {
+      "epoch": 0.9682151589242054,
+      "grad_norm": 0.09765648955425381,
+      "learning_rate": 4.948774963514685e-06,
+      "loss": 0.5131,
+      "step": 396
+    },
+    {
+      "epoch": 0.9706601466992665,
+      "grad_norm": 0.09417538561326143,
+      "learning_rate": 4.948516817303138e-06,
+      "loss": 0.5324,
+      "step": 397
+    },
+    {
+      "epoch": 0.9731051344743277,
+      "grad_norm": 0.0996378873170343,
+      "learning_rate": 4.9482580290344325e-06,
+      "loss": 0.5323,
+      "step": 398
+    },
+    {
+      "epoch": 0.9755501222493888,
+      "grad_norm": 0.0888330405479318,
+      "learning_rate": 4.947998598776429e-06,
+      "loss": 0.5733,
+      "step": 399
+    },
+    {
+      "epoch": 0.9779951100244498,
+      "grad_norm": 0.07794414942120746,
+      "learning_rate": 4.9477385265971534e-06,
+      "loss": 0.5127,
+      "step": 400
+    },
+    {
+      "epoch": 0.980440097799511,
+      "grad_norm": 0.5580974031036784,
+      "learning_rate": 4.947477812564804e-06,
+      "loss": 0.9767,
+      "step": 401
+    },
+    {
+      "epoch": 0.9828850855745721,
+      "grad_norm": 0.09663749476561667,
+      "learning_rate": 4.9472164567477465e-06,
+      "loss": 1.1788,
+      "step": 402
+    },
+    {
+      "epoch": 0.9853300733496333,
+      "grad_norm": 0.09105593405749489,
+      "learning_rate": 4.946954459214514e-06,
+      "loss": 0.917,
+      "step": 403
+    },
+    {
+      "epoch": 0.9877750611246944,
+      "grad_norm": 0.08597644599446,
+      "learning_rate": 4.946691820033807e-06,
+      "loss": 0.5118,
+      "step": 404
+    },
+    {
+      "epoch": 0.9902200488997555,
+      "grad_norm": 0.09511456844930492,
+      "learning_rate": 4.946428539274498e-06,
+      "loss": 0.5629,
+      "step": 405
+    },
+    {
+      "epoch": 0.9926650366748166,
+      "grad_norm": 0.0992683587423073,
+      "learning_rate": 4.9461646170056235e-06,
+      "loss": 1.0581,
+      "step": 406
+    },
+    {
+      "epoch": 0.9951100244498777,
+      "grad_norm": 0.08940455356676082,
+      "learning_rate": 4.94590005329639e-06,
+      "loss": 0.4939,
+      "step": 407
+    },
+    {
+      "epoch": 0.9975550122249389,
+      "grad_norm": 0.07104513302102258,
+      "learning_rate": 4.945634848216173e-06,
+      "loss": 0.6716,
+      "step": 408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.06934468396267954,
+      "learning_rate": 4.9453690018345144e-06,
+      "loss": 0.5411,
+      "step": 409
+    },
+    {
+      "epoch": 1.0024449877750612,
+      "grad_norm": 0.09105941539916199,
+      "learning_rate": 4.945102514221126e-06,
+      "loss": 0.7026,
+      "step": 410
+    },
+    {
+      "epoch": 1.0048899755501222,
+      "grad_norm": 0.08924714864807391,
+      "learning_rate": 4.944835385445886e-06,
+      "loss": 0.5374,
+      "step": 411
+    },
+    {
+      "epoch": 1.0073349633251834,
+      "grad_norm": 0.07192871393722046,
+      "learning_rate": 4.944567615578844e-06,
+      "loss": 0.656,
+      "step": 412
+    },
+    {
+      "epoch": 1.0097799511002445,
+      "grad_norm": 0.0868990642167748,
+      "learning_rate": 4.944299204690212e-06,
+      "loss": 0.5648,
+      "step": 413
+    },
+    {
+      "epoch": 1.0122249388753057,
+      "grad_norm": 0.07614894534555937,
+      "learning_rate": 4.944030152850377e-06,
+      "loss": 0.9752,
+      "step": 414
+    },
+    {
+      "epoch": 1.0146699266503667,
+      "grad_norm": 0.0771935688711757,
+      "learning_rate": 4.943760460129887e-06,
+      "loss": 0.5689,
+      "step": 415
+    },
+    {
+      "epoch": 1.017114914425428,
+      "grad_norm": 0.10450598696220409,
+      "learning_rate": 4.943490126599463e-06,
+      "loss": 0.8646,
+      "step": 416
+    },
+    {
+      "epoch": 1.019559902200489,
+      "grad_norm": 0.09160376010054183,
+      "learning_rate": 4.943219152329994e-06,
+      "loss": 0.9548,
+      "step": 417
+    },
+    {
+      "epoch": 1.02200488997555,
+      "grad_norm": 0.07862226233200668,
+      "learning_rate": 4.942947537392535e-06,
+      "loss": 0.4724,
+      "step": 418
+    },
+    {
+      "epoch": 1.0244498777506112,
+      "grad_norm": 0.08603212800037052,
+      "learning_rate": 4.942675281858308e-06,
+      "loss": 0.6607,
+      "step": 419
+    },
+    {
+      "epoch": 1.0268948655256724,
+      "grad_norm": 0.1251801147919071,
+      "learning_rate": 4.942402385798707e-06,
+      "loss": 0.9799,
+      "step": 420
+    },
+    {
+      "epoch": 1.0293398533007334,
+      "grad_norm": 0.104080517337458,
+      "learning_rate": 4.9421288492852905e-06,
+      "loss": 0.5324,
+      "step": 421
+    },
+    {
+      "epoch": 1.0317848410757946,
+      "grad_norm": 0.11161571103050913,
+      "learning_rate": 4.941854672389786e-06,
+      "loss": 0.6577,
+      "step": 422
+    },
+    {
+      "epoch": 1.0342298288508558,
+      "grad_norm": 0.12021884481536534,
+      "learning_rate": 4.941579855184089e-06,
+      "loss": 0.9195,
+      "step": 423
+    },
+    {
+      "epoch": 1.0366748166259168,
+      "grad_norm": 0.07288618577939406,
+      "learning_rate": 4.941304397740262e-06,
+      "loss": 0.5586,
+      "step": 424
+    },
+    {
+      "epoch": 1.039119804400978,
+      "grad_norm": 0.10488610513725002,
+      "learning_rate": 4.941028300130538e-06,
+      "loss": 1.3365,
+      "step": 425
+    },
+    {
+      "epoch": 1.0415647921760391,
+      "grad_norm": 0.11976183890406934,
+      "learning_rate": 4.940751562427315e-06,
+      "loss": 1.037,
+      "step": 426
+    },
+    {
+      "epoch": 1.0440097799511003,
+      "grad_norm": 0.11851888661317067,
+      "learning_rate": 4.94047418470316e-06,
+      "loss": 0.5093,
+      "step": 427
+    },
+    {
+      "epoch": 1.0464547677261613,
+      "grad_norm": 0.12888387709230967,
+      "learning_rate": 4.940196167030808e-06,
+      "loss": 0.6735,
+      "step": 428
+    },
+    {
+      "epoch": 1.0488997555012225,
+      "grad_norm": 0.07891732461627987,
+      "learning_rate": 4.939917509483162e-06,
+      "loss": 0.5831,
+      "step": 429
+    },
+    {
+      "epoch": 1.0513447432762837,
+      "grad_norm": 0.06648253960400075,
+      "learning_rate": 4.939638212133292e-06,
+      "loss": 0.5962,
+      "step": 430
+    },
+    {
+      "epoch": 1.0537897310513447,
+      "grad_norm": 0.10996548239770788,
+      "learning_rate": 4.9393582750544355e-06,
+      "loss": 0.8285,
+      "step": 431
+    },
+    {
+      "epoch": 1.0562347188264058,
+      "grad_norm": 0.08636236789345893,
+      "learning_rate": 4.93907769832e-06,
+      "loss": 0.5662,
+      "step": 432
+    },
+    {
+      "epoch": 1.058679706601467,
+      "grad_norm": 0.08195324240018571,
+      "learning_rate": 4.938796482003557e-06,
+      "loss": 0.6888,
+      "step": 433
+    },
+    {
+      "epoch": 1.061124694376528,
+      "grad_norm": 0.08104043991888836,
+      "learning_rate": 4.93851462617885e-06,
+      "loss": 0.7051,
+      "step": 434
+    },
+    {
+      "epoch": 1.0635696821515892,
+      "grad_norm": 0.11111075681798355,
+      "learning_rate": 4.938232130919786e-06,
+      "loss": 0.7417,
+      "step": 435
+    },
+    {
+      "epoch": 1.0660146699266504,
+      "grad_norm": 0.08823093697470814,
+      "learning_rate": 4.9379489963004435e-06,
+      "loss": 1.0322,
+      "step": 436
+    },
+    {
+      "epoch": 1.0684596577017116,
+      "grad_norm": 0.0990873186490748,
+      "learning_rate": 4.937665222395066e-06,
+      "loss": 0.4236,
+      "step": 437
+    },
+    {
+      "epoch": 1.0709046454767726,
+      "grad_norm": 0.07693311404671367,
+      "learning_rate": 4.937380809278066e-06,
+      "loss": 0.5643,
+      "step": 438
+    },
+    {
+      "epoch": 1.0733496332518337,
+      "grad_norm": 0.11993647913661866,
+      "learning_rate": 4.937095757024023e-06,
+      "loss": 0.9462,
+      "step": 439
+    },
+    {
+      "epoch": 1.075794621026895,
+      "grad_norm": 0.13528409042624007,
+      "learning_rate": 4.936810065707684e-06,
+      "loss": 0.8804,
+      "step": 440
+    },
+    {
+      "epoch": 1.078239608801956,
+      "grad_norm": 0.09500668501845402,
+      "learning_rate": 4.936523735403963e-06,
+      "loss": 0.6563,
+      "step": 441
+    },
+    {
+      "epoch": 1.080684596577017,
+      "grad_norm": 0.08985865165384338,
+      "learning_rate": 4.936236766187944e-06,
+      "loss": 0.5655,
+      "step": 442
+    },
+    {
+      "epoch": 1.0831295843520783,
+      "grad_norm": 0.06979213353817815,
+      "learning_rate": 4.935949158134876e-06,
+      "loss": 0.4365,
+      "step": 443
+    },
+    {
+      "epoch": 1.0855745721271393,
+      "grad_norm": 0.08587094419976782,
+      "learning_rate": 4.935660911320176e-06,
+      "loss": 0.5458,
+      "step": 444
+    },
+    {
+      "epoch": 1.0880195599022005,
+      "grad_norm": 0.06459945875041573,
+      "learning_rate": 4.9353720258194305e-06,
+      "loss": 0.3973,
+      "step": 445
+    },
+    {
+      "epoch": 1.0904645476772616,
+      "grad_norm": 0.06338378792968512,
+      "learning_rate": 4.93508250170839e-06,
+      "loss": 0.4729,
+      "step": 446
+    },
+    {
+      "epoch": 1.0929095354523228,
+      "grad_norm": 0.08758125427073629,
+      "learning_rate": 4.934792339062975e-06,
+      "loss": 0.9112,
+      "step": 447
+    },
+    {
+      "epoch": 1.0953545232273838,
+      "grad_norm": 0.08612473092257603,
+      "learning_rate": 4.934501537959274e-06,
+      "loss": 0.6525,
+      "step": 448
+    },
+    {
+      "epoch": 1.097799511002445,
+      "grad_norm": 0.08764721761343358,
+      "learning_rate": 4.93421009847354e-06,
+      "loss": 0.7525,
+      "step": 449
+    },
+    {
+      "epoch": 1.1002444987775062,
+      "grad_norm": 0.09500426925935404,
+      "learning_rate": 4.933918020682196e-06,
+      "loss": 0.7781,
+      "step": 450
+    },
+    {
+      "epoch": 1.1026894865525672,
+      "grad_norm": 0.13051747246092554,
+      "learning_rate": 4.933625304661831e-06,
+      "loss": 0.5881,
+      "step": 451
+    },
+    {
+      "epoch": 1.1051344743276283,
+      "grad_norm": 0.11079305835216474,
+      "learning_rate": 4.933331950489202e-06,
+      "loss": 0.6278,
+      "step": 452
+    },
+    {
+      "epoch": 1.1075794621026895,
+      "grad_norm": 0.1226532214058774,
+      "learning_rate": 4.933037958241233e-06,
+      "loss": 1.0762,
+      "step": 453
+    },
+    {
+      "epoch": 1.1100244498777505,
+      "grad_norm": 0.08619532896324963,
+      "learning_rate": 4.932743327995017e-06,
+      "loss": 0.7636,
+      "step": 454
+    },
+    {
+      "epoch": 1.1124694376528117,
+      "grad_norm": 0.08527134420681212,
+      "learning_rate": 4.932448059827811e-06,
+      "loss": 0.5058,
+      "step": 455
+    },
+    {
+      "epoch": 1.114914425427873,
+      "grad_norm": 0.08535253600194993,
+      "learning_rate": 4.9321521538170416e-06,
+      "loss": 0.9615,
+      "step": 456
+    },
+    {
+      "epoch": 1.117359413202934,
+      "grad_norm": 0.08863816004879815,
+      "learning_rate": 4.931855610040302e-06,
+      "loss": 1.0531,
+      "step": 457
+    },
+    {
+      "epoch": 1.119804400977995,
+      "grad_norm": 0.09487267031600953,
+      "learning_rate": 4.931558428575353e-06,
+      "loss": 0.8523,
+      "step": 458
+    },
+    {
+      "epoch": 1.1222493887530562,
+      "grad_norm": 0.09118134439196415,
+      "learning_rate": 4.931260609500122e-06,
+      "loss": 0.5716,
+      "step": 459
+    },
+    {
+      "epoch": 1.1246943765281174,
+      "grad_norm": 0.08850224361698028,
+      "learning_rate": 4.930962152892705e-06,
+      "loss": 0.5606,
+      "step": 460
+    },
+    {
+      "epoch": 1.1271393643031784,
+      "grad_norm": 0.07368599896584542,
+      "learning_rate": 4.930663058831363e-06,
+      "loss": 0.6512,
+      "step": 461
+    },
+    {
+      "epoch": 1.1295843520782396,
+      "grad_norm": 0.15835534489259462,
+      "learning_rate": 4.930363327394526e-06,
+      "loss": 0.483,
+      "step": 462
+    },
+    {
+      "epoch": 1.1320293398533008,
+      "grad_norm": 0.09847635370746623,
+      "learning_rate": 4.930062958660791e-06,
+      "loss": 0.6378,
+      "step": 463
+    },
+    {
+      "epoch": 1.1344743276283618,
+      "grad_norm": 0.12826222819889038,
+      "learning_rate": 4.929761952708919e-06,
+      "loss": 0.6869,
+      "step": 464
+    },
+    {
+      "epoch": 1.136919315403423,
+      "grad_norm": 0.0984962566017247,
+      "learning_rate": 4.929460309617844e-06,
+      "loss": 0.6264,
+      "step": 465
+    },
+    {
+      "epoch": 1.1393643031784841,
+      "grad_norm": 0.06699741722565294,
+      "learning_rate": 4.92915802946666e-06,
+      "loss": 0.5402,
+      "step": 466
+    },
+    {
+      "epoch": 1.141809290953545,
+      "grad_norm": 0.06976404766756682,
+      "learning_rate": 4.928855112334637e-06,
+      "loss": 0.5558,
+      "step": 467
+    },
+    {
+      "epoch": 1.1442542787286063,
+      "grad_norm": 0.09089804323140854,
+      "learning_rate": 4.928551558301202e-06,
+      "loss": 0.5615,
+      "step": 468
+    },
+    {
+      "epoch": 1.1466992665036675,
+      "grad_norm": 0.0728901145580466,
+      "learning_rate": 4.928247367445955e-06,
+      "loss": 0.5883,
+      "step": 469
+    },
+    {
+      "epoch": 1.1491442542787287,
+      "grad_norm": 0.07749424237991626,
+      "learning_rate": 4.927942539848664e-06,
+      "loss": 0.4626,
+      "step": 470
+    },
+    {
+      "epoch": 1.1515892420537897,
+      "grad_norm": 0.10542078728716303,
+      "learning_rate": 4.927637075589259e-06,
+      "loss": 0.4384,
+      "step": 471
+    },
+    {
+      "epoch": 1.1540342298288508,
+      "grad_norm": 0.09666276313939713,
+      "learning_rate": 4.9273309747478405e-06,
+      "loss": 0.8628,
+      "step": 472
+    },
+    {
+      "epoch": 1.156479217603912,
+      "grad_norm": 0.08803531242886628,
+      "learning_rate": 4.927024237404676e-06,
+      "loss": 0.9808,
+      "step": 473
+    },
+    {
+      "epoch": 1.1589242053789732,
+      "grad_norm": 0.06543209808717486,
+      "learning_rate": 4.926716863640198e-06,
+      "loss": 0.7855,
+      "step": 474
+    },
+    {
+      "epoch": 1.1613691931540342,
+      "grad_norm": 0.07860052143999972,
+      "learning_rate": 4.926408853535009e-06,
+      "loss": 0.8213,
+      "step": 475
+    },
+    {
+      "epoch": 1.1638141809290954,
+      "grad_norm": 0.10931352468951576,
+      "learning_rate": 4.926100207169874e-06,
+      "loss": 0.581,
+      "step": 476
+    },
+    {
+      "epoch": 1.1662591687041566,
+      "grad_norm": 0.15010409191628882,
+      "learning_rate": 4.925790924625727e-06,
+      "loss": 0.6576,
+      "step": 477
+    },
+    {
+      "epoch": 1.1687041564792175,
+      "grad_norm": 0.09910094501518846,
+      "learning_rate": 4.925481005983671e-06,
+      "loss": 0.5308,
+      "step": 478
+    },
+    {
+      "epoch": 1.1711491442542787,
+      "grad_norm": 0.09770242061791468,
+      "learning_rate": 4.925170451324973e-06,
+      "loss": 0.9874,
+      "step": 479
+    },
+    {
+      "epoch": 1.17359413202934,
+      "grad_norm": 0.08030153268201529,
+      "learning_rate": 4.924859260731067e-06,
+      "loss": 0.6364,
+      "step": 480
+    },
+    {
+      "epoch": 1.176039119804401,
+      "grad_norm": 0.06928780822098991,
+      "learning_rate": 4.924547434283553e-06,
+      "loss": 0.5545,
+      "step": 481
+    },
+    {
+      "epoch": 1.178484107579462,
+      "grad_norm": 0.09237649881859512,
+      "learning_rate": 4.924234972064202e-06,
+      "loss": 0.5735,
+      "step": 482
+    },
+    {
+      "epoch": 1.1809290953545233,
+      "grad_norm": 0.08688040667426593,
+      "learning_rate": 4.923921874154947e-06,
+      "loss": 0.6035,
+      "step": 483
+    },
+    {
+      "epoch": 1.1833740831295843,
+      "grad_norm": 0.11807118434345489,
+      "learning_rate": 4.9236081406378896e-06,
+      "loss": 0.828,
+      "step": 484
+    },
+    {
+      "epoch": 1.1858190709046454,
+      "grad_norm": 0.08796747058149752,
+      "learning_rate": 4.923293771595298e-06,
+      "loss": 0.6264,
+      "step": 485
+    },
+    {
+      "epoch": 1.1882640586797066,
+      "grad_norm": 0.09561220786072974,
+      "learning_rate": 4.922978767109606e-06,
+      "loss": 0.7769,
+      "step": 486
+    },
+    {
+      "epoch": 1.1907090464547676,
+      "grad_norm": 0.09088930540866695,
+      "learning_rate": 4.922663127263417e-06,
+      "loss": 0.8633,
+      "step": 487
+    },
+    {
+      "epoch": 1.1931540342298288,
+      "grad_norm": 0.10340377119760334,
+      "learning_rate": 4.922346852139498e-06,
+      "loss": 0.5747,
+      "step": 488
+    },
+    {
+      "epoch": 1.19559902200489,
+      "grad_norm": 0.11309677235280163,
+      "learning_rate": 4.922029941820784e-06,
+      "loss": 0.7417,
+      "step": 489
+    },
+    {
+      "epoch": 1.1980440097799512,
+      "grad_norm": 0.08953581187302126,
+      "learning_rate": 4.921712396390375e-06,
+      "loss": 0.7057,
+      "step": 490
+    },
+    {
+      "epoch": 1.2004889975550121,
+      "grad_norm": 0.09542073194832858,
+      "learning_rate": 4.92139421593154e-06,
+      "loss": 1.0115,
+      "step": 491
+    },
+    {
+      "epoch": 1.2029339853300733,
+      "grad_norm": 0.09435097212535767,
+      "learning_rate": 4.921075400527712e-06,
+      "loss": 0.8826,
+      "step": 492
+    },
+    {
+      "epoch": 1.2053789731051345,
+      "grad_norm": 0.11780954250490948,
+      "learning_rate": 4.920755950262491e-06,
+      "loss": 0.6608,
+      "step": 493
+    },
+    {
+      "epoch": 1.2078239608801957,
+      "grad_norm": 0.07779457888373696,
+      "learning_rate": 4.920435865219647e-06,
+      "loss": 0.4154,
+      "step": 494
+    },
+    {
+      "epoch": 1.2102689486552567,
+      "grad_norm": 0.11527553832608461,
+      "learning_rate": 4.920115145483112e-06,
+      "loss": 0.5662,
+      "step": 495
+    },
+    {
+      "epoch": 1.2127139364303179,
+      "grad_norm": 0.0888650857447355,
+      "learning_rate": 4.919793791136986e-06,
+      "loss": 0.9286,
+      "step": 496
+    },
+    {
+      "epoch": 1.215158924205379,
+      "grad_norm": 0.07589598115227027,
+      "learning_rate": 4.919471802265536e-06,
+      "loss": 0.5071,
+      "step": 497
+    },
+    {
+      "epoch": 1.21760391198044,
+      "grad_norm": 0.0964212187879595,
+      "learning_rate": 4.919149178953193e-06,
+      "loss": 0.952,
+      "step": 498
+    },
+    {
+      "epoch": 1.2200488997555012,
+      "grad_norm": 0.0983068903375565,
+      "learning_rate": 4.918825921284559e-06,
+      "loss": 0.8473,
+      "step": 499
+    },
+    {
+      "epoch": 1.2224938875305624,
+      "grad_norm": 0.11017880721530393,
+      "learning_rate": 4.918502029344398e-06,
+      "loss": 0.5354,
+      "step": 500
+    },
+    {
+      "epoch": 1.2249388753056234,
+      "grad_norm": 0.08369505279974075,
+      "learning_rate": 4.918177503217642e-06,
+      "loss": 0.5319,
+      "step": 501
+    },
+    {
+      "epoch": 1.2273838630806846,
+      "grad_norm": 0.08956559278845494,
+      "learning_rate": 4.917852342989389e-06,
+      "loss": 0.8995,
+      "step": 502
+    },
+    {
+      "epoch": 1.2298288508557458,
+      "grad_norm": 0.07862102544640515,
+      "learning_rate": 4.917526548744903e-06,
+      "loss": 0.6915,
+      "step": 503
+    },
+    {
+      "epoch": 1.2322738386308068,
+      "grad_norm": 0.11242447186742278,
+      "learning_rate": 4.917200120569617e-06,
+      "loss": 0.5762,
+      "step": 504
+    },
+    {
+      "epoch": 1.234718826405868,
+      "grad_norm": 0.08847837399366476,
+      "learning_rate": 4.916873058549125e-06,
+      "loss": 0.5408,
+      "step": 505
+    },
+    {
+      "epoch": 1.2371638141809291,
+      "grad_norm": 0.09974872244595148,
+      "learning_rate": 4.916545362769193e-06,
+      "loss": 0.7934,
+      "step": 506
+    },
+    {
+      "epoch": 1.23960880195599,
+      "grad_norm": 0.0912495810073242,
+      "learning_rate": 4.9162170333157475e-06,
+      "loss": 0.6672,
+      "step": 507
+    },
+    {
+      "epoch": 1.2420537897310513,
+      "grad_norm": 0.08109256305667999,
+      "learning_rate": 4.915888070274886e-06,
+      "loss": 0.5375,
+      "step": 508
+    },
+    {
+      "epoch": 1.2444987775061125,
+      "grad_norm": 0.07759709958745548,
+      "learning_rate": 4.915558473732869e-06,
+      "loss": 0.6376,
+      "step": 509
+    },
+    {
+      "epoch": 1.2469437652811737,
+      "grad_norm": 0.10773079697506899,
+      "learning_rate": 4.915228243776125e-06,
+      "loss": 0.7018,
+      "step": 510
+    },
+    {
+      "epoch": 1.2493887530562346,
+      "grad_norm": 0.06967041925635098,
+      "learning_rate": 4.914897380491248e-06,
+      "loss": 0.3793,
+      "step": 511
+    },
+    {
+      "epoch": 1.2518337408312958,
+      "grad_norm": 0.11057503599973109,
+      "learning_rate": 4.914565883964997e-06,
+      "loss": 0.5856,
+      "step": 512
+    },
+    {
+      "epoch": 1.254278728606357,
+      "grad_norm": 0.14273770447725426,
+      "learning_rate": 4.914233754284299e-06,
+      "loss": 0.8803,
+      "step": 513
+    },
+    {
+      "epoch": 1.2567237163814182,
+      "grad_norm": 0.11294192465425508,
+      "learning_rate": 4.913900991536245e-06,
+      "loss": 0.7196,
+      "step": 514
+    },
+    {
+      "epoch": 1.2591687041564792,
+      "grad_norm": 0.107378850412448,
+      "learning_rate": 4.913567595808094e-06,
+      "loss": 0.9583,
+      "step": 515
+    },
+    {
+      "epoch": 1.2616136919315404,
+      "grad_norm": 0.11842710303845422,
+      "learning_rate": 4.913233567187269e-06,
+      "loss": 0.5769,
+      "step": 516
+    },
+    {
+      "epoch": 1.2640586797066016,
+      "grad_norm": 0.10554500812774531,
+      "learning_rate": 4.91289890576136e-06,
+      "loss": 1.0385,
+      "step": 517
+    },
+    {
+      "epoch": 1.2665036674816625,
+      "grad_norm": 0.09190779802084394,
+      "learning_rate": 4.912563611618125e-06,
+      "loss": 0.5658,
+      "step": 518
+    },
+    {
+      "epoch": 1.2689486552567237,
+      "grad_norm": 0.07373464166101175,
+      "learning_rate": 4.912227684845483e-06,
+      "loss": 0.5163,
+      "step": 519
+    },
+    {
+      "epoch": 1.271393643031785,
+      "grad_norm": 0.1363402262737353,
+      "learning_rate": 4.911891125531524e-06,
+      "loss": 0.5468,
+      "step": 520
+    },
+    {
+      "epoch": 1.273838630806846,
+      "grad_norm": 0.07531750339186945,
+      "learning_rate": 4.911553933764499e-06,
+      "loss": 0.5689,
+      "step": 521
+    },
+    {
+      "epoch": 1.276283618581907,
+      "grad_norm": 0.11737246092971938,
+      "learning_rate": 4.911216109632829e-06,
+      "loss": 0.6143,
+      "step": 522
+    },
+    {
+      "epoch": 1.2787286063569683,
+      "grad_norm": 0.10621515820557399,
+      "learning_rate": 4.9108776532251e-06,
+      "loss": 0.5884,
+      "step": 523
+    },
+    {
+      "epoch": 1.2811735941320292,
+      "grad_norm": 0.10331919975435638,
+      "learning_rate": 4.910538564630061e-06,
+      "loss": 0.9745,
+      "step": 524
+    },
+    {
+      "epoch": 1.2836185819070904,
+      "grad_norm": 0.05899657654770767,
+      "learning_rate": 4.91019884393663e-06,
+      "loss": 0.3933,
+      "step": 525
+    },
+    {
+      "epoch": 1.2860635696821516,
+      "grad_norm": 0.09123789489194563,
+      "learning_rate": 4.90985849123389e-06,
+      "loss": 0.7584,
+      "step": 526
+    },
+    {
+      "epoch": 1.2885085574572126,
+      "grad_norm": 0.1295666077133763,
+      "learning_rate": 4.909517506611088e-06,
+      "loss": 1.1315,
+      "step": 527
+    },
+    {
+      "epoch": 1.2909535452322738,
+      "grad_norm": 0.08248827441123903,
+      "learning_rate": 4.909175890157639e-06,
+      "loss": 0.9952,
+      "step": 528
+    },
+    {
+      "epoch": 1.293398533007335,
+      "grad_norm": 0.08797103703404537,
+      "learning_rate": 4.908833641963122e-06,
+      "loss": 0.6969,
+      "step": 529
+    },
+    {
+      "epoch": 1.295843520782396,
+      "grad_norm": 0.1240972475850901,
+      "learning_rate": 4.9084907621172826e-06,
+      "loss": 0.6394,
+      "step": 530
+    },
+    {
+      "epoch": 1.2982885085574571,
+      "grad_norm": 0.09819301167927469,
+      "learning_rate": 4.908147250710032e-06,
+      "loss": 0.4723,
+      "step": 531
+    },
+    {
+      "epoch": 1.3007334963325183,
+      "grad_norm": 0.11175425875316981,
+      "learning_rate": 4.907803107831446e-06,
+      "loss": 0.9053,
+      "step": 532
+    },
+    {
+      "epoch": 1.3031784841075795,
+      "grad_norm": 0.10047112342425647,
+      "learning_rate": 4.907458333571767e-06,
+      "loss": 0.5744,
+      "step": 533
+    },
+    {
+      "epoch": 1.3056234718826407,
+      "grad_norm": 0.11503275720822878,
+      "learning_rate": 4.907112928021404e-06,
+      "loss": 0.5232,
+      "step": 534
+    },
+    {
+      "epoch": 1.3080684596577017,
+      "grad_norm": 0.0715738413310953,
+      "learning_rate": 4.906766891270929e-06,
+      "loss": 0.5465,
+      "step": 535
+    },
+    {
+      "epoch": 1.3105134474327629,
+      "grad_norm": 0.10741727799875068,
+      "learning_rate": 4.906420223411081e-06,
+      "loss": 0.5217,
+      "step": 536
+    },
+    {
+      "epoch": 1.312958435207824,
+      "grad_norm": 0.10568838429719278,
+      "learning_rate": 4.906072924532763e-06,
+      "loss": 0.8003,
+      "step": 537
+    },
+    {
+      "epoch": 1.315403422982885,
+      "grad_norm": 0.08231711039006366,
+      "learning_rate": 4.905724994727047e-06,
+      "loss": 0.4224,
+      "step": 538
+    },
+    {
+      "epoch": 1.3178484107579462,
+      "grad_norm": 0.08272248563292553,
+      "learning_rate": 4.905376434085166e-06,
+      "loss": 0.5293,
+      "step": 539
+    },
+    {
+      "epoch": 1.3202933985330074,
+      "grad_norm": 0.14225219728070146,
+      "learning_rate": 4.905027242698523e-06,
+      "loss": 0.7543,
+      "step": 540
+    },
+    {
+      "epoch": 1.3227383863080684,
+      "grad_norm": 0.08937075513214697,
+      "learning_rate": 4.904677420658681e-06,
+      "loss": 0.7368,
+      "step": 541
+    },
+    {
+      "epoch": 1.3251833740831296,
+      "grad_norm": 0.10592659035193994,
+      "learning_rate": 4.904326968057375e-06,
+      "loss": 0.9604,
+      "step": 542
+    },
+    {
+      "epoch": 1.3276283618581908,
+      "grad_norm": 0.11904973404505163,
+      "learning_rate": 4.903975884986498e-06,
+      "loss": 0.6545,
+      "step": 543
+    },
+    {
+      "epoch": 1.3300733496332517,
+      "grad_norm": 0.14959675224549268,
+      "learning_rate": 4.903624171538114e-06,
+      "loss": 0.528,
+      "step": 544
+    },
+    {
+      "epoch": 1.332518337408313,
+      "grad_norm": 0.09726056784858919,
+      "learning_rate": 4.90327182780445e-06,
+      "loss": 0.9798,
+      "step": 545
+    },
+    {
+      "epoch": 1.3349633251833741,
+      "grad_norm": 0.1177418094033225,
+      "learning_rate": 4.902918853877899e-06,
+      "loss": 0.929,
+      "step": 546
+    },
+    {
+      "epoch": 1.337408312958435,
+      "grad_norm": 0.08955869555869381,
+      "learning_rate": 4.902565249851018e-06,
+      "loss": 0.5253,
+      "step": 547
+    },
+    {
+      "epoch": 1.3398533007334963,
+      "grad_norm": 0.14758677449990193,
+      "learning_rate": 4.9022110158165305e-06,
+      "loss": 1.174,
+      "step": 548
+    },
+    {
+      "epoch": 1.3422982885085575,
+      "grad_norm": 0.10098649490249025,
+      "learning_rate": 4.901856151867325e-06,
+      "loss": 1.0261,
+      "step": 549
+    },
+    {
+      "epoch": 1.3447432762836184,
+      "grad_norm": 0.09395978178833167,
+      "learning_rate": 4.901500658096454e-06,
+      "loss": 1.0083,
+      "step": 550
+    },
+    {
+      "epoch": 1.3471882640586796,
+      "grad_norm": 0.09996676907015078,
+      "learning_rate": 4.901144534597137e-06,
+      "loss": 0.6274,
+      "step": 551
+    },
+    {
+      "epoch": 1.3496332518337408,
+      "grad_norm": 0.08130887512859405,
+      "learning_rate": 4.900787781462759e-06,
+      "loss": 0.5326,
+      "step": 552
+    },
+    {
+      "epoch": 1.352078239608802,
+      "grad_norm": 0.10394170759786459,
+      "learning_rate": 4.9004303987868666e-06,
+      "loss": 1.2606,
+      "step": 553
+    },
+    {
+      "epoch": 1.3545232273838632,
+      "grad_norm": 0.13384230535890962,
+      "learning_rate": 4.900072386663174e-06,
+      "loss": 0.397,
+      "step": 554
+    },
+    {
+      "epoch": 1.3569682151589242,
+      "grad_norm": 0.09223993894602162,
+      "learning_rate": 4.89971374518556e-06,
+      "loss": 0.6086,
+      "step": 555
+    },
+    {
+      "epoch": 1.3594132029339854,
+      "grad_norm": 0.14212086767132762,
+      "learning_rate": 4.899354474448071e-06,
+      "loss": 0.5452,
+      "step": 556
+    },
+    {
+      "epoch": 1.3618581907090466,
+      "grad_norm": 0.12688109844870638,
+      "learning_rate": 4.898994574544913e-06,
+      "loss": 0.9662,
+      "step": 557
+    },
+    {
+      "epoch": 1.3643031784841075,
+      "grad_norm": 0.10037622149272513,
+      "learning_rate": 4.898634045570462e-06,
+      "loss": 1.0087,
+      "step": 558
+    },
+    {
+      "epoch": 1.3667481662591687,
+      "grad_norm": 0.09468301175713245,
+      "learning_rate": 4.898272887619256e-06,
+      "loss": 0.6379,
+      "step": 559
+    },
+    {
+      "epoch": 1.36919315403423,
+      "grad_norm": 0.07645506116702744,
+      "learning_rate": 4.897911100785999e-06,
+      "loss": 0.51,
+      "step": 560
+    },
+    {
+      "epoch": 1.371638141809291,
+      "grad_norm": 0.1022377367280307,
+      "learning_rate": 4.89754868516556e-06,
+      "loss": 0.7167,
+      "step": 561
+    },
+    {
+      "epoch": 1.374083129584352,
+      "grad_norm": 0.08250956797244187,
+      "learning_rate": 4.897185640852973e-06,
+      "loss": 0.4682,
+      "step": 562
+    },
+    {
+      "epoch": 1.3765281173594133,
+      "grad_norm": 0.10906077574384798,
+      "learning_rate": 4.8968219679434355e-06,
+      "loss": 0.5498,
+      "step": 563
+    },
+    {
+      "epoch": 1.3789731051344742,
+      "grad_norm": 0.1050286030157204,
+      "learning_rate": 4.896457666532312e-06,
+      "loss": 0.9047,
+      "step": 564
+    },
+    {
+      "epoch": 1.3814180929095354,
+      "grad_norm": 0.11088361772351074,
+      "learning_rate": 4.89609273671513e-06,
+      "loss": 0.4295,
+      "step": 565
+    },
+    {
+      "epoch": 1.3838630806845966,
+      "grad_norm": 0.1169530036601329,
+      "learning_rate": 4.895727178587582e-06,
+      "loss": 0.483,
+      "step": 566
+    },
+    {
+      "epoch": 1.3863080684596576,
+      "grad_norm": 0.10731107916852321,
+      "learning_rate": 4.895360992245527e-06,
+      "loss": 0.5084,
+      "step": 567
+    },
+    {
+      "epoch": 1.3887530562347188,
+      "grad_norm": 0.08644998920942834,
+      "learning_rate": 4.894994177784986e-06,
+      "loss": 0.6619,
+      "step": 568
+    },
+    {
+      "epoch": 1.39119804400978,
+      "grad_norm": 0.09434787955252864,
+      "learning_rate": 4.894626735302148e-06,
+      "loss": 0.9048,
+      "step": 569
+    },
+    {
+      "epoch": 1.393643031784841,
+      "grad_norm": 0.13113408831860734,
+      "learning_rate": 4.894258664893363e-06,
+      "loss": 0.6551,
+      "step": 570
+    },
+    {
+      "epoch": 1.3960880195599021,
+      "grad_norm": 0.09423882718762316,
+      "learning_rate": 4.893889966655148e-06,
+      "loss": 0.7766,
+      "step": 571
+    },
+    {
+      "epoch": 1.3985330073349633,
+      "grad_norm": 0.09437923132162945,
+      "learning_rate": 4.893520640684185e-06,
+      "loss": 0.5685,
+      "step": 572
+    },
+    {
+      "epoch": 1.4009779951100245,
+      "grad_norm": 0.11353033262784303,
+      "learning_rate": 4.893150687077318e-06,
+      "loss": 0.4965,
+      "step": 573
+    },
+    {
+      "epoch": 1.4034229828850857,
+      "grad_norm": 0.08549231850722978,
+      "learning_rate": 4.892780105931559e-06,
+      "loss": 0.5272,
+      "step": 574
+    },
+    {
+      "epoch": 1.4058679706601467,
+      "grad_norm": 0.11421060288616502,
+      "learning_rate": 4.892408897344083e-06,
+      "loss": 0.6486,
+      "step": 575
+    },
+    {
+      "epoch": 1.4083129584352079,
+      "grad_norm": 0.1331043299439341,
+      "learning_rate": 4.8920370614122285e-06,
+      "loss": 0.5588,
+      "step": 576
+    },
+    {
+      "epoch": 1.410757946210269,
+      "grad_norm": 0.1976210080602442,
+      "learning_rate": 4.891664598233499e-06,
+      "loss": 0.7242,
+      "step": 577
+    },
+    {
+      "epoch": 1.41320293398533,
+      "grad_norm": 0.13434955998938355,
+      "learning_rate": 4.891291507905562e-06,
+      "loss": 0.4834,
+      "step": 578
+    },
+    {
+      "epoch": 1.4156479217603912,
+      "grad_norm": 0.12494403475060947,
+      "learning_rate": 4.8909177905262535e-06,
+      "loss": 1.1105,
+      "step": 579
+    },
+    {
+      "epoch": 1.4180929095354524,
+      "grad_norm": 0.11365029821295738,
+      "learning_rate": 4.8905434461935685e-06,
+      "loss": 0.4461,
+      "step": 580
+    },
+    {
+      "epoch": 1.4205378973105134,
+      "grad_norm": 0.10077186189663338,
+      "learning_rate": 4.890168475005669e-06,
+      "loss": 0.8464,
+      "step": 581
+    },
+    {
+      "epoch": 1.4229828850855746,
+      "grad_norm": 0.12555527740911893,
+      "learning_rate": 4.889792877060881e-06,
+      "loss": 0.5611,
+      "step": 582
+    },
+    {
+      "epoch": 1.4254278728606358,
+      "grad_norm": 0.09187134597673266,
+      "learning_rate": 4.889416652457695e-06,
+      "loss": 0.4763,
+      "step": 583
+    },
+    {
+      "epoch": 1.4278728606356967,
+      "grad_norm": 0.09586209515602012,
+      "learning_rate": 4.889039801294766e-06,
+      "loss": 0.6886,
+      "step": 584
+    },
+    {
+      "epoch": 1.430317848410758,
+      "grad_norm": 0.10092738627949223,
+      "learning_rate": 4.888662323670913e-06,
+      "loss": 0.8033,
+      "step": 585
+    },
+    {
+      "epoch": 1.4327628361858191,
+      "grad_norm": 0.11759176095909574,
+      "learning_rate": 4.888284219685118e-06,
+      "loss": 0.4628,
+      "step": 586
+    },
+    {
+      "epoch": 1.43520782396088,
+      "grad_norm": 0.09084250273361362,
+      "learning_rate": 4.8879054894365304e-06,
+      "loss": 0.6345,
+      "step": 587
+    },
+    {
+      "epoch": 1.4376528117359413,
+      "grad_norm": 0.11195802313909085,
+      "learning_rate": 4.887526133024461e-06,
+      "loss": 0.5431,
+      "step": 588
+    },
+    {
+      "epoch": 1.4400977995110025,
+      "grad_norm": 0.07866952681745683,
+      "learning_rate": 4.887146150548385e-06,
+      "loss": 0.5937,
+      "step": 589
+    },
+    {
+      "epoch": 1.4425427872860634,
+      "grad_norm": 0.1047830764367085,
+      "learning_rate": 4.886765542107944e-06,
+      "loss": 0.6097,
+      "step": 590
+    },
+    {
+      "epoch": 1.4449877750611246,
+      "grad_norm": 0.11716144373604559,
+      "learning_rate": 4.886384307802942e-06,
+      "loss": 0.5488,
+      "step": 591
+    },
+    {
+      "epoch": 1.4474327628361858,
+      "grad_norm": 0.10157336232284198,
+      "learning_rate": 4.8860024477333454e-06,
+      "loss": 0.4825,
+      "step": 592
+    },
+    {
+      "epoch": 1.449877750611247,
+      "grad_norm": 0.09591034445057087,
+      "learning_rate": 4.88561996199929e-06,
+      "loss": 0.7181,
+      "step": 593
+    },
+    {
+      "epoch": 1.452322738386308,
+      "grad_norm": 0.12073546731090626,
+      "learning_rate": 4.885236850701069e-06,
+      "loss": 0.6599,
+      "step": 594
+    },
+    {
+      "epoch": 1.4547677261613692,
+      "grad_norm": 0.12359034156662388,
+      "learning_rate": 4.8848531139391445e-06,
+      "loss": 0.5033,
+      "step": 595
+    },
+    {
+      "epoch": 1.4572127139364304,
+      "grad_norm": 0.11139842337838901,
+      "learning_rate": 4.884468751814141e-06,
+      "loss": 0.6316,
+      "step": 596
+    },
+    {
+      "epoch": 1.4596577017114916,
+      "grad_norm": 0.09633096943614694,
+      "learning_rate": 4.884083764426848e-06,
+      "loss": 0.5619,
+      "step": 597
+    },
+    {
+      "epoch": 1.4621026894865525,
+      "grad_norm": 0.12047753581913498,
+      "learning_rate": 4.883698151878216e-06,
+      "loss": 0.8111,
+      "step": 598
+    },
+    {
+      "epoch": 1.4645476772616137,
+      "grad_norm": 0.08539043747485742,
+      "learning_rate": 4.883311914269363e-06,
+      "loss": 0.499,
+      "step": 599
+    },
+    {
+      "epoch": 1.466992665036675,
+      "grad_norm": 0.0853362857331394,
+      "learning_rate": 4.882925051701568e-06,
+      "loss": 0.5415,
+      "step": 600
+    },
+    {
+      "epoch": 1.4694376528117359,
+      "grad_norm": 0.07729888911689278,
+      "learning_rate": 4.882537564276277e-06,
+      "loss": 0.502,
+      "step": 601
+    },
+    {
+      "epoch": 1.471882640586797,
+      "grad_norm": 0.119589983288633,
+      "learning_rate": 4.882149452095098e-06,
+      "loss": 0.5199,
+      "step": 602
+    },
+    {
+      "epoch": 1.4743276283618583,
+      "grad_norm": 0.10830874637390676,
+      "learning_rate": 4.8817607152598e-06,
+      "loss": 0.7056,
+      "step": 603
+    },
+    {
+      "epoch": 1.4767726161369192,
+      "grad_norm": 0.11304500383396528,
+      "learning_rate": 4.881371353872321e-06,
+      "loss": 0.5552,
+      "step": 604
+    },
+    {
+      "epoch": 1.4792176039119804,
+      "grad_norm": 0.14761148604727087,
+      "learning_rate": 4.88098136803476e-06,
+      "loss": 0.4864,
+      "step": 605
+    },
+    {
+      "epoch": 1.4816625916870416,
+      "grad_norm": 0.08616052223843693,
+      "learning_rate": 4.8805907578493815e-06,
+      "loss": 0.6932,
+      "step": 606
+    },
+    {
+      "epoch": 1.4841075794621026,
+      "grad_norm": 0.0969487209157877,
+      "learning_rate": 4.880199523418612e-06,
+      "loss": 0.5311,
+      "step": 607
+    },
+    {
+      "epoch": 1.4865525672371638,
+      "grad_norm": 0.1708784553402437,
+      "learning_rate": 4.87980766484504e-06,
+      "loss": 0.7008,
+      "step": 608
+    },
+    {
+      "epoch": 1.488997555012225,
+      "grad_norm": 0.1516345812888604,
+      "learning_rate": 4.879415182231421e-06,
+      "loss": 0.9707,
+      "step": 609
+    },
+    {
+      "epoch": 1.491442542787286,
+      "grad_norm": 0.10924570175359778,
+      "learning_rate": 4.879022075680674e-06,
+      "loss": 0.4253,
+      "step": 610
+    },
+    {
+      "epoch": 1.4938875305623471,
+      "grad_norm": 0.09878909510845578,
+      "learning_rate": 4.87862834529588e-06,
+      "loss": 0.9475,
+      "step": 611
+    },
+    {
+      "epoch": 1.4963325183374083,
+      "grad_norm": 0.10440136070009452,
+      "learning_rate": 4.878233991180282e-06,
+      "loss": 0.507,
+      "step": 612
+    },
+    {
+      "epoch": 1.4987775061124695,
+      "grad_norm": 0.14019649232764822,
+      "learning_rate": 4.877839013437292e-06,
+      "loss": 0.5253,
+      "step": 613
+    },
+    {
+      "epoch": 1.5012224938875307,
+      "grad_norm": 0.1657457171820653,
+      "learning_rate": 4.87744341217048e-06,
+      "loss": 0.9782,
+      "step": 614
+    },
+    {
+      "epoch": 1.5036674816625917,
+      "grad_norm": 0.10151625603491048,
+      "learning_rate": 4.877047187483583e-06,
+      "loss": 0.6492,
+      "step": 615
+    },
+    {
+      "epoch": 1.5061124694376526,
+      "grad_norm": 0.07992442002773023,
+      "learning_rate": 4.8766503394804996e-06,
+      "loss": 0.4593,
+      "step": 616
+    },
+    {
+      "epoch": 1.508557457212714,
+      "grad_norm": 0.1881540397932995,
+      "learning_rate": 4.876252868265292e-06,
+      "loss": 0.5718,
+      "step": 617
+    },
+    {
+      "epoch": 1.511002444987775,
+      "grad_norm": 0.14995548029453482,
+      "learning_rate": 4.875854773942187e-06,
+      "loss": 1.1579,
+      "step": 618
+    },
+    {
+      "epoch": 1.5134474327628362,
+      "grad_norm": 0.1251722466892237,
+      "learning_rate": 4.8754560566155736e-06,
+      "loss": 0.5597,
+      "step": 619
+    },
+    {
+      "epoch": 1.5158924205378974,
+      "grad_norm": 0.11206684872942402,
+      "learning_rate": 4.875056716390004e-06,
+      "loss": 0.5726,
+      "step": 620
+    },
+    {
+      "epoch": 1.5183374083129584,
+      "grad_norm": 0.13186339836517458,
+      "learning_rate": 4.874656753370196e-06,
+      "loss": 0.4605,
+      "step": 621
+    },
+    {
+      "epoch": 1.5207823960880196,
+      "grad_norm": 0.10769843328151026,
+      "learning_rate": 4.874256167661028e-06,
+      "loss": 0.4862,
+      "step": 622
+    },
+    {
+      "epoch": 1.5232273838630808,
+      "grad_norm": 0.16229971977633795,
+      "learning_rate": 4.873854959367543e-06,
+      "loss": 0.5198,
+      "step": 623
+    },
+    {
+      "epoch": 1.5256723716381417,
+      "grad_norm": 0.0718356191915513,
+      "learning_rate": 4.873453128594947e-06,
+      "loss": 0.4432,
+      "step": 624
+    },
+    {
+      "epoch": 1.528117359413203,
+      "grad_norm": 0.14969965259276746,
+      "learning_rate": 4.873050675448608e-06,
+      "loss": 0.72,
+      "step": 625
+    },
+    {
+      "epoch": 1.5305623471882641,
+      "grad_norm": 0.1438543773392057,
+      "learning_rate": 4.8726476000340605e-06,
+      "loss": 1.0206,
+      "step": 626
+    },
+    {
+      "epoch": 1.533007334963325,
+      "grad_norm": 0.1090563727566657,
+      "learning_rate": 4.872243902456999e-06,
+      "loss": 0.8143,
+      "step": 627
+    },
+    {
+      "epoch": 1.5354523227383863,
+      "grad_norm": 0.12047679246999657,
+      "learning_rate": 4.8718395828232825e-06,
+      "loss": 1.0677,
+      "step": 628
+    },
+    {
+      "epoch": 1.5378973105134475,
+      "grad_norm": 0.18664156072454702,
+      "learning_rate": 4.871434641238932e-06,
+      "loss": 0.8253,
+      "step": 629
+    },
+    {
+      "epoch": 1.5403422982885084,
+      "grad_norm": 0.09963534330499423,
+      "learning_rate": 4.871029077810134e-06,
+      "loss": 0.6614,
+      "step": 630
+    },
+    {
+      "epoch": 1.5427872860635696,
+      "grad_norm": 0.11440225158856199,
+      "learning_rate": 4.870622892643234e-06,
+      "loss": 0.81,
+      "step": 631
+    },
+    {
+      "epoch": 1.5452322738386308,
+      "grad_norm": 0.7695658222676224,
+      "learning_rate": 4.870216085844747e-06,
+      "loss": 1.0181,
+      "step": 632
+    },
+    {
+      "epoch": 1.5476772616136918,
+      "grad_norm": 0.09633012470089278,
+      "learning_rate": 4.869808657521344e-06,
+      "loss": 0.5962,
+      "step": 633
+    },
+    {
+      "epoch": 1.5501222493887532,
+      "grad_norm": 0.10346289251632876,
+      "learning_rate": 4.869400607779863e-06,
+      "loss": 0.6151,
+      "step": 634
+    },
+    {
+      "epoch": 1.5525672371638142,
+      "grad_norm": 0.152883330338306,
+      "learning_rate": 4.868991936727303e-06,
+      "loss": 0.8945,
+      "step": 635
+    },
+    {
+      "epoch": 1.5550122249388751,
+      "grad_norm": 0.10790987901824774,
+      "learning_rate": 4.868582644470829e-06,
+      "loss": 0.7651,
+      "step": 636
+    },
+    {
+      "epoch": 1.5574572127139366,
+      "grad_norm": 0.12241414703705915,
+      "learning_rate": 4.868172731117764e-06,
+      "loss": 0.9482,
+      "step": 637
+    },
+    {
+      "epoch": 1.5599022004889975,
+      "grad_norm": 0.10669500105409589,
+      "learning_rate": 4.867762196775599e-06,
+      "loss": 0.71,
+      "step": 638
+    },
+    {
+      "epoch": 1.5623471882640587,
+      "grad_norm": 0.09792055398862386,
+      "learning_rate": 4.867351041551984e-06,
+      "loss": 0.6317,
+      "step": 639
+    },
+    {
+      "epoch": 1.56479217603912,
+      "grad_norm": 0.09256610738995151,
+      "learning_rate": 4.866939265554735e-06,
+      "loss": 0.4865,
+      "step": 640
+    },
+    {
+      "epoch": 1.5672371638141809,
+      "grad_norm": 0.15834405522303402,
+      "learning_rate": 4.866526868891828e-06,
+      "loss": 0.9382,
+      "step": 641
+    },
+    {
+      "epoch": 1.569682151589242,
+      "grad_norm": 0.10862668475006454,
+      "learning_rate": 4.866113851671403e-06,
+      "loss": 0.596,
+      "step": 642
+    },
+    {
+      "epoch": 1.5721271393643033,
+      "grad_norm": 0.10906456716259076,
+      "learning_rate": 4.865700214001762e-06,
+      "loss": 0.4508,
+      "step": 643
+    },
+    {
+      "epoch": 1.5745721271393642,
+      "grad_norm": 0.10754506423305892,
+      "learning_rate": 4.865285955991371e-06,
+      "loss": 0.6482,
+      "step": 644
+    },
+    {
+      "epoch": 1.5770171149144254,
+      "grad_norm": 0.10263451111083424,
+      "learning_rate": 4.864871077748857e-06,
+      "loss": 0.5227,
+      "step": 645
+    },
+    {
+      "epoch": 1.5794621026894866,
+      "grad_norm": 0.09591355620951793,
+      "learning_rate": 4.864455579383013e-06,
+      "loss": 0.5327,
+      "step": 646
+    },
+    {
+      "epoch": 1.5819070904645476,
+      "grad_norm": 0.10028149108562139,
+      "learning_rate": 4.8640394610027895e-06,
+      "loss": 0.716,
+      "step": 647
+    },
+    {
+      "epoch": 1.5843520782396088,
+      "grad_norm": 0.12897076554561426,
+      "learning_rate": 4.863622722717303e-06,
+      "loss": 0.9392,
+      "step": 648
+    },
+    {
+      "epoch": 1.58679706601467,
+      "grad_norm": 0.1370399105921739,
+      "learning_rate": 4.863205364635833e-06,
+      "loss": 0.4962,
+      "step": 649
+    },
+    {
+      "epoch": 1.589242053789731,
+      "grad_norm": 0.13232056999970115,
+      "learning_rate": 4.8627873868678175e-06,
+      "loss": 0.4834,
+      "step": 650
+    },
+    {
+      "epoch": 1.5916870415647921,
+      "grad_norm": 0.09440559206872451,
+      "learning_rate": 4.862368789522864e-06,
+      "loss": 0.652,
+      "step": 651
+    },
+    {
+      "epoch": 1.5941320293398533,
+      "grad_norm": 0.11018987161159294,
+      "learning_rate": 4.861949572710734e-06,
+      "loss": 0.8007,
+      "step": 652
+    },
+    {
+      "epoch": 1.5965770171149143,
+      "grad_norm": 0.08912551249463314,
+      "learning_rate": 4.8615297365413585e-06,
+      "loss": 0.4337,
+      "step": 653
+    },
+    {
+      "epoch": 1.5990220048899757,
+      "grad_norm": 0.13250698685524578,
+      "learning_rate": 4.861109281124828e-06,
+      "loss": 0.4639,
+      "step": 654
+    },
+    {
+      "epoch": 1.6014669926650367,
+      "grad_norm": 0.11707705855611844,
+      "learning_rate": 4.8606882065713946e-06,
+      "loss": 0.9195,
+      "step": 655
+    },
+    {
+      "epoch": 1.6039119804400976,
+      "grad_norm": 0.19145744294286274,
+      "learning_rate": 4.860266512991474e-06,
+      "loss": 0.4865,
+      "step": 656
+    },
+    {
+      "epoch": 1.606356968215159,
+      "grad_norm": 0.17011802967734546,
+      "learning_rate": 4.859844200495644e-06,
+      "loss": 0.5022,
+      "step": 657
+    },
+    {
+      "epoch": 1.60880195599022,
+      "grad_norm": 0.0906458080351694,
+      "learning_rate": 4.859421269194644e-06,
+      "loss": 0.4341,
+      "step": 658
+    },
+    {
+      "epoch": 1.6112469437652812,
+      "grad_norm": 0.099955504189599,
+      "learning_rate": 4.8589977191993774e-06,
+      "loss": 0.5138,
+      "step": 659
+    },
+    {
+      "epoch": 1.6136919315403424,
+      "grad_norm": 0.12079758445894387,
+      "learning_rate": 4.8585735506209085e-06,
+      "loss": 0.5769,
+      "step": 660
+    },
+    {
+      "epoch": 1.6161369193154034,
+      "grad_norm": 0.11353390882262827,
+      "learning_rate": 4.858148763570463e-06,
+      "loss": 0.8745,
+      "step": 661
+    },
+    {
+      "epoch": 1.6185819070904646,
+      "grad_norm": 0.1038161883749101,
+      "learning_rate": 4.8577233581594306e-06,
+      "loss": 0.7222,
+      "step": 662
+    },
+    {
+      "epoch": 1.6210268948655258,
+      "grad_norm": 0.14748619844276883,
+      "learning_rate": 4.857297334499362e-06,
+      "loss": 0.595,
+      "step": 663
+    },
+    {
+      "epoch": 1.6234718826405867,
+      "grad_norm": 0.14399985929179457,
+      "learning_rate": 4.856870692701972e-06,
+      "loss": 0.8044,
+      "step": 664
+    },
+    {
+      "epoch": 1.625916870415648,
+      "grad_norm": 0.09579065432453852,
+      "learning_rate": 4.856443432879134e-06,
+      "loss": 0.5055,
+      "step": 665
+    },
+    {
+      "epoch": 1.628361858190709,
+      "grad_norm": 0.10542201469767118,
+      "learning_rate": 4.856015555142884e-06,
+      "loss": 0.6142,
+      "step": 666
+    },
+    {
+      "epoch": 1.63080684596577,
+      "grad_norm": 0.1590815488156907,
+      "learning_rate": 4.855587059605424e-06,
+      "loss": 0.6915,
+      "step": 667
+    },
+    {
+      "epoch": 1.6332518337408313,
+      "grad_norm": 0.1145729050653525,
+      "learning_rate": 4.855157946379114e-06,
+      "loss": 0.4971,
+      "step": 668
+    },
+    {
+      "epoch": 1.6356968215158925,
+      "grad_norm": 0.12269124317193993,
+      "learning_rate": 4.854728215576479e-06,
+      "loss": 1.0122,
+      "step": 669
+    },
+    {
+      "epoch": 1.6381418092909534,
+      "grad_norm": 0.11115980535243569,
+      "learning_rate": 4.8542978673102015e-06,
+      "loss": 0.9206,
+      "step": 670
+    },
+    {
+      "epoch": 1.6405867970660146,
+      "grad_norm": 0.13340921198939854,
+      "learning_rate": 4.853866901693131e-06,
+      "loss": 1.0498,
+      "step": 671
+    },
+    {
+      "epoch": 1.6430317848410758,
+      "grad_norm": 0.10597716474785111,
+      "learning_rate": 4.853435318838274e-06,
+      "loss": 0.5938,
+      "step": 672
+    },
+    {
+      "epoch": 1.6454767726161368,
+      "grad_norm": 0.0963222664089259,
+      "learning_rate": 4.853003118858805e-06,
+      "loss": 0.5491,
+      "step": 673
+    },
+    {
+      "epoch": 1.6479217603911982,
+      "grad_norm": 0.10252375914532312,
+      "learning_rate": 4.852570301868053e-06,
+      "loss": 0.4956,
+      "step": 674
+    },
+    {
+      "epoch": 1.6503667481662592,
+      "grad_norm": 0.06516883895046549,
+      "learning_rate": 4.852136867979515e-06,
+      "loss": 0.3496,
+      "step": 675
+    },
+    {
+      "epoch": 1.6528117359413201,
+      "grad_norm": 0.11019712518768765,
+      "learning_rate": 4.851702817306848e-06,
+      "loss": 0.5599,
+      "step": 676
+    },
+    {
+      "epoch": 1.6552567237163816,
+      "grad_norm": 0.08023275962675919,
+      "learning_rate": 4.851268149963867e-06,
+      "loss": 0.456,
+      "step": 677
+    },
+    {
+      "epoch": 1.6577017114914425,
+      "grad_norm": 0.12889745988554777,
+      "learning_rate": 4.850832866064554e-06,
+      "loss": 0.8283,
+      "step": 678
+    },
+    {
+      "epoch": 1.6601466992665037,
+      "grad_norm": 0.10399112768826765,
+      "learning_rate": 4.85039696572305e-06,
+      "loss": 0.5162,
+      "step": 679
+    },
+    {
+      "epoch": 1.662591687041565,
+      "grad_norm": 0.13260070586917,
+      "learning_rate": 4.849960449053656e-06,
+      "loss": 0.6224,
+      "step": 680
+    },
+    {
+      "epoch": 1.6650366748166259,
+      "grad_norm": 0.13257132298385343,
+      "learning_rate": 4.849523316170841e-06,
+      "loss": 0.6098,
+      "step": 681
+    },
+    {
+      "epoch": 1.667481662591687,
+      "grad_norm": 0.10854327220032024,
+      "learning_rate": 4.849085567189227e-06,
+      "loss": 0.6587,
+      "step": 682
+    },
+    {
+      "epoch": 1.6699266503667483,
+      "grad_norm": 0.1336869953402008,
+      "learning_rate": 4.848647202223604e-06,
+      "loss": 0.531,
+      "step": 683
+    },
+    {
+      "epoch": 1.6723716381418092,
+      "grad_norm": 0.12036566125205295,
+      "learning_rate": 4.848208221388922e-06,
+      "loss": 0.7005,
+      "step": 684
+    },
+    {
+      "epoch": 1.6748166259168704,
+      "grad_norm": 0.17965483234072654,
+      "learning_rate": 4.847768624800289e-06,
+      "loss": 0.5561,
+      "step": 685
+    },
+    {
+      "epoch": 1.6772616136919316,
+      "grad_norm": 0.15513040979227,
+      "learning_rate": 4.847328412572979e-06,
+      "loss": 0.7079,
+      "step": 686
+    },
+    {
+      "epoch": 1.6797066014669926,
+      "grad_norm": 0.12647590000250364,
+      "learning_rate": 4.846887584822426e-06,
+      "loss": 0.596,
+      "step": 687
+    },
+    {
+      "epoch": 1.6821515892420538,
+      "grad_norm": 0.12898244125643704,
+      "learning_rate": 4.846446141664225e-06,
+      "loss": 0.4973,
+      "step": 688
+    },
+    {
+      "epoch": 1.684596577017115,
+      "grad_norm": 0.14708750883534086,
+      "learning_rate": 4.846004083214132e-06,
+      "loss": 0.5912,
+      "step": 689
+    },
+    {
+      "epoch": 1.687041564792176,
+      "grad_norm": 0.08923262598268303,
+      "learning_rate": 4.845561409588065e-06,
+      "loss": 0.3206,
+      "step": 690
+    },
+    {
+      "epoch": 1.6894865525672371,
+      "grad_norm": 0.10013333489447876,
+      "learning_rate": 4.8451181209021046e-06,
+      "loss": 0.656,
+      "step": 691
+    },
+    {
+      "epoch": 1.6919315403422983,
+      "grad_norm": 0.08596983742011301,
+      "learning_rate": 4.844674217272488e-06,
+      "loss": 0.4029,
+      "step": 692
+    },
+    {
+      "epoch": 1.6943765281173593,
+      "grad_norm": 0.15451618563853647,
+      "learning_rate": 4.844229698815621e-06,
+      "loss": 1.3626,
+      "step": 693
+    },
+    {
+      "epoch": 1.6968215158924207,
+      "grad_norm": 0.13761627324353434,
+      "learning_rate": 4.843784565648063e-06,
+      "loss": 0.855,
+      "step": 694
+    },
+    {
+      "epoch": 1.6992665036674817,
+      "grad_norm": 0.12673090169921572,
+      "learning_rate": 4.843338817886539e-06,
+      "loss": 0.8405,
+      "step": 695
+    },
+    {
+      "epoch": 1.7017114914425426,
+      "grad_norm": 0.11822363087418883,
+      "learning_rate": 4.842892455647935e-06,
+      "loss": 0.5759,
+      "step": 696
+    },
+    {
+      "epoch": 1.704156479217604,
+      "grad_norm": 0.11608723785933603,
+      "learning_rate": 4.842445479049298e-06,
+      "loss": 0.4765,
+      "step": 697
+    },
+    {
+      "epoch": 1.706601466992665,
+      "grad_norm": 0.12431526901571709,
+      "learning_rate": 4.841997888207834e-06,
+      "loss": 0.5168,
+      "step": 698
+    },
+    {
+      "epoch": 1.7090464547677262,
+      "grad_norm": 0.1511962223294636,
+      "learning_rate": 4.841549683240912e-06,
+      "loss": 0.5864,
+      "step": 699
+    },
+    {
+      "epoch": 1.7114914425427874,
+      "grad_norm": 0.09940499078026584,
+      "learning_rate": 4.841100864266063e-06,
+      "loss": 0.5639,
+      "step": 700
+    },
+    {
+      "epoch": 1.7139364303178484,
+      "grad_norm": 0.08216183752276986,
+      "learning_rate": 4.840651431400975e-06,
+      "loss": 0.4249,
+      "step": 701
+    },
+    {
+      "epoch": 1.7163814180929096,
+      "grad_norm": 0.16990803333881208,
+      "learning_rate": 4.840201384763501e-06,
+      "loss": 0.5247,
+      "step": 702
+    },
+    {
+      "epoch": 1.7188264058679708,
+      "grad_norm": 0.09956485159354872,
+      "learning_rate": 4.839750724471654e-06,
+      "loss": 0.76,
+      "step": 703
+    },
+    {
+      "epoch": 1.7212713936430317,
+      "grad_norm": 0.09117376701958932,
+      "learning_rate": 4.839299450643608e-06,
+      "loss": 0.4501,
+      "step": 704
+    },
+    {
+      "epoch": 1.723716381418093,
+      "grad_norm": 0.1495785025028801,
+      "learning_rate": 4.838847563397694e-06,
+      "loss": 0.5441,
+      "step": 705
+    },
+    {
+      "epoch": 1.726161369193154,
+      "grad_norm": 0.1363291441701734,
+      "learning_rate": 4.838395062852411e-06,
+      "loss": 0.5831,
+      "step": 706
+    },
+    {
+      "epoch": 1.728606356968215,
+      "grad_norm": 0.1342577920804179,
+      "learning_rate": 4.837941949126413e-06,
+      "loss": 0.5628,
+      "step": 707
+    },
+    {
+      "epoch": 1.7310513447432763,
+      "grad_norm": 0.12277509365567624,
+      "learning_rate": 4.837488222338517e-06,
+      "loss": 0.5569,
+      "step": 708
+    },
+    {
+      "epoch": 1.7334963325183375,
+      "grad_norm": 0.17138058565246797,
+      "learning_rate": 4.8370338826077016e-06,
+      "loss": 0.6677,
+      "step": 709
+    },
+    {
+      "epoch": 1.7359413202933984,
+      "grad_norm": 0.1621979769406644,
+      "learning_rate": 4.836578930053104e-06,
+      "loss": 0.5889,
+      "step": 710
+    },
+    {
+      "epoch": 1.7383863080684596,
+      "grad_norm": 0.20299917175052679,
+      "learning_rate": 4.836123364794023e-06,
+      "loss": 0.5004,
+      "step": 711
+    },
+    {
+      "epoch": 1.7408312958435208,
+      "grad_norm": 0.18306590170774154,
+      "learning_rate": 4.835667186949918e-06,
+      "loss": 0.4694,
+      "step": 712
+    },
+    {
+      "epoch": 1.7432762836185818,
+      "grad_norm": 0.11741062071005252,
+      "learning_rate": 4.835210396640412e-06,
+      "loss": 0.9465,
+      "step": 713
+    },
+    {
+      "epoch": 1.7457212713936432,
+      "grad_norm": 0.11523234138292135,
+      "learning_rate": 4.834752993985282e-06,
+      "loss": 0.8555,
+      "step": 714
+    },
+    {
+      "epoch": 1.7481662591687042,
+      "grad_norm": 0.10149021095616566,
+      "learning_rate": 4.834294979104472e-06,
+      "loss": 0.4816,
+      "step": 715
+    },
+    {
+      "epoch": 1.7506112469437651,
+      "grad_norm": 0.17718400299002438,
+      "learning_rate": 4.833836352118084e-06,
+      "loss": 0.7864,
+      "step": 716
+    },
+    {
+      "epoch": 1.7530562347188265,
+      "grad_norm": 0.10049599079078834,
+      "learning_rate": 4.833377113146379e-06,
+      "loss": 0.4811,
+      "step": 717
+    },
+    {
+      "epoch": 1.7555012224938875,
+      "grad_norm": 0.1542851289951912,
+      "learning_rate": 4.832917262309781e-06,
+      "loss": 1.1234,
+      "step": 718
+    },
+    {
+      "epoch": 1.7579462102689487,
+      "grad_norm": 0.12182326363857869,
+      "learning_rate": 4.832456799728873e-06,
+      "loss": 0.5557,
+      "step": 719
+    },
+    {
+      "epoch": 1.76039119804401,
+      "grad_norm": 0.10784170758062846,
+      "learning_rate": 4.831995725524399e-06,
+      "loss": 0.5179,
+      "step": 720
+    },
+    {
+      "epoch": 1.7628361858190709,
+      "grad_norm": 0.13804408323952488,
+      "learning_rate": 4.831534039817263e-06,
+      "loss": 0.4202,
+      "step": 721
+    },
+    {
+      "epoch": 1.765281173594132,
+      "grad_norm": 0.11764823403614114,
+      "learning_rate": 4.83107174272853e-06,
+      "loss": 0.4983,
+      "step": 722
+    },
+    {
+      "epoch": 1.7677261613691932,
+      "grad_norm": 0.14996593574917372,
+      "learning_rate": 4.830608834379425e-06,
+      "loss": 0.6088,
+      "step": 723
+    },
+    {
+      "epoch": 1.7701711491442542,
+      "grad_norm": 0.1450566340673448,
+      "learning_rate": 4.8301453148913315e-06,
+      "loss": 1.0136,
+      "step": 724
+    },
+    {
+      "epoch": 1.7726161369193154,
+      "grad_norm": 0.10361854705671132,
+      "learning_rate": 4.829681184385797e-06,
+      "loss": 0.6096,
+      "step": 725
+    },
+    {
+      "epoch": 1.7750611246943766,
+      "grad_norm": 0.09918891795291579,
+      "learning_rate": 4.829216442984527e-06,
+      "loss": 0.7074,
+      "step": 726
+    },
+    {
+      "epoch": 1.7775061124694376,
+      "grad_norm": 0.0973180623640307,
+      "learning_rate": 4.828751090809386e-06,
+      "loss": 0.5953,
+      "step": 727
+    },
+    {
+      "epoch": 1.7799511002444988,
+      "grad_norm": 0.12789101585716287,
+      "learning_rate": 4.828285127982401e-06,
+      "loss": 0.7094,
+      "step": 728
+    },
+    {
+      "epoch": 1.78239608801956,
+      "grad_norm": 0.07630922705512852,
+      "learning_rate": 4.827818554625757e-06,
+      "loss": 0.5477,
+      "step": 729
+    },
+    {
+      "epoch": 1.784841075794621,
+      "grad_norm": 0.156995071857263,
+      "learning_rate": 4.827351370861802e-06,
+      "loss": 0.7341,
+      "step": 730
+    },
+    {
+      "epoch": 1.7872860635696821,
+      "grad_norm": 0.1444320392801839,
+      "learning_rate": 4.8268835768130416e-06,
+      "loss": 0.4963,
+      "step": 731
+    },
+    {
+      "epoch": 1.7897310513447433,
+      "grad_norm": 0.09681528053510027,
+      "learning_rate": 4.826415172602141e-06,
+      "loss": 0.5473,
+      "step": 732
+    },
+    {
+      "epoch": 1.7921760391198043,
+      "grad_norm": 0.12584394861018136,
+      "learning_rate": 4.8259461583519275e-06,
+      "loss": 0.8137,
+      "step": 733
+    },
+    {
+      "epoch": 1.7946210268948657,
+      "grad_norm": 0.1316575626084965,
+      "learning_rate": 4.825476534185387e-06,
+      "loss": 0.7689,
+      "step": 734
+    },
+    {
+      "epoch": 1.7970660146699267,
+      "grad_norm": 0.13746165912076075,
+      "learning_rate": 4.8250063002256665e-06,
+      "loss": 1.1722,
+      "step": 735
+    },
+    {
+      "epoch": 1.7995110024449876,
+      "grad_norm": 0.12278482239621036,
+      "learning_rate": 4.8245354565960715e-06,
+      "loss": 0.5363,
+      "step": 736
+    },
+    {
+      "epoch": 1.801955990220049,
+      "grad_norm": 0.10831116479829957,
+      "learning_rate": 4.824064003420067e-06,
+      "loss": 0.4675,
+      "step": 737
+    },
+    {
+      "epoch": 1.80440097799511,
+      "grad_norm": 0.13601052407661196,
+      "learning_rate": 4.8235919408212805e-06,
+      "loss": 0.6535,
+      "step": 738
+    },
+    {
+      "epoch": 1.8068459657701712,
+      "grad_norm": 0.1410565398203603,
+      "learning_rate": 4.823119268923496e-06,
+      "loss": 0.4656,
+      "step": 739
+    },
+    {
+      "epoch": 1.8092909535452324,
+      "grad_norm": 0.14323506634884267,
+      "learning_rate": 4.822645987850662e-06,
+      "loss": 0.4327,
+      "step": 740
+    },
+    {
+      "epoch": 1.8117359413202934,
+      "grad_norm": 0.1551502268756037,
+      "learning_rate": 4.82217209772688e-06,
+      "loss": 0.5046,
+      "step": 741
+    },
+    {
+      "epoch": 1.8141809290953546,
+      "grad_norm": 0.16024173573383538,
+      "learning_rate": 4.821697598676416e-06,
+      "loss": 0.5508,
+      "step": 742
+    },
+    {
+      "epoch": 1.8166259168704157,
+      "grad_norm": 0.12314984515616777,
+      "learning_rate": 4.821222490823695e-06,
+      "loss": 0.5043,
+      "step": 743
+    },
+    {
+      "epoch": 1.8190709046454767,
+      "grad_norm": 0.1307407136558782,
+      "learning_rate": 4.820746774293301e-06,
+      "loss": 0.9258,
+      "step": 744
+    },
+    {
+      "epoch": 1.821515892420538,
+      "grad_norm": 0.1222848554854843,
+      "learning_rate": 4.820270449209979e-06,
+      "loss": 0.6961,
+      "step": 745
+    },
+    {
+      "epoch": 1.823960880195599,
+      "grad_norm": 0.15014357077629403,
+      "learning_rate": 4.819793515698629e-06,
+      "loss": 0.6656,
+      "step": 746
+    },
+    {
+      "epoch": 1.82640586797066,
+      "grad_norm": 0.10984734986693803,
+      "learning_rate": 4.819315973884317e-06,
+      "loss": 0.5384,
+      "step": 747
+    },
+    {
+      "epoch": 1.8288508557457213,
+      "grad_norm": 0.12184988588281984,
+      "learning_rate": 4.818837823892264e-06,
+      "loss": 0.8951,
+      "step": 748
+    },
+    {
+      "epoch": 1.8312958435207825,
+      "grad_norm": 0.11079195537156254,
+      "learning_rate": 4.8183590658478516e-06,
+      "loss": 0.7624,
+      "step": 749
+    },
+    {
+      "epoch": 1.8337408312958434,
+      "grad_norm": 0.11008485332204336,
+      "learning_rate": 4.817879699876623e-06,
+      "loss": 0.5527,
+      "step": 750
+    },
+    {
+      "epoch": 1.8361858190709046,
+      "grad_norm": 0.1415032346155075,
+      "learning_rate": 4.817399726104277e-06,
+      "loss": 0.577,
+      "step": 751
+    },
+    {
+      "epoch": 1.8386308068459658,
+      "grad_norm": 0.1158999823098757,
+      "learning_rate": 4.8169191446566746e-06,
+      "loss": 0.6826,
+      "step": 752
+    },
+    {
+      "epoch": 1.8410757946210268,
+      "grad_norm": 0.09991612935598868,
+      "learning_rate": 4.816437955659835e-06,
+      "loss": 0.3888,
+      "step": 753
+    },
+    {
+      "epoch": 1.843520782396088,
+      "grad_norm": 0.14642084595183114,
+      "learning_rate": 4.815956159239937e-06,
+      "loss": 1.0093,
+      "step": 754
+    },
+    {
+      "epoch": 1.8459657701711492,
+      "grad_norm": 0.12277287073197717,
+      "learning_rate": 4.815473755523319e-06,
+      "loss": 0.4162,
+      "step": 755
+    },
+    {
+      "epoch": 1.8484107579462101,
+      "grad_norm": 0.15622600367136993,
+      "learning_rate": 4.8149907446364774e-06,
+      "loss": 0.3899,
+      "step": 756
+    },
+    {
+      "epoch": 1.8508557457212715,
+      "grad_norm": 0.15908980682987184,
+      "learning_rate": 4.81450712670607e-06,
+      "loss": 0.8689,
+      "step": 757
+    },
+    {
+      "epoch": 1.8533007334963325,
+      "grad_norm": 0.1910220045335047,
+      "learning_rate": 4.81402290185891e-06,
+      "loss": 1.0322,
+      "step": 758
+    },
+    {
+      "epoch": 1.8557457212713935,
+      "grad_norm": 0.13781731232482605,
+      "learning_rate": 4.813538070221976e-06,
+      "loss": 0.6689,
+      "step": 759
+    },
+    {
+      "epoch": 1.858190709046455,
+      "grad_norm": 0.12187660472100656,
+      "learning_rate": 4.813052631922399e-06,
+      "loss": 0.5701,
+      "step": 760
+    },
+    {
+      "epoch": 1.8606356968215159,
+      "grad_norm": 0.10704190521669231,
+      "learning_rate": 4.812566587087473e-06,
+      "loss": 0.541,
+      "step": 761
+    },
+    {
+      "epoch": 1.863080684596577,
+      "grad_norm": 0.12089889772706895,
+      "learning_rate": 4.81207993584465e-06,
+      "loss": 0.7169,
+      "step": 762
+    },
+    {
+      "epoch": 1.8655256723716382,
+      "grad_norm": 0.17754599312710506,
+      "learning_rate": 4.811592678321541e-06,
+      "loss": 0.5148,
+      "step": 763
+    },
+    {
+      "epoch": 1.8679706601466992,
+      "grad_norm": 0.10337394968121207,
+      "learning_rate": 4.811104814645916e-06,
+      "loss": 0.6708,
+      "step": 764
+    },
+    {
+      "epoch": 1.8704156479217604,
+      "grad_norm": 0.15990897703899912,
+      "learning_rate": 4.810616344945706e-06,
+      "loss": 0.9321,
+      "step": 765
+    },
+    {
+      "epoch": 1.8728606356968216,
+      "grad_norm": 0.12330790293414083,
+      "learning_rate": 4.810127269348995e-06,
+      "loss": 0.6211,
+      "step": 766
+    },
+    {
+      "epoch": 1.8753056234718826,
+      "grad_norm": 0.09210771404422552,
+      "learning_rate": 4.809637587984033e-06,
+      "loss": 0.6974,
+      "step": 767
+    },
+    {
+      "epoch": 1.8777506112469438,
+      "grad_norm": 0.168545999335004,
+      "learning_rate": 4.809147300979225e-06,
+      "loss": 1.1879,
+      "step": 768
+    },
+    {
+      "epoch": 1.880195599022005,
+      "grad_norm": 0.11482535276554853,
+      "learning_rate": 4.808656408463134e-06,
+      "loss": 0.7129,
+      "step": 769
+    },
+    {
+      "epoch": 1.882640586797066,
+      "grad_norm": 0.19798126409860214,
+      "learning_rate": 4.808164910564486e-06,
+      "loss": 0.8188,
+      "step": 770
+    },
+    {
+      "epoch": 1.885085574572127,
+      "grad_norm": 0.15737377378099945,
+      "learning_rate": 4.80767280741216e-06,
+      "loss": 0.729,
+      "step": 771
+    },
+    {
+      "epoch": 1.8875305623471883,
+      "grad_norm": 0.10255358358507011,
+      "learning_rate": 4.807180099135199e-06,
+      "loss": 0.5795,
+      "step": 772
+    },
+    {
+      "epoch": 1.8899755501222493,
+      "grad_norm": 0.12399009514435624,
+      "learning_rate": 4.806686785862801e-06,
+      "loss": 0.5496,
+      "step": 773
+    },
+    {
+      "epoch": 1.8924205378973105,
+      "grad_norm": 0.12451001518390688,
+      "learning_rate": 4.8061928677243244e-06,
+      "loss": 0.8979,
+      "step": 774
+    },
+    {
+      "epoch": 1.8948655256723717,
+      "grad_norm": 0.16118155814169954,
+      "learning_rate": 4.805698344849286e-06,
+      "loss": 0.8007,
+      "step": 775
+    },
+    {
+      "epoch": 1.8973105134474326,
+      "grad_norm": 0.11970416495016106,
+      "learning_rate": 4.805203217367362e-06,
+      "loss": 0.6259,
+      "step": 776
+    },
+    {
+      "epoch": 1.899755501222494,
+      "grad_norm": 0.15685960511820304,
+      "learning_rate": 4.804707485408384e-06,
+      "loss": 0.8075,
+      "step": 777
+    },
+    {
+      "epoch": 1.902200488997555,
+      "grad_norm": 0.1578579719899877,
+      "learning_rate": 4.8042111491023455e-06,
+      "loss": 0.4877,
+      "step": 778
+    },
+    {
+      "epoch": 1.904645476772616,
+      "grad_norm": 0.11629835768054726,
+      "learning_rate": 4.803714208579398e-06,
+      "loss": 0.5222,
+      "step": 779
+    },
+    {
+      "epoch": 1.9070904645476774,
+      "grad_norm": 0.11701472311495625,
+      "learning_rate": 4.803216663969849e-06,
+      "loss": 0.8609,
+      "step": 780
+    },
+    {
+      "epoch": 1.9095354523227384,
+      "grad_norm": 0.10103024460163296,
+      "learning_rate": 4.802718515404168e-06,
+      "loss": 0.5263,
+      "step": 781
+    },
+    {
+      "epoch": 1.9119804400977995,
+      "grad_norm": 0.1518993777182326,
+      "learning_rate": 4.802219763012978e-06,
+      "loss": 0.6717,
+      "step": 782
+    },
+    {
+      "epoch": 1.9144254278728607,
+      "grad_norm": 0.13697745718100693,
+      "learning_rate": 4.801720406927067e-06,
+      "loss": 0.5797,
+      "step": 783
+    },
+    {
+      "epoch": 1.9168704156479217,
+      "grad_norm": 0.1336832074610851,
+      "learning_rate": 4.8012204472773745e-06,
+      "loss": 0.607,
+      "step": 784
+    },
+    {
+      "epoch": 1.919315403422983,
+      "grad_norm": 0.1286659830336251,
+      "learning_rate": 4.800719884195003e-06,
+      "loss": 0.3785,
+      "step": 785
+    },
+    {
+      "epoch": 1.921760391198044,
+      "grad_norm": 0.13505916681429353,
+      "learning_rate": 4.800218717811212e-06,
+      "loss": 0.4645,
+      "step": 786
+    },
+    {
+      "epoch": 1.924205378973105,
+      "grad_norm": 0.11679295475938653,
+      "learning_rate": 4.799716948257418e-06,
+      "loss": 0.7331,
+      "step": 787
+    },
+    {
+      "epoch": 1.9266503667481663,
+      "grad_norm": 0.10508732261419586,
+      "learning_rate": 4.799214575665196e-06,
+      "loss": 0.5308,
+      "step": 788
+    },
+    {
+      "epoch": 1.9290953545232274,
+      "grad_norm": 0.12161540145949726,
+      "learning_rate": 4.79871160016628e-06,
+      "loss": 0.8243,
+      "step": 789
+    },
+    {
+      "epoch": 1.9315403422982884,
+      "grad_norm": 0.13688323244743186,
+      "learning_rate": 4.798208021892562e-06,
+      "loss": 0.6833,
+      "step": 790
+    },
+    {
+      "epoch": 1.9339853300733496,
+      "grad_norm": 0.10398439966137216,
+      "learning_rate": 4.797703840976091e-06,
+      "loss": 0.5062,
+      "step": 791
+    },
+    {
+      "epoch": 1.9364303178484108,
+      "grad_norm": 0.1408936864234124,
+      "learning_rate": 4.797199057549076e-06,
+      "loss": 0.5466,
+      "step": 792
+    },
+    {
+      "epoch": 1.9388753056234718,
+      "grad_norm": 0.11584925478575149,
+      "learning_rate": 4.796693671743882e-06,
+      "loss": 0.4612,
+      "step": 793
+    },
+    {
+      "epoch": 1.941320293398533,
+      "grad_norm": 0.08981749019854286,
+      "learning_rate": 4.7961876836930324e-06,
+      "loss": 0.4266,
+      "step": 794
+    },
+    {
+      "epoch": 1.9437652811735942,
+      "grad_norm": 0.18994463221787466,
+      "learning_rate": 4.79568109352921e-06,
+      "loss": 0.9827,
+      "step": 795
+    },
+    {
+      "epoch": 1.9462102689486551,
+      "grad_norm": 0.16703185893755695,
+      "learning_rate": 4.795173901385253e-06,
+      "loss": 0.4553,
+      "step": 796
+    },
+    {
+      "epoch": 1.9486552567237165,
+      "grad_norm": 0.10413546765513156,
+      "learning_rate": 4.794666107394159e-06,
+      "loss": 0.6008,
+      "step": 797
+    },
+    {
+      "epoch": 1.9511002444987775,
+      "grad_norm": 0.14401763284737917,
+      "learning_rate": 4.794157711689084e-06,
+      "loss": 0.6441,
+      "step": 798
+    },
+    {
+      "epoch": 1.9535452322738385,
+      "grad_norm": 0.14523066687569391,
+      "learning_rate": 4.7936487144033415e-06,
+      "loss": 0.5983,
+      "step": 799
+    },
+    {
+      "epoch": 1.9559902200488999,
+      "grad_norm": 0.11476848790727802,
+      "learning_rate": 4.7931391156704e-06,
+      "loss": 0.849,
+      "step": 800
+    },
+    {
+      "epoch": 1.9584352078239609,
+      "grad_norm": 0.13416802946301673,
+      "learning_rate": 4.792628915623889e-06,
+      "loss": 0.8281,
+      "step": 801
+    },
+    {
+      "epoch": 1.960880195599022,
+      "grad_norm": 0.14401738017231736,
+      "learning_rate": 4.792118114397597e-06,
+      "loss": 0.6789,
+      "step": 802
+    },
+    {
+      "epoch": 1.9633251833740832,
+      "grad_norm": 0.11154655375586213,
+      "learning_rate": 4.791606712125464e-06,
+      "loss": 0.5504,
+      "step": 803
+    },
+    {
+      "epoch": 1.9657701711491442,
+      "grad_norm": 0.17603785657545673,
+      "learning_rate": 4.791094708941594e-06,
+      "loss": 0.5241,
+      "step": 804
+    },
+    {
+      "epoch": 1.9682151589242054,
+      "grad_norm": 0.10288638917315537,
+      "learning_rate": 4.790582104980244e-06,
+      "loss": 0.4565,
+      "step": 805
+    },
+    {
+      "epoch": 1.9706601466992666,
+      "grad_norm": 0.11130311416716382,
+      "learning_rate": 4.790068900375831e-06,
+      "loss": 0.8371,
+      "step": 806
+    },
+    {
+      "epoch": 1.9731051344743276,
+      "grad_norm": 0.12434686315986627,
+      "learning_rate": 4.789555095262932e-06,
+      "loss": 0.5618,
+      "step": 807
+    },
+    {
+      "epoch": 1.9755501222493888,
+      "grad_norm": 0.1265132607261373,
+      "learning_rate": 4.789040689776274e-06,
+      "loss": 0.6798,
+      "step": 808
+    },
+    {
+      "epoch": 1.97799511002445,
+      "grad_norm": 0.1792243095832928,
+      "learning_rate": 4.788525684050749e-06,
+      "loss": 0.6625,
+      "step": 809
+    },
+    {
+      "epoch": 1.980440097799511,
+      "grad_norm": 0.11837845869571981,
+      "learning_rate": 4.788010078221401e-06,
+      "loss": 0.5964,
+      "step": 810
+    },
+    {
+      "epoch": 1.982885085574572,
+      "grad_norm": 0.14304280258710575,
+      "learning_rate": 4.787493872423435e-06,
+      "loss": 0.8614,
+      "step": 811
+    },
+    {
+      "epoch": 1.9853300733496333,
+      "grad_norm": 0.11049639213632467,
+      "learning_rate": 4.786977066792212e-06,
+      "loss": 0.4396,
+      "step": 812
+    },
+    {
+      "epoch": 1.9877750611246943,
+      "grad_norm": 0.15406352563106712,
+      "learning_rate": 4.786459661463251e-06,
+      "loss": 0.9546,
+      "step": 813
+    },
+    {
+      "epoch": 1.9902200488997555,
+      "grad_norm": 0.15202454763031117,
+      "learning_rate": 4.785941656572225e-06,
+      "loss": 1.1633,
+      "step": 814
+    },
+    {
+      "epoch": 1.9926650366748166,
+      "grad_norm": 0.15407890432406673,
+      "learning_rate": 4.78542305225497e-06,
+      "loss": 0.6896,
+      "step": 815
+    },
+    {
+      "epoch": 1.9951100244498776,
+      "grad_norm": 0.13348299433406074,
+      "learning_rate": 4.784903848647473e-06,
+      "loss": 0.6009,
+      "step": 816
+    },
+    {
+      "epoch": 1.997555012224939,
+      "grad_norm": 0.10584937182336218,
+      "learning_rate": 4.784384045885883e-06,
+      "loss": 0.4955,
+      "step": 817
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.147434937419781,
+      "learning_rate": 4.783863644106502e-06,
+      "loss": 0.879,
+      "step": 818
+    },
+    {
+      "epoch": 2.002444987775061,
+      "grad_norm": 0.14706158144953857,
+      "learning_rate": 4.7833426434457945e-06,
+      "loss": 0.5248,
+      "step": 819
+    },
+    {
+      "epoch": 2.0048899755501224,
+      "grad_norm": 0.15922707965360527,
+      "learning_rate": 4.782821044040376e-06,
+      "loss": 0.4911,
+      "step": 820
+    },
+    {
+      "epoch": 2.0073349633251834,
+      "grad_norm": 0.11953551149768692,
+      "learning_rate": 4.782298846027024e-06,
+      "loss": 0.8306,
+      "step": 821
+    },
+    {
+      "epoch": 2.0097799511002443,
+      "grad_norm": 0.10233519689936803,
+      "learning_rate": 4.781776049542669e-06,
+      "loss": 0.5911,
+      "step": 822
+    },
+    {
+      "epoch": 2.0122249388753057,
+      "grad_norm": 0.20332745990141157,
+      "learning_rate": 4.781252654724401e-06,
+      "loss": 0.5306,
+      "step": 823
+    },
+    {
+      "epoch": 2.0146699266503667,
+      "grad_norm": 0.07298230084059151,
+      "learning_rate": 4.780728661709465e-06,
+      "loss": 0.3727,
+      "step": 824
+    },
+    {
+      "epoch": 2.0171149144254277,
+      "grad_norm": 0.19146845969916104,
+      "learning_rate": 4.7802040706352655e-06,
+      "loss": 0.6415,
+      "step": 825
+    },
+    {
+      "epoch": 2.019559902200489,
+      "grad_norm": 0.17102622997073183,
+      "learning_rate": 4.779678881639363e-06,
+      "loss": 0.7184,
+      "step": 826
+    },
+    {
+      "epoch": 2.02200488997555,
+      "grad_norm": 0.13068934102332075,
+      "learning_rate": 4.779153094859471e-06,
+      "loss": 0.5366,
+      "step": 827
+    },
+    {
+      "epoch": 2.0244498777506115,
+      "grad_norm": 0.16215713018093997,
+      "learning_rate": 4.778626710433465e-06,
+      "loss": 0.8962,
+      "step": 828
+    },
+    {
+      "epoch": 2.0268948655256724,
+      "grad_norm": 0.13228487122095742,
+      "learning_rate": 4.778099728499375e-06,
+      "loss": 0.5542,
+      "step": 829
+    },
+    {
+      "epoch": 2.0293398533007334,
+      "grad_norm": 0.12567736735023852,
+      "learning_rate": 4.777572149195387e-06,
+      "loss": 0.8489,
+      "step": 830
+    },
+    {
+      "epoch": 2.031784841075795,
+      "grad_norm": 0.1278871828157041,
+      "learning_rate": 4.777043972659845e-06,
+      "loss": 0.7214,
+      "step": 831
+    },
+    {
+      "epoch": 2.034229828850856,
+      "grad_norm": 0.12969716220662594,
+      "learning_rate": 4.776515199031248e-06,
+      "loss": 0.8372,
+      "step": 832
+    },
+    {
+      "epoch": 2.0366748166259168,
+      "grad_norm": 0.12442518835234063,
+      "learning_rate": 4.7759858284482544e-06,
+      "loss": 0.7463,
+      "step": 833
+    },
+    {
+      "epoch": 2.039119804400978,
+      "grad_norm": 0.12009626731653386,
+      "learning_rate": 4.775455861049676e-06,
+      "loss": 0.5872,
+      "step": 834
+    },
+    {
+      "epoch": 2.041564792176039,
+      "grad_norm": 0.1029923305504545,
+      "learning_rate": 4.774925296974482e-06,
+      "loss": 0.4112,
+      "step": 835
+    },
+    {
+      "epoch": 2.0440097799511,
+      "grad_norm": 0.19129615884375242,
+      "learning_rate": 4.774394136361799e-06,
+      "loss": 0.6626,
+      "step": 836
+    },
+    {
+      "epoch": 2.0464547677261615,
+      "grad_norm": 0.1230039057494566,
+      "learning_rate": 4.77386237935091e-06,
+      "loss": 0.8634,
+      "step": 837
+    },
+    {
+      "epoch": 2.0488997555012225,
+      "grad_norm": 0.12053378320852179,
+      "learning_rate": 4.7733300260812525e-06,
+      "loss": 1.0506,
+      "step": 838
+    },
+    {
+      "epoch": 2.0513447432762835,
+      "grad_norm": 0.17000545122958616,
+      "learning_rate": 4.7727970766924234e-06,
+      "loss": 1.1707,
+      "step": 839
+    },
+    {
+      "epoch": 2.053789731051345,
+      "grad_norm": 0.13108008962080583,
+      "learning_rate": 4.772263531324173e-06,
+      "loss": 0.5354,
+      "step": 840
+    },
+    {
+      "epoch": 2.056234718826406,
+      "grad_norm": 0.14682324754138906,
+      "learning_rate": 4.771729390116409e-06,
+      "loss": 0.8349,
+      "step": 841
+    },
+    {
+      "epoch": 2.058679706601467,
+      "grad_norm": 0.10463182967955202,
+      "learning_rate": 4.771194653209197e-06,
+      "loss": 0.4461,
+      "step": 842
+    },
+    {
+      "epoch": 2.0611246943765282,
+      "grad_norm": 0.14846004991553632,
+      "learning_rate": 4.770659320742755e-06,
+      "loss": 0.548,
+      "step": 843
+    },
+    {
+      "epoch": 2.063569682151589,
+      "grad_norm": 0.17413156200466165,
+      "learning_rate": 4.770123392857461e-06,
+      "loss": 0.5997,
+      "step": 844
+    },
+    {
+      "epoch": 2.06601466992665,
+      "grad_norm": 0.10301084940909348,
+      "learning_rate": 4.769586869693848e-06,
+      "loss": 0.5678,
+      "step": 845
+    },
+    {
+      "epoch": 2.0684596577017116,
+      "grad_norm": 0.15027344077997845,
+      "learning_rate": 4.7690497513926025e-06,
+      "loss": 0.4973,
+      "step": 846
+    },
+    {
+      "epoch": 2.0709046454767726,
+      "grad_norm": 0.14456329482622987,
+      "learning_rate": 4.768512038094571e-06,
+      "loss": 0.8347,
+      "step": 847
+    },
+    {
+      "epoch": 2.0733496332518335,
+      "grad_norm": 0.09476506032195829,
+      "learning_rate": 4.767973729940754e-06,
+      "loss": 0.5279,
+      "step": 848
+    },
+    {
+      "epoch": 2.075794621026895,
+      "grad_norm": 0.1281703164245905,
+      "learning_rate": 4.767434827072308e-06,
+      "loss": 0.6836,
+      "step": 849
+    },
+    {
+      "epoch": 2.078239608801956,
+      "grad_norm": 0.11499591039702096,
+      "learning_rate": 4.766895329630545e-06,
+      "loss": 0.5449,
+      "step": 850
+    },
+    {
+      "epoch": 2.0806845965770173,
+      "grad_norm": 0.12725902925028126,
+      "learning_rate": 4.766355237756936e-06,
+      "loss": 0.5523,
+      "step": 851
+    },
+    {
+      "epoch": 2.0831295843520783,
+      "grad_norm": 0.11967575642898146,
+      "learning_rate": 4.765814551593102e-06,
+      "loss": 0.5467,
+      "step": 852
+    },
+    {
+      "epoch": 2.0855745721271393,
+      "grad_norm": 0.13449356042079105,
+      "learning_rate": 4.765273271280825e-06,
+      "loss": 0.7627,
+      "step": 853
+    },
+    {
+      "epoch": 2.0880195599022007,
+      "grad_norm": 0.1545773457575999,
+      "learning_rate": 4.7647313969620425e-06,
+      "loss": 1.1005,
+      "step": 854
+    },
+    {
+      "epoch": 2.0904645476772616,
+      "grad_norm": 0.13926844739111374,
+      "learning_rate": 4.764188928778843e-06,
+      "loss": 0.6687,
+      "step": 855
+    },
+    {
+      "epoch": 2.0929095354523226,
+      "grad_norm": 0.13175102178710443,
+      "learning_rate": 4.763645866873478e-06,
+      "loss": 0.7133,
+      "step": 856
+    },
+    {
+      "epoch": 2.095354523227384,
+      "grad_norm": 0.14983695092211466,
+      "learning_rate": 4.7631022113883475e-06,
+      "loss": 0.376,
+      "step": 857
+    },
+    {
+      "epoch": 2.097799511002445,
+      "grad_norm": 0.15965998543574816,
+      "learning_rate": 4.762557962466011e-06,
+      "loss": 0.7313,
+      "step": 858
+    },
+    {
+      "epoch": 2.100244498777506,
+      "grad_norm": 0.11394697192652084,
+      "learning_rate": 4.762013120249185e-06,
+      "loss": 0.7536,
+      "step": 859
+    },
+    {
+      "epoch": 2.1026894865525674,
+      "grad_norm": 0.13115564134488228,
+      "learning_rate": 4.761467684880738e-06,
+      "loss": 0.4737,
+      "step": 860
+    },
+    {
+      "epoch": 2.1051344743276283,
+      "grad_norm": 0.13628436164284463,
+      "learning_rate": 4.760921656503696e-06,
+      "loss": 0.8629,
+      "step": 861
+    },
+    {
+      "epoch": 2.1075794621026893,
+      "grad_norm": 0.16944637192952391,
+      "learning_rate": 4.76037503526124e-06,
+      "loss": 0.4788,
+      "step": 862
+    },
+    {
+      "epoch": 2.1100244498777507,
+      "grad_norm": 0.17113161308688937,
+      "learning_rate": 4.759827821296706e-06,
+      "loss": 0.6829,
+      "step": 863
+    },
+    {
+      "epoch": 2.1124694376528117,
+      "grad_norm": 0.12499755808514822,
+      "learning_rate": 4.7592800147535875e-06,
+      "loss": 0.7053,
+      "step": 864
+    },
+    {
+      "epoch": 2.1149144254278727,
+      "grad_norm": 0.1958786675928173,
+      "learning_rate": 4.7587316157755305e-06,
+      "loss": 1.2257,
+      "step": 865
+    },
+    {
+      "epoch": 2.117359413202934,
+      "grad_norm": 0.17356508282630306,
+      "learning_rate": 4.758182624506339e-06,
+      "loss": 0.459,
+      "step": 866
+    },
+    {
+      "epoch": 2.119804400977995,
+      "grad_norm": 0.09330789803328009,
+      "learning_rate": 4.757633041089971e-06,
+      "loss": 0.4654,
+      "step": 867
+    },
+    {
+      "epoch": 2.122249388753056,
+      "grad_norm": 0.16023903984593896,
+      "learning_rate": 4.757082865670538e-06,
+      "loss": 0.5512,
+      "step": 868
+    },
+    {
+      "epoch": 2.1246943765281174,
+      "grad_norm": 0.18351489392869802,
+      "learning_rate": 4.75653209839231e-06,
+      "loss": 0.5189,
+      "step": 869
+    },
+    {
+      "epoch": 2.1271393643031784,
+      "grad_norm": 0.15640933622450706,
+      "learning_rate": 4.755980739399712e-06,
+      "loss": 0.3466,
+      "step": 870
+    },
+    {
+      "epoch": 2.12958435207824,
+      "grad_norm": 0.1444257407646091,
+      "learning_rate": 4.755428788837321e-06,
+      "loss": 0.7519,
+      "step": 871
+    },
+    {
+      "epoch": 2.132029339853301,
+      "grad_norm": 0.14595829245939346,
+      "learning_rate": 4.754876246849871e-06,
+      "loss": 0.4077,
+      "step": 872
+    },
+    {
+      "epoch": 2.1344743276283618,
+      "grad_norm": 0.13451553604309138,
+      "learning_rate": 4.754323113582253e-06,
+      "loss": 0.8014,
+      "step": 873
+    },
+    {
+      "epoch": 2.136919315403423,
+      "grad_norm": 0.1657308840208002,
+      "learning_rate": 4.753769389179509e-06,
+      "loss": 0.5881,
+      "step": 874
+    },
+    {
+      "epoch": 2.139364303178484,
+      "grad_norm": 0.11880111012595808,
+      "learning_rate": 4.75321507378684e-06,
+      "loss": 0.7447,
+      "step": 875
+    },
+    {
+      "epoch": 2.141809290953545,
+      "grad_norm": 0.10161063275077391,
+      "learning_rate": 4.752660167549599e-06,
+      "loss": 0.4295,
+      "step": 876
+    },
+    {
+      "epoch": 2.1442542787286065,
+      "grad_norm": 0.14444496812933494,
+      "learning_rate": 4.752104670613296e-06,
+      "loss": 0.3627,
+      "step": 877
+    },
+    {
+      "epoch": 2.1466992665036675,
+      "grad_norm": 0.10630406958696703,
+      "learning_rate": 4.751548583123595e-06,
+      "loss": 0.4577,
+      "step": 878
+    },
+    {
+      "epoch": 2.1491442542787285,
+      "grad_norm": 0.09966748882034411,
+      "learning_rate": 4.750991905226313e-06,
+      "loss": 0.652,
+      "step": 879
+    },
+    {
+      "epoch": 2.15158924205379,
+      "grad_norm": 0.1293482303798882,
+      "learning_rate": 4.750434637067427e-06,
+      "loss": 0.9405,
+      "step": 880
+    },
+    {
+      "epoch": 2.154034229828851,
+      "grad_norm": 0.1529528807983986,
+      "learning_rate": 4.749876778793062e-06,
+      "loss": 0.4113,
+      "step": 881
+    },
+    {
+      "epoch": 2.156479217603912,
+      "grad_norm": 0.21978250281734155,
+      "learning_rate": 4.749318330549504e-06,
+      "loss": 1.009,
+      "step": 882
+    },
+    {
+      "epoch": 2.1589242053789732,
+      "grad_norm": 0.24391905542317935,
+      "learning_rate": 4.748759292483189e-06,
+      "loss": 0.775,
+      "step": 883
+    },
+    {
+      "epoch": 2.161369193154034,
+      "grad_norm": 0.1337623060502161,
+      "learning_rate": 4.748199664740711e-06,
+      "loss": 0.4591,
+      "step": 884
+    },
+    {
+      "epoch": 2.163814180929095,
+      "grad_norm": 0.1581243389135384,
+      "learning_rate": 4.747639447468816e-06,
+      "loss": 0.7211,
+      "step": 885
+    },
+    {
+      "epoch": 2.1662591687041566,
+      "grad_norm": 0.16282780623445856,
+      "learning_rate": 4.747078640814407e-06,
+      "loss": 0.4574,
+      "step": 886
+    },
+    {
+      "epoch": 2.1687041564792175,
+      "grad_norm": 0.11872462173956766,
+      "learning_rate": 4.7465172449245395e-06,
+      "loss": 0.4549,
+      "step": 887
+    },
+    {
+      "epoch": 2.1711491442542785,
+      "grad_norm": 0.1499580439233291,
+      "learning_rate": 4.745955259946424e-06,
+      "loss": 0.9132,
+      "step": 888
+    },
+    {
+      "epoch": 2.17359413202934,
+      "grad_norm": 0.12523862110866715,
+      "learning_rate": 4.745392686027427e-06,
+      "loss": 0.5837,
+      "step": 889
+    },
+    {
+      "epoch": 2.176039119804401,
+      "grad_norm": 0.1787126704314975,
+      "learning_rate": 4.744829523315067e-06,
+      "loss": 0.623,
+      "step": 890
+    },
+    {
+      "epoch": 2.178484107579462,
+      "grad_norm": 0.1568240826353576,
+      "learning_rate": 4.744265771957019e-06,
+      "loss": 0.5474,
+      "step": 891
+    },
+    {
+      "epoch": 2.1809290953545233,
+      "grad_norm": 0.18830288131446615,
+      "learning_rate": 4.743701432101111e-06,
+      "loss": 0.8836,
+      "step": 892
+    },
+    {
+      "epoch": 2.1833740831295843,
+      "grad_norm": 0.14104979392400585,
+      "learning_rate": 4.743136503895327e-06,
+      "loss": 0.6751,
+      "step": 893
+    },
+    {
+      "epoch": 2.1858190709046457,
+      "grad_norm": 0.17488735388963428,
+      "learning_rate": 4.742570987487802e-06,
+      "loss": 0.5186,
+      "step": 894
+    },
+    {
+      "epoch": 2.1882640586797066,
+      "grad_norm": 0.12165539735442335,
+      "learning_rate": 4.7420048830268295e-06,
+      "loss": 0.7064,
+      "step": 895
+    },
+    {
+      "epoch": 2.1907090464547676,
+      "grad_norm": 0.16492081834271125,
+      "learning_rate": 4.741438190660854e-06,
+      "loss": 0.7279,
+      "step": 896
+    },
+    {
+      "epoch": 2.193154034229829,
+      "grad_norm": 0.16121870729834303,
+      "learning_rate": 4.740870910538475e-06,
+      "loss": 0.803,
+      "step": 897
+    },
+    {
+      "epoch": 2.19559902200489,
+      "grad_norm": 0.1620792840653181,
+      "learning_rate": 4.740303042808447e-06,
+      "loss": 0.5295,
+      "step": 898
+    },
+    {
+      "epoch": 2.198044009779951,
+      "grad_norm": 0.11649367121019195,
+      "learning_rate": 4.739734587619677e-06,
+      "loss": 0.3964,
+      "step": 899
+    },
+    {
+      "epoch": 2.2004889975550124,
+      "grad_norm": 0.13711991196472734,
+      "learning_rate": 4.739165545121228e-06,
+      "loss": 0.5327,
+      "step": 900
+    },
+    {
+      "epoch": 2.2029339853300733,
+      "grad_norm": 0.08865038921741594,
+      "learning_rate": 4.738595915462316e-06,
+      "loss": 0.4779,
+      "step": 901
+    },
+    {
+      "epoch": 2.2053789731051343,
+      "grad_norm": 0.12785264003277,
+      "learning_rate": 4.7380256987923104e-06,
+      "loss": 0.5592,
+      "step": 902
+    },
+    {
+      "epoch": 2.2078239608801957,
+      "grad_norm": 0.1294170638533336,
+      "learning_rate": 4.737454895260734e-06,
+      "loss": 0.5868,
+      "step": 903
+    },
+    {
+      "epoch": 2.2102689486552567,
+      "grad_norm": 0.11187402251584612,
+      "learning_rate": 4.736883505017267e-06,
+      "loss": 0.776,
+      "step": 904
+    },
+    {
+      "epoch": 2.2127139364303177,
+      "grad_norm": 0.10857150942150316,
+      "learning_rate": 4.736311528211738e-06,
+      "loss": 0.4057,
+      "step": 905
+    },
+    {
+      "epoch": 2.215158924205379,
+      "grad_norm": 0.1819830411382696,
+      "learning_rate": 4.735738964994135e-06,
+      "loss": 0.4603,
+      "step": 906
+    },
+    {
+      "epoch": 2.21760391198044,
+      "grad_norm": 0.1769365870947913,
+      "learning_rate": 4.735165815514596e-06,
+      "loss": 1.2002,
+      "step": 907
+    },
+    {
+      "epoch": 2.220048899755501,
+      "grad_norm": 0.10794380648866674,
+      "learning_rate": 4.734592079923414e-06,
+      "loss": 0.7213,
+      "step": 908
+    },
+    {
+      "epoch": 2.2224938875305624,
+      "grad_norm": 0.15834685157815057,
+      "learning_rate": 4.734017758371037e-06,
+      "loss": 0.439,
+      "step": 909
+    },
+    {
+      "epoch": 2.2249388753056234,
+      "grad_norm": 0.15262814891961574,
+      "learning_rate": 4.733442851008063e-06,
+      "loss": 0.6084,
+      "step": 910
+    },
+    {
+      "epoch": 2.227383863080685,
+      "grad_norm": 0.15008262053687232,
+      "learning_rate": 4.732867357985248e-06,
+      "loss": 0.3913,
+      "step": 911
+    },
+    {
+      "epoch": 2.229828850855746,
+      "grad_norm": 0.13663814374154887,
+      "learning_rate": 4.732291279453497e-06,
+      "loss": 0.7689,
+      "step": 912
+    },
+    {
+      "epoch": 2.2322738386308068,
+      "grad_norm": 0.21862191649934534,
+      "learning_rate": 4.7317146155638725e-06,
+      "loss": 0.4013,
+      "step": 913
+    },
+    {
+      "epoch": 2.234718826405868,
+      "grad_norm": 0.11337906385319835,
+      "learning_rate": 4.73113736646759e-06,
+      "loss": 0.4518,
+      "step": 914
+    },
+    {
+      "epoch": 2.237163814180929,
+      "grad_norm": 0.12287967933339534,
+      "learning_rate": 4.730559532316015e-06,
+      "loss": 0.6515,
+      "step": 915
+    },
+    {
+      "epoch": 2.23960880195599,
+      "grad_norm": 0.20033366922416965,
+      "learning_rate": 4.72998111326067e-06,
+      "loss": 0.6552,
+      "step": 916
+    },
+    {
+      "epoch": 2.2420537897310515,
+      "grad_norm": 0.15508185938460795,
+      "learning_rate": 4.72940210945323e-06,
+      "loss": 0.4936,
+      "step": 917
+    },
+    {
+      "epoch": 2.2444987775061125,
+      "grad_norm": 0.11590757225801478,
+      "learning_rate": 4.728822521045523e-06,
+      "loss": 0.599,
+      "step": 918
+    },
+    {
+      "epoch": 2.2469437652811735,
+      "grad_norm": 0.11662840836827894,
+      "learning_rate": 4.728242348189531e-06,
+      "loss": 0.5455,
+      "step": 919
+    },
+    {
+      "epoch": 2.249388753056235,
+      "grad_norm": 0.1226509694511489,
+      "learning_rate": 4.727661591037387e-06,
+      "loss": 0.3578,
+      "step": 920
+    },
+    {
+      "epoch": 2.251833740831296,
+      "grad_norm": 0.12986041073194354,
+      "learning_rate": 4.727080249741379e-06,
+      "loss": 0.5854,
+      "step": 921
+    },
+    {
+      "epoch": 2.254278728606357,
+      "grad_norm": 0.1450672009254647,
+      "learning_rate": 4.726498324453949e-06,
+      "loss": 0.4936,
+      "step": 922
+    },
+    {
+      "epoch": 2.256723716381418,
+      "grad_norm": 0.1458859513236837,
+      "learning_rate": 4.725915815327692e-06,
+      "loss": 0.4962,
+      "step": 923
+    },
+    {
+      "epoch": 2.259168704156479,
+      "grad_norm": 0.11725198765875824,
+      "learning_rate": 4.725332722515352e-06,
+      "loss": 0.408,
+      "step": 924
+    },
+    {
+      "epoch": 2.26161369193154,
+      "grad_norm": 0.13243842444222276,
+      "learning_rate": 4.724749046169833e-06,
+      "loss": 0.7271,
+      "step": 925
+    },
+    {
+      "epoch": 2.2640586797066016,
+      "grad_norm": 0.13751507762577303,
+      "learning_rate": 4.724164786444185e-06,
+      "loss": 0.6217,
+      "step": 926
+    },
+    {
+      "epoch": 2.2665036674816625,
+      "grad_norm": 0.14330890183748815,
+      "learning_rate": 4.723579943491618e-06,
+      "loss": 0.706,
+      "step": 927
+    },
+    {
+      "epoch": 2.2689486552567235,
+      "grad_norm": 0.14210027436540335,
+      "learning_rate": 4.722994517465487e-06,
+      "loss": 0.7109,
+      "step": 928
+    },
+    {
+      "epoch": 2.271393643031785,
+      "grad_norm": 0.21532180195609618,
+      "learning_rate": 4.722408508519306e-06,
+      "loss": 0.6747,
+      "step": 929
+    },
+    {
+      "epoch": 2.273838630806846,
+      "grad_norm": 0.16318400866093985,
+      "learning_rate": 4.721821916806741e-06,
+      "loss": 0.7447,
+      "step": 930
+    },
+    {
+      "epoch": 2.276283618581907,
+      "grad_norm": 0.12780152224285085,
+      "learning_rate": 4.721234742481609e-06,
+      "loss": 0.4632,
+      "step": 931
+    },
+    {
+      "epoch": 2.2787286063569683,
+      "grad_norm": 0.1792340002645962,
+      "learning_rate": 4.72064698569788e-06,
+      "loss": 1.0131,
+      "step": 932
+    },
+    {
+      "epoch": 2.2811735941320292,
+      "grad_norm": 0.19779190322598758,
+      "learning_rate": 4.720058646609677e-06,
+      "loss": 0.5975,
+      "step": 933
+    },
+    {
+      "epoch": 2.28361858190709,
+      "grad_norm": 0.22120562477086841,
+      "learning_rate": 4.719469725371279e-06,
+      "loss": 0.7179,
+      "step": 934
+    },
+    {
+      "epoch": 2.2860635696821516,
+      "grad_norm": 0.12942999875436037,
+      "learning_rate": 4.71888022213711e-06,
+      "loss": 0.8739,
+      "step": 935
+    },
+    {
+      "epoch": 2.2885085574572126,
+      "grad_norm": 0.175888536192299,
+      "learning_rate": 4.718290137061755e-06,
+      "loss": 0.5433,
+      "step": 936
+    },
+    {
+      "epoch": 2.290953545232274,
+      "grad_norm": 0.1510107746602996,
+      "learning_rate": 4.717699470299946e-06,
+      "loss": 1.2176,
+      "step": 937
+    },
+    {
+      "epoch": 2.293398533007335,
+      "grad_norm": 0.11915330226089885,
+      "learning_rate": 4.717108222006569e-06,
+      "loss": 0.5377,
+      "step": 938
+    },
+    {
+      "epoch": 2.295843520782396,
+      "grad_norm": 0.13364831604716457,
+      "learning_rate": 4.716516392336664e-06,
+      "loss": 0.7308,
+      "step": 939
+    },
+    {
+      "epoch": 2.2982885085574574,
+      "grad_norm": 0.17623058189430155,
+      "learning_rate": 4.715923981445421e-06,
+      "loss": 1.1305,
+      "step": 940
+    },
+    {
+      "epoch": 2.3007334963325183,
+      "grad_norm": 0.15473015727939723,
+      "learning_rate": 4.715330989488186e-06,
+      "loss": 0.575,
+      "step": 941
+    },
+    {
+      "epoch": 2.3031784841075793,
+      "grad_norm": 0.12013438697202018,
+      "learning_rate": 4.714737416620453e-06,
+      "loss": 0.5961,
+      "step": 942
+    },
+    {
+      "epoch": 2.3056234718826407,
+      "grad_norm": 0.16772194657296288,
+      "learning_rate": 4.714143262997871e-06,
+      "loss": 0.823,
+      "step": 943
+    },
+    {
+      "epoch": 2.3080684596577017,
+      "grad_norm": 0.15155374028723395,
+      "learning_rate": 4.71354852877624e-06,
+      "loss": 0.4549,
+      "step": 944
+    },
+    {
+      "epoch": 2.3105134474327627,
+      "grad_norm": 0.17999504929451104,
+      "learning_rate": 4.712953214111516e-06,
+      "loss": 0.7118,
+      "step": 945
+    },
+    {
+      "epoch": 2.312958435207824,
+      "grad_norm": 0.13059050021859978,
+      "learning_rate": 4.712357319159799e-06,
+      "loss": 0.5235,
+      "step": 946
+    },
+    {
+      "epoch": 2.315403422982885,
+      "grad_norm": 0.09224310871861825,
+      "learning_rate": 4.7117608440773496e-06,
+      "loss": 0.399,
+      "step": 947
+    },
+    {
+      "epoch": 2.3178484107579465,
+      "grad_norm": 0.11596373574185828,
+      "learning_rate": 4.7111637890205775e-06,
+      "loss": 0.5031,
+      "step": 948
+    },
+    {
+      "epoch": 2.3202933985330074,
+      "grad_norm": 0.11438015839359254,
+      "learning_rate": 4.710566154146044e-06,
+      "loss": 0.7046,
+      "step": 949
+    },
+    {
+      "epoch": 2.3227383863080684,
+      "grad_norm": 0.12551851594910046,
+      "learning_rate": 4.7099679396104605e-06,
+      "loss": 0.6393,
+      "step": 950
+    },
+    {
+      "epoch": 2.32518337408313,
+      "grad_norm": 0.12653923848609436,
+      "learning_rate": 4.709369145570694e-06,
+      "loss": 0.7521,
+      "step": 951
+    },
+    {
+      "epoch": 2.3276283618581908,
+      "grad_norm": 0.10619069024153853,
+      "learning_rate": 4.708769772183763e-06,
+      "loss": 0.7489,
+      "step": 952
+    },
+    {
+      "epoch": 2.3300733496332517,
+      "grad_norm": 0.14512024517229347,
+      "learning_rate": 4.7081698196068356e-06,
+      "loss": 0.6433,
+      "step": 953
+    },
+    {
+      "epoch": 2.332518337408313,
+      "grad_norm": 0.14017676886713798,
+      "learning_rate": 4.707569287997233e-06,
+      "loss": 0.5963,
+      "step": 954
+    },
+    {
+      "epoch": 2.334963325183374,
+      "grad_norm": 0.18220370081871406,
+      "learning_rate": 4.70696817751243e-06,
+      "loss": 0.8583,
+      "step": 955
+    },
+    {
+      "epoch": 2.337408312958435,
+      "grad_norm": 0.14414716774013517,
+      "learning_rate": 4.706366488310049e-06,
+      "loss": 0.8359,
+      "step": 956
+    },
+    {
+      "epoch": 2.3398533007334965,
+      "grad_norm": 0.12027221973726152,
+      "learning_rate": 4.705764220547868e-06,
+      "loss": 0.5259,
+      "step": 957
+    },
+    {
+      "epoch": 2.3422982885085575,
+      "grad_norm": 0.1431167847846667,
+      "learning_rate": 4.705161374383816e-06,
+      "loss": 0.5169,
+      "step": 958
+    },
+    {
+      "epoch": 2.3447432762836184,
+      "grad_norm": 0.17822753630378646,
+      "learning_rate": 4.704557949975973e-06,
+      "loss": 0.6285,
+      "step": 959
+    },
+    {
+      "epoch": 2.34718826405868,
+      "grad_norm": 0.11114969810413733,
+      "learning_rate": 4.703953947482569e-06,
+      "loss": 0.4389,
+      "step": 960
+    },
+    {
+      "epoch": 2.349633251833741,
+      "grad_norm": 0.2697790285068632,
+      "learning_rate": 4.70334936706199e-06,
+      "loss": 0.6521,
+      "step": 961
+    },
+    {
+      "epoch": 2.352078239608802,
+      "grad_norm": 0.1445098025634498,
+      "learning_rate": 4.7027442088727685e-06,
+      "loss": 0.626,
+      "step": 962
+    },
+    {
+      "epoch": 2.354523227383863,
+      "grad_norm": 0.12773159187186936,
+      "learning_rate": 4.702138473073591e-06,
+      "loss": 0.419,
+      "step": 963
+    },
+    {
+      "epoch": 2.356968215158924,
+      "grad_norm": 0.12429265324879119,
+      "learning_rate": 4.701532159823296e-06,
+      "loss": 0.5937,
+      "step": 964
+    },
+    {
+      "epoch": 2.359413202933985,
+      "grad_norm": 0.11163996494575917,
+      "learning_rate": 4.700925269280873e-06,
+      "loss": 0.6526,
+      "step": 965
+    },
+    {
+      "epoch": 2.3618581907090466,
+      "grad_norm": 0.16217787992982088,
+      "learning_rate": 4.700317801605463e-06,
+      "loss": 0.9111,
+      "step": 966
+    },
+    {
+      "epoch": 2.3643031784841075,
+      "grad_norm": 0.16188261942479187,
+      "learning_rate": 4.6997097569563556e-06,
+      "loss": 0.7366,
+      "step": 967
+    },
+    {
+      "epoch": 2.3667481662591685,
+      "grad_norm": 0.13222198974900093,
+      "learning_rate": 4.699101135492996e-06,
+      "loss": 0.4229,
+      "step": 968
+    },
+    {
+      "epoch": 2.36919315403423,
+      "grad_norm": 0.13167972641871356,
+      "learning_rate": 4.698491937374979e-06,
+      "loss": 0.6368,
+      "step": 969
+    },
+    {
+      "epoch": 2.371638141809291,
+      "grad_norm": 0.1918013332396153,
+      "learning_rate": 4.697882162762048e-06,
+      "loss": 0.528,
+      "step": 970
+    },
+    {
+      "epoch": 2.374083129584352,
+      "grad_norm": 0.11124572883435012,
+      "learning_rate": 4.697271811814104e-06,
+      "loss": 0.7842,
+      "step": 971
+    },
+    {
+      "epoch": 2.3765281173594133,
+      "grad_norm": 0.2558350325922806,
+      "learning_rate": 4.69666088469119e-06,
+      "loss": 0.8491,
+      "step": 972
+    },
+    {
+      "epoch": 2.3789731051344742,
+      "grad_norm": 0.09707668449837198,
+      "learning_rate": 4.696049381553509e-06,
+      "loss": 0.5957,
+      "step": 973
+    },
+    {
+      "epoch": 2.381418092909535,
+      "grad_norm": 0.13278229718999052,
+      "learning_rate": 4.695437302561408e-06,
+      "loss": 0.7013,
+      "step": 974
+    },
+    {
+      "epoch": 2.3838630806845966,
+      "grad_norm": 0.15091163013536923,
+      "learning_rate": 4.694824647875391e-06,
+      "loss": 0.5969,
+      "step": 975
+    },
+    {
+      "epoch": 2.3863080684596576,
+      "grad_norm": 0.1923126066806301,
+      "learning_rate": 4.6942114176561084e-06,
+      "loss": 0.531,
+      "step": 976
+    },
+    {
+      "epoch": 2.388753056234719,
+      "grad_norm": 0.1271699189668431,
+      "learning_rate": 4.693597612064363e-06,
+      "loss": 0.8707,
+      "step": 977
+    },
+    {
+      "epoch": 2.39119804400978,
+      "grad_norm": 0.14677604098361635,
+      "learning_rate": 4.69298323126111e-06,
+      "loss": 0.4934,
+      "step": 978
+    },
+    {
+      "epoch": 2.393643031784841,
+      "grad_norm": 0.13290736420022167,
+      "learning_rate": 4.692368275407454e-06,
+      "loss": 1.0197,
+      "step": 979
+    },
+    {
+      "epoch": 2.3960880195599024,
+      "grad_norm": 0.18127629850782703,
+      "learning_rate": 4.691752744664649e-06,
+      "loss": 0.924,
+      "step": 980
+    },
+    {
+      "epoch": 2.3985330073349633,
+      "grad_norm": 0.12153102637028339,
+      "learning_rate": 4.691136639194102e-06,
+      "loss": 0.8316,
+      "step": 981
+    },
+    {
+      "epoch": 2.4009779951100243,
+      "grad_norm": 0.146796926374947,
+      "learning_rate": 4.6905199591573706e-06,
+      "loss": 0.5197,
+      "step": 982
+    },
+    {
+      "epoch": 2.4034229828850857,
+      "grad_norm": 0.12258372469337145,
+      "learning_rate": 4.6899027047161615e-06,
+      "loss": 0.408,
+      "step": 983
+    },
+    {
+      "epoch": 2.4058679706601467,
+      "grad_norm": 0.13064284406398707,
+      "learning_rate": 4.689284876032334e-06,
+      "loss": 0.501,
+      "step": 984
+    },
+    {
+      "epoch": 2.4083129584352077,
+      "grad_norm": 0.13816045644304425,
+      "learning_rate": 4.688666473267896e-06,
+      "loss": 0.5305,
+      "step": 985
+    },
+    {
+      "epoch": 2.410757946210269,
+      "grad_norm": 0.1429045056588975,
+      "learning_rate": 4.688047496585007e-06,
+      "loss": 0.5871,
+      "step": 986
+    },
+    {
+      "epoch": 2.41320293398533,
+      "grad_norm": 0.15697968858913497,
+      "learning_rate": 4.687427946145976e-06,
+      "loss": 0.7086,
+      "step": 987
+    },
+    {
+      "epoch": 2.4156479217603914,
+      "grad_norm": 0.14853228297253498,
+      "learning_rate": 4.686807822113265e-06,
+      "loss": 0.7972,
+      "step": 988
+    },
+    {
+      "epoch": 2.4180929095354524,
+      "grad_norm": 0.1187897684974105,
+      "learning_rate": 4.686187124649485e-06,
+      "loss": 0.7091,
+      "step": 989
+    },
+    {
+      "epoch": 2.4205378973105134,
+      "grad_norm": 0.10890781120831121,
+      "learning_rate": 4.685565853917395e-06,
+      "loss": 0.5262,
+      "step": 990
+    },
+    {
+      "epoch": 2.422982885085575,
+      "grad_norm": 0.19300820999198887,
+      "learning_rate": 4.684944010079908e-06,
+      "loss": 1.0204,
+      "step": 991
+    },
+    {
+      "epoch": 2.4254278728606358,
+      "grad_norm": 0.13237191204695267,
+      "learning_rate": 4.684321593300085e-06,
+      "loss": 0.4703,
+      "step": 992
+    },
+    {
+      "epoch": 2.4278728606356967,
+      "grad_norm": 0.12070992681263393,
+      "learning_rate": 4.6836986037411384e-06,
+      "loss": 0.4505,
+      "step": 993
+    },
+    {
+      "epoch": 2.430317848410758,
+      "grad_norm": 0.11625392508908099,
+      "learning_rate": 4.683075041566429e-06,
+      "loss": 0.5003,
+      "step": 994
+    },
+    {
+      "epoch": 2.432762836185819,
+      "grad_norm": 0.1438745412848109,
+      "learning_rate": 4.682450906939472e-06,
+      "loss": 0.7978,
+      "step": 995
+    },
+    {
+      "epoch": 2.43520782396088,
+      "grad_norm": 0.17819620163152924,
+      "learning_rate": 4.681826200023927e-06,
+      "loss": 0.4904,
+      "step": 996
+    },
+    {
+      "epoch": 2.4376528117359415,
+      "grad_norm": 0.15526300365703913,
+      "learning_rate": 4.681200920983607e-06,
+      "loss": 0.8612,
+      "step": 997
+    },
+    {
+      "epoch": 2.4400977995110025,
+      "grad_norm": 0.13594017934348038,
+      "learning_rate": 4.680575069982475e-06,
+      "loss": 0.4693,
+      "step": 998
+    },
+    {
+      "epoch": 2.4425427872860634,
+      "grad_norm": 0.13653953028606336,
+      "learning_rate": 4.679948647184644e-06,
+      "loss": 0.8228,
+      "step": 999
+    },
+    {
+      "epoch": 2.444987775061125,
+      "grad_norm": 0.16892391473239243,
+      "learning_rate": 4.679321652754375e-06,
+      "loss": 0.6108,
+      "step": 1000
+    },
+    {
+      "epoch": 2.447432762836186,
+      "grad_norm": 0.1801068582953552,
+      "learning_rate": 4.678694086856081e-06,
+      "loss": 0.4818,
+      "step": 1001
+    },
+    {
+      "epoch": 2.449877750611247,
+      "grad_norm": 0.16708118134265817,
+      "learning_rate": 4.678065949654324e-06,
+      "loss": 0.5227,
+      "step": 1002
+    },
+    {
+      "epoch": 2.452322738386308,
+      "grad_norm": 0.11393260708516915,
+      "learning_rate": 4.677437241313816e-06,
+      "loss": 0.5652,
+      "step": 1003
+    },
+    {
+      "epoch": 2.454767726161369,
+      "grad_norm": 0.1404136854199711,
+      "learning_rate": 4.6768079619994195e-06,
+      "loss": 0.6254,
+      "step": 1004
+    },
+    {
+      "epoch": 2.45721271393643,
+      "grad_norm": 0.1603197520447684,
+      "learning_rate": 4.676178111876145e-06,
+      "loss": 0.8834,
+      "step": 1005
+    },
+    {
+      "epoch": 2.4596577017114916,
+      "grad_norm": 0.11340308063412807,
+      "learning_rate": 4.675547691109154e-06,
+      "loss": 0.4282,
+      "step": 1006
+    },
+    {
+      "epoch": 2.4621026894865525,
+      "grad_norm": 0.1431421510508806,
+      "learning_rate": 4.674916699863756e-06,
+      "loss": 0.5582,
+      "step": 1007
+    },
+    {
+      "epoch": 2.4645476772616135,
+      "grad_norm": 0.09815779582924551,
+      "learning_rate": 4.674285138305414e-06,
+      "loss": 0.3488,
+      "step": 1008
+    },
+    {
+      "epoch": 2.466992665036675,
+      "grad_norm": 0.17378708860918254,
+      "learning_rate": 4.6736530065997356e-06,
+      "loss": 0.4958,
+      "step": 1009
+    },
+    {
+      "epoch": 2.469437652811736,
+      "grad_norm": 0.2074167538585944,
+      "learning_rate": 4.6730203049124814e-06,
+      "loss": 0.476,
+      "step": 1010
+    },
+    {
+      "epoch": 2.471882640586797,
+      "grad_norm": 0.15955371484594874,
+      "learning_rate": 4.672387033409559e-06,
+      "loss": 0.4671,
+      "step": 1011
+    },
+    {
+      "epoch": 2.4743276283618583,
+      "grad_norm": 0.149707285495147,
+      "learning_rate": 4.671753192257028e-06,
+      "loss": 0.404,
+      "step": 1012
+    },
+    {
+      "epoch": 2.4767726161369192,
+      "grad_norm": 0.12917962775273079,
+      "learning_rate": 4.6711187816210965e-06,
+      "loss": 0.6948,
+      "step": 1013
+    },
+    {
+      "epoch": 2.47921760391198,
+      "grad_norm": 0.15545400159453354,
+      "learning_rate": 4.67048380166812e-06,
+      "loss": 0.7672,
+      "step": 1014
+    },
+    {
+      "epoch": 2.4816625916870416,
+      "grad_norm": 0.11432517987596699,
+      "learning_rate": 4.669848252564605e-06,
+      "loss": 0.5068,
+      "step": 1015
+    },
+    {
+      "epoch": 2.4841075794621026,
+      "grad_norm": 0.18610809818951876,
+      "learning_rate": 4.669212134477208e-06,
+      "loss": 0.5838,
+      "step": 1016
+    },
+    {
+      "epoch": 2.486552567237164,
+      "grad_norm": 0.1603389560247637,
+      "learning_rate": 4.6685754475727315e-06,
+      "loss": 0.4616,
+      "step": 1017
+    },
+    {
+      "epoch": 2.488997555012225,
+      "grad_norm": 0.17459829017724168,
+      "learning_rate": 4.667938192018133e-06,
+      "loss": 1.0443,
+      "step": 1018
+    },
+    {
+      "epoch": 2.491442542787286,
+      "grad_norm": 0.15081780987153742,
+      "learning_rate": 4.667300367980512e-06,
+      "loss": 0.7724,
+      "step": 1019
+    },
+    {
+      "epoch": 2.4938875305623474,
+      "grad_norm": 0.11953483673601678,
+      "learning_rate": 4.666661975627123e-06,
+      "loss": 0.541,
+      "step": 1020
+    },
+    {
+      "epoch": 2.4963325183374083,
+      "grad_norm": 0.18042774162777217,
+      "learning_rate": 4.666023015125366e-06,
+      "loss": 0.5689,
+      "step": 1021
+    },
+    {
+      "epoch": 2.4987775061124693,
+      "grad_norm": 0.22952574694175681,
+      "learning_rate": 4.665383486642791e-06,
+      "loss": 0.9075,
+      "step": 1022
+    },
+    {
+      "epoch": 2.5012224938875307,
+      "grad_norm": 0.12555143621518156,
+      "learning_rate": 4.664743390347098e-06,
+      "loss": 0.7301,
+      "step": 1023
+    },
+    {
+      "epoch": 2.5036674816625917,
+      "grad_norm": 0.14648773488525046,
+      "learning_rate": 4.664102726406132e-06,
+      "loss": 1.0722,
+      "step": 1024
+    },
+    {
+      "epoch": 2.5061124694376526,
+      "grad_norm": 0.20242907172936675,
+      "learning_rate": 4.663461494987894e-06,
+      "loss": 0.6039,
+      "step": 1025
+    },
+    {
+      "epoch": 2.508557457212714,
+      "grad_norm": 0.15438751303198492,
+      "learning_rate": 4.662819696260527e-06,
+      "loss": 0.5306,
+      "step": 1026
+    },
+    {
+      "epoch": 2.511002444987775,
+      "grad_norm": 0.1552410765148403,
+      "learning_rate": 4.662177330392324e-06,
+      "loss": 0.4432,
+      "step": 1027
+    },
+    {
+      "epoch": 2.5134474327628364,
+      "grad_norm": 0.23341387331635374,
+      "learning_rate": 4.661534397551732e-06,
+      "loss": 0.7674,
+      "step": 1028
+    },
+    {
+      "epoch": 2.5158924205378974,
+      "grad_norm": 0.1761720270268955,
+      "learning_rate": 4.66089089790734e-06,
+      "loss": 0.9345,
+      "step": 1029
+    },
+    {
+      "epoch": 2.5183374083129584,
+      "grad_norm": 0.21442705492206302,
+      "learning_rate": 4.660246831627888e-06,
+      "loss": 0.397,
+      "step": 1030
+    },
+    {
+      "epoch": 2.52078239608802,
+      "grad_norm": 0.15063333108197974,
+      "learning_rate": 4.659602198882266e-06,
+      "loss": 0.6694,
+      "step": 1031
+    },
+    {
+      "epoch": 2.5232273838630808,
+      "grad_norm": 0.13793425707722298,
+      "learning_rate": 4.658956999839512e-06,
+      "loss": 0.5444,
+      "step": 1032
+    },
+    {
+      "epoch": 2.5256723716381417,
+      "grad_norm": 0.14919245311740761,
+      "learning_rate": 4.658311234668811e-06,
+      "loss": 0.5382,
+      "step": 1033
+    },
+    {
+      "epoch": 2.528117359413203,
+      "grad_norm": 0.1552781122188146,
+      "learning_rate": 4.657664903539497e-06,
+      "loss": 0.7202,
+      "step": 1034
+    },
+    {
+      "epoch": 2.530562347188264,
+      "grad_norm": 0.1361755501567377,
+      "learning_rate": 4.657018006621054e-06,
+      "loss": 0.6824,
+      "step": 1035
+    },
+    {
+      "epoch": 2.533007334963325,
+      "grad_norm": 0.17129766514631103,
+      "learning_rate": 4.656370544083112e-06,
+      "loss": 0.5598,
+      "step": 1036
+    },
+    {
+      "epoch": 2.5354523227383865,
+      "grad_norm": 0.16876796946532477,
+      "learning_rate": 4.6557225160954506e-06,
+      "loss": 0.9978,
+      "step": 1037
+    },
+    {
+      "epoch": 2.5378973105134475,
+      "grad_norm": 0.16195199884778624,
+      "learning_rate": 4.655073922827998e-06,
+      "loss": 0.5292,
+      "step": 1038
+    },
+    {
+      "epoch": 2.5403422982885084,
+      "grad_norm": 0.14150547349395673,
+      "learning_rate": 4.654424764450832e-06,
+      "loss": 0.6811,
+      "step": 1039
+    },
+    {
+      "epoch": 2.54278728606357,
+      "grad_norm": 0.2344239648151507,
+      "learning_rate": 4.653775041134174e-06,
+      "loss": 0.4769,
+      "step": 1040
+    },
+    {
+      "epoch": 2.545232273838631,
+      "grad_norm": 0.2007802792542489,
+      "learning_rate": 4.653124753048397e-06,
+      "loss": 0.661,
+      "step": 1041
+    },
+    {
+      "epoch": 2.547677261613692,
+      "grad_norm": 0.16131630634799532,
+      "learning_rate": 4.652473900364021e-06,
+      "loss": 0.8791,
+      "step": 1042
+    },
+    {
+      "epoch": 2.550122249388753,
+      "grad_norm": 0.14814968797381928,
+      "learning_rate": 4.6518224832517174e-06,
+      "loss": 0.4839,
+      "step": 1043
+    },
+    {
+      "epoch": 2.552567237163814,
+      "grad_norm": 0.15743299706469807,
+      "learning_rate": 4.651170501882299e-06,
+      "loss": 0.5036,
+      "step": 1044
+    },
+    {
+      "epoch": 2.555012224938875,
+      "grad_norm": 0.17773198221053674,
+      "learning_rate": 4.650517956426732e-06,
+      "loss": 0.4923,
+      "step": 1045
+    },
+    {
+      "epoch": 2.5574572127139366,
+      "grad_norm": 0.24212561295642518,
+      "learning_rate": 4.649864847056129e-06,
+      "loss": 0.5449,
+      "step": 1046
+    },
+    {
+      "epoch": 2.5599022004889975,
+      "grad_norm": 0.2316556041739059,
+      "learning_rate": 4.64921117394175e-06,
+      "loss": 0.8427,
+      "step": 1047
+    },
+    {
+      "epoch": 2.5623471882640585,
+      "grad_norm": 0.09738370266496339,
+      "learning_rate": 4.6485569372550025e-06,
+      "loss": 0.4653,
+      "step": 1048
+    },
+    {
+      "epoch": 2.56479217603912,
+      "grad_norm": 0.16681168705222116,
+      "learning_rate": 4.647902137167443e-06,
+      "loss": 0.5186,
+      "step": 1049
+    },
+    {
+      "epoch": 2.567237163814181,
+      "grad_norm": 0.1499951678784469,
+      "learning_rate": 4.647246773850773e-06,
+      "loss": 0.4371,
+      "step": 1050
+    },
+    {
+      "epoch": 2.569682151589242,
+      "grad_norm": 0.16475374429973771,
+      "learning_rate": 4.646590847476847e-06,
+      "loss": 0.5339,
+      "step": 1051
+    },
+    {
+      "epoch": 2.5721271393643033,
+      "grad_norm": 0.19656045166488556,
+      "learning_rate": 4.645934358217662e-06,
+      "loss": 0.4902,
+      "step": 1052
+    },
+    {
+      "epoch": 2.5745721271393642,
+      "grad_norm": 0.13851180565946436,
+      "learning_rate": 4.645277306245365e-06,
+      "loss": 0.5153,
+      "step": 1053
+    },
+    {
+      "epoch": 2.577017114914425,
+      "grad_norm": 0.13941503480662112,
+      "learning_rate": 4.64461969173225e-06,
+      "loss": 0.7362,
+      "step": 1054
+    },
+    {
+      "epoch": 2.5794621026894866,
+      "grad_norm": 0.1317585043398215,
+      "learning_rate": 4.643961514850759e-06,
+      "loss": 0.4666,
+      "step": 1055
+    },
+    {
+      "epoch": 2.5819070904645476,
+      "grad_norm": 0.13810522049144833,
+      "learning_rate": 4.64330277577348e-06,
+      "loss": 0.3837,
+      "step": 1056
+    },
+    {
+      "epoch": 2.5843520782396086,
+      "grad_norm": 0.13132952247335872,
+      "learning_rate": 4.64264347467315e-06,
+      "loss": 0.6076,
+      "step": 1057
+    },
+    {
+      "epoch": 2.58679706601467,
+      "grad_norm": 0.21038598840295386,
+      "learning_rate": 4.641983611722653e-06,
+      "loss": 0.5085,
+      "step": 1058
+    },
+    {
+      "epoch": 2.589242053789731,
+      "grad_norm": 0.1263050206040351,
+      "learning_rate": 4.6413231870950204e-06,
+      "loss": 0.9666,
+      "step": 1059
+    },
+    {
+      "epoch": 2.591687041564792,
+      "grad_norm": 0.17622449682485733,
+      "learning_rate": 4.64066220096343e-06,
+      "loss": 0.4403,
+      "step": 1060
+    },
+    {
+      "epoch": 2.5941320293398533,
+      "grad_norm": 0.13355327315207532,
+      "learning_rate": 4.640000653501207e-06,
+      "loss": 0.5231,
+      "step": 1061
+    },
+    {
+      "epoch": 2.5965770171149143,
+      "grad_norm": 0.20810238917678098,
+      "learning_rate": 4.639338544881826e-06,
+      "loss": 0.6895,
+      "step": 1062
+    },
+    {
+      "epoch": 2.5990220048899757,
+      "grad_norm": 0.16758404947862493,
+      "learning_rate": 4.638675875278907e-06,
+      "loss": 0.8839,
+      "step": 1063
+    },
+    {
+      "epoch": 2.6014669926650367,
+      "grad_norm": 0.21635251669690372,
+      "learning_rate": 4.638012644866216e-06,
+      "loss": 1.1101,
+      "step": 1064
+    },
+    {
+      "epoch": 2.6039119804400976,
+      "grad_norm": 0.19783844597685704,
+      "learning_rate": 4.637348853817666e-06,
+      "loss": 1.3538,
+      "step": 1065
+    },
+    {
+      "epoch": 2.606356968215159,
+      "grad_norm": 0.20553761607024784,
+      "learning_rate": 4.636684502307321e-06,
+      "loss": 0.4355,
+      "step": 1066
+    },
+    {
+      "epoch": 2.60880195599022,
+      "grad_norm": 0.1605473761595738,
+      "learning_rate": 4.636019590509387e-06,
+      "loss": 0.5234,
+      "step": 1067
+    },
+    {
+      "epoch": 2.6112469437652814,
+      "grad_norm": 0.16229957508562184,
+      "learning_rate": 4.635354118598221e-06,
+      "loss": 0.71,
+      "step": 1068
+    },
+    {
+      "epoch": 2.6136919315403424,
+      "grad_norm": 0.2628449579611908,
+      "learning_rate": 4.634688086748323e-06,
+      "loss": 0.7098,
+      "step": 1069
+    },
+    {
+      "epoch": 2.6161369193154034,
+      "grad_norm": 0.11654349733319179,
+      "learning_rate": 4.634021495134343e-06,
+      "loss": 0.4696,
+      "step": 1070
+    },
+    {
+      "epoch": 2.618581907090465,
+      "grad_norm": 0.18229759695237086,
+      "learning_rate": 4.633354343931077e-06,
+      "loss": 1.0007,
+      "step": 1071
+    },
+    {
+      "epoch": 2.6210268948655258,
+      "grad_norm": 0.12092147841102768,
+      "learning_rate": 4.6326866333134655e-06,
+      "loss": 0.4601,
+      "step": 1072
+    },
+    {
+      "epoch": 2.6234718826405867,
+      "grad_norm": 0.1198755721681329,
+      "learning_rate": 4.6320183634565996e-06,
+      "loss": 0.4213,
+      "step": 1073
+    },
+    {
+      "epoch": 2.625916870415648,
+      "grad_norm": 0.1413613396448796,
+      "learning_rate": 4.631349534535713e-06,
+      "loss": 0.4565,
+      "step": 1074
+    },
+    {
+      "epoch": 2.628361858190709,
+      "grad_norm": 0.15358467095315684,
+      "learning_rate": 4.6306801467261895e-06,
+      "loss": 0.5086,
+      "step": 1075
+    },
+    {
+      "epoch": 2.63080684596577,
+      "grad_norm": 0.21728528200081973,
+      "learning_rate": 4.6300102002035576e-06,
+      "loss": 0.7187,
+      "step": 1076
+    },
+    {
+      "epoch": 2.6332518337408315,
+      "grad_norm": 0.15114237598030636,
+      "learning_rate": 4.629339695143492e-06,
+      "loss": 0.5394,
+      "step": 1077
+    },
+    {
+      "epoch": 2.6356968215158925,
+      "grad_norm": 0.21030276424096803,
+      "learning_rate": 4.628668631721816e-06,
+      "loss": 0.5601,
+      "step": 1078
+    },
+    {
+      "epoch": 2.6381418092909534,
+      "grad_norm": 0.39810798703626527,
+      "learning_rate": 4.627997010114495e-06,
+      "loss": 0.923,
+      "step": 1079
+    },
+    {
+      "epoch": 2.640586797066015,
+      "grad_norm": 0.1353956511691117,
+      "learning_rate": 4.627324830497646e-06,
+      "loss": 0.5467,
+      "step": 1080
+    },
+    {
+      "epoch": 2.643031784841076,
+      "grad_norm": 0.1808890853882477,
+      "learning_rate": 4.6266520930475286e-06,
+      "loss": 0.7682,
+      "step": 1081
+    },
+    {
+      "epoch": 2.645476772616137,
+      "grad_norm": 0.1256481379291353,
+      "learning_rate": 4.62597879794055e-06,
+      "loss": 0.6591,
+      "step": 1082
+    },
+    {
+      "epoch": 2.647921760391198,
+      "grad_norm": 0.12709757847018727,
+      "learning_rate": 4.625304945353265e-06,
+      "loss": 0.7998,
+      "step": 1083
+    },
+    {
+      "epoch": 2.650366748166259,
+      "grad_norm": 0.18860867427527078,
+      "learning_rate": 4.6246305354623725e-06,
+      "loss": 0.9052,
+      "step": 1084
+    },
+    {
+      "epoch": 2.65281173594132,
+      "grad_norm": 0.10716849762092664,
+      "learning_rate": 4.623955568444717e-06,
+      "loss": 0.4708,
+      "step": 1085
+    },
+    {
+      "epoch": 2.6552567237163816,
+      "grad_norm": 0.14974019719972007,
+      "learning_rate": 4.623280044477291e-06,
+      "loss": 1.0287,
+      "step": 1086
+    },
+    {
+      "epoch": 2.6577017114914425,
+      "grad_norm": 0.1771362167640652,
+      "learning_rate": 4.6226039637372334e-06,
+      "loss": 0.7266,
+      "step": 1087
+    },
+    {
+      "epoch": 2.6601466992665035,
+      "grad_norm": 0.17744649800600945,
+      "learning_rate": 4.6219273264018265e-06,
+      "loss": 0.5589,
+      "step": 1088
+    },
+    {
+      "epoch": 2.662591687041565,
+      "grad_norm": 0.1414630445393385,
+      "learning_rate": 4.621250132648501e-06,
+      "loss": 0.4179,
+      "step": 1089
+    },
+    {
+      "epoch": 2.665036674816626,
+      "grad_norm": 0.14623572467848908,
+      "learning_rate": 4.620572382654833e-06,
+      "loss": 0.5305,
+      "step": 1090
+    },
+    {
+      "epoch": 2.667481662591687,
+      "grad_norm": 0.1627982292451145,
+      "learning_rate": 4.619894076598543e-06,
+      "loss": 0.5862,
+      "step": 1091
+    },
+    {
+      "epoch": 2.6699266503667483,
+      "grad_norm": 0.19089055860318369,
+      "learning_rate": 4.619215214657498e-06,
+      "loss": 1.0751,
+      "step": 1092
+    },
+    {
+      "epoch": 2.6723716381418092,
+      "grad_norm": 0.12607217717478977,
+      "learning_rate": 4.618535797009713e-06,
+      "loss": 0.6387,
+      "step": 1093
+    },
+    {
+      "epoch": 2.67481662591687,
+      "grad_norm": 0.12552754367747399,
+      "learning_rate": 4.6178558238333456e-06,
+      "loss": 0.7387,
+      "step": 1094
+    },
+    {
+      "epoch": 2.6772616136919316,
+      "grad_norm": 0.11711762571840323,
+      "learning_rate": 4.617175295306701e-06,
+      "loss": 0.4759,
+      "step": 1095
+    },
+    {
+      "epoch": 2.6797066014669926,
+      "grad_norm": 0.16928081239958062,
+      "learning_rate": 4.616494211608229e-06,
+      "loss": 0.4545,
+      "step": 1096
+    },
+    {
+      "epoch": 2.6821515892420535,
+      "grad_norm": 0.16753995890049916,
+      "learning_rate": 4.615812572916525e-06,
+      "loss": 0.5309,
+      "step": 1097
+    },
+    {
+      "epoch": 2.684596577017115,
+      "grad_norm": 0.15483599059983444,
+      "learning_rate": 4.615130379410329e-06,
+      "loss": 0.6783,
+      "step": 1098
+    },
+    {
+      "epoch": 2.687041564792176,
+      "grad_norm": 0.22104396671414372,
+      "learning_rate": 4.614447631268531e-06,
+      "loss": 0.6145,
+      "step": 1099
+    },
+    {
+      "epoch": 2.689486552567237,
+      "grad_norm": 0.1786122056076665,
+      "learning_rate": 4.613764328670161e-06,
+      "loss": 0.9847,
+      "step": 1100
+    },
+    {
+      "epoch": 2.6919315403422983,
+      "grad_norm": 0.1639304987605364,
+      "learning_rate": 4.613080471794397e-06,
+      "loss": 0.7752,
+      "step": 1101
+    },
+    {
+      "epoch": 2.6943765281173593,
+      "grad_norm": 0.21666933161378235,
+      "learning_rate": 4.612396060820563e-06,
+      "loss": 0.4715,
+      "step": 1102
+    },
+    {
+      "epoch": 2.6968215158924207,
+      "grad_norm": 0.1406384182804471,
+      "learning_rate": 4.611711095928126e-06,
+      "loss": 0.4199,
+      "step": 1103
+    },
+    {
+      "epoch": 2.6992665036674817,
+      "grad_norm": 0.14032307272962788,
+      "learning_rate": 4.611025577296699e-06,
+      "loss": 0.4079,
+      "step": 1104
+    },
+    {
+      "epoch": 2.7017114914425426,
+      "grad_norm": 0.1735388645254333,
+      "learning_rate": 4.610339505106042e-06,
+      "loss": 0.5341,
+      "step": 1105
+    },
+    {
+      "epoch": 2.704156479217604,
+      "grad_norm": 0.13987110208172146,
+      "learning_rate": 4.6096528795360575e-06,
+      "loss": 0.7478,
+      "step": 1106
+    },
+    {
+      "epoch": 2.706601466992665,
+      "grad_norm": 0.12405469393257719,
+      "learning_rate": 4.608965700766795e-06,
+      "loss": 0.5862,
+      "step": 1107
+    },
+    {
+      "epoch": 2.7090464547677264,
+      "grad_norm": 0.13439305148331876,
+      "learning_rate": 4.608277968978449e-06,
+      "loss": 0.472,
+      "step": 1108
+    },
+    {
+      "epoch": 2.7114914425427874,
+      "grad_norm": 0.22771469787635684,
+      "learning_rate": 4.607589684351358e-06,
+      "loss": 0.6159,
+      "step": 1109
+    },
+    {
+      "epoch": 2.7139364303178484,
+      "grad_norm": 0.20777836853947337,
+      "learning_rate": 4.606900847066007e-06,
+      "loss": 0.9789,
+      "step": 1110
+    },
+    {
+      "epoch": 2.71638141809291,
+      "grad_norm": 0.10579540328389136,
+      "learning_rate": 4.606211457303023e-06,
+      "loss": 0.4085,
+      "step": 1111
+    },
+    {
+      "epoch": 2.7188264058679708,
+      "grad_norm": 0.12372849376020849,
+      "learning_rate": 4.605521515243181e-06,
+      "loss": 0.3852,
+      "step": 1112
+    },
+    {
+      "epoch": 2.7212713936430317,
+      "grad_norm": 0.18214940924175324,
+      "learning_rate": 4.604831021067399e-06,
+      "loss": 0.7262,
+      "step": 1113
+    },
+    {
+      "epoch": 2.723716381418093,
+      "grad_norm": 0.16508046445229,
+      "learning_rate": 4.604139974956742e-06,
+      "loss": 1.1808,
+      "step": 1114
+    },
+    {
+      "epoch": 2.726161369193154,
+      "grad_norm": 0.14812349353016127,
+      "learning_rate": 4.603448377092417e-06,
+      "loss": 0.6326,
+      "step": 1115
+    },
+    {
+      "epoch": 2.728606356968215,
+      "grad_norm": 0.1524642289589814,
+      "learning_rate": 4.602756227655775e-06,
+      "loss": 0.848,
+      "step": 1116
+    },
+    {
+      "epoch": 2.7310513447432765,
+      "grad_norm": 0.14841344133862183,
+      "learning_rate": 4.602063526828317e-06,
+      "loss": 0.4519,
+      "step": 1117
+    },
+    {
+      "epoch": 2.7334963325183375,
+      "grad_norm": 0.22498396643855229,
+      "learning_rate": 4.601370274791683e-06,
+      "loss": 0.6302,
+      "step": 1118
+    },
+    {
+      "epoch": 2.7359413202933984,
+      "grad_norm": 0.17427705509509117,
+      "learning_rate": 4.600676471727659e-06,
+      "loss": 0.4823,
+      "step": 1119
+    },
+    {
+      "epoch": 2.73838630806846,
+      "grad_norm": 0.11765019267284965,
+      "learning_rate": 4.599982117818177e-06,
+      "loss": 0.6129,
+      "step": 1120
+    },
+    {
+      "epoch": 2.740831295843521,
+      "grad_norm": 0.12678908999019384,
+      "learning_rate": 4.599287213245312e-06,
+      "loss": 0.4959,
+      "step": 1121
+    },
+    {
+      "epoch": 2.743276283618582,
+      "grad_norm": 0.17136478844031794,
+      "learning_rate": 4.598591758191284e-06,
+      "loss": 0.492,
+      "step": 1122
+    },
+    {
+      "epoch": 2.745721271393643,
+      "grad_norm": 0.1619824238914578,
+      "learning_rate": 4.597895752838458e-06,
+      "loss": 1.0815,
+      "step": 1123
+    },
+    {
+      "epoch": 2.748166259168704,
+      "grad_norm": 0.134354982306176,
+      "learning_rate": 4.597199197369342e-06,
+      "loss": 0.7212,
+      "step": 1124
+    },
+    {
+      "epoch": 2.750611246943765,
+      "grad_norm": 0.1791360354603443,
+      "learning_rate": 4.5965020919665874e-06,
+      "loss": 0.5534,
+      "step": 1125
+    },
+    {
+      "epoch": 2.7530562347188265,
+      "grad_norm": 0.16845668333686556,
+      "learning_rate": 4.595804436812994e-06,
+      "loss": 0.7317,
+      "step": 1126
+    },
+    {
+      "epoch": 2.7555012224938875,
+      "grad_norm": 0.24891756650069807,
+      "learning_rate": 4.595106232091501e-06,
+      "loss": 0.3743,
+      "step": 1127
+    },
+    {
+      "epoch": 2.7579462102689485,
+      "grad_norm": 0.12827044760700881,
+      "learning_rate": 4.594407477985194e-06,
+      "loss": 0.5523,
+      "step": 1128
+    },
+    {
+      "epoch": 2.76039119804401,
+      "grad_norm": 0.1535366582898294,
+      "learning_rate": 4.593708174677302e-06,
+      "loss": 0.9273,
+      "step": 1129
+    },
+    {
+      "epoch": 2.762836185819071,
+      "grad_norm": 0.13908411961849743,
+      "learning_rate": 4.593008322351199e-06,
+      "loss": 0.5235,
+      "step": 1130
+    },
+    {
+      "epoch": 2.765281173594132,
+      "grad_norm": 0.15128583163765236,
+      "learning_rate": 4.592307921190402e-06,
+      "loss": 0.6772,
+      "step": 1131
+    },
+    {
+      "epoch": 2.7677261613691932,
+      "grad_norm": 0.20381586142103972,
+      "learning_rate": 4.591606971378573e-06,
+      "loss": 0.4186,
+      "step": 1132
+    },
+    {
+      "epoch": 2.770171149144254,
+      "grad_norm": 0.19922371363212993,
+      "learning_rate": 4.590905473099515e-06,
+      "loss": 0.4984,
+      "step": 1133
+    },
+    {
+      "epoch": 2.772616136919315,
+      "grad_norm": 0.1656824219964035,
+      "learning_rate": 4.590203426537179e-06,
+      "loss": 0.8546,
+      "step": 1134
+    },
+    {
+      "epoch": 2.7750611246943766,
+      "grad_norm": 0.18635990870109867,
+      "learning_rate": 4.589500831875658e-06,
+      "loss": 0.4809,
+      "step": 1135
+    },
+    {
+      "epoch": 2.7775061124694376,
+      "grad_norm": 0.11909794478489337,
+      "learning_rate": 4.588797689299185e-06,
+      "loss": 0.5238,
+      "step": 1136
+    },
+    {
+      "epoch": 2.7799511002444985,
+      "grad_norm": 0.09250628919307029,
+      "learning_rate": 4.588093998992144e-06,
+      "loss": 0.391,
+      "step": 1137
+    },
+    {
+      "epoch": 2.78239608801956,
+      "grad_norm": 0.2038924536008576,
+      "learning_rate": 4.587389761139057e-06,
+      "loss": 0.8296,
+      "step": 1138
+    },
+    {
+      "epoch": 2.784841075794621,
+      "grad_norm": 0.20391590091864686,
+      "learning_rate": 4.586684975924591e-06,
+      "loss": 1.0367,
+      "step": 1139
+    },
+    {
+      "epoch": 2.787286063569682,
+      "grad_norm": 0.19815045096211195,
+      "learning_rate": 4.585979643533558e-06,
+      "loss": 0.9047,
+      "step": 1140
+    },
+    {
+      "epoch": 2.7897310513447433,
+      "grad_norm": 0.180186434042982,
+      "learning_rate": 4.585273764150911e-06,
+      "loss": 0.8345,
+      "step": 1141
+    },
+    {
+      "epoch": 2.7921760391198043,
+      "grad_norm": 0.18892634971252834,
+      "learning_rate": 4.584567337961748e-06,
+      "loss": 0.8406,
+      "step": 1142
+    },
+    {
+      "epoch": 2.7946210268948657,
+      "grad_norm": 0.15912441421470258,
+      "learning_rate": 4.583860365151311e-06,
+      "loss": 0.4768,
+      "step": 1143
+    },
+    {
+      "epoch": 2.7970660146699267,
+      "grad_norm": 0.1454674981094276,
+      "learning_rate": 4.583152845904984e-06,
+      "loss": 0.5008,
+      "step": 1144
+    },
+    {
+      "epoch": 2.7995110024449876,
+      "grad_norm": 0.18786730789366027,
+      "learning_rate": 4.582444780408294e-06,
+      "loss": 1.0029,
+      "step": 1145
+    },
+    {
+      "epoch": 2.801955990220049,
+      "grad_norm": 0.18217041426647468,
+      "learning_rate": 4.581736168846913e-06,
+      "loss": 0.8548,
+      "step": 1146
+    },
+    {
+      "epoch": 2.80440097799511,
+      "grad_norm": 0.12938018093067463,
+      "learning_rate": 4.581027011406654e-06,
+      "loss": 0.4258,
+      "step": 1147
+    },
+    {
+      "epoch": 2.8068459657701714,
+      "grad_norm": 0.21399634838588283,
+      "learning_rate": 4.580317308273476e-06,
+      "loss": 0.5438,
+      "step": 1148
+    },
+    {
+      "epoch": 2.8092909535452324,
+      "grad_norm": 0.18062754053107197,
+      "learning_rate": 4.579607059633478e-06,
+      "loss": 1.2083,
+      "step": 1149
+    },
+    {
+      "epoch": 2.8117359413202934,
+      "grad_norm": 0.13329013577527737,
+      "learning_rate": 4.578896265672904e-06,
+      "loss": 0.805,
+      "step": 1150
+    },
+    {
+      "epoch": 2.8141809290953548,
+      "grad_norm": 0.1907589633661021,
+      "learning_rate": 4.57818492657814e-06,
+      "loss": 0.6109,
+      "step": 1151
+    },
+    {
+      "epoch": 2.8166259168704157,
+      "grad_norm": 0.1955894405461504,
+      "learning_rate": 4.577473042535716e-06,
+      "loss": 0.4771,
+      "step": 1152
+    },
+    {
+      "epoch": 2.8190709046454767,
+      "grad_norm": 0.16305278316672941,
+      "learning_rate": 4.576760613732304e-06,
+      "loss": 0.7663,
+      "step": 1153
+    },
+    {
+      "epoch": 2.821515892420538,
+      "grad_norm": 0.18503576796762433,
+      "learning_rate": 4.5760476403547195e-06,
+      "loss": 0.54,
+      "step": 1154
+    },
+    {
+      "epoch": 2.823960880195599,
+      "grad_norm": 0.13403390245576055,
+      "learning_rate": 4.57533412258992e-06,
+      "loss": 0.795,
+      "step": 1155
+    },
+    {
+      "epoch": 2.82640586797066,
+      "grad_norm": 0.1058672590575123,
+      "learning_rate": 4.574620060625007e-06,
+      "loss": 0.4006,
+      "step": 1156
+    },
+    {
+      "epoch": 2.8288508557457215,
+      "grad_norm": 0.2619085689200034,
+      "learning_rate": 4.573905454647221e-06,
+      "loss": 0.5499,
+      "step": 1157
+    },
+    {
+      "epoch": 2.8312958435207825,
+      "grad_norm": 0.20133711206782226,
+      "learning_rate": 4.573190304843953e-06,
+      "loss": 0.4823,
+      "step": 1158
+    },
+    {
+      "epoch": 2.8337408312958434,
+      "grad_norm": 0.176477271768202,
+      "learning_rate": 4.572474611402727e-06,
+      "loss": 0.4592,
+      "step": 1159
+    },
+    {
+      "epoch": 2.836185819070905,
+      "grad_norm": 0.15256435679263647,
+      "learning_rate": 4.571758374511217e-06,
+      "loss": 0.6061,
+      "step": 1160
+    },
+    {
+      "epoch": 2.838630806845966,
+      "grad_norm": 0.1578203928635288,
+      "learning_rate": 4.571041594357236e-06,
+      "loss": 0.7555,
+      "step": 1161
+    },
+    {
+      "epoch": 2.8410757946210268,
+      "grad_norm": 0.18603253858672134,
+      "learning_rate": 4.5703242711287394e-06,
+      "loss": 0.4318,
+      "step": 1162
+    },
+    {
+      "epoch": 2.843520782396088,
+      "grad_norm": 0.20892142538076658,
+      "learning_rate": 4.569606405013827e-06,
+      "loss": 1.1209,
+      "step": 1163
+    },
+    {
+      "epoch": 2.845965770171149,
+      "grad_norm": 0.15200102966958237,
+      "learning_rate": 4.568887996200738e-06,
+      "loss": 0.6955,
+      "step": 1164
+    },
+    {
+      "epoch": 2.84841075794621,
+      "grad_norm": 0.1435574938658663,
+      "learning_rate": 4.568169044877857e-06,
+      "loss": 0.4129,
+      "step": 1165
+    },
+    {
+      "epoch": 2.8508557457212715,
+      "grad_norm": 0.13225672577335346,
+      "learning_rate": 4.56744955123371e-06,
+      "loss": 0.4034,
+      "step": 1166
+    },
+    {
+      "epoch": 2.8533007334963325,
+      "grad_norm": 0.1323564354039993,
+      "learning_rate": 4.5667295154569635e-06,
+      "loss": 0.5591,
+      "step": 1167
+    },
+    {
+      "epoch": 2.8557457212713935,
+      "grad_norm": 0.12474035239423872,
+      "learning_rate": 4.566008937736428e-06,
+      "loss": 0.723,
+      "step": 1168
+    },
+    {
+      "epoch": 2.858190709046455,
+      "grad_norm": 0.15188252590376547,
+      "learning_rate": 4.5652878182610545e-06,
+      "loss": 0.5545,
+      "step": 1169
+    },
+    {
+      "epoch": 2.860635696821516,
+      "grad_norm": 0.14490741605601506,
+      "learning_rate": 4.564566157219938e-06,
+      "loss": 0.4108,
+      "step": 1170
+    },
+    {
+      "epoch": 2.863080684596577,
+      "grad_norm": 0.13716370715185708,
+      "learning_rate": 4.563843954802315e-06,
+      "loss": 0.5045,
+      "step": 1171
+    },
+    {
+      "epoch": 2.8655256723716382,
+      "grad_norm": 0.14640593017626594,
+      "learning_rate": 4.563121211197562e-06,
+      "loss": 0.4527,
+      "step": 1172
+    },
+    {
+      "epoch": 2.867970660146699,
+      "grad_norm": 0.15428150761732115,
+      "learning_rate": 4.562397926595202e-06,
+      "loss": 0.5177,
+      "step": 1173
+    },
+    {
+      "epoch": 2.87041564792176,
+      "grad_norm": 0.17295647417142113,
+      "learning_rate": 4.561674101184891e-06,
+      "loss": 0.4766,
+      "step": 1174
+    },
+    {
+      "epoch": 2.8728606356968216,
+      "grad_norm": 0.18327794991110657,
+      "learning_rate": 4.560949735156438e-06,
+      "loss": 0.3641,
+      "step": 1175
+    },
+    {
+      "epoch": 2.8753056234718826,
+      "grad_norm": 0.16893337654175863,
+      "learning_rate": 4.560224828699786e-06,
+      "loss": 0.5531,
+      "step": 1176
+    },
+    {
+      "epoch": 2.8777506112469435,
+      "grad_norm": 0.11770490149142304,
+      "learning_rate": 4.5594993820050236e-06,
+      "loss": 0.5493,
+      "step": 1177
+    },
+    {
+      "epoch": 2.880195599022005,
+      "grad_norm": 0.13852478666511223,
+      "learning_rate": 4.558773395262378e-06,
+      "loss": 0.4518,
+      "step": 1178
+    },
+    {
+      "epoch": 2.882640586797066,
+      "grad_norm": 0.16954234879886837,
+      "learning_rate": 4.55804686866222e-06,
+      "loss": 0.5094,
+      "step": 1179
+    },
+    {
+      "epoch": 2.885085574572127,
+      "grad_norm": 0.1553310105636114,
+      "learning_rate": 4.557319802395061e-06,
+      "loss": 0.8109,
+      "step": 1180
+    },
+    {
+      "epoch": 2.8875305623471883,
+      "grad_norm": 0.13436894812173883,
+      "learning_rate": 4.556592196651556e-06,
+      "loss": 0.5678,
+      "step": 1181
+    },
+    {
+      "epoch": 2.8899755501222493,
+      "grad_norm": 0.11568213342939627,
+      "learning_rate": 4.555864051622497e-06,
+      "loss": 0.4483,
+      "step": 1182
+    },
+    {
+      "epoch": 2.8924205378973102,
+      "grad_norm": 0.2173866815381029,
+      "learning_rate": 4.555135367498824e-06,
+      "loss": 0.6624,
+      "step": 1183
+    },
+    {
+      "epoch": 2.8948655256723717,
+      "grad_norm": 0.11794853163201566,
+      "learning_rate": 4.554406144471612e-06,
+      "loss": 0.4613,
+      "step": 1184
+    },
+    {
+      "epoch": 2.8973105134474326,
+      "grad_norm": 0.11660525071756107,
+      "learning_rate": 4.553676382732081e-06,
+      "loss": 0.6771,
+      "step": 1185
+    },
+    {
+      "epoch": 2.899755501222494,
+      "grad_norm": 0.17638501613839083,
+      "learning_rate": 4.552946082471591e-06,
+      "loss": 0.5546,
+      "step": 1186
+    },
+    {
+      "epoch": 2.902200488997555,
+      "grad_norm": 0.1388948745682831,
+      "learning_rate": 4.5522152438816424e-06,
+      "loss": 0.5451,
+      "step": 1187
+    },
+    {
+      "epoch": 2.904645476772616,
+      "grad_norm": 0.19680729431463675,
+      "learning_rate": 4.55148386715388e-06,
+      "loss": 1.0006,
+      "step": 1188
+    },
+    {
+      "epoch": 2.9070904645476774,
+      "grad_norm": 0.15430184633797764,
+      "learning_rate": 4.550751952480086e-06,
+      "loss": 0.4794,
+      "step": 1189
+    },
+    {
+      "epoch": 2.9095354523227384,
+      "grad_norm": 0.19232521175241812,
+      "learning_rate": 4.550019500052186e-06,
+      "loss": 0.4749,
+      "step": 1190
+    },
+    {
+      "epoch": 2.9119804400977998,
+      "grad_norm": 0.10232736836420402,
+      "learning_rate": 4.549286510062245e-06,
+      "loss": 0.2845,
+      "step": 1191
+    },
+    {
+      "epoch": 2.9144254278728607,
+      "grad_norm": 0.12211782695807297,
+      "learning_rate": 4.54855298270247e-06,
+      "loss": 0.5865,
+      "step": 1192
+    },
+    {
+      "epoch": 2.9168704156479217,
+      "grad_norm": 0.219019708747965,
+      "learning_rate": 4.547818918165209e-06,
+      "loss": 0.4852,
+      "step": 1193
+    },
+    {
+      "epoch": 2.919315403422983,
+      "grad_norm": 0.21299814297361516,
+      "learning_rate": 4.54708431664295e-06,
+      "loss": 1.1898,
+      "step": 1194
+    },
+    {
+      "epoch": 2.921760391198044,
+      "grad_norm": 0.26406013347393353,
+      "learning_rate": 4.546349178328322e-06,
+      "loss": 0.7653,
+      "step": 1195
+    },
+    {
+      "epoch": 2.924205378973105,
+      "grad_norm": 0.26119771654190144,
+      "learning_rate": 4.545613503414097e-06,
+      "loss": 0.4015,
+      "step": 1196
+    },
+    {
+      "epoch": 2.9266503667481665,
+      "grad_norm": 0.12431891239243097,
+      "learning_rate": 4.5448772920931835e-06,
+      "loss": 0.4894,
+      "step": 1197
+    },
+    {
+      "epoch": 2.9290953545232274,
+      "grad_norm": 0.1393440146323019,
+      "learning_rate": 4.544140544558634e-06,
+      "loss": 0.6626,
+      "step": 1198
+    },
+    {
+      "epoch": 2.9315403422982884,
+      "grad_norm": 0.15876534649796736,
+      "learning_rate": 4.543403261003642e-06,
+      "loss": 0.5137,
+      "step": 1199
+    },
+    {
+      "epoch": 2.93398533007335,
+      "grad_norm": 0.20863670265453652,
+      "learning_rate": 4.542665441621537e-06,
+      "loss": 0.8357,
+      "step": 1200
+    },
+    {
+      "epoch": 2.936430317848411,
+      "grad_norm": 0.17858106152611558,
+      "learning_rate": 4.5419270866057944e-06,
+      "loss": 0.7318,
+      "step": 1201
+    },
+    {
+      "epoch": 2.9388753056234718,
+      "grad_norm": 0.17963165759212266,
+      "learning_rate": 4.541188196150027e-06,
+      "loss": 0.4322,
+      "step": 1202
+    },
+    {
+      "epoch": 2.941320293398533,
+      "grad_norm": 0.0869756874887522,
+      "learning_rate": 4.540448770447989e-06,
+      "loss": 0.4098,
+      "step": 1203
+    },
+    {
+      "epoch": 2.943765281173594,
+      "grad_norm": 0.2315108456952936,
+      "learning_rate": 4.539708809693575e-06,
+      "loss": 1.0613,
+      "step": 1204
+    },
+    {
+      "epoch": 2.946210268948655,
+      "grad_norm": 0.22025547790682168,
+      "learning_rate": 4.53896831408082e-06,
+      "loss": 0.5298,
+      "step": 1205
+    },
+    {
+      "epoch": 2.9486552567237165,
+      "grad_norm": 0.16179186752052616,
+      "learning_rate": 4.538227283803897e-06,
+      "loss": 0.6957,
+      "step": 1206
+    },
+    {
+      "epoch": 2.9511002444987775,
+      "grad_norm": 0.2619071044019018,
+      "learning_rate": 4.537485719057123e-06,
+      "loss": 0.8555,
+      "step": 1207
+    },
+    {
+      "epoch": 2.9535452322738385,
+      "grad_norm": 0.11835943841204094,
+      "learning_rate": 4.536743620034952e-06,
+      "loss": 0.4937,
+      "step": 1208
+    },
+    {
+      "epoch": 2.9559902200489,
+      "grad_norm": 0.1297255155184526,
+      "learning_rate": 4.536000986931981e-06,
+      "loss": 0.7769,
+      "step": 1209
+    },
+    {
+      "epoch": 2.958435207823961,
+      "grad_norm": 0.2595960675736413,
+      "learning_rate": 4.535257819942944e-06,
+      "loss": 0.542,
+      "step": 1210
+    },
+    {
+      "epoch": 2.960880195599022,
+      "grad_norm": 0.19099562458836739,
+      "learning_rate": 4.534514119262717e-06,
+      "loss": 0.5471,
+      "step": 1211
+    },
+    {
+      "epoch": 2.9633251833740832,
+      "grad_norm": 0.14805890495778595,
+      "learning_rate": 4.533769885086315e-06,
+      "loss": 0.556,
+      "step": 1212
+    },
+    {
+      "epoch": 2.965770171149144,
+      "grad_norm": 0.2310952050169163,
+      "learning_rate": 4.533025117608893e-06,
+      "loss": 0.4327,
+      "step": 1213
+    },
+    {
+      "epoch": 2.968215158924205,
+      "grad_norm": 0.13329291154467213,
+      "learning_rate": 4.532279817025746e-06,
+      "loss": 0.5053,
+      "step": 1214
+    },
+    {
+      "epoch": 2.9706601466992666,
+      "grad_norm": 0.16297651212210837,
+      "learning_rate": 4.53153398353231e-06,
+      "loss": 0.4704,
+      "step": 1215
+    },
+    {
+      "epoch": 2.9731051344743276,
+      "grad_norm": 0.15181748206516257,
+      "learning_rate": 4.53078761732416e-06,
+      "loss": 0.4626,
+      "step": 1216
+    },
+    {
+      "epoch": 2.9755501222493885,
+      "grad_norm": 0.18555734990057685,
+      "learning_rate": 4.530040718597008e-06,
+      "loss": 0.4609,
+      "step": 1217
+    },
+    {
+      "epoch": 2.97799511002445,
+      "grad_norm": 0.2179231804133384,
+      "learning_rate": 4.529293287546709e-06,
+      "loss": 0.7654,
+      "step": 1218
+    },
+    {
+      "epoch": 2.980440097799511,
+      "grad_norm": 0.14695249721169837,
+      "learning_rate": 4.528545324369257e-06,
+      "loss": 0.6068,
+      "step": 1219
+    },
+    {
+      "epoch": 2.982885085574572,
+      "grad_norm": 0.17154534442966718,
+      "learning_rate": 4.527796829260784e-06,
+      "loss": 0.5653,
+      "step": 1220
+    },
+    {
+      "epoch": 2.9853300733496333,
+      "grad_norm": 0.1884027033641388,
+      "learning_rate": 4.527047802417565e-06,
+      "loss": 0.7649,
+      "step": 1221
+    },
+    {
+      "epoch": 2.9877750611246943,
+      "grad_norm": 0.24625535458810247,
+      "learning_rate": 4.526298244036009e-06,
+      "loss": 0.5049,
+      "step": 1222
+    },
+    {
+      "epoch": 2.9902200488997552,
+      "grad_norm": 0.11502702460276133,
+      "learning_rate": 4.525548154312669e-06,
+      "loss": 0.301,
+      "step": 1223
+    },
+    {
+      "epoch": 2.9926650366748166,
+      "grad_norm": 0.1405433891124375,
+      "learning_rate": 4.5247975334442354e-06,
+      "loss": 0.4516,
+      "step": 1224
+    },
+    {
+      "epoch": 2.9951100244498776,
+      "grad_norm": 0.12385925983270608,
+      "learning_rate": 4.5240463816275386e-06,
+      "loss": 0.6481,
+      "step": 1225
+    },
+    {
+      "epoch": 2.997555012224939,
+      "grad_norm": 0.17135102010592884,
+      "learning_rate": 4.523294699059546e-06,
+      "loss": 0.4397,
+      "step": 1226
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.17556446279068616,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.4461,
+      "step": 1227
+    },
+    {
+      "epoch": 3.002444987775061,
+      "grad_norm": 0.22017811012032287,
+      "learning_rate": 4.521789742458252e-06,
+      "loss": 0.7488,
+      "step": 1228
+    },
+    {
+      "epoch": 3.0048899755501224,
+      "grad_norm": 0.14308431153355256,
+      "learning_rate": 4.521036468819585e-06,
+      "loss": 0.3856,
+      "step": 1229
+    },
+    {
+      "epoch": 3.0073349633251834,
+      "grad_norm": 0.13413740267855528,
+      "learning_rate": 4.520282665218889e-06,
+      "loss": 0.5433,
+      "step": 1230
+    },
+    {
+      "epoch": 3.0097799511002443,
+      "grad_norm": 0.13691744977686368,
+      "learning_rate": 4.519528331853833e-06,
+      "loss": 0.6822,
+      "step": 1231
+    },
+    {
+      "epoch": 3.0122249388753057,
+      "grad_norm": 0.14155039512126036,
+      "learning_rate": 4.518773468922219e-06,
+      "loss": 0.6341,
+      "step": 1232
+    },
+    {
+      "epoch": 3.0146699266503667,
+      "grad_norm": 0.17764974230942743,
+      "learning_rate": 4.518018076621987e-06,
+      "loss": 0.744,
+      "step": 1233
+    },
+    {
+      "epoch": 3.0171149144254277,
+      "grad_norm": 0.11359776688505224,
+      "learning_rate": 4.517262155151222e-06,
+      "loss": 0.3236,
+      "step": 1234
+    },
+    {
+      "epoch": 3.019559902200489,
+      "grad_norm": 0.17191503122410867,
+      "learning_rate": 4.516505704708141e-06,
+      "loss": 0.5528,
+      "step": 1235
+    },
+    {
+      "epoch": 3.02200488997555,
+      "grad_norm": 0.18983494459128966,
+      "learning_rate": 4.515748725491103e-06,
+      "loss": 0.4683,
+      "step": 1236
+    },
+    {
+      "epoch": 3.0244498777506115,
+      "grad_norm": 0.23179954830868868,
+      "learning_rate": 4.514991217698606e-06,
+      "loss": 0.4708,
+      "step": 1237
+    },
+    {
+      "epoch": 3.0268948655256724,
+      "grad_norm": 0.12989589330009677,
+      "learning_rate": 4.514233181529287e-06,
+      "loss": 0.6847,
+      "step": 1238
+    },
+    {
+      "epoch": 3.0293398533007334,
+      "grad_norm": 0.1625443865259557,
+      "learning_rate": 4.51347461718192e-06,
+      "loss": 0.7301,
+      "step": 1239
+    },
+    {
+      "epoch": 3.031784841075795,
+      "grad_norm": 0.1443303118696757,
+      "learning_rate": 4.512715524855415e-06,
+      "loss": 0.6634,
+      "step": 1240
+    },
+    {
+      "epoch": 3.034229828850856,
+      "grad_norm": 0.1800621731396222,
+      "learning_rate": 4.511955904748827e-06,
+      "loss": 0.8164,
+      "step": 1241
+    },
+    {
+      "epoch": 3.0366748166259168,
+      "grad_norm": 0.1524683088085164,
+      "learning_rate": 4.511195757061344e-06,
+      "loss": 0.5134,
+      "step": 1242
+    },
+    {
+      "epoch": 3.039119804400978,
+      "grad_norm": 0.23900869253038692,
+      "learning_rate": 4.510435081992295e-06,
+      "loss": 1.0166,
+      "step": 1243
+    },
+    {
+      "epoch": 3.041564792176039,
+      "grad_norm": 0.1339810297782538,
+      "learning_rate": 4.5096738797411445e-06,
+      "loss": 0.3627,
+      "step": 1244
+    },
+    {
+      "epoch": 3.0440097799511,
+      "grad_norm": 0.22190862069788,
+      "learning_rate": 4.508912150507499e-06,
+      "loss": 0.4595,
+      "step": 1245
+    },
+    {
+      "epoch": 3.0464547677261615,
+      "grad_norm": 0.09981991541901061,
+      "learning_rate": 4.508149894491102e-06,
+      "loss": 0.4481,
+      "step": 1246
+    },
+    {
+      "epoch": 3.0488997555012225,
+      "grad_norm": 0.12651881824722214,
+      "learning_rate": 4.507387111891832e-06,
+      "loss": 0.4101,
+      "step": 1247
+    },
+    {
+      "epoch": 3.0513447432762835,
+      "grad_norm": 0.17516120702364987,
+      "learning_rate": 4.506623802909711e-06,
+      "loss": 0.6584,
+      "step": 1248
+    },
+    {
+      "epoch": 3.053789731051345,
+      "grad_norm": 0.14734649247028397,
+      "learning_rate": 4.505859967744893e-06,
+      "loss": 0.3531,
+      "step": 1249
+    },
+    {
+      "epoch": 3.056234718826406,
+      "grad_norm": 0.1609056089426988,
+      "learning_rate": 4.505095606597675e-06,
+      "loss": 0.5998,
+      "step": 1250
+    },
+    {
+      "epoch": 3.058679706601467,
+      "grad_norm": 0.17704999441932256,
+      "learning_rate": 4.504330719668489e-06,
+      "loss": 0.7456,
+      "step": 1251
+    },
+    {
+      "epoch": 3.0611246943765282,
+      "grad_norm": 0.2341202807189493,
+      "learning_rate": 4.503565307157908e-06,
+      "loss": 0.9408,
+      "step": 1252
+    },
+    {
+      "epoch": 3.063569682151589,
+      "grad_norm": 0.15203998286845186,
+      "learning_rate": 4.502799369266638e-06,
+      "loss": 0.4612,
+      "step": 1253
+    },
+    {
+      "epoch": 3.06601466992665,
+      "grad_norm": 0.16767054732418843,
+      "learning_rate": 4.502032906195527e-06,
+      "loss": 0.6231,
+      "step": 1254
+    },
+    {
+      "epoch": 3.0684596577017116,
+      "grad_norm": 0.18127768128806845,
+      "learning_rate": 4.501265918145559e-06,
+      "loss": 0.4582,
+      "step": 1255
+    },
+    {
+      "epoch": 3.0709046454767726,
+      "grad_norm": 0.11480184930677406,
+      "learning_rate": 4.500498405317855e-06,
+      "loss": 0.5081,
+      "step": 1256
+    },
+    {
+      "epoch": 3.0733496332518335,
+      "grad_norm": 0.1903991347399298,
+      "learning_rate": 4.499730367913674e-06,
+      "loss": 0.4764,
+      "step": 1257
+    },
+    {
+      "epoch": 3.075794621026895,
+      "grad_norm": 0.2165356089746921,
+      "learning_rate": 4.498961806134416e-06,
+      "loss": 0.7513,
+      "step": 1258
+    },
+    {
+      "epoch": 3.078239608801956,
+      "grad_norm": 0.19112686737044238,
+      "learning_rate": 4.498192720181613e-06,
+      "loss": 0.5037,
+      "step": 1259
+    },
+    {
+      "epoch": 3.0806845965770173,
+      "grad_norm": 0.16812609180766974,
+      "learning_rate": 4.497423110256936e-06,
+      "loss": 0.5316,
+      "step": 1260
+    },
+    {
+      "epoch": 3.0831295843520783,
+      "grad_norm": 0.1739822648682703,
+      "learning_rate": 4.496652976562197e-06,
+      "loss": 0.7983,
+      "step": 1261
+    },
+    {
+      "epoch": 3.0855745721271393,
+      "grad_norm": 0.08510611049043375,
+      "learning_rate": 4.495882319299341e-06,
+      "loss": 0.3519,
+      "step": 1262
+    },
+    {
+      "epoch": 3.0880195599022007,
+      "grad_norm": 0.14984509142203617,
+      "learning_rate": 4.495111138670453e-06,
+      "loss": 0.4065,
+      "step": 1263
+    },
+    {
+      "epoch": 3.0904645476772616,
+      "grad_norm": 0.13732612078852077,
+      "learning_rate": 4.494339434877752e-06,
+      "loss": 0.5618,
+      "step": 1264
+    },
+    {
+      "epoch": 3.0929095354523226,
+      "grad_norm": 0.16662068437497599,
+      "learning_rate": 4.493567208123599e-06,
+      "loss": 0.875,
+      "step": 1265
+    },
+    {
+      "epoch": 3.095354523227384,
+      "grad_norm": 0.21586131738190212,
+      "learning_rate": 4.492794458610489e-06,
+      "loss": 0.579,
+      "step": 1266
+    },
+    {
+      "epoch": 3.097799511002445,
+      "grad_norm": 0.16120181201505424,
+      "learning_rate": 4.492021186541054e-06,
+      "loss": 0.3902,
+      "step": 1267
+    },
+    {
+      "epoch": 3.100244498777506,
+      "grad_norm": 0.1710438386660066,
+      "learning_rate": 4.491247392118064e-06,
+      "loss": 0.7774,
+      "step": 1268
+    },
+    {
+      "epoch": 3.1026894865525674,
+      "grad_norm": 0.1781472626823595,
+      "learning_rate": 4.490473075544425e-06,
+      "loss": 0.4266,
+      "step": 1269
+    },
+    {
+      "epoch": 3.1051344743276283,
+      "grad_norm": 0.15279216474346197,
+      "learning_rate": 4.4896982370231814e-06,
+      "loss": 0.5912,
+      "step": 1270
+    },
+    {
+      "epoch": 3.1075794621026893,
+      "grad_norm": 0.2513092610834609,
+      "learning_rate": 4.4889228767575125e-06,
+      "loss": 0.5008,
+      "step": 1271
+    },
+    {
+      "epoch": 3.1100244498777507,
+      "grad_norm": 0.1334857450489098,
+      "learning_rate": 4.488146994950736e-06,
+      "loss": 0.4526,
+      "step": 1272
+    },
+    {
+      "epoch": 3.1124694376528117,
+      "grad_norm": 0.15430155417138328,
+      "learning_rate": 4.487370591806307e-06,
+      "loss": 0.5617,
+      "step": 1273
+    },
+    {
+      "epoch": 3.1149144254278727,
+      "grad_norm": 0.18054157010670074,
+      "learning_rate": 4.4865936675278146e-06,
+      "loss": 0.4552,
+      "step": 1274
+    },
+    {
+      "epoch": 3.117359413202934,
+      "grad_norm": 0.15352619819594984,
+      "learning_rate": 4.485816222318986e-06,
+      "loss": 0.5506,
+      "step": 1275
+    },
+    {
+      "epoch": 3.119804400977995,
+      "grad_norm": 0.16326556676056037,
+      "learning_rate": 4.4850382563836866e-06,
+      "loss": 0.5217,
+      "step": 1276
+    },
+    {
+      "epoch": 3.122249388753056,
+      "grad_norm": 0.20044497673056416,
+      "learning_rate": 4.484259769925917e-06,
+      "loss": 0.9806,
+      "step": 1277
+    },
+    {
+      "epoch": 3.1246943765281174,
+      "grad_norm": 0.21235811569160104,
+      "learning_rate": 4.483480763149812e-06,
+      "loss": 0.5584,
+      "step": 1278
+    },
+    {
+      "epoch": 3.1271393643031784,
+      "grad_norm": 0.1276886046716671,
+      "learning_rate": 4.482701236259647e-06,
+      "loss": 0.4395,
+      "step": 1279
+    },
+    {
+      "epoch": 3.12958435207824,
+      "grad_norm": 0.12622011128561542,
+      "learning_rate": 4.48192118945983e-06,
+      "loss": 0.4504,
+      "step": 1280
+    },
+    {
+      "epoch": 3.132029339853301,
+      "grad_norm": 0.1577682178159191,
+      "learning_rate": 4.48114062295491e-06,
+      "loss": 0.9822,
+      "step": 1281
+    },
+    {
+      "epoch": 3.1344743276283618,
+      "grad_norm": 0.15604296175203658,
+      "learning_rate": 4.480359536949566e-06,
+      "loss": 0.5444,
+      "step": 1282
+    },
+    {
+      "epoch": 3.136919315403423,
+      "grad_norm": 0.20696327454044353,
+      "learning_rate": 4.47957793164862e-06,
+      "loss": 1.1878,
+      "step": 1283
+    },
+    {
+      "epoch": 3.139364303178484,
+      "grad_norm": 0.2562488852692517,
+      "learning_rate": 4.478795807257024e-06,
+      "loss": 1.2973,
+      "step": 1284
+    },
+    {
+      "epoch": 3.141809290953545,
+      "grad_norm": 0.1441785371357995,
+      "learning_rate": 4.4780131639798695e-06,
+      "loss": 0.6982,
+      "step": 1285
+    },
+    {
+      "epoch": 3.1442542787286065,
+      "grad_norm": 0.1385038619084508,
+      "learning_rate": 4.477230002022385e-06,
+      "loss": 0.5676,
+      "step": 1286
+    },
+    {
+      "epoch": 3.1466992665036675,
+      "grad_norm": 0.17215628950115078,
+      "learning_rate": 4.476446321589932e-06,
+      "loss": 0.4852,
+      "step": 1287
+    },
+    {
+      "epoch": 3.1491442542787285,
+      "grad_norm": 0.18127179781756417,
+      "learning_rate": 4.4756621228880105e-06,
+      "loss": 0.4488,
+      "step": 1288
+    },
+    {
+      "epoch": 3.15158924205379,
+      "grad_norm": 0.16891594311283806,
+      "learning_rate": 4.474877406122254e-06,
+      "loss": 0.5933,
+      "step": 1289
+    },
+    {
+      "epoch": 3.154034229828851,
+      "grad_norm": 0.14345758158197966,
+      "learning_rate": 4.4740921714984346e-06,
+      "loss": 0.7536,
+      "step": 1290
+    },
+    {
+      "epoch": 3.156479217603912,
+      "grad_norm": 0.1642820279512776,
+      "learning_rate": 4.473306419222458e-06,
+      "loss": 0.8624,
+      "step": 1291
+    },
+    {
+      "epoch": 3.1589242053789732,
+      "grad_norm": 0.15664337588829666,
+      "learning_rate": 4.472520149500365e-06,
+      "loss": 0.5104,
+      "step": 1292
+    },
+    {
+      "epoch": 3.161369193154034,
+      "grad_norm": 0.19938166287186623,
+      "learning_rate": 4.471733362538337e-06,
+      "loss": 0.9768,
+      "step": 1293
+    },
+    {
+      "epoch": 3.163814180929095,
+      "grad_norm": 0.19656281672816017,
+      "learning_rate": 4.470946058542685e-06,
+      "loss": 0.6707,
+      "step": 1294
+    },
+    {
+      "epoch": 3.1662591687041566,
+      "grad_norm": 0.18351179065718753,
+      "learning_rate": 4.4701582377198594e-06,
+      "loss": 0.4688,
+      "step": 1295
+    },
+    {
+      "epoch": 3.1687041564792175,
+      "grad_norm": 0.15249767677811685,
+      "learning_rate": 4.469369900276443e-06,
+      "loss": 0.4937,
+      "step": 1296
+    },
+    {
+      "epoch": 3.1711491442542785,
+      "grad_norm": 0.17664107501039134,
+      "learning_rate": 4.4685810464191584e-06,
+      "loss": 0.8433,
+      "step": 1297
+    },
+    {
+      "epoch": 3.17359413202934,
+      "grad_norm": 0.17313713762115512,
+      "learning_rate": 4.46779167635486e-06,
+      "loss": 0.6598,
+      "step": 1298
+    },
+    {
+      "epoch": 3.176039119804401,
+      "grad_norm": 0.1207350934618037,
+      "learning_rate": 4.467001790290538e-06,
+      "loss": 0.4687,
+      "step": 1299
+    },
+    {
+      "epoch": 3.178484107579462,
+      "grad_norm": 0.16127282387547937,
+      "learning_rate": 4.466211388433319e-06,
+      "loss": 0.4434,
+      "step": 1300
+    },
+    {
+      "epoch": 3.1809290953545233,
+      "grad_norm": 0.16511133323502677,
+      "learning_rate": 4.465420470990466e-06,
+      "loss": 0.6124,
+      "step": 1301
+    },
+    {
+      "epoch": 3.1833740831295843,
+      "grad_norm": 0.13963470303586623,
+      "learning_rate": 4.464629038169373e-06,
+      "loss": 0.3694,
+      "step": 1302
+    },
+    {
+      "epoch": 3.1858190709046457,
+      "grad_norm": 0.22506221024699372,
+      "learning_rate": 4.463837090177573e-06,
+      "loss": 1.0833,
+      "step": 1303
+    },
+    {
+      "epoch": 3.1882640586797066,
+      "grad_norm": 0.21540378537989388,
+      "learning_rate": 4.463044627222734e-06,
+      "loss": 0.614,
+      "step": 1304
+    },
+    {
+      "epoch": 3.1907090464547676,
+      "grad_norm": 0.1427782779960142,
+      "learning_rate": 4.462251649512656e-06,
+      "loss": 0.5039,
+      "step": 1305
+    },
+    {
+      "epoch": 3.193154034229829,
+      "grad_norm": 0.20557753007232604,
+      "learning_rate": 4.461458157255278e-06,
+      "loss": 0.9929,
+      "step": 1306
+    },
+    {
+      "epoch": 3.19559902200489,
+      "grad_norm": 0.18081999185390915,
+      "learning_rate": 4.46066415065867e-06,
+      "loss": 0.4553,
+      "step": 1307
+    },
+    {
+      "epoch": 3.198044009779951,
+      "grad_norm": 0.11179310092695381,
+      "learning_rate": 4.459869629931039e-06,
+      "loss": 0.4595,
+      "step": 1308
+    },
+    {
+      "epoch": 3.2004889975550124,
+      "grad_norm": 0.1900949783350105,
+      "learning_rate": 4.4590745952807275e-06,
+      "loss": 0.6683,
+      "step": 1309
+    },
+    {
+      "epoch": 3.2029339853300733,
+      "grad_norm": 0.25915575506080024,
+      "learning_rate": 4.4582790469162116e-06,
+      "loss": 0.5207,
+      "step": 1310
+    },
+    {
+      "epoch": 3.2053789731051343,
+      "grad_norm": 0.17578677687295075,
+      "learning_rate": 4.457482985046101e-06,
+      "loss": 0.7039,
+      "step": 1311
+    },
+    {
+      "epoch": 3.2078239608801957,
+      "grad_norm": 0.1584048253811272,
+      "learning_rate": 4.456686409879143e-06,
+      "loss": 0.704,
+      "step": 1312
+    },
+    {
+      "epoch": 3.2102689486552567,
+      "grad_norm": 0.10762403748793255,
+      "learning_rate": 4.455889321624217e-06,
+      "loss": 0.8104,
+      "step": 1313
+    },
+    {
+      "epoch": 3.2127139364303177,
+      "grad_norm": 0.1385164779105021,
+      "learning_rate": 4.4550917204903384e-06,
+      "loss": 0.9023,
+      "step": 1314
+    },
+    {
+      "epoch": 3.215158924205379,
+      "grad_norm": 0.26520266059550096,
+      "learning_rate": 4.454293606686655e-06,
+      "loss": 0.4816,
+      "step": 1315
+    },
+    {
+      "epoch": 3.21760391198044,
+      "grad_norm": 0.18442310939643727,
+      "learning_rate": 4.4534949804224516e-06,
+      "loss": 0.463,
+      "step": 1316
+    },
+    {
+      "epoch": 3.220048899755501,
+      "grad_norm": 0.15599800908581243,
+      "learning_rate": 4.452695841907146e-06,
+      "loss": 0.4833,
+      "step": 1317
+    },
+    {
+      "epoch": 3.2224938875305624,
+      "grad_norm": 0.14586960791085962,
+      "learning_rate": 4.451896191350292e-06,
+      "loss": 0.6932,
+      "step": 1318
+    },
+    {
+      "epoch": 3.2249388753056234,
+      "grad_norm": 0.12815509749195497,
+      "learning_rate": 4.451096028961574e-06,
+      "loss": 0.5276,
+      "step": 1319
+    },
+    {
+      "epoch": 3.227383863080685,
+      "grad_norm": 0.19605937567101867,
+      "learning_rate": 4.450295354950814e-06,
+      "loss": 0.4353,
+      "step": 1320
+    },
+    {
+      "epoch": 3.229828850855746,
+      "grad_norm": 0.21004343334280354,
+      "learning_rate": 4.449494169527967e-06,
+      "loss": 0.444,
+      "step": 1321
+    },
+    {
+      "epoch": 3.2322738386308068,
+      "grad_norm": 0.13561312683289953,
+      "learning_rate": 4.448692472903123e-06,
+      "loss": 0.7136,
+      "step": 1322
+    },
+    {
+      "epoch": 3.234718826405868,
+      "grad_norm": 0.15234110978656384,
+      "learning_rate": 4.447890265286504e-06,
+      "loss": 0.445,
+      "step": 1323
+    },
+    {
+      "epoch": 3.237163814180929,
+      "grad_norm": 0.24344737661470825,
+      "learning_rate": 4.447087546888467e-06,
+      "loss": 0.4139,
+      "step": 1324
+    },
+    {
+      "epoch": 3.23960880195599,
+      "grad_norm": 0.18254591203805187,
+      "learning_rate": 4.446284317919503e-06,
+      "loss": 0.7247,
+      "step": 1325
+    },
+    {
+      "epoch": 3.2420537897310515,
+      "grad_norm": 0.23369416459868653,
+      "learning_rate": 4.445480578590239e-06,
+      "loss": 0.7487,
+      "step": 1326
+    },
+    {
+      "epoch": 3.2444987775061125,
+      "grad_norm": 0.15565741297396668,
+      "learning_rate": 4.444676329111432e-06,
+      "loss": 0.7647,
+      "step": 1327
+    },
+    {
+      "epoch": 3.2469437652811735,
+      "grad_norm": 0.1189697691681801,
+      "learning_rate": 4.443871569693976e-06,
+      "loss": 0.4374,
+      "step": 1328
+    },
+    {
+      "epoch": 3.249388753056235,
+      "grad_norm": 0.14522022741890428,
+      "learning_rate": 4.443066300548896e-06,
+      "loss": 0.4156,
+      "step": 1329
+    },
+    {
+      "epoch": 3.251833740831296,
+      "grad_norm": 0.16398341849098805,
+      "learning_rate": 4.442260521887353e-06,
+      "loss": 0.8357,
+      "step": 1330
+    },
+    {
+      "epoch": 3.254278728606357,
+      "grad_norm": 0.1762782723475786,
+      "learning_rate": 4.44145423392064e-06,
+      "loss": 0.5304,
+      "step": 1331
+    },
+    {
+      "epoch": 3.256723716381418,
+      "grad_norm": 0.1922090385737338,
+      "learning_rate": 4.440647436860184e-06,
+      "loss": 0.5482,
+      "step": 1332
+    },
+    {
+      "epoch": 3.259168704156479,
+      "grad_norm": 0.12490568420733943,
+      "learning_rate": 4.439840130917546e-06,
+      "loss": 0.4291,
+      "step": 1333
+    },
+    {
+      "epoch": 3.26161369193154,
+      "grad_norm": 0.15498586251161028,
+      "learning_rate": 4.43903231630442e-06,
+      "loss": 0.4267,
+      "step": 1334
+    },
+    {
+      "epoch": 3.2640586797066016,
+      "grad_norm": 0.20029005869070945,
+      "learning_rate": 4.4382239932326344e-06,
+      "loss": 1.0601,
+      "step": 1335
+    },
+    {
+      "epoch": 3.2665036674816625,
+      "grad_norm": 0.15807953353238713,
+      "learning_rate": 4.437415161914148e-06,
+      "loss": 0.4628,
+      "step": 1336
+    },
+    {
+      "epoch": 3.2689486552567235,
+      "grad_norm": 0.1333337603647761,
+      "learning_rate": 4.436605822561056e-06,
+      "loss": 0.4927,
+      "step": 1337
+    },
+    {
+      "epoch": 3.271393643031785,
+      "grad_norm": 0.17488025046141334,
+      "learning_rate": 4.435795975385587e-06,
+      "loss": 0.4456,
+      "step": 1338
+    },
+    {
+      "epoch": 3.273838630806846,
+      "grad_norm": 0.21241879714365483,
+      "learning_rate": 4.4349856206001e-06,
+      "loss": 0.4399,
+      "step": 1339
+    },
+    {
+      "epoch": 3.276283618581907,
+      "grad_norm": 0.1749284252524475,
+      "learning_rate": 4.434174758417088e-06,
+      "loss": 0.4009,
+      "step": 1340
+    },
+    {
+      "epoch": 3.2787286063569683,
+      "grad_norm": 0.3018037054563145,
+      "learning_rate": 4.4333633890491786e-06,
+      "loss": 1.102,
+      "step": 1341
+    },
+    {
+      "epoch": 3.2811735941320292,
+      "grad_norm": 0.15889734041520165,
+      "learning_rate": 4.432551512709132e-06,
+      "loss": 0.4829,
+      "step": 1342
+    },
+    {
+      "epoch": 3.28361858190709,
+      "grad_norm": 0.1820426500272768,
+      "learning_rate": 4.4317391296098385e-06,
+      "loss": 0.5921,
+      "step": 1343
+    },
+    {
+      "epoch": 3.2860635696821516,
+      "grad_norm": 0.15985652528319985,
+      "learning_rate": 4.430926239964327e-06,
+      "loss": 0.7949,
+      "step": 1344
+    },
+    {
+      "epoch": 3.2885085574572126,
+      "grad_norm": 0.21200904226124137,
+      "learning_rate": 4.430112843985753e-06,
+      "loss": 0.542,
+      "step": 1345
+    },
+    {
+      "epoch": 3.290953545232274,
+      "grad_norm": 0.2019499855173706,
+      "learning_rate": 4.429298941887408e-06,
+      "loss": 0.3833,
+      "step": 1346
+    },
+    {
+      "epoch": 3.293398533007335,
+      "grad_norm": 0.20393170602186828,
+      "learning_rate": 4.4284845338827165e-06,
+      "loss": 0.5783,
+      "step": 1347
+    },
+    {
+      "epoch": 3.295843520782396,
+      "grad_norm": 0.28244006800033244,
+      "learning_rate": 4.4276696201852346e-06,
+      "loss": 0.4809,
+      "step": 1348
+    },
+    {
+      "epoch": 3.2982885085574574,
+      "grad_norm": 0.1301736319744728,
+      "learning_rate": 4.426854201008651e-06,
+      "loss": 0.4721,
+      "step": 1349
+    },
+    {
+      "epoch": 3.3007334963325183,
+      "grad_norm": 0.20962482316826106,
+      "learning_rate": 4.426038276566788e-06,
+      "loss": 0.3922,
+      "step": 1350
+    },
+    {
+      "epoch": 3.3031784841075793,
+      "grad_norm": 0.16291490072675338,
+      "learning_rate": 4.4252218470735984e-06,
+      "loss": 0.3659,
+      "step": 1351
+    },
+    {
+      "epoch": 3.3056234718826407,
+      "grad_norm": 0.29003709329839317,
+      "learning_rate": 4.424404912743171e-06,
+      "loss": 1.108,
+      "step": 1352
+    },
+    {
+      "epoch": 3.3080684596577017,
+      "grad_norm": 0.13014565347766072,
+      "learning_rate": 4.423587473789722e-06,
+      "loss": 0.4388,
+      "step": 1353
+    },
+    {
+      "epoch": 3.3105134474327627,
+      "grad_norm": 0.15602013644278834,
+      "learning_rate": 4.422769530427604e-06,
+      "loss": 0.4812,
+      "step": 1354
+    },
+    {
+      "epoch": 3.312958435207824,
+      "grad_norm": 0.17247109398420665,
+      "learning_rate": 4.421951082871301e-06,
+      "loss": 0.5419,
+      "step": 1355
+    },
+    {
+      "epoch": 3.315403422982885,
+      "grad_norm": 0.24803878066460436,
+      "learning_rate": 4.421132131335427e-06,
+      "loss": 0.7113,
+      "step": 1356
+    },
+    {
+      "epoch": 3.3178484107579465,
+      "grad_norm": 0.21826640030444933,
+      "learning_rate": 4.420312676034731e-06,
+      "loss": 0.5719,
+      "step": 1357
+    },
+    {
+      "epoch": 3.3202933985330074,
+      "grad_norm": 0.1518586778362871,
+      "learning_rate": 4.419492717184094e-06,
+      "loss": 0.7889,
+      "step": 1358
+    },
+    {
+      "epoch": 3.3227383863080684,
+      "grad_norm": 0.20829789104766844,
+      "learning_rate": 4.418672254998525e-06,
+      "loss": 0.9048,
+      "step": 1359
+    },
+    {
+      "epoch": 3.32518337408313,
+      "grad_norm": 0.1768582651782789,
+      "learning_rate": 4.417851289693171e-06,
+      "loss": 0.8189,
+      "step": 1360
+    },
+    {
+      "epoch": 3.3276283618581908,
+      "grad_norm": 0.1534543141728541,
+      "learning_rate": 4.417029821483306e-06,
+      "loss": 0.4563,
+      "step": 1361
+    },
+    {
+      "epoch": 3.3300733496332517,
+      "grad_norm": 0.1055761917647611,
+      "learning_rate": 4.416207850584339e-06,
+      "loss": 0.4424,
+      "step": 1362
+    },
+    {
+      "epoch": 3.332518337408313,
+      "grad_norm": 0.1940248670005149,
+      "learning_rate": 4.415385377211808e-06,
+      "loss": 0.7029,
+      "step": 1363
+    },
+    {
+      "epoch": 3.334963325183374,
+      "grad_norm": 0.18171624743522638,
+      "learning_rate": 4.414562401581386e-06,
+      "loss": 0.6187,
+      "step": 1364
+    },
+    {
+      "epoch": 3.337408312958435,
+      "grad_norm": 0.22204221011849848,
+      "learning_rate": 4.4137389239088744e-06,
+      "loss": 0.5287,
+      "step": 1365
+    },
+    {
+      "epoch": 3.3398533007334965,
+      "grad_norm": 0.2805067684006206,
+      "learning_rate": 4.41291494441021e-06,
+      "loss": 0.7874,
+      "step": 1366
+    },
+    {
+      "epoch": 3.3422982885085575,
+      "grad_norm": 0.11870399869303958,
+      "learning_rate": 4.412090463301458e-06,
+      "loss": 0.5957,
+      "step": 1367
+    },
+    {
+      "epoch": 3.3447432762836184,
+      "grad_norm": 0.20919219321658356,
+      "learning_rate": 4.411265480798815e-06,
+      "loss": 0.6323,
+      "step": 1368
+    },
+    {
+      "epoch": 3.34718826405868,
+      "grad_norm": 0.15914712958604574,
+      "learning_rate": 4.410439997118612e-06,
+      "loss": 0.5404,
+      "step": 1369
+    },
+    {
+      "epoch": 3.349633251833741,
+      "grad_norm": 0.17930234472139145,
+      "learning_rate": 4.4096140124773095e-06,
+      "loss": 0.7741,
+      "step": 1370
+    },
+    {
+      "epoch": 3.352078239608802,
+      "grad_norm": 0.17758547784514836,
+      "learning_rate": 4.408787527091499e-06,
+      "loss": 0.9643,
+      "step": 1371
+    },
+    {
+      "epoch": 3.354523227383863,
+      "grad_norm": 0.1785453425744948,
+      "learning_rate": 4.407960541177905e-06,
+      "loss": 0.4483,
+      "step": 1372
+    },
+    {
+      "epoch": 3.356968215158924,
+      "grad_norm": 0.16515287143659235,
+      "learning_rate": 4.40713305495338e-06,
+      "loss": 1.0549,
+      "step": 1373
+    },
+    {
+      "epoch": 3.359413202933985,
+      "grad_norm": 0.16622841635302463,
+      "learning_rate": 4.406305068634912e-06,
+      "loss": 0.5289,
+      "step": 1374
+    },
+    {
+      "epoch": 3.3618581907090466,
+      "grad_norm": 0.18395560324980878,
+      "learning_rate": 4.405476582439616e-06,
+      "loss": 0.7516,
+      "step": 1375
+    },
+    {
+      "epoch": 3.3643031784841075,
+      "grad_norm": 0.21607564782044592,
+      "learning_rate": 4.404647596584742e-06,
+      "loss": 0.4733,
+      "step": 1376
+    },
+    {
+      "epoch": 3.3667481662591685,
+      "grad_norm": 0.1526065309931781,
+      "learning_rate": 4.403818111287668e-06,
+      "loss": 0.7207,
+      "step": 1377
+    },
+    {
+      "epoch": 3.36919315403423,
+      "grad_norm": 0.19394622016390767,
+      "learning_rate": 4.4029881267659045e-06,
+      "loss": 0.921,
+      "step": 1378
+    },
+    {
+      "epoch": 3.371638141809291,
+      "grad_norm": 0.20355571954244783,
+      "learning_rate": 4.402157643237092e-06,
+      "loss": 0.9869,
+      "step": 1379
+    },
+    {
+      "epoch": 3.374083129584352,
+      "grad_norm": 0.12071461333941827,
+      "learning_rate": 4.401326660919002e-06,
+      "loss": 0.4069,
+      "step": 1380
+    },
+    {
+      "epoch": 3.3765281173594133,
+      "grad_norm": 0.16293483783139062,
+      "learning_rate": 4.400495180029537e-06,
+      "loss": 0.4753,
+      "step": 1381
+    },
+    {
+      "epoch": 3.3789731051344742,
+      "grad_norm": 0.18484211920235352,
+      "learning_rate": 4.399663200786733e-06,
+      "loss": 0.6018,
+      "step": 1382
+    },
+    {
+      "epoch": 3.381418092909535,
+      "grad_norm": 0.10182079121309406,
+      "learning_rate": 4.39883072340875e-06,
+      "loss": 0.486,
+      "step": 1383
+    },
+    {
+      "epoch": 3.3838630806845966,
+      "grad_norm": 0.17739761366535337,
+      "learning_rate": 4.397997748113884e-06,
+      "loss": 0.88,
+      "step": 1384
+    },
+    {
+      "epoch": 3.3863080684596576,
+      "grad_norm": 0.20190398997845,
+      "learning_rate": 4.3971642751205626e-06,
+      "loss": 0.6489,
+      "step": 1385
+    },
+    {
+      "epoch": 3.388753056234719,
+      "grad_norm": 0.20790052550815696,
+      "learning_rate": 4.3963303046473385e-06,
+      "loss": 0.7959,
+      "step": 1386
+    },
+    {
+      "epoch": 3.39119804400978,
+      "grad_norm": 0.20335046330718368,
+      "learning_rate": 4.395495836912899e-06,
+      "loss": 0.6075,
+      "step": 1387
+    },
+    {
+      "epoch": 3.393643031784841,
+      "grad_norm": 0.18343092134458344,
+      "learning_rate": 4.394660872136061e-06,
+      "loss": 0.9653,
+      "step": 1388
+    },
+    {
+      "epoch": 3.3960880195599024,
+      "grad_norm": 0.2260283261789901,
+      "learning_rate": 4.393825410535771e-06,
+      "loss": 0.5119,
+      "step": 1389
+    },
+    {
+      "epoch": 3.3985330073349633,
+      "grad_norm": 0.22771194568008446,
+      "learning_rate": 4.392989452331106e-06,
+      "loss": 0.3854,
+      "step": 1390
+    },
+    {
+      "epoch": 3.4009779951100243,
+      "grad_norm": 0.23458777399375302,
+      "learning_rate": 4.392152997741273e-06,
+      "loss": 0.5985,
+      "step": 1391
+    },
+    {
+      "epoch": 3.4034229828850857,
+      "grad_norm": 0.14535138495432257,
+      "learning_rate": 4.391316046985611e-06,
+      "loss": 0.3749,
+      "step": 1392
+    },
+    {
+      "epoch": 3.4058679706601467,
+      "grad_norm": 0.18522645316667558,
+      "learning_rate": 4.390478600283585e-06,
+      "loss": 0.8911,
+      "step": 1393
+    },
+    {
+      "epoch": 3.4083129584352077,
+      "grad_norm": 0.22187619657363306,
+      "learning_rate": 4.389640657854796e-06,
+      "loss": 0.6332,
+      "step": 1394
+    },
+    {
+      "epoch": 3.410757946210269,
+      "grad_norm": 0.12981434073399714,
+      "learning_rate": 4.388802219918969e-06,
+      "loss": 0.3989,
+      "step": 1395
+    },
+    {
+      "epoch": 3.41320293398533,
+      "grad_norm": 0.17414209909242584,
+      "learning_rate": 4.387963286695963e-06,
+      "loss": 0.6862,
+      "step": 1396
+    },
+    {
+      "epoch": 3.4156479217603914,
+      "grad_norm": 0.16355705303495935,
+      "learning_rate": 4.387123858405765e-06,
+      "loss": 0.6819,
+      "step": 1397
+    },
+    {
+      "epoch": 3.4180929095354524,
+      "grad_norm": 0.21839448707169096,
+      "learning_rate": 4.386283935268493e-06,
+      "loss": 0.6606,
+      "step": 1398
+    },
+    {
+      "epoch": 3.4205378973105134,
+      "grad_norm": 0.2005445773900621,
+      "learning_rate": 4.385443517504393e-06,
+      "loss": 0.762,
+      "step": 1399
+    },
+    {
+      "epoch": 3.422982885085575,
+      "grad_norm": 0.14701066238117477,
+      "learning_rate": 4.384602605333843e-06,
+      "loss": 0.4893,
+      "step": 1400
+    },
+    {
+      "epoch": 3.4254278728606358,
+      "grad_norm": 0.1590540649426254,
+      "learning_rate": 4.3837611989773485e-06,
+      "loss": 0.5255,
+      "step": 1401
+    },
+    {
+      "epoch": 3.4278728606356967,
+      "grad_norm": 0.20913179942040158,
+      "learning_rate": 4.3829192986555445e-06,
+      "loss": 0.3938,
+      "step": 1402
+    },
+    {
+      "epoch": 3.430317848410758,
+      "grad_norm": 0.1958948456807502,
+      "learning_rate": 4.3820769045891995e-06,
+      "loss": 0.9963,
+      "step": 1403
+    },
+    {
+      "epoch": 3.432762836185819,
+      "grad_norm": 0.2076553151679762,
+      "learning_rate": 4.381234016999206e-06,
+      "loss": 0.4865,
+      "step": 1404
+    },
+    {
+      "epoch": 3.43520782396088,
+      "grad_norm": 0.17323831054579475,
+      "learning_rate": 4.38039063610659e-06,
+      "loss": 0.699,
+      "step": 1405
+    },
+    {
+      "epoch": 3.4376528117359415,
+      "grad_norm": 0.20331146573962627,
+      "learning_rate": 4.379546762132504e-06,
+      "loss": 0.376,
+      "step": 1406
+    },
+    {
+      "epoch": 3.4400977995110025,
+      "grad_norm": 0.1267542509895385,
+      "learning_rate": 4.378702395298231e-06,
+      "loss": 0.617,
+      "step": 1407
+    },
+    {
+      "epoch": 3.4425427872860634,
+      "grad_norm": 0.1640927808711629,
+      "learning_rate": 4.3778575358251844e-06,
+      "loss": 0.3967,
+      "step": 1408
+    },
+    {
+      "epoch": 3.444987775061125,
+      "grad_norm": 0.1281459689292886,
+      "learning_rate": 4.3770121839349054e-06,
+      "loss": 0.5522,
+      "step": 1409
+    },
+    {
+      "epoch": 3.447432762836186,
+      "grad_norm": 0.20616684405351077,
+      "learning_rate": 4.376166339849064e-06,
+      "loss": 0.5627,
+      "step": 1410
+    },
+    {
+      "epoch": 3.449877750611247,
+      "grad_norm": 0.24047842981071962,
+      "learning_rate": 4.37532000378946e-06,
+      "loss": 0.4852,
+      "step": 1411
+    },
+    {
+      "epoch": 3.452322738386308,
+      "grad_norm": 0.11413467377691602,
+      "learning_rate": 4.374473175978023e-06,
+      "loss": 0.4811,
+      "step": 1412
+    },
+    {
+      "epoch": 3.454767726161369,
+      "grad_norm": 0.12112066177938999,
+      "learning_rate": 4.37362585663681e-06,
+      "loss": 0.5288,
+      "step": 1413
+    },
+    {
+      "epoch": 3.45721271393643,
+      "grad_norm": 0.23508199707061983,
+      "learning_rate": 4.372778045988008e-06,
+      "loss": 0.6079,
+      "step": 1414
+    },
+    {
+      "epoch": 3.4596577017114916,
+      "grad_norm": 0.20379458878680404,
+      "learning_rate": 4.37192974425393e-06,
+      "loss": 0.5999,
+      "step": 1415
+    },
+    {
+      "epoch": 3.4621026894865525,
+      "grad_norm": 0.17966541922919488,
+      "learning_rate": 4.371080951657024e-06,
+      "loss": 0.736,
+      "step": 1416
+    },
+    {
+      "epoch": 3.4645476772616135,
+      "grad_norm": 0.18645930734314087,
+      "learning_rate": 4.37023166841986e-06,
+      "loss": 0.8258,
+      "step": 1417
+    },
+    {
+      "epoch": 3.466992665036675,
+      "grad_norm": 0.27363349253423064,
+      "learning_rate": 4.369381894765141e-06,
+      "loss": 0.4445,
+      "step": 1418
+    },
+    {
+      "epoch": 3.469437652811736,
+      "grad_norm": 0.21538590190623094,
+      "learning_rate": 4.3685316309156955e-06,
+      "loss": 0.7469,
+      "step": 1419
+    },
+    {
+      "epoch": 3.471882640586797,
+      "grad_norm": 0.20406088705742215,
+      "learning_rate": 4.367680877094484e-06,
+      "loss": 0.4606,
+      "step": 1420
+    },
+    {
+      "epoch": 3.4743276283618583,
+      "grad_norm": 0.21520826602228588,
+      "learning_rate": 4.366829633524592e-06,
+      "loss": 0.4542,
+      "step": 1421
+    },
+    {
+      "epoch": 3.4767726161369192,
+      "grad_norm": 0.16909965092155824,
+      "learning_rate": 4.3659779004292366e-06,
+      "loss": 0.4435,
+      "step": 1422
+    },
+    {
+      "epoch": 3.47921760391198,
+      "grad_norm": 0.17452577866518698,
+      "learning_rate": 4.3651256780317605e-06,
+      "loss": 0.7729,
+      "step": 1423
+    },
+    {
+      "epoch": 3.4816625916870416,
+      "grad_norm": 0.12607983488825988,
+      "learning_rate": 4.364272966555636e-06,
+      "loss": 0.7839,
+      "step": 1424
+    },
+    {
+      "epoch": 3.4841075794621026,
+      "grad_norm": 0.15244942520596705,
+      "learning_rate": 4.363419766224465e-06,
+      "loss": 0.644,
+      "step": 1425
+    },
+    {
+      "epoch": 3.486552567237164,
+      "grad_norm": 0.15320036776048773,
+      "learning_rate": 4.362566077261974e-06,
+      "loss": 0.6334,
+      "step": 1426
+    },
+    {
+      "epoch": 3.488997555012225,
+      "grad_norm": 0.19369292462420964,
+      "learning_rate": 4.361711899892021e-06,
+      "loss": 0.8962,
+      "step": 1427
+    },
+    {
+      "epoch": 3.491442542787286,
+      "grad_norm": 0.24324497766246367,
+      "learning_rate": 4.360857234338591e-06,
+      "loss": 0.5653,
+      "step": 1428
+    },
+    {
+      "epoch": 3.4938875305623474,
+      "grad_norm": 0.15193539668516265,
+      "learning_rate": 4.360002080825796e-06,
+      "loss": 0.4622,
+      "step": 1429
+    },
+    {
+      "epoch": 3.4963325183374083,
+      "grad_norm": 0.15479630971448796,
+      "learning_rate": 4.359146439577878e-06,
+      "loss": 0.4407,
+      "step": 1430
+    },
+    {
+      "epoch": 3.4987775061124693,
+      "grad_norm": 0.13791943750594668,
+      "learning_rate": 4.3582903108192054e-06,
+      "loss": 0.6699,
+      "step": 1431
+    },
+    {
+      "epoch": 3.5012224938875307,
+      "grad_norm": 0.14650506014495693,
+      "learning_rate": 4.357433694774274e-06,
+      "loss": 0.499,
+      "step": 1432
+    },
+    {
+      "epoch": 3.5036674816625917,
+      "grad_norm": 0.1871743483050339,
+      "learning_rate": 4.356576591667709e-06,
+      "loss": 0.4916,
+      "step": 1433
+    },
+    {
+      "epoch": 3.5061124694376526,
+      "grad_norm": 0.17092922306443895,
+      "learning_rate": 4.355719001724262e-06,
+      "loss": 0.7783,
+      "step": 1434
+    },
+    {
+      "epoch": 3.508557457212714,
+      "grad_norm": 0.1954119187353881,
+      "learning_rate": 4.354860925168812e-06,
+      "loss": 0.5707,
+      "step": 1435
+    },
+    {
+      "epoch": 3.511002444987775,
+      "grad_norm": 0.21981753743401947,
+      "learning_rate": 4.354002362226368e-06,
+      "loss": 1.6729,
+      "step": 1436
+    },
+    {
+      "epoch": 3.5134474327628364,
+      "grad_norm": 0.26154835211336896,
+      "learning_rate": 4.353143313122064e-06,
+      "loss": 0.9926,
+      "step": 1437
+    },
+    {
+      "epoch": 3.5158924205378974,
+      "grad_norm": 0.14187237410065198,
+      "learning_rate": 4.352283778081162e-06,
+      "loss": 0.7691,
+      "step": 1438
+    },
+    {
+      "epoch": 3.5183374083129584,
+      "grad_norm": 0.18628350673981284,
+      "learning_rate": 4.351423757329052e-06,
+      "loss": 0.8613,
+      "step": 1439
+    },
+    {
+      "epoch": 3.52078239608802,
+      "grad_norm": 0.12831278154072343,
+      "learning_rate": 4.350563251091252e-06,
+      "loss": 0.8102,
+      "step": 1440
+    },
+    {
+      "epoch": 3.5232273838630808,
+      "grad_norm": 0.2233819312228632,
+      "learning_rate": 4.349702259593405e-06,
+      "loss": 0.427,
+      "step": 1441
+    },
+    {
+      "epoch": 3.5256723716381417,
+      "grad_norm": 0.2875855231881401,
+      "learning_rate": 4.348840783061284e-06,
+      "loss": 1.0445,
+      "step": 1442
+    },
+    {
+      "epoch": 3.528117359413203,
+      "grad_norm": 0.2569706177012072,
+      "learning_rate": 4.347978821720786e-06,
+      "loss": 0.4943,
+      "step": 1443
+    },
+    {
+      "epoch": 3.530562347188264,
+      "grad_norm": 0.20567981083111123,
+      "learning_rate": 4.347116375797938e-06,
+      "loss": 1.0003,
+      "step": 1444
+    },
+    {
+      "epoch": 3.533007334963325,
+      "grad_norm": 0.21219055670552875,
+      "learning_rate": 4.346253445518894e-06,
+      "loss": 0.7194,
+      "step": 1445
+    },
+    {
+      "epoch": 3.5354523227383865,
+      "grad_norm": 0.2748775990899456,
+      "learning_rate": 4.345390031109934e-06,
+      "loss": 0.7129,
+      "step": 1446
+    },
+    {
+      "epoch": 3.5378973105134475,
+      "grad_norm": 0.15102110166666918,
+      "learning_rate": 4.344526132797463e-06,
+      "loss": 0.5414,
+      "step": 1447
+    },
+    {
+      "epoch": 3.5403422982885084,
+      "grad_norm": 0.18082100796509593,
+      "learning_rate": 4.343661750808017e-06,
+      "loss": 0.8485,
+      "step": 1448
+    },
+    {
+      "epoch": 3.54278728606357,
+      "grad_norm": 0.13442170133735099,
+      "learning_rate": 4.342796885368256e-06,
+      "loss": 0.6917,
+      "step": 1449
+    },
+    {
+      "epoch": 3.545232273838631,
+      "grad_norm": 0.19256661819102777,
+      "learning_rate": 4.341931536704968e-06,
+      "loss": 0.5569,
+      "step": 1450
+    },
+    {
+      "epoch": 3.547677261613692,
+      "grad_norm": 0.16459209263237912,
+      "learning_rate": 4.341065705045066e-06,
+      "loss": 0.4166,
+      "step": 1451
+    },
+    {
+      "epoch": 3.550122249388753,
+      "grad_norm": 0.183932147676097,
+      "learning_rate": 4.3401993906155914e-06,
+      "loss": 0.6807,
+      "step": 1452
+    },
+    {
+      "epoch": 3.552567237163814,
+      "grad_norm": 0.1900098132990076,
+      "learning_rate": 4.339332593643713e-06,
+      "loss": 0.8729,
+      "step": 1453
+    },
+    {
+      "epoch": 3.555012224938875,
+      "grad_norm": 0.20799056978977354,
+      "learning_rate": 4.338465314356722e-06,
+      "loss": 0.3431,
+      "step": 1454
+    },
+    {
+      "epoch": 3.5574572127139366,
+      "grad_norm": 0.1287927261339702,
+      "learning_rate": 4.337597552982042e-06,
+      "loss": 0.6634,
+      "step": 1455
+    },
+    {
+      "epoch": 3.5599022004889975,
+      "grad_norm": 0.1769892306400033,
+      "learning_rate": 4.3367293097472176e-06,
+      "loss": 0.9828,
+      "step": 1456
+    },
+    {
+      "epoch": 3.5623471882640585,
+      "grad_norm": 0.22673138798462789,
+      "learning_rate": 4.335860584879924e-06,
+      "loss": 0.4221,
+      "step": 1457
+    },
+    {
+      "epoch": 3.56479217603912,
+      "grad_norm": 0.19840215749516363,
+      "learning_rate": 4.334991378607959e-06,
+      "loss": 0.4405,
+      "step": 1458
+    },
+    {
+      "epoch": 3.567237163814181,
+      "grad_norm": 0.171654173913462,
+      "learning_rate": 4.3341216911592496e-06,
+      "loss": 0.5189,
+      "step": 1459
+    },
+    {
+      "epoch": 3.569682151589242,
+      "grad_norm": 0.16664268588769726,
+      "learning_rate": 4.333251522761847e-06,
+      "loss": 0.5313,
+      "step": 1460
+    },
+    {
+      "epoch": 3.5721271393643033,
+      "grad_norm": 0.2067790030698619,
+      "learning_rate": 4.332380873643929e-06,
+      "loss": 0.5817,
+      "step": 1461
+    },
+    {
+      "epoch": 3.5745721271393642,
+      "grad_norm": 0.1486872688804778,
+      "learning_rate": 4.3315097440338006e-06,
+      "loss": 0.8625,
+      "step": 1462
+    },
+    {
+      "epoch": 3.577017114914425,
+      "grad_norm": 0.2499377110971087,
+      "learning_rate": 4.330638134159891e-06,
+      "loss": 0.446,
+      "step": 1463
+    },
+    {
+      "epoch": 3.5794621026894866,
+      "grad_norm": 0.2094586283328521,
+      "learning_rate": 4.329766044250757e-06,
+      "loss": 0.8141,
+      "step": 1464
+    },
+    {
+      "epoch": 3.5819070904645476,
+      "grad_norm": 0.11548997375784534,
+      "learning_rate": 4.328893474535081e-06,
+      "loss": 0.5048,
+      "step": 1465
+    },
+    {
+      "epoch": 3.5843520782396086,
+      "grad_norm": 0.1619720756009413,
+      "learning_rate": 4.328020425241669e-06,
+      "loss": 0.5402,
+      "step": 1466
+    },
+    {
+      "epoch": 3.58679706601467,
+      "grad_norm": 0.1884784234051161,
+      "learning_rate": 4.3271468965994566e-06,
+      "loss": 0.8024,
+      "step": 1467
+    },
+    {
+      "epoch": 3.589242053789731,
+      "grad_norm": 0.25480939961537113,
+      "learning_rate": 4.3262728888375024e-06,
+      "loss": 0.3462,
+      "step": 1468
+    },
+    {
+      "epoch": 3.591687041564792,
+      "grad_norm": 0.18468870483292496,
+      "learning_rate": 4.32539840218499e-06,
+      "loss": 0.4549,
+      "step": 1469
+    },
+    {
+      "epoch": 3.5941320293398533,
+      "grad_norm": 0.21589180748163994,
+      "learning_rate": 4.324523436871231e-06,
+      "loss": 0.6731,
+      "step": 1470
+    },
+    {
+      "epoch": 3.5965770171149143,
+      "grad_norm": 0.15857567521917573,
+      "learning_rate": 4.323647993125661e-06,
+      "loss": 0.552,
+      "step": 1471
+    },
+    {
+      "epoch": 3.5990220048899757,
+      "grad_norm": 0.2198527865014562,
+      "learning_rate": 4.322772071177842e-06,
+      "loss": 0.672,
+      "step": 1472
+    },
+    {
+      "epoch": 3.6014669926650367,
+      "grad_norm": 0.21414227829709978,
+      "learning_rate": 4.321895671257459e-06,
+      "loss": 0.5018,
+      "step": 1473
+    },
+    {
+      "epoch": 3.6039119804400976,
+      "grad_norm": 0.17079431168079148,
+      "learning_rate": 4.321018793594326e-06,
+      "loss": 0.5151,
+      "step": 1474
+    },
+    {
+      "epoch": 3.606356968215159,
+      "grad_norm": 0.14659845528289905,
+      "learning_rate": 4.320141438418379e-06,
+      "loss": 0.4638,
+      "step": 1475
+    },
+    {
+      "epoch": 3.60880195599022,
+      "grad_norm": 0.18733563204604847,
+      "learning_rate": 4.319263605959683e-06,
+      "loss": 0.469,
+      "step": 1476
+    },
+    {
+      "epoch": 3.6112469437652814,
+      "grad_norm": 0.17218919258897608,
+      "learning_rate": 4.318385296448422e-06,
+      "loss": 0.5544,
+      "step": 1477
+    },
+    {
+      "epoch": 3.6136919315403424,
+      "grad_norm": 0.14140450197392046,
+      "learning_rate": 4.317506510114913e-06,
+      "loss": 0.5253,
+      "step": 1478
+    },
+    {
+      "epoch": 3.6161369193154034,
+      "grad_norm": 0.1396760457464423,
+      "learning_rate": 4.316627247189589e-06,
+      "loss": 0.4702,
+      "step": 1479
+    },
+    {
+      "epoch": 3.618581907090465,
+      "grad_norm": 0.12013345575866503,
+      "learning_rate": 4.315747507903018e-06,
+      "loss": 0.4641,
+      "step": 1480
+    },
+    {
+      "epoch": 3.6210268948655258,
+      "grad_norm": 0.15888546585628682,
+      "learning_rate": 4.3148672924858855e-06,
+      "loss": 0.7532,
+      "step": 1481
+    },
+    {
+      "epoch": 3.6234718826405867,
+      "grad_norm": 0.16868996795096083,
+      "learning_rate": 4.313986601169003e-06,
+      "loss": 0.4134,
+      "step": 1482
+    },
+    {
+      "epoch": 3.625916870415648,
+      "grad_norm": 0.28188336926929525,
+      "learning_rate": 4.31310543418331e-06,
+      "loss": 0.4692,
+      "step": 1483
+    },
+    {
+      "epoch": 3.628361858190709,
+      "grad_norm": 0.1863411648408465,
+      "learning_rate": 4.312223791759868e-06,
+      "loss": 0.3785,
+      "step": 1484
+    },
+    {
+      "epoch": 3.63080684596577,
+      "grad_norm": 0.20316204591819487,
+      "learning_rate": 4.311341674129862e-06,
+      "loss": 0.4637,
+      "step": 1485
+    },
+    {
+      "epoch": 3.6332518337408315,
+      "grad_norm": 0.18107639605614553,
+      "learning_rate": 4.310459081524606e-06,
+      "loss": 0.8133,
+      "step": 1486
+    },
+    {
+      "epoch": 3.6356968215158925,
+      "grad_norm": 0.19032454987562653,
+      "learning_rate": 4.309576014175535e-06,
+      "loss": 0.4337,
+      "step": 1487
+    },
+    {
+      "epoch": 3.6381418092909534,
+      "grad_norm": 0.19454702114028524,
+      "learning_rate": 4.308692472314209e-06,
+      "loss": 0.467,
+      "step": 1488
+    },
+    {
+      "epoch": 3.640586797066015,
+      "grad_norm": 0.13856707864158854,
+      "learning_rate": 4.3078084561723134e-06,
+      "loss": 0.4939,
+      "step": 1489
+    },
+    {
+      "epoch": 3.643031784841076,
+      "grad_norm": 0.18387632838792337,
+      "learning_rate": 4.306923965981657e-06,
+      "loss": 0.3799,
+      "step": 1490
+    },
+    {
+      "epoch": 3.645476772616137,
+      "grad_norm": 0.24807917153353443,
+      "learning_rate": 4.3060390019741746e-06,
+      "loss": 1.0492,
+      "step": 1491
+    },
+    {
+      "epoch": 3.647921760391198,
+      "grad_norm": 0.18846077517423573,
+      "learning_rate": 4.3051535643819225e-06,
+      "loss": 0.381,
+      "step": 1492
+    },
+    {
+      "epoch": 3.650366748166259,
+      "grad_norm": 0.25066795363480215,
+      "learning_rate": 4.304267653437083e-06,
+      "loss": 0.5585,
+      "step": 1493
+    },
+    {
+      "epoch": 3.65281173594132,
+      "grad_norm": 0.22392544902309183,
+      "learning_rate": 4.303381269371964e-06,
+      "loss": 0.4853,
+      "step": 1494
+    },
+    {
+      "epoch": 3.6552567237163816,
+      "grad_norm": 0.15902572689501818,
+      "learning_rate": 4.302494412418994e-06,
+      "loss": 0.7105,
+      "step": 1495
+    },
+    {
+      "epoch": 3.6577017114914425,
+      "grad_norm": 0.23361247651338574,
+      "learning_rate": 4.301607082810727e-06,
+      "loss": 0.6046,
+      "step": 1496
+    },
+    {
+      "epoch": 3.6601466992665035,
+      "grad_norm": 0.19385865405435335,
+      "learning_rate": 4.300719280779842e-06,
+      "loss": 1.009,
+      "step": 1497
+    },
+    {
+      "epoch": 3.662591687041565,
+      "grad_norm": 0.19916281625821344,
+      "learning_rate": 4.29983100655914e-06,
+      "loss": 0.5657,
+      "step": 1498
+    },
+    {
+      "epoch": 3.665036674816626,
+      "grad_norm": 0.207806164119596,
+      "learning_rate": 4.298942260381548e-06,
+      "loss": 0.6715,
+      "step": 1499
+    },
+    {
+      "epoch": 3.667481662591687,
+      "grad_norm": 0.1942795111829436,
+      "learning_rate": 4.2980530424801145e-06,
+      "loss": 0.444,
+      "step": 1500
+    },
+    {
+      "epoch": 3.6699266503667483,
+      "grad_norm": 0.1270088727281754,
+      "learning_rate": 4.297163353088014e-06,
+      "loss": 0.4151,
+      "step": 1501
+    },
+    {
+      "epoch": 3.6723716381418092,
+      "grad_norm": 0.1700411507143267,
+      "learning_rate": 4.296273192438543e-06,
+      "loss": 0.5041,
+      "step": 1502
+    },
+    {
+      "epoch": 3.67481662591687,
+      "grad_norm": 0.2903104934413432,
+      "learning_rate": 4.295382560765121e-06,
+      "loss": 0.4802,
+      "step": 1503
+    },
+    {
+      "epoch": 3.6772616136919316,
+      "grad_norm": 0.18035000692924077,
+      "learning_rate": 4.294491458301293e-06,
+      "loss": 0.9768,
+      "step": 1504
+    },
+    {
+      "epoch": 3.6797066014669926,
+      "grad_norm": 0.27120066828265305,
+      "learning_rate": 4.293599885280727e-06,
+      "loss": 0.7155,
+      "step": 1505
+    },
+    {
+      "epoch": 3.6821515892420535,
+      "grad_norm": 0.15064437492867336,
+      "learning_rate": 4.292707841937213e-06,
+      "loss": 0.4737,
+      "step": 1506
+    },
+    {
+      "epoch": 3.684596577017115,
+      "grad_norm": 0.31568396148302774,
+      "learning_rate": 4.291815328504665e-06,
+      "loss": 0.6019,
+      "step": 1507
+    },
+    {
+      "epoch": 3.687041564792176,
+      "grad_norm": 0.20728697775484933,
+      "learning_rate": 4.29092234521712e-06,
+      "loss": 0.7537,
+      "step": 1508
+    },
+    {
+      "epoch": 3.689486552567237,
+      "grad_norm": 0.2290461934726355,
+      "learning_rate": 4.29002889230874e-06,
+      "loss": 0.4736,
+      "step": 1509
+    },
+    {
+      "epoch": 3.6919315403422983,
+      "grad_norm": 0.17725863292730443,
+      "learning_rate": 4.289134970013809e-06,
+      "loss": 0.4207,
+      "step": 1510
+    },
+    {
+      "epoch": 3.6943765281173593,
+      "grad_norm": 0.21184428169776348,
+      "learning_rate": 4.288240578566732e-06,
+      "loss": 0.3837,
+      "step": 1511
+    },
+    {
+      "epoch": 3.6968215158924207,
+      "grad_norm": 0.19281424747423478,
+      "learning_rate": 4.287345718202041e-06,
+      "loss": 0.7442,
+      "step": 1512
+    },
+    {
+      "epoch": 3.6992665036674817,
+      "grad_norm": 0.20792668583419546,
+      "learning_rate": 4.286450389154387e-06,
+      "loss": 0.878,
+      "step": 1513
+    },
+    {
+      "epoch": 3.7017114914425426,
+      "grad_norm": 0.16574980397941944,
+      "learning_rate": 4.285554591658546e-06,
+      "loss": 0.4383,
+      "step": 1514
+    },
+    {
+      "epoch": 3.704156479217604,
+      "grad_norm": 0.16584324283169455,
+      "learning_rate": 4.284658325949419e-06,
+      "loss": 0.5184,
+      "step": 1515
+    },
+    {
+      "epoch": 3.706601466992665,
+      "grad_norm": 0.2141895469207335,
+      "learning_rate": 4.283761592262025e-06,
+      "loss": 0.8406,
+      "step": 1516
+    },
+    {
+      "epoch": 3.7090464547677264,
+      "grad_norm": 0.2120433571963209,
+      "learning_rate": 4.282864390831508e-06,
+      "loss": 0.6474,
+      "step": 1517
+    },
+    {
+      "epoch": 3.7114914425427874,
+      "grad_norm": 0.20612972130386734,
+      "learning_rate": 4.2819667218931375e-06,
+      "loss": 0.5167,
+      "step": 1518
+    },
+    {
+      "epoch": 3.7139364303178484,
+      "grad_norm": 0.23358016053152944,
+      "learning_rate": 4.2810685856823e-06,
+      "loss": 0.4651,
+      "step": 1519
+    },
+    {
+      "epoch": 3.71638141809291,
+      "grad_norm": 0.26729679763270525,
+      "learning_rate": 4.2801699824345074e-06,
+      "loss": 1.0619,
+      "step": 1520
+    },
+    {
+      "epoch": 3.7188264058679708,
+      "grad_norm": 0.18570088279451943,
+      "learning_rate": 4.279270912385395e-06,
+      "loss": 0.6004,
+      "step": 1521
+    },
+    {
+      "epoch": 3.7212713936430317,
+      "grad_norm": 0.18522366998539547,
+      "learning_rate": 4.27837137577072e-06,
+      "loss": 1.0134,
+      "step": 1522
+    },
+    {
+      "epoch": 3.723716381418093,
+      "grad_norm": 0.21074140330349975,
+      "learning_rate": 4.27747137282636e-06,
+      "loss": 0.507,
+      "step": 1523
+    },
+    {
+      "epoch": 3.726161369193154,
+      "grad_norm": 0.26696686775835965,
+      "learning_rate": 4.276570903788318e-06,
+      "loss": 0.6401,
+      "step": 1524
+    },
+    {
+      "epoch": 3.728606356968215,
+      "grad_norm": 0.16500629376413725,
+      "learning_rate": 4.2756699688927176e-06,
+      "loss": 0.6937,
+      "step": 1525
+    },
+    {
+      "epoch": 3.7310513447432765,
+      "grad_norm": 0.13921129365137652,
+      "learning_rate": 4.2747685683758036e-06,
+      "loss": 0.7923,
+      "step": 1526
+    },
+    {
+      "epoch": 3.7334963325183375,
+      "grad_norm": 0.1824163193267647,
+      "learning_rate": 4.273866702473943e-06,
+      "loss": 0.4964,
+      "step": 1527
+    },
+    {
+      "epoch": 3.7359413202933984,
+      "grad_norm": 0.20847943266358018,
+      "learning_rate": 4.272964371423628e-06,
+      "loss": 0.6435,
+      "step": 1528
+    },
+    {
+      "epoch": 3.73838630806846,
+      "grad_norm": 0.17606801285512352,
+      "learning_rate": 4.272061575461469e-06,
+      "loss": 0.9311,
+      "step": 1529
+    },
+    {
+      "epoch": 3.740831295843521,
+      "grad_norm": 0.2299405760726726,
+      "learning_rate": 4.2711583148241995e-06,
+      "loss": 0.9356,
+      "step": 1530
+    },
+    {
+      "epoch": 3.743276283618582,
+      "grad_norm": 0.16092293121770962,
+      "learning_rate": 4.270254589748675e-06,
+      "loss": 0.5201,
+      "step": 1531
+    },
+    {
+      "epoch": 3.745721271393643,
+      "grad_norm": 0.18385487811520876,
+      "learning_rate": 4.269350400471874e-06,
+      "loss": 1.1429,
+      "step": 1532
+    },
+    {
+      "epoch": 3.748166259168704,
+      "grad_norm": 0.2607026126241708,
+      "learning_rate": 4.268445747230896e-06,
+      "loss": 1.1623,
+      "step": 1533
+    },
+    {
+      "epoch": 3.750611246943765,
+      "grad_norm": 0.14452905365906135,
+      "learning_rate": 4.267540630262961e-06,
+      "loss": 0.5079,
+      "step": 1534
+    },
+    {
+      "epoch": 3.7530562347188265,
+      "grad_norm": 0.17244045236361516,
+      "learning_rate": 4.26663504980541e-06,
+      "loss": 0.3563,
+      "step": 1535
+    },
+    {
+      "epoch": 3.7555012224938875,
+      "grad_norm": 0.18368827771342403,
+      "learning_rate": 4.265729006095709e-06,
+      "loss": 0.4127,
+      "step": 1536
+    },
+    {
+      "epoch": 3.7579462102689485,
+      "grad_norm": 0.12539516835635212,
+      "learning_rate": 4.264822499371442e-06,
+      "loss": 0.4719,
+      "step": 1537
+    },
+    {
+      "epoch": 3.76039119804401,
+      "grad_norm": 0.17665950638073558,
+      "learning_rate": 4.263915529870319e-06,
+      "loss": 0.4709,
+      "step": 1538
+    },
+    {
+      "epoch": 3.762836185819071,
+      "grad_norm": 0.19173385021846664,
+      "learning_rate": 4.263008097830162e-06,
+      "loss": 0.4406,
+      "step": 1539
+    },
+    {
+      "epoch": 3.765281173594132,
+      "grad_norm": 0.1658000759947098,
+      "learning_rate": 4.262100203488927e-06,
+      "loss": 0.7088,
+      "step": 1540
+    },
+    {
+      "epoch": 3.7677261613691932,
+      "grad_norm": 0.22846945384107484,
+      "learning_rate": 4.26119184708468e-06,
+      "loss": 0.623,
+      "step": 1541
+    },
+    {
+      "epoch": 3.770171149144254,
+      "grad_norm": 0.19206946579087833,
+      "learning_rate": 4.2602830288556155e-06,
+      "loss": 0.5145,
+      "step": 1542
+    },
+    {
+      "epoch": 3.772616136919315,
+      "grad_norm": 0.24032548699199668,
+      "learning_rate": 4.259373749040045e-06,
+      "loss": 0.4808,
+      "step": 1543
+    },
+    {
+      "epoch": 3.7750611246943766,
+      "grad_norm": 0.1404873040384191,
+      "learning_rate": 4.2584640078764025e-06,
+      "loss": 0.5355,
+      "step": 1544
+    },
+    {
+      "epoch": 3.7775061124694376,
+      "grad_norm": 0.16166719602009264,
+      "learning_rate": 4.257553805603245e-06,
+      "loss": 0.6952,
+      "step": 1545
+    },
+    {
+      "epoch": 3.7799511002444985,
+      "grad_norm": 0.22202483071601753,
+      "learning_rate": 4.256643142459246e-06,
+      "loss": 0.6393,
+      "step": 1546
+    },
+    {
+      "epoch": 3.78239608801956,
+      "grad_norm": 0.20887696487231108,
+      "learning_rate": 4.255732018683204e-06,
+      "loss": 0.8471,
+      "step": 1547
+    },
+    {
+      "epoch": 3.784841075794621,
+      "grad_norm": 0.15497906654704485,
+      "learning_rate": 4.254820434514034e-06,
+      "loss": 0.5881,
+      "step": 1548
+    },
+    {
+      "epoch": 3.787286063569682,
+      "grad_norm": 0.12984721685974215,
+      "learning_rate": 4.253908390190777e-06,
+      "loss": 0.5935,
+      "step": 1549
+    },
+    {
+      "epoch": 3.7897310513447433,
+      "grad_norm": 0.21225482954029468,
+      "learning_rate": 4.25299588595259e-06,
+      "loss": 0.5,
+      "step": 1550
+    },
+    {
+      "epoch": 3.7921760391198043,
+      "grad_norm": 0.20013195653995045,
+      "learning_rate": 4.2520829220387535e-06,
+      "loss": 0.3674,
+      "step": 1551
+    },
+    {
+      "epoch": 3.7946210268948657,
+      "grad_norm": 0.20720309807292175,
+      "learning_rate": 4.251169498688668e-06,
+      "loss": 0.3565,
+      "step": 1552
+    },
+    {
+      "epoch": 3.7970660146699267,
+      "grad_norm": 0.14003994861678892,
+      "learning_rate": 4.2502556161418515e-06,
+      "loss": 0.5815,
+      "step": 1553
+    },
+    {
+      "epoch": 3.7995110024449876,
+      "grad_norm": 0.22282677147792743,
+      "learning_rate": 4.2493412746379485e-06,
+      "loss": 0.4971,
+      "step": 1554
+    },
+    {
+      "epoch": 3.801955990220049,
+      "grad_norm": 0.1399559531940187,
+      "learning_rate": 4.248426474416718e-06,
+      "loss": 0.5482,
+      "step": 1555
+    },
+    {
+      "epoch": 3.80440097799511,
+      "grad_norm": 0.24546203825299318,
+      "learning_rate": 4.247511215718042e-06,
+      "loss": 0.734,
+      "step": 1556
+    },
+    {
+      "epoch": 3.8068459657701714,
+      "grad_norm": 0.18488732525386353,
+      "learning_rate": 4.246595498781922e-06,
+      "loss": 0.8227,
+      "step": 1557
+    },
+    {
+      "epoch": 3.8092909535452324,
+      "grad_norm": 0.1622969373666004,
+      "learning_rate": 4.24567932384848e-06,
+      "loss": 0.516,
+      "step": 1558
+    },
+    {
+      "epoch": 3.8117359413202934,
+      "grad_norm": 0.23809265583221384,
+      "learning_rate": 4.244762691157959e-06,
+      "loss": 0.6903,
+      "step": 1559
+    },
+    {
+      "epoch": 3.8141809290953548,
+      "grad_norm": 0.19879265661555484,
+      "learning_rate": 4.24384560095072e-06,
+      "loss": 0.8092,
+      "step": 1560
+    },
+    {
+      "epoch": 3.8166259168704157,
+      "grad_norm": 0.20594885494352225,
+      "learning_rate": 4.242928053467246e-06,
+      "loss": 0.7784,
+      "step": 1561
+    },
+    {
+      "epoch": 3.8190709046454767,
+      "grad_norm": 0.17180817524216707,
+      "learning_rate": 4.242010048948138e-06,
+      "loss": 0.4345,
+      "step": 1562
+    },
+    {
+      "epoch": 3.821515892420538,
+      "grad_norm": 0.17764534261214263,
+      "learning_rate": 4.241091587634118e-06,
+      "loss": 0.4455,
+      "step": 1563
+    },
+    {
+      "epoch": 3.823960880195599,
+      "grad_norm": 0.19811064694886915,
+      "learning_rate": 4.240172669766027e-06,
+      "loss": 0.8163,
+      "step": 1564
+    },
+    {
+      "epoch": 3.82640586797066,
+      "grad_norm": 0.14952906767358026,
+      "learning_rate": 4.239253295584828e-06,
+      "loss": 0.5155,
+      "step": 1565
+    },
+    {
+      "epoch": 3.8288508557457215,
+      "grad_norm": 0.19168341891658247,
+      "learning_rate": 4.238333465331601e-06,
+      "loss": 0.8454,
+      "step": 1566
+    },
+    {
+      "epoch": 3.8312958435207825,
+      "grad_norm": 0.30231206111676673,
+      "learning_rate": 4.237413179247546e-06,
+      "loss": 0.4726,
+      "step": 1567
+    },
+    {
+      "epoch": 3.8337408312958434,
+      "grad_norm": 0.13233568890046865,
+      "learning_rate": 4.236492437573985e-06,
+      "loss": 0.2259,
+      "step": 1568
+    },
+    {
+      "epoch": 3.836185819070905,
+      "grad_norm": 0.17937110840805964,
+      "learning_rate": 4.235571240552355e-06,
+      "loss": 0.6933,
+      "step": 1569
+    },
+    {
+      "epoch": 3.838630806845966,
+      "grad_norm": 0.16919566440693068,
+      "learning_rate": 4.234649588424216e-06,
+      "loss": 0.789,
+      "step": 1570
+    },
+    {
+      "epoch": 3.8410757946210268,
+      "grad_norm": 0.15430861584473152,
+      "learning_rate": 4.233727481431247e-06,
+      "loss": 0.5344,
+      "step": 1571
+    },
+    {
+      "epoch": 3.843520782396088,
+      "grad_norm": 0.22215341790533846,
+      "learning_rate": 4.232804919815245e-06,
+      "loss": 1.3468,
+      "step": 1572
+    },
+    {
+      "epoch": 3.845965770171149,
+      "grad_norm": 0.1902720520035249,
+      "learning_rate": 4.2318819038181264e-06,
+      "loss": 0.4194,
+      "step": 1573
+    },
+    {
+      "epoch": 3.84841075794621,
+      "grad_norm": 0.23471616108777327,
+      "learning_rate": 4.230958433681928e-06,
+      "loss": 0.4626,
+      "step": 1574
+    },
+    {
+      "epoch": 3.8508557457212715,
+      "grad_norm": 0.1949645465012799,
+      "learning_rate": 4.230034509648803e-06,
+      "loss": 0.7755,
+      "step": 1575
+    },
+    {
+      "epoch": 3.8533007334963325,
+      "grad_norm": 0.1422097616776362,
+      "learning_rate": 4.229110131961029e-06,
+      "loss": 0.5084,
+      "step": 1576
+    },
+    {
+      "epoch": 3.8557457212713935,
+      "grad_norm": 0.23504625714161964,
+      "learning_rate": 4.228185300860994e-06,
+      "loss": 0.521,
+      "step": 1577
+    },
+    {
+      "epoch": 3.858190709046455,
+      "grad_norm": 0.17536469979245362,
+      "learning_rate": 4.2272600165912146e-06,
+      "loss": 0.5521,
+      "step": 1578
+    },
+    {
+      "epoch": 3.860635696821516,
+      "grad_norm": 0.2015376955427226,
+      "learning_rate": 4.2263342793943184e-06,
+      "loss": 0.471,
+      "step": 1579
+    },
+    {
+      "epoch": 3.863080684596577,
+      "grad_norm": 0.1796933714389003,
+      "learning_rate": 4.225408089513057e-06,
+      "loss": 0.6722,
+      "step": 1580
+    },
+    {
+      "epoch": 3.8655256723716382,
+      "grad_norm": 0.293466873610754,
+      "learning_rate": 4.224481447190296e-06,
+      "loss": 0.5212,
+      "step": 1581
+    },
+    {
+      "epoch": 3.867970660146699,
+      "grad_norm": 0.19800107930789396,
+      "learning_rate": 4.223554352669026e-06,
+      "loss": 0.4922,
+      "step": 1582
+    },
+    {
+      "epoch": 3.87041564792176,
+      "grad_norm": 0.25473286406229834,
+      "learning_rate": 4.222626806192348e-06,
+      "loss": 0.4795,
+      "step": 1583
+    },
+    {
+      "epoch": 3.8728606356968216,
+      "grad_norm": 0.16630353602020626,
+      "learning_rate": 4.2216988080034894e-06,
+      "loss": 0.776,
+      "step": 1584
+    },
+    {
+      "epoch": 3.8753056234718826,
+      "grad_norm": 0.2119233515040073,
+      "learning_rate": 4.220770358345791e-06,
+      "loss": 0.5058,
+      "step": 1585
+    },
+    {
+      "epoch": 3.8777506112469435,
+      "grad_norm": 0.1903820401212916,
+      "learning_rate": 4.2198414574627135e-06,
+      "loss": 0.4093,
+      "step": 1586
+    },
+    {
+      "epoch": 3.880195599022005,
+      "grad_norm": 0.2262167579208167,
+      "learning_rate": 4.218912105597838e-06,
+      "loss": 0.3843,
+      "step": 1587
+    },
+    {
+      "epoch": 3.882640586797066,
+      "grad_norm": 0.15087880821374797,
+      "learning_rate": 4.2179823029948585e-06,
+      "loss": 0.3799,
+      "step": 1588
+    },
+    {
+      "epoch": 3.885085574572127,
+      "grad_norm": 0.14002032366094444,
+      "learning_rate": 4.217052049897592e-06,
+      "loss": 0.7532,
+      "step": 1589
+    },
+    {
+      "epoch": 3.8875305623471883,
+      "grad_norm": 0.1027091568738711,
+      "learning_rate": 4.216121346549974e-06,
+      "loss": 0.392,
+      "step": 1590
+    },
+    {
+      "epoch": 3.8899755501222493,
+      "grad_norm": 0.16200987371775113,
+      "learning_rate": 4.215190193196053e-06,
+      "loss": 0.5896,
+      "step": 1591
+    },
+    {
+      "epoch": 3.8924205378973102,
+      "grad_norm": 0.1284166160874503,
+      "learning_rate": 4.21425859008e-06,
+      "loss": 0.6224,
+      "step": 1592
+    },
+    {
+      "epoch": 3.8948655256723717,
+      "grad_norm": 0.21395288701094411,
+      "learning_rate": 4.213326537446104e-06,
+      "loss": 0.7552,
+      "step": 1593
+    },
+    {
+      "epoch": 3.8973105134474326,
+      "grad_norm": 0.12981866925141236,
+      "learning_rate": 4.212394035538769e-06,
+      "loss": 0.614,
+      "step": 1594
+    },
+    {
+      "epoch": 3.899755501222494,
+      "grad_norm": 0.19813644440009445,
+      "learning_rate": 4.211461084602519e-06,
+      "loss": 0.5935,
+      "step": 1595
+    },
+    {
+      "epoch": 3.902200488997555,
+      "grad_norm": 0.2708515972073224,
+      "learning_rate": 4.210527684881996e-06,
+      "loss": 0.9627,
+      "step": 1596
+    },
+    {
+      "epoch": 3.904645476772616,
+      "grad_norm": 0.23673999506549334,
+      "learning_rate": 4.209593836621957e-06,
+      "loss": 1.0825,
+      "step": 1597
+    },
+    {
+      "epoch": 3.9070904645476774,
+      "grad_norm": 0.16888179114165103,
+      "learning_rate": 4.208659540067278e-06,
+      "loss": 0.7053,
+      "step": 1598
+    },
+    {
+      "epoch": 3.9095354523227384,
+      "grad_norm": 0.2120324456302641,
+      "learning_rate": 4.207724795462955e-06,
+      "loss": 0.5489,
+      "step": 1599
+    },
+    {
+      "epoch": 3.9119804400977998,
+      "grad_norm": 0.3608041373714785,
+      "learning_rate": 4.206789603054099e-06,
+      "loss": 1.1897,
+      "step": 1600
+    },
+    {
+      "epoch": 3.9144254278728607,
+      "grad_norm": 0.23182534157722756,
+      "learning_rate": 4.205853963085937e-06,
+      "loss": 0.5483,
+      "step": 1601
+    },
+    {
+      "epoch": 3.9168704156479217,
+      "grad_norm": 0.23511133959491076,
+      "learning_rate": 4.204917875803818e-06,
+      "loss": 0.4389,
+      "step": 1602
+    },
+    {
+      "epoch": 3.919315403422983,
+      "grad_norm": 0.15927780478854633,
+      "learning_rate": 4.203981341453202e-06,
+      "loss": 0.7396,
+      "step": 1603
+    },
+    {
+      "epoch": 3.921760391198044,
+      "grad_norm": 0.22949192444221084,
+      "learning_rate": 4.203044360279674e-06,
+      "loss": 0.5043,
+      "step": 1604
+    },
+    {
+      "epoch": 3.924205378973105,
+      "grad_norm": 0.20490558445871518,
+      "learning_rate": 4.202106932528929e-06,
+      "loss": 0.6175,
+      "step": 1605
+    },
+    {
+      "epoch": 3.9266503667481665,
+      "grad_norm": 0.18046069191629607,
+      "learning_rate": 4.201169058446783e-06,
+      "loss": 0.399,
+      "step": 1606
+    },
+    {
+      "epoch": 3.9290953545232274,
+      "grad_norm": 0.15374193138552703,
+      "learning_rate": 4.200230738279167e-06,
+      "loss": 0.4391,
+      "step": 1607
+    },
+    {
+      "epoch": 3.9315403422982884,
+      "grad_norm": 0.20078915755772558,
+      "learning_rate": 4.199291972272131e-06,
+      "loss": 0.287,
+      "step": 1608
+    },
+    {
+      "epoch": 3.93398533007335,
+      "grad_norm": 0.22147256550198563,
+      "learning_rate": 4.198352760671841e-06,
+      "loss": 0.492,
+      "step": 1609
+    },
+    {
+      "epoch": 3.936430317848411,
+      "grad_norm": 0.16574030049361732,
+      "learning_rate": 4.197413103724579e-06,
+      "loss": 0.5324,
+      "step": 1610
+    },
+    {
+      "epoch": 3.9388753056234718,
+      "grad_norm": 0.2103753507382826,
+      "learning_rate": 4.1964730016767465e-06,
+      "loss": 0.7365,
+      "step": 1611
+    },
+    {
+      "epoch": 3.941320293398533,
+      "grad_norm": 0.1518267580262356,
+      "learning_rate": 4.1955324547748565e-06,
+      "loss": 0.8013,
+      "step": 1612
+    },
+    {
+      "epoch": 3.943765281173594,
+      "grad_norm": 0.20154510576343865,
+      "learning_rate": 4.1945914632655445e-06,
+      "loss": 0.7363,
+      "step": 1613
+    },
+    {
+      "epoch": 3.946210268948655,
+      "grad_norm": 0.15890745318421953,
+      "learning_rate": 4.1936500273955604e-06,
+      "loss": 0.3269,
+      "step": 1614
+    },
+    {
+      "epoch": 3.9486552567237165,
+      "grad_norm": 0.16580897328637217,
+      "learning_rate": 4.192708147411768e-06,
+      "loss": 0.7251,
+      "step": 1615
+    },
+    {
+      "epoch": 3.9511002444987775,
+      "grad_norm": 0.1795353181782988,
+      "learning_rate": 4.191765823561151e-06,
+      "loss": 0.4769,
+      "step": 1616
+    },
+    {
+      "epoch": 3.9535452322738385,
+      "grad_norm": 0.16144794702069534,
+      "learning_rate": 4.190823056090808e-06,
+      "loss": 0.5483,
+      "step": 1617
+    },
+    {
+      "epoch": 3.9559902200489,
+      "grad_norm": 0.16557851212954114,
+      "learning_rate": 4.189879845247955e-06,
+      "loss": 0.8062,
+      "step": 1618
+    },
+    {
+      "epoch": 3.958435207823961,
+      "grad_norm": 0.2464569364178437,
+      "learning_rate": 4.1889361912799236e-06,
+      "loss": 0.4014,
+      "step": 1619
+    },
+    {
+      "epoch": 3.960880195599022,
+      "grad_norm": 0.14265732039902237,
+      "learning_rate": 4.18799209443416e-06,
+      "loss": 0.4374,
+      "step": 1620
+    },
+    {
+      "epoch": 3.9633251833740832,
+      "grad_norm": 0.22339158613662402,
+      "learning_rate": 4.187047554958229e-06,
+      "loss": 0.5925,
+      "step": 1621
+    },
+    {
+      "epoch": 3.965770171149144,
+      "grad_norm": 0.16802303875551336,
+      "learning_rate": 4.18610257309981e-06,
+      "loss": 0.6077,
+      "step": 1622
+    },
+    {
+      "epoch": 3.968215158924205,
+      "grad_norm": 0.19415483104611594,
+      "learning_rate": 4.185157149106699e-06,
+      "loss": 0.6042,
+      "step": 1623
+    },
+    {
+      "epoch": 3.9706601466992666,
+      "grad_norm": 0.19962803492883538,
+      "learning_rate": 4.184211283226808e-06,
+      "loss": 0.5957,
+      "step": 1624
+    },
+    {
+      "epoch": 3.9731051344743276,
+      "grad_norm": 0.18497252565632344,
+      "learning_rate": 4.183264975708165e-06,
+      "loss": 0.5733,
+      "step": 1625
+    },
+    {
+      "epoch": 3.9755501222493885,
+      "grad_norm": 0.14586139866041184,
+      "learning_rate": 4.1823182267989134e-06,
+      "loss": 0.7121,
+      "step": 1626
+    },
+    {
+      "epoch": 3.97799511002445,
+      "grad_norm": 0.22799748060588776,
+      "learning_rate": 4.181371036747312e-06,
+      "loss": 0.4812,
+      "step": 1627
+    },
+    {
+      "epoch": 3.980440097799511,
+      "grad_norm": 0.1820323704700886,
+      "learning_rate": 4.180423405801734e-06,
+      "loss": 0.516,
+      "step": 1628
+    },
+    {
+      "epoch": 3.982885085574572,
+      "grad_norm": 0.20330517608379037,
+      "learning_rate": 4.179475334210674e-06,
+      "loss": 0.3732,
+      "step": 1629
+    },
+    {
+      "epoch": 3.9853300733496333,
+      "grad_norm": 0.15813036453233392,
+      "learning_rate": 4.178526822222736e-06,
+      "loss": 0.6198,
+      "step": 1630
+    },
+    {
+      "epoch": 3.9877750611246943,
+      "grad_norm": 0.15778655137904618,
+      "learning_rate": 4.177577870086641e-06,
+      "loss": 0.8369,
+      "step": 1631
+    },
+    {
+      "epoch": 3.9902200488997552,
+      "grad_norm": 0.3010724095591919,
+      "learning_rate": 4.176628478051225e-06,
+      "loss": 0.4517,
+      "step": 1632
+    },
+    {
+      "epoch": 3.9926650366748166,
+      "grad_norm": 0.20983829947240318,
+      "learning_rate": 4.175678646365443e-06,
+      "loss": 0.5051,
+      "step": 1633
+    },
+    {
+      "epoch": 3.9951100244498776,
+      "grad_norm": 0.2001018715260085,
+      "learning_rate": 4.1747283752783614e-06,
+      "loss": 0.5244,
+      "step": 1634
+    },
+    {
+      "epoch": 3.997555012224939,
+      "grad_norm": 0.21021590865627596,
+      "learning_rate": 4.173777665039163e-06,
+      "loss": 0.6324,
+      "step": 1635
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.3458555843589508,
+      "learning_rate": 4.172826515897146e-06,
+      "loss": 0.736,
+      "step": 1636
+    },
+    {
+      "epoch": 4.002444987775061,
+      "grad_norm": 0.1518077119406904,
+      "learning_rate": 4.171874928101723e-06,
+      "loss": 0.792,
+      "step": 1637
+    },
+    {
+      "epoch": 4.004889975550122,
+      "grad_norm": 0.1421089320209172,
+      "learning_rate": 4.170922901902422e-06,
+      "loss": 0.5526,
+      "step": 1638
+    },
+    {
+      "epoch": 4.007334963325183,
+      "grad_norm": 0.1388042758702329,
+      "learning_rate": 4.169970437548887e-06,
+      "loss": 0.6591,
+      "step": 1639
+    },
+    {
+      "epoch": 4.009779951100245,
+      "grad_norm": 0.20370445584125405,
+      "learning_rate": 4.1690175352908745e-06,
+      "loss": 0.4398,
+      "step": 1640
+    },
+    {
+      "epoch": 4.012224938875305,
+      "grad_norm": 0.1708535653334295,
+      "learning_rate": 4.1680641953782584e-06,
+      "loss": 0.4895,
+      "step": 1641
+    },
+    {
+      "epoch": 4.014669926650367,
+      "grad_norm": 0.18480442299047597,
+      "learning_rate": 4.167110418061027e-06,
+      "loss": 0.6549,
+      "step": 1642
+    },
+    {
+      "epoch": 4.017114914425428,
+      "grad_norm": 0.24273002542773492,
+      "learning_rate": 4.1661562035892805e-06,
+      "loss": 0.4971,
+      "step": 1643
+    },
+    {
+      "epoch": 4.019559902200489,
+      "grad_norm": 0.17093116699432037,
+      "learning_rate": 4.165201552213237e-06,
+      "loss": 0.5171,
+      "step": 1644
+    },
+    {
+      "epoch": 4.02200488997555,
+      "grad_norm": 0.17132151584278324,
+      "learning_rate": 4.1642464641832275e-06,
+      "loss": 0.4778,
+      "step": 1645
+    },
+    {
+      "epoch": 4.0244498777506115,
+      "grad_norm": 0.1905381437255112,
+      "learning_rate": 4.163290939749699e-06,
+      "loss": 0.8282,
+      "step": 1646
+    },
+    {
+      "epoch": 4.026894865525672,
+      "grad_norm": 0.18941523453512035,
+      "learning_rate": 4.16233497916321e-06,
+      "loss": 0.5959,
+      "step": 1647
+    },
+    {
+      "epoch": 4.029339853300733,
+      "grad_norm": 0.15737630515843087,
+      "learning_rate": 4.161378582674438e-06,
+      "loss": 0.4567,
+      "step": 1648
+    },
+    {
+      "epoch": 4.031784841075795,
+      "grad_norm": 0.22320279396684584,
+      "learning_rate": 4.160421750534169e-06,
+      "loss": 0.46,
+      "step": 1649
+    },
+    {
+      "epoch": 4.034229828850855,
+      "grad_norm": 0.15903555038672185,
+      "learning_rate": 4.159464482993308e-06,
+      "loss": 0.4813,
+      "step": 1650
+    },
+    {
+      "epoch": 4.036674816625917,
+      "grad_norm": 0.1594686881803847,
+      "learning_rate": 4.158506780302872e-06,
+      "loss": 0.7616,
+      "step": 1651
+    },
+    {
+      "epoch": 4.039119804400978,
+      "grad_norm": 0.25305541297434053,
+      "learning_rate": 4.157548642713993e-06,
+      "loss": 1.1283,
+      "step": 1652
+    },
+    {
+      "epoch": 4.041564792176039,
+      "grad_norm": 0.2789869036025608,
+      "learning_rate": 4.156590070477916e-06,
+      "loss": 0.4036,
+      "step": 1653
+    },
+    {
+      "epoch": 4.0440097799511,
+      "grad_norm": 0.14314037616237907,
+      "learning_rate": 4.1556310638460015e-06,
+      "loss": 0.4155,
+      "step": 1654
+    },
+    {
+      "epoch": 4.0464547677261615,
+      "grad_norm": 0.20900868903680211,
+      "learning_rate": 4.154671623069723e-06,
+      "loss": 0.849,
+      "step": 1655
+    },
+    {
+      "epoch": 4.048899755501223,
+      "grad_norm": 0.19439041231223247,
+      "learning_rate": 4.153711748400666e-06,
+      "loss": 0.7261,
+      "step": 1656
+    },
+    {
+      "epoch": 4.0513447432762835,
+      "grad_norm": 0.17187170420442915,
+      "learning_rate": 4.152751440090533e-06,
+      "loss": 0.6516,
+      "step": 1657
+    },
+    {
+      "epoch": 4.053789731051345,
+      "grad_norm": 0.17641294394021761,
+      "learning_rate": 4.151790698391139e-06,
+      "loss": 0.8281,
+      "step": 1658
+    },
+    {
+      "epoch": 4.056234718826406,
+      "grad_norm": 0.14973407190393406,
+      "learning_rate": 4.150829523554412e-06,
+      "loss": 0.4672,
+      "step": 1659
+    },
+    {
+      "epoch": 4.058679706601467,
+      "grad_norm": 0.17623868288621014,
+      "learning_rate": 4.149867915832394e-06,
+      "loss": 0.6132,
+      "step": 1660
+    },
+    {
+      "epoch": 4.061124694376528,
+      "grad_norm": 0.18759592893698873,
+      "learning_rate": 4.148905875477241e-06,
+      "loss": 0.5966,
+      "step": 1661
+    },
+    {
+      "epoch": 4.06356968215159,
+      "grad_norm": 0.213524041968869,
+      "learning_rate": 4.147943402741221e-06,
+      "loss": 0.4755,
+      "step": 1662
+    },
+    {
+      "epoch": 4.06601466992665,
+      "grad_norm": 0.13691961930828037,
+      "learning_rate": 4.146980497876718e-06,
+      "loss": 0.654,
+      "step": 1663
+    },
+    {
+      "epoch": 4.068459657701712,
+      "grad_norm": 0.1382214264948139,
+      "learning_rate": 4.146017161136226e-06,
+      "loss": 0.4311,
+      "step": 1664
+    },
+    {
+      "epoch": 4.070904645476773,
+      "grad_norm": 0.1579008065563269,
+      "learning_rate": 4.145053392772357e-06,
+      "loss": 0.5373,
+      "step": 1665
+    },
+    {
+      "epoch": 4.0733496332518335,
+      "grad_norm": 0.2314255331933333,
+      "learning_rate": 4.144089193037828e-06,
+      "loss": 1.5379,
+      "step": 1666
+    },
+    {
+      "epoch": 4.075794621026895,
+      "grad_norm": 0.14734056300997364,
+      "learning_rate": 4.143124562185479e-06,
+      "loss": 0.5351,
+      "step": 1667
+    },
+    {
+      "epoch": 4.078239608801956,
+      "grad_norm": 0.16489353002071347,
+      "learning_rate": 4.142159500468256e-06,
+      "loss": 0.5031,
+      "step": 1668
+    },
+    {
+      "epoch": 4.080684596577017,
+      "grad_norm": 0.18067862871934579,
+      "learning_rate": 4.14119400813922e-06,
+      "loss": 0.7994,
+      "step": 1669
+    },
+    {
+      "epoch": 4.083129584352078,
+      "grad_norm": 0.2007292764285633,
+      "learning_rate": 4.140228085451547e-06,
+      "loss": 0.4887,
+      "step": 1670
+    },
+    {
+      "epoch": 4.08557457212714,
+      "grad_norm": 0.13194276336074412,
+      "learning_rate": 4.1392617326585215e-06,
+      "loss": 0.3955,
+      "step": 1671
+    },
+    {
+      "epoch": 4.0880195599022,
+      "grad_norm": 0.17421264694468763,
+      "learning_rate": 4.138294950013546e-06,
+      "loss": 0.3632,
+      "step": 1672
+    },
+    {
+      "epoch": 4.090464547677262,
+      "grad_norm": 0.15223383470903787,
+      "learning_rate": 4.137327737770131e-06,
+      "loss": 0.3526,
+      "step": 1673
+    },
+    {
+      "epoch": 4.092909535452323,
+      "grad_norm": 0.15296515038485883,
+      "learning_rate": 4.136360096181904e-06,
+      "loss": 0.509,
+      "step": 1674
+    },
+    {
+      "epoch": 4.095354523227384,
+      "grad_norm": 0.17636666370307155,
+      "learning_rate": 4.1353920255026e-06,
+      "loss": 0.8884,
+      "step": 1675
+    },
+    {
+      "epoch": 4.097799511002445,
+      "grad_norm": 0.21033548482512168,
+      "learning_rate": 4.13442352598607e-06,
+      "loss": 0.8026,
+      "step": 1676
+    },
+    {
+      "epoch": 4.100244498777506,
+      "grad_norm": 0.1768482151374831,
+      "learning_rate": 4.133454597886278e-06,
+      "loss": 0.4585,
+      "step": 1677
+    },
+    {
+      "epoch": 4.102689486552567,
+      "grad_norm": 0.18983964173445708,
+      "learning_rate": 4.132485241457299e-06,
+      "loss": 0.721,
+      "step": 1678
+    },
+    {
+      "epoch": 4.105134474327628,
+      "grad_norm": 0.1591509256280448,
+      "learning_rate": 4.131515456953319e-06,
+      "loss": 0.5851,
+      "step": 1679
+    },
+    {
+      "epoch": 4.10757946210269,
+      "grad_norm": 0.12720504523789747,
+      "learning_rate": 4.130545244628639e-06,
+      "loss": 0.4402,
+      "step": 1680
+    },
+    {
+      "epoch": 4.11002444987775,
+      "grad_norm": 0.3026554285814813,
+      "learning_rate": 4.129574604737671e-06,
+      "loss": 0.7289,
+      "step": 1681
+    },
+    {
+      "epoch": 4.112469437652812,
+      "grad_norm": 0.16282817183258416,
+      "learning_rate": 4.128603537534936e-06,
+      "loss": 0.6524,
+      "step": 1682
+    },
+    {
+      "epoch": 4.114914425427873,
+      "grad_norm": 0.18469978114627153,
+      "learning_rate": 4.127632043275075e-06,
+      "loss": 0.3755,
+      "step": 1683
+    },
+    {
+      "epoch": 4.117359413202934,
+      "grad_norm": 0.24026038701837185,
+      "learning_rate": 4.126660122212833e-06,
+      "loss": 0.6782,
+      "step": 1684
+    },
+    {
+      "epoch": 4.119804400977995,
+      "grad_norm": 0.15762160010421056,
+      "learning_rate": 4.125687774603071e-06,
+      "loss": 0.3026,
+      "step": 1685
+    },
+    {
+      "epoch": 4.1222493887530565,
+      "grad_norm": 0.22776534075144217,
+      "learning_rate": 4.124715000700759e-06,
+      "loss": 0.7973,
+      "step": 1686
+    },
+    {
+      "epoch": 4.124694376528117,
+      "grad_norm": 0.17985068128101256,
+      "learning_rate": 4.123741800760982e-06,
+      "loss": 0.3162,
+      "step": 1687
+    },
+    {
+      "epoch": 4.127139364303178,
+      "grad_norm": 0.18723261069880706,
+      "learning_rate": 4.1227681750389356e-06,
+      "loss": 1.0402,
+      "step": 1688
+    },
+    {
+      "epoch": 4.12958435207824,
+      "grad_norm": 0.15943760525014972,
+      "learning_rate": 4.121794123789926e-06,
+      "loss": 0.681,
+      "step": 1689
+    },
+    {
+      "epoch": 4.1320293398533,
+      "grad_norm": 0.14188143452333535,
+      "learning_rate": 4.120819647269373e-06,
+      "loss": 0.4526,
+      "step": 1690
+    },
+    {
+      "epoch": 4.134474327628362,
+      "grad_norm": 0.22423321228533125,
+      "learning_rate": 4.119844745732805e-06,
+      "loss": 0.9314,
+      "step": 1691
+    },
+    {
+      "epoch": 4.136919315403423,
+      "grad_norm": 0.1983247750301809,
+      "learning_rate": 4.118869419435864e-06,
+      "loss": 0.4461,
+      "step": 1692
+    },
+    {
+      "epoch": 4.139364303178484,
+      "grad_norm": 0.13899832920549743,
+      "learning_rate": 4.1178936686343035e-06,
+      "loss": 0.4981,
+      "step": 1693
+    },
+    {
+      "epoch": 4.141809290953545,
+      "grad_norm": 0.2773548838963686,
+      "learning_rate": 4.116917493583987e-06,
+      "loss": 0.369,
+      "step": 1694
+    },
+    {
+      "epoch": 4.1442542787286065,
+      "grad_norm": 0.16370477728609112,
+      "learning_rate": 4.115940894540891e-06,
+      "loss": 0.675,
+      "step": 1695
+    },
+    {
+      "epoch": 4.146699266503667,
+      "grad_norm": 0.19750941908226533,
+      "learning_rate": 4.114963871761101e-06,
+      "loss": 1.0292,
+      "step": 1696
+    },
+    {
+      "epoch": 4.1491442542787285,
+      "grad_norm": 0.17229118212678296,
+      "learning_rate": 4.113986425500814e-06,
+      "loss": 0.5258,
+      "step": 1697
+    },
+    {
+      "epoch": 4.15158924205379,
+      "grad_norm": 0.2563897044610342,
+      "learning_rate": 4.113008556016341e-06,
+      "loss": 1.0661,
+      "step": 1698
+    },
+    {
+      "epoch": 4.15403422982885,
+      "grad_norm": 0.2079464054693061,
+      "learning_rate": 4.1120302635641e-06,
+      "loss": 0.6705,
+      "step": 1699
+    },
+    {
+      "epoch": 4.156479217603912,
+      "grad_norm": 0.21156203157025513,
+      "learning_rate": 4.111051548400623e-06,
+      "loss": 0.7136,
+      "step": 1700
+    },
+    {
+      "epoch": 4.158924205378973,
+      "grad_norm": 0.24742298365123258,
+      "learning_rate": 4.11007241078255e-06,
+      "loss": 0.8015,
+      "step": 1701
+    },
+    {
+      "epoch": 4.161369193154035,
+      "grad_norm": 0.2352102139484552,
+      "learning_rate": 4.109092850966635e-06,
+      "loss": 0.3871,
+      "step": 1702
+    },
+    {
+      "epoch": 4.163814180929095,
+      "grad_norm": 0.32180295269890397,
+      "learning_rate": 4.10811286920974e-06,
+      "loss": 0.6764,
+      "step": 1703
+    },
+    {
+      "epoch": 4.166259168704157,
+      "grad_norm": 0.2982592472054046,
+      "learning_rate": 4.107132465768839e-06,
+      "loss": 0.5148,
+      "step": 1704
+    },
+    {
+      "epoch": 4.168704156479218,
+      "grad_norm": 0.17478824715792302,
+      "learning_rate": 4.106151640901015e-06,
+      "loss": 0.528,
+      "step": 1705
+    },
+    {
+      "epoch": 4.1711491442542785,
+      "grad_norm": 0.16932053451028353,
+      "learning_rate": 4.105170394863465e-06,
+      "loss": 0.6537,
+      "step": 1706
+    },
+    {
+      "epoch": 4.17359413202934,
+      "grad_norm": 0.2728418361198744,
+      "learning_rate": 4.104188727913492e-06,
+      "loss": 0.9982,
+      "step": 1707
+    },
+    {
+      "epoch": 4.176039119804401,
+      "grad_norm": 0.2484005884172726,
+      "learning_rate": 4.103206640308513e-06,
+      "loss": 0.4627,
+      "step": 1708
+    },
+    {
+      "epoch": 4.178484107579462,
+      "grad_norm": 0.1624321113208904,
+      "learning_rate": 4.102224132306053e-06,
+      "loss": 0.4261,
+      "step": 1709
+    },
+    {
+      "epoch": 4.180929095354523,
+      "grad_norm": 0.18311486500358004,
+      "learning_rate": 4.101241204163748e-06,
+      "loss": 0.7649,
+      "step": 1710
+    },
+    {
+      "epoch": 4.183374083129585,
+      "grad_norm": 0.1878438592262303,
+      "learning_rate": 4.100257856139345e-06,
+      "loss": 0.5934,
+      "step": 1711
+    },
+    {
+      "epoch": 4.185819070904645,
+      "grad_norm": 0.134180268795672,
+      "learning_rate": 4.0992740884907005e-06,
+      "loss": 0.3716,
+      "step": 1712
+    },
+    {
+      "epoch": 4.188264058679707,
+      "grad_norm": 0.19271192195398162,
+      "learning_rate": 4.098289901475781e-06,
+      "loss": 1.0834,
+      "step": 1713
+    },
+    {
+      "epoch": 4.190709046454768,
+      "grad_norm": 0.1801323161805221,
+      "learning_rate": 4.0973052953526604e-06,
+      "loss": 0.8863,
+      "step": 1714
+    },
+    {
+      "epoch": 4.193154034229829,
+      "grad_norm": 0.19938665333273717,
+      "learning_rate": 4.096320270379528e-06,
+      "loss": 0.8924,
+      "step": 1715
+    },
+    {
+      "epoch": 4.19559902200489,
+      "grad_norm": 0.18916681893391316,
+      "learning_rate": 4.095334826814678e-06,
+      "loss": 0.8121,
+      "step": 1716
+    },
+    {
+      "epoch": 4.198044009779951,
+      "grad_norm": 0.1939948039664555,
+      "learning_rate": 4.094348964916518e-06,
+      "loss": 0.5514,
+      "step": 1717
+    },
+    {
+      "epoch": 4.200488997555012,
+      "grad_norm": 0.24374525762095,
+      "learning_rate": 4.093362684943563e-06,
+      "loss": 0.5261,
+      "step": 1718
+    },
+    {
+      "epoch": 4.202933985330073,
+      "grad_norm": 0.10104504785597727,
+      "learning_rate": 4.092375987154436e-06,
+      "loss": 0.2721,
+      "step": 1719
+    },
+    {
+      "epoch": 4.205378973105135,
+      "grad_norm": 0.2090159826655009,
+      "learning_rate": 4.091388871807875e-06,
+      "loss": 0.5686,
+      "step": 1720
+    },
+    {
+      "epoch": 4.207823960880195,
+      "grad_norm": 0.24106039689704994,
+      "learning_rate": 4.0904013391627215e-06,
+      "loss": 0.6725,
+      "step": 1721
+    },
+    {
+      "epoch": 4.210268948655257,
+      "grad_norm": 0.1711164393090134,
+      "learning_rate": 4.089413389477932e-06,
+      "loss": 0.6855,
+      "step": 1722
+    },
+    {
+      "epoch": 4.212713936430318,
+      "grad_norm": 0.20650419400661307,
+      "learning_rate": 4.088425023012567e-06,
+      "loss": 0.3486,
+      "step": 1723
+    },
+    {
+      "epoch": 4.215158924205379,
+      "grad_norm": 0.19761023162015257,
+      "learning_rate": 4.087436240025799e-06,
+      "loss": 0.5559,
+      "step": 1724
+    },
+    {
+      "epoch": 4.21760391198044,
+      "grad_norm": 0.19256799855029696,
+      "learning_rate": 4.086447040776911e-06,
+      "loss": 0.518,
+      "step": 1725
+    },
+    {
+      "epoch": 4.2200488997555015,
+      "grad_norm": 0.24045062378784746,
+      "learning_rate": 4.0854574255252936e-06,
+      "loss": 0.5368,
+      "step": 1726
+    },
+    {
+      "epoch": 4.222493887530562,
+      "grad_norm": 0.19615511642049846,
+      "learning_rate": 4.084467394530446e-06,
+      "loss": 0.6047,
+      "step": 1727
+    },
+    {
+      "epoch": 4.224938875305623,
+      "grad_norm": 0.19862164662345816,
+      "learning_rate": 4.083476948051975e-06,
+      "loss": 0.4562,
+      "step": 1728
+    },
+    {
+      "epoch": 4.227383863080685,
+      "grad_norm": 0.24290472576999142,
+      "learning_rate": 4.082486086349603e-06,
+      "loss": 0.4321,
+      "step": 1729
+    },
+    {
+      "epoch": 4.229828850855745,
+      "grad_norm": 0.24874969368093675,
+      "learning_rate": 4.0814948096831515e-06,
+      "loss": 0.9137,
+      "step": 1730
+    },
+    {
+      "epoch": 4.232273838630807,
+      "grad_norm": 0.2004430668876246,
+      "learning_rate": 4.080503118312559e-06,
+      "loss": 0.4105,
+      "step": 1731
+    },
+    {
+      "epoch": 4.234718826405868,
+      "grad_norm": 0.30042391552745845,
+      "learning_rate": 4.07951101249787e-06,
+      "loss": 0.4581,
+      "step": 1732
+    },
+    {
+      "epoch": 4.237163814180929,
+      "grad_norm": 0.15593661106081655,
+      "learning_rate": 4.078518492499235e-06,
+      "loss": 0.5084,
+      "step": 1733
+    },
+    {
+      "epoch": 4.23960880195599,
+      "grad_norm": 0.13460768923376595,
+      "learning_rate": 4.077525558576918e-06,
+      "loss": 0.3965,
+      "step": 1734
+    },
+    {
+      "epoch": 4.2420537897310515,
+      "grad_norm": 0.2806401285595934,
+      "learning_rate": 4.076532210991287e-06,
+      "loss": 0.4495,
+      "step": 1735
+    },
+    {
+      "epoch": 4.244498777506112,
+      "grad_norm": 0.17670674889130006,
+      "learning_rate": 4.075538450002822e-06,
+      "loss": 0.5333,
+      "step": 1736
+    },
+    {
+      "epoch": 4.2469437652811735,
+      "grad_norm": 0.1669677734742625,
+      "learning_rate": 4.074544275872109e-06,
+      "loss": 1.1966,
+      "step": 1737
+    },
+    {
+      "epoch": 4.249388753056235,
+      "grad_norm": 0.1980139139445526,
+      "learning_rate": 4.0735496888598425e-06,
+      "loss": 0.9163,
+      "step": 1738
+    },
+    {
+      "epoch": 4.251833740831296,
+      "grad_norm": 0.24699550494926706,
+      "learning_rate": 4.072554689226828e-06,
+      "loss": 0.5292,
+      "step": 1739
+    },
+    {
+      "epoch": 4.254278728606357,
+      "grad_norm": 0.12767823783425794,
+      "learning_rate": 4.071559277233975e-06,
+      "loss": 0.4056,
+      "step": 1740
+    },
+    {
+      "epoch": 4.256723716381418,
+      "grad_norm": 0.18410892576079202,
+      "learning_rate": 4.070563453142305e-06,
+      "loss": 0.5237,
+      "step": 1741
+    },
+    {
+      "epoch": 4.25916870415648,
+      "grad_norm": 0.1264051982372328,
+      "learning_rate": 4.069567217212945e-06,
+      "loss": 0.3116,
+      "step": 1742
+    },
+    {
+      "epoch": 4.26161369193154,
+      "grad_norm": 0.18251878071261948,
+      "learning_rate": 4.068570569707132e-06,
+      "loss": 0.5484,
+      "step": 1743
+    },
+    {
+      "epoch": 4.264058679706602,
+      "grad_norm": 0.13456310387728304,
+      "learning_rate": 4.067573510886208e-06,
+      "loss": 0.379,
+      "step": 1744
+    },
+    {
+      "epoch": 4.266503667481663,
+      "grad_norm": 0.17153722104478095,
+      "learning_rate": 4.066576041011626e-06,
+      "loss": 0.562,
+      "step": 1745
+    },
+    {
+      "epoch": 4.2689486552567235,
+      "grad_norm": 0.16674707226468977,
+      "learning_rate": 4.065578160344944e-06,
+      "loss": 0.4714,
+      "step": 1746
+    },
+    {
+      "epoch": 4.271393643031785,
+      "grad_norm": 0.18346192347526105,
+      "learning_rate": 4.064579869147831e-06,
+      "loss": 0.5058,
+      "step": 1747
+    },
+    {
+      "epoch": 4.273838630806846,
+      "grad_norm": 0.15110590133481344,
+      "learning_rate": 4.063581167682062e-06,
+      "loss": 0.4163,
+      "step": 1748
+    },
+    {
+      "epoch": 4.276283618581907,
+      "grad_norm": 0.18313462366880673,
+      "learning_rate": 4.062582056209516e-06,
+      "loss": 0.5325,
+      "step": 1749
+    },
+    {
+      "epoch": 4.278728606356968,
+      "grad_norm": 0.2365198025369478,
+      "learning_rate": 4.061582534992187e-06,
+      "loss": 0.5257,
+      "step": 1750
+    },
+    {
+      "epoch": 4.28117359413203,
+      "grad_norm": 0.1604786642045765,
+      "learning_rate": 4.060582604292169e-06,
+      "loss": 0.8026,
+      "step": 1751
+    },
+    {
+      "epoch": 4.28361858190709,
+      "grad_norm": 0.1863887245925862,
+      "learning_rate": 4.05958226437167e-06,
+      "loss": 0.4638,
+      "step": 1752
+    },
+    {
+      "epoch": 4.286063569682152,
+      "grad_norm": 0.23044308543333392,
+      "learning_rate": 4.058581515492999e-06,
+      "loss": 0.4691,
+      "step": 1753
+    },
+    {
+      "epoch": 4.288508557457213,
+      "grad_norm": 0.1892552082385137,
+      "learning_rate": 4.057580357918577e-06,
+      "loss": 0.4189,
+      "step": 1754
+    },
+    {
+      "epoch": 4.290953545232274,
+      "grad_norm": 0.21843822922905395,
+      "learning_rate": 4.056578791910931e-06,
+      "loss": 0.6979,
+      "step": 1755
+    },
+    {
+      "epoch": 4.293398533007335,
+      "grad_norm": 0.20821046782495747,
+      "learning_rate": 4.055576817732693e-06,
+      "loss": 0.4588,
+      "step": 1756
+    },
+    {
+      "epoch": 4.295843520782396,
+      "grad_norm": 0.1719093538131664,
+      "learning_rate": 4.054574435646604e-06,
+      "loss": 0.4315,
+      "step": 1757
+    },
+    {
+      "epoch": 4.298288508557457,
+      "grad_norm": 0.3132694618291291,
+      "learning_rate": 4.0535716459155105e-06,
+      "loss": 0.5837,
+      "step": 1758
+    },
+    {
+      "epoch": 4.300733496332518,
+      "grad_norm": 0.20402567179431644,
+      "learning_rate": 4.052568448802369e-06,
+      "loss": 0.9718,
+      "step": 1759
+    },
+    {
+      "epoch": 4.30317848410758,
+      "grad_norm": 0.14336884048577295,
+      "learning_rate": 4.051564844570239e-06,
+      "loss": 0.4744,
+      "step": 1760
+    },
+    {
+      "epoch": 4.30562347188264,
+      "grad_norm": 0.22774705291664313,
+      "learning_rate": 4.050560833482291e-06,
+      "loss": 0.4027,
+      "step": 1761
+    },
+    {
+      "epoch": 4.308068459657702,
+      "grad_norm": 0.6390468195830807,
+      "learning_rate": 4.0495564158017964e-06,
+      "loss": 0.8346,
+      "step": 1762
+    },
+    {
+      "epoch": 4.310513447432763,
+      "grad_norm": 0.2645342446436882,
+      "learning_rate": 4.048551591792139e-06,
+      "loss": 0.6561,
+      "step": 1763
+    },
+    {
+      "epoch": 4.312958435207824,
+      "grad_norm": 0.21334544502180697,
+      "learning_rate": 4.047546361716806e-06,
+      "loss": 0.3755,
+      "step": 1764
+    },
+    {
+      "epoch": 4.315403422982885,
+      "grad_norm": 0.1988327070302909,
+      "learning_rate": 4.0465407258393905e-06,
+      "loss": 0.4531,
+      "step": 1765
+    },
+    {
+      "epoch": 4.3178484107579465,
+      "grad_norm": 0.20237744161343457,
+      "learning_rate": 4.045534684423596e-06,
+      "loss": 0.6549,
+      "step": 1766
+    },
+    {
+      "epoch": 4.320293398533007,
+      "grad_norm": 0.24120107619189118,
+      "learning_rate": 4.044528237733226e-06,
+      "loss": 0.5196,
+      "step": 1767
+    },
+    {
+      "epoch": 4.322738386308068,
+      "grad_norm": 0.17158185891352729,
+      "learning_rate": 4.043521386032196e-06,
+      "loss": 0.4657,
+      "step": 1768
+    },
+    {
+      "epoch": 4.32518337408313,
+      "grad_norm": 0.22935870136502498,
+      "learning_rate": 4.042514129584526e-06,
+      "loss": 0.5189,
+      "step": 1769
+    },
+    {
+      "epoch": 4.32762836185819,
+      "grad_norm": 0.14087522812290942,
+      "learning_rate": 4.04150646865434e-06,
+      "loss": 0.6564,
+      "step": 1770
+    },
+    {
+      "epoch": 4.330073349633252,
+      "grad_norm": 0.17686764150528303,
+      "learning_rate": 4.040498403505872e-06,
+      "loss": 0.6632,
+      "step": 1771
+    },
+    {
+      "epoch": 4.332518337408313,
+      "grad_norm": 0.19924420889254563,
+      "learning_rate": 4.039489934403458e-06,
+      "loss": 0.725,
+      "step": 1772
+    },
+    {
+      "epoch": 4.334963325183374,
+      "grad_norm": 0.2177841923449248,
+      "learning_rate": 4.038481061611542e-06,
+      "loss": 0.5316,
+      "step": 1773
+    },
+    {
+      "epoch": 4.337408312958435,
+      "grad_norm": 0.201290162040878,
+      "learning_rate": 4.037471785394673e-06,
+      "loss": 1.0606,
+      "step": 1774
+    },
+    {
+      "epoch": 4.3398533007334965,
+      "grad_norm": 0.18398617109739293,
+      "learning_rate": 4.036462106017508e-06,
+      "loss": 0.383,
+      "step": 1775
+    },
+    {
+      "epoch": 4.342298288508557,
+      "grad_norm": 0.3274435806789136,
+      "learning_rate": 4.035452023744807e-06,
+      "loss": 0.9509,
+      "step": 1776
+    },
+    {
+      "epoch": 4.3447432762836184,
+      "grad_norm": 0.22585707793721407,
+      "learning_rate": 4.034441538841437e-06,
+      "loss": 0.4751,
+      "step": 1777
+    },
+    {
+      "epoch": 4.34718826405868,
+      "grad_norm": 0.17996294049670775,
+      "learning_rate": 4.033430651572369e-06,
+      "loss": 0.6846,
+      "step": 1778
+    },
+    {
+      "epoch": 4.34963325183374,
+      "grad_norm": 0.1579337410408365,
+      "learning_rate": 4.032419362202683e-06,
+      "loss": 0.3962,
+      "step": 1779
+    },
+    {
+      "epoch": 4.352078239608802,
+      "grad_norm": 0.18352501604993987,
+      "learning_rate": 4.03140767099756e-06,
+      "loss": 0.7938,
+      "step": 1780
+    },
+    {
+      "epoch": 4.354523227383863,
+      "grad_norm": 0.3073501777722171,
+      "learning_rate": 4.0303955782222905e-06,
+      "loss": 0.4474,
+      "step": 1781
+    },
+    {
+      "epoch": 4.356968215158924,
+      "grad_norm": 0.20519229629056135,
+      "learning_rate": 4.029383084142268e-06,
+      "loss": 0.4718,
+      "step": 1782
+    },
+    {
+      "epoch": 4.359413202933985,
+      "grad_norm": 0.17426326249792826,
+      "learning_rate": 4.02837018902299e-06,
+      "loss": 0.4655,
+      "step": 1783
+    },
+    {
+      "epoch": 4.361858190709047,
+      "grad_norm": 0.20204797164590707,
+      "learning_rate": 4.027356893130062e-06,
+      "loss": 0.451,
+      "step": 1784
+    },
+    {
+      "epoch": 4.364303178484108,
+      "grad_norm": 0.24220725507855714,
+      "learning_rate": 4.026343196729194e-06,
+      "loss": 0.5278,
+      "step": 1785
+    },
+    {
+      "epoch": 4.3667481662591685,
+      "grad_norm": 0.16983783158252388,
+      "learning_rate": 4.0253291000861995e-06,
+      "loss": 0.6553,
+      "step": 1786
+    },
+    {
+      "epoch": 4.36919315403423,
+      "grad_norm": 0.2216676257208036,
+      "learning_rate": 4.0243146034669975e-06,
+      "loss": 0.9118,
+      "step": 1787
+    },
+    {
+      "epoch": 4.371638141809291,
+      "grad_norm": 0.19887613832894624,
+      "learning_rate": 4.023299707137613e-06,
+      "loss": 0.4599,
+      "step": 1788
+    },
+    {
+      "epoch": 4.374083129584352,
+      "grad_norm": 0.17733841855995897,
+      "learning_rate": 4.022284411364176e-06,
+      "loss": 0.4989,
+      "step": 1789
+    },
+    {
+      "epoch": 4.376528117359413,
+      "grad_norm": 0.2700709655444306,
+      "learning_rate": 4.021268716412918e-06,
+      "loss": 0.4295,
+      "step": 1790
+    },
+    {
+      "epoch": 4.378973105134475,
+      "grad_norm": 0.18641684424913657,
+      "learning_rate": 4.020252622550179e-06,
+      "loss": 0.8216,
+      "step": 1791
+    },
+    {
+      "epoch": 4.381418092909535,
+      "grad_norm": 0.25579577790537766,
+      "learning_rate": 4.0192361300424024e-06,
+      "loss": 0.5524,
+      "step": 1792
+    },
+    {
+      "epoch": 4.383863080684597,
+      "grad_norm": 0.1704170266308227,
+      "learning_rate": 4.018219239156134e-06,
+      "loss": 0.4501,
+      "step": 1793
+    },
+    {
+      "epoch": 4.386308068459658,
+      "grad_norm": 0.30986540950700525,
+      "learning_rate": 4.0172019501580275e-06,
+      "loss": 0.5034,
+      "step": 1794
+    },
+    {
+      "epoch": 4.388753056234719,
+      "grad_norm": 0.19879134398740553,
+      "learning_rate": 4.016184263314839e-06,
+      "loss": 0.4996,
+      "step": 1795
+    },
+    {
+      "epoch": 4.39119804400978,
+      "grad_norm": 0.15689007778639283,
+      "learning_rate": 4.015166178893428e-06,
+      "loss": 0.5767,
+      "step": 1796
+    },
+    {
+      "epoch": 4.393643031784841,
+      "grad_norm": 0.22665268849645154,
+      "learning_rate": 4.0141476971607626e-06,
+      "loss": 0.5678,
+      "step": 1797
+    },
+    {
+      "epoch": 4.396088019559902,
+      "grad_norm": 0.1612728689316623,
+      "learning_rate": 4.013128818383909e-06,
+      "loss": 0.5074,
+      "step": 1798
+    },
+    {
+      "epoch": 4.398533007334963,
+      "grad_norm": 0.19454716973470845,
+      "learning_rate": 4.012109542830042e-06,
+      "loss": 0.4193,
+      "step": 1799
+    },
+    {
+      "epoch": 4.400977995110025,
+      "grad_norm": 0.2900809894034894,
+      "learning_rate": 4.011089870766437e-06,
+      "loss": 0.7492,
+      "step": 1800
+    },
+    {
+      "epoch": 4.403422982885085,
+      "grad_norm": 0.17017235076254456,
+      "learning_rate": 4.010069802460479e-06,
+      "loss": 0.4003,
+      "step": 1801
+    },
+    {
+      "epoch": 4.405867970660147,
+      "grad_norm": 0.11863350690988382,
+      "learning_rate": 4.00904933817965e-06,
+      "loss": 0.4585,
+      "step": 1802
+    },
+    {
+      "epoch": 4.408312958435208,
+      "grad_norm": 0.201116212157392,
+      "learning_rate": 4.008028478191541e-06,
+      "loss": 0.3439,
+      "step": 1803
+    },
+    {
+      "epoch": 4.410757946210269,
+      "grad_norm": 0.18935881941306143,
+      "learning_rate": 4.007007222763844e-06,
+      "loss": 0.5309,
+      "step": 1804
+    },
+    {
+      "epoch": 4.41320293398533,
+      "grad_norm": 0.21402870084554498,
+      "learning_rate": 4.005985572164356e-06,
+      "loss": 0.9903,
+      "step": 1805
+    },
+    {
+      "epoch": 4.4156479217603914,
+      "grad_norm": 0.2234688076653641,
+      "learning_rate": 4.004963526660977e-06,
+      "loss": 0.3975,
+      "step": 1806
+    },
+    {
+      "epoch": 4.418092909535452,
+      "grad_norm": 0.20426539988315862,
+      "learning_rate": 4.00394108652171e-06,
+      "loss": 0.6047,
+      "step": 1807
+    },
+    {
+      "epoch": 4.420537897310513,
+      "grad_norm": 0.20070267367364727,
+      "learning_rate": 4.002918252014664e-06,
+      "loss": 0.5787,
+      "step": 1808
+    },
+    {
+      "epoch": 4.422982885085575,
+      "grad_norm": 0.2610563973291742,
+      "learning_rate": 4.001895023408047e-06,
+      "loss": 1.1268,
+      "step": 1809
+    },
+    {
+      "epoch": 4.425427872860635,
+      "grad_norm": 0.20834543458077015,
+      "learning_rate": 4.000871400970175e-06,
+      "loss": 0.5251,
+      "step": 1810
+    },
+    {
+      "epoch": 4.427872860635697,
+      "grad_norm": 0.27337758605500256,
+      "learning_rate": 3.999847384969465e-06,
+      "loss": 0.7662,
+      "step": 1811
+    },
+    {
+      "epoch": 4.430317848410758,
+      "grad_norm": 0.2397016683189441,
+      "learning_rate": 3.998822975674437e-06,
+      "loss": 0.6643,
+      "step": 1812
+    },
+    {
+      "epoch": 4.432762836185819,
+      "grad_norm": 0.1684599967440021,
+      "learning_rate": 3.997798173353714e-06,
+      "loss": 0.7524,
+      "step": 1813
+    },
+    {
+      "epoch": 4.43520782396088,
+      "grad_norm": 0.1904418463849341,
+      "learning_rate": 3.996772978276025e-06,
+      "loss": 0.4726,
+      "step": 1814
+    },
+    {
+      "epoch": 4.4376528117359415,
+      "grad_norm": 0.22925054435124603,
+      "learning_rate": 3.995747390710196e-06,
+      "loss": 0.5277,
+      "step": 1815
+    },
+    {
+      "epoch": 4.440097799511002,
+      "grad_norm": 0.16079024094680633,
+      "learning_rate": 3.994721410925162e-06,
+      "loss": 0.6566,
+      "step": 1816
+    },
+    {
+      "epoch": 4.442542787286063,
+      "grad_norm": 0.2463820434037348,
+      "learning_rate": 3.993695039189957e-06,
+      "loss": 0.7487,
+      "step": 1817
+    },
+    {
+      "epoch": 4.444987775061125,
+      "grad_norm": 0.22008504254238753,
+      "learning_rate": 3.992668275773722e-06,
+      "loss": 0.6086,
+      "step": 1818
+    },
+    {
+      "epoch": 4.447432762836186,
+      "grad_norm": 0.1828612602387857,
+      "learning_rate": 3.9916411209456955e-06,
+      "loss": 0.5678,
+      "step": 1819
+    },
+    {
+      "epoch": 4.449877750611247,
+      "grad_norm": 0.1686064422524913,
+      "learning_rate": 3.99061357497522e-06,
+      "loss": 0.3714,
+      "step": 1820
+    },
+    {
+      "epoch": 4.452322738386308,
+      "grad_norm": 0.23980047548549416,
+      "learning_rate": 3.989585638131744e-06,
+      "loss": 0.609,
+      "step": 1821
+    },
+    {
+      "epoch": 4.45476772616137,
+      "grad_norm": 0.17846990185641398,
+      "learning_rate": 3.988557310684815e-06,
+      "loss": 0.5094,
+      "step": 1822
+    },
+    {
+      "epoch": 4.45721271393643,
+      "grad_norm": 0.21079232393269018,
+      "learning_rate": 3.987528592904084e-06,
+      "loss": 0.6343,
+      "step": 1823
+    },
+    {
+      "epoch": 4.459657701711492,
+      "grad_norm": 0.14001613922297487,
+      "learning_rate": 3.986499485059304e-06,
+      "loss": 0.5633,
+      "step": 1824
+    },
+    {
+      "epoch": 4.462102689486553,
+      "grad_norm": 0.19152436557825264,
+      "learning_rate": 3.985469987420331e-06,
+      "loss": 0.6092,
+      "step": 1825
+    },
+    {
+      "epoch": 4.4645476772616135,
+      "grad_norm": 0.1980740598537105,
+      "learning_rate": 3.984440100257123e-06,
+      "loss": 0.4678,
+      "step": 1826
+    },
+    {
+      "epoch": 4.466992665036675,
+      "grad_norm": 0.2710528962094933,
+      "learning_rate": 3.9834098238397396e-06,
+      "loss": 1.1581,
+      "step": 1827
+    },
+    {
+      "epoch": 4.469437652811736,
+      "grad_norm": 0.22415951174185758,
+      "learning_rate": 3.982379158438344e-06,
+      "loss": 0.4822,
+      "step": 1828
+    },
+    {
+      "epoch": 4.471882640586797,
+      "grad_norm": 0.16679266462959624,
+      "learning_rate": 3.9813481043231985e-06,
+      "loss": 0.421,
+      "step": 1829
+    },
+    {
+      "epoch": 4.474327628361858,
+      "grad_norm": 0.2112976244305874,
+      "learning_rate": 3.980316661764671e-06,
+      "loss": 1.1342,
+      "step": 1830
+    },
+    {
+      "epoch": 4.47677261613692,
+      "grad_norm": 0.24289323730858625,
+      "learning_rate": 3.9792848310332264e-06,
+      "loss": 0.5676,
+      "step": 1831
+    },
+    {
+      "epoch": 4.47921760391198,
+      "grad_norm": 0.22967380686068448,
+      "learning_rate": 3.978252612399437e-06,
+      "loss": 0.509,
+      "step": 1832
+    },
+    {
+      "epoch": 4.481662591687042,
+      "grad_norm": 0.1821182498724709,
+      "learning_rate": 3.977220006133974e-06,
+      "loss": 0.4698,
+      "step": 1833
+    },
+    {
+      "epoch": 4.484107579462103,
+      "grad_norm": 0.14355172592759233,
+      "learning_rate": 3.97618701250761e-06,
+      "loss": 0.4379,
+      "step": 1834
+    },
+    {
+      "epoch": 4.486552567237164,
+      "grad_norm": 0.19990549910478775,
+      "learning_rate": 3.975153631791219e-06,
+      "loss": 0.5842,
+      "step": 1835
+    },
+    {
+      "epoch": 4.488997555012225,
+      "grad_norm": 0.16541384726173608,
+      "learning_rate": 3.9741198642557775e-06,
+      "loss": 0.4023,
+      "step": 1836
+    },
+    {
+      "epoch": 4.491442542787286,
+      "grad_norm": 0.1588940364784366,
+      "learning_rate": 3.973085710172363e-06,
+      "loss": 0.902,
+      "step": 1837
+    },
+    {
+      "epoch": 4.493887530562347,
+      "grad_norm": 0.15597887789605894,
+      "learning_rate": 3.9720511698121545e-06,
+      "loss": 0.4805,
+      "step": 1838
+    },
+    {
+      "epoch": 4.496332518337408,
+      "grad_norm": 0.29329667943109233,
+      "learning_rate": 3.971016243446432e-06,
+      "loss": 1.0666,
+      "step": 1839
+    },
+    {
+      "epoch": 4.49877750611247,
+      "grad_norm": 0.22005140845318127,
+      "learning_rate": 3.969980931346577e-06,
+      "loss": 0.6153,
+      "step": 1840
+    },
+    {
+      "epoch": 4.50122249388753,
+      "grad_norm": 0.1852103371068952,
+      "learning_rate": 3.968945233784071e-06,
+      "loss": 0.465,
+      "step": 1841
+    },
+    {
+      "epoch": 4.503667481662592,
+      "grad_norm": 0.2446345860628552,
+      "learning_rate": 3.967909151030499e-06,
+      "loss": 0.5681,
+      "step": 1842
+    },
+    {
+      "epoch": 4.506112469437653,
+      "grad_norm": 0.24009461048981703,
+      "learning_rate": 3.9668726833575455e-06,
+      "loss": 0.7541,
+      "step": 1843
+    },
+    {
+      "epoch": 4.508557457212714,
+      "grad_norm": 0.14015557083908609,
+      "learning_rate": 3.965835831036995e-06,
+      "loss": 0.6138,
+      "step": 1844
+    },
+    {
+      "epoch": 4.511002444987775,
+      "grad_norm": 0.22249708612642274,
+      "learning_rate": 3.9647985943407354e-06,
+      "loss": 0.3902,
+      "step": 1845
+    },
+    {
+      "epoch": 4.513447432762836,
+      "grad_norm": 0.3066803834758657,
+      "learning_rate": 3.963760973540753e-06,
+      "loss": 0.5112,
+      "step": 1846
+    },
+    {
+      "epoch": 4.515892420537897,
+      "grad_norm": 0.1606742484773916,
+      "learning_rate": 3.962722968909136e-06,
+      "loss": 0.4571,
+      "step": 1847
+    },
+    {
+      "epoch": 4.518337408312958,
+      "grad_norm": 0.18699561638523465,
+      "learning_rate": 3.961684580718072e-06,
+      "loss": 0.5717,
+      "step": 1848
+    },
+    {
+      "epoch": 4.52078239608802,
+      "grad_norm": 0.20952396515852445,
+      "learning_rate": 3.960645809239851e-06,
+      "loss": 0.6622,
+      "step": 1849
+    },
+    {
+      "epoch": 4.52322738386308,
+      "grad_norm": 0.2017567125761346,
+      "learning_rate": 3.959606654746864e-06,
+      "loss": 0.6138,
+      "step": 1850
+    },
+    {
+      "epoch": 4.525672371638142,
+      "grad_norm": 0.1642699895413477,
+      "learning_rate": 3.958567117511598e-06,
+      "loss": 0.6044,
+      "step": 1851
+    },
+    {
+      "epoch": 4.528117359413203,
+      "grad_norm": 0.23514934733691917,
+      "learning_rate": 3.957527197806647e-06,
+      "loss": 0.9963,
+      "step": 1852
+    },
+    {
+      "epoch": 4.530562347188264,
+      "grad_norm": 0.2635508464939336,
+      "learning_rate": 3.956486895904699e-06,
+      "loss": 0.803,
+      "step": 1853
+    },
+    {
+      "epoch": 4.533007334963325,
+      "grad_norm": 0.30614670873905137,
+      "learning_rate": 3.955446212078545e-06,
+      "loss": 0.9068,
+      "step": 1854
+    },
+    {
+      "epoch": 4.5354523227383865,
+      "grad_norm": 0.24150155579032728,
+      "learning_rate": 3.954405146601079e-06,
+      "loss": 0.6967,
+      "step": 1855
+    },
+    {
+      "epoch": 4.537897310513447,
+      "grad_norm": 0.16581157252024892,
+      "learning_rate": 3.953363699745289e-06,
+      "loss": 0.5162,
+      "step": 1856
+    },
+    {
+      "epoch": 4.540342298288508,
+      "grad_norm": 0.2387610117554493,
+      "learning_rate": 3.9523218717842675e-06,
+      "loss": 0.6121,
+      "step": 1857
+    },
+    {
+      "epoch": 4.54278728606357,
+      "grad_norm": 0.17651754729692953,
+      "learning_rate": 3.951279662991206e-06,
+      "loss": 0.4259,
+      "step": 1858
+    },
+    {
+      "epoch": 4.54523227383863,
+      "grad_norm": 0.14499539782092472,
+      "learning_rate": 3.950237073639395e-06,
+      "loss": 0.4067,
+      "step": 1859
+    },
+    {
+      "epoch": 4.547677261613692,
+      "grad_norm": 0.23292429591774189,
+      "learning_rate": 3.949194104002224e-06,
+      "loss": 0.5699,
+      "step": 1860
+    },
+    {
+      "epoch": 4.550122249388753,
+      "grad_norm": 0.1339464462617127,
+      "learning_rate": 3.948150754353187e-06,
+      "loss": 0.3574,
+      "step": 1861
+    },
+    {
+      "epoch": 4.552567237163814,
+      "grad_norm": 0.18376054811835527,
+      "learning_rate": 3.947107024965872e-06,
+      "loss": 0.7236,
+      "step": 1862
+    },
+    {
+      "epoch": 4.555012224938875,
+      "grad_norm": 0.21717306051206267,
+      "learning_rate": 3.946062916113967e-06,
+      "loss": 0.8475,
+      "step": 1863
+    },
+    {
+      "epoch": 4.557457212713937,
+      "grad_norm": 0.216445086430067,
+      "learning_rate": 3.945018428071264e-06,
+      "loss": 0.5497,
+      "step": 1864
+    },
+    {
+      "epoch": 4.559902200488997,
+      "grad_norm": 0.1862245673034554,
+      "learning_rate": 3.943973561111651e-06,
+      "loss": 0.6632,
+      "step": 1865
+    },
+    {
+      "epoch": 4.5623471882640585,
+      "grad_norm": 0.1458842636032291,
+      "learning_rate": 3.942928315509115e-06,
+      "loss": 0.4285,
+      "step": 1866
+    },
+    {
+      "epoch": 4.56479217603912,
+      "grad_norm": 0.14425197581722418,
+      "learning_rate": 3.941882691537745e-06,
+      "loss": 0.5793,
+      "step": 1867
+    },
+    {
+      "epoch": 4.56723716381418,
+      "grad_norm": 0.2811144462792972,
+      "learning_rate": 3.940836689471726e-06,
+      "loss": 0.3542,
+      "step": 1868
+    },
+    {
+      "epoch": 4.569682151589242,
+      "grad_norm": 0.16165048847183508,
+      "learning_rate": 3.939790309585344e-06,
+      "loss": 0.5952,
+      "step": 1869
+    },
+    {
+      "epoch": 4.572127139364303,
+      "grad_norm": 0.3270001859298466,
+      "learning_rate": 3.938743552152985e-06,
+      "loss": 0.4275,
+      "step": 1870
+    },
+    {
+      "epoch": 4.574572127139365,
+      "grad_norm": 0.20009795157423668,
+      "learning_rate": 3.937696417449131e-06,
+      "loss": 0.5025,
+      "step": 1871
+    },
+    {
+      "epoch": 4.577017114914425,
+      "grad_norm": 0.30726422777285173,
+      "learning_rate": 3.936648905748366e-06,
+      "loss": 0.4029,
+      "step": 1872
+    },
+    {
+      "epoch": 4.579462102689487,
+      "grad_norm": 0.1451535724279507,
+      "learning_rate": 3.9356010173253714e-06,
+      "loss": 0.4649,
+      "step": 1873
+    },
+    {
+      "epoch": 4.581907090464548,
+      "grad_norm": 0.17192979659205962,
+      "learning_rate": 3.934552752454926e-06,
+      "loss": 0.7784,
+      "step": 1874
+    },
+    {
+      "epoch": 4.5843520782396086,
+      "grad_norm": 0.23242553728874452,
+      "learning_rate": 3.93350411141191e-06,
+      "loss": 0.525,
+      "step": 1875
+    },
+    {
+      "epoch": 4.58679706601467,
+      "grad_norm": 0.22615294912869327,
+      "learning_rate": 3.932455094471303e-06,
+      "loss": 0.4938,
+      "step": 1876
+    },
+    {
+      "epoch": 4.589242053789731,
+      "grad_norm": 0.16984893052173727,
+      "learning_rate": 3.931405701908177e-06,
+      "loss": 0.5479,
+      "step": 1877
+    },
+    {
+      "epoch": 4.591687041564792,
+      "grad_norm": 0.19651716525114596,
+      "learning_rate": 3.9303559339977114e-06,
+      "loss": 0.7397,
+      "step": 1878
+    },
+    {
+      "epoch": 4.594132029339853,
+      "grad_norm": 0.1595045025638794,
+      "learning_rate": 3.9293057910151765e-06,
+      "loss": 0.7098,
+      "step": 1879
+    },
+    {
+      "epoch": 4.596577017114915,
+      "grad_norm": 0.2809518858234133,
+      "learning_rate": 3.928255273235943e-06,
+      "loss": 0.6728,
+      "step": 1880
+    },
+    {
+      "epoch": 4.599022004889975,
+      "grad_norm": 0.3305036299706136,
+      "learning_rate": 3.927204380935483e-06,
+      "loss": 0.5246,
+      "step": 1881
+    },
+    {
+      "epoch": 4.601466992665037,
+      "grad_norm": 0.22271749803837831,
+      "learning_rate": 3.926153114389364e-06,
+      "loss": 0.782,
+      "step": 1882
+    },
+    {
+      "epoch": 4.603911980440098,
+      "grad_norm": 0.20451195584354562,
+      "learning_rate": 3.925101473873252e-06,
+      "loss": 0.6166,
+      "step": 1883
+    },
+    {
+      "epoch": 4.606356968215159,
+      "grad_norm": 0.19031764205979998,
+      "learning_rate": 3.924049459662911e-06,
+      "loss": 0.4109,
+      "step": 1884
+    },
+    {
+      "epoch": 4.60880195599022,
+      "grad_norm": 0.23794814895868074,
+      "learning_rate": 3.922997072034202e-06,
+      "loss": 0.3295,
+      "step": 1885
+    },
+    {
+      "epoch": 4.611246943765281,
+      "grad_norm": 0.1779918921973669,
+      "learning_rate": 3.921944311263088e-06,
+      "loss": 0.4323,
+      "step": 1886
+    },
+    {
+      "epoch": 4.613691931540342,
+      "grad_norm": 0.20230169512440785,
+      "learning_rate": 3.920891177625623e-06,
+      "loss": 0.441,
+      "step": 1887
+    },
+    {
+      "epoch": 4.616136919315403,
+      "grad_norm": 0.20682415426651563,
+      "learning_rate": 3.9198376713979675e-06,
+      "loss": 0.8208,
+      "step": 1888
+    },
+    {
+      "epoch": 4.618581907090465,
+      "grad_norm": 0.21065620274465463,
+      "learning_rate": 3.9187837928563725e-06,
+      "loss": 0.9154,
+      "step": 1889
+    },
+    {
+      "epoch": 4.621026894865525,
+      "grad_norm": 0.257671479348446,
+      "learning_rate": 3.917729542277188e-06,
+      "loss": 0.5919,
+      "step": 1890
+    },
+    {
+      "epoch": 4.623471882640587,
+      "grad_norm": 0.1594716575405407,
+      "learning_rate": 3.916674919936864e-06,
+      "loss": 0.4284,
+      "step": 1891
+    },
+    {
+      "epoch": 4.625916870415648,
+      "grad_norm": 0.1444027691432101,
+      "learning_rate": 3.9156199261119455e-06,
+      "loss": 0.6073,
+      "step": 1892
+    },
+    {
+      "epoch": 4.628361858190709,
+      "grad_norm": 0.18558313884434222,
+      "learning_rate": 3.914564561079078e-06,
+      "loss": 0.898,
+      "step": 1893
+    },
+    {
+      "epoch": 4.63080684596577,
+      "grad_norm": 0.3672970379679832,
+      "learning_rate": 3.913508825115001e-06,
+      "loss": 0.3401,
+      "step": 1894
+    },
+    {
+      "epoch": 4.6332518337408315,
+      "grad_norm": 0.18765576347420218,
+      "learning_rate": 3.912452718496554e-06,
+      "loss": 0.5971,
+      "step": 1895
+    },
+    {
+      "epoch": 4.635696821515893,
+      "grad_norm": 0.18746003819447513,
+      "learning_rate": 3.911396241500669e-06,
+      "loss": 0.5445,
+      "step": 1896
+    },
+    {
+      "epoch": 4.638141809290953,
+      "grad_norm": 0.15935878787748933,
+      "learning_rate": 3.910339394404382e-06,
+      "loss": 0.4754,
+      "step": 1897
+    },
+    {
+      "epoch": 4.640586797066015,
+      "grad_norm": 0.2236216344747909,
+      "learning_rate": 3.909282177484821e-06,
+      "loss": 0.4526,
+      "step": 1898
+    },
+    {
+      "epoch": 4.643031784841076,
+      "grad_norm": 0.22629157971195515,
+      "learning_rate": 3.908224591019213e-06,
+      "loss": 0.4919,
+      "step": 1899
+    },
+    {
+      "epoch": 4.645476772616137,
+      "grad_norm": 0.1903794903127352,
+      "learning_rate": 3.907166635284881e-06,
+      "loss": 0.9723,
+      "step": 1900
+    },
+    {
+      "epoch": 4.647921760391198,
+      "grad_norm": 0.2997041759211504,
+      "learning_rate": 3.9061083105592455e-06,
+      "loss": 0.7076,
+      "step": 1901
+    },
+    {
+      "epoch": 4.65036674816626,
+      "grad_norm": 0.15998817367315762,
+      "learning_rate": 3.905049617119822e-06,
+      "loss": 0.9578,
+      "step": 1902
+    },
+    {
+      "epoch": 4.65281173594132,
+      "grad_norm": 0.2623170736236951,
+      "learning_rate": 3.903990555244227e-06,
+      "loss": 0.5527,
+      "step": 1903
+    },
+    {
+      "epoch": 4.6552567237163816,
+      "grad_norm": 0.2216462606613719,
+      "learning_rate": 3.902931125210168e-06,
+      "loss": 0.8426,
+      "step": 1904
+    },
+    {
+      "epoch": 4.657701711491443,
+      "grad_norm": 0.20229204987759317,
+      "learning_rate": 3.901871327295453e-06,
+      "loss": 0.4246,
+      "step": 1905
+    },
+    {
+      "epoch": 4.6601466992665035,
+      "grad_norm": 0.26263612350646326,
+      "learning_rate": 3.9008111617779855e-06,
+      "loss": 0.7432,
+      "step": 1906
+    },
+    {
+      "epoch": 4.662591687041565,
+      "grad_norm": 0.2947146796861689,
+      "learning_rate": 3.899750628935764e-06,
+      "loss": 0.8694,
+      "step": 1907
+    },
+    {
+      "epoch": 4.665036674816626,
+      "grad_norm": 0.18774811915838951,
+      "learning_rate": 3.8986897290468875e-06,
+      "loss": 0.5394,
+      "step": 1908
+    },
+    {
+      "epoch": 4.667481662591687,
+      "grad_norm": 0.21467511827136018,
+      "learning_rate": 3.897628462389544e-06,
+      "loss": 0.6262,
+      "step": 1909
+    },
+    {
+      "epoch": 4.669926650366748,
+      "grad_norm": 0.1456336371248034,
+      "learning_rate": 3.896566829242025e-06,
+      "loss": 0.8374,
+      "step": 1910
+    },
+    {
+      "epoch": 4.67237163814181,
+      "grad_norm": 0.2131324176040151,
+      "learning_rate": 3.895504829882715e-06,
+      "loss": 0.3847,
+      "step": 1911
+    },
+    {
+      "epoch": 4.67481662591687,
+      "grad_norm": 0.15910066798828792,
+      "learning_rate": 3.894442464590092e-06,
+      "loss": 0.39,
+      "step": 1912
+    },
+    {
+      "epoch": 4.677261613691932,
+      "grad_norm": 0.18842476919132325,
+      "learning_rate": 3.8933797336427345e-06,
+      "loss": 0.5324,
+      "step": 1913
+    },
+    {
+      "epoch": 4.679706601466993,
+      "grad_norm": 0.2005519473645647,
+      "learning_rate": 3.8923166373193146e-06,
+      "loss": 0.5339,
+      "step": 1914
+    },
+    {
+      "epoch": 4.6821515892420535,
+      "grad_norm": 0.1973795478010137,
+      "learning_rate": 3.891253175898601e-06,
+      "loss": 0.7942,
+      "step": 1915
+    },
+    {
+      "epoch": 4.684596577017115,
+      "grad_norm": 0.2510740378252299,
+      "learning_rate": 3.8901893496594565e-06,
+      "loss": 0.5394,
+      "step": 1916
+    },
+    {
+      "epoch": 4.687041564792176,
+      "grad_norm": 0.23551798025366968,
+      "learning_rate": 3.889125158880841e-06,
+      "loss": 0.4284,
+      "step": 1917
+    },
+    {
+      "epoch": 4.689486552567237,
+      "grad_norm": 0.24952782997015438,
+      "learning_rate": 3.88806060384181e-06,
+      "loss": 0.4171,
+      "step": 1918
+    },
+    {
+      "epoch": 4.691931540342298,
+      "grad_norm": 0.2479449163759259,
+      "learning_rate": 3.886995684821514e-06,
+      "loss": 0.4515,
+      "step": 1919
+    },
+    {
+      "epoch": 4.69437652811736,
+      "grad_norm": 0.1931792617800815,
+      "learning_rate": 3.8859304020992e-06,
+      "loss": 0.6077,
+      "step": 1920
+    },
+    {
+      "epoch": 4.69682151589242,
+      "grad_norm": 0.19937564088534307,
+      "learning_rate": 3.884864755954207e-06,
+      "loss": 0.8667,
+      "step": 1921
+    },
+    {
+      "epoch": 4.699266503667482,
+      "grad_norm": 0.21468890912171396,
+      "learning_rate": 3.883798746665975e-06,
+      "loss": 0.4879,
+      "step": 1922
+    },
+    {
+      "epoch": 4.701711491442543,
+      "grad_norm": 0.21647069204607092,
+      "learning_rate": 3.882732374514033e-06,
+      "loss": 0.4702,
+      "step": 1923
+    },
+    {
+      "epoch": 4.704156479217604,
+      "grad_norm": 0.18320323555742968,
+      "learning_rate": 3.881665639778012e-06,
+      "loss": 0.4358,
+      "step": 1924
+    },
+    {
+      "epoch": 4.706601466992665,
+      "grad_norm": 0.2699457173555042,
+      "learning_rate": 3.88059854273763e-06,
+      "loss": 0.6853,
+      "step": 1925
+    },
+    {
+      "epoch": 4.709046454767726,
+      "grad_norm": 0.14644264747846136,
+      "learning_rate": 3.879531083672708e-06,
+      "loss": 0.4755,
+      "step": 1926
+    },
+    {
+      "epoch": 4.711491442542787,
+      "grad_norm": 0.23900530595740624,
+      "learning_rate": 3.8784632628631555e-06,
+      "loss": 0.6236,
+      "step": 1927
+    },
+    {
+      "epoch": 4.713936430317848,
+      "grad_norm": 0.22967985501134436,
+      "learning_rate": 3.87739508058898e-06,
+      "loss": 0.8071,
+      "step": 1928
+    },
+    {
+      "epoch": 4.71638141809291,
+      "grad_norm": 0.16133319269491517,
+      "learning_rate": 3.876326537130285e-06,
+      "loss": 0.4893,
+      "step": 1929
+    },
+    {
+      "epoch": 4.71882640586797,
+      "grad_norm": 0.20475268684471065,
+      "learning_rate": 3.875257632767267e-06,
+      "loss": 0.4853,
+      "step": 1930
+    },
+    {
+      "epoch": 4.721271393643032,
+      "grad_norm": 0.2417358906988055,
+      "learning_rate": 3.874188367780214e-06,
+      "loss": 0.3741,
+      "step": 1931
+    },
+    {
+      "epoch": 4.723716381418093,
+      "grad_norm": 0.1880112607770472,
+      "learning_rate": 3.873118742449515e-06,
+      "loss": 0.5762,
+      "step": 1932
+    },
+    {
+      "epoch": 4.726161369193154,
+      "grad_norm": 0.1882466959111549,
+      "learning_rate": 3.8720487570556485e-06,
+      "loss": 0.6927,
+      "step": 1933
+    },
+    {
+      "epoch": 4.728606356968215,
+      "grad_norm": 0.14870883943225105,
+      "learning_rate": 3.87097841187919e-06,
+      "loss": 0.3591,
+      "step": 1934
+    },
+    {
+      "epoch": 4.7310513447432765,
+      "grad_norm": 0.20881363835655278,
+      "learning_rate": 3.869907707200809e-06,
+      "loss": 0.5028,
+      "step": 1935
+    },
+    {
+      "epoch": 4.733496332518337,
+      "grad_norm": 0.17361856461676922,
+      "learning_rate": 3.868836643301268e-06,
+      "loss": 0.5445,
+      "step": 1936
+    },
+    {
+      "epoch": 4.735941320293398,
+      "grad_norm": 0.16035092661048211,
+      "learning_rate": 3.867765220461424e-06,
+      "loss": 0.6175,
+      "step": 1937
+    },
+    {
+      "epoch": 4.73838630806846,
+      "grad_norm": 0.23681248214655376,
+      "learning_rate": 3.866693438962229e-06,
+      "loss": 0.7777,
+      "step": 1938
+    },
+    {
+      "epoch": 4.74083129584352,
+      "grad_norm": 0.16443145304661844,
+      "learning_rate": 3.86562129908473e-06,
+      "loss": 0.7213,
+      "step": 1939
+    },
+    {
+      "epoch": 4.743276283618582,
+      "grad_norm": 0.23362913143257044,
+      "learning_rate": 3.864548801110064e-06,
+      "loss": 0.697,
+      "step": 1940
+    },
+    {
+      "epoch": 4.745721271393643,
+      "grad_norm": 0.17138185489462232,
+      "learning_rate": 3.863475945319467e-06,
+      "loss": 0.5086,
+      "step": 1941
+    },
+    {
+      "epoch": 4.748166259168704,
+      "grad_norm": 0.27202414886424636,
+      "learning_rate": 3.8624027319942646e-06,
+      "loss": 0.4491,
+      "step": 1942
+    },
+    {
+      "epoch": 4.750611246943765,
+      "grad_norm": 0.22256250420000104,
+      "learning_rate": 3.86132916141588e-06,
+      "loss": 0.6502,
+      "step": 1943
+    },
+    {
+      "epoch": 4.7530562347188265,
+      "grad_norm": 0.21566192500470766,
+      "learning_rate": 3.860255233865825e-06,
+      "loss": 0.7884,
+      "step": 1944
+    },
+    {
+      "epoch": 4.755501222493887,
+      "grad_norm": 0.19536353880397456,
+      "learning_rate": 3.8591809496257105e-06,
+      "loss": 0.484,
+      "step": 1945
+    },
+    {
+      "epoch": 4.7579462102689485,
+      "grad_norm": 0.21020061158050243,
+      "learning_rate": 3.858106308977236e-06,
+      "loss": 0.3889,
+      "step": 1946
+    },
+    {
+      "epoch": 4.76039119804401,
+      "grad_norm": 0.2123902171862699,
+      "learning_rate": 3.857031312202201e-06,
+      "loss": 0.7079,
+      "step": 1947
+    },
+    {
+      "epoch": 4.76283618581907,
+      "grad_norm": 0.15343300741797006,
+      "learning_rate": 3.855955959582489e-06,
+      "loss": 0.5779,
+      "step": 1948
+    },
+    {
+      "epoch": 4.765281173594132,
+      "grad_norm": 0.2455924513677434,
+      "learning_rate": 3.8548802514000865e-06,
+      "loss": 0.9137,
+      "step": 1949
+    },
+    {
+      "epoch": 4.767726161369193,
+      "grad_norm": 0.15175211624113477,
+      "learning_rate": 3.853804187937066e-06,
+      "loss": 0.6061,
+      "step": 1950
+    },
+    {
+      "epoch": 4.770171149144255,
+      "grad_norm": 0.2901670525217353,
+      "learning_rate": 3.852727769475598e-06,
+      "loss": 0.4708,
+      "step": 1951
+    },
+    {
+      "epoch": 4.772616136919315,
+      "grad_norm": 0.2189426443335556,
+      "learning_rate": 3.851650996297942e-06,
+      "loss": 0.4331,
+      "step": 1952
+    },
+    {
+      "epoch": 4.775061124694377,
+      "grad_norm": 0.21269033874983664,
+      "learning_rate": 3.850573868686455e-06,
+      "loss": 0.4559,
+      "step": 1953
+    },
+    {
+      "epoch": 4.777506112469438,
+      "grad_norm": 0.20700304799388738,
+      "learning_rate": 3.849496386923581e-06,
+      "loss": 0.8294,
+      "step": 1954
+    },
+    {
+      "epoch": 4.7799511002444985,
+      "grad_norm": 0.2575916673336584,
+      "learning_rate": 3.848418551291863e-06,
+      "loss": 0.6847,
+      "step": 1955
+    },
+    {
+      "epoch": 4.78239608801956,
+      "grad_norm": 0.19198169268911439,
+      "learning_rate": 3.847340362073936e-06,
+      "loss": 1.0064,
+      "step": 1956
+    },
+    {
+      "epoch": 4.784841075794621,
+      "grad_norm": 0.21795549797020738,
+      "learning_rate": 3.846261819552522e-06,
+      "loss": 0.4946,
+      "step": 1957
+    },
+    {
+      "epoch": 4.787286063569682,
+      "grad_norm": 0.26371608561823984,
+      "learning_rate": 3.845182924010442e-06,
+      "loss": 0.4436,
+      "step": 1958
+    },
+    {
+      "epoch": 4.789731051344743,
+      "grad_norm": 0.2056800555113254,
+      "learning_rate": 3.844103675730605e-06,
+      "loss": 0.8141,
+      "step": 1959
+    },
+    {
+      "epoch": 4.792176039119805,
+      "grad_norm": 0.11243758951674035,
+      "learning_rate": 3.843024074996016e-06,
+      "loss": 0.3891,
+      "step": 1960
+    },
+    {
+      "epoch": 4.794621026894865,
+      "grad_norm": 0.26235840690026363,
+      "learning_rate": 3.841944122089771e-06,
+      "loss": 0.4628,
+      "step": 1961
+    },
+    {
+      "epoch": 4.797066014669927,
+      "grad_norm": 0.12164715714691905,
+      "learning_rate": 3.840863817295058e-06,
+      "loss": 0.5938,
+      "step": 1962
+    },
+    {
+      "epoch": 4.799511002444988,
+      "grad_norm": 0.21038809430988087,
+      "learning_rate": 3.83978316089516e-06,
+      "loss": 0.6312,
+      "step": 1963
+    },
+    {
+      "epoch": 4.801955990220049,
+      "grad_norm": 0.3013022160166754,
+      "learning_rate": 3.838702153173445e-06,
+      "loss": 0.5416,
+      "step": 1964
+    },
+    {
+      "epoch": 4.80440097799511,
+      "grad_norm": 0.261301971960218,
+      "learning_rate": 3.837620794413382e-06,
+      "loss": 0.5817,
+      "step": 1965
+    },
+    {
+      "epoch": 4.806845965770171,
+      "grad_norm": 0.2614020785175297,
+      "learning_rate": 3.836539084898526e-06,
+      "loss": 0.4594,
+      "step": 1966
+    },
+    {
+      "epoch": 4.809290953545232,
+      "grad_norm": 0.19587007139984464,
+      "learning_rate": 3.835457024912528e-06,
+      "loss": 0.6377,
+      "step": 1967
+    },
+    {
+      "epoch": 4.811735941320293,
+      "grad_norm": 0.17601643732403502,
+      "learning_rate": 3.8343746147391266e-06,
+      "loss": 0.9869,
+      "step": 1968
+    },
+    {
+      "epoch": 4.814180929095355,
+      "grad_norm": 0.22464474292812306,
+      "learning_rate": 3.833291854662156e-06,
+      "loss": 0.6477,
+      "step": 1969
+    },
+    {
+      "epoch": 4.816625916870415,
+      "grad_norm": 0.21213694340223635,
+      "learning_rate": 3.8322087449655395e-06,
+      "loss": 1.0335,
+      "step": 1970
+    },
+    {
+      "epoch": 4.819070904645477,
+      "grad_norm": 0.2874830695096747,
+      "learning_rate": 3.8311252859332945e-06,
+      "loss": 0.5767,
+      "step": 1971
+    },
+    {
+      "epoch": 4.821515892420538,
+      "grad_norm": 0.20176036799039174,
+      "learning_rate": 3.8300414778495285e-06,
+      "loss": 0.5538,
+      "step": 1972
+    },
+    {
+      "epoch": 4.823960880195599,
+      "grad_norm": 0.10918902686791593,
+      "learning_rate": 3.828957320998441e-06,
+      "loss": 0.3323,
+      "step": 1973
+    },
+    {
+      "epoch": 4.82640586797066,
+      "grad_norm": 0.15742143109472015,
+      "learning_rate": 3.8278728156643216e-06,
+      "loss": 0.4958,
+      "step": 1974
+    },
+    {
+      "epoch": 4.8288508557457215,
+      "grad_norm": 0.17326474685642712,
+      "learning_rate": 3.826787962131553e-06,
+      "loss": 0.4217,
+      "step": 1975
+    },
+    {
+      "epoch": 4.831295843520783,
+      "grad_norm": 0.22096924192095757,
+      "learning_rate": 3.82570276068461e-06,
+      "loss": 1.1507,
+      "step": 1976
+    },
+    {
+      "epoch": 4.833740831295843,
+      "grad_norm": 0.16739412855243058,
+      "learning_rate": 3.824617211608054e-06,
+      "loss": 0.6572,
+      "step": 1977
+    },
+    {
+      "epoch": 4.836185819070905,
+      "grad_norm": 0.19427156035346388,
+      "learning_rate": 3.823531315186544e-06,
+      "loss": 0.6499,
+      "step": 1978
+    },
+    {
+      "epoch": 4.838630806845966,
+      "grad_norm": 0.2020682508750726,
+      "learning_rate": 3.822445071704826e-06,
+      "loss": 0.6886,
+      "step": 1979
+    },
+    {
+      "epoch": 4.841075794621027,
+      "grad_norm": 0.21305664021715248,
+      "learning_rate": 3.8213584814477365e-06,
+      "loss": 0.4027,
+      "step": 1980
+    },
+    {
+      "epoch": 4.843520782396088,
+      "grad_norm": 0.19729619654737715,
+      "learning_rate": 3.8202715447002055e-06,
+      "loss": 0.7101,
+      "step": 1981
+    },
+    {
+      "epoch": 4.84596577017115,
+      "grad_norm": 0.32905442981255023,
+      "learning_rate": 3.8191842617472535e-06,
+      "loss": 0.6772,
+      "step": 1982
+    },
+    {
+      "epoch": 4.84841075794621,
+      "grad_norm": 0.14138651999902113,
+      "learning_rate": 3.81809663287399e-06,
+      "loss": 0.494,
+      "step": 1983
+    },
+    {
+      "epoch": 4.8508557457212715,
+      "grad_norm": 0.25353772888397397,
+      "learning_rate": 3.817008658365617e-06,
+      "loss": 0.438,
+      "step": 1984
+    },
+    {
+      "epoch": 4.853300733496333,
+      "grad_norm": 0.21700538081063245,
+      "learning_rate": 3.815920338507424e-06,
+      "loss": 0.7767,
+      "step": 1985
+    },
+    {
+      "epoch": 4.8557457212713935,
+      "grad_norm": 0.24209812091480867,
+      "learning_rate": 3.814831673584796e-06,
+      "loss": 0.6428,
+      "step": 1986
+    },
+    {
+      "epoch": 4.858190709046455,
+      "grad_norm": 0.2539413766704324,
+      "learning_rate": 3.813742663883204e-06,
+      "loss": 0.618,
+      "step": 1987
+    },
+    {
+      "epoch": 4.860635696821516,
+      "grad_norm": 0.15096846661600963,
+      "learning_rate": 3.8126533096882133e-06,
+      "loss": 0.4192,
+      "step": 1988
+    },
+    {
+      "epoch": 4.863080684596577,
+      "grad_norm": 0.17490886731874755,
+      "learning_rate": 3.811563611285476e-06,
+      "loss": 0.4625,
+      "step": 1989
+    },
+    {
+      "epoch": 4.865525672371638,
+      "grad_norm": 0.31723157663273943,
+      "learning_rate": 3.8104735689607366e-06,
+      "loss": 1.1832,
+      "step": 1990
+    },
+    {
+      "epoch": 4.8679706601467,
+      "grad_norm": 0.22811169507031812,
+      "learning_rate": 3.8093831829998285e-06,
+      "loss": 0.714,
+      "step": 1991
+    },
+    {
+      "epoch": 4.87041564792176,
+      "grad_norm": 0.241903334113044,
+      "learning_rate": 3.808292453688677e-06,
+      "loss": 0.8457,
+      "step": 1992
+    },
+    {
+      "epoch": 4.872860635696822,
+      "grad_norm": 0.2608841564707084,
+      "learning_rate": 3.8072013813132946e-06,
+      "loss": 1.2257,
+      "step": 1993
+    },
+    {
+      "epoch": 4.875305623471883,
+      "grad_norm": 0.20358449819275207,
+      "learning_rate": 3.8061099661597862e-06,
+      "loss": 1.1289,
+      "step": 1994
+    },
+    {
+      "epoch": 4.8777506112469435,
+      "grad_norm": 0.19425777775823166,
+      "learning_rate": 3.8050182085143468e-06,
+      "loss": 0.6455,
+      "step": 1995
+    },
+    {
+      "epoch": 4.880195599022005,
+      "grad_norm": 0.17592971668666774,
+      "learning_rate": 3.8039261086632594e-06,
+      "loss": 0.5491,
+      "step": 1996
+    },
+    {
+      "epoch": 4.882640586797066,
+      "grad_norm": 0.1708452357251292,
+      "learning_rate": 3.802833666892898e-06,
+      "loss": 0.5469,
+      "step": 1997
+    },
+    {
+      "epoch": 4.885085574572127,
+      "grad_norm": 0.2092071207069566,
+      "learning_rate": 3.801740883489724e-06,
+      "loss": 0.5994,
+      "step": 1998
+    },
+    {
+      "epoch": 4.887530562347188,
+      "grad_norm": 0.1986278689043537,
+      "learning_rate": 3.8006477587402923e-06,
+      "loss": 0.6345,
+      "step": 1999
+    },
+    {
+      "epoch": 4.88997555012225,
+      "grad_norm": 0.3686688065604612,
+      "learning_rate": 3.7995542929312456e-06,
+      "loss": 0.5568,
+      "step": 2000
+    },
+    {
+      "epoch": 4.89242053789731,
+      "grad_norm": 0.1557762142140868,
+      "learning_rate": 3.798460486349314e-06,
+      "loss": 0.6145,
+      "step": 2001
+    },
+    {
+      "epoch": 4.894865525672372,
+      "grad_norm": 0.15066400384436882,
+      "learning_rate": 3.79736633928132e-06,
+      "loss": 0.4007,
+      "step": 2002
+    },
+    {
+      "epoch": 4.897310513447433,
+      "grad_norm": 0.33131998490164694,
+      "learning_rate": 3.796271852014172e-06,
+      "loss": 0.4809,
+      "step": 2003
+    },
+    {
+      "epoch": 4.899755501222494,
+      "grad_norm": 0.23293494050755983,
+      "learning_rate": 3.795177024834872e-06,
+      "loss": 0.4936,
+      "step": 2004
+    },
+    {
+      "epoch": 4.902200488997555,
+      "grad_norm": 0.2000899059101259,
+      "learning_rate": 3.794081858030507e-06,
+      "loss": 0.8239,
+      "step": 2005
+    },
+    {
+      "epoch": 4.904645476772616,
+      "grad_norm": 0.307332489168978,
+      "learning_rate": 3.792986351888257e-06,
+      "loss": 0.3519,
+      "step": 2006
+    },
+    {
+      "epoch": 4.907090464547677,
+      "grad_norm": 0.30272302421376335,
+      "learning_rate": 3.791890506695386e-06,
+      "loss": 0.648,
+      "step": 2007
+    },
+    {
+      "epoch": 4.909535452322738,
+      "grad_norm": 0.19490624767310388,
+      "learning_rate": 3.790794322739251e-06,
+      "loss": 0.5777,
+      "step": 2008
+    },
+    {
+      "epoch": 4.9119804400978,
+      "grad_norm": 0.26806768413369647,
+      "learning_rate": 3.789697800307297e-06,
+      "loss": 0.6835,
+      "step": 2009
+    },
+    {
+      "epoch": 4.91442542787286,
+      "grad_norm": 0.20663552404159946,
+      "learning_rate": 3.788600939687057e-06,
+      "loss": 0.8178,
+      "step": 2010
+    },
+    {
+      "epoch": 4.916870415647922,
+      "grad_norm": 0.19750992003682344,
+      "learning_rate": 3.7875037411661524e-06,
+      "loss": 0.587,
+      "step": 2011
+    },
+    {
+      "epoch": 4.919315403422983,
+      "grad_norm": 0.21485029330342917,
+      "learning_rate": 3.7864062050322946e-06,
+      "loss": 0.5488,
+      "step": 2012
+    },
+    {
+      "epoch": 4.921760391198044,
+      "grad_norm": 0.26926451689765024,
+      "learning_rate": 3.785308331573282e-06,
+      "loss": 0.654,
+      "step": 2013
+    },
+    {
+      "epoch": 4.924205378973105,
+      "grad_norm": 0.18553515270981613,
+      "learning_rate": 3.7842101210770015e-06,
+      "loss": 0.7023,
+      "step": 2014
+    },
+    {
+      "epoch": 4.9266503667481665,
+      "grad_norm": 0.21393325805719873,
+      "learning_rate": 3.7831115738314304e-06,
+      "loss": 0.3538,
+      "step": 2015
+    },
+    {
+      "epoch": 4.929095354523227,
+      "grad_norm": 0.23201438187328913,
+      "learning_rate": 3.7820126901246333e-06,
+      "loss": 0.6286,
+      "step": 2016
+    },
+    {
+      "epoch": 4.931540342298288,
+      "grad_norm": 0.21047620000930423,
+      "learning_rate": 3.7809134702447612e-06,
+      "loss": 0.3687,
+      "step": 2017
+    },
+    {
+      "epoch": 4.93398533007335,
+      "grad_norm": 0.406652349391441,
+      "learning_rate": 3.7798139144800556e-06,
+      "loss": 0.5549,
+      "step": 2018
+    },
+    {
+      "epoch": 4.93643031784841,
+      "grad_norm": 0.17129572422517605,
+      "learning_rate": 3.7787140231188447e-06,
+      "loss": 0.6089,
+      "step": 2019
+    },
+    {
+      "epoch": 4.938875305623472,
+      "grad_norm": 0.262309863813515,
+      "learning_rate": 3.777613796449545e-06,
+      "loss": 0.5803,
+      "step": 2020
+    },
+    {
+      "epoch": 4.941320293398533,
+      "grad_norm": 0.17737294663134487,
+      "learning_rate": 3.7765132347606624e-06,
+      "loss": 0.9221,
+      "step": 2021
+    },
+    {
+      "epoch": 4.943765281173594,
+      "grad_norm": 0.23111015373190685,
+      "learning_rate": 3.775412338340788e-06,
+      "loss": 0.8123,
+      "step": 2022
+    },
+    {
+      "epoch": 4.946210268948655,
+      "grad_norm": 0.18297921749076518,
+      "learning_rate": 3.774311107478603e-06,
+      "loss": 0.4031,
+      "step": 2023
+    },
+    {
+      "epoch": 4.9486552567237165,
+      "grad_norm": 0.2586420993427602,
+      "learning_rate": 3.7732095424628745e-06,
+      "loss": 0.3515,
+      "step": 2024
+    },
+    {
+      "epoch": 4.951100244498777,
+      "grad_norm": 0.1539691933071811,
+      "learning_rate": 3.772107643582459e-06,
+      "loss": 0.5956,
+      "step": 2025
+    },
+    {
+      "epoch": 4.9535452322738385,
+      "grad_norm": 0.24451172645146627,
+      "learning_rate": 3.7710054111262983e-06,
+      "loss": 0.5956,
+      "step": 2026
+    },
+    {
+      "epoch": 4.9559902200489,
+      "grad_norm": 0.27555798407071436,
+      "learning_rate": 3.769902845383424e-06,
+      "loss": 0.452,
+      "step": 2027
+    },
+    {
+      "epoch": 4.95843520782396,
+      "grad_norm": 0.17389396108444852,
+      "learning_rate": 3.7687999466429537e-06,
+      "loss": 0.5294,
+      "step": 2028
+    },
+    {
+      "epoch": 4.960880195599022,
+      "grad_norm": 0.16614482374299225,
+      "learning_rate": 3.7676967151940934e-06,
+      "loss": 0.578,
+      "step": 2029
+    },
+    {
+      "epoch": 4.963325183374083,
+      "grad_norm": 0.21186167741105447,
+      "learning_rate": 3.7665931513261343e-06,
+      "loss": 0.6392,
+      "step": 2030
+    },
+    {
+      "epoch": 4.965770171149144,
+      "grad_norm": 0.15095146052310937,
+      "learning_rate": 3.7654892553284572e-06,
+      "loss": 0.5462,
+      "step": 2031
+    },
+    {
+      "epoch": 4.968215158924205,
+      "grad_norm": 0.19698066864023028,
+      "learning_rate": 3.764385027490529e-06,
+      "loss": 0.9083,
+      "step": 2032
+    },
+    {
+      "epoch": 4.970660146699267,
+      "grad_norm": 0.21849600590034604,
+      "learning_rate": 3.7632804681019026e-06,
+      "loss": 0.8427,
+      "step": 2033
+    },
+    {
+      "epoch": 4.973105134474328,
+      "grad_norm": 0.1444160031193256,
+      "learning_rate": 3.7621755774522194e-06,
+      "loss": 0.5039,
+      "step": 2034
+    },
+    {
+      "epoch": 4.9755501222493885,
+      "grad_norm": 0.1708550844064277,
+      "learning_rate": 3.7610703558312066e-06,
+      "loss": 0.6111,
+      "step": 2035
+    },
+    {
+      "epoch": 4.97799511002445,
+      "grad_norm": 0.16775642790513573,
+      "learning_rate": 3.7599648035286793e-06,
+      "loss": 0.5445,
+      "step": 2036
+    },
+    {
+      "epoch": 4.980440097799511,
+      "grad_norm": 0.1707128075406305,
+      "learning_rate": 3.758858920834538e-06,
+      "loss": 0.6843,
+      "step": 2037
+    },
+    {
+      "epoch": 4.982885085574572,
+      "grad_norm": 0.1677513351740558,
+      "learning_rate": 3.757752708038771e-06,
+      "loss": 0.5329,
+      "step": 2038
+    },
+    {
+      "epoch": 4.985330073349633,
+      "grad_norm": 0.18443435104057523,
+      "learning_rate": 3.7566461654314517e-06,
+      "loss": 0.5066,
+      "step": 2039
+    },
+    {
+      "epoch": 4.987775061124695,
+      "grad_norm": 0.20616232490108505,
+      "learning_rate": 3.7555392933027423e-06,
+      "loss": 0.3955,
+      "step": 2040
+    },
+    {
+      "epoch": 4.990220048899755,
+      "grad_norm": 0.17286930939597117,
+      "learning_rate": 3.754432091942888e-06,
+      "loss": 0.4773,
+      "step": 2041
+    },
+    {
+      "epoch": 4.992665036674817,
+      "grad_norm": 0.19628475350325839,
+      "learning_rate": 3.7533245616422244e-06,
+      "loss": 0.8423,
+      "step": 2042
+    },
+    {
+      "epoch": 4.995110024449878,
+      "grad_norm": 0.2452645530827968,
+      "learning_rate": 3.75221670269117e-06,
+      "loss": 0.7313,
+      "step": 2043
+    },
+    {
+      "epoch": 4.997555012224939,
+      "grad_norm": 0.2615524126775721,
+      "learning_rate": 3.7511085153802314e-06,
+      "loss": 0.3855,
+      "step": 2044
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.2434355251250328,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.7845,
+      "step": 2045
+    },
+    {
+      "epoch": 5.002444987775061,
+      "grad_norm": 0.16109563114864184,
+      "learning_rate": 3.7488911568411553e-06,
+      "loss": 0.7569,
+      "step": 2046
+    },
+    {
+      "epoch": 5.004889975550122,
+      "grad_norm": 0.15735684935296101,
+      "learning_rate": 3.7477819861944597e-06,
+      "loss": 0.3686,
+      "step": 2047
+    },
+    {
+      "epoch": 5.007334963325183,
+      "grad_norm": 0.21194573241183948,
+      "learning_rate": 3.746672488350764e-06,
+      "loss": 0.4379,
+      "step": 2048
+    },
+    {
+      "epoch": 5.009779951100245,
+      "grad_norm": 0.17339994080447857,
+      "learning_rate": 3.7455626636010046e-06,
+      "loss": 0.8235,
+      "step": 2049
+    },
+    {
+      "epoch": 5.012224938875305,
+      "grad_norm": 0.31944516023074276,
+      "learning_rate": 3.7444525122362015e-06,
+      "loss": 0.7996,
+      "step": 2050
+    },
+    {
+      "epoch": 5.014669926650367,
+      "grad_norm": 0.22939504986346795,
+      "learning_rate": 3.7433420345474624e-06,
+      "loss": 0.7048,
+      "step": 2051
+    },
+    {
+      "epoch": 5.017114914425428,
+      "grad_norm": 0.1618363442223097,
+      "learning_rate": 3.74223123082598e-06,
+      "loss": 1.008,
+      "step": 2052
+    },
+    {
+      "epoch": 5.019559902200489,
+      "grad_norm": 0.14721689164750812,
+      "learning_rate": 3.741120101363033e-06,
+      "loss": 0.5197,
+      "step": 2053
+    },
+    {
+      "epoch": 5.02200488997555,
+      "grad_norm": 0.1694875910171043,
+      "learning_rate": 3.740008646449984e-06,
+      "loss": 0.3399,
+      "step": 2054
+    },
+    {
+      "epoch": 5.0244498777506115,
+      "grad_norm": 0.2092040762878438,
+      "learning_rate": 3.7388968663782825e-06,
+      "loss": 0.5061,
+      "step": 2055
+    },
+    {
+      "epoch": 5.026894865525672,
+      "grad_norm": 0.1808907995750573,
+      "learning_rate": 3.737784761439463e-06,
+      "loss": 0.6191,
+      "step": 2056
+    },
+    {
+      "epoch": 5.029339853300733,
+      "grad_norm": 0.2167372295433262,
+      "learning_rate": 3.736672331925144e-06,
+      "loss": 0.903,
+      "step": 2057
+    },
+    {
+      "epoch": 5.031784841075795,
+      "grad_norm": 0.19735001348887968,
+      "learning_rate": 3.7355595781270304e-06,
+      "loss": 0.8945,
+      "step": 2058
+    },
+    {
+      "epoch": 5.034229828850855,
+      "grad_norm": 0.16513299651852423,
+      "learning_rate": 3.734446500336912e-06,
+      "loss": 0.4226,
+      "step": 2059
+    },
+    {
+      "epoch": 5.036674816625917,
+      "grad_norm": 0.20150106643164067,
+      "learning_rate": 3.733333098846663e-06,
+      "loss": 0.6902,
+      "step": 2060
+    },
+    {
+      "epoch": 5.039119804400978,
+      "grad_norm": 0.255453117342702,
+      "learning_rate": 3.732219373948243e-06,
+      "loss": 0.5388,
+      "step": 2061
+    },
+    {
+      "epoch": 5.041564792176039,
+      "grad_norm": 0.21904539621790262,
+      "learning_rate": 3.7311053259336953e-06,
+      "loss": 0.3778,
+      "step": 2062
+    },
+    {
+      "epoch": 5.0440097799511,
+      "grad_norm": 0.2536145338282659,
+      "learning_rate": 3.7299909550951495e-06,
+      "loss": 0.8119,
+      "step": 2063
+    },
+    {
+      "epoch": 5.0464547677261615,
+      "grad_norm": 0.18498787151540202,
+      "learning_rate": 3.72887626172482e-06,
+      "loss": 0.4255,
+      "step": 2064
+    },
+    {
+      "epoch": 5.048899755501223,
+      "grad_norm": 0.18836031374252818,
+      "learning_rate": 3.7277612461150037e-06,
+      "loss": 0.4448,
+      "step": 2065
+    },
+    {
+      "epoch": 5.0513447432762835,
+      "grad_norm": 0.18298151989927058,
+      "learning_rate": 3.7266459085580834e-06,
+      "loss": 0.6381,
+      "step": 2066
+    },
+    {
+      "epoch": 5.053789731051345,
+      "grad_norm": 0.25343862114536697,
+      "learning_rate": 3.7255302493465264e-06,
+      "loss": 0.5411,
+      "step": 2067
+    },
+    {
+      "epoch": 5.056234718826406,
+      "grad_norm": 0.2842673456838669,
+      "learning_rate": 3.7244142687728836e-06,
+      "loss": 0.8084,
+      "step": 2068
+    },
+    {
+      "epoch": 5.058679706601467,
+      "grad_norm": 0.23676455103581842,
+      "learning_rate": 3.7232979671297908e-06,
+      "loss": 1.305,
+      "step": 2069
+    },
+    {
+      "epoch": 5.061124694376528,
+      "grad_norm": 0.1812856147831502,
+      "learning_rate": 3.7221813447099693e-06,
+      "loss": 0.769,
+      "step": 2070
+    },
+    {
+      "epoch": 5.06356968215159,
+      "grad_norm": 0.22498980848527977,
+      "learning_rate": 3.721064401806221e-06,
+      "loss": 0.764,
+      "step": 2071
+    },
+    {
+      "epoch": 5.06601466992665,
+      "grad_norm": 0.22482236418218207,
+      "learning_rate": 3.7199471387114343e-06,
+      "loss": 0.6532,
+      "step": 2072
+    },
+    {
+      "epoch": 5.068459657701712,
+      "grad_norm": 0.18125859229578006,
+      "learning_rate": 3.7188295557185817e-06,
+      "loss": 0.3887,
+      "step": 2073
+    },
+    {
+      "epoch": 5.070904645476773,
+      "grad_norm": 0.23672909655585825,
+      "learning_rate": 3.717711653120719e-06,
+      "loss": 0.465,
+      "step": 2074
+    },
+    {
+      "epoch": 5.0733496332518335,
+      "grad_norm": 0.14252514902839977,
+      "learning_rate": 3.7165934312109863e-06,
+      "loss": 0.7082,
+      "step": 2075
+    },
+    {
+      "epoch": 5.075794621026895,
+      "grad_norm": 0.1619753205830732,
+      "learning_rate": 3.7154748902826066e-06,
+      "loss": 0.9648,
+      "step": 2076
+    },
+    {
+      "epoch": 5.078239608801956,
+      "grad_norm": 0.16355380726896754,
+      "learning_rate": 3.7143560306288863e-06,
+      "loss": 0.4784,
+      "step": 2077
+    },
+    {
+      "epoch": 5.080684596577017,
+      "grad_norm": 0.19496912463798421,
+      "learning_rate": 3.713236852543216e-06,
+      "loss": 0.7877,
+      "step": 2078
+    },
+    {
+      "epoch": 5.083129584352078,
+      "grad_norm": 0.17263534827473018,
+      "learning_rate": 3.7121173563190704e-06,
+      "loss": 0.4282,
+      "step": 2079
+    },
+    {
+      "epoch": 5.08557457212714,
+      "grad_norm": 0.28495006712161103,
+      "learning_rate": 3.7109975422500076e-06,
+      "loss": 0.6145,
+      "step": 2080
+    },
+    {
+      "epoch": 5.0880195599022,
+      "grad_norm": 0.211043385597544,
+      "learning_rate": 3.709877410629668e-06,
+      "loss": 0.7415,
+      "step": 2081
+    },
+    {
+      "epoch": 5.090464547677262,
+      "grad_norm": 0.21493335041067174,
+      "learning_rate": 3.7087569617517748e-06,
+      "loss": 0.602,
+      "step": 2082
+    },
+    {
+      "epoch": 5.092909535452323,
+      "grad_norm": 0.24835588154747737,
+      "learning_rate": 3.707636195910137e-06,
+      "loss": 0.8523,
+      "step": 2083
+    },
+    {
+      "epoch": 5.095354523227384,
+      "grad_norm": 0.1714890789052701,
+      "learning_rate": 3.7065151133986434e-06,
+      "loss": 0.9308,
+      "step": 2084
+    },
+    {
+      "epoch": 5.097799511002445,
+      "grad_norm": 0.16482953190126545,
+      "learning_rate": 3.7053937145112682e-06,
+      "loss": 0.5792,
+      "step": 2085
+    },
+    {
+      "epoch": 5.100244498777506,
+      "grad_norm": 0.24576765578946383,
+      "learning_rate": 3.704271999542069e-06,
+      "loss": 0.6846,
+      "step": 2086
+    },
+    {
+      "epoch": 5.102689486552567,
+      "grad_norm": 0.2284261127763889,
+      "learning_rate": 3.7031499687851828e-06,
+      "loss": 0.5299,
+      "step": 2087
+    },
+    {
+      "epoch": 5.105134474327628,
+      "grad_norm": 0.1697789895836933,
+      "learning_rate": 3.702027622534834e-06,
+      "loss": 0.6674,
+      "step": 2088
+    },
+    {
+      "epoch": 5.10757946210269,
+      "grad_norm": 0.20660629650210033,
+      "learning_rate": 3.700904961085326e-06,
+      "loss": 1.0055,
+      "step": 2089
+    },
+    {
+      "epoch": 5.11002444987775,
+      "grad_norm": 0.23226771914771305,
+      "learning_rate": 3.6997819847310475e-06,
+      "loss": 0.6362,
+      "step": 2090
+    },
+    {
+      "epoch": 5.112469437652812,
+      "grad_norm": 0.18380249217489605,
+      "learning_rate": 3.6986586937664672e-06,
+      "loss": 0.4715,
+      "step": 2091
+    },
+    {
+      "epoch": 5.114914425427873,
+      "grad_norm": 0.12450233399837017,
+      "learning_rate": 3.6975350884861384e-06,
+      "loss": 0.7729,
+      "step": 2092
+    },
+    {
+      "epoch": 5.117359413202934,
+      "grad_norm": 0.21849488853786664,
+      "learning_rate": 3.6964111691846973e-06,
+      "loss": 0.4097,
+      "step": 2093
+    },
+    {
+      "epoch": 5.119804400977995,
+      "grad_norm": 0.18452563369418593,
+      "learning_rate": 3.6952869361568587e-06,
+      "loss": 0.8212,
+      "step": 2094
+    },
+    {
+      "epoch": 5.1222493887530565,
+      "grad_norm": 0.29796345709010247,
+      "learning_rate": 3.6941623896974245e-06,
+      "loss": 0.4151,
+      "step": 2095
+    },
+    {
+      "epoch": 5.124694376528117,
+      "grad_norm": 0.229999026641762,
+      "learning_rate": 3.6930375301012755e-06,
+      "loss": 0.6615,
+      "step": 2096
+    },
+    {
+      "epoch": 5.127139364303178,
+      "grad_norm": 0.21413008497904962,
+      "learning_rate": 3.691912357663376e-06,
+      "loss": 0.6919,
+      "step": 2097
+    },
+    {
+      "epoch": 5.12958435207824,
+      "grad_norm": 0.15737818578336213,
+      "learning_rate": 3.6907868726787717e-06,
+      "loss": 0.4261,
+      "step": 2098
+    },
+    {
+      "epoch": 5.1320293398533,
+      "grad_norm": 0.28068837012746667,
+      "learning_rate": 3.689661075442591e-06,
+      "loss": 0.5471,
+      "step": 2099
+    },
+    {
+      "epoch": 5.134474327628362,
+      "grad_norm": 0.33290811587452424,
+      "learning_rate": 3.6885349662500423e-06,
+      "loss": 0.5975,
+      "step": 2100
+    },
+    {
+      "epoch": 5.136919315403423,
+      "grad_norm": 0.1786004918209996,
+      "learning_rate": 3.687408545396419e-06,
+      "loss": 0.7505,
+      "step": 2101
+    },
+    {
+      "epoch": 5.139364303178484,
+      "grad_norm": 0.20152443049657645,
+      "learning_rate": 3.686281813177094e-06,
+      "loss": 0.7161,
+      "step": 2102
+    },
+    {
+      "epoch": 5.141809290953545,
+      "grad_norm": 0.259211002387912,
+      "learning_rate": 3.6851547698875216e-06,
+      "loss": 0.5705,
+      "step": 2103
+    },
+    {
+      "epoch": 5.1442542787286065,
+      "grad_norm": 0.26085245813654273,
+      "learning_rate": 3.684027415823239e-06,
+      "loss": 0.4421,
+      "step": 2104
+    },
+    {
+      "epoch": 5.146699266503667,
+      "grad_norm": 0.22527074714918777,
+      "learning_rate": 3.6828997512798635e-06,
+      "loss": 0.5591,
+      "step": 2105
+    },
+    {
+      "epoch": 5.1491442542787285,
+      "grad_norm": 0.1854647928967856,
+      "learning_rate": 3.6817717765530953e-06,
+      "loss": 1.0413,
+      "step": 2106
+    },
+    {
+      "epoch": 5.15158924205379,
+      "grad_norm": 0.2058749798127868,
+      "learning_rate": 3.680643491938715e-06,
+      "loss": 0.5927,
+      "step": 2107
+    },
+    {
+      "epoch": 5.15403422982885,
+      "grad_norm": 0.11999034936399257,
+      "learning_rate": 3.6795148977325846e-06,
+      "loss": 0.3803,
+      "step": 2108
+    },
+    {
+      "epoch": 5.156479217603912,
+      "grad_norm": 0.21600662594533374,
+      "learning_rate": 3.678385994230647e-06,
+      "loss": 0.4288,
+      "step": 2109
+    },
+    {
+      "epoch": 5.158924205378973,
+      "grad_norm": 0.25378512829289,
+      "learning_rate": 3.6772567817289275e-06,
+      "loss": 0.5874,
+      "step": 2110
+    },
+    {
+      "epoch": 5.161369193154035,
+      "grad_norm": 0.24987066388456966,
+      "learning_rate": 3.67612726052353e-06,
+      "loss": 0.4971,
+      "step": 2111
+    },
+    {
+      "epoch": 5.163814180929095,
+      "grad_norm": 0.14943094983788496,
+      "learning_rate": 3.674997430910642e-06,
+      "loss": 0.5019,
+      "step": 2112
+    },
+    {
+      "epoch": 5.166259168704157,
+      "grad_norm": 0.1704947679311722,
+      "learning_rate": 3.6738672931865297e-06,
+      "loss": 0.9122,
+      "step": 2113
+    },
+    {
+      "epoch": 5.168704156479218,
+      "grad_norm": 0.24898697634952338,
+      "learning_rate": 3.6727368476475427e-06,
+      "loss": 0.6099,
+      "step": 2114
+    },
+    {
+      "epoch": 5.1711491442542785,
+      "grad_norm": 0.2667313247007851,
+      "learning_rate": 3.6716060945901084e-06,
+      "loss": 0.9603,
+      "step": 2115
+    },
+    {
+      "epoch": 5.17359413202934,
+      "grad_norm": 0.18768121598491,
+      "learning_rate": 3.670475034310736e-06,
+      "loss": 0.5143,
+      "step": 2116
+    },
+    {
+      "epoch": 5.176039119804401,
+      "grad_norm": 0.15997224837678883,
+      "learning_rate": 3.6693436671060158e-06,
+      "loss": 0.6646,
+      "step": 2117
+    },
+    {
+      "epoch": 5.178484107579462,
+      "grad_norm": 0.2172855730645409,
+      "learning_rate": 3.668211993272619e-06,
+      "loss": 0.6451,
+      "step": 2118
+    },
+    {
+      "epoch": 5.180929095354523,
+      "grad_norm": 0.1901992657389288,
+      "learning_rate": 3.6670800131072948e-06,
+      "loss": 0.4664,
+      "step": 2119
+    },
+    {
+      "epoch": 5.183374083129585,
+      "grad_norm": 0.19258303832498572,
+      "learning_rate": 3.6659477269068754e-06,
+      "loss": 0.5509,
+      "step": 2120
+    },
+    {
+      "epoch": 5.185819070904645,
+      "grad_norm": 0.2276976577830496,
+      "learning_rate": 3.664815134968272e-06,
+      "loss": 0.4729,
+      "step": 2121
+    },
+    {
+      "epoch": 5.188264058679707,
+      "grad_norm": 0.1614653089989081,
+      "learning_rate": 3.6636822375884757e-06,
+      "loss": 0.5779,
+      "step": 2122
+    },
+    {
+      "epoch": 5.190709046454768,
+      "grad_norm": 0.4340789950657901,
+      "learning_rate": 3.662549035064559e-06,
+      "loss": 0.4417,
+      "step": 2123
+    },
+    {
+      "epoch": 5.193154034229829,
+      "grad_norm": 0.22815455022546718,
+      "learning_rate": 3.6614155276936724e-06,
+      "loss": 0.3982,
+      "step": 2124
+    },
+    {
+      "epoch": 5.19559902200489,
+      "grad_norm": 0.1579487667006239,
+      "learning_rate": 3.6602817157730486e-06,
+      "loss": 0.4922,
+      "step": 2125
+    },
+    {
+      "epoch": 5.198044009779951,
+      "grad_norm": 0.2209374313725165,
+      "learning_rate": 3.659147599599998e-06,
+      "loss": 0.4823,
+      "step": 2126
+    },
+    {
+      "epoch": 5.200488997555012,
+      "grad_norm": 0.19335759246157247,
+      "learning_rate": 3.658013179471912e-06,
+      "loss": 0.4809,
+      "step": 2127
+    },
+    {
+      "epoch": 5.202933985330073,
+      "grad_norm": 0.2272717050360955,
+      "learning_rate": 3.6568784556862622e-06,
+      "loss": 0.8961,
+      "step": 2128
+    },
+    {
+      "epoch": 5.205378973105135,
+      "grad_norm": 0.23036034916843798,
+      "learning_rate": 3.655743428540599e-06,
+      "loss": 0.573,
+      "step": 2129
+    },
+    {
+      "epoch": 5.207823960880195,
+      "grad_norm": 0.2362346540137818,
+      "learning_rate": 3.654608098332553e-06,
+      "loss": 0.4361,
+      "step": 2130
+    },
+    {
+      "epoch": 5.210268948655257,
+      "grad_norm": 0.21135474452709654,
+      "learning_rate": 3.6534724653598312e-06,
+      "loss": 0.4775,
+      "step": 2131
+    },
+    {
+      "epoch": 5.212713936430318,
+      "grad_norm": 0.23048788040404924,
+      "learning_rate": 3.6523365299202255e-06,
+      "loss": 0.603,
+      "step": 2132
+    },
+    {
+      "epoch": 5.215158924205379,
+      "grad_norm": 0.29023952096835604,
+      "learning_rate": 3.651200292311603e-06,
+      "loss": 0.716,
+      "step": 2133
+    },
+    {
+      "epoch": 5.21760391198044,
+      "grad_norm": 0.22837550170067364,
+      "learning_rate": 3.6500637528319115e-06,
+      "loss": 0.4567,
+      "step": 2134
+    },
+    {
+      "epoch": 5.2200488997555015,
+      "grad_norm": 0.22903019299457172,
+      "learning_rate": 3.6489269117791782e-06,
+      "loss": 1.0954,
+      "step": 2135
+    },
+    {
+      "epoch": 5.222493887530562,
+      "grad_norm": 0.1909403867781725,
+      "learning_rate": 3.6477897694515063e-06,
+      "loss": 0.659,
+      "step": 2136
+    },
+    {
+      "epoch": 5.224938875305623,
+      "grad_norm": 0.21263450208630374,
+      "learning_rate": 3.6466523261470837e-06,
+      "loss": 0.7687,
+      "step": 2137
+    },
+    {
+      "epoch": 5.227383863080685,
+      "grad_norm": 0.21628289821966418,
+      "learning_rate": 3.6455145821641726e-06,
+      "loss": 0.5646,
+      "step": 2138
+    },
+    {
+      "epoch": 5.229828850855745,
+      "grad_norm": 0.21116576678531623,
+      "learning_rate": 3.6443765378011154e-06,
+      "loss": 0.3851,
+      "step": 2139
+    },
+    {
+      "epoch": 5.232273838630807,
+      "grad_norm": 0.17493305020936142,
+      "learning_rate": 3.6432381933563337e-06,
+      "loss": 0.5168,
+      "step": 2140
+    },
+    {
+      "epoch": 5.234718826405868,
+      "grad_norm": 0.27078908826322523,
+      "learning_rate": 3.6420995491283274e-06,
+      "loss": 1.2432,
+      "step": 2141
+    },
+    {
+      "epoch": 5.237163814180929,
+      "grad_norm": 0.20666443048296643,
+      "learning_rate": 3.6409606054156747e-06,
+      "loss": 0.6603,
+      "step": 2142
+    },
+    {
+      "epoch": 5.23960880195599,
+      "grad_norm": 0.1859391395608223,
+      "learning_rate": 3.639821362517033e-06,
+      "loss": 0.6153,
+      "step": 2143
+    },
+    {
+      "epoch": 5.2420537897310515,
+      "grad_norm": 0.16753322963300563,
+      "learning_rate": 3.638681820731139e-06,
+      "loss": 0.5872,
+      "step": 2144
+    },
+    {
+      "epoch": 5.244498777506112,
+      "grad_norm": 0.2258477684421268,
+      "learning_rate": 3.637541980356805e-06,
+      "loss": 0.5507,
+      "step": 2145
+    },
+    {
+      "epoch": 5.2469437652811735,
+      "grad_norm": 0.1784607257063599,
+      "learning_rate": 3.6364018416929236e-06,
+      "loss": 0.5821,
+      "step": 2146
+    },
+    {
+      "epoch": 5.249388753056235,
+      "grad_norm": 0.15719212308487598,
+      "learning_rate": 3.6352614050384665e-06,
+      "loss": 0.438,
+      "step": 2147
+    },
+    {
+      "epoch": 5.251833740831296,
+      "grad_norm": 0.31697331904796755,
+      "learning_rate": 3.6341206706924802e-06,
+      "loss": 1.1717,
+      "step": 2148
+    },
+    {
+      "epoch": 5.254278728606357,
+      "grad_norm": 0.22757953455097363,
+      "learning_rate": 3.6329796389540935e-06,
+      "loss": 0.6083,
+      "step": 2149
+    },
+    {
+      "epoch": 5.256723716381418,
+      "grad_norm": 0.25453999116716636,
+      "learning_rate": 3.6318383101225097e-06,
+      "loss": 0.6664,
+      "step": 2150
+    },
+    {
+      "epoch": 5.25916870415648,
+      "grad_norm": 0.27613469474801966,
+      "learning_rate": 3.630696684497011e-06,
+      "loss": 0.8694,
+      "step": 2151
+    },
+    {
+      "epoch": 5.26161369193154,
+      "grad_norm": 0.21425127867120816,
+      "learning_rate": 3.629554762376959e-06,
+      "loss": 0.8508,
+      "step": 2152
+    },
+    {
+      "epoch": 5.264058679706602,
+      "grad_norm": 0.15046011626706285,
+      "learning_rate": 3.628412544061791e-06,
+      "loss": 0.8085,
+      "step": 2153
+    },
+    {
+      "epoch": 5.266503667481663,
+      "grad_norm": 0.17180039915197945,
+      "learning_rate": 3.6272700298510233e-06,
+      "loss": 0.8843,
+      "step": 2154
+    },
+    {
+      "epoch": 5.2689486552567235,
+      "grad_norm": 0.18419989808008455,
+      "learning_rate": 3.626127220044249e-06,
+      "loss": 0.5409,
+      "step": 2155
+    },
+    {
+      "epoch": 5.271393643031785,
+      "grad_norm": 0.29349251179461483,
+      "learning_rate": 3.6249841149411385e-06,
+      "loss": 0.4038,
+      "step": 2156
+    },
+    {
+      "epoch": 5.273838630806846,
+      "grad_norm": 0.19601519037995702,
+      "learning_rate": 3.6238407148414413e-06,
+      "loss": 0.4002,
+      "step": 2157
+    },
+    {
+      "epoch": 5.276283618581907,
+      "grad_norm": 0.21844047622123686,
+      "learning_rate": 3.6226970200449824e-06,
+      "loss": 0.5217,
+      "step": 2158
+    },
+    {
+      "epoch": 5.278728606356968,
+      "grad_norm": 0.22321897660964074,
+      "learning_rate": 3.621553030851664e-06,
+      "loss": 0.6151,
+      "step": 2159
+    },
+    {
+      "epoch": 5.28117359413203,
+      "grad_norm": 0.20922811172847908,
+      "learning_rate": 3.620408747561468e-06,
+      "loss": 0.3649,
+      "step": 2160
+    },
+    {
+      "epoch": 5.28361858190709,
+      "grad_norm": 0.23511713245526447,
+      "learning_rate": 3.6192641704744495e-06,
+      "loss": 0.569,
+      "step": 2161
+    },
+    {
+      "epoch": 5.286063569682152,
+      "grad_norm": 0.21968237424631146,
+      "learning_rate": 3.6181192998907453e-06,
+      "loss": 0.3507,
+      "step": 2162
+    },
+    {
+      "epoch": 5.288508557457213,
+      "grad_norm": 0.18599691794295706,
+      "learning_rate": 3.616974136110565e-06,
+      "loss": 0.6965,
+      "step": 2163
+    },
+    {
+      "epoch": 5.290953545232274,
+      "grad_norm": 0.25258341891239905,
+      "learning_rate": 3.615828679434197e-06,
+      "loss": 0.6335,
+      "step": 2164
+    },
+    {
+      "epoch": 5.293398533007335,
+      "grad_norm": 0.1727851775207428,
+      "learning_rate": 3.614682930162006e-06,
+      "loss": 0.4426,
+      "step": 2165
+    },
+    {
+      "epoch": 5.295843520782396,
+      "grad_norm": 0.21024083727583648,
+      "learning_rate": 3.6135368885944344e-06,
+      "loss": 0.5258,
+      "step": 2166
+    },
+    {
+      "epoch": 5.298288508557457,
+      "grad_norm": 0.14742386508035366,
+      "learning_rate": 3.6123905550320006e-06,
+      "loss": 0.4863,
+      "step": 2167
+    },
+    {
+      "epoch": 5.300733496332518,
+      "grad_norm": 0.1802107637301803,
+      "learning_rate": 3.611243929775299e-06,
+      "loss": 0.6268,
+      "step": 2168
+    },
+    {
+      "epoch": 5.30317848410758,
+      "grad_norm": 0.2764852618844023,
+      "learning_rate": 3.6100970131250013e-06,
+      "loss": 0.6348,
+      "step": 2169
+    },
+    {
+      "epoch": 5.30562347188264,
+      "grad_norm": 0.2430505816519341,
+      "learning_rate": 3.6089498053818543e-06,
+      "loss": 0.4831,
+      "step": 2170
+    },
+    {
+      "epoch": 5.308068459657702,
+      "grad_norm": 0.232395854986584,
+      "learning_rate": 3.6078023068466832e-06,
+      "loss": 0.4745,
+      "step": 2171
+    },
+    {
+      "epoch": 5.310513447432763,
+      "grad_norm": 0.20798738703477299,
+      "learning_rate": 3.606654517820389e-06,
+      "loss": 0.4821,
+      "step": 2172
+    },
+    {
+      "epoch": 5.312958435207824,
+      "grad_norm": 0.30277682647660076,
+      "learning_rate": 3.6055064386039467e-06,
+      "loss": 0.3754,
+      "step": 2173
+    },
+    {
+      "epoch": 5.315403422982885,
+      "grad_norm": 0.19687269394567342,
+      "learning_rate": 3.60435806949841e-06,
+      "loss": 0.4324,
+      "step": 2174
+    },
+    {
+      "epoch": 5.3178484107579465,
+      "grad_norm": 0.24015875000730597,
+      "learning_rate": 3.6032094108049063e-06,
+      "loss": 0.4981,
+      "step": 2175
+    },
+    {
+      "epoch": 5.320293398533007,
+      "grad_norm": 0.18127294191034132,
+      "learning_rate": 3.6020604628246418e-06,
+      "loss": 0.5015,
+      "step": 2176
+    },
+    {
+      "epoch": 5.322738386308068,
+      "grad_norm": 0.28965796298554247,
+      "learning_rate": 3.600911225858896e-06,
+      "loss": 0.449,
+      "step": 2177
+    },
+    {
+      "epoch": 5.32518337408313,
+      "grad_norm": 0.2591795274334654,
+      "learning_rate": 3.5997617002090264e-06,
+      "loss": 0.6724,
+      "step": 2178
+    },
+    {
+      "epoch": 5.32762836185819,
+      "grad_norm": 0.22570884879404238,
+      "learning_rate": 3.598611886176463e-06,
+      "loss": 0.4894,
+      "step": 2179
+    },
+    {
+      "epoch": 5.330073349633252,
+      "grad_norm": 0.13608595120491174,
+      "learning_rate": 3.5974617840627145e-06,
+      "loss": 0.382,
+      "step": 2180
+    },
+    {
+      "epoch": 5.332518337408313,
+      "grad_norm": 0.21459449247210435,
+      "learning_rate": 3.596311394169363e-06,
+      "loss": 0.5245,
+      "step": 2181
+    },
+    {
+      "epoch": 5.334963325183374,
+      "grad_norm": 0.1630110178679035,
+      "learning_rate": 3.5951607167980695e-06,
+      "loss": 0.4677,
+      "step": 2182
+    },
+    {
+      "epoch": 5.337408312958435,
+      "grad_norm": 0.20975214105427356,
+      "learning_rate": 3.5940097522505655e-06,
+      "loss": 0.631,
+      "step": 2183
+    },
+    {
+      "epoch": 5.3398533007334965,
+      "grad_norm": 0.19356520671954475,
+      "learning_rate": 3.5928585008286614e-06,
+      "loss": 0.9404,
+      "step": 2184
+    },
+    {
+      "epoch": 5.342298288508557,
+      "grad_norm": 0.21928050038486538,
+      "learning_rate": 3.591706962834241e-06,
+      "loss": 0.6218,
+      "step": 2185
+    },
+    {
+      "epoch": 5.3447432762836184,
+      "grad_norm": 0.1862271096654394,
+      "learning_rate": 3.590555138569264e-06,
+      "loss": 0.3539,
+      "step": 2186
+    },
+    {
+      "epoch": 5.34718826405868,
+      "grad_norm": 0.18313391349123137,
+      "learning_rate": 3.5894030283357663e-06,
+      "loss": 0.3692,
+      "step": 2187
+    },
+    {
+      "epoch": 5.34963325183374,
+      "grad_norm": 0.20052656395762344,
+      "learning_rate": 3.5882506324358557e-06,
+      "loss": 0.4359,
+      "step": 2188
+    },
+    {
+      "epoch": 5.352078239608802,
+      "grad_norm": 0.2085014969718404,
+      "learning_rate": 3.5870979511717184e-06,
+      "loss": 0.6994,
+      "step": 2189
+    },
+    {
+      "epoch": 5.354523227383863,
+      "grad_norm": 0.2930754367421236,
+      "learning_rate": 3.585944984845613e-06,
+      "loss": 0.6871,
+      "step": 2190
+    },
+    {
+      "epoch": 5.356968215158924,
+      "grad_norm": 0.29896306658933597,
+      "learning_rate": 3.584791733759873e-06,
+      "loss": 0.4254,
+      "step": 2191
+    },
+    {
+      "epoch": 5.359413202933985,
+      "grad_norm": 0.18337083746557345,
+      "learning_rate": 3.5836381982169088e-06,
+      "loss": 0.4719,
+      "step": 2192
+    },
+    {
+      "epoch": 5.361858190709047,
+      "grad_norm": 0.20541684678338798,
+      "learning_rate": 3.582484378519203e-06,
+      "loss": 0.3929,
+      "step": 2193
+    },
+    {
+      "epoch": 5.364303178484108,
+      "grad_norm": 0.20922361278730478,
+      "learning_rate": 3.5813302749693143e-06,
+      "loss": 0.7873,
+      "step": 2194
+    },
+    {
+      "epoch": 5.3667481662591685,
+      "grad_norm": 0.17489750910655175,
+      "learning_rate": 3.5801758878698735e-06,
+      "loss": 0.5274,
+      "step": 2195
+    },
+    {
+      "epoch": 5.36919315403423,
+      "grad_norm": 0.13452382413424466,
+      "learning_rate": 3.5790212175235885e-06,
+      "loss": 0.4157,
+      "step": 2196
+    },
+    {
+      "epoch": 5.371638141809291,
+      "grad_norm": 0.14172307205017176,
+      "learning_rate": 3.5778662642332395e-06,
+      "loss": 0.4088,
+      "step": 2197
+    },
+    {
+      "epoch": 5.374083129584352,
+      "grad_norm": 0.33251437361504943,
+      "learning_rate": 3.5767110283016824e-06,
+      "loss": 0.4245,
+      "step": 2198
+    },
+    {
+      "epoch": 5.376528117359413,
+      "grad_norm": 0.2661282517138751,
+      "learning_rate": 3.5755555100318466e-06,
+      "loss": 0.4009,
+      "step": 2199
+    },
+    {
+      "epoch": 5.378973105134475,
+      "grad_norm": 0.32569982209414106,
+      "learning_rate": 3.5743997097267345e-06,
+      "loss": 0.8575,
+      "step": 2200
+    },
+    {
+      "epoch": 5.381418092909535,
+      "grad_norm": 0.2802546545487847,
+      "learning_rate": 3.573243627689425e-06,
+      "loss": 0.7675,
+      "step": 2201
+    },
+    {
+      "epoch": 5.383863080684597,
+      "grad_norm": 0.2809570042766992,
+      "learning_rate": 3.5720872642230674e-06,
+      "loss": 0.6308,
+      "step": 2202
+    },
+    {
+      "epoch": 5.386308068459658,
+      "grad_norm": 0.13076090260361758,
+      "learning_rate": 3.5709306196308885e-06,
+      "loss": 0.4768,
+      "step": 2203
+    },
+    {
+      "epoch": 5.388753056234719,
+      "grad_norm": 0.21494050203518575,
+      "learning_rate": 3.569773694216186e-06,
+      "loss": 0.4779,
+      "step": 2204
+    },
+    {
+      "epoch": 5.39119804400978,
+      "grad_norm": 0.18163821000288635,
+      "learning_rate": 3.5686164882823314e-06,
+      "loss": 0.6205,
+      "step": 2205
+    },
+    {
+      "epoch": 5.393643031784841,
+      "grad_norm": 0.290270247174451,
+      "learning_rate": 3.5674590021327723e-06,
+      "loss": 0.4559,
+      "step": 2206
+    },
+    {
+      "epoch": 5.396088019559902,
+      "grad_norm": 0.2746632414942217,
+      "learning_rate": 3.5663012360710266e-06,
+      "loss": 0.4281,
+      "step": 2207
+    },
+    {
+      "epoch": 5.398533007334963,
+      "grad_norm": 0.3135986818857818,
+      "learning_rate": 3.565143190400688e-06,
+      "loss": 0.4237,
+      "step": 2208
+    },
+    {
+      "epoch": 5.400977995110025,
+      "grad_norm": 0.22444419257332893,
+      "learning_rate": 3.563984865425422e-06,
+      "loss": 0.7947,
+      "step": 2209
+    },
+    {
+      "epoch": 5.403422982885085,
+      "grad_norm": 0.2714304106414512,
+      "learning_rate": 3.5628262614489684e-06,
+      "loss": 1.6441,
+      "step": 2210
+    },
+    {
+      "epoch": 5.405867970660147,
+      "grad_norm": 0.17090607325258253,
+      "learning_rate": 3.56166737877514e-06,
+      "loss": 0.382,
+      "step": 2211
+    },
+    {
+      "epoch": 5.408312958435208,
+      "grad_norm": 0.2359439768148487,
+      "learning_rate": 3.5605082177078215e-06,
+      "loss": 0.5672,
+      "step": 2212
+    },
+    {
+      "epoch": 5.410757946210269,
+      "grad_norm": 0.2053767338420804,
+      "learning_rate": 3.5593487785509717e-06,
+      "loss": 0.493,
+      "step": 2213
+    },
+    {
+      "epoch": 5.41320293398533,
+      "grad_norm": 0.3124420531563946,
+      "learning_rate": 3.5581890616086225e-06,
+      "loss": 0.4293,
+      "step": 2214
+    },
+    {
+      "epoch": 5.4156479217603914,
+      "grad_norm": 0.1826239169923515,
+      "learning_rate": 3.5570290671848773e-06,
+      "loss": 0.7154,
+      "step": 2215
+    },
+    {
+      "epoch": 5.418092909535452,
+      "grad_norm": 0.27511127399033364,
+      "learning_rate": 3.5558687955839146e-06,
+      "loss": 0.5191,
+      "step": 2216
+    },
+    {
+      "epoch": 5.420537897310513,
+      "grad_norm": 0.2598960192566311,
+      "learning_rate": 3.554708247109984e-06,
+      "loss": 0.648,
+      "step": 2217
+    },
+    {
+      "epoch": 5.422982885085575,
+      "grad_norm": 0.24429128901864908,
+      "learning_rate": 3.5535474220674054e-06,
+      "loss": 0.3936,
+      "step": 2218
+    },
+    {
+      "epoch": 5.425427872860635,
+      "grad_norm": 0.22226652699452956,
+      "learning_rate": 3.5523863207605767e-06,
+      "loss": 0.6975,
+      "step": 2219
+    },
+    {
+      "epoch": 5.427872860635697,
+      "grad_norm": 0.20384416918918755,
+      "learning_rate": 3.5512249434939637e-06,
+      "loss": 0.9031,
+      "step": 2220
+    },
+    {
+      "epoch": 5.430317848410758,
+      "grad_norm": 0.2581983614004652,
+      "learning_rate": 3.5500632905721074e-06,
+      "loss": 1.2965,
+      "step": 2221
+    },
+    {
+      "epoch": 5.432762836185819,
+      "grad_norm": 0.15815602385438232,
+      "learning_rate": 3.548901362299618e-06,
+      "loss": 0.4571,
+      "step": 2222
+    },
+    {
+      "epoch": 5.43520782396088,
+      "grad_norm": 0.23083962056025792,
+      "learning_rate": 3.5477391589811817e-06,
+      "loss": 1.0567,
+      "step": 2223
+    },
+    {
+      "epoch": 5.4376528117359415,
+      "grad_norm": 0.1943489918572033,
+      "learning_rate": 3.546576680921553e-06,
+      "loss": 0.4071,
+      "step": 2224
+    },
+    {
+      "epoch": 5.440097799511002,
+      "grad_norm": 0.19311107817517553,
+      "learning_rate": 3.5454139284255605e-06,
+      "loss": 0.5258,
+      "step": 2225
+    },
+    {
+      "epoch": 5.442542787286063,
+      "grad_norm": 0.24669553035443006,
+      "learning_rate": 3.544250901798106e-06,
+      "loss": 0.6146,
+      "step": 2226
+    },
+    {
+      "epoch": 5.444987775061125,
+      "grad_norm": 0.2187002169037261,
+      "learning_rate": 3.5430876013441607e-06,
+      "loss": 1.167,
+      "step": 2227
+    },
+    {
+      "epoch": 5.447432762836186,
+      "grad_norm": 0.24262848916232485,
+      "learning_rate": 3.5419240273687684e-06,
+      "loss": 1.0638,
+      "step": 2228
+    },
+    {
+      "epoch": 5.449877750611247,
+      "grad_norm": 0.275059167435587,
+      "learning_rate": 3.5407601801770443e-06,
+      "loss": 0.7149,
+      "step": 2229
+    },
+    {
+      "epoch": 5.452322738386308,
+      "grad_norm": 0.23862185311846906,
+      "learning_rate": 3.539596060074177e-06,
+      "loss": 1.1739,
+      "step": 2230
+    },
+    {
+      "epoch": 5.45476772616137,
+      "grad_norm": 0.30645879840030904,
+      "learning_rate": 3.538431667365425e-06,
+      "loss": 0.5851,
+      "step": 2231
+    },
+    {
+      "epoch": 5.45721271393643,
+      "grad_norm": 0.16890222894354093,
+      "learning_rate": 3.537267002356119e-06,
+      "loss": 0.5203,
+      "step": 2232
+    },
+    {
+      "epoch": 5.459657701711492,
+      "grad_norm": 0.17209024615549923,
+      "learning_rate": 3.536102065351661e-06,
+      "loss": 0.8368,
+      "step": 2233
+    },
+    {
+      "epoch": 5.462102689486553,
+      "grad_norm": 0.18832669033785032,
+      "learning_rate": 3.5349368566575226e-06,
+      "loss": 0.4282,
+      "step": 2234
+    },
+    {
+      "epoch": 5.4645476772616135,
+      "grad_norm": 0.23389315248914386,
+      "learning_rate": 3.5337713765792493e-06,
+      "loss": 0.5663,
+      "step": 2235
+    },
+    {
+      "epoch": 5.466992665036675,
+      "grad_norm": 0.21298358173669918,
+      "learning_rate": 3.532605625422457e-06,
+      "loss": 0.652,
+      "step": 2236
+    },
+    {
+      "epoch": 5.469437652811736,
+      "grad_norm": 0.22462721438526495,
+      "learning_rate": 3.531439603492832e-06,
+      "loss": 0.6973,
+      "step": 2237
+    },
+    {
+      "epoch": 5.471882640586797,
+      "grad_norm": 0.20557470482948537,
+      "learning_rate": 3.5302733110961316e-06,
+      "loss": 0.4573,
+      "step": 2238
+    },
+    {
+      "epoch": 5.474327628361858,
+      "grad_norm": 0.23949063014426908,
+      "learning_rate": 3.5291067485381854e-06,
+      "loss": 0.413,
+      "step": 2239
+    },
+    {
+      "epoch": 5.47677261613692,
+      "grad_norm": 0.22312051049735845,
+      "learning_rate": 3.5279399161248904e-06,
+      "loss": 0.3989,
+      "step": 2240
+    },
+    {
+      "epoch": 5.47921760391198,
+      "grad_norm": 0.16551310206286843,
+      "learning_rate": 3.5267728141622205e-06,
+      "loss": 0.7857,
+      "step": 2241
+    },
+    {
+      "epoch": 5.481662591687042,
+      "grad_norm": 0.19896641660693515,
+      "learning_rate": 3.5256054429562136e-06,
+      "loss": 0.7117,
+      "step": 2242
+    },
+    {
+      "epoch": 5.484107579462103,
+      "grad_norm": 0.2758334907879952,
+      "learning_rate": 3.524437802812982e-06,
+      "loss": 0.3938,
+      "step": 2243
+    },
+    {
+      "epoch": 5.486552567237164,
+      "grad_norm": 0.20712906497050476,
+      "learning_rate": 3.5232698940387078e-06,
+      "loss": 0.4382,
+      "step": 2244
+    },
+    {
+      "epoch": 5.488997555012225,
+      "grad_norm": 0.19328280667723455,
+      "learning_rate": 3.5221017169396433e-06,
+      "loss": 0.3808,
+      "step": 2245
+    },
+    {
+      "epoch": 5.491442542787286,
+      "grad_norm": 0.25045253035904314,
+      "learning_rate": 3.5209332718221112e-06,
+      "loss": 0.8375,
+      "step": 2246
+    },
+    {
+      "epoch": 5.493887530562347,
+      "grad_norm": 0.22663208500839396,
+      "learning_rate": 3.519764558992505e-06,
+      "loss": 0.7346,
+      "step": 2247
+    },
+    {
+      "epoch": 5.496332518337408,
+      "grad_norm": 0.18821415071283903,
+      "learning_rate": 3.518595578757288e-06,
+      "loss": 0.3688,
+      "step": 2248
+    },
+    {
+      "epoch": 5.49877750611247,
+      "grad_norm": 0.2773602702366634,
+      "learning_rate": 3.5174263314229916e-06,
+      "loss": 0.6316,
+      "step": 2249
+    },
+    {
+      "epoch": 5.50122249388753,
+      "grad_norm": 0.1532534544817341,
+      "learning_rate": 3.516256817296222e-06,
+      "loss": 0.3511,
+      "step": 2250
+    },
+    {
+      "epoch": 5.503667481662592,
+      "grad_norm": 0.17489945834887297,
+      "learning_rate": 3.5150870366836513e-06,
+      "loss": 0.7129,
+      "step": 2251
+    },
+    {
+      "epoch": 5.506112469437653,
+      "grad_norm": 0.21826621247297684,
+      "learning_rate": 3.513916989892023e-06,
+      "loss": 0.4151,
+      "step": 2252
+    },
+    {
+      "epoch": 5.508557457212714,
+      "grad_norm": 0.21626687541904638,
+      "learning_rate": 3.51274667722815e-06,
+      "loss": 0.6042,
+      "step": 2253
+    },
+    {
+      "epoch": 5.511002444987775,
+      "grad_norm": 0.21136335403517106,
+      "learning_rate": 3.5115760989989143e-06,
+      "loss": 0.5028,
+      "step": 2254
+    },
+    {
+      "epoch": 5.513447432762836,
+      "grad_norm": 0.22664864938328647,
+      "learning_rate": 3.51040525551127e-06,
+      "loss": 0.5551,
+      "step": 2255
+    },
+    {
+      "epoch": 5.515892420537897,
+      "grad_norm": 0.2856589393689792,
+      "learning_rate": 3.509234147072238e-06,
+      "loss": 0.4004,
+      "step": 2256
+    },
+    {
+      "epoch": 5.518337408312958,
+      "grad_norm": 0.16216507863084234,
+      "learning_rate": 3.5080627739889096e-06,
+      "loss": 0.3514,
+      "step": 2257
+    },
+    {
+      "epoch": 5.52078239608802,
+      "grad_norm": 0.18189162316301738,
+      "learning_rate": 3.5068911365684462e-06,
+      "loss": 0.4052,
+      "step": 2258
+    },
+    {
+      "epoch": 5.52322738386308,
+      "grad_norm": 0.22215664723077658,
+      "learning_rate": 3.505719235118077e-06,
+      "loss": 0.6929,
+      "step": 2259
+    },
+    {
+      "epoch": 5.525672371638142,
+      "grad_norm": 0.2012573269615039,
+      "learning_rate": 3.504547069945103e-06,
+      "loss": 0.6347,
+      "step": 2260
+    },
+    {
+      "epoch": 5.528117359413203,
+      "grad_norm": 0.16888406794469107,
+      "learning_rate": 3.503374641356891e-06,
+      "loss": 0.6199,
+      "step": 2261
+    },
+    {
+      "epoch": 5.530562347188264,
+      "grad_norm": 0.22464729008898413,
+      "learning_rate": 3.50220194966088e-06,
+      "loss": 0.6904,
+      "step": 2262
+    },
+    {
+      "epoch": 5.533007334963325,
+      "grad_norm": 0.19609021693176726,
+      "learning_rate": 3.501028995164576e-06,
+      "loss": 0.4726,
+      "step": 2263
+    },
+    {
+      "epoch": 5.5354523227383865,
+      "grad_norm": 0.2156451675238026,
+      "learning_rate": 3.499855778175554e-06,
+      "loss": 0.339,
+      "step": 2264
+    },
+    {
+      "epoch": 5.537897310513447,
+      "grad_norm": 0.2347227027702509,
+      "learning_rate": 3.4986822990014598e-06,
+      "loss": 0.692,
+      "step": 2265
+    },
+    {
+      "epoch": 5.540342298288508,
+      "grad_norm": 0.1787402255681198,
+      "learning_rate": 3.4975085579500046e-06,
+      "loss": 0.4092,
+      "step": 2266
+    },
+    {
+      "epoch": 5.54278728606357,
+      "grad_norm": 0.21335863348260645,
+      "learning_rate": 3.496334555328972e-06,
+      "loss": 0.4477,
+      "step": 2267
+    },
+    {
+      "epoch": 5.54523227383863,
+      "grad_norm": 0.33863929079874755,
+      "learning_rate": 3.495160291446211e-06,
+      "loss": 0.4267,
+      "step": 2268
+    },
+    {
+      "epoch": 5.547677261613692,
+      "grad_norm": 0.19779648068455502,
+      "learning_rate": 3.493985766609641e-06,
+      "loss": 0.4418,
+      "step": 2269
+    },
+    {
+      "epoch": 5.550122249388753,
+      "grad_norm": 0.2354920358820682,
+      "learning_rate": 3.492810981127249e-06,
+      "loss": 0.5292,
+      "step": 2270
+    },
+    {
+      "epoch": 5.552567237163814,
+      "grad_norm": 0.21692553317605812,
+      "learning_rate": 3.4916359353070924e-06,
+      "loss": 1.1664,
+      "step": 2271
+    },
+    {
+      "epoch": 5.555012224938875,
+      "grad_norm": 0.19653955060153305,
+      "learning_rate": 3.490460629457292e-06,
+      "loss": 0.7786,
+      "step": 2272
+    },
+    {
+      "epoch": 5.557457212713937,
+      "grad_norm": 0.19393087272557635,
+      "learning_rate": 3.4892850638860425e-06,
+      "loss": 0.5937,
+      "step": 2273
+    },
+    {
+      "epoch": 5.559902200488997,
+      "grad_norm": 0.2249400587845235,
+      "learning_rate": 3.4881092389016026e-06,
+      "loss": 0.4538,
+      "step": 2274
+    },
+    {
+      "epoch": 5.5623471882640585,
+      "grad_norm": 0.21143317314017557,
+      "learning_rate": 3.4869331548123014e-06,
+      "loss": 0.4862,
+      "step": 2275
+    },
+    {
+      "epoch": 5.56479217603912,
+      "grad_norm": 0.3041515457346303,
+      "learning_rate": 3.485756811926535e-06,
+      "loss": 1.0056,
+      "step": 2276
+    },
+    {
+      "epoch": 5.56723716381418,
+      "grad_norm": 0.13574752049043348,
+      "learning_rate": 3.484580210552767e-06,
+      "loss": 0.4825,
+      "step": 2277
+    },
+    {
+      "epoch": 5.569682151589242,
+      "grad_norm": 0.15425320873268222,
+      "learning_rate": 3.483403350999529e-06,
+      "loss": 0.469,
+      "step": 2278
+    },
+    {
+      "epoch": 5.572127139364303,
+      "grad_norm": 0.1868166851290588,
+      "learning_rate": 3.482226233575421e-06,
+      "loss": 0.3887,
+      "step": 2279
+    },
+    {
+      "epoch": 5.574572127139365,
+      "grad_norm": 0.3065152082082507,
+      "learning_rate": 3.4810488585891108e-06,
+      "loss": 0.6598,
+      "step": 2280
+    },
+    {
+      "epoch": 5.577017114914425,
+      "grad_norm": 0.18584043951062992,
+      "learning_rate": 3.4798712263493316e-06,
+      "loss": 0.3838,
+      "step": 2281
+    },
+    {
+      "epoch": 5.579462102689487,
+      "grad_norm": 0.18302250257697428,
+      "learning_rate": 3.478693337164887e-06,
+      "loss": 0.4542,
+      "step": 2282
+    },
+    {
+      "epoch": 5.581907090464548,
+      "grad_norm": 0.2198460604033387,
+      "learning_rate": 3.477515191344645e-06,
+      "loss": 0.4596,
+      "step": 2283
+    },
+    {
+      "epoch": 5.5843520782396086,
+      "grad_norm": 0.2589073095649969,
+      "learning_rate": 3.476336789197543e-06,
+      "loss": 0.6148,
+      "step": 2284
+    },
+    {
+      "epoch": 5.58679706601467,
+      "grad_norm": 0.25995060385363744,
+      "learning_rate": 3.475158131032586e-06,
+      "loss": 0.5695,
+      "step": 2285
+    },
+    {
+      "epoch": 5.589242053789731,
+      "grad_norm": 0.19766335673366794,
+      "learning_rate": 3.4739792171588445e-06,
+      "loss": 0.4749,
+      "step": 2286
+    },
+    {
+      "epoch": 5.591687041564792,
+      "grad_norm": 0.2783649648695533,
+      "learning_rate": 3.4728000478854564e-06,
+      "loss": 0.6517,
+      "step": 2287
+    },
+    {
+      "epoch": 5.594132029339853,
+      "grad_norm": 0.2536550589136312,
+      "learning_rate": 3.471620623521626e-06,
+      "loss": 0.4347,
+      "step": 2288
+    },
+    {
+      "epoch": 5.596577017114915,
+      "grad_norm": 0.2733002774272313,
+      "learning_rate": 3.470440944376626e-06,
+      "loss": 0.697,
+      "step": 2289
+    },
+    {
+      "epoch": 5.599022004889975,
+      "grad_norm": 0.18664770028280567,
+      "learning_rate": 3.4692610107597972e-06,
+      "loss": 0.4176,
+      "step": 2290
+    },
+    {
+      "epoch": 5.601466992665037,
+      "grad_norm": 0.17207840144638994,
+      "learning_rate": 3.4680808229805435e-06,
+      "loss": 0.4735,
+      "step": 2291
+    },
+    {
+      "epoch": 5.603911980440098,
+      "grad_norm": 0.16852075324321733,
+      "learning_rate": 3.466900381348337e-06,
+      "loss": 0.4496,
+      "step": 2292
+    },
+    {
+      "epoch": 5.606356968215159,
+      "grad_norm": 0.2238344788411242,
+      "learning_rate": 3.4657196861727165e-06,
+      "loss": 0.518,
+      "step": 2293
+    },
+    {
+      "epoch": 5.60880195599022,
+      "grad_norm": 0.19941653823702699,
+      "learning_rate": 3.464538737763288e-06,
+      "loss": 0.5941,
+      "step": 2294
+    },
+    {
+      "epoch": 5.611246943765281,
+      "grad_norm": 0.205900444381756,
+      "learning_rate": 3.463357536429723e-06,
+      "loss": 0.4351,
+      "step": 2295
+    },
+    {
+      "epoch": 5.613691931540342,
+      "grad_norm": 0.23487522768185495,
+      "learning_rate": 3.4621760824817596e-06,
+      "loss": 0.9335,
+      "step": 2296
+    },
+    {
+      "epoch": 5.616136919315403,
+      "grad_norm": 0.17050435137936504,
+      "learning_rate": 3.4609943762292015e-06,
+      "loss": 0.6002,
+      "step": 2297
+    },
+    {
+      "epoch": 5.618581907090465,
+      "grad_norm": 0.1711676849426343,
+      "learning_rate": 3.459812417981919e-06,
+      "loss": 0.6122,
+      "step": 2298
+    },
+    {
+      "epoch": 5.621026894865525,
+      "grad_norm": 0.22733038272996806,
+      "learning_rate": 3.458630208049849e-06,
+      "loss": 0.822,
+      "step": 2299
+    },
+    {
+      "epoch": 5.623471882640587,
+      "grad_norm": 0.19948333506378357,
+      "learning_rate": 3.457447746742995e-06,
+      "loss": 0.6581,
+      "step": 2300
+    },
+    {
+      "epoch": 5.625916870415648,
+      "grad_norm": 0.1325592647462057,
+      "learning_rate": 3.4562650343714242e-06,
+      "loss": 0.6559,
+      "step": 2301
+    },
+    {
+      "epoch": 5.628361858190709,
+      "grad_norm": 0.31221786315430433,
+      "learning_rate": 3.4550820712452716e-06,
+      "loss": 0.5888,
+      "step": 2302
+    },
+    {
+      "epoch": 5.63080684596577,
+      "grad_norm": 0.22947079647496033,
+      "learning_rate": 3.4538988576747355e-06,
+      "loss": 0.5025,
+      "step": 2303
+    },
+    {
+      "epoch": 5.6332518337408315,
+      "grad_norm": 0.15431899545948138,
+      "learning_rate": 3.4527153939700838e-06,
+      "loss": 0.5115,
+      "step": 2304
+    },
+    {
+      "epoch": 5.635696821515893,
+      "grad_norm": 0.2189846114351366,
+      "learning_rate": 3.4515316804416464e-06,
+      "loss": 0.5722,
+      "step": 2305
+    },
+    {
+      "epoch": 5.638141809290953,
+      "grad_norm": 0.2011928578909688,
+      "learning_rate": 3.450347717399821e-06,
+      "loss": 0.398,
+      "step": 2306
+    },
+    {
+      "epoch": 5.640586797066015,
+      "grad_norm": 0.16117320531938067,
+      "learning_rate": 3.4491635051550697e-06,
+      "loss": 0.3987,
+      "step": 2307
+    },
+    {
+      "epoch": 5.643031784841076,
+      "grad_norm": 0.3559818285440626,
+      "learning_rate": 3.447979044017918e-06,
+      "loss": 0.3819,
+      "step": 2308
+    },
+    {
+      "epoch": 5.645476772616137,
+      "grad_norm": 0.17453653781201287,
+      "learning_rate": 3.4467943342989617e-06,
+      "loss": 0.7916,
+      "step": 2309
+    },
+    {
+      "epoch": 5.647921760391198,
+      "grad_norm": 0.22425214333448937,
+      "learning_rate": 3.4456093763088577e-06,
+      "loss": 0.8948,
+      "step": 2310
+    },
+    {
+      "epoch": 5.65036674816626,
+      "grad_norm": 0.20817775833429783,
+      "learning_rate": 3.444424170358329e-06,
+      "loss": 0.365,
+      "step": 2311
+    },
+    {
+      "epoch": 5.65281173594132,
+      "grad_norm": 0.3149536892096343,
+      "learning_rate": 3.4432387167581637e-06,
+      "loss": 0.3978,
+      "step": 2312
+    },
+    {
+      "epoch": 5.6552567237163816,
+      "grad_norm": 0.23229871051620887,
+      "learning_rate": 3.442053015819214e-06,
+      "loss": 0.4871,
+      "step": 2313
+    },
+    {
+      "epoch": 5.657701711491443,
+      "grad_norm": 0.23252590902316755,
+      "learning_rate": 3.4408670678523994e-06,
+      "loss": 0.5863,
+      "step": 2314
+    },
+    {
+      "epoch": 5.6601466992665035,
+      "grad_norm": 0.24001265838313057,
+      "learning_rate": 3.439680873168701e-06,
+      "loss": 0.6851,
+      "step": 2315
+    },
+    {
+      "epoch": 5.662591687041565,
+      "grad_norm": 0.15895325123358917,
+      "learning_rate": 3.438494432079168e-06,
+      "loss": 0.5993,
+      "step": 2316
+    },
+    {
+      "epoch": 5.665036674816626,
+      "grad_norm": 0.357380995846018,
+      "learning_rate": 3.437307744894911e-06,
+      "loss": 0.4389,
+      "step": 2317
+    },
+    {
+      "epoch": 5.667481662591687,
+      "grad_norm": 0.15978268934021794,
+      "learning_rate": 3.4361208119271074e-06,
+      "loss": 0.4446,
+      "step": 2318
+    },
+    {
+      "epoch": 5.669926650366748,
+      "grad_norm": 0.19611973995521184,
+      "learning_rate": 3.4349336334869976e-06,
+      "loss": 0.4373,
+      "step": 2319
+    },
+    {
+      "epoch": 5.67237163814181,
+      "grad_norm": 0.18070571348711656,
+      "learning_rate": 3.433746209885887e-06,
+      "loss": 0.4942,
+      "step": 2320
+    },
+    {
+      "epoch": 5.67481662591687,
+      "grad_norm": 0.281780647619913,
+      "learning_rate": 3.432558541435146e-06,
+      "loss": 0.8002,
+      "step": 2321
+    },
+    {
+      "epoch": 5.677261613691932,
+      "grad_norm": 0.2535031880235516,
+      "learning_rate": 3.4313706284462077e-06,
+      "loss": 0.3914,
+      "step": 2322
+    },
+    {
+      "epoch": 5.679706601466993,
+      "grad_norm": 0.24631698730373758,
+      "learning_rate": 3.43018247123057e-06,
+      "loss": 0.4547,
+      "step": 2323
+    },
+    {
+      "epoch": 5.6821515892420535,
+      "grad_norm": 0.2523638451053032,
+      "learning_rate": 3.4289940700997953e-06,
+      "loss": 0.4387,
+      "step": 2324
+    },
+    {
+      "epoch": 5.684596577017115,
+      "grad_norm": 0.1785939487744636,
+      "learning_rate": 3.4278054253655092e-06,
+      "loss": 0.8406,
+      "step": 2325
+    },
+    {
+      "epoch": 5.687041564792176,
+      "grad_norm": 0.18286204645114806,
+      "learning_rate": 3.4266165373394024e-06,
+      "loss": 0.3072,
+      "step": 2326
+    },
+    {
+      "epoch": 5.689486552567237,
+      "grad_norm": 0.23259953466410532,
+      "learning_rate": 3.425427406333228e-06,
+      "loss": 0.4584,
+      "step": 2327
+    },
+    {
+      "epoch": 5.691931540342298,
+      "grad_norm": 0.1034310274458396,
+      "learning_rate": 3.424238032658803e-06,
+      "loss": 0.2987,
+      "step": 2328
+    },
+    {
+      "epoch": 5.69437652811736,
+      "grad_norm": 0.29110742561310504,
+      "learning_rate": 3.423048416628009e-06,
+      "loss": 0.4139,
+      "step": 2329
+    },
+    {
+      "epoch": 5.69682151589242,
+      "grad_norm": 0.30323392285102774,
+      "learning_rate": 3.42185855855279e-06,
+      "loss": 0.4489,
+      "step": 2330
+    },
+    {
+      "epoch": 5.699266503667482,
+      "grad_norm": 0.2699791636125737,
+      "learning_rate": 3.4206684587451542e-06,
+      "loss": 0.6126,
+      "step": 2331
+    },
+    {
+      "epoch": 5.701711491442543,
+      "grad_norm": 0.21679526599851004,
+      "learning_rate": 3.419478117517174e-06,
+      "loss": 0.4635,
+      "step": 2332
+    },
+    {
+      "epoch": 5.704156479217604,
+      "grad_norm": 0.28562832309040437,
+      "learning_rate": 3.418287535180982e-06,
+      "loss": 0.7509,
+      "step": 2333
+    },
+    {
+      "epoch": 5.706601466992665,
+      "grad_norm": 0.20567200023337673,
+      "learning_rate": 3.4170967120487787e-06,
+      "loss": 0.5643,
+      "step": 2334
+    },
+    {
+      "epoch": 5.709046454767726,
+      "grad_norm": 0.16727561433355023,
+      "learning_rate": 3.4159056484328236e-06,
+      "loss": 0.5666,
+      "step": 2335
+    },
+    {
+      "epoch": 5.711491442542787,
+      "grad_norm": 0.21633055378235785,
+      "learning_rate": 3.41471434464544e-06,
+      "loss": 0.5114,
+      "step": 2336
+    },
+    {
+      "epoch": 5.713936430317848,
+      "grad_norm": 0.22522864376940144,
+      "learning_rate": 3.4135228009990175e-06,
+      "loss": 0.4444,
+      "step": 2337
+    },
+    {
+      "epoch": 5.71638141809291,
+      "grad_norm": 0.1595446128902458,
+      "learning_rate": 3.4123310178060045e-06,
+      "loss": 0.4456,
+      "step": 2338
+    },
+    {
+      "epoch": 5.71882640586797,
+      "grad_norm": 0.2513658426774408,
+      "learning_rate": 3.4111389953789138e-06,
+      "loss": 0.4686,
+      "step": 2339
+    },
+    {
+      "epoch": 5.721271393643032,
+      "grad_norm": 0.22831297887049987,
+      "learning_rate": 3.4099467340303217e-06,
+      "loss": 0.6991,
+      "step": 2340
+    },
+    {
+      "epoch": 5.723716381418093,
+      "grad_norm": 0.25064053731396574,
+      "learning_rate": 3.4087542340728657e-06,
+      "loss": 0.6031,
+      "step": 2341
+    },
+    {
+      "epoch": 5.726161369193154,
+      "grad_norm": 0.2348242841412849,
+      "learning_rate": 3.4075614958192464e-06,
+      "loss": 0.9422,
+      "step": 2342
+    },
+    {
+      "epoch": 5.728606356968215,
+      "grad_norm": 0.32432060976957644,
+      "learning_rate": 3.4063685195822284e-06,
+      "loss": 0.3888,
+      "step": 2343
+    },
+    {
+      "epoch": 5.7310513447432765,
+      "grad_norm": 0.25680401587189,
+      "learning_rate": 3.405175305674637e-06,
+      "loss": 0.5747,
+      "step": 2344
+    },
+    {
+      "epoch": 5.733496332518337,
+      "grad_norm": 0.16128843814426544,
+      "learning_rate": 3.4039818544093587e-06,
+      "loss": 0.5815,
+      "step": 2345
+    },
+    {
+      "epoch": 5.735941320293398,
+      "grad_norm": 0.19500002871723887,
+      "learning_rate": 3.4027881660993454e-06,
+      "loss": 0.6205,
+      "step": 2346
+    },
+    {
+      "epoch": 5.73838630806846,
+      "grad_norm": 0.16938211466865477,
+      "learning_rate": 3.401594241057608e-06,
+      "loss": 0.6935,
+      "step": 2347
+    },
+    {
+      "epoch": 5.74083129584352,
+      "grad_norm": 0.27400129801822554,
+      "learning_rate": 3.400400079597223e-06,
+      "loss": 0.6052,
+      "step": 2348
+    },
+    {
+      "epoch": 5.743276283618582,
+      "grad_norm": 0.19611389587401065,
+      "learning_rate": 3.399205682031326e-06,
+      "loss": 0.5777,
+      "step": 2349
+    },
+    {
+      "epoch": 5.745721271393643,
+      "grad_norm": 0.17681117528036347,
+      "learning_rate": 3.398011048673114e-06,
+      "loss": 0.4115,
+      "step": 2350
+    },
+    {
+      "epoch": 5.748166259168704,
+      "grad_norm": 0.2375043940666574,
+      "learning_rate": 3.396816179835849e-06,
+      "loss": 0.446,
+      "step": 2351
+    },
+    {
+      "epoch": 5.750611246943765,
+      "grad_norm": 0.20185581146427406,
+      "learning_rate": 3.395621075832851e-06,
+      "loss": 0.6884,
+      "step": 2352
+    },
+    {
+      "epoch": 5.7530562347188265,
+      "grad_norm": 0.24129838127533926,
+      "learning_rate": 3.3944257369775063e-06,
+      "loss": 0.4558,
+      "step": 2353
+    },
+    {
+      "epoch": 5.755501222493887,
+      "grad_norm": 0.16402889712190388,
+      "learning_rate": 3.3932301635832586e-06,
+      "loss": 0.5296,
+      "step": 2354
+    },
+    {
+      "epoch": 5.7579462102689485,
+      "grad_norm": 0.2113663175776669,
+      "learning_rate": 3.392034355963615e-06,
+      "loss": 0.5456,
+      "step": 2355
+    },
+    {
+      "epoch": 5.76039119804401,
+      "grad_norm": 0.2918791855462405,
+      "learning_rate": 3.390838314432143e-06,
+      "loss": 0.5995,
+      "step": 2356
+    },
+    {
+      "epoch": 5.76283618581907,
+      "grad_norm": 0.24424203231717198,
+      "learning_rate": 3.3896420393024715e-06,
+      "loss": 0.8061,
+      "step": 2357
+    },
+    {
+      "epoch": 5.765281173594132,
+      "grad_norm": 0.18538240695553068,
+      "learning_rate": 3.3884455308882934e-06,
+      "loss": 0.4448,
+      "step": 2358
+    },
+    {
+      "epoch": 5.767726161369193,
+      "grad_norm": 0.2745739035076842,
+      "learning_rate": 3.387248789503359e-06,
+      "loss": 0.8291,
+      "step": 2359
+    },
+    {
+      "epoch": 5.770171149144255,
+      "grad_norm": 0.27461866254046174,
+      "learning_rate": 3.386051815461482e-06,
+      "loss": 0.5828,
+      "step": 2360
+    },
+    {
+      "epoch": 5.772616136919315,
+      "grad_norm": 0.23155057276696558,
+      "learning_rate": 3.384854609076536e-06,
+      "loss": 0.6884,
+      "step": 2361
+    },
+    {
+      "epoch": 5.775061124694377,
+      "grad_norm": 0.19867924418971772,
+      "learning_rate": 3.3836571706624553e-06,
+      "loss": 0.4309,
+      "step": 2362
+    },
+    {
+      "epoch": 5.777506112469438,
+      "grad_norm": 0.2259343515177491,
+      "learning_rate": 3.3824595005332367e-06,
+      "loss": 0.267,
+      "step": 2363
+    },
+    {
+      "epoch": 5.7799511002444985,
+      "grad_norm": 0.2029600668067855,
+      "learning_rate": 3.381261599002936e-06,
+      "loss": 0.5574,
+      "step": 2364
+    },
+    {
+      "epoch": 5.78239608801956,
+      "grad_norm": 0.18730033859746603,
+      "learning_rate": 3.3800634663856718e-06,
+      "loss": 0.3505,
+      "step": 2365
+    },
+    {
+      "epoch": 5.784841075794621,
+      "grad_norm": 0.21496789442092193,
+      "learning_rate": 3.3788651029956205e-06,
+      "loss": 0.6177,
+      "step": 2366
+    },
+    {
+      "epoch": 5.787286063569682,
+      "grad_norm": 0.2821557809849861,
+      "learning_rate": 3.37766650914702e-06,
+      "loss": 0.775,
+      "step": 2367
+    },
+    {
+      "epoch": 5.789731051344743,
+      "grad_norm": 0.20019550564283553,
+      "learning_rate": 3.37646768515417e-06,
+      "loss": 0.4221,
+      "step": 2368
+    },
+    {
+      "epoch": 5.792176039119805,
+      "grad_norm": 0.2764469422138519,
+      "learning_rate": 3.37526863133143e-06,
+      "loss": 0.3791,
+      "step": 2369
+    },
+    {
+      "epoch": 5.794621026894865,
+      "grad_norm": 0.15641115271093253,
+      "learning_rate": 3.3740693479932184e-06,
+      "loss": 0.5918,
+      "step": 2370
+    },
+    {
+      "epoch": 5.797066014669927,
+      "grad_norm": 0.14546444337694192,
+      "learning_rate": 3.372869835454015e-06,
+      "loss": 0.4162,
+      "step": 2371
+    },
+    {
+      "epoch": 5.799511002444988,
+      "grad_norm": 0.33527827936283466,
+      "learning_rate": 3.37167009402836e-06,
+      "loss": 0.6958,
+      "step": 2372
+    },
+    {
+      "epoch": 5.801955990220049,
+      "grad_norm": 0.27792526545278035,
+      "learning_rate": 3.3704701240308525e-06,
+      "loss": 1.1211,
+      "step": 2373
+    },
+    {
+      "epoch": 5.80440097799511,
+      "grad_norm": 0.14290643840602424,
+      "learning_rate": 3.3692699257761517e-06,
+      "loss": 0.433,
+      "step": 2374
+    },
+    {
+      "epoch": 5.806845965770171,
+      "grad_norm": 0.19447193625277287,
+      "learning_rate": 3.3680694995789788e-06,
+      "loss": 0.5248,
+      "step": 2375
+    },
+    {
+      "epoch": 5.809290953545232,
+      "grad_norm": 0.24049011102098605,
+      "learning_rate": 3.3668688457541103e-06,
+      "loss": 0.6636,
+      "step": 2376
+    },
+    {
+      "epoch": 5.811735941320293,
+      "grad_norm": 0.2144616902603737,
+      "learning_rate": 3.3656679646163876e-06,
+      "loss": 0.7337,
+      "step": 2377
+    },
+    {
+      "epoch": 5.814180929095355,
+      "grad_norm": 0.28839858018112746,
+      "learning_rate": 3.3644668564807083e-06,
+      "loss": 0.4367,
+      "step": 2378
+    },
+    {
+      "epoch": 5.816625916870415,
+      "grad_norm": 0.22087501383301614,
+      "learning_rate": 3.3632655216620298e-06,
+      "loss": 0.9032,
+      "step": 2379
+    },
+    {
+      "epoch": 5.819070904645477,
+      "grad_norm": 0.23609888178573912,
+      "learning_rate": 3.3620639604753702e-06,
+      "loss": 0.4583,
+      "step": 2380
+    },
+    {
+      "epoch": 5.821515892420538,
+      "grad_norm": 0.3296580145779637,
+      "learning_rate": 3.3608621732358065e-06,
+      "loss": 0.4745,
+      "step": 2381
+    },
+    {
+      "epoch": 5.823960880195599,
+      "grad_norm": 0.19935878199940832,
+      "learning_rate": 3.359660160258475e-06,
+      "loss": 0.5053,
+      "step": 2382
+    },
+    {
+      "epoch": 5.82640586797066,
+      "grad_norm": 0.3043803220681292,
+      "learning_rate": 3.3584579218585706e-06,
+      "loss": 0.3901,
+      "step": 2383
+    },
+    {
+      "epoch": 5.8288508557457215,
+      "grad_norm": 0.2235664326070787,
+      "learning_rate": 3.357255458351347e-06,
+      "loss": 0.7055,
+      "step": 2384
+    },
+    {
+      "epoch": 5.831295843520783,
+      "grad_norm": 0.22455293351263333,
+      "learning_rate": 3.3560527700521196e-06,
+      "loss": 0.4853,
+      "step": 2385
+    },
+    {
+      "epoch": 5.833740831295843,
+      "grad_norm": 0.1839662483921858,
+      "learning_rate": 3.354849857276259e-06,
+      "loss": 0.4479,
+      "step": 2386
+    },
+    {
+      "epoch": 5.836185819070905,
+      "grad_norm": 0.16714724607182893,
+      "learning_rate": 3.3536467203391978e-06,
+      "loss": 0.4347,
+      "step": 2387
+    },
+    {
+      "epoch": 5.838630806845966,
+      "grad_norm": 0.20050619843367853,
+      "learning_rate": 3.3524433595564254e-06,
+      "loss": 0.8465,
+      "step": 2388
+    },
+    {
+      "epoch": 5.841075794621027,
+      "grad_norm": 0.20325125777278805,
+      "learning_rate": 3.35123977524349e-06,
+      "loss": 0.9144,
+      "step": 2389
+    },
+    {
+      "epoch": 5.843520782396088,
+      "grad_norm": 0.18782007021554803,
+      "learning_rate": 3.3500359677159993e-06,
+      "loss": 0.6118,
+      "step": 2390
+    },
+    {
+      "epoch": 5.84596577017115,
+      "grad_norm": 0.31675184819799923,
+      "learning_rate": 3.3488319372896193e-06,
+      "loss": 0.6424,
+      "step": 2391
+    },
+    {
+      "epoch": 5.84841075794621,
+      "grad_norm": 0.21448991271551274,
+      "learning_rate": 3.347627684280075e-06,
+      "loss": 0.6228,
+      "step": 2392
+    },
+    {
+      "epoch": 5.8508557457212715,
+      "grad_norm": 0.2339446009528034,
+      "learning_rate": 3.3464232090031485e-06,
+      "loss": 0.2827,
+      "step": 2393
+    },
+    {
+      "epoch": 5.853300733496333,
+      "grad_norm": 0.26442900891524973,
+      "learning_rate": 3.3452185117746803e-06,
+      "loss": 0.7434,
+      "step": 2394
+    },
+    {
+      "epoch": 5.8557457212713935,
+      "grad_norm": 0.23474950026903882,
+      "learning_rate": 3.3440135929105704e-06,
+      "loss": 0.6813,
+      "step": 2395
+    },
+    {
+      "epoch": 5.858190709046455,
+      "grad_norm": 0.2750358149936989,
+      "learning_rate": 3.3428084527267756e-06,
+      "loss": 0.6497,
+      "step": 2396
+    },
+    {
+      "epoch": 5.860635696821516,
+      "grad_norm": 0.28262994192473495,
+      "learning_rate": 3.341603091539312e-06,
+      "loss": 0.3396,
+      "step": 2397
+    },
+    {
+      "epoch": 5.863080684596577,
+      "grad_norm": 0.3141255163866676,
+      "learning_rate": 3.340397509664252e-06,
+      "loss": 0.706,
+      "step": 2398
+    },
+    {
+      "epoch": 5.865525672371638,
+      "grad_norm": 0.1986453908934641,
+      "learning_rate": 3.3391917074177277e-06,
+      "loss": 0.4916,
+      "step": 2399
+    },
+    {
+      "epoch": 5.8679706601467,
+      "grad_norm": 0.39451721311496163,
+      "learning_rate": 3.3379856851159263e-06,
+      "loss": 0.4414,
+      "step": 2400
+    },
+    {
+      "epoch": 5.87041564792176,
+      "grad_norm": 0.30057604698705626,
+      "learning_rate": 3.336779443075096e-06,
+      "loss": 0.5294,
+      "step": 2401
+    },
+    {
+      "epoch": 5.872860635696822,
+      "grad_norm": 0.2523414290743578,
+      "learning_rate": 3.3355729816115413e-06,
+      "loss": 1.1975,
+      "step": 2402
+    },
+    {
+      "epoch": 5.875305623471883,
+      "grad_norm": 0.21586962034332774,
+      "learning_rate": 3.3343663010416235e-06,
+      "loss": 0.4718,
+      "step": 2403
+    },
+    {
+      "epoch": 5.8777506112469435,
+      "grad_norm": 0.23634880431323999,
+      "learning_rate": 3.3331594016817613e-06,
+      "loss": 0.7717,
+      "step": 2404
+    },
+    {
+      "epoch": 5.880195599022005,
+      "grad_norm": 0.25680720406637997,
+      "learning_rate": 3.331952283848432e-06,
+      "loss": 0.4945,
+      "step": 2405
+    },
+    {
+      "epoch": 5.882640586797066,
+      "grad_norm": 0.1953450736442293,
+      "learning_rate": 3.3307449478581687e-06,
+      "loss": 0.7775,
+      "step": 2406
+    },
+    {
+      "epoch": 5.885085574572127,
+      "grad_norm": 0.22571656809171378,
+      "learning_rate": 3.3295373940275637e-06,
+      "loss": 0.5714,
+      "step": 2407
+    },
+    {
+      "epoch": 5.887530562347188,
+      "grad_norm": 0.1916419682898357,
+      "learning_rate": 3.3283296226732655e-06,
+      "loss": 0.7468,
+      "step": 2408
+    },
+    {
+      "epoch": 5.88997555012225,
+      "grad_norm": 0.2269492305225127,
+      "learning_rate": 3.3271216341119775e-06,
+      "loss": 0.4385,
+      "step": 2409
+    },
+    {
+      "epoch": 5.89242053789731,
+      "grad_norm": 0.26957986247964194,
+      "learning_rate": 3.325913428660463e-06,
+      "loss": 0.7777,
+      "step": 2410
+    },
+    {
+      "epoch": 5.894865525672372,
+      "grad_norm": 0.205830739107848,
+      "learning_rate": 3.3247050066355414e-06,
+      "loss": 0.6831,
+      "step": 2411
+    },
+    {
+      "epoch": 5.897310513447433,
+      "grad_norm": 0.27129806025979647,
+      "learning_rate": 3.323496368354089e-06,
+      "loss": 0.9132,
+      "step": 2412
+    },
+    {
+      "epoch": 5.899755501222494,
+      "grad_norm": 0.25437160611557746,
+      "learning_rate": 3.322287514133038e-06,
+      "loss": 0.4542,
+      "step": 2413
+    },
+    {
+      "epoch": 5.902200488997555,
+      "grad_norm": 0.22100498812937824,
+      "learning_rate": 3.3210784442893774e-06,
+      "loss": 0.7659,
+      "step": 2414
+    },
+    {
+      "epoch": 5.904645476772616,
+      "grad_norm": 0.21827314939109838,
+      "learning_rate": 3.319869159140152e-06,
+      "loss": 0.5557,
+      "step": 2415
+    },
+    {
+      "epoch": 5.907090464547677,
+      "grad_norm": 0.21848556949574652,
+      "learning_rate": 3.318659659002466e-06,
+      "loss": 0.4069,
+      "step": 2416
+    },
+    {
+      "epoch": 5.909535452322738,
+      "grad_norm": 0.24273525461267176,
+      "learning_rate": 3.3174499441934773e-06,
+      "loss": 0.6805,
+      "step": 2417
+    },
+    {
+      "epoch": 5.9119804400978,
+      "grad_norm": 0.2248193183476636,
+      "learning_rate": 3.316240015030401e-06,
+      "loss": 1.1249,
+      "step": 2418
+    },
+    {
+      "epoch": 5.91442542787286,
+      "grad_norm": 0.30453235129947015,
+      "learning_rate": 3.3150298718305074e-06,
+      "loss": 0.4837,
+      "step": 2419
+    },
+    {
+      "epoch": 5.916870415647922,
+      "grad_norm": 0.21179115452327044,
+      "learning_rate": 3.313819514911124e-06,
+      "loss": 1.205,
+      "step": 2420
+    },
+    {
+      "epoch": 5.919315403422983,
+      "grad_norm": 0.3438382236206335,
+      "learning_rate": 3.312608944589635e-06,
+      "loss": 0.5912,
+      "step": 2421
+    },
+    {
+      "epoch": 5.921760391198044,
+      "grad_norm": 0.27379111103155923,
+      "learning_rate": 3.3113981611834777e-06,
+      "loss": 0.5731,
+      "step": 2422
+    },
+    {
+      "epoch": 5.924205378973105,
+      "grad_norm": 0.2283416150203339,
+      "learning_rate": 3.3101871650101504e-06,
+      "loss": 0.5076,
+      "step": 2423
+    },
+    {
+      "epoch": 5.9266503667481665,
+      "grad_norm": 0.13232700038212505,
+      "learning_rate": 3.308975956387201e-06,
+      "loss": 0.3935,
+      "step": 2424
+    },
+    {
+      "epoch": 5.929095354523227,
+      "grad_norm": 0.1916659782669626,
+      "learning_rate": 3.3077645356322374e-06,
+      "loss": 0.5252,
+      "step": 2425
+    },
+    {
+      "epoch": 5.931540342298288,
+      "grad_norm": 0.1995413766740241,
+      "learning_rate": 3.3065529030629216e-06,
+      "loss": 0.3891,
+      "step": 2426
+    },
+    {
+      "epoch": 5.93398533007335,
+      "grad_norm": 0.22823590683835818,
+      "learning_rate": 3.3053410589969713e-06,
+      "loss": 0.6494,
+      "step": 2427
+    },
+    {
+      "epoch": 5.93643031784841,
+      "grad_norm": 0.28847211065857165,
+      "learning_rate": 3.3041290037521607e-06,
+      "loss": 1.2323,
+      "step": 2428
+    },
+    {
+      "epoch": 5.938875305623472,
+      "grad_norm": 0.20218801050946034,
+      "learning_rate": 3.3029167376463177e-06,
+      "loss": 0.4479,
+      "step": 2429
+    },
+    {
+      "epoch": 5.941320293398533,
+      "grad_norm": 0.1497490467100729,
+      "learning_rate": 3.3017042609973253e-06,
+      "loss": 0.4546,
+      "step": 2430
+    },
+    {
+      "epoch": 5.943765281173594,
+      "grad_norm": 0.35081121438773644,
+      "learning_rate": 3.300491574123125e-06,
+      "loss": 0.6256,
+      "step": 2431
+    },
+    {
+      "epoch": 5.946210268948655,
+      "grad_norm": 0.16232737418186707,
+      "learning_rate": 3.2992786773417083e-06,
+      "loss": 0.5296,
+      "step": 2432
+    },
+    {
+      "epoch": 5.9486552567237165,
+      "grad_norm": 0.24893341466363325,
+      "learning_rate": 3.2980655709711263e-06,
+      "loss": 0.53,
+      "step": 2433
+    },
+    {
+      "epoch": 5.951100244498777,
+      "grad_norm": 0.20435792058729257,
+      "learning_rate": 3.296852255329483e-06,
+      "loss": 0.4565,
+      "step": 2434
+    },
+    {
+      "epoch": 5.9535452322738385,
+      "grad_norm": 0.25615497450801267,
+      "learning_rate": 3.295638730734937e-06,
+      "loss": 0.4393,
+      "step": 2435
+    },
+    {
+      "epoch": 5.9559902200489,
+      "grad_norm": 0.25130964805989187,
+      "learning_rate": 3.2944249975057025e-06,
+      "loss": 0.5311,
+      "step": 2436
+    },
+    {
+      "epoch": 5.95843520782396,
+      "grad_norm": 0.24047857468989378,
+      "learning_rate": 3.2932110559600484e-06,
+      "loss": 0.7289,
+      "step": 2437
+    },
+    {
+      "epoch": 5.960880195599022,
+      "grad_norm": 0.23370984597690975,
+      "learning_rate": 3.2919969064162976e-06,
+      "loss": 0.47,
+      "step": 2438
+    },
+    {
+      "epoch": 5.963325183374083,
+      "grad_norm": 0.1543545392566495,
+      "learning_rate": 3.290782549192828e-06,
+      "loss": 0.4325,
+      "step": 2439
+    },
+    {
+      "epoch": 5.965770171149144,
+      "grad_norm": 0.31025893700427754,
+      "learning_rate": 3.289567984608072e-06,
+      "loss": 0.4936,
+      "step": 2440
+    },
+    {
+      "epoch": 5.968215158924205,
+      "grad_norm": 0.25288069852872647,
+      "learning_rate": 3.2883532129805163e-06,
+      "loss": 1.4045,
+      "step": 2441
+    },
+    {
+      "epoch": 5.970660146699267,
+      "grad_norm": 0.1238719909116039,
+      "learning_rate": 3.2871382346287016e-06,
+      "loss": 0.4157,
+      "step": 2442
+    },
+    {
+      "epoch": 5.973105134474328,
+      "grad_norm": 0.21633176972154994,
+      "learning_rate": 3.285923049871223e-06,
+      "loss": 0.4001,
+      "step": 2443
+    },
+    {
+      "epoch": 5.9755501222493885,
+      "grad_norm": 0.22771966866569052,
+      "learning_rate": 3.2847076590267303e-06,
+      "loss": 0.45,
+      "step": 2444
+    },
+    {
+      "epoch": 5.97799511002445,
+      "grad_norm": 0.2841526951188325,
+      "learning_rate": 3.2834920624139253e-06,
+      "loss": 0.4622,
+      "step": 2445
+    },
+    {
+      "epoch": 5.980440097799511,
+      "grad_norm": 0.32736960767062323,
+      "learning_rate": 3.282276260351567e-06,
+      "loss": 0.3976,
+      "step": 2446
+    },
+    {
+      "epoch": 5.982885085574572,
+      "grad_norm": 0.183518306178426,
+      "learning_rate": 3.2810602531584663e-06,
+      "loss": 0.4949,
+      "step": 2447
+    },
+    {
+      "epoch": 5.985330073349633,
+      "grad_norm": 0.16518993210911348,
+      "learning_rate": 3.2798440411534878e-06,
+      "loss": 0.3782,
+      "step": 2448
+    },
+    {
+      "epoch": 5.987775061124695,
+      "grad_norm": 0.2560545753932088,
+      "learning_rate": 3.2786276246555494e-06,
+      "loss": 0.8431,
+      "step": 2449
+    },
+    {
+      "epoch": 5.990220048899755,
+      "grad_norm": 0.23437409028110592,
+      "learning_rate": 3.2774110039836233e-06,
+      "loss": 0.6213,
+      "step": 2450
+    },
+    {
+      "epoch": 5.992665036674817,
+      "grad_norm": 0.19194742058526684,
+      "learning_rate": 3.276194179456737e-06,
+      "loss": 0.4524,
+      "step": 2451
+    },
+    {
+      "epoch": 5.995110024449878,
+      "grad_norm": 0.2782643495525404,
+      "learning_rate": 3.274977151393969e-06,
+      "loss": 0.9042,
+      "step": 2452
+    },
+    {
+      "epoch": 5.997555012224939,
+      "grad_norm": 0.3026044596217569,
+      "learning_rate": 3.273759920114451e-06,
+      "loss": 0.5031,
+      "step": 2453
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.23036025024613652,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.5018,
+      "step": 2454
+    },
+    {
+      "epoch": 6.002444987775061,
+      "grad_norm": 0.2513976929482409,
+      "learning_rate": 3.2713248491819627e-06,
+      "loss": 0.4424,
+      "step": 2455
+    },
+    {
+      "epoch": 6.004889975550122,
+      "grad_norm": 0.22148987313550889,
+      "learning_rate": 3.2701070101675245e-06,
+      "loss": 0.8285,
+      "step": 2456
+    },
+    {
+      "epoch": 6.007334963325183,
+      "grad_norm": 0.2279291324243467,
+      "learning_rate": 3.268888969213401e-06,
+      "loss": 0.4299,
+      "step": 2457
+    },
+    {
+      "epoch": 6.009779951100245,
+      "grad_norm": 0.2758919141498375,
+      "learning_rate": 3.2676707266389872e-06,
+      "loss": 0.6349,
+      "step": 2458
+    },
+    {
+      "epoch": 6.012224938875305,
+      "grad_norm": 0.28653927076381486,
+      "learning_rate": 3.2664522827637356e-06,
+      "loss": 0.762,
+      "step": 2459
+    },
+    {
+      "epoch": 6.014669926650367,
+      "grad_norm": 0.20640339419662412,
+      "learning_rate": 3.2652336379071507e-06,
+      "loss": 0.4848,
+      "step": 2460
+    },
+    {
+      "epoch": 6.017114914425428,
+      "grad_norm": 0.2532503963981896,
+      "learning_rate": 3.264014792388789e-06,
+      "loss": 0.4262,
+      "step": 2461
+    },
+    {
+      "epoch": 6.019559902200489,
+      "grad_norm": 0.19932666924854608,
+      "learning_rate": 3.2627957465282597e-06,
+      "loss": 0.7791,
+      "step": 2462
+    },
+    {
+      "epoch": 6.02200488997555,
+      "grad_norm": 0.2758667543455121,
+      "learning_rate": 3.2615765006452237e-06,
+      "loss": 0.8319,
+      "step": 2463
+    },
+    {
+      "epoch": 6.0244498777506115,
+      "grad_norm": 0.26867954379582326,
+      "learning_rate": 3.2603570550593956e-06,
+      "loss": 0.7497,
+      "step": 2464
+    },
+    {
+      "epoch": 6.026894865525672,
+      "grad_norm": 0.21691323347737473,
+      "learning_rate": 3.259137410090542e-06,
+      "loss": 0.5305,
+      "step": 2465
+    },
+    {
+      "epoch": 6.029339853300733,
+      "grad_norm": 0.22995714216288998,
+      "learning_rate": 3.2579175660584823e-06,
+      "loss": 0.7501,
+      "step": 2466
+    },
+    {
+      "epoch": 6.031784841075795,
+      "grad_norm": 0.14620688339538562,
+      "learning_rate": 3.2566975232830866e-06,
+      "loss": 0.3846,
+      "step": 2467
+    },
+    {
+      "epoch": 6.034229828850855,
+      "grad_norm": 0.2284254270817203,
+      "learning_rate": 3.2554772820842793e-06,
+      "loss": 0.9083,
+      "step": 2468
+    },
+    {
+      "epoch": 6.036674816625917,
+      "grad_norm": 0.2693249352597879,
+      "learning_rate": 3.2542568427820336e-06,
+      "loss": 1.1761,
+      "step": 2469
+    },
+    {
+      "epoch": 6.039119804400978,
+      "grad_norm": 0.3425602955655935,
+      "learning_rate": 3.2530362056963775e-06,
+      "loss": 0.3438,
+      "step": 2470
+    },
+    {
+      "epoch": 6.041564792176039,
+      "grad_norm": 0.17441389457322984,
+      "learning_rate": 3.2518153711473917e-06,
+      "loss": 0.619,
+      "step": 2471
+    },
+    {
+      "epoch": 6.0440097799511,
+      "grad_norm": 0.18718678685061632,
+      "learning_rate": 3.2505943394552046e-06,
+      "loss": 0.9092,
+      "step": 2472
+    },
+    {
+      "epoch": 6.0464547677261615,
+      "grad_norm": 0.29704072016157496,
+      "learning_rate": 3.24937311094e-06,
+      "loss": 0.4548,
+      "step": 2473
+    },
+    {
+      "epoch": 6.048899755501223,
+      "grad_norm": 0.2865220371401877,
+      "learning_rate": 3.2481516859220107e-06,
+      "loss": 0.5146,
+      "step": 2474
+    },
+    {
+      "epoch": 6.0513447432762835,
+      "grad_norm": 0.2783430806274797,
+      "learning_rate": 3.2469300647215234e-06,
+      "loss": 0.3906,
+      "step": 2475
+    },
+    {
+      "epoch": 6.053789731051345,
+      "grad_norm": 0.32744867983122616,
+      "learning_rate": 3.2457082476588752e-06,
+      "loss": 0.4666,
+      "step": 2476
+    },
+    {
+      "epoch": 6.056234718826406,
+      "grad_norm": 0.24895326915751598,
+      "learning_rate": 3.2444862350544546e-06,
+      "loss": 0.8376,
+      "step": 2477
+    },
+    {
+      "epoch": 6.058679706601467,
+      "grad_norm": 0.1849687168234643,
+      "learning_rate": 3.243264027228701e-06,
+      "loss": 0.8888,
+      "step": 2478
+    },
+    {
+      "epoch": 6.061124694376528,
+      "grad_norm": 0.25822622680820934,
+      "learning_rate": 3.2420416245021036e-06,
+      "loss": 0.4463,
+      "step": 2479
+    },
+    {
+      "epoch": 6.06356968215159,
+      "grad_norm": 0.23066420052494863,
+      "learning_rate": 3.2408190271952073e-06,
+      "loss": 0.5821,
+      "step": 2480
+    },
+    {
+      "epoch": 6.06601466992665,
+      "grad_norm": 0.22955840671648725,
+      "learning_rate": 3.2395962356286025e-06,
+      "loss": 0.4257,
+      "step": 2481
+    },
+    {
+      "epoch": 6.068459657701712,
+      "grad_norm": 0.1915321510386415,
+      "learning_rate": 3.2383732501229358e-06,
+      "loss": 0.6004,
+      "step": 2482
+    },
+    {
+      "epoch": 6.070904645476773,
+      "grad_norm": 0.18455112464336287,
+      "learning_rate": 3.2371500709989e-06,
+      "loss": 0.4722,
+      "step": 2483
+    },
+    {
+      "epoch": 6.0733496332518335,
+      "grad_norm": 0.166267426214357,
+      "learning_rate": 3.235926698577241e-06,
+      "loss": 0.37,
+      "step": 2484
+    },
+    {
+      "epoch": 6.075794621026895,
+      "grad_norm": 0.20769920510701356,
+      "learning_rate": 3.234703133178755e-06,
+      "loss": 0.8453,
+      "step": 2485
+    },
+    {
+      "epoch": 6.078239608801956,
+      "grad_norm": 0.27994606697243635,
+      "learning_rate": 3.2334793751242892e-06,
+      "loss": 0.7242,
+      "step": 2486
+    },
+    {
+      "epoch": 6.080684596577017,
+      "grad_norm": 0.250495774082677,
+      "learning_rate": 3.2322554247347414e-06,
+      "loss": 0.5065,
+      "step": 2487
+    },
+    {
+      "epoch": 6.083129584352078,
+      "grad_norm": 0.18909120342728641,
+      "learning_rate": 3.2310312823310584e-06,
+      "loss": 0.3028,
+      "step": 2488
+    },
+    {
+      "epoch": 6.08557457212714,
+      "grad_norm": 0.2518570256791643,
+      "learning_rate": 3.229806948234239e-06,
+      "loss": 0.4713,
+      "step": 2489
+    },
+    {
+      "epoch": 6.0880195599022,
+      "grad_norm": 0.15464732727439603,
+      "learning_rate": 3.228582422765332e-06,
+      "loss": 0.3873,
+      "step": 2490
+    },
+    {
+      "epoch": 6.090464547677262,
+      "grad_norm": 0.27351013014371445,
+      "learning_rate": 3.2273577062454353e-06,
+      "loss": 0.5432,
+      "step": 2491
+    },
+    {
+      "epoch": 6.092909535452323,
+      "grad_norm": 0.3104069600799555,
+      "learning_rate": 3.226132798995697e-06,
+      "loss": 0.3664,
+      "step": 2492
+    },
+    {
+      "epoch": 6.095354523227384,
+      "grad_norm": 0.2443331767595374,
+      "learning_rate": 3.2249077013373174e-06,
+      "loss": 0.4585,
+      "step": 2493
+    },
+    {
+      "epoch": 6.097799511002445,
+      "grad_norm": 0.2530429237483783,
+      "learning_rate": 3.2236824135915447e-06,
+      "loss": 0.3253,
+      "step": 2494
+    },
+    {
+      "epoch": 6.100244498777506,
+      "grad_norm": 0.18790324363781535,
+      "learning_rate": 3.2224569360796763e-06,
+      "loss": 0.5985,
+      "step": 2495
+    },
+    {
+      "epoch": 6.102689486552567,
+      "grad_norm": 0.19242122029131728,
+      "learning_rate": 3.221231269123062e-06,
+      "loss": 0.4852,
+      "step": 2496
+    },
+    {
+      "epoch": 6.105134474327628,
+      "grad_norm": 0.33081058547196296,
+      "learning_rate": 3.220005413043098e-06,
+      "loss": 0.5803,
+      "step": 2497
+    },
+    {
+      "epoch": 6.10757946210269,
+      "grad_norm": 0.29377223958527715,
+      "learning_rate": 3.2187793681612333e-06,
+      "loss": 0.3906,
+      "step": 2498
+    },
+    {
+      "epoch": 6.11002444987775,
+      "grad_norm": 0.19995748238701366,
+      "learning_rate": 3.217553134798964e-06,
+      "loss": 0.41,
+      "step": 2499
+    },
+    {
+      "epoch": 6.112469437652812,
+      "grad_norm": 0.32221375834820093,
+      "learning_rate": 3.2163267132778376e-06,
+      "loss": 0.4924,
+      "step": 2500
+    },
+    {
+      "epoch": 6.114914425427873,
+      "grad_norm": 0.2574256412043207,
+      "learning_rate": 3.2151001039194496e-06,
+      "loss": 0.5619,
+      "step": 2501
+    },
+    {
+      "epoch": 6.117359413202934,
+      "grad_norm": 0.22303369012575647,
+      "learning_rate": 3.213873307045444e-06,
+      "loss": 0.9861,
+      "step": 2502
+    },
+    {
+      "epoch": 6.119804400977995,
+      "grad_norm": 0.22929064114826783,
+      "learning_rate": 3.2126463229775163e-06,
+      "loss": 1.2961,
+      "step": 2503
+    },
+    {
+      "epoch": 6.1222493887530565,
+      "grad_norm": 0.2554991718668828,
+      "learning_rate": 3.2114191520374082e-06,
+      "loss": 0.6175,
+      "step": 2504
+    },
+    {
+      "epoch": 6.124694376528117,
+      "grad_norm": 0.17705901702793223,
+      "learning_rate": 3.2101917945469142e-06,
+      "loss": 0.7001,
+      "step": 2505
+    },
+    {
+      "epoch": 6.127139364303178,
+      "grad_norm": 0.1564626973162439,
+      "learning_rate": 3.208964250827874e-06,
+      "loss": 0.4367,
+      "step": 2506
+    },
+    {
+      "epoch": 6.12958435207824,
+      "grad_norm": 0.2090933123625577,
+      "learning_rate": 3.2077365212021783e-06,
+      "loss": 0.4469,
+      "step": 2507
+    },
+    {
+      "epoch": 6.1320293398533,
+      "grad_norm": 0.2725937049814604,
+      "learning_rate": 3.206508605991765e-06,
+      "loss": 0.7854,
+      "step": 2508
+    },
+    {
+      "epoch": 6.134474327628362,
+      "grad_norm": 0.2181447435168512,
+      "learning_rate": 3.2052805055186225e-06,
+      "loss": 0.9504,
+      "step": 2509
+    },
+    {
+      "epoch": 6.136919315403423,
+      "grad_norm": 0.21896244225451889,
+      "learning_rate": 3.2040522201047876e-06,
+      "loss": 0.5894,
+      "step": 2510
+    },
+    {
+      "epoch": 6.139364303178484,
+      "grad_norm": 0.22276593939660236,
+      "learning_rate": 3.202823750072343e-06,
+      "loss": 0.5784,
+      "step": 2511
+    },
+    {
+      "epoch": 6.141809290953545,
+      "grad_norm": 0.23124239978473743,
+      "learning_rate": 3.2015950957434227e-06,
+      "loss": 0.3995,
+      "step": 2512
+    },
+    {
+      "epoch": 6.1442542787286065,
+      "grad_norm": 0.300247092769625,
+      "learning_rate": 3.200366257440208e-06,
+      "loss": 0.5032,
+      "step": 2513
+    },
+    {
+      "epoch": 6.146699266503667,
+      "grad_norm": 0.17729510603123494,
+      "learning_rate": 3.199137235484927e-06,
+      "loss": 0.8367,
+      "step": 2514
+    },
+    {
+      "epoch": 6.1491442542787285,
+      "grad_norm": 0.3328879061878735,
+      "learning_rate": 3.19790803019986e-06,
+      "loss": 0.8704,
+      "step": 2515
+    },
+    {
+      "epoch": 6.15158924205379,
+      "grad_norm": 0.18416947990908444,
+      "learning_rate": 3.1966786419073313e-06,
+      "loss": 0.4846,
+      "step": 2516
+    },
+    {
+      "epoch": 6.15403422982885,
+      "grad_norm": 0.22164864846257565,
+      "learning_rate": 3.1954490709297146e-06,
+      "loss": 0.4616,
+      "step": 2517
+    },
+    {
+      "epoch": 6.156479217603912,
+      "grad_norm": 0.17376515725261277,
+      "learning_rate": 3.194219317589432e-06,
+      "loss": 0.6963,
+      "step": 2518
+    },
+    {
+      "epoch": 6.158924205378973,
+      "grad_norm": 0.24220704878122706,
+      "learning_rate": 3.192989382208952e-06,
+      "loss": 0.7876,
+      "step": 2519
+    },
+    {
+      "epoch": 6.161369193154035,
+      "grad_norm": 0.19372422328941805,
+      "learning_rate": 3.191759265110793e-06,
+      "loss": 0.6141,
+      "step": 2520
+    },
+    {
+      "epoch": 6.163814180929095,
+      "grad_norm": 0.24604473150736486,
+      "learning_rate": 3.1905289666175192e-06,
+      "loss": 0.3585,
+      "step": 2521
+    },
+    {
+      "epoch": 6.166259168704157,
+      "grad_norm": 0.3211571738257054,
+      "learning_rate": 3.189298487051744e-06,
+      "loss": 0.6562,
+      "step": 2522
+    },
+    {
+      "epoch": 6.168704156479218,
+      "grad_norm": 0.24270946462886525,
+      "learning_rate": 3.1880678267361254e-06,
+      "loss": 0.3482,
+      "step": 2523
+    },
+    {
+      "epoch": 6.1711491442542785,
+      "grad_norm": 0.2959602281507325,
+      "learning_rate": 3.186836985993372e-06,
+      "loss": 0.7735,
+      "step": 2524
+    },
+    {
+      "epoch": 6.17359413202934,
+      "grad_norm": 0.17218137517657722,
+      "learning_rate": 3.1856059651462385e-06,
+      "loss": 0.7214,
+      "step": 2525
+    },
+    {
+      "epoch": 6.176039119804401,
+      "grad_norm": 0.25829475154854653,
+      "learning_rate": 3.1843747645175264e-06,
+      "loss": 0.8829,
+      "step": 2526
+    },
+    {
+      "epoch": 6.178484107579462,
+      "grad_norm": 0.21937994400205987,
+      "learning_rate": 3.183143384430085e-06,
+      "loss": 0.4261,
+      "step": 2527
+    },
+    {
+      "epoch": 6.180929095354523,
+      "grad_norm": 0.30233322799547563,
+      "learning_rate": 3.18191182520681e-06,
+      "loss": 0.4959,
+      "step": 2528
+    },
+    {
+      "epoch": 6.183374083129585,
+      "grad_norm": 0.13198622409451818,
+      "learning_rate": 3.180680087170644e-06,
+      "loss": 0.3255,
+      "step": 2529
+    },
+    {
+      "epoch": 6.185819070904645,
+      "grad_norm": 0.2093383653349948,
+      "learning_rate": 3.1794481706445777e-06,
+      "loss": 0.4383,
+      "step": 2530
+    },
+    {
+      "epoch": 6.188264058679707,
+      "grad_norm": 0.16627778733948947,
+      "learning_rate": 3.178216075951648e-06,
+      "loss": 0.9405,
+      "step": 2531
+    },
+    {
+      "epoch": 6.190709046454768,
+      "grad_norm": 0.1900166950557964,
+      "learning_rate": 3.176983803414938e-06,
+      "loss": 0.5988,
+      "step": 2532
+    },
+    {
+      "epoch": 6.193154034229829,
+      "grad_norm": 0.25613220311532364,
+      "learning_rate": 3.175751353357577e-06,
+      "loss": 0.7824,
+      "step": 2533
+    },
+    {
+      "epoch": 6.19559902200489,
+      "grad_norm": 0.33933603601668194,
+      "learning_rate": 3.1745187261027424e-06,
+      "loss": 0.4298,
+      "step": 2534
+    },
+    {
+      "epoch": 6.198044009779951,
+      "grad_norm": 0.18346956908614717,
+      "learning_rate": 3.1732859219736574e-06,
+      "loss": 0.9447,
+      "step": 2535
+    },
+    {
+      "epoch": 6.200488997555012,
+      "grad_norm": 0.23632754281393847,
+      "learning_rate": 3.1720529412935913e-06,
+      "loss": 0.7936,
+      "step": 2536
+    },
+    {
+      "epoch": 6.202933985330073,
+      "grad_norm": 0.18023655928255033,
+      "learning_rate": 3.1708197843858603e-06,
+      "loss": 0.7261,
+      "step": 2537
+    },
+    {
+      "epoch": 6.205378973105135,
+      "grad_norm": 0.16676162520201446,
+      "learning_rate": 3.1695864515738244e-06,
+      "loss": 0.6606,
+      "step": 2538
+    },
+    {
+      "epoch": 6.207823960880195,
+      "grad_norm": 0.17347837612809358,
+      "learning_rate": 3.1683529431808946e-06,
+      "loss": 0.6643,
+      "step": 2539
+    },
+    {
+      "epoch": 6.210268948655257,
+      "grad_norm": 0.1548626809353046,
+      "learning_rate": 3.1671192595305224e-06,
+      "loss": 0.4237,
+      "step": 2540
+    },
+    {
+      "epoch": 6.212713936430318,
+      "grad_norm": 0.22254059636503765,
+      "learning_rate": 3.16588540094621e-06,
+      "loss": 0.3548,
+      "step": 2541
+    },
+    {
+      "epoch": 6.215158924205379,
+      "grad_norm": 0.25483058080035254,
+      "learning_rate": 3.1646513677515023e-06,
+      "loss": 0.6653,
+      "step": 2542
+    },
+    {
+      "epoch": 6.21760391198044,
+      "grad_norm": 0.22700774966605214,
+      "learning_rate": 3.1634171602699905e-06,
+      "loss": 0.379,
+      "step": 2543
+    },
+    {
+      "epoch": 6.2200488997555015,
+      "grad_norm": 0.193380285790183,
+      "learning_rate": 3.1621827788253133e-06,
+      "loss": 0.6955,
+      "step": 2544
+    },
+    {
+      "epoch": 6.222493887530562,
+      "grad_norm": 0.226972848027893,
+      "learning_rate": 3.1609482237411528e-06,
+      "loss": 0.8486,
+      "step": 2545
+    },
+    {
+      "epoch": 6.224938875305623,
+      "grad_norm": 0.2227604730664185,
+      "learning_rate": 3.1597134953412386e-06,
+      "loss": 0.5108,
+      "step": 2546
+    },
+    {
+      "epoch": 6.227383863080685,
+      "grad_norm": 0.27975335020918013,
+      "learning_rate": 3.158478593949344e-06,
+      "loss": 0.4804,
+      "step": 2547
+    },
+    {
+      "epoch": 6.229828850855745,
+      "grad_norm": 0.20594765239204993,
+      "learning_rate": 3.157243519889288e-06,
+      "loss": 0.9913,
+      "step": 2548
+    },
+    {
+      "epoch": 6.232273838630807,
+      "grad_norm": 0.11914015953993708,
+      "learning_rate": 3.1560082734849367e-06,
+      "loss": 0.3124,
+      "step": 2549
+    },
+    {
+      "epoch": 6.234718826405868,
+      "grad_norm": 0.22108511080294047,
+      "learning_rate": 3.154772855060199e-06,
+      "loss": 0.628,
+      "step": 2550
+    },
+    {
+      "epoch": 6.237163814180929,
+      "grad_norm": 0.18121199155274048,
+      "learning_rate": 3.15353726493903e-06,
+      "loss": 0.5399,
+      "step": 2551
+    },
+    {
+      "epoch": 6.23960880195599,
+      "grad_norm": 0.20882863961173595,
+      "learning_rate": 3.1523015034454296e-06,
+      "loss": 0.7777,
+      "step": 2552
+    },
+    {
+      "epoch": 6.2420537897310515,
+      "grad_norm": 0.17963013134617542,
+      "learning_rate": 3.151065570903443e-06,
+      "loss": 0.3443,
+      "step": 2553
+    },
+    {
+      "epoch": 6.244498777506112,
+      "grad_norm": 0.2973909004383695,
+      "learning_rate": 3.14982946763716e-06,
+      "loss": 0.9612,
+      "step": 2554
+    },
+    {
+      "epoch": 6.2469437652811735,
+      "grad_norm": 0.2756763464726199,
+      "learning_rate": 3.1485931939707154e-06,
+      "loss": 1.2303,
+      "step": 2555
+    },
+    {
+      "epoch": 6.249388753056235,
+      "grad_norm": 0.20623850445049904,
+      "learning_rate": 3.1473567502282874e-06,
+      "loss": 0.3758,
+      "step": 2556
+    },
+    {
+      "epoch": 6.251833740831296,
+      "grad_norm": 0.19694382871989244,
+      "learning_rate": 3.146120136734101e-06,
+      "loss": 0.4014,
+      "step": 2557
+    },
+    {
+      "epoch": 6.254278728606357,
+      "grad_norm": 0.25355379159144675,
+      "learning_rate": 3.144883353812424e-06,
+      "loss": 0.6491,
+      "step": 2558
+    },
+    {
+      "epoch": 6.256723716381418,
+      "grad_norm": 0.21359266073758404,
+      "learning_rate": 3.14364640178757e-06,
+      "loss": 0.6015,
+      "step": 2559
+    },
+    {
+      "epoch": 6.25916870415648,
+      "grad_norm": 0.26574542446944377,
+      "learning_rate": 3.1424092809838952e-06,
+      "loss": 0.4261,
+      "step": 2560
+    },
+    {
+      "epoch": 6.26161369193154,
+      "grad_norm": 0.15649352546963427,
+      "learning_rate": 3.141171991725801e-06,
+      "loss": 0.5906,
+      "step": 2561
+    },
+    {
+      "epoch": 6.264058679706602,
+      "grad_norm": 0.22242197179468262,
+      "learning_rate": 3.1399345343377335e-06,
+      "loss": 0.5517,
+      "step": 2562
+    },
+    {
+      "epoch": 6.266503667481663,
+      "grad_norm": 0.22402933204778624,
+      "learning_rate": 3.1386969091441818e-06,
+      "loss": 0.5985,
+      "step": 2563
+    },
+    {
+      "epoch": 6.2689486552567235,
+      "grad_norm": 0.12688411178481487,
+      "learning_rate": 3.1374591164696797e-06,
+      "loss": 0.2545,
+      "step": 2564
+    },
+    {
+      "epoch": 6.271393643031785,
+      "grad_norm": 0.26120048639574395,
+      "learning_rate": 3.1362211566388062e-06,
+      "loss": 0.5335,
+      "step": 2565
+    },
+    {
+      "epoch": 6.273838630806846,
+      "grad_norm": 0.2763981462653861,
+      "learning_rate": 3.1349830299761807e-06,
+      "loss": 0.6066,
+      "step": 2566
+    },
+    {
+      "epoch": 6.276283618581907,
+      "grad_norm": 0.35923355819891295,
+      "learning_rate": 3.1337447368064688e-06,
+      "loss": 0.3717,
+      "step": 2567
+    },
+    {
+      "epoch": 6.278728606356968,
+      "grad_norm": 0.29447716705757104,
+      "learning_rate": 3.1325062774543797e-06,
+      "loss": 0.7071,
+      "step": 2568
+    },
+    {
+      "epoch": 6.28117359413203,
+      "grad_norm": 0.22013361056074715,
+      "learning_rate": 3.1312676522446663e-06,
+      "loss": 0.4954,
+      "step": 2569
+    },
+    {
+      "epoch": 6.28361858190709,
+      "grad_norm": 0.17737047831735606,
+      "learning_rate": 3.1300288615021233e-06,
+      "loss": 0.4877,
+      "step": 2570
+    },
+    {
+      "epoch": 6.286063569682152,
+      "grad_norm": 0.23894168058087356,
+      "learning_rate": 3.1287899055515914e-06,
+      "loss": 0.6056,
+      "step": 2571
+    },
+    {
+      "epoch": 6.288508557457213,
+      "grad_norm": 0.2558137720883508,
+      "learning_rate": 3.1275507847179507e-06,
+      "loss": 0.7645,
+      "step": 2572
+    },
+    {
+      "epoch": 6.290953545232274,
+      "grad_norm": 0.16682292042884223,
+      "learning_rate": 3.12631149932613e-06,
+      "loss": 0.4147,
+      "step": 2573
+    },
+    {
+      "epoch": 6.293398533007335,
+      "grad_norm": 0.22948517031074533,
+      "learning_rate": 3.125072049701097e-06,
+      "loss": 0.5882,
+      "step": 2574
+    },
+    {
+      "epoch": 6.295843520782396,
+      "grad_norm": 0.2005988469138682,
+      "learning_rate": 3.123832436167864e-06,
+      "loss": 0.3994,
+      "step": 2575
+    },
+    {
+      "epoch": 6.298288508557457,
+      "grad_norm": 0.23239011232948692,
+      "learning_rate": 3.1225926590514853e-06,
+      "loss": 0.4304,
+      "step": 2576
+    },
+    {
+      "epoch": 6.300733496332518,
+      "grad_norm": 0.2552286923519756,
+      "learning_rate": 3.12135271867706e-06,
+      "loss": 0.7305,
+      "step": 2577
+    },
+    {
+      "epoch": 6.30317848410758,
+      "grad_norm": 0.2030008074198635,
+      "learning_rate": 3.1201126153697275e-06,
+      "loss": 0.5212,
+      "step": 2578
+    },
+    {
+      "epoch": 6.30562347188264,
+      "grad_norm": 0.24745038198436645,
+      "learning_rate": 3.1188723494546723e-06,
+      "loss": 1.2205,
+      "step": 2579
+    },
+    {
+      "epoch": 6.308068459657702,
+      "grad_norm": 0.15748317085593955,
+      "learning_rate": 3.117631921257121e-06,
+      "loss": 0.5496,
+      "step": 2580
+    },
+    {
+      "epoch": 6.310513447432763,
+      "grad_norm": 0.12852504029890663,
+      "learning_rate": 3.116391331102341e-06,
+      "loss": 0.4139,
+      "step": 2581
+    },
+    {
+      "epoch": 6.312958435207824,
+      "grad_norm": 0.22214521695113232,
+      "learning_rate": 3.1151505793156433e-06,
+      "loss": 0.5323,
+      "step": 2582
+    },
+    {
+      "epoch": 6.315403422982885,
+      "grad_norm": 0.2233074061278456,
+      "learning_rate": 3.1139096662223822e-06,
+      "loss": 0.5108,
+      "step": 2583
+    },
+    {
+      "epoch": 6.3178484107579465,
+      "grad_norm": 0.29352310305741575,
+      "learning_rate": 3.1126685921479545e-06,
+      "loss": 0.4481,
+      "step": 2584
+    },
+    {
+      "epoch": 6.320293398533007,
+      "grad_norm": 0.2514815794016727,
+      "learning_rate": 3.111427357417796e-06,
+      "loss": 0.5029,
+      "step": 2585
+    },
+    {
+      "epoch": 6.322738386308068,
+      "grad_norm": 0.19136425463549375,
+      "learning_rate": 3.1101859623573886e-06,
+      "loss": 0.7785,
+      "step": 2586
+    },
+    {
+      "epoch": 6.32518337408313,
+      "grad_norm": 0.16416622497684552,
+      "learning_rate": 3.1089444072922537e-06,
+      "loss": 0.4277,
+      "step": 2587
+    },
+    {
+      "epoch": 6.32762836185819,
+      "grad_norm": 0.20160817436562767,
+      "learning_rate": 3.1077026925479553e-06,
+      "loss": 0.9888,
+      "step": 2588
+    },
+    {
+      "epoch": 6.330073349633252,
+      "grad_norm": 0.1981144325444725,
+      "learning_rate": 3.1064608184501e-06,
+      "loss": 0.7836,
+      "step": 2589
+    },
+    {
+      "epoch": 6.332518337408313,
+      "grad_norm": 0.202957447876274,
+      "learning_rate": 3.105218785324336e-06,
+      "loss": 0.4073,
+      "step": 2590
+    },
+    {
+      "epoch": 6.334963325183374,
+      "grad_norm": 0.3023134816194395,
+      "learning_rate": 3.1039765934963515e-06,
+      "loss": 0.5137,
+      "step": 2591
+    },
+    {
+      "epoch": 6.337408312958435,
+      "grad_norm": 0.22151347639408875,
+      "learning_rate": 3.1027342432918784e-06,
+      "loss": 0.5849,
+      "step": 2592
+    },
+    {
+      "epoch": 6.3398533007334965,
+      "grad_norm": 0.26270029158589314,
+      "learning_rate": 3.1014917350366893e-06,
+      "loss": 0.6923,
+      "step": 2593
+    },
+    {
+      "epoch": 6.342298288508557,
+      "grad_norm": 0.29995123460797396,
+      "learning_rate": 3.1002490690565983e-06,
+      "loss": 0.4826,
+      "step": 2594
+    },
+    {
+      "epoch": 6.3447432762836184,
+      "grad_norm": 0.21975879882477406,
+      "learning_rate": 3.099006245677461e-06,
+      "loss": 0.6614,
+      "step": 2595
+    },
+    {
+      "epoch": 6.34718826405868,
+      "grad_norm": 0.24819246308470977,
+      "learning_rate": 3.0977632652251742e-06,
+      "loss": 0.8443,
+      "step": 2596
+    },
+    {
+      "epoch": 6.34963325183374,
+      "grad_norm": 0.2783422709337147,
+      "learning_rate": 3.0965201280256753e-06,
+      "loss": 0.584,
+      "step": 2597
+    },
+    {
+      "epoch": 6.352078239608802,
+      "grad_norm": 0.19643894490004454,
+      "learning_rate": 3.095276834404944e-06,
+      "loss": 0.518,
+      "step": 2598
+    },
+    {
+      "epoch": 6.354523227383863,
+      "grad_norm": 0.23907666706004613,
+      "learning_rate": 3.094033384688999e-06,
+      "loss": 0.859,
+      "step": 2599
+    },
+    {
+      "epoch": 6.356968215158924,
+      "grad_norm": 0.20978596056628784,
+      "learning_rate": 3.0927897792039034e-06,
+      "loss": 0.4227,
+      "step": 2600
+    },
+    {
+      "epoch": 6.359413202933985,
+      "grad_norm": 0.32676125304010617,
+      "learning_rate": 3.0915460182757586e-06,
+      "loss": 0.4254,
+      "step": 2601
+    },
+    {
+      "epoch": 6.361858190709047,
+      "grad_norm": 0.1976576170849751,
+      "learning_rate": 3.090302102230705e-06,
+      "loss": 0.5292,
+      "step": 2602
+    },
+    {
+      "epoch": 6.364303178484108,
+      "grad_norm": 0.19039374279580804,
+      "learning_rate": 3.089058031394928e-06,
+      "loss": 0.4586,
+      "step": 2603
+    },
+    {
+      "epoch": 6.3667481662591685,
+      "grad_norm": 0.20728358996109225,
+      "learning_rate": 3.0878138060946506e-06,
+      "loss": 0.5421,
+      "step": 2604
+    },
+    {
+      "epoch": 6.36919315403423,
+      "grad_norm": 0.25487566401951084,
+      "learning_rate": 3.0865694266561376e-06,
+      "loss": 0.7118,
+      "step": 2605
+    },
+    {
+      "epoch": 6.371638141809291,
+      "grad_norm": 0.25448513293231295,
+      "learning_rate": 3.0853248934056935e-06,
+      "loss": 0.4779,
+      "step": 2606
+    },
+    {
+      "epoch": 6.374083129584352,
+      "grad_norm": 0.22245020829264056,
+      "learning_rate": 3.084080206669663e-06,
+      "loss": 0.4621,
+      "step": 2607
+    },
+    {
+      "epoch": 6.376528117359413,
+      "grad_norm": 0.23893368835176065,
+      "learning_rate": 3.0828353667744326e-06,
+      "loss": 0.3396,
+      "step": 2608
+    },
+    {
+      "epoch": 6.378973105134475,
+      "grad_norm": 0.19805314554787354,
+      "learning_rate": 3.0815903740464266e-06,
+      "loss": 0.8295,
+      "step": 2609
+    },
+    {
+      "epoch": 6.381418092909535,
+      "grad_norm": 0.31429540250501625,
+      "learning_rate": 3.0803452288121116e-06,
+      "loss": 0.736,
+      "step": 2610
+    },
+    {
+      "epoch": 6.383863080684597,
+      "grad_norm": 0.3680373955125361,
+      "learning_rate": 3.079099931397993e-06,
+      "loss": 0.5517,
+      "step": 2611
+    },
+    {
+      "epoch": 6.386308068459658,
+      "grad_norm": 0.16574992332841876,
+      "learning_rate": 3.077854482130615e-06,
+      "loss": 0.5557,
+      "step": 2612
+    },
+    {
+      "epoch": 6.388753056234719,
+      "grad_norm": 0.24630814219741592,
+      "learning_rate": 3.0766088813365646e-06,
+      "loss": 0.6251,
+      "step": 2613
+    },
+    {
+      "epoch": 6.39119804400978,
+      "grad_norm": 0.2517810928564044,
+      "learning_rate": 3.0753631293424662e-06,
+      "loss": 0.5654,
+      "step": 2614
+    },
+    {
+      "epoch": 6.393643031784841,
+      "grad_norm": 0.24910906025312432,
+      "learning_rate": 3.0741172264749843e-06,
+      "loss": 0.3776,
+      "step": 2615
+    },
+    {
+      "epoch": 6.396088019559902,
+      "grad_norm": 0.27301101124350163,
+      "learning_rate": 3.072871173060823e-06,
+      "loss": 0.4949,
+      "step": 2616
+    },
+    {
+      "epoch": 6.398533007334963,
+      "grad_norm": 0.24345622147713936,
+      "learning_rate": 3.0716249694267265e-06,
+      "loss": 0.4231,
+      "step": 2617
+    },
+    {
+      "epoch": 6.400977995110025,
+      "grad_norm": 0.38778318143143486,
+      "learning_rate": 3.0703786158994773e-06,
+      "loss": 0.4189,
+      "step": 2618
+    },
+    {
+      "epoch": 6.403422982885085,
+      "grad_norm": 0.17208381711521498,
+      "learning_rate": 3.0691321128058993e-06,
+      "loss": 0.4015,
+      "step": 2619
+    },
+    {
+      "epoch": 6.405867970660147,
+      "grad_norm": 0.19602949419441026,
+      "learning_rate": 3.0678854604728524e-06,
+      "loss": 0.8393,
+      "step": 2620
+    },
+    {
+      "epoch": 6.408312958435208,
+      "grad_norm": 0.2506024382072708,
+      "learning_rate": 3.0666386592272383e-06,
+      "loss": 0.5956,
+      "step": 2621
+    },
+    {
+      "epoch": 6.410757946210269,
+      "grad_norm": 0.3161966822281516,
+      "learning_rate": 3.065391709395996e-06,
+      "loss": 0.8634,
+      "step": 2622
+    },
+    {
+      "epoch": 6.41320293398533,
+      "grad_norm": 0.2266830823934656,
+      "learning_rate": 3.0641446113061056e-06,
+      "loss": 1.142,
+      "step": 2623
+    },
+    {
+      "epoch": 6.4156479217603914,
+      "grad_norm": 0.28127582778881116,
+      "learning_rate": 3.0628973652845844e-06,
+      "loss": 0.3812,
+      "step": 2624
+    },
+    {
+      "epoch": 6.418092909535452,
+      "grad_norm": 0.15016607763061554,
+      "learning_rate": 3.061649971658488e-06,
+      "loss": 0.4832,
+      "step": 2625
+    },
+    {
+      "epoch": 6.420537897310513,
+      "grad_norm": 0.25080755257887083,
+      "learning_rate": 3.060402430754912e-06,
+      "loss": 0.5978,
+      "step": 2626
+    },
+    {
+      "epoch": 6.422982885085575,
+      "grad_norm": 0.19395190758050576,
+      "learning_rate": 3.0591547429009904e-06,
+      "loss": 0.4887,
+      "step": 2627
+    },
+    {
+      "epoch": 6.425427872860635,
+      "grad_norm": 0.24081837594303346,
+      "learning_rate": 3.0579069084238956e-06,
+      "loss": 0.4192,
+      "step": 2628
+    },
+    {
+      "epoch": 6.427872860635697,
+      "grad_norm": 0.2411738263676204,
+      "learning_rate": 3.0566589276508378e-06,
+      "loss": 0.5571,
+      "step": 2629
+    },
+    {
+      "epoch": 6.430317848410758,
+      "grad_norm": 0.28248861885516363,
+      "learning_rate": 3.0554108009090673e-06,
+      "loss": 0.3714,
+      "step": 2630
+    },
+    {
+      "epoch": 6.432762836185819,
+      "grad_norm": 0.2433957390618439,
+      "learning_rate": 3.0541625285258697e-06,
+      "loss": 0.7324,
+      "step": 2631
+    },
+    {
+      "epoch": 6.43520782396088,
+      "grad_norm": 0.2592034021177652,
+      "learning_rate": 3.0529141108285716e-06,
+      "loss": 0.4371,
+      "step": 2632
+    },
+    {
+      "epoch": 6.4376528117359415,
+      "grad_norm": 0.18264399442590148,
+      "learning_rate": 3.051665548144537e-06,
+      "loss": 0.7301,
+      "step": 2633
+    },
+    {
+      "epoch": 6.440097799511002,
+      "grad_norm": 0.3052687934894556,
+      "learning_rate": 3.0504168408011675e-06,
+      "loss": 0.647,
+      "step": 2634
+    },
+    {
+      "epoch": 6.442542787286063,
+      "grad_norm": 0.19463382291350106,
+      "learning_rate": 3.0491679891259028e-06,
+      "loss": 0.5206,
+      "step": 2635
+    },
+    {
+      "epoch": 6.444987775061125,
+      "grad_norm": 0.2236750490856417,
+      "learning_rate": 3.0479189934462193e-06,
+      "loss": 0.427,
+      "step": 2636
+    },
+    {
+      "epoch": 6.447432762836186,
+      "grad_norm": 0.18561564322400415,
+      "learning_rate": 3.046669854089633e-06,
+      "loss": 0.4594,
+      "step": 2637
+    },
+    {
+      "epoch": 6.449877750611247,
+      "grad_norm": 0.16007845207349647,
+      "learning_rate": 3.045420571383698e-06,
+      "loss": 0.6762,
+      "step": 2638
+    },
+    {
+      "epoch": 6.452322738386308,
+      "grad_norm": 0.23037334715367425,
+      "learning_rate": 3.044171145656003e-06,
+      "loss": 0.5492,
+      "step": 2639
+    },
+    {
+      "epoch": 6.45476772616137,
+      "grad_norm": 0.23897160401969714,
+      "learning_rate": 3.0429215772341774e-06,
+      "loss": 0.6096,
+      "step": 2640
+    },
+    {
+      "epoch": 6.45721271393643,
+      "grad_norm": 0.1514064211222451,
+      "learning_rate": 3.0416718664458856e-06,
+      "loss": 0.4867,
+      "step": 2641
+    },
+    {
+      "epoch": 6.459657701711492,
+      "grad_norm": 0.20611106386340128,
+      "learning_rate": 3.04042201361883e-06,
+      "loss": 0.7408,
+      "step": 2642
+    },
+    {
+      "epoch": 6.462102689486553,
+      "grad_norm": 0.28683155873909766,
+      "learning_rate": 3.0391720190807526e-06,
+      "loss": 0.6533,
+      "step": 2643
+    },
+    {
+      "epoch": 6.4645476772616135,
+      "grad_norm": 0.17145523096021667,
+      "learning_rate": 3.0379218831594294e-06,
+      "loss": 0.5002,
+      "step": 2644
+    },
+    {
+      "epoch": 6.466992665036675,
+      "grad_norm": 0.3013053151233808,
+      "learning_rate": 3.0366716061826747e-06,
+      "loss": 0.6106,
+      "step": 2645
+    },
+    {
+      "epoch": 6.469437652811736,
+      "grad_norm": 0.27490909468164043,
+      "learning_rate": 3.035421188478339e-06,
+      "loss": 0.6135,
+      "step": 2646
+    },
+    {
+      "epoch": 6.471882640586797,
+      "grad_norm": 0.23641831376824862,
+      "learning_rate": 3.034170630374311e-06,
+      "loss": 0.4138,
+      "step": 2647
+    },
+    {
+      "epoch": 6.474327628361858,
+      "grad_norm": 0.21608543498617774,
+      "learning_rate": 3.032919932198517e-06,
+      "loss": 0.6937,
+      "step": 2648
+    },
+    {
+      "epoch": 6.47677261613692,
+      "grad_norm": 0.2505049747415038,
+      "learning_rate": 3.0316690942789166e-06,
+      "loss": 0.5515,
+      "step": 2649
+    },
+    {
+      "epoch": 6.47921760391198,
+      "grad_norm": 0.21512728303048317,
+      "learning_rate": 3.03041811694351e-06,
+      "loss": 0.7614,
+      "step": 2650
+    },
+    {
+      "epoch": 6.481662591687042,
+      "grad_norm": 0.20998272842913945,
+      "learning_rate": 3.02916700052033e-06,
+      "loss": 0.6434,
+      "step": 2651
+    },
+    {
+      "epoch": 6.484107579462103,
+      "grad_norm": 0.2525983228130296,
+      "learning_rate": 3.027915745337449e-06,
+      "loss": 0.427,
+      "step": 2652
+    },
+    {
+      "epoch": 6.486552567237164,
+      "grad_norm": 0.38571700008245036,
+      "learning_rate": 3.0266643517229754e-06,
+      "loss": 0.8491,
+      "step": 2653
+    },
+    {
+      "epoch": 6.488997555012225,
+      "grad_norm": 0.20265827562066865,
+      "learning_rate": 3.025412820005053e-06,
+      "loss": 0.4355,
+      "step": 2654
+    },
+    {
+      "epoch": 6.491442542787286,
+      "grad_norm": 0.3051546548153466,
+      "learning_rate": 3.0241611505118617e-06,
+      "loss": 0.9612,
+      "step": 2655
+    },
+    {
+      "epoch": 6.493887530562347,
+      "grad_norm": 0.1964812738010243,
+      "learning_rate": 3.0229093435716177e-06,
+      "loss": 0.7518,
+      "step": 2656
+    },
+    {
+      "epoch": 6.496332518337408,
+      "grad_norm": 0.17488816197186746,
+      "learning_rate": 3.0216573995125743e-06,
+      "loss": 0.4341,
+      "step": 2657
+    },
+    {
+      "epoch": 6.49877750611247,
+      "grad_norm": 0.19993841109438337,
+      "learning_rate": 3.0204053186630184e-06,
+      "loss": 0.4158,
+      "step": 2658
+    },
+    {
+      "epoch": 6.50122249388753,
+      "grad_norm": 0.2579996171861024,
+      "learning_rate": 3.0191531013512754e-06,
+      "loss": 0.5411,
+      "step": 2659
+    },
+    {
+      "epoch": 6.503667481662592,
+      "grad_norm": 0.23437155010463537,
+      "learning_rate": 3.0179007479057054e-06,
+      "loss": 0.5494,
+      "step": 2660
+    },
+    {
+      "epoch": 6.506112469437653,
+      "grad_norm": 0.2698069491630078,
+      "learning_rate": 3.0166482586547035e-06,
+      "loss": 0.9092,
+      "step": 2661
+    },
+    {
+      "epoch": 6.508557457212714,
+      "grad_norm": 0.27178839949031947,
+      "learning_rate": 3.015395633926702e-06,
+      "loss": 0.6301,
+      "step": 2662
+    },
+    {
+      "epoch": 6.511002444987775,
+      "grad_norm": 0.33727667953944246,
+      "learning_rate": 3.014142874050166e-06,
+      "loss": 0.4578,
+      "step": 2663
+    },
+    {
+      "epoch": 6.513447432762836,
+      "grad_norm": 0.2233925985354532,
+      "learning_rate": 3.0128899793536004e-06,
+      "loss": 0.6856,
+      "step": 2664
+    },
+    {
+      "epoch": 6.515892420537897,
+      "grad_norm": 0.2893950352519993,
+      "learning_rate": 3.011636950165541e-06,
+      "loss": 0.5547,
+      "step": 2665
+    },
+    {
+      "epoch": 6.518337408312958,
+      "grad_norm": 0.24246571529305033,
+      "learning_rate": 3.01038378681456e-06,
+      "loss": 0.3582,
+      "step": 2666
+    },
+    {
+      "epoch": 6.52078239608802,
+      "grad_norm": 0.22241750449147513,
+      "learning_rate": 3.0091304896292685e-06,
+      "loss": 1.056,
+      "step": 2667
+    },
+    {
+      "epoch": 6.52322738386308,
+      "grad_norm": 0.20996244740252007,
+      "learning_rate": 3.0078770589383067e-06,
+      "loss": 0.605,
+      "step": 2668
+    },
+    {
+      "epoch": 6.525672371638142,
+      "grad_norm": 0.23586272198530978,
+      "learning_rate": 3.006623495070355e-06,
+      "loss": 0.8006,
+      "step": 2669
+    },
+    {
+      "epoch": 6.528117359413203,
+      "grad_norm": 0.22481452191041768,
+      "learning_rate": 3.0053697983541254e-06,
+      "loss": 0.6126,
+      "step": 2670
+    },
+    {
+      "epoch": 6.530562347188264,
+      "grad_norm": 0.2904064181232295,
+      "learning_rate": 3.004115969118365e-06,
+      "loss": 0.4795,
+      "step": 2671
+    },
+    {
+      "epoch": 6.533007334963325,
+      "grad_norm": 0.25087349970587886,
+      "learning_rate": 3.0028620076918586e-06,
+      "loss": 0.8884,
+      "step": 2672
+    },
+    {
+      "epoch": 6.5354523227383865,
+      "grad_norm": 0.27735085746983523,
+      "learning_rate": 3.0016079144034226e-06,
+      "loss": 0.5938,
+      "step": 2673
+    },
+    {
+      "epoch": 6.537897310513447,
+      "grad_norm": 0.1753704005772011,
+      "learning_rate": 3.0003536895819084e-06,
+      "loss": 0.3975,
+      "step": 2674
+    },
+    {
+      "epoch": 6.540342298288508,
+      "grad_norm": 0.1557245782344478,
+      "learning_rate": 2.9990993335562036e-06,
+      "loss": 0.4411,
+      "step": 2675
+    },
+    {
+      "epoch": 6.54278728606357,
+      "grad_norm": 0.26516177430613525,
+      "learning_rate": 2.997844846655227e-06,
+      "loss": 1.0124,
+      "step": 2676
+    },
+    {
+      "epoch": 6.54523227383863,
+      "grad_norm": 0.18843645144325918,
+      "learning_rate": 2.9965902292079375e-06,
+      "loss": 0.5905,
+      "step": 2677
+    },
+    {
+      "epoch": 6.547677261613692,
+      "grad_norm": 0.2047826022467107,
+      "learning_rate": 2.995335481543321e-06,
+      "loss": 0.4208,
+      "step": 2678
+    },
+    {
+      "epoch": 6.550122249388753,
+      "grad_norm": 0.19923272382698506,
+      "learning_rate": 2.9940806039904025e-06,
+      "loss": 0.4718,
+      "step": 2679
+    },
+    {
+      "epoch": 6.552567237163814,
+      "grad_norm": 0.2550964154261786,
+      "learning_rate": 2.9928255968782394e-06,
+      "loss": 0.5048,
+      "step": 2680
+    },
+    {
+      "epoch": 6.555012224938875,
+      "grad_norm": 0.22377564850326986,
+      "learning_rate": 2.991570460535923e-06,
+      "loss": 0.5758,
+      "step": 2681
+    },
+    {
+      "epoch": 6.557457212713937,
+      "grad_norm": 0.2311359389950381,
+      "learning_rate": 2.99031519529258e-06,
+      "loss": 0.7457,
+      "step": 2682
+    },
+    {
+      "epoch": 6.559902200488997,
+      "grad_norm": 0.17256223584885527,
+      "learning_rate": 2.989059801477369e-06,
+      "loss": 0.4341,
+      "step": 2683
+    },
+    {
+      "epoch": 6.5623471882640585,
+      "grad_norm": 0.14548126215494764,
+      "learning_rate": 2.9878042794194816e-06,
+      "loss": 0.4921,
+      "step": 2684
+    },
+    {
+      "epoch": 6.56479217603912,
+      "grad_norm": 0.24706978448397027,
+      "learning_rate": 2.986548629448146e-06,
+      "loss": 0.3941,
+      "step": 2685
+    },
+    {
+      "epoch": 6.56723716381418,
+      "grad_norm": 0.22201557159348928,
+      "learning_rate": 2.985292851892622e-06,
+      "loss": 0.8555,
+      "step": 2686
+    },
+    {
+      "epoch": 6.569682151589242,
+      "grad_norm": 0.2688686015455099,
+      "learning_rate": 2.9840369470822033e-06,
+      "loss": 0.951,
+      "step": 2687
+    },
+    {
+      "epoch": 6.572127139364303,
+      "grad_norm": 0.25251141761663465,
+      "learning_rate": 2.9827809153462175e-06,
+      "loss": 0.7463,
+      "step": 2688
+    },
+    {
+      "epoch": 6.574572127139365,
+      "grad_norm": 0.17171794332440718,
+      "learning_rate": 2.981524757014024e-06,
+      "loss": 0.7028,
+      "step": 2689
+    },
+    {
+      "epoch": 6.577017114914425,
+      "grad_norm": 0.3269105034715147,
+      "learning_rate": 2.9802684724150153e-06,
+      "loss": 0.5727,
+      "step": 2690
+    },
+    {
+      "epoch": 6.579462102689487,
+      "grad_norm": 0.2591005183102207,
+      "learning_rate": 2.9790120618786194e-06,
+      "loss": 0.8143,
+      "step": 2691
+    },
+    {
+      "epoch": 6.581907090464548,
+      "grad_norm": 0.20389425061484726,
+      "learning_rate": 2.9777555257342967e-06,
+      "loss": 0.7032,
+      "step": 2692
+    },
+    {
+      "epoch": 6.5843520782396086,
+      "grad_norm": 0.24826773708105962,
+      "learning_rate": 2.976498864311538e-06,
+      "loss": 0.4452,
+      "step": 2693
+    },
+    {
+      "epoch": 6.58679706601467,
+      "grad_norm": 0.24916078402016142,
+      "learning_rate": 2.97524207793987e-06,
+      "loss": 0.4132,
+      "step": 2694
+    },
+    {
+      "epoch": 6.589242053789731,
+      "grad_norm": 0.3055012626057449,
+      "learning_rate": 2.973985166948849e-06,
+      "loss": 0.5207,
+      "step": 2695
+    },
+    {
+      "epoch": 6.591687041564792,
+      "grad_norm": 0.20532255796825322,
+      "learning_rate": 2.972728131668068e-06,
+      "loss": 0.5616,
+      "step": 2696
+    },
+    {
+      "epoch": 6.594132029339853,
+      "grad_norm": 0.16931948927088916,
+      "learning_rate": 2.971470972427149e-06,
+      "loss": 0.4439,
+      "step": 2697
+    },
+    {
+      "epoch": 6.596577017114915,
+      "grad_norm": 0.224220677110465,
+      "learning_rate": 2.970213689555749e-06,
+      "loss": 0.4907,
+      "step": 2698
+    },
+    {
+      "epoch": 6.599022004889975,
+      "grad_norm": 0.23675718780016247,
+      "learning_rate": 2.9689562833835554e-06,
+      "loss": 0.4266,
+      "step": 2699
+    },
+    {
+      "epoch": 6.601466992665037,
+      "grad_norm": 0.2193748048340769,
+      "learning_rate": 2.967698754240289e-06,
+      "loss": 0.6999,
+      "step": 2700
+    },
+    {
+      "epoch": 6.603911980440098,
+      "grad_norm": 0.2623786958337752,
+      "learning_rate": 2.9664411024557028e-06,
+      "loss": 0.4871,
+      "step": 2701
+    },
+    {
+      "epoch": 6.606356968215159,
+      "grad_norm": 0.24164537885918347,
+      "learning_rate": 2.9651833283595825e-06,
+      "loss": 0.4935,
+      "step": 2702
+    },
+    {
+      "epoch": 6.60880195599022,
+      "grad_norm": 0.15721826415161416,
+      "learning_rate": 2.9639254322817452e-06,
+      "loss": 0.2632,
+      "step": 2703
+    },
+    {
+      "epoch": 6.611246943765281,
+      "grad_norm": 0.28990486519600583,
+      "learning_rate": 2.9626674145520395e-06,
+      "loss": 0.4998,
+      "step": 2704
+    },
+    {
+      "epoch": 6.613691931540342,
+      "grad_norm": 0.14627604416645817,
+      "learning_rate": 2.9614092755003458e-06,
+      "loss": 0.425,
+      "step": 2705
+    },
+    {
+      "epoch": 6.616136919315403,
+      "grad_norm": 0.21084199247646593,
+      "learning_rate": 2.960151015456578e-06,
+      "loss": 0.679,
+      "step": 2706
+    },
+    {
+      "epoch": 6.618581907090465,
+      "grad_norm": 0.23032614314046862,
+      "learning_rate": 2.958892634750681e-06,
+      "loss": 0.5144,
+      "step": 2707
+    },
+    {
+      "epoch": 6.621026894865525,
+      "grad_norm": 0.20662132597692687,
+      "learning_rate": 2.9576341337126307e-06,
+      "loss": 0.4933,
+      "step": 2708
+    },
+    {
+      "epoch": 6.623471882640587,
+      "grad_norm": 0.18756664168992057,
+      "learning_rate": 2.9563755126724348e-06,
+      "loss": 0.4756,
+      "step": 2709
+    },
+    {
+      "epoch": 6.625916870415648,
+      "grad_norm": 0.28438284606941416,
+      "learning_rate": 2.9551167719601316e-06,
+      "loss": 0.7255,
+      "step": 2710
+    },
+    {
+      "epoch": 6.628361858190709,
+      "grad_norm": 0.19967237692839773,
+      "learning_rate": 2.9538579119057926e-06,
+      "loss": 0.6223,
+      "step": 2711
+    },
+    {
+      "epoch": 6.63080684596577,
+      "grad_norm": 0.31763163723840154,
+      "learning_rate": 2.95259893283952e-06,
+      "loss": 0.3949,
+      "step": 2712
+    },
+    {
+      "epoch": 6.6332518337408315,
+      "grad_norm": 0.2155008705970982,
+      "learning_rate": 2.9513398350914467e-06,
+      "loss": 0.4276,
+      "step": 2713
+    },
+    {
+      "epoch": 6.635696821515893,
+      "grad_norm": 0.3625415938722578,
+      "learning_rate": 2.950080618991737e-06,
+      "loss": 0.4067,
+      "step": 2714
+    },
+    {
+      "epoch": 6.638141809290953,
+      "grad_norm": 0.2895888761224593,
+      "learning_rate": 2.9488212848705853e-06,
+      "loss": 0.6283,
+      "step": 2715
+    },
+    {
+      "epoch": 6.640586797066015,
+      "grad_norm": 0.26193784786323143,
+      "learning_rate": 2.947561833058219e-06,
+      "loss": 0.7917,
+      "step": 2716
+    },
+    {
+      "epoch": 6.643031784841076,
+      "grad_norm": 0.2492031825452439,
+      "learning_rate": 2.946302263884895e-06,
+      "loss": 0.461,
+      "step": 2717
+    },
+    {
+      "epoch": 6.645476772616137,
+      "grad_norm": 0.25541623107479616,
+      "learning_rate": 2.9450425776809005e-06,
+      "loss": 1.3095,
+      "step": 2718
+    },
+    {
+      "epoch": 6.647921760391198,
+      "grad_norm": 0.38052565682056416,
+      "learning_rate": 2.943782774776555e-06,
+      "loss": 0.456,
+      "step": 2719
+    },
+    {
+      "epoch": 6.65036674816626,
+      "grad_norm": 0.18461887494915855,
+      "learning_rate": 2.942522855502206e-06,
+      "loss": 0.5895,
+      "step": 2720
+    },
+    {
+      "epoch": 6.65281173594132,
+      "grad_norm": 0.2992363306762597,
+      "learning_rate": 2.941262820188236e-06,
+      "loss": 0.8049,
+      "step": 2721
+    },
+    {
+      "epoch": 6.6552567237163816,
+      "grad_norm": 0.30724322442935603,
+      "learning_rate": 2.9400026691650516e-06,
+      "loss": 0.6763,
+      "step": 2722
+    },
+    {
+      "epoch": 6.657701711491443,
+      "grad_norm": 0.21154868438804283,
+      "learning_rate": 2.938742402763097e-06,
+      "loss": 0.3707,
+      "step": 2723
+    },
+    {
+      "epoch": 6.6601466992665035,
+      "grad_norm": 0.19229092939292675,
+      "learning_rate": 2.9374820213128403e-06,
+      "loss": 0.5442,
+      "step": 2724
+    },
+    {
+      "epoch": 6.662591687041565,
+      "grad_norm": 0.2742948361715638,
+      "learning_rate": 2.9362215251447825e-06,
+      "loss": 0.4725,
+      "step": 2725
+    },
+    {
+      "epoch": 6.665036674816626,
+      "grad_norm": 0.3666432675646573,
+      "learning_rate": 2.9349609145894555e-06,
+      "loss": 0.6125,
+      "step": 2726
+    },
+    {
+      "epoch": 6.667481662591687,
+      "grad_norm": 0.27497653231406094,
+      "learning_rate": 2.9337001899774194e-06,
+      "loss": 0.5669,
+      "step": 2727
+    },
+    {
+      "epoch": 6.669926650366748,
+      "grad_norm": 0.2258595381231396,
+      "learning_rate": 2.932439351639266e-06,
+      "loss": 0.8942,
+      "step": 2728
+    },
+    {
+      "epoch": 6.67237163814181,
+      "grad_norm": 1.0147960895206174,
+      "learning_rate": 2.9311783999056155e-06,
+      "loss": 0.3725,
+      "step": 2729
+    },
+    {
+      "epoch": 6.67481662591687,
+      "grad_norm": 0.25218958077865056,
+      "learning_rate": 2.929917335107118e-06,
+      "loss": 0.5526,
+      "step": 2730
+    },
+    {
+      "epoch": 6.677261613691932,
+      "grad_norm": 0.16792298084885113,
+      "learning_rate": 2.9286561575744545e-06,
+      "loss": 0.5027,
+      "step": 2731
+    },
+    {
+      "epoch": 6.679706601466993,
+      "grad_norm": 0.19157012028497086,
+      "learning_rate": 2.9273948676383336e-06,
+      "loss": 0.7786,
+      "step": 2732
+    },
+    {
+      "epoch": 6.6821515892420535,
+      "grad_norm": 0.20042667939340975,
+      "learning_rate": 2.9261334656294937e-06,
+      "loss": 0.4554,
+      "step": 2733
+    },
+    {
+      "epoch": 6.684596577017115,
+      "grad_norm": 0.24652258489059,
+      "learning_rate": 2.9248719518787054e-06,
+      "loss": 0.7173,
+      "step": 2734
+    },
+    {
+      "epoch": 6.687041564792176,
+      "grad_norm": 0.28255033205159097,
+      "learning_rate": 2.9236103267167655e-06,
+      "loss": 0.5164,
+      "step": 2735
+    },
+    {
+      "epoch": 6.689486552567237,
+      "grad_norm": 0.2649261641607428,
+      "learning_rate": 2.922348590474501e-06,
+      "loss": 0.3937,
+      "step": 2736
+    },
+    {
+      "epoch": 6.691931540342298,
+      "grad_norm": 0.31930246367682547,
+      "learning_rate": 2.921086743482767e-06,
+      "loss": 0.4017,
+      "step": 2737
+    },
+    {
+      "epoch": 6.69437652811736,
+      "grad_norm": 0.2812279678445192,
+      "learning_rate": 2.9198247860724504e-06,
+      "loss": 0.8392,
+      "step": 2738
+    },
+    {
+      "epoch": 6.69682151589242,
+      "grad_norm": 0.18960881525836548,
+      "learning_rate": 2.918562718574464e-06,
+      "loss": 0.4895,
+      "step": 2739
+    },
+    {
+      "epoch": 6.699266503667482,
+      "grad_norm": 0.19709767991086352,
+      "learning_rate": 2.917300541319752e-06,
+      "loss": 0.4432,
+      "step": 2740
+    },
+    {
+      "epoch": 6.701711491442543,
+      "grad_norm": 0.19926816703391803,
+      "learning_rate": 2.916038254639285e-06,
+      "loss": 0.8889,
+      "step": 2741
+    },
+    {
+      "epoch": 6.704156479217604,
+      "grad_norm": 0.23519372874048605,
+      "learning_rate": 2.914775858864064e-06,
+      "loss": 0.3823,
+      "step": 2742
+    },
+    {
+      "epoch": 6.706601466992665,
+      "grad_norm": 0.16451325553688165,
+      "learning_rate": 2.913513354325119e-06,
+      "loss": 0.5356,
+      "step": 2743
+    },
+    {
+      "epoch": 6.709046454767726,
+      "grad_norm": 0.1727199848078932,
+      "learning_rate": 2.912250741353505e-06,
+      "loss": 0.589,
+      "step": 2744
+    },
+    {
+      "epoch": 6.711491442542787,
+      "grad_norm": 0.2781649841616548,
+      "learning_rate": 2.91098802028031e-06,
+      "loss": 0.4254,
+      "step": 2745
+    },
+    {
+      "epoch": 6.713936430317848,
+      "grad_norm": 0.2744545367492934,
+      "learning_rate": 2.9097251914366487e-06,
+      "loss": 0.8466,
+      "step": 2746
+    },
+    {
+      "epoch": 6.71638141809291,
+      "grad_norm": 0.21735279559568932,
+      "learning_rate": 2.908462255153663e-06,
+      "loss": 0.7528,
+      "step": 2747
+    },
+    {
+      "epoch": 6.71882640586797,
+      "grad_norm": 0.18385081154694746,
+      "learning_rate": 2.9071992117625237e-06,
+      "loss": 0.6961,
+      "step": 2748
+    },
+    {
+      "epoch": 6.721271393643032,
+      "grad_norm": 0.23424880872975262,
+      "learning_rate": 2.905936061594429e-06,
+      "loss": 0.8013,
+      "step": 2749
+    },
+    {
+      "epoch": 6.723716381418093,
+      "grad_norm": 0.370127315376921,
+      "learning_rate": 2.9046728049806066e-06,
+      "loss": 0.446,
+      "step": 2750
+    },
+    {
+      "epoch": 6.726161369193154,
+      "grad_norm": 0.19917553839152738,
+      "learning_rate": 2.903409442252312e-06,
+      "loss": 0.6816,
+      "step": 2751
+    },
+    {
+      "epoch": 6.728606356968215,
+      "grad_norm": 0.24988300606577307,
+      "learning_rate": 2.9021459737408264e-06,
+      "loss": 0.4542,
+      "step": 2752
+    },
+    {
+      "epoch": 6.7310513447432765,
+      "grad_norm": 0.198693501637806,
+      "learning_rate": 2.9008823997774617e-06,
+      "loss": 0.4508,
+      "step": 2753
+    },
+    {
+      "epoch": 6.733496332518337,
+      "grad_norm": 0.2987059254981822,
+      "learning_rate": 2.899618720693554e-06,
+      "loss": 0.7597,
+      "step": 2754
+    },
+    {
+      "epoch": 6.735941320293398,
+      "grad_norm": 0.16310346787961943,
+      "learning_rate": 2.89835493682047e-06,
+      "loss": 0.513,
+      "step": 2755
+    },
+    {
+      "epoch": 6.73838630806846,
+      "grad_norm": 0.2059609108865939,
+      "learning_rate": 2.897091048489603e-06,
+      "loss": 0.4991,
+      "step": 2756
+    },
+    {
+      "epoch": 6.74083129584352,
+      "grad_norm": 0.18402849371140545,
+      "learning_rate": 2.895827056032373e-06,
+      "loss": 0.3627,
+      "step": 2757
+    },
+    {
+      "epoch": 6.743276283618582,
+      "grad_norm": 0.24634234265554922,
+      "learning_rate": 2.894562959780228e-06,
+      "loss": 0.3881,
+      "step": 2758
+    },
+    {
+      "epoch": 6.745721271393643,
+      "grad_norm": 0.2990195768822749,
+      "learning_rate": 2.8932987600646426e-06,
+      "loss": 0.6247,
+      "step": 2759
+    },
+    {
+      "epoch": 6.748166259168704,
+      "grad_norm": 0.22955393709081237,
+      "learning_rate": 2.8920344572171195e-06,
+      "loss": 0.7099,
+      "step": 2760
+    },
+    {
+      "epoch": 6.750611246943765,
+      "grad_norm": 0.24466288292807217,
+      "learning_rate": 2.8907700515691877e-06,
+      "loss": 1.1852,
+      "step": 2761
+    },
+    {
+      "epoch": 6.7530562347188265,
+      "grad_norm": 0.30472795388821744,
+      "learning_rate": 2.8895055434524028e-06,
+      "loss": 0.4398,
+      "step": 2762
+    },
+    {
+      "epoch": 6.755501222493887,
+      "grad_norm": 0.20664302174119287,
+      "learning_rate": 2.8882409331983487e-06,
+      "loss": 0.8953,
+      "step": 2763
+    },
+    {
+      "epoch": 6.7579462102689485,
+      "grad_norm": 0.2328990996188546,
+      "learning_rate": 2.8869762211386335e-06,
+      "loss": 0.3735,
+      "step": 2764
+    },
+    {
+      "epoch": 6.76039119804401,
+      "grad_norm": 0.2732119444567348,
+      "learning_rate": 2.8857114076048948e-06,
+      "loss": 0.4616,
+      "step": 2765
+    },
+    {
+      "epoch": 6.76283618581907,
+      "grad_norm": 0.20187718409519675,
+      "learning_rate": 2.884446492928796e-06,
+      "loss": 0.5036,
+      "step": 2766
+    },
+    {
+      "epoch": 6.765281173594132,
+      "grad_norm": 0.2564015954416233,
+      "learning_rate": 2.8831814774420265e-06,
+      "loss": 0.6892,
+      "step": 2767
+    },
+    {
+      "epoch": 6.767726161369193,
+      "grad_norm": 0.36373327251497733,
+      "learning_rate": 2.8819163614763026e-06,
+      "loss": 0.4086,
+      "step": 2768
+    },
+    {
+      "epoch": 6.770171149144255,
+      "grad_norm": 0.33241102667606953,
+      "learning_rate": 2.8806511453633644e-06,
+      "loss": 0.5613,
+      "step": 2769
+    },
+    {
+      "epoch": 6.772616136919315,
+      "grad_norm": 0.21117446139519003,
+      "learning_rate": 2.879385829434983e-06,
+      "loss": 0.6965,
+      "step": 2770
+    },
+    {
+      "epoch": 6.775061124694377,
+      "grad_norm": 0.19735893864714288,
+      "learning_rate": 2.878120414022953e-06,
+      "loss": 0.4242,
+      "step": 2771
+    },
+    {
+      "epoch": 6.777506112469438,
+      "grad_norm": 0.2629117529727301,
+      "learning_rate": 2.8768548994590943e-06,
+      "loss": 0.6384,
+      "step": 2772
+    },
+    {
+      "epoch": 6.7799511002444985,
+      "grad_norm": 0.2816011177731885,
+      "learning_rate": 2.875589286075255e-06,
+      "loss": 0.4578,
+      "step": 2773
+    },
+    {
+      "epoch": 6.78239608801956,
+      "grad_norm": 0.3300256350434343,
+      "learning_rate": 2.874323574203306e-06,
+      "loss": 0.4178,
+      "step": 2774
+    },
+    {
+      "epoch": 6.784841075794621,
+      "grad_norm": 0.20584583164061476,
+      "learning_rate": 2.873057764175148e-06,
+      "loss": 0.9352,
+      "step": 2775
+    },
+    {
+      "epoch": 6.787286063569682,
+      "grad_norm": 0.4074261674837491,
+      "learning_rate": 2.8717918563227033e-06,
+      "loss": 0.3981,
+      "step": 2776
+    },
+    {
+      "epoch": 6.789731051344743,
+      "grad_norm": 0.25685936415912963,
+      "learning_rate": 2.870525850977924e-06,
+      "loss": 0.7321,
+      "step": 2777
+    },
+    {
+      "epoch": 6.792176039119805,
+      "grad_norm": 0.24280540137406426,
+      "learning_rate": 2.8692597484727845e-06,
+      "loss": 0.4047,
+      "step": 2778
+    },
+    {
+      "epoch": 6.794621026894865,
+      "grad_norm": 0.29043875898454935,
+      "learning_rate": 2.8679935491392858e-06,
+      "loss": 0.4859,
+      "step": 2779
+    },
+    {
+      "epoch": 6.797066014669927,
+      "grad_norm": 0.164345140183799,
+      "learning_rate": 2.8667272533094547e-06,
+      "loss": 0.8862,
+      "step": 2780
+    },
+    {
+      "epoch": 6.799511002444988,
+      "grad_norm": 0.21405966188164813,
+      "learning_rate": 2.865460861315343e-06,
+      "loss": 0.6696,
+      "step": 2781
+    },
+    {
+      "epoch": 6.801955990220049,
+      "grad_norm": 0.29081377205013303,
+      "learning_rate": 2.8641943734890275e-06,
+      "loss": 0.4452,
+      "step": 2782
+    },
+    {
+      "epoch": 6.80440097799511,
+      "grad_norm": 0.24536436426114933,
+      "learning_rate": 2.862927790162611e-06,
+      "loss": 0.4858,
+      "step": 2783
+    },
+    {
+      "epoch": 6.806845965770171,
+      "grad_norm": 0.3157539307182182,
+      "learning_rate": 2.8616611116682193e-06,
+      "loss": 0.4516,
+      "step": 2784
+    },
+    {
+      "epoch": 6.809290953545232,
+      "grad_norm": 0.2031114985673134,
+      "learning_rate": 2.8603943383380057e-06,
+      "loss": 0.8251,
+      "step": 2785
+    },
+    {
+      "epoch": 6.811735941320293,
+      "grad_norm": 0.279626524171972,
+      "learning_rate": 2.8591274705041453e-06,
+      "loss": 1.0458,
+      "step": 2786
+    },
+    {
+      "epoch": 6.814180929095355,
+      "grad_norm": 0.3242376603844505,
+      "learning_rate": 2.857860508498843e-06,
+      "loss": 0.7359,
+      "step": 2787
+    },
+    {
+      "epoch": 6.816625916870415,
+      "grad_norm": 0.20884423801849686,
+      "learning_rate": 2.8565934526543226e-06,
+      "loss": 0.68,
+      "step": 2788
+    },
+    {
+      "epoch": 6.819070904645477,
+      "grad_norm": 0.16276661716179386,
+      "learning_rate": 2.855326303302836e-06,
+      "loss": 0.5919,
+      "step": 2789
+    },
+    {
+      "epoch": 6.821515892420538,
+      "grad_norm": 0.24677520783481247,
+      "learning_rate": 2.8540590607766594e-06,
+      "loss": 0.5566,
+      "step": 2790
+    },
+    {
+      "epoch": 6.823960880195599,
+      "grad_norm": 0.16615301463507698,
+      "learning_rate": 2.8527917254080916e-06,
+      "loss": 0.4758,
+      "step": 2791
+    },
+    {
+      "epoch": 6.82640586797066,
+      "grad_norm": 0.2095359279365771,
+      "learning_rate": 2.851524297529457e-06,
+      "loss": 0.3986,
+      "step": 2792
+    },
+    {
+      "epoch": 6.8288508557457215,
+      "grad_norm": 0.27380082906367376,
+      "learning_rate": 2.850256777473106e-06,
+      "loss": 0.2664,
+      "step": 2793
+    },
+    {
+      "epoch": 6.831295843520783,
+      "grad_norm": 0.252485887655719,
+      "learning_rate": 2.848989165571409e-06,
+      "loss": 0.3373,
+      "step": 2794
+    },
+    {
+      "epoch": 6.833740831295843,
+      "grad_norm": 0.2752948505743244,
+      "learning_rate": 2.8477214621567646e-06,
+      "loss": 1.5643,
+      "step": 2795
+    },
+    {
+      "epoch": 6.836185819070905,
+      "grad_norm": 0.32543176822072833,
+      "learning_rate": 2.846453667561593e-06,
+      "loss": 0.5343,
+      "step": 2796
+    },
+    {
+      "epoch": 6.838630806845966,
+      "grad_norm": 0.3075030356717426,
+      "learning_rate": 2.8451857821183383e-06,
+      "loss": 0.9246,
+      "step": 2797
+    },
+    {
+      "epoch": 6.841075794621027,
+      "grad_norm": 0.21719772656087993,
+      "learning_rate": 2.84391780615947e-06,
+      "loss": 0.4434,
+      "step": 2798
+    },
+    {
+      "epoch": 6.843520782396088,
+      "grad_norm": 0.23325966516724952,
+      "learning_rate": 2.8426497400174803e-06,
+      "loss": 0.7113,
+      "step": 2799
+    },
+    {
+      "epoch": 6.84596577017115,
+      "grad_norm": 0.21647800007884022,
+      "learning_rate": 2.8413815840248855e-06,
+      "loss": 0.7238,
+      "step": 2800
+    },
+    {
+      "epoch": 6.84841075794621,
+      "grad_norm": 0.27048519433463364,
+      "learning_rate": 2.8401133385142244e-06,
+      "loss": 0.4382,
+      "step": 2801
+    },
+    {
+      "epoch": 6.8508557457212715,
+      "grad_norm": 0.18601097103292863,
+      "learning_rate": 2.8388450038180603e-06,
+      "loss": 0.2983,
+      "step": 2802
+    },
+    {
+      "epoch": 6.853300733496333,
+      "grad_norm": 0.21018733614206458,
+      "learning_rate": 2.837576580268979e-06,
+      "loss": 0.5112,
+      "step": 2803
+    },
+    {
+      "epoch": 6.8557457212713935,
+      "grad_norm": 0.2577631653028338,
+      "learning_rate": 2.836308068199591e-06,
+      "loss": 0.5394,
+      "step": 2804
+    },
+    {
+      "epoch": 6.858190709046455,
+      "grad_norm": 0.2606307959223212,
+      "learning_rate": 2.8350394679425292e-06,
+      "loss": 0.4073,
+      "step": 2805
+    },
+    {
+      "epoch": 6.860635696821516,
+      "grad_norm": 0.21119054002772333,
+      "learning_rate": 2.8337707798304496e-06,
+      "loss": 0.6045,
+      "step": 2806
+    },
+    {
+      "epoch": 6.863080684596577,
+      "grad_norm": 0.18527861172121762,
+      "learning_rate": 2.832502004196031e-06,
+      "loss": 0.6509,
+      "step": 2807
+    },
+    {
+      "epoch": 6.865525672371638,
+      "grad_norm": 0.4518854001438533,
+      "learning_rate": 2.8312331413719745e-06,
+      "loss": 0.3333,
+      "step": 2808
+    },
+    {
+      "epoch": 6.8679706601467,
+      "grad_norm": 0.2742515820423122,
+      "learning_rate": 2.829964191691007e-06,
+      "loss": 0.427,
+      "step": 2809
+    },
+    {
+      "epoch": 6.87041564792176,
+      "grad_norm": 0.27879854249861935,
+      "learning_rate": 2.828695155485875e-06,
+      "loss": 0.5961,
+      "step": 2810
+    },
+    {
+      "epoch": 6.872860635696822,
+      "grad_norm": 0.19661120593022052,
+      "learning_rate": 2.827426033089349e-06,
+      "loss": 0.3912,
+      "step": 2811
+    },
+    {
+      "epoch": 6.875305623471883,
+      "grad_norm": 0.20068993321178927,
+      "learning_rate": 2.826156824834222e-06,
+      "loss": 0.5391,
+      "step": 2812
+    },
+    {
+      "epoch": 6.8777506112469435,
+      "grad_norm": 0.18969871197521632,
+      "learning_rate": 2.824887531053308e-06,
+      "loss": 0.4676,
+      "step": 2813
+    },
+    {
+      "epoch": 6.880195599022005,
+      "grad_norm": 0.20889441044113433,
+      "learning_rate": 2.8236181520794476e-06,
+      "loss": 0.5877,
+      "step": 2814
+    },
+    {
+      "epoch": 6.882640586797066,
+      "grad_norm": 0.2364876938075912,
+      "learning_rate": 2.8223486882455e-06,
+      "loss": 0.6308,
+      "step": 2815
+    },
+    {
+      "epoch": 6.885085574572127,
+      "grad_norm": 0.25701925051803837,
+      "learning_rate": 2.8210791398843462e-06,
+      "loss": 0.4281,
+      "step": 2816
+    },
+    {
+      "epoch": 6.887530562347188,
+      "grad_norm": 0.19670853692453427,
+      "learning_rate": 2.8198095073288927e-06,
+      "loss": 0.3746,
+      "step": 2817
+    },
+    {
+      "epoch": 6.88997555012225,
+      "grad_norm": 0.18468337424437048,
+      "learning_rate": 2.8185397909120647e-06,
+      "loss": 0.4062,
+      "step": 2818
+    },
+    {
+      "epoch": 6.89242053789731,
+      "grad_norm": 0.26482491036256023,
+      "learning_rate": 2.817269990966812e-06,
+      "loss": 0.8197,
+      "step": 2819
+    },
+    {
+      "epoch": 6.894865525672372,
+      "grad_norm": 0.296123128878938,
+      "learning_rate": 2.8160001078261056e-06,
+      "loss": 0.663,
+      "step": 2820
+    },
+    {
+      "epoch": 6.897310513447433,
+      "grad_norm": 0.323392070416714,
+      "learning_rate": 2.8147301418229373e-06,
+      "loss": 0.3726,
+      "step": 2821
+    },
+    {
+      "epoch": 6.899755501222494,
+      "grad_norm": 0.2458417002010351,
+      "learning_rate": 2.813460093290321e-06,
+      "loss": 0.9851,
+      "step": 2822
+    },
+    {
+      "epoch": 6.902200488997555,
+      "grad_norm": 0.21911391718046638,
+      "learning_rate": 2.812189962561292e-06,
+      "loss": 0.4563,
+      "step": 2823
+    },
+    {
+      "epoch": 6.904645476772616,
+      "grad_norm": 0.2704856297046878,
+      "learning_rate": 2.8109197499689083e-06,
+      "loss": 0.5959,
+      "step": 2824
+    },
+    {
+      "epoch": 6.907090464547677,
+      "grad_norm": 0.2520313035666214,
+      "learning_rate": 2.80964945584625e-06,
+      "loss": 0.5115,
+      "step": 2825
+    },
+    {
+      "epoch": 6.909535452322738,
+      "grad_norm": 0.25291875628267846,
+      "learning_rate": 2.8083790805264153e-06,
+      "loss": 0.4304,
+      "step": 2826
+    },
+    {
+      "epoch": 6.9119804400978,
+      "grad_norm": 0.16666761708475025,
+      "learning_rate": 2.807108624342527e-06,
+      "loss": 0.5669,
+      "step": 2827
+    },
+    {
+      "epoch": 6.91442542787286,
+      "grad_norm": 0.17386870018651085,
+      "learning_rate": 2.8058380876277265e-06,
+      "loss": 0.5577,
+      "step": 2828
+    },
+    {
+      "epoch": 6.916870415647922,
+      "grad_norm": 0.17696185295288228,
+      "learning_rate": 2.804567470715178e-06,
+      "loss": 0.5349,
+      "step": 2829
+    },
+    {
+      "epoch": 6.919315403422983,
+      "grad_norm": 0.20857481435771388,
+      "learning_rate": 2.8032967739380678e-06,
+      "loss": 0.8764,
+      "step": 2830
+    },
+    {
+      "epoch": 6.921760391198044,
+      "grad_norm": 0.279417430575681,
+      "learning_rate": 2.8020259976296006e-06,
+      "loss": 0.7464,
+      "step": 2831
+    },
+    {
+      "epoch": 6.924205378973105,
+      "grad_norm": 0.3099109123344021,
+      "learning_rate": 2.800755142123003e-06,
+      "loss": 0.6357,
+      "step": 2832
+    },
+    {
+      "epoch": 6.9266503667481665,
+      "grad_norm": 0.16070560251874458,
+      "learning_rate": 2.7994842077515216e-06,
+      "loss": 0.5408,
+      "step": 2833
+    },
+    {
+      "epoch": 6.929095354523227,
+      "grad_norm": 0.23790132839621222,
+      "learning_rate": 2.798213194848426e-06,
+      "loss": 0.3719,
+      "step": 2834
+    },
+    {
+      "epoch": 6.931540342298288,
+      "grad_norm": 0.1416649488690476,
+      "learning_rate": 2.7969421037470037e-06,
+      "loss": 0.4203,
+      "step": 2835
+    },
+    {
+      "epoch": 6.93398533007335,
+      "grad_norm": 0.16597424079124218,
+      "learning_rate": 2.795670934780565e-06,
+      "loss": 0.2303,
+      "step": 2836
+    },
+    {
+      "epoch": 6.93643031784841,
+      "grad_norm": 0.17889120704636863,
+      "learning_rate": 2.794399688282438e-06,
+      "loss": 0.9721,
+      "step": 2837
+    },
+    {
+      "epoch": 6.938875305623472,
+      "grad_norm": 0.2186810405583919,
+      "learning_rate": 2.793128364585974e-06,
+      "loss": 0.5193,
+      "step": 2838
+    },
+    {
+      "epoch": 6.941320293398533,
+      "grad_norm": 0.3560748118375034,
+      "learning_rate": 2.7918569640245437e-06,
+      "loss": 0.5849,
+      "step": 2839
+    },
+    {
+      "epoch": 6.943765281173594,
+      "grad_norm": 0.2502336209852303,
+      "learning_rate": 2.7905854869315347e-06,
+      "loss": 0.6671,
+      "step": 2840
+    },
+    {
+      "epoch": 6.946210268948655,
+      "grad_norm": 0.3532463416050091,
+      "learning_rate": 2.78931393364036e-06,
+      "loss": 0.5823,
+      "step": 2841
+    },
+    {
+      "epoch": 6.9486552567237165,
+      "grad_norm": 0.18908993704629637,
+      "learning_rate": 2.788042304484449e-06,
+      "loss": 0.6043,
+      "step": 2842
+    },
+    {
+      "epoch": 6.951100244498777,
+      "grad_norm": 0.16310714500755577,
+      "learning_rate": 2.7867705997972523e-06,
+      "loss": 0.6818,
+      "step": 2843
+    },
+    {
+      "epoch": 6.9535452322738385,
+      "grad_norm": 0.24819913839430116,
+      "learning_rate": 2.78549881991224e-06,
+      "loss": 0.9136,
+      "step": 2844
+    },
+    {
+      "epoch": 6.9559902200489,
+      "grad_norm": 0.1912671282555043,
+      "learning_rate": 2.7842269651629008e-06,
+      "loss": 0.3358,
+      "step": 2845
+    },
+    {
+      "epoch": 6.95843520782396,
+      "grad_norm": 0.23391566127328417,
+      "learning_rate": 2.7829550358827455e-06,
+      "loss": 0.7382,
+      "step": 2846
+    },
+    {
+      "epoch": 6.960880195599022,
+      "grad_norm": 0.26613179607738646,
+      "learning_rate": 2.7816830324053026e-06,
+      "loss": 0.4464,
+      "step": 2847
+    },
+    {
+      "epoch": 6.963325183374083,
+      "grad_norm": 0.2181186003855521,
+      "learning_rate": 2.7804109550641205e-06,
+      "loss": 0.4594,
+      "step": 2848
+    },
+    {
+      "epoch": 6.965770171149144,
+      "grad_norm": 0.18790626434449093,
+      "learning_rate": 2.779138804192767e-06,
+      "loss": 0.877,
+      "step": 2849
+    },
+    {
+      "epoch": 6.968215158924205,
+      "grad_norm": 0.24991768155867214,
+      "learning_rate": 2.7778665801248294e-06,
+      "loss": 0.4746,
+      "step": 2850
+    },
+    {
+      "epoch": 6.970660146699267,
+      "grad_norm": 0.1995437717850146,
+      "learning_rate": 2.776594283193913e-06,
+      "loss": 0.6947,
+      "step": 2851
+    },
+    {
+      "epoch": 6.973105134474328,
+      "grad_norm": 0.21548287625896684,
+      "learning_rate": 2.775321913733644e-06,
+      "loss": 0.7359,
+      "step": 2852
+    },
+    {
+      "epoch": 6.9755501222493885,
+      "grad_norm": 0.2600861711723258,
+      "learning_rate": 2.7740494720776673e-06,
+      "loss": 0.4591,
+      "step": 2853
+    },
+    {
+      "epoch": 6.97799511002445,
+      "grad_norm": 0.19812222263502807,
+      "learning_rate": 2.772776958559646e-06,
+      "loss": 0.7522,
+      "step": 2854
+    },
+    {
+      "epoch": 6.980440097799511,
+      "grad_norm": 0.2274583461427906,
+      "learning_rate": 2.7715043735132615e-06,
+      "loss": 0.4138,
+      "step": 2855
+    },
+    {
+      "epoch": 6.982885085574572,
+      "grad_norm": 0.19302459766376526,
+      "learning_rate": 2.7702317172722143e-06,
+      "loss": 0.4676,
+      "step": 2856
+    },
+    {
+      "epoch": 6.985330073349633,
+      "grad_norm": 0.2437648146181491,
+      "learning_rate": 2.768958990170225e-06,
+      "loss": 0.4823,
+      "step": 2857
+    },
+    {
+      "epoch": 6.987775061124695,
+      "grad_norm": 0.2844858934455108,
+      "learning_rate": 2.7676861925410326e-06,
+      "loss": 0.3401,
+      "step": 2858
+    },
+    {
+      "epoch": 6.990220048899755,
+      "grad_norm": 0.2947085837465676,
+      "learning_rate": 2.7664133247183927e-06,
+      "loss": 0.6423,
+      "step": 2859
+    },
+    {
+      "epoch": 6.992665036674817,
+      "grad_norm": 0.28026310584723535,
+      "learning_rate": 2.7651403870360803e-06,
+      "loss": 0.6117,
+      "step": 2860
+    },
+    {
+      "epoch": 6.995110024449878,
+      "grad_norm": 0.16974081277840003,
+      "learning_rate": 2.7638673798278894e-06,
+      "loss": 0.8068,
+      "step": 2861
+    },
+    {
+      "epoch": 6.997555012224939,
+      "grad_norm": 0.25464037365789094,
+      "learning_rate": 2.76259430342763e-06,
+      "loss": 0.5885,
+      "step": 2862
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.16322625467003435,
+      "learning_rate": 2.761321158169134e-06,
+      "loss": 0.418,
+      "step": 2863
+    },
+    {
+      "epoch": 7.002444987775061,
+      "grad_norm": 0.2803745876184196,
+      "learning_rate": 2.7600479443862484e-06,
+      "loss": 0.5407,
+      "step": 2864
+    },
+    {
+      "epoch": 7.004889975550122,
+      "grad_norm": 0.23088286437399547,
+      "learning_rate": 2.7587746624128387e-06,
+      "loss": 0.7657,
+      "step": 2865
+    },
+    {
+      "epoch": 7.007334963325183,
+      "grad_norm": 0.211075860092187,
+      "learning_rate": 2.7575013125827887e-06,
+      "loss": 0.6514,
+      "step": 2866
+    },
+    {
+      "epoch": 7.009779951100245,
+      "grad_norm": 0.22747524363771285,
+      "learning_rate": 2.75622789523e-06,
+      "loss": 0.6311,
+      "step": 2867
+    },
+    {
+      "epoch": 7.012224938875305,
+      "grad_norm": 0.26568755421704543,
+      "learning_rate": 2.754954410688392e-06,
+      "loss": 0.9956,
+      "step": 2868
+    },
+    {
+      "epoch": 7.014669926650367,
+      "grad_norm": 0.26519804678688574,
+      "learning_rate": 2.7536808592919012e-06,
+      "loss": 0.5011,
+      "step": 2869
+    },
+    {
+      "epoch": 7.017114914425428,
+      "grad_norm": 0.2130430904612387,
+      "learning_rate": 2.7524072413744817e-06,
+      "loss": 1.2497,
+      "step": 2870
+    },
+    {
+      "epoch": 7.019559902200489,
+      "grad_norm": 0.22564891452350325,
+      "learning_rate": 2.751133557270106e-06,
+      "loss": 0.8184,
+      "step": 2871
+    },
+    {
+      "epoch": 7.02200488997555,
+      "grad_norm": 0.3505898093982352,
+      "learning_rate": 2.749859807312762e-06,
+      "loss": 0.7188,
+      "step": 2872
+    },
+    {
+      "epoch": 7.0244498777506115,
+      "grad_norm": 0.15052571682476995,
+      "learning_rate": 2.7485859918364577e-06,
+      "loss": 0.3633,
+      "step": 2873
+    },
+    {
+      "epoch": 7.026894865525672,
+      "grad_norm": 0.24242577175350447,
+      "learning_rate": 2.7473121111752155e-06,
+      "loss": 0.4543,
+      "step": 2874
+    },
+    {
+      "epoch": 7.029339853300733,
+      "grad_norm": 0.2626842315138586,
+      "learning_rate": 2.7460381656630768e-06,
+      "loss": 0.7194,
+      "step": 2875
+    },
+    {
+      "epoch": 7.031784841075795,
+      "grad_norm": 0.2732009367635997,
+      "learning_rate": 2.7447641556340987e-06,
+      "loss": 0.446,
+      "step": 2876
+    },
+    {
+      "epoch": 7.034229828850855,
+      "grad_norm": 0.18694373464323058,
+      "learning_rate": 2.743490081422356e-06,
+      "loss": 0.5515,
+      "step": 2877
+    },
+    {
+      "epoch": 7.036674816625917,
+      "grad_norm": 0.307523530491107,
+      "learning_rate": 2.74221594336194e-06,
+      "loss": 0.7568,
+      "step": 2878
+    },
+    {
+      "epoch": 7.039119804400978,
+      "grad_norm": 0.18285192211744286,
+      "learning_rate": 2.740941741786959e-06,
+      "loss": 0.5321,
+      "step": 2879
+    },
+    {
+      "epoch": 7.041564792176039,
+      "grad_norm": 0.2691301659888327,
+      "learning_rate": 2.739667477031538e-06,
+      "loss": 0.9386,
+      "step": 2880
+    },
+    {
+      "epoch": 7.0440097799511,
+      "grad_norm": 0.2584152581453055,
+      "learning_rate": 2.738393149429818e-06,
+      "loss": 0.5119,
+      "step": 2881
+    },
+    {
+      "epoch": 7.0464547677261615,
+      "grad_norm": 0.19308514640073143,
+      "learning_rate": 2.7371187593159577e-06,
+      "loss": 0.3915,
+      "step": 2882
+    },
+    {
+      "epoch": 7.048899755501223,
+      "grad_norm": 0.27246911867563556,
+      "learning_rate": 2.73584430702413e-06,
+      "loss": 0.4662,
+      "step": 2883
+    },
+    {
+      "epoch": 7.0513447432762835,
+      "grad_norm": 0.18812136140108413,
+      "learning_rate": 2.7345697928885263e-06,
+      "loss": 0.4338,
+      "step": 2884
+    },
+    {
+      "epoch": 7.053789731051345,
+      "grad_norm": 0.22176889392026958,
+      "learning_rate": 2.7332952172433535e-06,
+      "loss": 0.6034,
+      "step": 2885
+    },
+    {
+      "epoch": 7.056234718826406,
+      "grad_norm": 0.4688783208501427,
+      "learning_rate": 2.7320205804228337e-06,
+      "loss": 0.4336,
+      "step": 2886
+    },
+    {
+      "epoch": 7.058679706601467,
+      "grad_norm": 0.16761452844997968,
+      "learning_rate": 2.7307458827612075e-06,
+      "loss": 0.625,
+      "step": 2887
+    },
+    {
+      "epoch": 7.061124694376528,
+      "grad_norm": 0.33137347259019007,
+      "learning_rate": 2.729471124592728e-06,
+      "loss": 0.5385,
+      "step": 2888
+    },
+    {
+      "epoch": 7.06356968215159,
+      "grad_norm": 0.2702831926650596,
+      "learning_rate": 2.728196306251667e-06,
+      "loss": 1.0968,
+      "step": 2889
+    },
+    {
+      "epoch": 7.06601466992665,
+      "grad_norm": 0.2289703240948399,
+      "learning_rate": 2.726921428072311e-06,
+      "loss": 0.5042,
+      "step": 2890
+    },
+    {
+      "epoch": 7.068459657701712,
+      "grad_norm": 0.15061584641231482,
+      "learning_rate": 2.7256464903889614e-06,
+      "loss": 0.6853,
+      "step": 2891
+    },
+    {
+      "epoch": 7.070904645476773,
+      "grad_norm": 0.25300586316964646,
+      "learning_rate": 2.724371493535937e-06,
+      "loss": 0.4191,
+      "step": 2892
+    },
+    {
+      "epoch": 7.0733496332518335,
+      "grad_norm": 0.21906541187270398,
+      "learning_rate": 2.723096437847571e-06,
+      "loss": 0.8312,
+      "step": 2893
+    },
+    {
+      "epoch": 7.075794621026895,
+      "grad_norm": 0.19925159206363205,
+      "learning_rate": 2.7218213236582114e-06,
+      "loss": 0.7383,
+      "step": 2894
+    },
+    {
+      "epoch": 7.078239608801956,
+      "grad_norm": 0.31152798439838525,
+      "learning_rate": 2.7205461513022237e-06,
+      "loss": 0.4381,
+      "step": 2895
+    },
+    {
+      "epoch": 7.080684596577017,
+      "grad_norm": 0.3486666521629648,
+      "learning_rate": 2.7192709211139855e-06,
+      "loss": 0.3554,
+      "step": 2896
+    },
+    {
+      "epoch": 7.083129584352078,
+      "grad_norm": 0.26358769216485556,
+      "learning_rate": 2.7179956334278933e-06,
+      "loss": 0.6358,
+      "step": 2897
+    },
+    {
+      "epoch": 7.08557457212714,
+      "grad_norm": 0.27558719539655036,
+      "learning_rate": 2.7167202885783563e-06,
+      "loss": 0.4475,
+      "step": 2898
+    },
+    {
+      "epoch": 7.0880195599022,
+      "grad_norm": 0.13034945863927172,
+      "learning_rate": 2.7154448868997974e-06,
+      "loss": 0.3353,
+      "step": 2899
+    },
+    {
+      "epoch": 7.090464547677262,
+      "grad_norm": 0.21670235433208634,
+      "learning_rate": 2.714169428726658e-06,
+      "loss": 0.4382,
+      "step": 2900
+    },
+    {
+      "epoch": 7.092909535452323,
+      "grad_norm": 0.16897747468006827,
+      "learning_rate": 2.7128939143933914e-06,
+      "loss": 0.5487,
+      "step": 2901
+    },
+    {
+      "epoch": 7.095354523227384,
+      "grad_norm": 0.26073954062784044,
+      "learning_rate": 2.7116183442344672e-06,
+      "loss": 0.9542,
+      "step": 2902
+    },
+    {
+      "epoch": 7.097799511002445,
+      "grad_norm": 0.3093657401425847,
+      "learning_rate": 2.7103427185843687e-06,
+      "loss": 0.4449,
+      "step": 2903
+    },
+    {
+      "epoch": 7.100244498777506,
+      "grad_norm": 0.2155042553841293,
+      "learning_rate": 2.7090670377775947e-06,
+      "loss": 0.6862,
+      "step": 2904
+    },
+    {
+      "epoch": 7.102689486552567,
+      "grad_norm": 0.19141219379610852,
+      "learning_rate": 2.7077913021486574e-06,
+      "loss": 0.5114,
+      "step": 2905
+    },
+    {
+      "epoch": 7.105134474327628,
+      "grad_norm": 0.23082767452471636,
+      "learning_rate": 2.706515512032084e-06,
+      "loss": 0.5656,
+      "step": 2906
+    },
+    {
+      "epoch": 7.10757946210269,
+      "grad_norm": 0.25779437907799657,
+      "learning_rate": 2.705239667762416e-06,
+      "loss": 0.5559,
+      "step": 2907
+    },
+    {
+      "epoch": 7.11002444987775,
+      "grad_norm": 0.2695268250447425,
+      "learning_rate": 2.7039637696742095e-06,
+      "loss": 0.4092,
+      "step": 2908
+    },
+    {
+      "epoch": 7.112469437652812,
+      "grad_norm": 0.23132305638522085,
+      "learning_rate": 2.7026878181020328e-06,
+      "loss": 0.4901,
+      "step": 2909
+    },
+    {
+      "epoch": 7.114914425427873,
+      "grad_norm": 0.2665686056791714,
+      "learning_rate": 2.70141181338047e-06,
+      "loss": 0.3599,
+      "step": 2910
+    },
+    {
+      "epoch": 7.117359413202934,
+      "grad_norm": 0.1679841045410817,
+      "learning_rate": 2.700135755844119e-06,
+      "loss": 0.584,
+      "step": 2911
+    },
+    {
+      "epoch": 7.119804400977995,
+      "grad_norm": 0.2558683853890477,
+      "learning_rate": 2.698859645827593e-06,
+      "loss": 0.4002,
+      "step": 2912
+    },
+    {
+      "epoch": 7.1222493887530565,
+      "grad_norm": 0.30949856081948834,
+      "learning_rate": 2.6975834836655148e-06,
+      "loss": 0.4174,
+      "step": 2913
+    },
+    {
+      "epoch": 7.124694376528117,
+      "grad_norm": 0.3680640146746572,
+      "learning_rate": 2.6963072696925247e-06,
+      "loss": 0.3545,
+      "step": 2914
+    },
+    {
+      "epoch": 7.127139364303178,
+      "grad_norm": 0.2821535466417029,
+      "learning_rate": 2.6950310042432736e-06,
+      "loss": 0.8195,
+      "step": 2915
+    },
+    {
+      "epoch": 7.12958435207824,
+      "grad_norm": 0.32110940189330334,
+      "learning_rate": 2.693754687652429e-06,
+      "loss": 0.495,
+      "step": 2916
+    },
+    {
+      "epoch": 7.1320293398533,
+      "grad_norm": 0.27139841801002657,
+      "learning_rate": 2.692478320254671e-06,
+      "loss": 0.7873,
+      "step": 2917
+    },
+    {
+      "epoch": 7.134474327628362,
+      "grad_norm": 0.3627069503060499,
+      "learning_rate": 2.6912019023846907e-06,
+      "loss": 0.8211,
+      "step": 2918
+    },
+    {
+      "epoch": 7.136919315403423,
+      "grad_norm": 0.2770772859573193,
+      "learning_rate": 2.689925434377195e-06,
+      "loss": 0.6837,
+      "step": 2919
+    },
+    {
+      "epoch": 7.139364303178484,
+      "grad_norm": 0.22039436161782777,
+      "learning_rate": 2.688648916566902e-06,
+      "loss": 0.8853,
+      "step": 2920
+    },
+    {
+      "epoch": 7.141809290953545,
+      "grad_norm": 0.2875204119327131,
+      "learning_rate": 2.6873723492885446e-06,
+      "loss": 0.5072,
+      "step": 2921
+    },
+    {
+      "epoch": 7.1442542787286065,
+      "grad_norm": 0.2028580979142081,
+      "learning_rate": 2.686095732876869e-06,
+      "loss": 0.511,
+      "step": 2922
+    },
+    {
+      "epoch": 7.146699266503667,
+      "grad_norm": 0.22505244739592664,
+      "learning_rate": 2.684819067666632e-06,
+      "loss": 0.6521,
+      "step": 2923
+    },
+    {
+      "epoch": 7.1491442542787285,
+      "grad_norm": 0.2213350866866343,
+      "learning_rate": 2.683542353992604e-06,
+      "loss": 0.5769,
+      "step": 2924
+    },
+    {
+      "epoch": 7.15158924205379,
+      "grad_norm": 0.33583246074464174,
+      "learning_rate": 2.6822655921895696e-06,
+      "loss": 0.6502,
+      "step": 2925
+    },
+    {
+      "epoch": 7.15403422982885,
+      "grad_norm": 0.30042754931647303,
+      "learning_rate": 2.6809887825923246e-06,
+      "loss": 0.5514,
+      "step": 2926
+    },
+    {
+      "epoch": 7.156479217603912,
+      "grad_norm": 0.31121705307781394,
+      "learning_rate": 2.679711925535678e-06,
+      "loss": 0.3817,
+      "step": 2927
+    },
+    {
+      "epoch": 7.158924205378973,
+      "grad_norm": 0.24563668641683023,
+      "learning_rate": 2.6784350213544506e-06,
+      "loss": 0.5166,
+      "step": 2928
+    },
+    {
+      "epoch": 7.161369193154035,
+      "grad_norm": 0.2259532147608978,
+      "learning_rate": 2.6771580703834765e-06,
+      "loss": 0.7099,
+      "step": 2929
+    },
+    {
+      "epoch": 7.163814180929095,
+      "grad_norm": 0.22245470278937962,
+      "learning_rate": 2.6758810729575996e-06,
+      "loss": 0.4901,
+      "step": 2930
+    },
+    {
+      "epoch": 7.166259168704157,
+      "grad_norm": 0.24832582897620054,
+      "learning_rate": 2.67460402941168e-06,
+      "loss": 0.9396,
+      "step": 2931
+    },
+    {
+      "epoch": 7.168704156479218,
+      "grad_norm": 0.19387990831694119,
+      "learning_rate": 2.6733269400805873e-06,
+      "loss": 0.4139,
+      "step": 2932
+    },
+    {
+      "epoch": 7.1711491442542785,
+      "grad_norm": 0.20145170067938317,
+      "learning_rate": 2.6720498052992027e-06,
+      "loss": 0.682,
+      "step": 2933
+    },
+    {
+      "epoch": 7.17359413202934,
+      "grad_norm": 0.22821181254687828,
+      "learning_rate": 2.670772625402421e-06,
+      "loss": 0.6827,
+      "step": 2934
+    },
+    {
+      "epoch": 7.176039119804401,
+      "grad_norm": 0.2549198815844666,
+      "learning_rate": 2.6694954007251473e-06,
+      "loss": 0.7026,
+      "step": 2935
+    },
+    {
+      "epoch": 7.178484107579462,
+      "grad_norm": 0.27325397442169697,
+      "learning_rate": 2.6682181316022997e-06,
+      "loss": 0.4772,
+      "step": 2936
+    },
+    {
+      "epoch": 7.180929095354523,
+      "grad_norm": 0.26054492466893603,
+      "learning_rate": 2.6669408183688073e-06,
+      "loss": 0.5789,
+      "step": 2937
+    },
+    {
+      "epoch": 7.183374083129585,
+      "grad_norm": 0.313714585961581,
+      "learning_rate": 2.665663461359611e-06,
+      "loss": 0.6954,
+      "step": 2938
+    },
+    {
+      "epoch": 7.185819070904645,
+      "grad_norm": 0.20054052361492203,
+      "learning_rate": 2.6643860609096627e-06,
+      "loss": 0.6214,
+      "step": 2939
+    },
+    {
+      "epoch": 7.188264058679707,
+      "grad_norm": 0.22081203247082012,
+      "learning_rate": 2.663108617353926e-06,
+      "loss": 0.575,
+      "step": 2940
+    },
+    {
+      "epoch": 7.190709046454768,
+      "grad_norm": 0.16713815639057567,
+      "learning_rate": 2.661831131027377e-06,
+      "loss": 0.8715,
+      "step": 2941
+    },
+    {
+      "epoch": 7.193154034229829,
+      "grad_norm": 0.24276798644266592,
+      "learning_rate": 2.660553602265e-06,
+      "loss": 0.7361,
+      "step": 2942
+    },
+    {
+      "epoch": 7.19559902200489,
+      "grad_norm": 0.2605901492973475,
+      "learning_rate": 2.6592760314017927e-06,
+      "loss": 0.4941,
+      "step": 2943
+    },
+    {
+      "epoch": 7.198044009779951,
+      "grad_norm": 0.1786826033971783,
+      "learning_rate": 2.6579984187727653e-06,
+      "loss": 0.5443,
+      "step": 2944
+    },
+    {
+      "epoch": 7.200488997555012,
+      "grad_norm": 0.2606748743012102,
+      "learning_rate": 2.6567207647129343e-06,
+      "loss": 0.3753,
+      "step": 2945
+    },
+    {
+      "epoch": 7.202933985330073,
+      "grad_norm": 0.19414882758983018,
+      "learning_rate": 2.6554430695573323e-06,
+      "loss": 0.3664,
+      "step": 2946
+    },
+    {
+      "epoch": 7.205378973105135,
+      "grad_norm": 0.2264733794119016,
+      "learning_rate": 2.654165333640999e-06,
+      "loss": 0.6219,
+      "step": 2947
+    },
+    {
+      "epoch": 7.207823960880195,
+      "grad_norm": 0.19034874069577673,
+      "learning_rate": 2.652887557298986e-06,
+      "loss": 0.6806,
+      "step": 2948
+    },
+    {
+      "epoch": 7.210268948655257,
+      "grad_norm": 0.2752382280083933,
+      "learning_rate": 2.6516097408663554e-06,
+      "loss": 0.6122,
+      "step": 2949
+    },
+    {
+      "epoch": 7.212713936430318,
+      "grad_norm": 0.3066266089459573,
+      "learning_rate": 2.6503318846781806e-06,
+      "loss": 0.7968,
+      "step": 2950
+    },
+    {
+      "epoch": 7.215158924205379,
+      "grad_norm": 0.16419497836946748,
+      "learning_rate": 2.6490539890695444e-06,
+      "loss": 0.3694,
+      "step": 2951
+    },
+    {
+      "epoch": 7.21760391198044,
+      "grad_norm": 0.17491037978819649,
+      "learning_rate": 2.6477760543755397e-06,
+      "loss": 0.4807,
+      "step": 2952
+    },
+    {
+      "epoch": 7.2200488997555015,
+      "grad_norm": 0.24596675408316465,
+      "learning_rate": 2.6464980809312714e-06,
+      "loss": 0.4241,
+      "step": 2953
+    },
+    {
+      "epoch": 7.222493887530562,
+      "grad_norm": 0.21827950017245704,
+      "learning_rate": 2.645220069071852e-06,
+      "loss": 0.7709,
+      "step": 2954
+    },
+    {
+      "epoch": 7.224938875305623,
+      "grad_norm": 0.2578390717253169,
+      "learning_rate": 2.6439420191324068e-06,
+      "loss": 0.42,
+      "step": 2955
+    },
+    {
+      "epoch": 7.227383863080685,
+      "grad_norm": 0.2694821835003641,
+      "learning_rate": 2.642663931448069e-06,
+      "loss": 0.5034,
+      "step": 2956
+    },
+    {
+      "epoch": 7.229828850855745,
+      "grad_norm": 0.20612655245266992,
+      "learning_rate": 2.6413858063539825e-06,
+      "loss": 1.0793,
+      "step": 2957
+    },
+    {
+      "epoch": 7.232273838630807,
+      "grad_norm": 0.2426643406424326,
+      "learning_rate": 2.640107644185301e-06,
+      "loss": 0.5392,
+      "step": 2958
+    },
+    {
+      "epoch": 7.234718826405868,
+      "grad_norm": 0.16444199449606253,
+      "learning_rate": 2.6388294452771877e-06,
+      "loss": 0.4558,
+      "step": 2959
+    },
+    {
+      "epoch": 7.237163814180929,
+      "grad_norm": 0.2619059134833961,
+      "learning_rate": 2.637551209964816e-06,
+      "loss": 0.552,
+      "step": 2960
+    },
+    {
+      "epoch": 7.23960880195599,
+      "grad_norm": 0.18439667078066344,
+      "learning_rate": 2.636272938583369e-06,
+      "loss": 0.7145,
+      "step": 2961
+    },
+    {
+      "epoch": 7.2420537897310515,
+      "grad_norm": 0.23863734257390962,
+      "learning_rate": 2.6349946314680377e-06,
+      "loss": 0.4409,
+      "step": 2962
+    },
+    {
+      "epoch": 7.244498777506112,
+      "grad_norm": 0.2198552145322951,
+      "learning_rate": 2.633716288954023e-06,
+      "loss": 0.6169,
+      "step": 2963
+    },
+    {
+      "epoch": 7.2469437652811735,
+      "grad_norm": 0.22326119636185135,
+      "learning_rate": 2.6324379113765375e-06,
+      "loss": 0.4382,
+      "step": 2964
+    },
+    {
+      "epoch": 7.249388753056235,
+      "grad_norm": 0.27014790237821074,
+      "learning_rate": 2.631159499070799e-06,
+      "loss": 0.4749,
+      "step": 2965
+    },
+    {
+      "epoch": 7.251833740831296,
+      "grad_norm": 0.24081356311916513,
+      "learning_rate": 2.629881052372038e-06,
+      "loss": 0.6134,
+      "step": 2966
+    },
+    {
+      "epoch": 7.254278728606357,
+      "grad_norm": 0.17161117902722794,
+      "learning_rate": 2.628602571615492e-06,
+      "loss": 0.7874,
+      "step": 2967
+    },
+    {
+      "epoch": 7.256723716381418,
+      "grad_norm": 0.2398725251477415,
+      "learning_rate": 2.6273240571364085e-06,
+      "loss": 0.5432,
+      "step": 2968
+    },
+    {
+      "epoch": 7.25916870415648,
+      "grad_norm": 0.19982384076847412,
+      "learning_rate": 2.626045509270041e-06,
+      "loss": 0.5497,
+      "step": 2969
+    },
+    {
+      "epoch": 7.26161369193154,
+      "grad_norm": 0.1894593131207446,
+      "learning_rate": 2.6247669283516556e-06,
+      "loss": 0.3953,
+      "step": 2970
+    },
+    {
+      "epoch": 7.264058679706602,
+      "grad_norm": 0.1687955973744008,
+      "learning_rate": 2.623488314716527e-06,
+      "loss": 0.6704,
+      "step": 2971
+    },
+    {
+      "epoch": 7.266503667481663,
+      "grad_norm": 0.27730200852178416,
+      "learning_rate": 2.622209668699935e-06,
+      "loss": 0.8758,
+      "step": 2972
+    },
+    {
+      "epoch": 7.2689486552567235,
+      "grad_norm": 0.17811972796571376,
+      "learning_rate": 2.62093099063717e-06,
+      "loss": 0.8755,
+      "step": 2973
+    },
+    {
+      "epoch": 7.271393643031785,
+      "grad_norm": 0.21224666534498302,
+      "learning_rate": 2.6196522808635312e-06,
+      "loss": 0.9706,
+      "step": 2974
+    },
+    {
+      "epoch": 7.273838630806846,
+      "grad_norm": 0.32741921302892496,
+      "learning_rate": 2.6183735397143245e-06,
+      "loss": 0.5269,
+      "step": 2975
+    },
+    {
+      "epoch": 7.276283618581907,
+      "grad_norm": 0.274112751574388,
+      "learning_rate": 2.617094767524867e-06,
+      "loss": 0.5993,
+      "step": 2976
+    },
+    {
+      "epoch": 7.278728606356968,
+      "grad_norm": 0.3280601352893266,
+      "learning_rate": 2.615815964630481e-06,
+      "loss": 0.4208,
+      "step": 2977
+    },
+    {
+      "epoch": 7.28117359413203,
+      "grad_norm": 0.310599068435982,
+      "learning_rate": 2.6145371313664974e-06,
+      "loss": 0.5437,
+      "step": 2978
+    },
+    {
+      "epoch": 7.28361858190709,
+      "grad_norm": 0.2557657092366433,
+      "learning_rate": 2.6132582680682555e-06,
+      "loss": 0.5243,
+      "step": 2979
+    },
+    {
+      "epoch": 7.286063569682152,
+      "grad_norm": 0.2930215435854808,
+      "learning_rate": 2.6119793750711037e-06,
+      "loss": 0.4333,
+      "step": 2980
+    },
+    {
+      "epoch": 7.288508557457213,
+      "grad_norm": 0.17914355849993288,
+      "learning_rate": 2.610700452710397e-06,
+      "loss": 0.6089,
+      "step": 2981
+    },
+    {
+      "epoch": 7.290953545232274,
+      "grad_norm": 0.1947525712560195,
+      "learning_rate": 2.6094215013214968e-06,
+      "loss": 0.5047,
+      "step": 2982
+    },
+    {
+      "epoch": 7.293398533007335,
+      "grad_norm": 0.27300799867964864,
+      "learning_rate": 2.6081425212397743e-06,
+      "loss": 0.5266,
+      "step": 2983
+    },
+    {
+      "epoch": 7.295843520782396,
+      "grad_norm": 0.253010527314013,
+      "learning_rate": 2.606863512800607e-06,
+      "loss": 0.6578,
+      "step": 2984
+    },
+    {
+      "epoch": 7.298288508557457,
+      "grad_norm": 0.15130497976652685,
+      "learning_rate": 2.605584476339381e-06,
+      "loss": 0.3647,
+      "step": 2985
+    },
+    {
+      "epoch": 7.300733496332518,
+      "grad_norm": 0.2721948823865963,
+      "learning_rate": 2.6043054121914874e-06,
+      "loss": 0.6244,
+      "step": 2986
+    },
+    {
+      "epoch": 7.30317848410758,
+      "grad_norm": 0.26666221734950074,
+      "learning_rate": 2.603026320692328e-06,
+      "loss": 0.4325,
+      "step": 2987
+    },
+    {
+      "epoch": 7.30562347188264,
+      "grad_norm": 0.18867001521198126,
+      "learning_rate": 2.601747202177308e-06,
+      "loss": 0.5654,
+      "step": 2988
+    },
+    {
+      "epoch": 7.308068459657702,
+      "grad_norm": 0.16113026859571675,
+      "learning_rate": 2.600468056981843e-06,
+      "loss": 0.467,
+      "step": 2989
+    },
+    {
+      "epoch": 7.310513447432763,
+      "grad_norm": 0.23391896775037566,
+      "learning_rate": 2.599188885441353e-06,
+      "loss": 0.495,
+      "step": 2990
+    },
+    {
+      "epoch": 7.312958435207824,
+      "grad_norm": 0.24518063952663302,
+      "learning_rate": 2.5979096878912674e-06,
+      "loss": 0.642,
+      "step": 2991
+    },
+    {
+      "epoch": 7.315403422982885,
+      "grad_norm": 0.24886738902605468,
+      "learning_rate": 2.5966304646670205e-06,
+      "loss": 0.6152,
+      "step": 2992
+    },
+    {
+      "epoch": 7.3178484107579465,
+      "grad_norm": 0.38492841103188963,
+      "learning_rate": 2.595351216104054e-06,
+      "loss": 0.4035,
+      "step": 2993
+    },
+    {
+      "epoch": 7.320293398533007,
+      "grad_norm": 0.23033118490359544,
+      "learning_rate": 2.5940719425378146e-06,
+      "loss": 0.5255,
+      "step": 2994
+    },
+    {
+      "epoch": 7.322738386308068,
+      "grad_norm": 0.18682112409446436,
+      "learning_rate": 2.592792644303759e-06,
+      "loss": 0.3984,
+      "step": 2995
+    },
+    {
+      "epoch": 7.32518337408313,
+      "grad_norm": 0.2569320141099276,
+      "learning_rate": 2.5915133217373486e-06,
+      "loss": 0.4159,
+      "step": 2996
+    },
+    {
+      "epoch": 7.32762836185819,
+      "grad_norm": 0.20084271876316015,
+      "learning_rate": 2.5902339751740495e-06,
+      "loss": 0.4319,
+      "step": 2997
+    },
+    {
+      "epoch": 7.330073349633252,
+      "grad_norm": 0.21885052279069006,
+      "learning_rate": 2.588954604949337e-06,
+      "loss": 0.4847,
+      "step": 2998
+    },
+    {
+      "epoch": 7.332518337408313,
+      "grad_norm": 0.2607211894337378,
+      "learning_rate": 2.58767521139869e-06,
+      "loss": 0.6601,
+      "step": 2999
+    },
+    {
+      "epoch": 7.334963325183374,
+      "grad_norm": 0.18573235146562372,
+      "learning_rate": 2.586395794857597e-06,
+      "loss": 0.7328,
+      "step": 3000
+    },
+    {
+      "epoch": 7.337408312958435,
+      "grad_norm": 0.19264480822199387,
+      "learning_rate": 2.5851163556615477e-06,
+      "loss": 0.3931,
+      "step": 3001
+    },
+    {
+      "epoch": 7.3398533007334965,
+      "grad_norm": 0.15248358598988995,
+      "learning_rate": 2.5838368941460424e-06,
+      "loss": 0.401,
+      "step": 3002
+    },
+    {
+      "epoch": 7.342298288508557,
+      "grad_norm": 0.2635920588085101,
+      "learning_rate": 2.582557410646584e-06,
+      "loss": 0.4985,
+      "step": 3003
+    },
+    {
+      "epoch": 7.3447432762836184,
+      "grad_norm": 0.24724241576032033,
+      "learning_rate": 2.581277905498682e-06,
+      "loss": 0.5974,
+      "step": 3004
+    },
+    {
+      "epoch": 7.34718826405868,
+      "grad_norm": 0.3105746730882682,
+      "learning_rate": 2.579998379037853e-06,
+      "loss": 0.6453,
+      "step": 3005
+    },
+    {
+      "epoch": 7.34963325183374,
+      "grad_norm": 0.2837998957809835,
+      "learning_rate": 2.5787188315996174e-06,
+      "loss": 0.979,
+      "step": 3006
+    },
+    {
+      "epoch": 7.352078239608802,
+      "grad_norm": 0.17875702232896482,
+      "learning_rate": 2.5774392635195034e-06,
+      "loss": 0.3907,
+      "step": 3007
+    },
+    {
+      "epoch": 7.354523227383863,
+      "grad_norm": 0.22796479694519697,
+      "learning_rate": 2.576159675133041e-06,
+      "loss": 0.3553,
+      "step": 3008
+    },
+    {
+      "epoch": 7.356968215158924,
+      "grad_norm": 0.21862653271171265,
+      "learning_rate": 2.574880066775768e-06,
+      "loss": 0.8928,
+      "step": 3009
+    },
+    {
+      "epoch": 7.359413202933985,
+      "grad_norm": 0.22612851527264033,
+      "learning_rate": 2.573600438783228e-06,
+      "loss": 1.0224,
+      "step": 3010
+    },
+    {
+      "epoch": 7.361858190709047,
+      "grad_norm": 0.31028712900532934,
+      "learning_rate": 2.5723207914909683e-06,
+      "loss": 0.4225,
+      "step": 3011
+    },
+    {
+      "epoch": 7.364303178484108,
+      "grad_norm": 0.18582083063007593,
+      "learning_rate": 2.571041125234541e-06,
+      "loss": 0.6432,
+      "step": 3012
+    },
+    {
+      "epoch": 7.3667481662591685,
+      "grad_norm": 0.39592482881641045,
+      "learning_rate": 2.5697614403495047e-06,
+      "loss": 0.4614,
+      "step": 3013
+    },
+    {
+      "epoch": 7.36919315403423,
+      "grad_norm": 0.32490787442781377,
+      "learning_rate": 2.5684817371714215e-06,
+      "loss": 0.7597,
+      "step": 3014
+    },
+    {
+      "epoch": 7.371638141809291,
+      "grad_norm": 0.24597339568802967,
+      "learning_rate": 2.5672020160358594e-06,
+      "loss": 0.6171,
+      "step": 3015
+    },
+    {
+      "epoch": 7.374083129584352,
+      "grad_norm": 0.13920933361593787,
+      "learning_rate": 2.56592227727839e-06,
+      "loss": 0.4047,
+      "step": 3016
+    },
+    {
+      "epoch": 7.376528117359413,
+      "grad_norm": 0.3537658207729003,
+      "learning_rate": 2.564642521234591e-06,
+      "loss": 0.3657,
+      "step": 3017
+    },
+    {
+      "epoch": 7.378973105134475,
+      "grad_norm": 0.2527338755418982,
+      "learning_rate": 2.563362748240043e-06,
+      "loss": 0.4189,
+      "step": 3018
+    },
+    {
+      "epoch": 7.381418092909535,
+      "grad_norm": 0.28029910000597685,
+      "learning_rate": 2.562082958630332e-06,
+      "loss": 0.7656,
+      "step": 3019
+    },
+    {
+      "epoch": 7.383863080684597,
+      "grad_norm": 0.2868369160858378,
+      "learning_rate": 2.560803152741048e-06,
+      "loss": 0.3976,
+      "step": 3020
+    },
+    {
+      "epoch": 7.386308068459658,
+      "grad_norm": 0.33983470571230534,
+      "learning_rate": 2.559523330907786e-06,
+      "loss": 0.4035,
+      "step": 3021
+    },
+    {
+      "epoch": 7.388753056234719,
+      "grad_norm": 0.20548505186591842,
+      "learning_rate": 2.558243493466143e-06,
+      "loss": 0.505,
+      "step": 3022
+    },
+    {
+      "epoch": 7.39119804400978,
+      "grad_norm": 0.22560695512715223,
+      "learning_rate": 2.5569636407517236e-06,
+      "loss": 0.5681,
+      "step": 3023
+    },
+    {
+      "epoch": 7.393643031784841,
+      "grad_norm": 0.30957612893826314,
+      "learning_rate": 2.5556837731001327e-06,
+      "loss": 0.4167,
+      "step": 3024
+    },
+    {
+      "epoch": 7.396088019559902,
+      "grad_norm": 0.25213066813495016,
+      "learning_rate": 2.554403890846983e-06,
+      "loss": 0.4206,
+      "step": 3025
+    },
+    {
+      "epoch": 7.398533007334963,
+      "grad_norm": 0.29377561120107787,
+      "learning_rate": 2.5531239943278875e-06,
+      "loss": 0.4275,
+      "step": 3026
+    },
+    {
+      "epoch": 7.400977995110025,
+      "grad_norm": 0.24972624162853088,
+      "learning_rate": 2.551844083878464e-06,
+      "loss": 0.6882,
+      "step": 3027
+    },
+    {
+      "epoch": 7.403422982885085,
+      "grad_norm": 0.1800902171482418,
+      "learning_rate": 2.550564159834335e-06,
+      "loss": 0.387,
+      "step": 3028
+    },
+    {
+      "epoch": 7.405867970660147,
+      "grad_norm": 0.29228083711624503,
+      "learning_rate": 2.549284222531126e-06,
+      "loss": 0.4235,
+      "step": 3029
+    },
+    {
+      "epoch": 7.408312958435208,
+      "grad_norm": 0.2377301917286694,
+      "learning_rate": 2.548004272304466e-06,
+      "loss": 0.5987,
+      "step": 3030
+    },
+    {
+      "epoch": 7.410757946210269,
+      "grad_norm": 0.25098074720481267,
+      "learning_rate": 2.546724309489986e-06,
+      "loss": 0.7531,
+      "step": 3031
+    },
+    {
+      "epoch": 7.41320293398533,
+      "grad_norm": 0.1585232265398779,
+      "learning_rate": 2.5454443344233237e-06,
+      "loss": 0.4742,
+      "step": 3032
+    },
+    {
+      "epoch": 7.4156479217603914,
+      "grad_norm": 0.28072409603685156,
+      "learning_rate": 2.5441643474401154e-06,
+      "loss": 0.6785,
+      "step": 3033
+    },
+    {
+      "epoch": 7.418092909535452,
+      "grad_norm": 0.2604901554050247,
+      "learning_rate": 2.5428843488760047e-06,
+      "loss": 0.5039,
+      "step": 3034
+    },
+    {
+      "epoch": 7.420537897310513,
+      "grad_norm": 0.27961694355764205,
+      "learning_rate": 2.5416043390666366e-06,
+      "loss": 0.8087,
+      "step": 3035
+    },
+    {
+      "epoch": 7.422982885085575,
+      "grad_norm": 0.22774640603189428,
+      "learning_rate": 2.5403243183476583e-06,
+      "loss": 0.3912,
+      "step": 3036
+    },
+    {
+      "epoch": 7.425427872860635,
+      "grad_norm": 0.2501693371116074,
+      "learning_rate": 2.5390442870547204e-06,
+      "loss": 0.897,
+      "step": 3037
+    },
+    {
+      "epoch": 7.427872860635697,
+      "grad_norm": 0.16049190845485004,
+      "learning_rate": 2.5377642455234764e-06,
+      "loss": 0.641,
+      "step": 3038
+    },
+    {
+      "epoch": 7.430317848410758,
+      "grad_norm": 0.2490419741601378,
+      "learning_rate": 2.5364841940895824e-06,
+      "loss": 0.9267,
+      "step": 3039
+    },
+    {
+      "epoch": 7.432762836185819,
+      "grad_norm": 0.39942779544388296,
+      "learning_rate": 2.5352041330886983e-06,
+      "loss": 0.7282,
+      "step": 3040
+    },
+    {
+      "epoch": 7.43520782396088,
+      "grad_norm": 0.33902692858378725,
+      "learning_rate": 2.533924062856485e-06,
+      "loss": 0.4725,
+      "step": 3041
+    },
+    {
+      "epoch": 7.4376528117359415,
+      "grad_norm": 0.16852391063269725,
+      "learning_rate": 2.5326439837286056e-06,
+      "loss": 0.4283,
+      "step": 3042
+    },
+    {
+      "epoch": 7.440097799511002,
+      "grad_norm": 0.27402820569518577,
+      "learning_rate": 2.5313638960407257e-06,
+      "loss": 0.4424,
+      "step": 3043
+    },
+    {
+      "epoch": 7.442542787286063,
+      "grad_norm": 0.1691815766016492,
+      "learning_rate": 2.5300838001285138e-06,
+      "loss": 0.4659,
+      "step": 3044
+    },
+    {
+      "epoch": 7.444987775061125,
+      "grad_norm": 0.21312398871392432,
+      "learning_rate": 2.5288036963276418e-06,
+      "loss": 0.6387,
+      "step": 3045
+    },
+    {
+      "epoch": 7.447432762836186,
+      "grad_norm": 0.2580434180449843,
+      "learning_rate": 2.5275235849737816e-06,
+      "loss": 0.3497,
+      "step": 3046
+    },
+    {
+      "epoch": 7.449877750611247,
+      "grad_norm": 0.22142529211194736,
+      "learning_rate": 2.526243466402607e-06,
+      "loss": 0.5879,
+      "step": 3047
+    },
+    {
+      "epoch": 7.452322738386308,
+      "grad_norm": 0.28453282966537363,
+      "learning_rate": 2.524963340949795e-06,
+      "loss": 0.6065,
+      "step": 3048
+    },
+    {
+      "epoch": 7.45476772616137,
+      "grad_norm": 0.25583762718860864,
+      "learning_rate": 2.523683208951023e-06,
+      "loss": 0.5536,
+      "step": 3049
+    },
+    {
+      "epoch": 7.45721271393643,
+      "grad_norm": 0.2513097814568369,
+      "learning_rate": 2.522403070741971e-06,
+      "loss": 0.5697,
+      "step": 3050
+    },
+    {
+      "epoch": 7.459657701711492,
+      "grad_norm": 0.24574963961874632,
+      "learning_rate": 2.5211229266583227e-06,
+      "loss": 0.9711,
+      "step": 3051
+    },
+    {
+      "epoch": 7.462102689486553,
+      "grad_norm": 0.17398616180687465,
+      "learning_rate": 2.519842777035759e-06,
+      "loss": 0.8388,
+      "step": 3052
+    },
+    {
+      "epoch": 7.4645476772616135,
+      "grad_norm": 0.2483442806086505,
+      "learning_rate": 2.5185626222099634e-06,
+      "loss": 0.9939,
+      "step": 3053
+    },
+    {
+      "epoch": 7.466992665036675,
+      "grad_norm": 0.332713206201648,
+      "learning_rate": 2.5172824625166246e-06,
+      "loss": 0.6806,
+      "step": 3054
+    },
+    {
+      "epoch": 7.469437652811736,
+      "grad_norm": 0.1952017106663902,
+      "learning_rate": 2.5160022982914283e-06,
+      "loss": 0.4928,
+      "step": 3055
+    },
+    {
+      "epoch": 7.471882640586797,
+      "grad_norm": 0.23318188727798092,
+      "learning_rate": 2.514722129870063e-06,
+      "loss": 0.9781,
+      "step": 3056
+    },
+    {
+      "epoch": 7.474327628361858,
+      "grad_norm": 0.2955992672319146,
+      "learning_rate": 2.5134419575882186e-06,
+      "loss": 0.4968,
+      "step": 3057
+    },
+    {
+      "epoch": 7.47677261613692,
+      "grad_norm": 0.19652001076603975,
+      "learning_rate": 2.512161781781585e-06,
+      "loss": 0.6841,
+      "step": 3058
+    },
+    {
+      "epoch": 7.47921760391198,
+      "grad_norm": 0.27635151418185444,
+      "learning_rate": 2.5108816027858534e-06,
+      "loss": 0.4746,
+      "step": 3059
+    },
+    {
+      "epoch": 7.481662591687042,
+      "grad_norm": 0.20029156472941967,
+      "learning_rate": 2.509601420936717e-06,
+      "loss": 0.4187,
+      "step": 3060
+    },
+    {
+      "epoch": 7.484107579462103,
+      "grad_norm": 0.1940639684185768,
+      "learning_rate": 2.508321236569868e-06,
+      "loss": 0.4481,
+      "step": 3061
+    },
+    {
+      "epoch": 7.486552567237164,
+      "grad_norm": 0.3534214377274368,
+      "learning_rate": 2.507041050021001e-06,
+      "loss": 0.8606,
+      "step": 3062
+    },
+    {
+      "epoch": 7.488997555012225,
+      "grad_norm": 0.24147646593089964,
+      "learning_rate": 2.505760861625808e-06,
+      "loss": 0.389,
+      "step": 3063
+    },
+    {
+      "epoch": 7.491442542787286,
+      "grad_norm": 0.2827565574547815,
+      "learning_rate": 2.5044806717199867e-06,
+      "loss": 0.3694,
+      "step": 3064
+    },
+    {
+      "epoch": 7.493887530562347,
+      "grad_norm": 0.19286131188641506,
+      "learning_rate": 2.5032004806392296e-06,
+      "loss": 0.6074,
+      "step": 3065
+    },
+    {
+      "epoch": 7.496332518337408,
+      "grad_norm": 0.22638559719212145,
+      "learning_rate": 2.5019202887192335e-06,
+      "loss": 0.4574,
+      "step": 3066
+    },
+    {
+      "epoch": 7.49877750611247,
+      "grad_norm": 0.2197834027513471,
+      "learning_rate": 2.5006400962956946e-06,
+      "loss": 0.3832,
+      "step": 3067
+    },
+    {
+      "epoch": 7.50122249388753,
+      "grad_norm": 0.299025486887785,
+      "learning_rate": 2.4993599037043063e-06,
+      "loss": 0.6896,
+      "step": 3068
+    },
+    {
+      "epoch": 7.503667481662592,
+      "grad_norm": 0.2095614338150072,
+      "learning_rate": 2.498079711280767e-06,
+      "loss": 0.6673,
+      "step": 3069
+    },
+    {
+      "epoch": 7.506112469437653,
+      "grad_norm": 0.2958210297138419,
+      "learning_rate": 2.4967995193607713e-06,
+      "loss": 0.442,
+      "step": 3070
+    },
+    {
+      "epoch": 7.508557457212714,
+      "grad_norm": 0.20643013406650393,
+      "learning_rate": 2.495519328280015e-06,
+      "loss": 0.8165,
+      "step": 3071
+    },
+    {
+      "epoch": 7.511002444987775,
+      "grad_norm": 0.22687363631541038,
+      "learning_rate": 2.4942391383741925e-06,
+      "loss": 0.6408,
+      "step": 3072
+    },
+    {
+      "epoch": 7.513447432762836,
+      "grad_norm": 0.17539261811180762,
+      "learning_rate": 2.4929589499790004e-06,
+      "loss": 0.6777,
+      "step": 3073
+    },
+    {
+      "epoch": 7.515892420537897,
+      "grad_norm": 0.25795241122804413,
+      "learning_rate": 2.491678763430132e-06,
+      "loss": 0.5883,
+      "step": 3074
+    },
+    {
+      "epoch": 7.518337408312958,
+      "grad_norm": 0.2542978446735659,
+      "learning_rate": 2.4903985790632835e-06,
+      "loss": 0.7852,
+      "step": 3075
+    },
+    {
+      "epoch": 7.52078239608802,
+      "grad_norm": 0.23306459767291018,
+      "learning_rate": 2.489118397214147e-06,
+      "loss": 0.8453,
+      "step": 3076
+    },
+    {
+      "epoch": 7.52322738386308,
+      "grad_norm": 0.19316702078147083,
+      "learning_rate": 2.487838218218416e-06,
+      "loss": 0.5335,
+      "step": 3077
+    },
+    {
+      "epoch": 7.525672371638142,
+      "grad_norm": 0.2412896765339769,
+      "learning_rate": 2.486558042411782e-06,
+      "loss": 0.7603,
+      "step": 3078
+    },
+    {
+      "epoch": 7.528117359413203,
+      "grad_norm": 0.24760459132221221,
+      "learning_rate": 2.4852778701299375e-06,
+      "loss": 0.3275,
+      "step": 3079
+    },
+    {
+      "epoch": 7.530562347188264,
+      "grad_norm": 0.18954719812295986,
+      "learning_rate": 2.4839977017085726e-06,
+      "loss": 0.6596,
+      "step": 3080
+    },
+    {
+      "epoch": 7.533007334963325,
+      "grad_norm": 0.3242798383202998,
+      "learning_rate": 2.4827175374833767e-06,
+      "loss": 0.3886,
+      "step": 3081
+    },
+    {
+      "epoch": 7.5354523227383865,
+      "grad_norm": 0.16208012684464868,
+      "learning_rate": 2.481437377790037e-06,
+      "loss": 0.7222,
+      "step": 3082
+    },
+    {
+      "epoch": 7.537897310513447,
+      "grad_norm": 0.20860461337524006,
+      "learning_rate": 2.480157222964243e-06,
+      "loss": 0.4343,
+      "step": 3083
+    },
+    {
+      "epoch": 7.540342298288508,
+      "grad_norm": 0.21025417567686325,
+      "learning_rate": 2.4788770733416777e-06,
+      "loss": 0.4839,
+      "step": 3084
+    },
+    {
+      "epoch": 7.54278728606357,
+      "grad_norm": 0.31637854851569014,
+      "learning_rate": 2.4775969292580287e-06,
+      "loss": 0.3542,
+      "step": 3085
+    },
+    {
+      "epoch": 7.54523227383863,
+      "grad_norm": 0.27428815704601006,
+      "learning_rate": 2.4763167910489775e-06,
+      "loss": 0.6499,
+      "step": 3086
+    },
+    {
+      "epoch": 7.547677261613692,
+      "grad_norm": 0.1851423567150205,
+      "learning_rate": 2.4750366590502063e-06,
+      "loss": 0.5171,
+      "step": 3087
+    },
+    {
+      "epoch": 7.550122249388753,
+      "grad_norm": 0.3918767615715888,
+      "learning_rate": 2.4737565335973934e-06,
+      "loss": 0.584,
+      "step": 3088
+    },
+    {
+      "epoch": 7.552567237163814,
+      "grad_norm": 0.136371493376507,
+      "learning_rate": 2.4724764150262188e-06,
+      "loss": 0.4393,
+      "step": 3089
+    },
+    {
+      "epoch": 7.555012224938875,
+      "grad_norm": 0.21472356174138402,
+      "learning_rate": 2.4711963036723586e-06,
+      "loss": 0.4288,
+      "step": 3090
+    },
+    {
+      "epoch": 7.557457212713937,
+      "grad_norm": 1.1333014224540965,
+      "learning_rate": 2.4699161998714866e-06,
+      "loss": 0.6945,
+      "step": 3091
+    },
+    {
+      "epoch": 7.559902200488997,
+      "grad_norm": 0.21766224747912963,
+      "learning_rate": 2.4686361039592756e-06,
+      "loss": 0.4151,
+      "step": 3092
+    },
+    {
+      "epoch": 7.5623471882640585,
+      "grad_norm": 0.24123092111061495,
+      "learning_rate": 2.467356016271396e-06,
+      "loss": 0.3888,
+      "step": 3093
+    },
+    {
+      "epoch": 7.56479217603912,
+      "grad_norm": 0.2530918459185112,
+      "learning_rate": 2.466075937143517e-06,
+      "loss": 0.8052,
+      "step": 3094
+    },
+    {
+      "epoch": 7.56723716381418,
+      "grad_norm": 0.26064207752990876,
+      "learning_rate": 2.4647958669113017e-06,
+      "loss": 0.4154,
+      "step": 3095
+    },
+    {
+      "epoch": 7.569682151589242,
+      "grad_norm": 0.2969673794101501,
+      "learning_rate": 2.463515805910418e-06,
+      "loss": 1.0655,
+      "step": 3096
+    },
+    {
+      "epoch": 7.572127139364303,
+      "grad_norm": 0.29479804469298926,
+      "learning_rate": 2.4622357544765244e-06,
+      "loss": 0.5498,
+      "step": 3097
+    },
+    {
+      "epoch": 7.574572127139365,
+      "grad_norm": 0.2570075059243293,
+      "learning_rate": 2.46095571294528e-06,
+      "loss": 0.3113,
+      "step": 3098
+    },
+    {
+      "epoch": 7.577017114914425,
+      "grad_norm": 0.21717549246514384,
+      "learning_rate": 2.4596756816523425e-06,
+      "loss": 0.7294,
+      "step": 3099
+    },
+    {
+      "epoch": 7.579462102689487,
+      "grad_norm": 0.20854011607966358,
+      "learning_rate": 2.458395660933364e-06,
+      "loss": 0.5262,
+      "step": 3100
+    },
+    {
+      "epoch": 7.581907090464548,
+      "grad_norm": 0.20348584720553187,
+      "learning_rate": 2.4571156511239957e-06,
+      "loss": 0.7973,
+      "step": 3101
+    },
+    {
+      "epoch": 7.5843520782396086,
+      "grad_norm": 0.20996518095982614,
+      "learning_rate": 2.455835652559885e-06,
+      "loss": 0.5112,
+      "step": 3102
+    },
+    {
+      "epoch": 7.58679706601467,
+      "grad_norm": 0.2798613367066497,
+      "learning_rate": 2.4545556655766776e-06,
+      "loss": 0.4554,
+      "step": 3103
+    },
+    {
+      "epoch": 7.589242053789731,
+      "grad_norm": 0.21628406734140962,
+      "learning_rate": 2.4532756905100146e-06,
+      "loss": 0.8099,
+      "step": 3104
+    },
+    {
+      "epoch": 7.591687041564792,
+      "grad_norm": 0.24931404934942428,
+      "learning_rate": 2.4519957276955354e-06,
+      "loss": 1.1828,
+      "step": 3105
+    },
+    {
+      "epoch": 7.594132029339853,
+      "grad_norm": 0.2970211474157629,
+      "learning_rate": 2.4507157774688743e-06,
+      "loss": 0.6834,
+      "step": 3106
+    },
+    {
+      "epoch": 7.596577017114915,
+      "grad_norm": 0.18084727239322743,
+      "learning_rate": 2.4494358401656654e-06,
+      "loss": 0.3969,
+      "step": 3107
+    },
+    {
+      "epoch": 7.599022004889975,
+      "grad_norm": 0.25043634630076206,
+      "learning_rate": 2.4481559161215364e-06,
+      "loss": 0.3496,
+      "step": 3108
+    },
+    {
+      "epoch": 7.601466992665037,
+      "grad_norm": 0.23266827103149407,
+      "learning_rate": 2.4468760056721134e-06,
+      "loss": 0.6346,
+      "step": 3109
+    },
+    {
+      "epoch": 7.603911980440098,
+      "grad_norm": 0.2709130000379762,
+      "learning_rate": 2.445596109153018e-06,
+      "loss": 1.2344,
+      "step": 3110
+    },
+    {
+      "epoch": 7.606356968215159,
+      "grad_norm": 0.22920038629441936,
+      "learning_rate": 2.4443162268998677e-06,
+      "loss": 0.3802,
+      "step": 3111
+    },
+    {
+      "epoch": 7.60880195599022,
+      "grad_norm": 0.3238605862359861,
+      "learning_rate": 2.4430363592482772e-06,
+      "loss": 0.5317,
+      "step": 3112
+    },
+    {
+      "epoch": 7.611246943765281,
+      "grad_norm": 0.18694501470826708,
+      "learning_rate": 2.441756506533858e-06,
+      "loss": 0.6354,
+      "step": 3113
+    },
+    {
+      "epoch": 7.613691931540342,
+      "grad_norm": 0.12149681604714048,
+      "learning_rate": 2.4404766690922154e-06,
+      "loss": 0.2385,
+      "step": 3114
+    },
+    {
+      "epoch": 7.616136919315403,
+      "grad_norm": 0.20149419441038058,
+      "learning_rate": 2.439196847258953e-06,
+      "loss": 0.4682,
+      "step": 3115
+    },
+    {
+      "epoch": 7.618581907090465,
+      "grad_norm": 0.18316129613057858,
+      "learning_rate": 2.4379170413696683e-06,
+      "loss": 0.6557,
+      "step": 3116
+    },
+    {
+      "epoch": 7.621026894865525,
+      "grad_norm": 0.24878583243713595,
+      "learning_rate": 2.4366372517599573e-06,
+      "loss": 0.3668,
+      "step": 3117
+    },
+    {
+      "epoch": 7.623471882640587,
+      "grad_norm": 0.19572351963178702,
+      "learning_rate": 2.435357478765409e-06,
+      "loss": 0.5577,
+      "step": 3118
+    },
+    {
+      "epoch": 7.625916870415648,
+      "grad_norm": 0.24939790969271688,
+      "learning_rate": 2.4340777227216102e-06,
+      "loss": 0.7628,
+      "step": 3119
+    },
+    {
+      "epoch": 7.628361858190709,
+      "grad_norm": 0.29580147799702244,
+      "learning_rate": 2.4327979839641414e-06,
+      "loss": 1.1039,
+      "step": 3120
+    },
+    {
+      "epoch": 7.63080684596577,
+      "grad_norm": 0.24793205193479567,
+      "learning_rate": 2.4315182628285794e-06,
+      "loss": 1.0408,
+      "step": 3121
+    },
+    {
+      "epoch": 7.6332518337408315,
+      "grad_norm": 0.23509392737206194,
+      "learning_rate": 2.430238559650496e-06,
+      "loss": 0.4099,
+      "step": 3122
+    },
+    {
+      "epoch": 7.635696821515893,
+      "grad_norm": 0.24806324467933052,
+      "learning_rate": 2.42895887476546e-06,
+      "loss": 0.369,
+      "step": 3123
+    },
+    {
+      "epoch": 7.638141809290953,
+      "grad_norm": 0.2037916016765037,
+      "learning_rate": 2.427679208509033e-06,
+      "loss": 0.6029,
+      "step": 3124
+    },
+    {
+      "epoch": 7.640586797066015,
+      "grad_norm": 0.28029880706299704,
+      "learning_rate": 2.4263995612167733e-06,
+      "loss": 1.2196,
+      "step": 3125
+    },
+    {
+      "epoch": 7.643031784841076,
+      "grad_norm": 0.26595026334896166,
+      "learning_rate": 2.4251199332242326e-06,
+      "loss": 1.1411,
+      "step": 3126
+    },
+    {
+      "epoch": 7.645476772616137,
+      "grad_norm": 0.23682166046414943,
+      "learning_rate": 2.42384032486696e-06,
+      "loss": 0.4842,
+      "step": 3127
+    },
+    {
+      "epoch": 7.647921760391198,
+      "grad_norm": 0.25539084379869303,
+      "learning_rate": 2.422560736480497e-06,
+      "loss": 0.3856,
+      "step": 3128
+    },
+    {
+      "epoch": 7.65036674816626,
+      "grad_norm": 0.26142277643795847,
+      "learning_rate": 2.4212811684003826e-06,
+      "loss": 0.3374,
+      "step": 3129
+    },
+    {
+      "epoch": 7.65281173594132,
+      "grad_norm": 0.2257427591586434,
+      "learning_rate": 2.4200016209621473e-06,
+      "loss": 0.3485,
+      "step": 3130
+    },
+    {
+      "epoch": 7.6552567237163816,
+      "grad_norm": 0.33243368849347354,
+      "learning_rate": 2.4187220945013187e-06,
+      "loss": 0.4164,
+      "step": 3131
+    },
+    {
+      "epoch": 7.657701711491443,
+      "grad_norm": 0.18678351332570467,
+      "learning_rate": 2.417442589353417e-06,
+      "loss": 0.3696,
+      "step": 3132
+    },
+    {
+      "epoch": 7.6601466992665035,
+      "grad_norm": 0.18927821584272086,
+      "learning_rate": 2.4161631058539584e-06,
+      "loss": 0.6585,
+      "step": 3133
+    },
+    {
+      "epoch": 7.662591687041565,
+      "grad_norm": 0.26544126333496704,
+      "learning_rate": 2.414883644338453e-06,
+      "loss": 0.6283,
+      "step": 3134
+    },
+    {
+      "epoch": 7.665036674816626,
+      "grad_norm": 0.3545344959476537,
+      "learning_rate": 2.4136042051424044e-06,
+      "loss": 0.7948,
+      "step": 3135
+    },
+    {
+      "epoch": 7.667481662591687,
+      "grad_norm": 0.2381216169463349,
+      "learning_rate": 2.4123247886013103e-06,
+      "loss": 0.3896,
+      "step": 3136
+    },
+    {
+      "epoch": 7.669926650366748,
+      "grad_norm": 0.19688833541210066,
+      "learning_rate": 2.411045395050664e-06,
+      "loss": 0.4053,
+      "step": 3137
+    },
+    {
+      "epoch": 7.67237163814181,
+      "grad_norm": 0.2632038454551941,
+      "learning_rate": 2.4097660248259505e-06,
+      "loss": 0.7826,
+      "step": 3138
+    },
+    {
+      "epoch": 7.67481662591687,
+      "grad_norm": 0.24262971968419264,
+      "learning_rate": 2.408486678262652e-06,
+      "loss": 0.7517,
+      "step": 3139
+    },
+    {
+      "epoch": 7.677261613691932,
+      "grad_norm": 0.19645329809199671,
+      "learning_rate": 2.4072073556962413e-06,
+      "loss": 0.4941,
+      "step": 3140
+    },
+    {
+      "epoch": 7.679706601466993,
+      "grad_norm": 0.3042721898639073,
+      "learning_rate": 2.405928057462186e-06,
+      "loss": 0.4733,
+      "step": 3141
+    },
+    {
+      "epoch": 7.6821515892420535,
+      "grad_norm": 0.15354177989267473,
+      "learning_rate": 2.404648783895947e-06,
+      "loss": 0.6951,
+      "step": 3142
+    },
+    {
+      "epoch": 7.684596577017115,
+      "grad_norm": 0.24551650479184892,
+      "learning_rate": 2.40336953533298e-06,
+      "loss": 0.4915,
+      "step": 3143
+    },
+    {
+      "epoch": 7.687041564792176,
+      "grad_norm": 0.20497693746367474,
+      "learning_rate": 2.402090312108733e-06,
+      "loss": 0.5237,
+      "step": 3144
+    },
+    {
+      "epoch": 7.689486552567237,
+      "grad_norm": 0.23141152778065371,
+      "learning_rate": 2.4008111145586473e-06,
+      "loss": 0.3957,
+      "step": 3145
+    },
+    {
+      "epoch": 7.691931540342298,
+      "grad_norm": 0.24998408901827956,
+      "learning_rate": 2.399531943018158e-06,
+      "loss": 0.7646,
+      "step": 3146
+    },
+    {
+      "epoch": 7.69437652811736,
+      "grad_norm": 0.1812420897044591,
+      "learning_rate": 2.3982527978226928e-06,
+      "loss": 0.4292,
+      "step": 3147
+    },
+    {
+      "epoch": 7.69682151589242,
+      "grad_norm": 0.24762521513003954,
+      "learning_rate": 2.3969736793076724e-06,
+      "loss": 0.6067,
+      "step": 3148
+    },
+    {
+      "epoch": 7.699266503667482,
+      "grad_norm": 0.28776322788633224,
+      "learning_rate": 2.3956945878085126e-06,
+      "loss": 0.4628,
+      "step": 3149
+    },
+    {
+      "epoch": 7.701711491442543,
+      "grad_norm": 0.19873109913489917,
+      "learning_rate": 2.3944155236606196e-06,
+      "loss": 0.4561,
+      "step": 3150
+    },
+    {
+      "epoch": 7.704156479217604,
+      "grad_norm": 0.2718957934517454,
+      "learning_rate": 2.3931364871993936e-06,
+      "loss": 0.711,
+      "step": 3151
+    },
+    {
+      "epoch": 7.706601466992665,
+      "grad_norm": 0.3742228184716042,
+      "learning_rate": 2.391857478760226e-06,
+      "loss": 0.6987,
+      "step": 3152
+    },
+    {
+      "epoch": 7.709046454767726,
+      "grad_norm": 0.22814577541179415,
+      "learning_rate": 2.3905784986785036e-06,
+      "loss": 0.4985,
+      "step": 3153
+    },
+    {
+      "epoch": 7.711491442542787,
+      "grad_norm": 0.29769822406075214,
+      "learning_rate": 2.3892995472896043e-06,
+      "loss": 0.5882,
+      "step": 3154
+    },
+    {
+      "epoch": 7.713936430317848,
+      "grad_norm": 0.26045635976734643,
+      "learning_rate": 2.388020624928897e-06,
+      "loss": 0.3918,
+      "step": 3155
+    },
+    {
+      "epoch": 7.71638141809291,
+      "grad_norm": 0.2316618379215414,
+      "learning_rate": 2.386741731931745e-06,
+      "loss": 0.6423,
+      "step": 3156
+    },
+    {
+      "epoch": 7.71882640586797,
+      "grad_norm": 0.17477930855745918,
+      "learning_rate": 2.385462868633504e-06,
+      "loss": 0.4273,
+      "step": 3157
+    },
+    {
+      "epoch": 7.721271393643032,
+      "grad_norm": 0.30484145992022255,
+      "learning_rate": 2.3841840353695204e-06,
+      "loss": 0.6654,
+      "step": 3158
+    },
+    {
+      "epoch": 7.723716381418093,
+      "grad_norm": 0.19539622918562286,
+      "learning_rate": 2.3829052324751334e-06,
+      "loss": 0.8415,
+      "step": 3159
+    },
+    {
+      "epoch": 7.726161369193154,
+      "grad_norm": 0.38854475430340346,
+      "learning_rate": 2.3816264602856755e-06,
+      "loss": 0.9295,
+      "step": 3160
+    },
+    {
+      "epoch": 7.728606356968215,
+      "grad_norm": 0.2634508293024988,
+      "learning_rate": 2.38034771913647e-06,
+      "loss": 0.8087,
+      "step": 3161
+    },
+    {
+      "epoch": 7.7310513447432765,
+      "grad_norm": 0.28004455048678034,
+      "learning_rate": 2.3790690093628303e-06,
+      "loss": 0.5717,
+      "step": 3162
+    },
+    {
+      "epoch": 7.733496332518337,
+      "grad_norm": 0.2457687233142582,
+      "learning_rate": 2.3777903313000658e-06,
+      "loss": 0.4699,
+      "step": 3163
+    },
+    {
+      "epoch": 7.735941320293398,
+      "grad_norm": 0.26529724540965743,
+      "learning_rate": 2.376511685283474e-06,
+      "loss": 0.6874,
+      "step": 3164
+    },
+    {
+      "epoch": 7.73838630806846,
+      "grad_norm": 0.22724169588157916,
+      "learning_rate": 2.375233071648345e-06,
+      "loss": 0.4439,
+      "step": 3165
+    },
+    {
+      "epoch": 7.74083129584352,
+      "grad_norm": 0.2543940101695516,
+      "learning_rate": 2.37395449072996e-06,
+      "loss": 0.635,
+      "step": 3166
+    },
+    {
+      "epoch": 7.743276283618582,
+      "grad_norm": 0.23534333702766563,
+      "learning_rate": 2.3726759428635932e-06,
+      "loss": 0.5333,
+      "step": 3167
+    },
+    {
+      "epoch": 7.745721271393643,
+      "grad_norm": 0.29225670957661554,
+      "learning_rate": 2.371397428384509e-06,
+      "loss": 0.3734,
+      "step": 3168
+    },
+    {
+      "epoch": 7.748166259168704,
+      "grad_norm": 0.2908015095675076,
+      "learning_rate": 2.370118947627963e-06,
+      "loss": 0.5655,
+      "step": 3169
+    },
+    {
+      "epoch": 7.750611246943765,
+      "grad_norm": 0.3161393612001849,
+      "learning_rate": 2.368840500929201e-06,
+      "loss": 0.5279,
+      "step": 3170
+    },
+    {
+      "epoch": 7.7530562347188265,
+      "grad_norm": 0.23357802992854354,
+      "learning_rate": 2.3675620886234634e-06,
+      "loss": 0.6345,
+      "step": 3171
+    },
+    {
+      "epoch": 7.755501222493887,
+      "grad_norm": 0.3005708796550451,
+      "learning_rate": 2.366283711045977e-06,
+      "loss": 0.2983,
+      "step": 3172
+    },
+    {
+      "epoch": 7.7579462102689485,
+      "grad_norm": 0.24185348150688785,
+      "learning_rate": 2.365005368531963e-06,
+      "loss": 0.6028,
+      "step": 3173
+    },
+    {
+      "epoch": 7.76039119804401,
+      "grad_norm": 0.26559729372548946,
+      "learning_rate": 2.363727061416632e-06,
+      "loss": 0.3782,
+      "step": 3174
+    },
+    {
+      "epoch": 7.76283618581907,
+      "grad_norm": 0.2391491338159959,
+      "learning_rate": 2.3624487900351843e-06,
+      "loss": 0.7549,
+      "step": 3175
+    },
+    {
+      "epoch": 7.765281173594132,
+      "grad_norm": 0.2831720594614054,
+      "learning_rate": 2.3611705547228127e-06,
+      "loss": 0.6142,
+      "step": 3176
+    },
+    {
+      "epoch": 7.767726161369193,
+      "grad_norm": 0.19778844692055547,
+      "learning_rate": 2.3598923558147e-06,
+      "loss": 0.6207,
+      "step": 3177
+    },
+    {
+      "epoch": 7.770171149144255,
+      "grad_norm": 0.21700360897866425,
+      "learning_rate": 2.3586141936460188e-06,
+      "loss": 0.8188,
+      "step": 3178
+    },
+    {
+      "epoch": 7.772616136919315,
+      "grad_norm": 0.20446390751614257,
+      "learning_rate": 2.3573360685519324e-06,
+      "loss": 0.419,
+      "step": 3179
+    },
+    {
+      "epoch": 7.775061124694377,
+      "grad_norm": 0.2146638171160273,
+      "learning_rate": 2.3560579808675945e-06,
+      "loss": 0.4322,
+      "step": 3180
+    },
+    {
+      "epoch": 7.777506112469438,
+      "grad_norm": 0.2395970243405885,
+      "learning_rate": 2.3547799309281483e-06,
+      "loss": 0.9569,
+      "step": 3181
+    },
+    {
+      "epoch": 7.7799511002444985,
+      "grad_norm": 0.21724586072791693,
+      "learning_rate": 2.353501919068729e-06,
+      "loss": 0.3865,
+      "step": 3182
+    },
+    {
+      "epoch": 7.78239608801956,
+      "grad_norm": 0.2821998542033634,
+      "learning_rate": 2.3522239456244607e-06,
+      "loss": 0.8293,
+      "step": 3183
+    },
+    {
+      "epoch": 7.784841075794621,
+      "grad_norm": 0.3151105134185895,
+      "learning_rate": 2.3509460109304564e-06,
+      "loss": 0.4647,
+      "step": 3184
+    },
+    {
+      "epoch": 7.787286063569682,
+      "grad_norm": 0.2243361820710655,
+      "learning_rate": 2.34966811532182e-06,
+      "loss": 0.705,
+      "step": 3185
+    },
+    {
+      "epoch": 7.789731051344743,
+      "grad_norm": 0.3260290177692361,
+      "learning_rate": 2.348390259133645e-06,
+      "loss": 0.7008,
+      "step": 3186
+    },
+    {
+      "epoch": 7.792176039119805,
+      "grad_norm": 0.21681754867041578,
+      "learning_rate": 2.3471124427010146e-06,
+      "loss": 0.4151,
+      "step": 3187
+    },
+    {
+      "epoch": 7.794621026894865,
+      "grad_norm": 0.30992780644441953,
+      "learning_rate": 2.345834666359002e-06,
+      "loss": 0.5602,
+      "step": 3188
+    },
+    {
+      "epoch": 7.797066014669927,
+      "grad_norm": 0.2811711957974211,
+      "learning_rate": 2.344556930442669e-06,
+      "loss": 0.4655,
+      "step": 3189
+    },
+    {
+      "epoch": 7.799511002444988,
+      "grad_norm": 0.3994448716397458,
+      "learning_rate": 2.343279235287066e-06,
+      "loss": 1.112,
+      "step": 3190
+    },
+    {
+      "epoch": 7.801955990220049,
+      "grad_norm": 0.18318610060823692,
+      "learning_rate": 2.3420015812272355e-06,
+      "loss": 0.6131,
+      "step": 3191
+    },
+    {
+      "epoch": 7.80440097799511,
+      "grad_norm": 0.2694047533989525,
+      "learning_rate": 2.340723968598207e-06,
+      "loss": 0.38,
+      "step": 3192
+    },
+    {
+      "epoch": 7.806845965770171,
+      "grad_norm": 0.34058683775136644,
+      "learning_rate": 2.3394463977350006e-06,
+      "loss": 0.4159,
+      "step": 3193
+    },
+    {
+      "epoch": 7.809290953545232,
+      "grad_norm": 0.20797645109804502,
+      "learning_rate": 2.338168868972624e-06,
+      "loss": 0.3349,
+      "step": 3194
+    },
+    {
+      "epoch": 7.811735941320293,
+      "grad_norm": 0.2280936349902658,
+      "learning_rate": 2.3368913826460747e-06,
+      "loss": 0.3447,
+      "step": 3195
+    },
+    {
+      "epoch": 7.814180929095355,
+      "grad_norm": 0.3069921428014094,
+      "learning_rate": 2.3356139390903377e-06,
+      "loss": 0.794,
+      "step": 3196
+    },
+    {
+      "epoch": 7.816625916870415,
+      "grad_norm": 0.31277784198727426,
+      "learning_rate": 2.3343365386403895e-06,
+      "loss": 0.5676,
+      "step": 3197
+    },
+    {
+      "epoch": 7.819070904645477,
+      "grad_norm": 0.27399383642020914,
+      "learning_rate": 2.3330591816311935e-06,
+      "loss": 0.4982,
+      "step": 3198
+    },
+    {
+      "epoch": 7.821515892420538,
+      "grad_norm": 0.28448812562833087,
+      "learning_rate": 2.331781868397701e-06,
+      "loss": 0.3521,
+      "step": 3199
+    },
+    {
+      "epoch": 7.823960880195599,
+      "grad_norm": 0.3810281906669007,
+      "learning_rate": 2.3305045992748535e-06,
+      "loss": 0.7999,
+      "step": 3200
+    },
+    {
+      "epoch": 7.82640586797066,
+      "grad_norm": 0.3215459745154409,
+      "learning_rate": 2.32922737459758e-06,
+      "loss": 0.5269,
+      "step": 3201
+    },
+    {
+      "epoch": 7.8288508557457215,
+      "grad_norm": 0.27854362919366615,
+      "learning_rate": 2.3279501947007973e-06,
+      "loss": 0.6282,
+      "step": 3202
+    },
+    {
+      "epoch": 7.831295843520783,
+      "grad_norm": 0.29443093097456946,
+      "learning_rate": 2.326673059919413e-06,
+      "loss": 0.4309,
+      "step": 3203
+    },
+    {
+      "epoch": 7.833740831295843,
+      "grad_norm": 0.30518041813809504,
+      "learning_rate": 2.3253959705883205e-06,
+      "loss": 0.3812,
+      "step": 3204
+    },
+    {
+      "epoch": 7.836185819070905,
+      "grad_norm": 0.32789582768920966,
+      "learning_rate": 2.324118927042401e-06,
+      "loss": 0.6941,
+      "step": 3205
+    },
+    {
+      "epoch": 7.838630806845966,
+      "grad_norm": 0.3312699980873302,
+      "learning_rate": 2.3228419296165243e-06,
+      "loss": 0.8747,
+      "step": 3206
+    },
+    {
+      "epoch": 7.841075794621027,
+      "grad_norm": 0.16034457439578667,
+      "learning_rate": 2.32156497864555e-06,
+      "loss": 0.487,
+      "step": 3207
+    },
+    {
+      "epoch": 7.843520782396088,
+      "grad_norm": 0.17820043916342254,
+      "learning_rate": 2.320288074464323e-06,
+      "loss": 0.412,
+      "step": 3208
+    },
+    {
+      "epoch": 7.84596577017115,
+      "grad_norm": 0.13428850030942405,
+      "learning_rate": 2.319011217407676e-06,
+      "loss": 0.306,
+      "step": 3209
+    },
+    {
+      "epoch": 7.84841075794621,
+      "grad_norm": 0.2769796786345839,
+      "learning_rate": 2.317734407810431e-06,
+      "loss": 0.5125,
+      "step": 3210
+    },
+    {
+      "epoch": 7.8508557457212715,
+      "grad_norm": 0.2091112752025157,
+      "learning_rate": 2.3164576460073968e-06,
+      "loss": 0.9005,
+      "step": 3211
+    },
+    {
+      "epoch": 7.853300733496333,
+      "grad_norm": 0.19643153493471402,
+      "learning_rate": 2.3151809323333695e-06,
+      "loss": 0.3908,
+      "step": 3212
+    },
+    {
+      "epoch": 7.8557457212713935,
+      "grad_norm": 0.15205628903794896,
+      "learning_rate": 2.3139042671231314e-06,
+      "loss": 0.754,
+      "step": 3213
+    },
+    {
+      "epoch": 7.858190709046455,
+      "grad_norm": 0.17306439900821313,
+      "learning_rate": 2.3126276507114554e-06,
+      "loss": 0.4363,
+      "step": 3214
+    },
+    {
+      "epoch": 7.860635696821516,
+      "grad_norm": 0.28057051863693905,
+      "learning_rate": 2.3113510834330987e-06,
+      "loss": 0.5,
+      "step": 3215
+    },
+    {
+      "epoch": 7.863080684596577,
+      "grad_norm": 0.20826017686313453,
+      "learning_rate": 2.3100745656228056e-06,
+      "loss": 0.7434,
+      "step": 3216
+    },
+    {
+      "epoch": 7.865525672371638,
+      "grad_norm": 0.31609110995425865,
+      "learning_rate": 2.30879809761531e-06,
+      "loss": 0.8658,
+      "step": 3217
+    },
+    {
+      "epoch": 7.8679706601467,
+      "grad_norm": 0.3029161121405763,
+      "learning_rate": 2.30752167974533e-06,
+      "loss": 0.4779,
+      "step": 3218
+    },
+    {
+      "epoch": 7.87041564792176,
+      "grad_norm": 0.2418983548846202,
+      "learning_rate": 2.3062453123475716e-06,
+      "loss": 0.9823,
+      "step": 3219
+    },
+    {
+      "epoch": 7.872860635696822,
+      "grad_norm": 0.18393846840435493,
+      "learning_rate": 2.304968995756727e-06,
+      "loss": 0.4982,
+      "step": 3220
+    },
+    {
+      "epoch": 7.875305623471883,
+      "grad_norm": 0.2974337019145741,
+      "learning_rate": 2.303692730307477e-06,
+      "loss": 0.5631,
+      "step": 3221
+    },
+    {
+      "epoch": 7.8777506112469435,
+      "grad_norm": 0.21946059741874746,
+      "learning_rate": 2.3024165163344865e-06,
+      "loss": 0.3665,
+      "step": 3222
+    },
+    {
+      "epoch": 7.880195599022005,
+      "grad_norm": 0.25182871704461207,
+      "learning_rate": 2.3011403541724084e-06,
+      "loss": 0.9837,
+      "step": 3223
+    },
+    {
+      "epoch": 7.882640586797066,
+      "grad_norm": 0.255778740461956,
+      "learning_rate": 2.299864244155881e-06,
+      "loss": 0.6821,
+      "step": 3224
+    },
+    {
+      "epoch": 7.885085574572127,
+      "grad_norm": 0.35564665562460485,
+      "learning_rate": 2.2985881866195305e-06,
+      "loss": 0.4816,
+      "step": 3225
+    },
+    {
+      "epoch": 7.887530562347188,
+      "grad_norm": 0.21068956579030376,
+      "learning_rate": 2.297312181897968e-06,
+      "loss": 0.4047,
+      "step": 3226
+    },
+    {
+      "epoch": 7.88997555012225,
+      "grad_norm": 0.2656364748045013,
+      "learning_rate": 2.296036230325792e-06,
+      "loss": 0.6599,
+      "step": 3227
+    },
+    {
+      "epoch": 7.89242053789731,
+      "grad_norm": 0.22146173859832524,
+      "learning_rate": 2.294760332237585e-06,
+      "loss": 0.5366,
+      "step": 3228
+    },
+    {
+      "epoch": 7.894865525672372,
+      "grad_norm": 0.18617491550036522,
+      "learning_rate": 2.2934844879679167e-06,
+      "loss": 0.5427,
+      "step": 3229
+    },
+    {
+      "epoch": 7.897310513447433,
+      "grad_norm": 0.27111302158369094,
+      "learning_rate": 2.2922086978513434e-06,
+      "loss": 0.776,
+      "step": 3230
+    },
+    {
+      "epoch": 7.899755501222494,
+      "grad_norm": 0.24665173031329884,
+      "learning_rate": 2.290932962222406e-06,
+      "loss": 0.4599,
+      "step": 3231
+    },
+    {
+      "epoch": 7.902200488997555,
+      "grad_norm": 0.24451211112846627,
+      "learning_rate": 2.289657281415632e-06,
+      "loss": 0.4712,
+      "step": 3232
+    },
+    {
+      "epoch": 7.904645476772616,
+      "grad_norm": 0.3948875188231338,
+      "learning_rate": 2.288381655765534e-06,
+      "loss": 0.5739,
+      "step": 3233
+    },
+    {
+      "epoch": 7.907090464547677,
+      "grad_norm": 0.39050795022372625,
+      "learning_rate": 2.287106085606609e-06,
+      "loss": 0.4051,
+      "step": 3234
+    },
+    {
+      "epoch": 7.909535452322738,
+      "grad_norm": 0.31680193181172295,
+      "learning_rate": 2.285830571273343e-06,
+      "loss": 0.563,
+      "step": 3235
+    },
+    {
+      "epoch": 7.9119804400978,
+      "grad_norm": 0.33373864718743146,
+      "learning_rate": 2.284555113100203e-06,
+      "loss": 0.4937,
+      "step": 3236
+    },
+    {
+      "epoch": 7.91442542787286,
+      "grad_norm": 0.24548489686152225,
+      "learning_rate": 2.2832797114216445e-06,
+      "loss": 0.7264,
+      "step": 3237
+    },
+    {
+      "epoch": 7.916870415647922,
+      "grad_norm": 0.18656614994953785,
+      "learning_rate": 2.282004366572107e-06,
+      "loss": 0.7359,
+      "step": 3238
+    },
+    {
+      "epoch": 7.919315403422983,
+      "grad_norm": 0.26130581292892213,
+      "learning_rate": 2.280729078886015e-06,
+      "loss": 0.5839,
+      "step": 3239
+    },
+    {
+      "epoch": 7.921760391198044,
+      "grad_norm": 0.228604495419214,
+      "learning_rate": 2.2794538486977767e-06,
+      "loss": 0.6304,
+      "step": 3240
+    },
+    {
+      "epoch": 7.924205378973105,
+      "grad_norm": 0.2797424308407201,
+      "learning_rate": 2.2781786763417894e-06,
+      "loss": 0.3997,
+      "step": 3241
+    },
+    {
+      "epoch": 7.9266503667481665,
+      "grad_norm": 0.2383957439768026,
+      "learning_rate": 2.27690356215243e-06,
+      "loss": 1.1731,
+      "step": 3242
+    },
+    {
+      "epoch": 7.929095354523227,
+      "grad_norm": 0.21404238275490492,
+      "learning_rate": 2.2756285064640645e-06,
+      "loss": 0.5782,
+      "step": 3243
+    },
+    {
+      "epoch": 7.931540342298288,
+      "grad_norm": 0.15510037145429412,
+      "learning_rate": 2.2743535096110394e-06,
+      "loss": 0.439,
+      "step": 3244
+    },
+    {
+      "epoch": 7.93398533007335,
+      "grad_norm": 0.2777662451941109,
+      "learning_rate": 2.2730785719276893e-06,
+      "loss": 0.6197,
+      "step": 3245
+    },
+    {
+      "epoch": 7.93643031784841,
+      "grad_norm": 0.20822673822689605,
+      "learning_rate": 2.271803693748333e-06,
+      "loss": 0.4595,
+      "step": 3246
+    },
+    {
+      "epoch": 7.938875305623472,
+      "grad_norm": 0.21044260981543203,
+      "learning_rate": 2.2705288754072724e-06,
+      "loss": 0.6432,
+      "step": 3247
+    },
+    {
+      "epoch": 7.941320293398533,
+      "grad_norm": 0.3093710070447849,
+      "learning_rate": 2.2692541172387933e-06,
+      "loss": 0.4343,
+      "step": 3248
+    },
+    {
+      "epoch": 7.943765281173594,
+      "grad_norm": 0.3270340408205683,
+      "learning_rate": 2.2679794195771667e-06,
+      "loss": 0.4525,
+      "step": 3249
+    },
+    {
+      "epoch": 7.946210268948655,
+      "grad_norm": 0.26553413509787116,
+      "learning_rate": 2.266704782756647e-06,
+      "loss": 0.5808,
+      "step": 3250
+    },
+    {
+      "epoch": 7.9486552567237165,
+      "grad_norm": 0.17919662978193723,
+      "learning_rate": 2.2654302071114746e-06,
+      "loss": 0.5499,
+      "step": 3251
+    },
+    {
+      "epoch": 7.951100244498777,
+      "grad_norm": 0.19762948581713405,
+      "learning_rate": 2.2641556929758714e-06,
+      "loss": 0.5006,
+      "step": 3252
+    },
+    {
+      "epoch": 7.9535452322738385,
+      "grad_norm": 0.2026054606749769,
+      "learning_rate": 2.262881240684044e-06,
+      "loss": 0.7326,
+      "step": 3253
+    },
+    {
+      "epoch": 7.9559902200489,
+      "grad_norm": 0.20306689087156107,
+      "learning_rate": 2.2616068505701825e-06,
+      "loss": 0.4867,
+      "step": 3254
+    },
+    {
+      "epoch": 7.95843520782396,
+      "grad_norm": 0.43497799236102697,
+      "learning_rate": 2.260332522968463e-06,
+      "loss": 0.5227,
+      "step": 3255
+    },
+    {
+      "epoch": 7.960880195599022,
+      "grad_norm": 0.25885414864675954,
+      "learning_rate": 2.259058258213041e-06,
+      "loss": 0.4592,
+      "step": 3256
+    },
+    {
+      "epoch": 7.963325183374083,
+      "grad_norm": 0.280833723150612,
+      "learning_rate": 2.2577840566380606e-06,
+      "loss": 0.3524,
+      "step": 3257
+    },
+    {
+      "epoch": 7.965770171149144,
+      "grad_norm": 0.23013473488437863,
+      "learning_rate": 2.256509918577645e-06,
+      "loss": 0.9061,
+      "step": 3258
+    },
+    {
+      "epoch": 7.968215158924205,
+      "grad_norm": 0.22034064583006813,
+      "learning_rate": 2.255235844365902e-06,
+      "loss": 0.5337,
+      "step": 3259
+    },
+    {
+      "epoch": 7.970660146699267,
+      "grad_norm": 0.31109651275930195,
+      "learning_rate": 2.2539618343369236e-06,
+      "loss": 0.6184,
+      "step": 3260
+    },
+    {
+      "epoch": 7.973105134474328,
+      "grad_norm": 0.2963571529769519,
+      "learning_rate": 2.252687888824785e-06,
+      "loss": 0.3998,
+      "step": 3261
+    },
+    {
+      "epoch": 7.9755501222493885,
+      "grad_norm": 0.26493851618912834,
+      "learning_rate": 2.251414008163543e-06,
+      "loss": 0.4372,
+      "step": 3262
+    },
+    {
+      "epoch": 7.97799511002445,
+      "grad_norm": 0.2589890421631379,
+      "learning_rate": 2.250140192687239e-06,
+      "loss": 0.4442,
+      "step": 3263
+    },
+    {
+      "epoch": 7.980440097799511,
+      "grad_norm": 0.23826811148711516,
+      "learning_rate": 2.2488664427298953e-06,
+      "loss": 0.6974,
+      "step": 3264
+    },
+    {
+      "epoch": 7.982885085574572,
+      "grad_norm": 0.29756901477902586,
+      "learning_rate": 2.2475927586255195e-06,
+      "loss": 0.4962,
+      "step": 3265
+    },
+    {
+      "epoch": 7.985330073349633,
+      "grad_norm": 0.18120215433475353,
+      "learning_rate": 2.246319140708099e-06,
+      "loss": 0.4656,
+      "step": 3266
+    },
+    {
+      "epoch": 7.987775061124695,
+      "grad_norm": 0.34522225863218786,
+      "learning_rate": 2.2450455893116084e-06,
+      "loss": 0.5383,
+      "step": 3267
+    },
+    {
+      "epoch": 7.990220048899755,
+      "grad_norm": 0.3162570980239538,
+      "learning_rate": 2.2437721047700004e-06,
+      "loss": 0.7995,
+      "step": 3268
+    },
+    {
+      "epoch": 7.992665036674817,
+      "grad_norm": 0.25755895302927456,
+      "learning_rate": 2.242498687417212e-06,
+      "loss": 0.6439,
+      "step": 3269
+    },
+    {
+      "epoch": 7.995110024449878,
+      "grad_norm": 0.17985710227091536,
+      "learning_rate": 2.241225337587162e-06,
+      "loss": 0.497,
+      "step": 3270
+    },
+    {
+      "epoch": 7.997555012224939,
+      "grad_norm": 0.2793589515870948,
+      "learning_rate": 2.2399520556137524e-06,
+      "loss": 0.9224,
+      "step": 3271
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.16483927629548378,
+      "learning_rate": 2.238678841830867e-06,
+      "loss": 0.4565,
+      "step": 3272
+    },
+    {
+      "epoch": 8.00244498777506,
+      "grad_norm": 0.20966436356805296,
+      "learning_rate": 2.237405696572371e-06,
+      "loss": 0.6797,
+      "step": 3273
+    },
+    {
+      "epoch": 8.004889975550123,
+      "grad_norm": 0.3033158216942618,
+      "learning_rate": 2.236132620172112e-06,
+      "loss": 0.4024,
+      "step": 3274
+    },
+    {
+      "epoch": 8.007334963325183,
+      "grad_norm": 0.2934947694569621,
+      "learning_rate": 2.234859612963921e-06,
+      "loss": 0.5648,
+      "step": 3275
+    },
+    {
+      "epoch": 8.009779951100244,
+      "grad_norm": 0.24567766275521513,
+      "learning_rate": 2.2335866752816085e-06,
+      "loss": 0.5651,
+      "step": 3276
+    },
+    {
+      "epoch": 8.012224938875306,
+      "grad_norm": 0.28683418643508823,
+      "learning_rate": 2.2323138074589673e-06,
+      "loss": 0.3594,
+      "step": 3277
+    },
+    {
+      "epoch": 8.014669926650367,
+      "grad_norm": 0.34829030597822436,
+      "learning_rate": 2.2310410098297746e-06,
+      "loss": 0.7458,
+      "step": 3278
+    },
+    {
+      "epoch": 8.017114914425427,
+      "grad_norm": 0.22910489223440902,
+      "learning_rate": 2.2297682827277857e-06,
+      "loss": 0.765,
+      "step": 3279
+    },
+    {
+      "epoch": 8.01955990220049,
+      "grad_norm": 0.29258807743206355,
+      "learning_rate": 2.2284956264867394e-06,
+      "loss": 0.5015,
+      "step": 3280
+    },
+    {
+      "epoch": 8.02200488997555,
+      "grad_norm": 0.2869198065234492,
+      "learning_rate": 2.227223041440355e-06,
+      "loss": 0.3421,
+      "step": 3281
+    },
+    {
+      "epoch": 8.02444987775061,
+      "grad_norm": 0.19175866844328074,
+      "learning_rate": 2.2259505279223335e-06,
+      "loss": 0.6012,
+      "step": 3282
+    },
+    {
+      "epoch": 8.026894865525673,
+      "grad_norm": 0.1944543923646408,
+      "learning_rate": 2.2246780862663567e-06,
+      "loss": 0.3996,
+      "step": 3283
+    },
+    {
+      "epoch": 8.029339853300733,
+      "grad_norm": 0.20996111324000724,
+      "learning_rate": 2.2234057168060875e-06,
+      "loss": 0.6123,
+      "step": 3284
+    },
+    {
+      "epoch": 8.031784841075794,
+      "grad_norm": 0.2589008677014097,
+      "learning_rate": 2.222133419875172e-06,
+      "loss": 0.8901,
+      "step": 3285
+    },
+    {
+      "epoch": 8.034229828850856,
+      "grad_norm": 0.17460551725506557,
+      "learning_rate": 2.2208611958072344e-06,
+      "loss": 0.5167,
+      "step": 3286
+    },
+    {
+      "epoch": 8.036674816625917,
+      "grad_norm": 0.2680406664567495,
+      "learning_rate": 2.2195890449358807e-06,
+      "loss": 0.6727,
+      "step": 3287
+    },
+    {
+      "epoch": 8.039119804400977,
+      "grad_norm": 0.1965472965844495,
+      "learning_rate": 2.218316967594698e-06,
+      "loss": 0.7956,
+      "step": 3288
+    },
+    {
+      "epoch": 8.04156479217604,
+      "grad_norm": 0.26894507983137267,
+      "learning_rate": 2.2170449641172545e-06,
+      "loss": 0.4245,
+      "step": 3289
+    },
+    {
+      "epoch": 8.0440097799511,
+      "grad_norm": 0.2614031105042637,
+      "learning_rate": 2.215773034837099e-06,
+      "loss": 0.407,
+      "step": 3290
+    },
+    {
+      "epoch": 8.04645476772616,
+      "grad_norm": 0.28128198962168793,
+      "learning_rate": 2.214501180087761e-06,
+      "loss": 0.4564,
+      "step": 3291
+    },
+    {
+      "epoch": 8.048899755501223,
+      "grad_norm": 0.37807689248766874,
+      "learning_rate": 2.213229400202748e-06,
+      "loss": 0.3756,
+      "step": 3292
+    },
+    {
+      "epoch": 8.051344743276283,
+      "grad_norm": 0.24016845795409608,
+      "learning_rate": 2.211957695515552e-06,
+      "loss": 0.8035,
+      "step": 3293
+    },
+    {
+      "epoch": 8.053789731051344,
+      "grad_norm": 0.23773057321647217,
+      "learning_rate": 2.2106860663596404e-06,
+      "loss": 0.8945,
+      "step": 3294
+    },
+    {
+      "epoch": 8.056234718826406,
+      "grad_norm": 0.2945353519289615,
+      "learning_rate": 2.2094145130684657e-06,
+      "loss": 0.8274,
+      "step": 3295
+    },
+    {
+      "epoch": 8.058679706601467,
+      "grad_norm": 0.20002552096846707,
+      "learning_rate": 2.208143035975458e-06,
+      "loss": 0.6665,
+      "step": 3296
+    },
+    {
+      "epoch": 8.061124694376527,
+      "grad_norm": 0.2399752748552482,
+      "learning_rate": 2.206871635414027e-06,
+      "loss": 0.8377,
+      "step": 3297
+    },
+    {
+      "epoch": 8.06356968215159,
+      "grad_norm": 0.297450056383463,
+      "learning_rate": 2.2056003117175628e-06,
+      "loss": 0.4946,
+      "step": 3298
+    },
+    {
+      "epoch": 8.06601466992665,
+      "grad_norm": 0.24031986741383574,
+      "learning_rate": 2.2043290652194354e-06,
+      "loss": 0.6912,
+      "step": 3299
+    },
+    {
+      "epoch": 8.06845965770171,
+      "grad_norm": 0.3109731827357909,
+      "learning_rate": 2.2030578962529968e-06,
+      "loss": 1.1028,
+      "step": 3300
+    },
+    {
+      "epoch": 8.070904645476773,
+      "grad_norm": 0.2882215914484284,
+      "learning_rate": 2.201786805151575e-06,
+      "loss": 0.4467,
+      "step": 3301
+    },
+    {
+      "epoch": 8.073349633251834,
+      "grad_norm": 0.18925992291693913,
+      "learning_rate": 2.2005157922484793e-06,
+      "loss": 0.4117,
+      "step": 3302
+    },
+    {
+      "epoch": 8.075794621026894,
+      "grad_norm": 0.24852423088940928,
+      "learning_rate": 2.199244857876998e-06,
+      "loss": 0.4681,
+      "step": 3303
+    },
+    {
+      "epoch": 8.078239608801956,
+      "grad_norm": 0.17809239993774464,
+      "learning_rate": 2.1979740023704e-06,
+      "loss": 0.3467,
+      "step": 3304
+    },
+    {
+      "epoch": 8.080684596577017,
+      "grad_norm": 0.19610735340762786,
+      "learning_rate": 2.1967032260619326e-06,
+      "loss": 0.7391,
+      "step": 3305
+    },
+    {
+      "epoch": 8.083129584352077,
+      "grad_norm": 0.24257174498061096,
+      "learning_rate": 2.1954325292848224e-06,
+      "loss": 0.4964,
+      "step": 3306
+    },
+    {
+      "epoch": 8.08557457212714,
+      "grad_norm": 0.2676488731244071,
+      "learning_rate": 2.194161912372275e-06,
+      "loss": 0.3909,
+      "step": 3307
+    },
+    {
+      "epoch": 8.0880195599022,
+      "grad_norm": 0.1676290116316388,
+      "learning_rate": 2.1928913756574745e-06,
+      "loss": 0.4186,
+      "step": 3308
+    },
+    {
+      "epoch": 8.09046454767726,
+      "grad_norm": 0.3654213870694616,
+      "learning_rate": 2.1916209194735847e-06,
+      "loss": 0.2666,
+      "step": 3309
+    },
+    {
+      "epoch": 8.092909535452323,
+      "grad_norm": 0.35652037365058786,
+      "learning_rate": 2.1903505441537504e-06,
+      "loss": 0.4306,
+      "step": 3310
+    },
+    {
+      "epoch": 8.095354523227384,
+      "grad_norm": 0.21613964975179473,
+      "learning_rate": 2.1890802500310917e-06,
+      "loss": 0.9138,
+      "step": 3311
+    },
+    {
+      "epoch": 8.097799511002446,
+      "grad_norm": 0.3772319397740884,
+      "learning_rate": 2.1878100374387088e-06,
+      "loss": 0.3566,
+      "step": 3312
+    },
+    {
+      "epoch": 8.100244498777506,
+      "grad_norm": 0.1957001359507549,
+      "learning_rate": 2.18653990670968e-06,
+      "loss": 0.3775,
+      "step": 3313
+    },
+    {
+      "epoch": 8.102689486552567,
+      "grad_norm": 0.1562721947646665,
+      "learning_rate": 2.1852698581770635e-06,
+      "loss": 0.3096,
+      "step": 3314
+    },
+    {
+      "epoch": 8.10513447432763,
+      "grad_norm": 0.3130643678278964,
+      "learning_rate": 2.1839998921738952e-06,
+      "loss": 0.6474,
+      "step": 3315
+    },
+    {
+      "epoch": 8.10757946210269,
+      "grad_norm": 0.2728111836077955,
+      "learning_rate": 2.1827300090331882e-06,
+      "loss": 0.4167,
+      "step": 3316
+    },
+    {
+      "epoch": 8.11002444987775,
+      "grad_norm": 0.13170344815930668,
+      "learning_rate": 2.181460209087936e-06,
+      "loss": 0.3108,
+      "step": 3317
+    },
+    {
+      "epoch": 8.112469437652813,
+      "grad_norm": 0.1990113643389976,
+      "learning_rate": 2.1801904926711086e-06,
+      "loss": 0.902,
+      "step": 3318
+    },
+    {
+      "epoch": 8.114914425427873,
+      "grad_norm": 0.3466188007064488,
+      "learning_rate": 2.1789208601156546e-06,
+      "loss": 0.8578,
+      "step": 3319
+    },
+    {
+      "epoch": 8.117359413202934,
+      "grad_norm": 0.26783782390502947,
+      "learning_rate": 2.1776513117545005e-06,
+      "loss": 0.5227,
+      "step": 3320
+    },
+    {
+      "epoch": 8.119804400977996,
+      "grad_norm": 0.22102632472072922,
+      "learning_rate": 2.1763818479205524e-06,
+      "loss": 0.6049,
+      "step": 3321
+    },
+    {
+      "epoch": 8.122249388753056,
+      "grad_norm": 0.3476238586802372,
+      "learning_rate": 2.175112468946692e-06,
+      "loss": 0.5577,
+      "step": 3322
+    },
+    {
+      "epoch": 8.124694376528117,
+      "grad_norm": 0.32182590725407917,
+      "learning_rate": 2.1738431751657786e-06,
+      "loss": 0.6853,
+      "step": 3323
+    },
+    {
+      "epoch": 8.12713936430318,
+      "grad_norm": 0.2402886490455442,
+      "learning_rate": 2.1725739669106514e-06,
+      "loss": 0.3808,
+      "step": 3324
+    },
+    {
+      "epoch": 8.12958435207824,
+      "grad_norm": 0.28320159314786564,
+      "learning_rate": 2.1713048445141258e-06,
+      "loss": 0.676,
+      "step": 3325
+    },
+    {
+      "epoch": 8.1320293398533,
+      "grad_norm": 0.3037174715692259,
+      "learning_rate": 2.1700358083089937e-06,
+      "loss": 0.4805,
+      "step": 3326
+    },
+    {
+      "epoch": 8.134474327628363,
+      "grad_norm": 0.22168510148589918,
+      "learning_rate": 2.1687668586280264e-06,
+      "loss": 0.3918,
+      "step": 3327
+    },
+    {
+      "epoch": 8.136919315403423,
+      "grad_norm": 0.16482402201319796,
+      "learning_rate": 2.1674979958039702e-06,
+      "loss": 0.4132,
+      "step": 3328
+    },
+    {
+      "epoch": 8.139364303178484,
+      "grad_norm": 0.25571129611684856,
+      "learning_rate": 2.1662292201695517e-06,
+      "loss": 0.8492,
+      "step": 3329
+    },
+    {
+      "epoch": 8.141809290953546,
+      "grad_norm": 0.2024948433133835,
+      "learning_rate": 2.1649605320574716e-06,
+      "loss": 0.8337,
+      "step": 3330
+    },
+    {
+      "epoch": 8.144254278728607,
+      "grad_norm": 0.23716956650170803,
+      "learning_rate": 2.1636919318004095e-06,
+      "loss": 0.945,
+      "step": 3331
+    },
+    {
+      "epoch": 8.146699266503667,
+      "grad_norm": 0.22239581704359357,
+      "learning_rate": 2.162423419731022e-06,
+      "loss": 0.5735,
+      "step": 3332
+    },
+    {
+      "epoch": 8.14914425427873,
+      "grad_norm": 0.27260363303174445,
+      "learning_rate": 2.1611549961819405e-06,
+      "loss": 0.4612,
+      "step": 3333
+    },
+    {
+      "epoch": 8.15158924205379,
+      "grad_norm": 0.2045198430600031,
+      "learning_rate": 2.1598866614857764e-06,
+      "loss": 0.9936,
+      "step": 3334
+    },
+    {
+      "epoch": 8.15403422982885,
+      "grad_norm": 0.260414411549753,
+      "learning_rate": 2.1586184159751153e-06,
+      "loss": 0.6533,
+      "step": 3335
+    },
+    {
+      "epoch": 8.156479217603913,
+      "grad_norm": 0.22760596742884684,
+      "learning_rate": 2.15735025998252e-06,
+      "loss": 1.1346,
+      "step": 3336
+    },
+    {
+      "epoch": 8.158924205378973,
+      "grad_norm": 0.2593018291387644,
+      "learning_rate": 2.1560821938405304e-06,
+      "loss": 0.511,
+      "step": 3337
+    },
+    {
+      "epoch": 8.161369193154034,
+      "grad_norm": 0.2185253968022397,
+      "learning_rate": 2.1548142178816625e-06,
+      "loss": 0.5877,
+      "step": 3338
+    },
+    {
+      "epoch": 8.163814180929096,
+      "grad_norm": 0.26085755194928273,
+      "learning_rate": 2.1535463324384083e-06,
+      "loss": 0.4246,
+      "step": 3339
+    },
+    {
+      "epoch": 8.166259168704157,
+      "grad_norm": 0.26407723762602575,
+      "learning_rate": 2.1522785378432366e-06,
+      "loss": 0.9678,
+      "step": 3340
+    },
+    {
+      "epoch": 8.168704156479217,
+      "grad_norm": 0.23066863650752187,
+      "learning_rate": 2.1510108344285923e-06,
+      "loss": 0.4286,
+      "step": 3341
+    },
+    {
+      "epoch": 8.17114914425428,
+      "grad_norm": 0.3533463113288656,
+      "learning_rate": 2.149743222526895e-06,
+      "loss": 0.449,
+      "step": 3342
+    },
+    {
+      "epoch": 8.17359413202934,
+      "grad_norm": 0.32523346008625154,
+      "learning_rate": 2.148475702470543e-06,
+      "loss": 0.6602,
+      "step": 3343
+    },
+    {
+      "epoch": 8.1760391198044,
+      "grad_norm": 0.2218149508281891,
+      "learning_rate": 2.147208274591909e-06,
+      "loss": 0.3142,
+      "step": 3344
+    },
+    {
+      "epoch": 8.178484107579463,
+      "grad_norm": 0.27337454581191045,
+      "learning_rate": 2.1459409392233414e-06,
+      "loss": 0.3963,
+      "step": 3345
+    },
+    {
+      "epoch": 8.180929095354523,
+      "grad_norm": 0.33572395258140497,
+      "learning_rate": 2.1446736966971647e-06,
+      "loss": 0.4172,
+      "step": 3346
+    },
+    {
+      "epoch": 8.183374083129584,
+      "grad_norm": 0.18174562819312717,
+      "learning_rate": 2.1434065473456778e-06,
+      "loss": 0.5893,
+      "step": 3347
+    },
+    {
+      "epoch": 8.185819070904646,
+      "grad_norm": 0.3187460703982711,
+      "learning_rate": 2.142139491501158e-06,
+      "loss": 0.3628,
+      "step": 3348
+    },
+    {
+      "epoch": 8.188264058679707,
+      "grad_norm": 0.229579750361833,
+      "learning_rate": 2.140872529495855e-06,
+      "loss": 0.6271,
+      "step": 3349
+    },
+    {
+      "epoch": 8.190709046454767,
+      "grad_norm": 0.21759990670649823,
+      "learning_rate": 2.139605661661996e-06,
+      "loss": 0.4626,
+      "step": 3350
+    },
+    {
+      "epoch": 8.19315403422983,
+      "grad_norm": 0.2425253018586069,
+      "learning_rate": 2.138338888331782e-06,
+      "loss": 0.5089,
+      "step": 3351
+    },
+    {
+      "epoch": 8.19559902200489,
+      "grad_norm": 0.2755832683214657,
+      "learning_rate": 2.1370722098373896e-06,
+      "loss": 0.6217,
+      "step": 3352
+    },
+    {
+      "epoch": 8.19804400977995,
+      "grad_norm": 0.1948822978017345,
+      "learning_rate": 2.1358056265109724e-06,
+      "loss": 0.5413,
+      "step": 3353
+    },
+    {
+      "epoch": 8.200488997555013,
+      "grad_norm": 0.25747125288334,
+      "learning_rate": 2.134539138684657e-06,
+      "loss": 0.4911,
+      "step": 3354
+    },
+    {
+      "epoch": 8.202933985330073,
+      "grad_norm": 0.33537374928304037,
+      "learning_rate": 2.1332727466905457e-06,
+      "loss": 0.5571,
+      "step": 3355
+    },
+    {
+      "epoch": 8.205378973105134,
+      "grad_norm": 0.300579616927296,
+      "learning_rate": 2.132006450860715e-06,
+      "loss": 0.6801,
+      "step": 3356
+    },
+    {
+      "epoch": 8.207823960880196,
+      "grad_norm": 0.22554301242116664,
+      "learning_rate": 2.1307402515272164e-06,
+      "loss": 0.5255,
+      "step": 3357
+    },
+    {
+      "epoch": 8.210268948655257,
+      "grad_norm": 0.1858291724045866,
+      "learning_rate": 2.129474149022077e-06,
+      "loss": 0.5286,
+      "step": 3358
+    },
+    {
+      "epoch": 8.212713936430317,
+      "grad_norm": 0.18648916050311798,
+      "learning_rate": 2.1282081436772975e-06,
+      "loss": 0.5646,
+      "step": 3359
+    },
+    {
+      "epoch": 8.21515892420538,
+      "grad_norm": 0.25587507494362327,
+      "learning_rate": 2.1269422358248537e-06,
+      "loss": 0.4062,
+      "step": 3360
+    },
+    {
+      "epoch": 8.21760391198044,
+      "grad_norm": 0.2925418272781181,
+      "learning_rate": 2.125676425796695e-06,
+      "loss": 0.4998,
+      "step": 3361
+    },
+    {
+      "epoch": 8.2200488997555,
+      "grad_norm": 0.2932527219095575,
+      "learning_rate": 2.1244107139247467e-06,
+      "loss": 0.597,
+      "step": 3362
+    },
+    {
+      "epoch": 8.222493887530563,
+      "grad_norm": 0.1951304442942124,
+      "learning_rate": 2.1231451005409057e-06,
+      "loss": 0.4422,
+      "step": 3363
+    },
+    {
+      "epoch": 8.224938875305623,
+      "grad_norm": 0.19665287770304266,
+      "learning_rate": 2.1218795859770473e-06,
+      "loss": 0.5439,
+      "step": 3364
+    },
+    {
+      "epoch": 8.227383863080684,
+      "grad_norm": 0.44932913170944594,
+      "learning_rate": 2.120614170565017e-06,
+      "loss": 0.5757,
+      "step": 3365
+    },
+    {
+      "epoch": 8.229828850855746,
+      "grad_norm": 0.2970313936281067,
+      "learning_rate": 2.119348854636636e-06,
+      "loss": 0.4703,
+      "step": 3366
+    },
+    {
+      "epoch": 8.232273838630807,
+      "grad_norm": 0.4727216724489321,
+      "learning_rate": 2.1180836385236987e-06,
+      "loss": 0.6709,
+      "step": 3367
+    },
+    {
+      "epoch": 8.234718826405867,
+      "grad_norm": 0.21503357278070417,
+      "learning_rate": 2.116818522557974e-06,
+      "loss": 0.6792,
+      "step": 3368
+    },
+    {
+      "epoch": 8.23716381418093,
+      "grad_norm": 0.20375855782870206,
+      "learning_rate": 2.1155535070712043e-06,
+      "loss": 0.8595,
+      "step": 3369
+    },
+    {
+      "epoch": 8.23960880195599,
+      "grad_norm": 0.22397210605989334,
+      "learning_rate": 2.114288592395106e-06,
+      "loss": 0.381,
+      "step": 3370
+    },
+    {
+      "epoch": 8.24205378973105,
+      "grad_norm": 0.19497901458370331,
+      "learning_rate": 2.1130237788613673e-06,
+      "loss": 0.6178,
+      "step": 3371
+    },
+    {
+      "epoch": 8.244498777506113,
+      "grad_norm": 0.30867577787826517,
+      "learning_rate": 2.1117590668016525e-06,
+      "loss": 0.6962,
+      "step": 3372
+    },
+    {
+      "epoch": 8.246943765281173,
+      "grad_norm": 0.19803752531891974,
+      "learning_rate": 2.110494456547598e-06,
+      "loss": 0.4969,
+      "step": 3373
+    },
+    {
+      "epoch": 8.249388753056234,
+      "grad_norm": 0.2169016215344782,
+      "learning_rate": 2.1092299484308127e-06,
+      "loss": 0.7604,
+      "step": 3374
+    },
+    {
+      "epoch": 8.251833740831296,
+      "grad_norm": 0.33423973079752384,
+      "learning_rate": 2.107965542782881e-06,
+      "loss": 0.5164,
+      "step": 3375
+    },
+    {
+      "epoch": 8.254278728606357,
+      "grad_norm": 0.30088352178476596,
+      "learning_rate": 2.1067012399353578e-06,
+      "loss": 0.7118,
+      "step": 3376
+    },
+    {
+      "epoch": 8.256723716381417,
+      "grad_norm": 0.3345517177433198,
+      "learning_rate": 2.1054370402197723e-06,
+      "loss": 0.512,
+      "step": 3377
+    },
+    {
+      "epoch": 8.25916870415648,
+      "grad_norm": 0.19397875087021696,
+      "learning_rate": 2.1041729439676275e-06,
+      "loss": 0.6044,
+      "step": 3378
+    },
+    {
+      "epoch": 8.26161369193154,
+      "grad_norm": 0.2796766997854982,
+      "learning_rate": 2.1029089515103977e-06,
+      "loss": 0.9442,
+      "step": 3379
+    },
+    {
+      "epoch": 8.2640586797066,
+      "grad_norm": 0.2510208419655631,
+      "learning_rate": 2.101645063179531e-06,
+      "loss": 0.484,
+      "step": 3380
+    },
+    {
+      "epoch": 8.266503667481663,
+      "grad_norm": 0.2829611678227683,
+      "learning_rate": 2.100381279306447e-06,
+      "loss": 0.3747,
+      "step": 3381
+    },
+    {
+      "epoch": 8.268948655256724,
+      "grad_norm": 0.1780699095012843,
+      "learning_rate": 2.09911760022254e-06,
+      "loss": 0.3773,
+      "step": 3382
+    },
+    {
+      "epoch": 8.271393643031784,
+      "grad_norm": 0.2202648710978936,
+      "learning_rate": 2.0978540262591745e-06,
+      "loss": 0.8265,
+      "step": 3383
+    },
+    {
+      "epoch": 8.273838630806846,
+      "grad_norm": 0.26987873896587145,
+      "learning_rate": 2.0965905577476886e-06,
+      "loss": 0.8246,
+      "step": 3384
+    },
+    {
+      "epoch": 8.276283618581907,
+      "grad_norm": 0.2071182980440185,
+      "learning_rate": 2.095327195019394e-06,
+      "loss": 0.378,
+      "step": 3385
+    },
+    {
+      "epoch": 8.278728606356967,
+      "grad_norm": 0.22715977351774883,
+      "learning_rate": 2.094063938405572e-06,
+      "loss": 0.5309,
+      "step": 3386
+    },
+    {
+      "epoch": 8.28117359413203,
+      "grad_norm": 0.2091454551842603,
+      "learning_rate": 2.0928007882374775e-06,
+      "loss": 0.3418,
+      "step": 3387
+    },
+    {
+      "epoch": 8.28361858190709,
+      "grad_norm": 0.23582710552066402,
+      "learning_rate": 2.0915377448463383e-06,
+      "loss": 0.6727,
+      "step": 3388
+    },
+    {
+      "epoch": 8.28606356968215,
+      "grad_norm": 0.2877486045425973,
+      "learning_rate": 2.090274808563352e-06,
+      "loss": 0.594,
+      "step": 3389
+    },
+    {
+      "epoch": 8.288508557457213,
+      "grad_norm": 0.22243680534312027,
+      "learning_rate": 2.0890119797196906e-06,
+      "loss": 0.5697,
+      "step": 3390
+    },
+    {
+      "epoch": 8.290953545232274,
+      "grad_norm": 0.3072022173931324,
+      "learning_rate": 2.0877492586464955e-06,
+      "loss": 0.8175,
+      "step": 3391
+    },
+    {
+      "epoch": 8.293398533007334,
+      "grad_norm": 0.31225453177874035,
+      "learning_rate": 2.086486645674883e-06,
+      "loss": 0.6288,
+      "step": 3392
+    },
+    {
+      "epoch": 8.295843520782396,
+      "grad_norm": 0.2338806616940788,
+      "learning_rate": 2.085224141135937e-06,
+      "loss": 0.5839,
+      "step": 3393
+    },
+    {
+      "epoch": 8.298288508557457,
+      "grad_norm": 0.3356195927529672,
+      "learning_rate": 2.0839617453607165e-06,
+      "loss": 0.4523,
+      "step": 3394
+    },
+    {
+      "epoch": 8.30073349633252,
+      "grad_norm": 0.3090457311145743,
+      "learning_rate": 2.0826994586802483e-06,
+      "loss": 0.7139,
+      "step": 3395
+    },
+    {
+      "epoch": 8.30317848410758,
+      "grad_norm": 0.21844868274931487,
+      "learning_rate": 2.0814372814255363e-06,
+      "loss": 0.3581,
+      "step": 3396
+    },
+    {
+      "epoch": 8.30562347188264,
+      "grad_norm": 0.32577250562808757,
+      "learning_rate": 2.08017521392755e-06,
+      "loss": 0.3167,
+      "step": 3397
+    },
+    {
+      "epoch": 8.3080684596577,
+      "grad_norm": 0.23337886064912508,
+      "learning_rate": 2.0789132565172334e-06,
+      "loss": 0.6412,
+      "step": 3398
+    },
+    {
+      "epoch": 8.310513447432763,
+      "grad_norm": 0.17878435674897483,
+      "learning_rate": 2.0776514095255004e-06,
+      "loss": 0.6377,
+      "step": 3399
+    },
+    {
+      "epoch": 8.312958435207824,
+      "grad_norm": 0.31572463304418785,
+      "learning_rate": 2.076389673283236e-06,
+      "loss": 0.9649,
+      "step": 3400
+    },
+    {
+      "epoch": 8.315403422982886,
+      "grad_norm": 0.2675760035307563,
+      "learning_rate": 2.075128048121295e-06,
+      "loss": 0.8356,
+      "step": 3401
+    },
+    {
+      "epoch": 8.317848410757946,
+      "grad_norm": 0.2952103395822672,
+      "learning_rate": 2.0738665343705067e-06,
+      "loss": 0.6255,
+      "step": 3402
+    },
+    {
+      "epoch": 8.320293398533007,
+      "grad_norm": 0.16655698430568028,
+      "learning_rate": 2.072605132361668e-06,
+      "loss": 0.4255,
+      "step": 3403
+    },
+    {
+      "epoch": 8.32273838630807,
+      "grad_norm": 0.2050784296849249,
+      "learning_rate": 2.0713438424255472e-06,
+      "loss": 0.4992,
+      "step": 3404
+    },
+    {
+      "epoch": 8.32518337408313,
+      "grad_norm": 0.2273616927984424,
+      "learning_rate": 2.070082664892883e-06,
+      "loss": 0.4345,
+      "step": 3405
+    },
+    {
+      "epoch": 8.32762836185819,
+      "grad_norm": 0.28025948483705754,
+      "learning_rate": 2.068821600094385e-06,
+      "loss": 0.4779,
+      "step": 3406
+    },
+    {
+      "epoch": 8.330073349633253,
+      "grad_norm": 0.27508873012165336,
+      "learning_rate": 2.0675606483607343e-06,
+      "loss": 0.8361,
+      "step": 3407
+    },
+    {
+      "epoch": 8.332518337408313,
+      "grad_norm": 0.1781506826075628,
+      "learning_rate": 2.0662998100225806e-06,
+      "loss": 0.3488,
+      "step": 3408
+    },
+    {
+      "epoch": 8.334963325183374,
+      "grad_norm": 0.16844370223706442,
+      "learning_rate": 2.0650390854105454e-06,
+      "loss": 0.3299,
+      "step": 3409
+    },
+    {
+      "epoch": 8.337408312958436,
+      "grad_norm": 0.32402860328147687,
+      "learning_rate": 2.0637784748552183e-06,
+      "loss": 0.445,
+      "step": 3410
+    },
+    {
+      "epoch": 8.339853300733497,
+      "grad_norm": 0.2819396934826585,
+      "learning_rate": 2.062517978687161e-06,
+      "loss": 0.4548,
+      "step": 3411
+    },
+    {
+      "epoch": 8.342298288508557,
+      "grad_norm": 0.3211030707301094,
+      "learning_rate": 2.061257597236904e-06,
+      "loss": 0.4618,
+      "step": 3412
+    },
+    {
+      "epoch": 8.34474327628362,
+      "grad_norm": 0.19514637441586846,
+      "learning_rate": 2.0599973308349488e-06,
+      "loss": 0.5277,
+      "step": 3413
+    },
+    {
+      "epoch": 8.34718826405868,
+      "grad_norm": 0.2529062358165802,
+      "learning_rate": 2.0587371798117656e-06,
+      "loss": 0.4326,
+      "step": 3414
+    },
+    {
+      "epoch": 8.34963325183374,
+      "grad_norm": 0.3118818630202157,
+      "learning_rate": 2.0574771444977944e-06,
+      "loss": 0.6284,
+      "step": 3415
+    },
+    {
+      "epoch": 8.352078239608803,
+      "grad_norm": 0.29677829176884524,
+      "learning_rate": 2.0562172252234465e-06,
+      "loss": 0.6335,
+      "step": 3416
+    },
+    {
+      "epoch": 8.354523227383863,
+      "grad_norm": 0.19344817468517547,
+      "learning_rate": 2.0549574223191e-06,
+      "loss": 0.6191,
+      "step": 3417
+    },
+    {
+      "epoch": 8.356968215158924,
+      "grad_norm": 0.25026792360165995,
+      "learning_rate": 2.0536977361151056e-06,
+      "loss": 0.6727,
+      "step": 3418
+    },
+    {
+      "epoch": 8.359413202933986,
+      "grad_norm": 0.17765642151833175,
+      "learning_rate": 2.0524381669417814e-06,
+      "loss": 0.8783,
+      "step": 3419
+    },
+    {
+      "epoch": 8.361858190709047,
+      "grad_norm": 0.20665953394130498,
+      "learning_rate": 2.0511787151294155e-06,
+      "loss": 1.0192,
+      "step": 3420
+    },
+    {
+      "epoch": 8.364303178484107,
+      "grad_norm": 0.23322126659676687,
+      "learning_rate": 2.049919381008264e-06,
+      "loss": 0.4594,
+      "step": 3421
+    },
+    {
+      "epoch": 8.36674816625917,
+      "grad_norm": 0.22952978689422676,
+      "learning_rate": 2.048660164908554e-06,
+      "loss": 0.8095,
+      "step": 3422
+    },
+    {
+      "epoch": 8.36919315403423,
+      "grad_norm": 0.1647336948038653,
+      "learning_rate": 2.047401067160481e-06,
+      "loss": 0.4817,
+      "step": 3423
+    },
+    {
+      "epoch": 8.37163814180929,
+      "grad_norm": 0.22749716011141102,
+      "learning_rate": 2.0461420880942083e-06,
+      "loss": 0.6213,
+      "step": 3424
+    },
+    {
+      "epoch": 8.374083129584353,
+      "grad_norm": 0.3723366899042355,
+      "learning_rate": 2.0448832280398692e-06,
+      "loss": 0.4867,
+      "step": 3425
+    },
+    {
+      "epoch": 8.376528117359413,
+      "grad_norm": 0.38826891638444855,
+      "learning_rate": 2.0436244873275665e-06,
+      "loss": 0.3349,
+      "step": 3426
+    },
+    {
+      "epoch": 8.378973105134474,
+      "grad_norm": 0.23517561514999716,
+      "learning_rate": 2.0423658662873692e-06,
+      "loss": 0.4286,
+      "step": 3427
+    },
+    {
+      "epoch": 8.381418092909536,
+      "grad_norm": 0.2427328008607717,
+      "learning_rate": 2.041107365249319e-06,
+      "loss": 0.3834,
+      "step": 3428
+    },
+    {
+      "epoch": 8.383863080684597,
+      "grad_norm": 0.2808909523934966,
+      "learning_rate": 2.0398489845434218e-06,
+      "loss": 0.545,
+      "step": 3429
+    },
+    {
+      "epoch": 8.386308068459657,
+      "grad_norm": 0.22841285788294444,
+      "learning_rate": 2.0385907244996546e-06,
+      "loss": 0.7469,
+      "step": 3430
+    },
+    {
+      "epoch": 8.38875305623472,
+      "grad_norm": 0.22083418274828928,
+      "learning_rate": 2.0373325854479613e-06,
+      "loss": 0.4581,
+      "step": 3431
+    },
+    {
+      "epoch": 8.39119804400978,
+      "grad_norm": 0.2040634099796583,
+      "learning_rate": 2.036074567718255e-06,
+      "loss": 0.3826,
+      "step": 3432
+    },
+    {
+      "epoch": 8.39364303178484,
+      "grad_norm": 0.29919685022484543,
+      "learning_rate": 2.034816671640418e-06,
+      "loss": 0.5029,
+      "step": 3433
+    },
+    {
+      "epoch": 8.396088019559903,
+      "grad_norm": 0.3630374426822063,
+      "learning_rate": 2.0335588975442976e-06,
+      "loss": 0.7438,
+      "step": 3434
+    },
+    {
+      "epoch": 8.398533007334963,
+      "grad_norm": 0.26018754246078385,
+      "learning_rate": 2.0323012457597117e-06,
+      "loss": 0.4561,
+      "step": 3435
+    },
+    {
+      "epoch": 8.400977995110024,
+      "grad_norm": 0.3552612284614733,
+      "learning_rate": 2.0310437166164455e-06,
+      "loss": 0.4431,
+      "step": 3436
+    },
+    {
+      "epoch": 8.403422982885086,
+      "grad_norm": 0.21130114213829806,
+      "learning_rate": 2.0297863104442524e-06,
+      "loss": 0.7768,
+      "step": 3437
+    },
+    {
+      "epoch": 8.405867970660147,
+      "grad_norm": 0.27095450763579304,
+      "learning_rate": 2.028529027572851e-06,
+      "loss": 0.7652,
+      "step": 3438
+    },
+    {
+      "epoch": 8.408312958435207,
+      "grad_norm": 0.23179304885781649,
+      "learning_rate": 2.0272718683319324e-06,
+      "loss": 0.591,
+      "step": 3439
+    },
+    {
+      "epoch": 8.41075794621027,
+      "grad_norm": 0.22587948916367723,
+      "learning_rate": 2.0260148330511516e-06,
+      "loss": 0.6797,
+      "step": 3440
+    },
+    {
+      "epoch": 8.41320293398533,
+      "grad_norm": 0.1663087905902326,
+      "learning_rate": 2.024757922060131e-06,
+      "loss": 0.6021,
+      "step": 3441
+    },
+    {
+      "epoch": 8.41564792176039,
+      "grad_norm": 0.23632069414957482,
+      "learning_rate": 2.0235011356884626e-06,
+      "loss": 0.4362,
+      "step": 3442
+    },
+    {
+      "epoch": 8.418092909535453,
+      "grad_norm": 0.2929898914490159,
+      "learning_rate": 2.022244474265704e-06,
+      "loss": 0.4028,
+      "step": 3443
+    },
+    {
+      "epoch": 8.420537897310513,
+      "grad_norm": 0.23262149448981922,
+      "learning_rate": 2.020987938121381e-06,
+      "loss": 0.7849,
+      "step": 3444
+    },
+    {
+      "epoch": 8.422982885085574,
+      "grad_norm": 0.2327857269211227,
+      "learning_rate": 2.019731527584985e-06,
+      "loss": 0.4748,
+      "step": 3445
+    },
+    {
+      "epoch": 8.425427872860636,
+      "grad_norm": 0.2738723621020514,
+      "learning_rate": 2.0184752429859773e-06,
+      "loss": 0.3634,
+      "step": 3446
+    },
+    {
+      "epoch": 8.427872860635697,
+      "grad_norm": 0.21150595229738153,
+      "learning_rate": 2.0172190846537838e-06,
+      "loss": 0.7538,
+      "step": 3447
+    },
+    {
+      "epoch": 8.430317848410757,
+      "grad_norm": 0.29809412622435655,
+      "learning_rate": 2.0159630529177975e-06,
+      "loss": 0.5818,
+      "step": 3448
+    },
+    {
+      "epoch": 8.43276283618582,
+      "grad_norm": 0.2140657593249767,
+      "learning_rate": 2.0147071481073784e-06,
+      "loss": 0.9393,
+      "step": 3449
+    },
+    {
+      "epoch": 8.43520782396088,
+      "grad_norm": 0.25685127894637233,
+      "learning_rate": 2.0134513705518545e-06,
+      "loss": 0.4599,
+      "step": 3450
+    },
+    {
+      "epoch": 8.43765281173594,
+      "grad_norm": 0.3010748797119755,
+      "learning_rate": 2.012195720580519e-06,
+      "loss": 0.7133,
+      "step": 3451
+    },
+    {
+      "epoch": 8.440097799511003,
+      "grad_norm": 0.28334849859695055,
+      "learning_rate": 2.0109401985226325e-06,
+      "loss": 0.5436,
+      "step": 3452
+    },
+    {
+      "epoch": 8.442542787286063,
+      "grad_norm": 0.2844726500699958,
+      "learning_rate": 2.0096848047074207e-06,
+      "loss": 0.7464,
+      "step": 3453
+    },
+    {
+      "epoch": 8.444987775061124,
+      "grad_norm": 0.2720857460371934,
+      "learning_rate": 2.0084295394640777e-06,
+      "loss": 0.8159,
+      "step": 3454
+    },
+    {
+      "epoch": 8.447432762836186,
+      "grad_norm": 0.15513839200603916,
+      "learning_rate": 2.0071744031217614e-06,
+      "loss": 0.555,
+      "step": 3455
+    },
+    {
+      "epoch": 8.449877750611247,
+      "grad_norm": 0.19640942693321717,
+      "learning_rate": 2.0059193960095983e-06,
+      "loss": 0.5562,
+      "step": 3456
+    },
+    {
+      "epoch": 8.452322738386307,
+      "grad_norm": 0.26330371947760495,
+      "learning_rate": 2.0046645184566802e-06,
+      "loss": 0.3978,
+      "step": 3457
+    },
+    {
+      "epoch": 8.45476772616137,
+      "grad_norm": 0.24033992797584902,
+      "learning_rate": 2.003409770792064e-06,
+      "loss": 0.6038,
+      "step": 3458
+    },
+    {
+      "epoch": 8.45721271393643,
+      "grad_norm": 0.18700285439486017,
+      "learning_rate": 2.0021551533447724e-06,
+      "loss": 0.5415,
+      "step": 3459
+    },
+    {
+      "epoch": 8.45965770171149,
+      "grad_norm": 0.22936795936443075,
+      "learning_rate": 2.0009006664437973e-06,
+      "loss": 1.1404,
+      "step": 3460
+    },
+    {
+      "epoch": 8.462102689486553,
+      "grad_norm": 0.29091826064494175,
+      "learning_rate": 1.999646310418092e-06,
+      "loss": 0.8825,
+      "step": 3461
+    },
+    {
+      "epoch": 8.464547677261614,
+      "grad_norm": 0.1876587646209225,
+      "learning_rate": 1.998392085596578e-06,
+      "loss": 0.628,
+      "step": 3462
+    },
+    {
+      "epoch": 8.466992665036674,
+      "grad_norm": 0.21211714009735663,
+      "learning_rate": 1.997137992308142e-06,
+      "loss": 0.5015,
+      "step": 3463
+    },
+    {
+      "epoch": 8.469437652811736,
+      "grad_norm": 0.2487273800376849,
+      "learning_rate": 1.9958840308816356e-06,
+      "loss": 0.378,
+      "step": 3464
+    },
+    {
+      "epoch": 8.471882640586797,
+      "grad_norm": 0.2617102295436207,
+      "learning_rate": 1.994630201645876e-06,
+      "loss": 0.396,
+      "step": 3465
+    },
+    {
+      "epoch": 8.474327628361857,
+      "grad_norm": 0.46322961722230716,
+      "learning_rate": 1.993376504929646e-06,
+      "loss": 0.7566,
+      "step": 3466
+    },
+    {
+      "epoch": 8.47677261613692,
+      "grad_norm": 0.3227873329415639,
+      "learning_rate": 1.992122941061694e-06,
+      "loss": 0.5154,
+      "step": 3467
+    },
+    {
+      "epoch": 8.47921760391198,
+      "grad_norm": 0.25595505532255103,
+      "learning_rate": 1.9908695103707328e-06,
+      "loss": 0.4571,
+      "step": 3468
+    },
+    {
+      "epoch": 8.48166259168704,
+      "grad_norm": 0.25688017162328164,
+      "learning_rate": 1.9896162131854403e-06,
+      "loss": 0.8192,
+      "step": 3469
+    },
+    {
+      "epoch": 8.484107579462103,
+      "grad_norm": 0.2534667793603329,
+      "learning_rate": 1.98836304983446e-06,
+      "loss": 0.4847,
+      "step": 3470
+    },
+    {
+      "epoch": 8.486552567237164,
+      "grad_norm": 0.19904472783850125,
+      "learning_rate": 1.9871100206464e-06,
+      "loss": 0.6501,
+      "step": 3471
+    },
+    {
+      "epoch": 8.488997555012224,
+      "grad_norm": 0.2426345197295938,
+      "learning_rate": 1.9858571259498337e-06,
+      "loss": 0.8233,
+      "step": 3472
+    },
+    {
+      "epoch": 8.491442542787286,
+      "grad_norm": 0.3098659857495727,
+      "learning_rate": 1.984604366073299e-06,
+      "loss": 0.5228,
+      "step": 3473
+    },
+    {
+      "epoch": 8.493887530562347,
+      "grad_norm": 0.23323184752158677,
+      "learning_rate": 1.9833517413452973e-06,
+      "loss": 1.165,
+      "step": 3474
+    },
+    {
+      "epoch": 8.496332518337407,
+      "grad_norm": 0.22369321632061312,
+      "learning_rate": 1.982099252094295e-06,
+      "loss": 0.3996,
+      "step": 3475
+    },
+    {
+      "epoch": 8.49877750611247,
+      "grad_norm": 0.24718967362547936,
+      "learning_rate": 1.980846898648725e-06,
+      "loss": 0.5318,
+      "step": 3476
+    },
+    {
+      "epoch": 8.50122249388753,
+      "grad_norm": 0.17581178230781888,
+      "learning_rate": 1.9795946813369825e-06,
+      "loss": 0.6048,
+      "step": 3477
+    },
+    {
+      "epoch": 8.503667481662593,
+      "grad_norm": 0.2326584189182196,
+      "learning_rate": 1.9783426004874274e-06,
+      "loss": 0.5056,
+      "step": 3478
+    },
+    {
+      "epoch": 8.506112469437653,
+      "grad_norm": 0.3021945024852293,
+      "learning_rate": 1.9770906564283828e-06,
+      "loss": 0.8403,
+      "step": 3479
+    },
+    {
+      "epoch": 8.508557457212714,
+      "grad_norm": 0.3463324406417476,
+      "learning_rate": 1.975838849488139e-06,
+      "loss": 0.4415,
+      "step": 3480
+    },
+    {
+      "epoch": 8.511002444987774,
+      "grad_norm": 0.3697625240035511,
+      "learning_rate": 1.974587179994947e-06,
+      "loss": 0.8968,
+      "step": 3481
+    },
+    {
+      "epoch": 8.513447432762836,
+      "grad_norm": 0.17100665827733202,
+      "learning_rate": 1.973335648277024e-06,
+      "loss": 0.6178,
+      "step": 3482
+    },
+    {
+      "epoch": 8.515892420537897,
+      "grad_norm": 0.31159033677965736,
+      "learning_rate": 1.972084254662551e-06,
+      "loss": 0.9117,
+      "step": 3483
+    },
+    {
+      "epoch": 8.51833740831296,
+      "grad_norm": 0.22136098447305175,
+      "learning_rate": 1.9708329994796707e-06,
+      "loss": 0.8505,
+      "step": 3484
+    },
+    {
+      "epoch": 8.52078239608802,
+      "grad_norm": 0.26238335861967294,
+      "learning_rate": 1.969581883056491e-06,
+      "loss": 0.4293,
+      "step": 3485
+    },
+    {
+      "epoch": 8.52322738386308,
+      "grad_norm": 0.20846764913793728,
+      "learning_rate": 1.968330905721084e-06,
+      "loss": 0.3413,
+      "step": 3486
+    },
+    {
+      "epoch": 8.525672371638143,
+      "grad_norm": 0.2462977978996641,
+      "learning_rate": 1.967080067801484e-06,
+      "loss": 0.7905,
+      "step": 3487
+    },
+    {
+      "epoch": 8.528117359413203,
+      "grad_norm": 0.24242753971382194,
+      "learning_rate": 1.9658293696256893e-06,
+      "loss": 0.4386,
+      "step": 3488
+    },
+    {
+      "epoch": 8.530562347188264,
+      "grad_norm": 0.2526537077565564,
+      "learning_rate": 1.9645788115216614e-06,
+      "loss": 0.3959,
+      "step": 3489
+    },
+    {
+      "epoch": 8.533007334963326,
+      "grad_norm": 0.2529371426102152,
+      "learning_rate": 1.9633283938173265e-06,
+      "loss": 0.4898,
+      "step": 3490
+    },
+    {
+      "epoch": 8.535452322738386,
+      "grad_norm": 0.18851739780634966,
+      "learning_rate": 1.962078116840572e-06,
+      "loss": 0.9808,
+      "step": 3491
+    },
+    {
+      "epoch": 8.537897310513447,
+      "grad_norm": 0.28748391227190234,
+      "learning_rate": 1.9608279809192474e-06,
+      "loss": 0.542,
+      "step": 3492
+    },
+    {
+      "epoch": 8.54034229828851,
+      "grad_norm": 0.1649721954538384,
+      "learning_rate": 1.9595779863811697e-06,
+      "loss": 0.4289,
+      "step": 3493
+    },
+    {
+      "epoch": 8.54278728606357,
+      "grad_norm": 0.2221071967383684,
+      "learning_rate": 1.9583281335541152e-06,
+      "loss": 0.645,
+      "step": 3494
+    },
+    {
+      "epoch": 8.54523227383863,
+      "grad_norm": 0.25201814625354385,
+      "learning_rate": 1.9570784227658234e-06,
+      "loss": 0.4984,
+      "step": 3495
+    },
+    {
+      "epoch": 8.547677261613693,
+      "grad_norm": 0.21172051968087555,
+      "learning_rate": 1.9558288543439973e-06,
+      "loss": 0.5099,
+      "step": 3496
+    },
+    {
+      "epoch": 8.550122249388753,
+      "grad_norm": 0.20991644584812197,
+      "learning_rate": 1.9545794286163027e-06,
+      "loss": 0.5876,
+      "step": 3497
+    },
+    {
+      "epoch": 8.552567237163814,
+      "grad_norm": 0.20930536048535836,
+      "learning_rate": 1.9533301459103677e-06,
+      "loss": 0.4725,
+      "step": 3498
+    },
+    {
+      "epoch": 8.555012224938876,
+      "grad_norm": 0.21017253622942617,
+      "learning_rate": 1.9520810065537815e-06,
+      "loss": 0.9064,
+      "step": 3499
+    },
+    {
+      "epoch": 8.557457212713937,
+      "grad_norm": 0.29022485731845893,
+      "learning_rate": 1.9508320108740985e-06,
+      "loss": 0.4033,
+      "step": 3500
+    },
+    {
+      "epoch": 8.559902200488997,
+      "grad_norm": 0.27130152721460377,
+      "learning_rate": 1.9495831591988333e-06,
+      "loss": 0.4446,
+      "step": 3501
+    },
+    {
+      "epoch": 8.56234718826406,
+      "grad_norm": 0.28189091625406437,
+      "learning_rate": 1.9483344518554632e-06,
+      "loss": 0.3453,
+      "step": 3502
+    },
+    {
+      "epoch": 8.56479217603912,
+      "grad_norm": 0.20880743274939603,
+      "learning_rate": 1.947085889171429e-06,
+      "loss": 0.6393,
+      "step": 3503
+    },
+    {
+      "epoch": 8.56723716381418,
+      "grad_norm": 0.2846776112193436,
+      "learning_rate": 1.945837471474131e-06,
+      "loss": 0.5495,
+      "step": 3504
+    },
+    {
+      "epoch": 8.569682151589243,
+      "grad_norm": 0.2429138249945023,
+      "learning_rate": 1.9445891990909335e-06,
+      "loss": 0.4722,
+      "step": 3505
+    },
+    {
+      "epoch": 8.572127139364303,
+      "grad_norm": 0.2495870228371255,
+      "learning_rate": 1.9433410723491627e-06,
+      "loss": 0.4256,
+      "step": 3506
+    },
+    {
+      "epoch": 8.574572127139364,
+      "grad_norm": 0.3697444792189972,
+      "learning_rate": 1.9420930915761053e-06,
+      "loss": 0.5657,
+      "step": 3507
+    },
+    {
+      "epoch": 8.577017114914426,
+      "grad_norm": 0.20177825430554475,
+      "learning_rate": 1.9408452570990104e-06,
+      "loss": 0.5508,
+      "step": 3508
+    },
+    {
+      "epoch": 8.579462102689487,
+      "grad_norm": 0.25046160816116525,
+      "learning_rate": 1.9395975692450886e-06,
+      "loss": 0.8596,
+      "step": 3509
+    },
+    {
+      "epoch": 8.581907090464547,
+      "grad_norm": 0.17699240289447116,
+      "learning_rate": 1.938350028341513e-06,
+      "loss": 0.4062,
+      "step": 3510
+    },
+    {
+      "epoch": 8.58435207823961,
+      "grad_norm": 0.18910200825032344,
+      "learning_rate": 1.937102634715417e-06,
+      "loss": 1.0002,
+      "step": 3511
+    },
+    {
+      "epoch": 8.58679706601467,
+      "grad_norm": 0.23239455434756917,
+      "learning_rate": 1.9358553886938952e-06,
+      "loss": 0.3862,
+      "step": 3512
+    },
+    {
+      "epoch": 8.58924205378973,
+      "grad_norm": 0.24163750549647728,
+      "learning_rate": 1.934608290604004e-06,
+      "loss": 0.4689,
+      "step": 3513
+    },
+    {
+      "epoch": 8.591687041564793,
+      "grad_norm": 0.16407933128204633,
+      "learning_rate": 1.9333613407727625e-06,
+      "loss": 0.6044,
+      "step": 3514
+    },
+    {
+      "epoch": 8.594132029339853,
+      "grad_norm": 0.20372024372004213,
+      "learning_rate": 1.932114539527148e-06,
+      "loss": 0.3832,
+      "step": 3515
+    },
+    {
+      "epoch": 8.596577017114914,
+      "grad_norm": 0.404097054579615,
+      "learning_rate": 1.9308678871941015e-06,
+      "loss": 0.4433,
+      "step": 3516
+    },
+    {
+      "epoch": 8.599022004889976,
+      "grad_norm": 0.2042974117653784,
+      "learning_rate": 1.929621384100523e-06,
+      "loss": 0.5233,
+      "step": 3517
+    },
+    {
+      "epoch": 8.601466992665037,
+      "grad_norm": 0.26809622469822847,
+      "learning_rate": 1.9283750305732747e-06,
+      "loss": 0.3648,
+      "step": 3518
+    },
+    {
+      "epoch": 8.603911980440097,
+      "grad_norm": 0.35456141185891565,
+      "learning_rate": 1.9271288269391773e-06,
+      "loss": 0.4103,
+      "step": 3519
+    },
+    {
+      "epoch": 8.60635696821516,
+      "grad_norm": 0.27894431595251695,
+      "learning_rate": 1.925882773525017e-06,
+      "loss": 0.3682,
+      "step": 3520
+    },
+    {
+      "epoch": 8.60880195599022,
+      "grad_norm": 0.3000402634321122,
+      "learning_rate": 1.924636870657535e-06,
+      "loss": 0.4124,
+      "step": 3521
+    },
+    {
+      "epoch": 8.61124694376528,
+      "grad_norm": 0.28394657959099134,
+      "learning_rate": 1.9233911186634367e-06,
+      "loss": 0.7768,
+      "step": 3522
+    },
+    {
+      "epoch": 8.613691931540343,
+      "grad_norm": 0.28027335269104686,
+      "learning_rate": 1.9221455178693858e-06,
+      "loss": 0.8926,
+      "step": 3523
+    },
+    {
+      "epoch": 8.616136919315403,
+      "grad_norm": 0.29206777414134233,
+      "learning_rate": 1.920900068602008e-06,
+      "loss": 0.5853,
+      "step": 3524
+    },
+    {
+      "epoch": 8.618581907090464,
+      "grad_norm": 0.2876923464255937,
+      "learning_rate": 1.9196547711878884e-06,
+      "loss": 0.6171,
+      "step": 3525
+    },
+    {
+      "epoch": 8.621026894865526,
+      "grad_norm": 0.3861950516456796,
+      "learning_rate": 1.9184096259535734e-06,
+      "loss": 0.7075,
+      "step": 3526
+    },
+    {
+      "epoch": 8.623471882640587,
+      "grad_norm": 0.26286690408869506,
+      "learning_rate": 1.9171646332255682e-06,
+      "loss": 0.8028,
+      "step": 3527
+    },
+    {
+      "epoch": 8.625916870415647,
+      "grad_norm": 0.27436038805453095,
+      "learning_rate": 1.9159197933303376e-06,
+      "loss": 0.7116,
+      "step": 3528
+    },
+    {
+      "epoch": 8.62836185819071,
+      "grad_norm": 0.22404469343288536,
+      "learning_rate": 1.914675106594307e-06,
+      "loss": 0.9514,
+      "step": 3529
+    },
+    {
+      "epoch": 8.63080684596577,
+      "grad_norm": 0.25793644445227665,
+      "learning_rate": 1.913430573343863e-06,
+      "loss": 0.5078,
+      "step": 3530
+    },
+    {
+      "epoch": 8.63325183374083,
+      "grad_norm": 0.27709460765148414,
+      "learning_rate": 1.9121861939053506e-06,
+      "loss": 0.4828,
+      "step": 3531
+    },
+    {
+      "epoch": 8.635696821515893,
+      "grad_norm": 0.18781442664342946,
+      "learning_rate": 1.9109419686050733e-06,
+      "loss": 0.6476,
+      "step": 3532
+    },
+    {
+      "epoch": 8.638141809290953,
+      "grad_norm": 0.17832987303233894,
+      "learning_rate": 1.909697897769296e-06,
+      "loss": 0.6052,
+      "step": 3533
+    },
+    {
+      "epoch": 8.640586797066014,
+      "grad_norm": 0.34380449294481163,
+      "learning_rate": 1.908453981724243e-06,
+      "loss": 0.5571,
+      "step": 3534
+    },
+    {
+      "epoch": 8.643031784841076,
+      "grad_norm": 0.3291903673278451,
+      "learning_rate": 1.9072102207960964e-06,
+      "loss": 0.4599,
+      "step": 3535
+    },
+    {
+      "epoch": 8.645476772616137,
+      "grad_norm": 0.2474979389749802,
+      "learning_rate": 1.9059666153110006e-06,
+      "loss": 0.6129,
+      "step": 3536
+    },
+    {
+      "epoch": 8.647921760391197,
+      "grad_norm": 0.2733955512016475,
+      "learning_rate": 1.9047231655950568e-06,
+      "loss": 0.6318,
+      "step": 3537
+    },
+    {
+      "epoch": 8.65036674816626,
+      "grad_norm": 0.2476370397992299,
+      "learning_rate": 1.9034798719743253e-06,
+      "loss": 0.413,
+      "step": 3538
+    },
+    {
+      "epoch": 8.65281173594132,
+      "grad_norm": 0.21549234716228272,
+      "learning_rate": 1.9022367347748264e-06,
+      "loss": 0.6836,
+      "step": 3539
+    },
+    {
+      "epoch": 8.65525672371638,
+      "grad_norm": 0.23096758833599285,
+      "learning_rate": 1.9009937543225397e-06,
+      "loss": 0.6656,
+      "step": 3540
+    },
+    {
+      "epoch": 8.657701711491443,
+      "grad_norm": 0.20289537968424193,
+      "learning_rate": 1.8997509309434023e-06,
+      "loss": 0.4325,
+      "step": 3541
+    },
+    {
+      "epoch": 8.660146699266503,
+      "grad_norm": 0.16243383038822015,
+      "learning_rate": 1.8985082649633115e-06,
+      "loss": 0.5114,
+      "step": 3542
+    },
+    {
+      "epoch": 8.662591687041564,
+      "grad_norm": 0.3310952435561895,
+      "learning_rate": 1.8972657567081224e-06,
+      "loss": 0.5086,
+      "step": 3543
+    },
+    {
+      "epoch": 8.665036674816626,
+      "grad_norm": 0.25989128465423195,
+      "learning_rate": 1.8960234065036495e-06,
+      "loss": 0.7747,
+      "step": 3544
+    },
+    {
+      "epoch": 8.667481662591687,
+      "grad_norm": 0.3117353261024849,
+      "learning_rate": 1.8947812146756645e-06,
+      "loss": 0.4592,
+      "step": 3545
+    },
+    {
+      "epoch": 8.669926650366747,
+      "grad_norm": 0.291014918859351,
+      "learning_rate": 1.8935391815498998e-06,
+      "loss": 0.7344,
+      "step": 3546
+    },
+    {
+      "epoch": 8.67237163814181,
+      "grad_norm": 0.29487924447870956,
+      "learning_rate": 1.8922973074520451e-06,
+      "loss": 0.4169,
+      "step": 3547
+    },
+    {
+      "epoch": 8.67481662591687,
+      "grad_norm": 0.2106458073265303,
+      "learning_rate": 1.8910555927077471e-06,
+      "loss": 0.4149,
+      "step": 3548
+    },
+    {
+      "epoch": 8.67726161369193,
+      "grad_norm": 0.23161344642698872,
+      "learning_rate": 1.8898140376426118e-06,
+      "loss": 0.4984,
+      "step": 3549
+    },
+    {
+      "epoch": 8.679706601466993,
+      "grad_norm": 0.2920424107188469,
+      "learning_rate": 1.8885726425822044e-06,
+      "loss": 0.5177,
+      "step": 3550
+    },
+    {
+      "epoch": 8.682151589242054,
+      "grad_norm": 0.20094702529289873,
+      "learning_rate": 1.8873314078520466e-06,
+      "loss": 0.4431,
+      "step": 3551
+    },
+    {
+      "epoch": 8.684596577017114,
+      "grad_norm": 0.26483461221684934,
+      "learning_rate": 1.8860903337776182e-06,
+      "loss": 0.831,
+      "step": 3552
+    },
+    {
+      "epoch": 8.687041564792176,
+      "grad_norm": 0.30028524212987706,
+      "learning_rate": 1.8848494206843574e-06,
+      "loss": 0.5183,
+      "step": 3553
+    },
+    {
+      "epoch": 8.689486552567237,
+      "grad_norm": 0.3340435237146265,
+      "learning_rate": 1.8836086688976605e-06,
+      "loss": 0.4664,
+      "step": 3554
+    },
+    {
+      "epoch": 8.6919315403423,
+      "grad_norm": 0.24168574000166768,
+      "learning_rate": 1.8823680787428805e-06,
+      "loss": 0.6054,
+      "step": 3555
+    },
+    {
+      "epoch": 8.69437652811736,
+      "grad_norm": 0.18809454416548121,
+      "learning_rate": 1.8811276505453277e-06,
+      "loss": 0.377,
+      "step": 3556
+    },
+    {
+      "epoch": 8.69682151589242,
+      "grad_norm": 0.2791547233976903,
+      "learning_rate": 1.879887384630273e-06,
+      "loss": 0.5847,
+      "step": 3557
+    },
+    {
+      "epoch": 8.69926650366748,
+      "grad_norm": 0.15870852643894778,
+      "learning_rate": 1.8786472813229408e-06,
+      "loss": 0.3086,
+      "step": 3558
+    },
+    {
+      "epoch": 8.701711491442543,
+      "grad_norm": 0.2661033255739425,
+      "learning_rate": 1.8774073409485149e-06,
+      "loss": 0.461,
+      "step": 3559
+    },
+    {
+      "epoch": 8.704156479217604,
+      "grad_norm": 0.1954390829026525,
+      "learning_rate": 1.876167563832137e-06,
+      "loss": 0.5994,
+      "step": 3560
+    },
+    {
+      "epoch": 8.706601466992666,
+      "grad_norm": 0.2209018184894722,
+      "learning_rate": 1.8749279502989037e-06,
+      "loss": 0.4499,
+      "step": 3561
+    },
+    {
+      "epoch": 8.709046454767726,
+      "grad_norm": 0.2579623621111341,
+      "learning_rate": 1.8736885006738708e-06,
+      "loss": 0.7331,
+      "step": 3562
+    },
+    {
+      "epoch": 8.711491442542787,
+      "grad_norm": 0.19758607921099702,
+      "learning_rate": 1.8724492152820495e-06,
+      "loss": 0.4074,
+      "step": 3563
+    },
+    {
+      "epoch": 8.713936430317847,
+      "grad_norm": 0.20391215032763163,
+      "learning_rate": 1.87121009444841e-06,
+      "loss": 1.1082,
+      "step": 3564
+    },
+    {
+      "epoch": 8.71638141809291,
+      "grad_norm": 0.1999394149660434,
+      "learning_rate": 1.8699711384978781e-06,
+      "loss": 0.8326,
+      "step": 3565
+    },
+    {
+      "epoch": 8.71882640586797,
+      "grad_norm": 0.32513091702383184,
+      "learning_rate": 1.8687323477553354e-06,
+      "loss": 0.6517,
+      "step": 3566
+    },
+    {
+      "epoch": 8.721271393643033,
+      "grad_norm": 0.1843675802642467,
+      "learning_rate": 1.8674937225456207e-06,
+      "loss": 0.3082,
+      "step": 3567
+    },
+    {
+      "epoch": 8.723716381418093,
+      "grad_norm": 0.19508695125818107,
+      "learning_rate": 1.866255263193532e-06,
+      "loss": 0.3884,
+      "step": 3568
+    },
+    {
+      "epoch": 8.726161369193154,
+      "grad_norm": 0.2795456489020726,
+      "learning_rate": 1.86501697002382e-06,
+      "loss": 0.7367,
+      "step": 3569
+    },
+    {
+      "epoch": 8.728606356968216,
+      "grad_norm": 0.2638947225523053,
+      "learning_rate": 1.8637788433611948e-06,
+      "loss": 0.5503,
+      "step": 3570
+    },
+    {
+      "epoch": 8.731051344743276,
+      "grad_norm": 0.2984692933370658,
+      "learning_rate": 1.8625408835303205e-06,
+      "loss": 0.4435,
+      "step": 3571
+    },
+    {
+      "epoch": 8.733496332518337,
+      "grad_norm": 0.1680395301381349,
+      "learning_rate": 1.861303090855819e-06,
+      "loss": 0.5352,
+      "step": 3572
+    },
+    {
+      "epoch": 8.7359413202934,
+      "grad_norm": 0.18032241384480857,
+      "learning_rate": 1.8600654656622675e-06,
+      "loss": 0.4048,
+      "step": 3573
+    },
+    {
+      "epoch": 8.73838630806846,
+      "grad_norm": 0.18451461045994727,
+      "learning_rate": 1.8588280082741999e-06,
+      "loss": 0.3899,
+      "step": 3574
+    },
+    {
+      "epoch": 8.74083129584352,
+      "grad_norm": 0.19278557122166826,
+      "learning_rate": 1.857590719016106e-06,
+      "loss": 0.4494,
+      "step": 3575
+    },
+    {
+      "epoch": 8.743276283618583,
+      "grad_norm": 0.2670709610212619,
+      "learning_rate": 1.8563535982124314e-06,
+      "loss": 0.8321,
+      "step": 3576
+    },
+    {
+      "epoch": 8.745721271393643,
+      "grad_norm": 0.2369556950403544,
+      "learning_rate": 1.8551166461875761e-06,
+      "loss": 0.7048,
+      "step": 3577
+    },
+    {
+      "epoch": 8.748166259168704,
+      "grad_norm": 0.19792029199228522,
+      "learning_rate": 1.8538798632658994e-06,
+      "loss": 0.3984,
+      "step": 3578
+    },
+    {
+      "epoch": 8.750611246943766,
+      "grad_norm": 0.20337492509153593,
+      "learning_rate": 1.8526432497717128e-06,
+      "loss": 0.5945,
+      "step": 3579
+    },
+    {
+      "epoch": 8.753056234718827,
+      "grad_norm": 0.27425567054014555,
+      "learning_rate": 1.8514068060292856e-06,
+      "loss": 0.4128,
+      "step": 3580
+    },
+    {
+      "epoch": 8.755501222493887,
+      "grad_norm": 0.2374559670832558,
+      "learning_rate": 1.8501705323628405e-06,
+      "loss": 0.6566,
+      "step": 3581
+    },
+    {
+      "epoch": 8.75794621026895,
+      "grad_norm": 0.2271241467376953,
+      "learning_rate": 1.8489344290965578e-06,
+      "loss": 0.7612,
+      "step": 3582
+    },
+    {
+      "epoch": 8.76039119804401,
+      "grad_norm": 0.17744630326547883,
+      "learning_rate": 1.8476984965545708e-06,
+      "loss": 0.4367,
+      "step": 3583
+    },
+    {
+      "epoch": 8.76283618581907,
+      "grad_norm": 0.29642515539022385,
+      "learning_rate": 1.846462735060971e-06,
+      "loss": 0.8888,
+      "step": 3584
+    },
+    {
+      "epoch": 8.765281173594133,
+      "grad_norm": 0.2956561935943152,
+      "learning_rate": 1.8452271449398018e-06,
+      "loss": 0.6676,
+      "step": 3585
+    },
+    {
+      "epoch": 8.767726161369193,
+      "grad_norm": 0.3481583250220868,
+      "learning_rate": 1.8439917265150641e-06,
+      "loss": 0.4657,
+      "step": 3586
+    },
+    {
+      "epoch": 8.770171149144254,
+      "grad_norm": 0.17487483307159501,
+      "learning_rate": 1.8427564801107123e-06,
+      "loss": 0.3737,
+      "step": 3587
+    },
+    {
+      "epoch": 8.772616136919316,
+      "grad_norm": 0.3204048859324367,
+      "learning_rate": 1.8415214060506568e-06,
+      "loss": 0.3674,
+      "step": 3588
+    },
+    {
+      "epoch": 8.775061124694377,
+      "grad_norm": 0.2737594055342559,
+      "learning_rate": 1.8402865046587618e-06,
+      "loss": 0.7675,
+      "step": 3589
+    },
+    {
+      "epoch": 8.777506112469437,
+      "grad_norm": 0.19345359628899317,
+      "learning_rate": 1.8390517762588472e-06,
+      "loss": 0.514,
+      "step": 3590
+    },
+    {
+      "epoch": 8.7799511002445,
+      "grad_norm": 0.21344587476710078,
+      "learning_rate": 1.8378172211746875e-06,
+      "loss": 0.7672,
+      "step": 3591
+    },
+    {
+      "epoch": 8.78239608801956,
+      "grad_norm": 0.25356720115852227,
+      "learning_rate": 1.8365828397300103e-06,
+      "loss": 0.6055,
+      "step": 3592
+    },
+    {
+      "epoch": 8.78484107579462,
+      "grad_norm": 0.25905199134501394,
+      "learning_rate": 1.8353486322484988e-06,
+      "loss": 0.4837,
+      "step": 3593
+    },
+    {
+      "epoch": 8.787286063569683,
+      "grad_norm": 0.17390934203312977,
+      "learning_rate": 1.834114599053791e-06,
+      "loss": 0.7742,
+      "step": 3594
+    },
+    {
+      "epoch": 8.789731051344743,
+      "grad_norm": 0.24831720076112573,
+      "learning_rate": 1.8328807404694785e-06,
+      "loss": 0.4213,
+      "step": 3595
+    },
+    {
+      "epoch": 8.792176039119804,
+      "grad_norm": 0.27056706558634086,
+      "learning_rate": 1.8316470568191069e-06,
+      "loss": 0.5978,
+      "step": 3596
+    },
+    {
+      "epoch": 8.794621026894866,
+      "grad_norm": 0.23147927279437583,
+      "learning_rate": 1.8304135484261763e-06,
+      "loss": 0.6103,
+      "step": 3597
+    },
+    {
+      "epoch": 8.797066014669927,
+      "grad_norm": 0.33076328852681885,
+      "learning_rate": 1.8291802156141414e-06,
+      "loss": 0.7115,
+      "step": 3598
+    },
+    {
+      "epoch": 8.799511002444987,
+      "grad_norm": 0.3902915468455055,
+      "learning_rate": 1.8279470587064089e-06,
+      "loss": 0.7379,
+      "step": 3599
+    },
+    {
+      "epoch": 8.80195599022005,
+      "grad_norm": 0.19933831404052776,
+      "learning_rate": 1.8267140780263426e-06,
+      "loss": 0.5269,
+      "step": 3600
+    },
+    {
+      "epoch": 8.80440097799511,
+      "grad_norm": 0.25602405112150456,
+      "learning_rate": 1.8254812738972576e-06,
+      "loss": 0.5156,
+      "step": 3601
+    },
+    {
+      "epoch": 8.80684596577017,
+      "grad_norm": 0.3238957148239051,
+      "learning_rate": 1.8242486466424236e-06,
+      "loss": 0.5862,
+      "step": 3602
+    },
+    {
+      "epoch": 8.809290953545233,
+      "grad_norm": 0.2650617154850533,
+      "learning_rate": 1.8230161965850626e-06,
+      "loss": 1.0186,
+      "step": 3603
+    },
+    {
+      "epoch": 8.811735941320293,
+      "grad_norm": 0.2126378721161555,
+      "learning_rate": 1.8217839240483527e-06,
+      "loss": 0.4014,
+      "step": 3604
+    },
+    {
+      "epoch": 8.814180929095354,
+      "grad_norm": 0.32915470229914523,
+      "learning_rate": 1.8205518293554227e-06,
+      "loss": 0.7862,
+      "step": 3605
+    },
+    {
+      "epoch": 8.816625916870416,
+      "grad_norm": 0.3288675811089684,
+      "learning_rate": 1.819319912829357e-06,
+      "loss": 0.4115,
+      "step": 3606
+    },
+    {
+      "epoch": 8.819070904645477,
+      "grad_norm": 0.27349779861110085,
+      "learning_rate": 1.8180881747931912e-06,
+      "loss": 0.6449,
+      "step": 3607
+    },
+    {
+      "epoch": 8.821515892420537,
+      "grad_norm": 0.3366097350630797,
+      "learning_rate": 1.8168566155699161e-06,
+      "loss": 0.4608,
+      "step": 3608
+    },
+    {
+      "epoch": 8.8239608801956,
+      "grad_norm": 0.2398592912437631,
+      "learning_rate": 1.8156252354824746e-06,
+      "loss": 0.5226,
+      "step": 3609
+    },
+    {
+      "epoch": 8.82640586797066,
+      "grad_norm": 0.25126285784834845,
+      "learning_rate": 1.8143940348537619e-06,
+      "loss": 1.1248,
+      "step": 3610
+    },
+    {
+      "epoch": 8.82885085574572,
+      "grad_norm": 0.20497404551365436,
+      "learning_rate": 1.8131630140066284e-06,
+      "loss": 0.4013,
+      "step": 3611
+    },
+    {
+      "epoch": 8.831295843520783,
+      "grad_norm": 0.17657946261528143,
+      "learning_rate": 1.8119321732638756e-06,
+      "loss": 0.3747,
+      "step": 3612
+    },
+    {
+      "epoch": 8.833740831295843,
+      "grad_norm": 0.2112306497322087,
+      "learning_rate": 1.8107015129482568e-06,
+      "loss": 0.4105,
+      "step": 3613
+    },
+    {
+      "epoch": 8.836185819070904,
+      "grad_norm": 0.22649199266512973,
+      "learning_rate": 1.8094710333824812e-06,
+      "loss": 0.8881,
+      "step": 3614
+    },
+    {
+      "epoch": 8.838630806845966,
+      "grad_norm": 0.17546627290486372,
+      "learning_rate": 1.8082407348892078e-06,
+      "loss": 0.3832,
+      "step": 3615
+    },
+    {
+      "epoch": 8.841075794621027,
+      "grad_norm": 0.2826507214981961,
+      "learning_rate": 1.8070106177910488e-06,
+      "loss": 1.0206,
+      "step": 3616
+    },
+    {
+      "epoch": 8.843520782396087,
+      "grad_norm": 0.30125207639250673,
+      "learning_rate": 1.8057806824105689e-06,
+      "loss": 0.5728,
+      "step": 3617
+    },
+    {
+      "epoch": 8.84596577017115,
+      "grad_norm": 0.22873835128877792,
+      "learning_rate": 1.8045509290702862e-06,
+      "loss": 0.34,
+      "step": 3618
+    },
+    {
+      "epoch": 8.84841075794621,
+      "grad_norm": 0.4193489042586497,
+      "learning_rate": 1.8033213580926696e-06,
+      "loss": 0.68,
+      "step": 3619
+    },
+    {
+      "epoch": 8.85085574572127,
+      "grad_norm": 0.2858781203351356,
+      "learning_rate": 1.80209196980014e-06,
+      "loss": 0.4504,
+      "step": 3620
+    },
+    {
+      "epoch": 8.853300733496333,
+      "grad_norm": 0.2958818440981415,
+      "learning_rate": 1.8008627645150727e-06,
+      "loss": 0.6342,
+      "step": 3621
+    },
+    {
+      "epoch": 8.855745721271393,
+      "grad_norm": 0.28012187701396,
+      "learning_rate": 1.7996337425597932e-06,
+      "loss": 0.8687,
+      "step": 3622
+    },
+    {
+      "epoch": 8.858190709046454,
+      "grad_norm": 0.3814683350693685,
+      "learning_rate": 1.7984049042565777e-06,
+      "loss": 0.4411,
+      "step": 3623
+    },
+    {
+      "epoch": 8.860635696821516,
+      "grad_norm": 0.2640221794278391,
+      "learning_rate": 1.7971762499276576e-06,
+      "loss": 0.5539,
+      "step": 3624
+    },
+    {
+      "epoch": 8.863080684596577,
+      "grad_norm": 0.20738841897062282,
+      "learning_rate": 1.7959477798952135e-06,
+      "loss": 0.6179,
+      "step": 3625
+    },
+    {
+      "epoch": 8.865525672371637,
+      "grad_norm": 0.3262876550672289,
+      "learning_rate": 1.7947194944813779e-06,
+      "loss": 0.4065,
+      "step": 3626
+    },
+    {
+      "epoch": 8.8679706601467,
+      "grad_norm": 0.4221722697792616,
+      "learning_rate": 1.7934913940082355e-06,
+      "loss": 0.4429,
+      "step": 3627
+    },
+    {
+      "epoch": 8.87041564792176,
+      "grad_norm": 0.2229330865453247,
+      "learning_rate": 1.7922634787978227e-06,
+      "loss": 0.8856,
+      "step": 3628
+    },
+    {
+      "epoch": 8.87286063569682,
+      "grad_norm": 0.25642928739595605,
+      "learning_rate": 1.791035749172127e-06,
+      "loss": 0.8247,
+      "step": 3629
+    },
+    {
+      "epoch": 8.875305623471883,
+      "grad_norm": 0.26881034316757757,
+      "learning_rate": 1.789808205453087e-06,
+      "loss": 0.8383,
+      "step": 3630
+    },
+    {
+      "epoch": 8.877750611246944,
+      "grad_norm": 0.23335074805961314,
+      "learning_rate": 1.788580847962592e-06,
+      "loss": 0.5438,
+      "step": 3631
+    },
+    {
+      "epoch": 8.880195599022004,
+      "grad_norm": 0.20788433365989567,
+      "learning_rate": 1.7873536770224847e-06,
+      "loss": 0.4696,
+      "step": 3632
+    },
+    {
+      "epoch": 8.882640586797066,
+      "grad_norm": 0.29934702043428696,
+      "learning_rate": 1.7861266929545561e-06,
+      "loss": 0.4366,
+      "step": 3633
+    },
+    {
+      "epoch": 8.885085574572127,
+      "grad_norm": 0.22062837977260127,
+      "learning_rate": 1.7848998960805515e-06,
+      "loss": 0.7434,
+      "step": 3634
+    },
+    {
+      "epoch": 8.887530562347187,
+      "grad_norm": 0.13359902350951794,
+      "learning_rate": 1.783673286722163e-06,
+      "loss": 0.4307,
+      "step": 3635
+    },
+    {
+      "epoch": 8.88997555012225,
+      "grad_norm": 0.35273608991692523,
+      "learning_rate": 1.7824468652010366e-06,
+      "loss": 0.6067,
+      "step": 3636
+    },
+    {
+      "epoch": 8.89242053789731,
+      "grad_norm": 0.20696255054202498,
+      "learning_rate": 1.7812206318387674e-06,
+      "loss": 0.87,
+      "step": 3637
+    },
+    {
+      "epoch": 8.894865525672373,
+      "grad_norm": 0.2399146374565134,
+      "learning_rate": 1.779994586956903e-06,
+      "loss": 0.4494,
+      "step": 3638
+    },
+    {
+      "epoch": 8.897310513447433,
+      "grad_norm": 0.2326272397900849,
+      "learning_rate": 1.7787687308769394e-06,
+      "loss": 0.4104,
+      "step": 3639
+    },
+    {
+      "epoch": 8.899755501222494,
+      "grad_norm": 0.2801005163790302,
+      "learning_rate": 1.777543063920325e-06,
+      "loss": 0.5831,
+      "step": 3640
+    },
+    {
+      "epoch": 8.902200488997554,
+      "grad_norm": 0.25036850775354236,
+      "learning_rate": 1.7763175864084563e-06,
+      "loss": 0.8582,
+      "step": 3641
+    },
+    {
+      "epoch": 8.904645476772616,
+      "grad_norm": 0.2770522128615707,
+      "learning_rate": 1.7750922986626828e-06,
+      "loss": 0.4781,
+      "step": 3642
+    },
+    {
+      "epoch": 8.907090464547677,
+      "grad_norm": 0.35283891841933107,
+      "learning_rate": 1.773867201004303e-06,
+      "loss": 0.3935,
+      "step": 3643
+    },
+    {
+      "epoch": 8.90953545232274,
+      "grad_norm": 0.2634912974010283,
+      "learning_rate": 1.7726422937545653e-06,
+      "loss": 0.8075,
+      "step": 3644
+    },
+    {
+      "epoch": 8.9119804400978,
+      "grad_norm": 0.20034590320795434,
+      "learning_rate": 1.7714175772346686e-06,
+      "loss": 0.4773,
+      "step": 3645
+    },
+    {
+      "epoch": 8.91442542787286,
+      "grad_norm": 0.2097835495187987,
+      "learning_rate": 1.7701930517657615e-06,
+      "loss": 0.3724,
+      "step": 3646
+    },
+    {
+      "epoch": 8.91687041564792,
+      "grad_norm": 0.44833926481000913,
+      "learning_rate": 1.7689687176689418e-06,
+      "loss": 0.5903,
+      "step": 3647
+    },
+    {
+      "epoch": 8.919315403422983,
+      "grad_norm": 0.21785872488112026,
+      "learning_rate": 1.7677445752652594e-06,
+      "loss": 0.3931,
+      "step": 3648
+    },
+    {
+      "epoch": 8.921760391198044,
+      "grad_norm": 0.4219317473076237,
+      "learning_rate": 1.7665206248757116e-06,
+      "loss": 0.4517,
+      "step": 3649
+    },
+    {
+      "epoch": 8.924205378973106,
+      "grad_norm": 0.23288670727080563,
+      "learning_rate": 1.7652968668212462e-06,
+      "loss": 1.0049,
+      "step": 3650
+    },
+    {
+      "epoch": 8.926650366748166,
+      "grad_norm": 0.3105504328060746,
+      "learning_rate": 1.76407330142276e-06,
+      "loss": 0.6928,
+      "step": 3651
+    },
+    {
+      "epoch": 8.929095354523227,
+      "grad_norm": 0.2448374618102944,
+      "learning_rate": 1.7628499290011014e-06,
+      "loss": 0.8116,
+      "step": 3652
+    },
+    {
+      "epoch": 8.93154034229829,
+      "grad_norm": 0.3930291356974308,
+      "learning_rate": 1.7616267498770644e-06,
+      "loss": 0.9907,
+      "step": 3653
+    },
+    {
+      "epoch": 8.93398533007335,
+      "grad_norm": 0.20867402496773718,
+      "learning_rate": 1.7604037643713973e-06,
+      "loss": 0.4163,
+      "step": 3654
+    },
+    {
+      "epoch": 8.93643031784841,
+      "grad_norm": 0.3972959669402503,
+      "learning_rate": 1.7591809728047934e-06,
+      "loss": 0.7532,
+      "step": 3655
+    },
+    {
+      "epoch": 8.938875305623473,
+      "grad_norm": 0.39822776357470846,
+      "learning_rate": 1.7579583754978968e-06,
+      "loss": 0.3948,
+      "step": 3656
+    },
+    {
+      "epoch": 8.941320293398533,
+      "grad_norm": 0.37464443040936035,
+      "learning_rate": 1.7567359727713e-06,
+      "loss": 0.3813,
+      "step": 3657
+    },
+    {
+      "epoch": 8.943765281173594,
+      "grad_norm": 0.2734178913534183,
+      "learning_rate": 1.7555137649455462e-06,
+      "loss": 0.6545,
+      "step": 3658
+    },
+    {
+      "epoch": 8.946210268948656,
+      "grad_norm": 0.23281213624014122,
+      "learning_rate": 1.7542917523411252e-06,
+      "loss": 0.6916,
+      "step": 3659
+    },
+    {
+      "epoch": 8.948655256723717,
+      "grad_norm": 0.27526596644154083,
+      "learning_rate": 1.7530699352784772e-06,
+      "loss": 0.7506,
+      "step": 3660
+    },
+    {
+      "epoch": 8.951100244498777,
+      "grad_norm": 0.2202602641955853,
+      "learning_rate": 1.7518483140779895e-06,
+      "loss": 0.5424,
+      "step": 3661
+    },
+    {
+      "epoch": 8.95354523227384,
+      "grad_norm": 0.3514005572126245,
+      "learning_rate": 1.750626889060001e-06,
+      "loss": 0.4009,
+      "step": 3662
+    },
+    {
+      "epoch": 8.9559902200489,
+      "grad_norm": 0.18485340344583961,
+      "learning_rate": 1.7494056605447956e-06,
+      "loss": 0.4246,
+      "step": 3663
+    },
+    {
+      "epoch": 8.95843520782396,
+      "grad_norm": 0.19706697309874063,
+      "learning_rate": 1.748184628852609e-06,
+      "loss": 0.5644,
+      "step": 3664
+    },
+    {
+      "epoch": 8.960880195599023,
+      "grad_norm": 0.23839806335486152,
+      "learning_rate": 1.7469637943036223e-06,
+      "loss": 0.5784,
+      "step": 3665
+    },
+    {
+      "epoch": 8.963325183374083,
+      "grad_norm": 0.32638494809175844,
+      "learning_rate": 1.7457431572179672e-06,
+      "loss": 0.6155,
+      "step": 3666
+    },
+    {
+      "epoch": 8.965770171149144,
+      "grad_norm": 0.21498940232261818,
+      "learning_rate": 1.7445227179157215e-06,
+      "loss": 0.4677,
+      "step": 3667
+    },
+    {
+      "epoch": 8.968215158924206,
+      "grad_norm": 0.3010381909948852,
+      "learning_rate": 1.7433024767169138e-06,
+      "loss": 0.2288,
+      "step": 3668
+    },
+    {
+      "epoch": 8.970660146699267,
+      "grad_norm": 0.25056184123472935,
+      "learning_rate": 1.7420824339415181e-06,
+      "loss": 0.6996,
+      "step": 3669
+    },
+    {
+      "epoch": 8.973105134474327,
+      "grad_norm": 0.4725189358441569,
+      "learning_rate": 1.7408625899094587e-06,
+      "loss": 0.447,
+      "step": 3670
+    },
+    {
+      "epoch": 8.97555012224939,
+      "grad_norm": 0.26131544802175893,
+      "learning_rate": 1.7396429449406049e-06,
+      "loss": 0.8393,
+      "step": 3671
+    },
+    {
+      "epoch": 8.97799511002445,
+      "grad_norm": 0.17502893508039313,
+      "learning_rate": 1.7384234993547772e-06,
+      "loss": 0.6801,
+      "step": 3672
+    },
+    {
+      "epoch": 8.98044009779951,
+      "grad_norm": 0.31462350844714715,
+      "learning_rate": 1.7372042534717414e-06,
+      "loss": 0.4786,
+      "step": 3673
+    },
+    {
+      "epoch": 8.982885085574573,
+      "grad_norm": 0.2007139309560344,
+      "learning_rate": 1.735985207611211e-06,
+      "loss": 0.5823,
+      "step": 3674
+    },
+    {
+      "epoch": 8.985330073349633,
+      "grad_norm": 0.2681694303288693,
+      "learning_rate": 1.7347663620928495e-06,
+      "loss": 0.5757,
+      "step": 3675
+    },
+    {
+      "epoch": 8.987775061124694,
+      "grad_norm": 0.2402669852138904,
+      "learning_rate": 1.7335477172362648e-06,
+      "loss": 0.3457,
+      "step": 3676
+    },
+    {
+      "epoch": 8.990220048899756,
+      "grad_norm": 0.3139336397866662,
+      "learning_rate": 1.7323292733610136e-06,
+      "loss": 0.4574,
+      "step": 3677
+    },
+    {
+      "epoch": 8.992665036674817,
+      "grad_norm": 0.31739826620724554,
+      "learning_rate": 1.7311110307866003e-06,
+      "loss": 0.4143,
+      "step": 3678
+    },
+    {
+      "epoch": 8.995110024449877,
+      "grad_norm": 0.1857274865172723,
+      "learning_rate": 1.7298929898324757e-06,
+      "loss": 0.5478,
+      "step": 3679
+    },
+    {
+      "epoch": 8.99755501222494,
+      "grad_norm": 0.2418060200439178,
+      "learning_rate": 1.7286751508180377e-06,
+      "loss": 1.0249,
+      "step": 3680
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.2897884308653675,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.3965,
+      "step": 3681
+    },
+    {
+      "epoch": 9.00244498777506,
+      "grad_norm": 0.3228548685613579,
+      "learning_rate": 1.7262400798855506e-06,
+      "loss": 0.4992,
+      "step": 3682
+    },
+    {
+      "epoch": 9.004889975550123,
+      "grad_norm": 0.2840729176965229,
+      "learning_rate": 1.7250228486060327e-06,
+      "loss": 0.3513,
+      "step": 3683
+    },
+    {
+      "epoch": 9.007334963325183,
+      "grad_norm": 0.2740017510081472,
+      "learning_rate": 1.7238058205432645e-06,
+      "loss": 1.3055,
+      "step": 3684
+    },
+    {
+      "epoch": 9.009779951100244,
+      "grad_norm": 0.32951515299543316,
+      "learning_rate": 1.7225889960163769e-06,
+      "loss": 0.496,
+      "step": 3685
+    },
+    {
+      "epoch": 9.012224938875306,
+      "grad_norm": 0.34827803623910913,
+      "learning_rate": 1.7213723753444517e-06,
+      "loss": 0.7676,
+      "step": 3686
+    },
+    {
+      "epoch": 9.014669926650367,
+      "grad_norm": 0.34380633238781994,
+      "learning_rate": 1.720155958846513e-06,
+      "loss": 0.67,
+      "step": 3687
+    },
+    {
+      "epoch": 9.017114914425427,
+      "grad_norm": 0.19034952715698975,
+      "learning_rate": 1.7189397468415342e-06,
+      "loss": 0.4378,
+      "step": 3688
+    },
+    {
+      "epoch": 9.01955990220049,
+      "grad_norm": 0.19875623691519614,
+      "learning_rate": 1.7177237396484333e-06,
+      "loss": 0.4273,
+      "step": 3689
+    },
+    {
+      "epoch": 9.02200488997555,
+      "grad_norm": 0.21094727794638565,
+      "learning_rate": 1.7165079375860755e-06,
+      "loss": 0.372,
+      "step": 3690
+    },
+    {
+      "epoch": 9.02444987775061,
+      "grad_norm": 0.22221619136656934,
+      "learning_rate": 1.7152923409732708e-06,
+      "loss": 0.6271,
+      "step": 3691
+    },
+    {
+      "epoch": 9.026894865525673,
+      "grad_norm": 0.2648316594089732,
+      "learning_rate": 1.7140769501287779e-06,
+      "loss": 0.6083,
+      "step": 3692
+    },
+    {
+      "epoch": 9.029339853300733,
+      "grad_norm": 0.17813606891911415,
+      "learning_rate": 1.7128617653712995e-06,
+      "loss": 0.5181,
+      "step": 3693
+    },
+    {
+      "epoch": 9.031784841075794,
+      "grad_norm": 0.25785981157895854,
+      "learning_rate": 1.7116467870194848e-06,
+      "loss": 0.7364,
+      "step": 3694
+    },
+    {
+      "epoch": 9.034229828850856,
+      "grad_norm": 0.20646755174345485,
+      "learning_rate": 1.7104320153919282e-06,
+      "loss": 0.4577,
+      "step": 3695
+    },
+    {
+      "epoch": 9.036674816625917,
+      "grad_norm": 0.26057312865616206,
+      "learning_rate": 1.7092174508071723e-06,
+      "loss": 0.3341,
+      "step": 3696
+    },
+    {
+      "epoch": 9.039119804400977,
+      "grad_norm": 0.3090236312135023,
+      "learning_rate": 1.7080030935837028e-06,
+      "loss": 0.4987,
+      "step": 3697
+    },
+    {
+      "epoch": 9.04156479217604,
+      "grad_norm": 0.19374664850984177,
+      "learning_rate": 1.706788944039952e-06,
+      "loss": 0.4914,
+      "step": 3698
+    },
+    {
+      "epoch": 9.0440097799511,
+      "grad_norm": 0.17954118068271388,
+      "learning_rate": 1.705575002494298e-06,
+      "loss": 0.5855,
+      "step": 3699
+    },
+    {
+      "epoch": 9.04645476772616,
+      "grad_norm": 0.3357297595797145,
+      "learning_rate": 1.704361269265064e-06,
+      "loss": 0.7081,
+      "step": 3700
+    },
+    {
+      "epoch": 9.048899755501223,
+      "grad_norm": 0.22778642926523404,
+      "learning_rate": 1.7031477446705177e-06,
+      "loss": 0.4111,
+      "step": 3701
+    },
+    {
+      "epoch": 9.051344743276283,
+      "grad_norm": 0.29308149555047897,
+      "learning_rate": 1.7019344290288745e-06,
+      "loss": 0.6798,
+      "step": 3702
+    },
+    {
+      "epoch": 9.053789731051344,
+      "grad_norm": 0.2262081971621139,
+      "learning_rate": 1.7007213226582925e-06,
+      "loss": 1.1001,
+      "step": 3703
+    },
+    {
+      "epoch": 9.056234718826406,
+      "grad_norm": 0.2887851866884744,
+      "learning_rate": 1.6995084258768768e-06,
+      "loss": 0.6088,
+      "step": 3704
+    },
+    {
+      "epoch": 9.058679706601467,
+      "grad_norm": 0.17916203359554328,
+      "learning_rate": 1.698295739002675e-06,
+      "loss": 0.6397,
+      "step": 3705
+    },
+    {
+      "epoch": 9.061124694376527,
+      "grad_norm": 0.2160105868811308,
+      "learning_rate": 1.6970832623536831e-06,
+      "loss": 0.3685,
+      "step": 3706
+    },
+    {
+      "epoch": 9.06356968215159,
+      "grad_norm": 0.33142951810255217,
+      "learning_rate": 1.6958709962478393e-06,
+      "loss": 0.5572,
+      "step": 3707
+    },
+    {
+      "epoch": 9.06601466992665,
+      "grad_norm": 0.2087830952399488,
+      "learning_rate": 1.6946589410030287e-06,
+      "loss": 0.647,
+      "step": 3708
+    },
+    {
+      "epoch": 9.06845965770171,
+      "grad_norm": 0.20544846030078978,
+      "learning_rate": 1.693447096937079e-06,
+      "loss": 0.5347,
+      "step": 3709
+    },
+    {
+      "epoch": 9.070904645476773,
+      "grad_norm": 0.3136290222947587,
+      "learning_rate": 1.6922354643677635e-06,
+      "loss": 0.6184,
+      "step": 3710
+    },
+    {
+      "epoch": 9.073349633251834,
+      "grad_norm": 0.20037297567558165,
+      "learning_rate": 1.6910240436127998e-06,
+      "loss": 0.5755,
+      "step": 3711
+    },
+    {
+      "epoch": 9.075794621026894,
+      "grad_norm": 0.28240057248266964,
+      "learning_rate": 1.6898128349898506e-06,
+      "loss": 0.4307,
+      "step": 3712
+    },
+    {
+      "epoch": 9.078239608801956,
+      "grad_norm": 0.21289114597158304,
+      "learning_rate": 1.6886018388165225e-06,
+      "loss": 0.5909,
+      "step": 3713
+    },
+    {
+      "epoch": 9.080684596577017,
+      "grad_norm": 0.2560310542811439,
+      "learning_rate": 1.6873910554103665e-06,
+      "loss": 0.5309,
+      "step": 3714
+    },
+    {
+      "epoch": 9.083129584352077,
+      "grad_norm": 0.1899935635694557,
+      "learning_rate": 1.686180485088877e-06,
+      "loss": 0.4388,
+      "step": 3715
+    },
+    {
+      "epoch": 9.08557457212714,
+      "grad_norm": 0.17742918419061687,
+      "learning_rate": 1.6849701281694939e-06,
+      "loss": 0.4545,
+      "step": 3716
+    },
+    {
+      "epoch": 9.0880195599022,
+      "grad_norm": 0.2944865736054724,
+      "learning_rate": 1.6837599849695997e-06,
+      "loss": 1.0101,
+      "step": 3717
+    },
+    {
+      "epoch": 9.09046454767726,
+      "grad_norm": 0.23322332682570077,
+      "learning_rate": 1.6825500558065227e-06,
+      "loss": 0.6963,
+      "step": 3718
+    },
+    {
+      "epoch": 9.092909535452323,
+      "grad_norm": 0.20870916576435303,
+      "learning_rate": 1.6813403409975343e-06,
+      "loss": 0.7526,
+      "step": 3719
+    },
+    {
+      "epoch": 9.095354523227384,
+      "grad_norm": 0.21417039188762763,
+      "learning_rate": 1.6801308408598482e-06,
+      "loss": 0.5864,
+      "step": 3720
+    },
+    {
+      "epoch": 9.097799511002446,
+      "grad_norm": 0.23700313280661495,
+      "learning_rate": 1.6789215557106235e-06,
+      "loss": 0.508,
+      "step": 3721
+    },
+    {
+      "epoch": 9.100244498777506,
+      "grad_norm": 0.17053782317739405,
+      "learning_rate": 1.6777124858669627e-06,
+      "loss": 0.5249,
+      "step": 3722
+    },
+    {
+      "epoch": 9.102689486552567,
+      "grad_norm": 0.3076115590266008,
+      "learning_rate": 1.6765036316459116e-06,
+      "loss": 0.6926,
+      "step": 3723
+    },
+    {
+      "epoch": 9.10513447432763,
+      "grad_norm": 0.2842074746253879,
+      "learning_rate": 1.675294993364459e-06,
+      "loss": 0.3693,
+      "step": 3724
+    },
+    {
+      "epoch": 9.10757946210269,
+      "grad_norm": 0.35671516508547924,
+      "learning_rate": 1.6740865713395375e-06,
+      "loss": 0.5412,
+      "step": 3725
+    },
+    {
+      "epoch": 9.11002444987775,
+      "grad_norm": 0.3030085451625398,
+      "learning_rate": 1.6728783658880237e-06,
+      "loss": 0.6223,
+      "step": 3726
+    },
+    {
+      "epoch": 9.112469437652813,
+      "grad_norm": 0.364928485959764,
+      "learning_rate": 1.6716703773267362e-06,
+      "loss": 0.5489,
+      "step": 3727
+    },
+    {
+      "epoch": 9.114914425427873,
+      "grad_norm": 0.36402170204914,
+      "learning_rate": 1.6704626059724365e-06,
+      "loss": 0.4529,
+      "step": 3728
+    },
+    {
+      "epoch": 9.117359413202934,
+      "grad_norm": 0.32530071850513187,
+      "learning_rate": 1.6692550521418315e-06,
+      "loss": 0.795,
+      "step": 3729
+    },
+    {
+      "epoch": 9.119804400977996,
+      "grad_norm": 0.24370227121304872,
+      "learning_rate": 1.668047716151569e-06,
+      "loss": 0.6319,
+      "step": 3730
+    },
+    {
+      "epoch": 9.122249388753056,
+      "grad_norm": 0.23501026701979078,
+      "learning_rate": 1.6668405983182396e-06,
+      "loss": 0.8241,
+      "step": 3731
+    },
+    {
+      "epoch": 9.124694376528117,
+      "grad_norm": 0.21172114971586284,
+      "learning_rate": 1.6656336989583774e-06,
+      "loss": 0.837,
+      "step": 3732
+    },
+    {
+      "epoch": 9.12713936430318,
+      "grad_norm": 0.3014212597291553,
+      "learning_rate": 1.6644270183884591e-06,
+      "loss": 0.7785,
+      "step": 3733
+    },
+    {
+      "epoch": 9.12958435207824,
+      "grad_norm": 0.23603569185838502,
+      "learning_rate": 1.6632205569249043e-06,
+      "loss": 0.458,
+      "step": 3734
+    },
+    {
+      "epoch": 9.1320293398533,
+      "grad_norm": 0.19829630956475972,
+      "learning_rate": 1.6620143148840743e-06,
+      "loss": 0.6728,
+      "step": 3735
+    },
+    {
+      "epoch": 9.134474327628363,
+      "grad_norm": 0.30043491651235604,
+      "learning_rate": 1.6608082925822736e-06,
+      "loss": 0.9533,
+      "step": 3736
+    },
+    {
+      "epoch": 9.136919315403423,
+      "grad_norm": 0.2716964335105591,
+      "learning_rate": 1.659602490335749e-06,
+      "loss": 0.458,
+      "step": 3737
+    },
+    {
+      "epoch": 9.139364303178484,
+      "grad_norm": 0.2368201972049681,
+      "learning_rate": 1.6583969084606882e-06,
+      "loss": 0.7162,
+      "step": 3738
+    },
+    {
+      "epoch": 9.141809290953546,
+      "grad_norm": 0.22048558433269078,
+      "learning_rate": 1.6571915472732244e-06,
+      "loss": 0.6522,
+      "step": 3739
+    },
+    {
+      "epoch": 9.144254278728607,
+      "grad_norm": 0.1781830068774942,
+      "learning_rate": 1.6559864070894296e-06,
+      "loss": 0.4682,
+      "step": 3740
+    },
+    {
+      "epoch": 9.146699266503667,
+      "grad_norm": 0.20422760914985105,
+      "learning_rate": 1.6547814882253199e-06,
+      "loss": 0.3882,
+      "step": 3741
+    },
+    {
+      "epoch": 9.14914425427873,
+      "grad_norm": 0.24461098580943325,
+      "learning_rate": 1.6535767909968523e-06,
+      "loss": 0.8867,
+      "step": 3742
+    },
+    {
+      "epoch": 9.15158924205379,
+      "grad_norm": 0.26469061251149767,
+      "learning_rate": 1.6523723157199256e-06,
+      "loss": 0.8181,
+      "step": 3743
+    },
+    {
+      "epoch": 9.15403422982885,
+      "grad_norm": 0.16941790869842885,
+      "learning_rate": 1.6511680627103815e-06,
+      "loss": 0.744,
+      "step": 3744
+    },
+    {
+      "epoch": 9.156479217603913,
+      "grad_norm": 0.2265218410292478,
+      "learning_rate": 1.6499640322840015e-06,
+      "loss": 0.4978,
+      "step": 3745
+    },
+    {
+      "epoch": 9.158924205378973,
+      "grad_norm": 0.1738015802257966,
+      "learning_rate": 1.6487602247565113e-06,
+      "loss": 0.512,
+      "step": 3746
+    },
+    {
+      "epoch": 9.161369193154034,
+      "grad_norm": 0.17724390341372845,
+      "learning_rate": 1.647556640443576e-06,
+      "loss": 0.9639,
+      "step": 3747
+    },
+    {
+      "epoch": 9.163814180929096,
+      "grad_norm": 0.19530979704047577,
+      "learning_rate": 1.6463532796608033e-06,
+      "loss": 0.4142,
+      "step": 3748
+    },
+    {
+      "epoch": 9.166259168704157,
+      "grad_norm": 0.36818484818698294,
+      "learning_rate": 1.6451501427237409e-06,
+      "loss": 0.5216,
+      "step": 3749
+    },
+    {
+      "epoch": 9.168704156479217,
+      "grad_norm": 0.27025054447724406,
+      "learning_rate": 1.6439472299478804e-06,
+      "loss": 0.8026,
+      "step": 3750
+    },
+    {
+      "epoch": 9.17114914425428,
+      "grad_norm": 0.2535552631533771,
+      "learning_rate": 1.6427445416486526e-06,
+      "loss": 0.8205,
+      "step": 3751
+    },
+    {
+      "epoch": 9.17359413202934,
+      "grad_norm": 0.31793232299305474,
+      "learning_rate": 1.6415420781414298e-06,
+      "loss": 0.6869,
+      "step": 3752
+    },
+    {
+      "epoch": 9.1760391198044,
+      "grad_norm": 0.2601201240994496,
+      "learning_rate": 1.6403398397415257e-06,
+      "loss": 0.797,
+      "step": 3753
+    },
+    {
+      "epoch": 9.178484107579463,
+      "grad_norm": 0.2747501640079801,
+      "learning_rate": 1.639137826764194e-06,
+      "loss": 0.4444,
+      "step": 3754
+    },
+    {
+      "epoch": 9.180929095354523,
+      "grad_norm": 0.299737542945522,
+      "learning_rate": 1.6379360395246302e-06,
+      "loss": 0.4071,
+      "step": 3755
+    },
+    {
+      "epoch": 9.183374083129584,
+      "grad_norm": 0.22879057273475878,
+      "learning_rate": 1.6367344783379713e-06,
+      "loss": 0.5856,
+      "step": 3756
+    },
+    {
+      "epoch": 9.185819070904646,
+      "grad_norm": 0.23508058587943995,
+      "learning_rate": 1.6355331435192931e-06,
+      "loss": 0.7579,
+      "step": 3757
+    },
+    {
+      "epoch": 9.188264058679707,
+      "grad_norm": 0.2544698070829883,
+      "learning_rate": 1.6343320353836136e-06,
+      "loss": 0.5149,
+      "step": 3758
+    },
+    {
+      "epoch": 9.190709046454767,
+      "grad_norm": 0.26067841063301544,
+      "learning_rate": 1.6331311542458905e-06,
+      "loss": 0.6982,
+      "step": 3759
+    },
+    {
+      "epoch": 9.19315403422983,
+      "grad_norm": 0.36520995289866987,
+      "learning_rate": 1.6319305004210218e-06,
+      "loss": 0.3495,
+      "step": 3760
+    },
+    {
+      "epoch": 9.19559902200489,
+      "grad_norm": 0.1892094000401061,
+      "learning_rate": 1.6307300742238483e-06,
+      "loss": 0.5657,
+      "step": 3761
+    },
+    {
+      "epoch": 9.19804400977995,
+      "grad_norm": 0.2474017554639616,
+      "learning_rate": 1.629529875969148e-06,
+      "loss": 0.3428,
+      "step": 3762
+    },
+    {
+      "epoch": 9.200488997555013,
+      "grad_norm": 0.25912866439841126,
+      "learning_rate": 1.6283299059716407e-06,
+      "loss": 0.9238,
+      "step": 3763
+    },
+    {
+      "epoch": 9.202933985330073,
+      "grad_norm": 0.2650997651041663,
+      "learning_rate": 1.6271301645459858e-06,
+      "loss": 0.5867,
+      "step": 3764
+    },
+    {
+      "epoch": 9.205378973105134,
+      "grad_norm": 0.32228420738556607,
+      "learning_rate": 1.6259306520067822e-06,
+      "loss": 0.4173,
+      "step": 3765
+    },
+    {
+      "epoch": 9.207823960880196,
+      "grad_norm": 0.506277531583115,
+      "learning_rate": 1.6247313686685707e-06,
+      "loss": 0.7579,
+      "step": 3766
+    },
+    {
+      "epoch": 9.210268948655257,
+      "grad_norm": 0.1915610389020002,
+      "learning_rate": 1.6235323148458307e-06,
+      "loss": 0.3681,
+      "step": 3767
+    },
+    {
+      "epoch": 9.212713936430317,
+      "grad_norm": 0.1671853483903209,
+      "learning_rate": 1.622333490852981e-06,
+      "loss": 0.7176,
+      "step": 3768
+    },
+    {
+      "epoch": 9.21515892420538,
+      "grad_norm": 0.23659874535830658,
+      "learning_rate": 1.6211348970043807e-06,
+      "loss": 0.8612,
+      "step": 3769
+    },
+    {
+      "epoch": 9.21760391198044,
+      "grad_norm": 0.3086686194070926,
+      "learning_rate": 1.6199365336143297e-06,
+      "loss": 0.5305,
+      "step": 3770
+    },
+    {
+      "epoch": 9.2200488997555,
+      "grad_norm": 0.23544422666156783,
+      "learning_rate": 1.6187384009970639e-06,
+      "loss": 1.073,
+      "step": 3771
+    },
+    {
+      "epoch": 9.222493887530563,
+      "grad_norm": 0.31378595219232436,
+      "learning_rate": 1.6175404994667637e-06,
+      "loss": 0.5475,
+      "step": 3772
+    },
+    {
+      "epoch": 9.224938875305623,
+      "grad_norm": 0.2913281005225974,
+      "learning_rate": 1.6163428293375455e-06,
+      "loss": 0.4706,
+      "step": 3773
+    },
+    {
+      "epoch": 9.227383863080684,
+      "grad_norm": 0.2124143285837525,
+      "learning_rate": 1.6151453909234649e-06,
+      "loss": 0.6314,
+      "step": 3774
+    },
+    {
+      "epoch": 9.229828850855746,
+      "grad_norm": 0.17088870560903544,
+      "learning_rate": 1.6139481845385186e-06,
+      "loss": 0.6387,
+      "step": 3775
+    },
+    {
+      "epoch": 9.232273838630807,
+      "grad_norm": 0.22286428608958236,
+      "learning_rate": 1.6127512104966413e-06,
+      "loss": 0.437,
+      "step": 3776
+    },
+    {
+      "epoch": 9.234718826405867,
+      "grad_norm": 0.2856527412349572,
+      "learning_rate": 1.6115544691117074e-06,
+      "loss": 0.7422,
+      "step": 3777
+    },
+    {
+      "epoch": 9.23716381418093,
+      "grad_norm": 0.29922213258273284,
+      "learning_rate": 1.6103579606975291e-06,
+      "loss": 0.3501,
+      "step": 3778
+    },
+    {
+      "epoch": 9.23960880195599,
+      "grad_norm": 0.256104060447046,
+      "learning_rate": 1.6091616855678583e-06,
+      "loss": 0.4883,
+      "step": 3779
+    },
+    {
+      "epoch": 9.24205378973105,
+      "grad_norm": 0.3106199968381704,
+      "learning_rate": 1.6079656440363864e-06,
+      "loss": 0.4477,
+      "step": 3780
+    },
+    {
+      "epoch": 9.244498777506113,
+      "grad_norm": 0.24593255016935583,
+      "learning_rate": 1.6067698364167418e-06,
+      "loss": 0.3889,
+      "step": 3781
+    },
+    {
+      "epoch": 9.246943765281173,
+      "grad_norm": 0.13855123344890902,
+      "learning_rate": 1.6055742630224939e-06,
+      "loss": 0.3928,
+      "step": 3782
+    },
+    {
+      "epoch": 9.249388753056234,
+      "grad_norm": 0.20610135502255728,
+      "learning_rate": 1.6043789241671489e-06,
+      "loss": 0.8226,
+      "step": 3783
+    },
+    {
+      "epoch": 9.251833740831296,
+      "grad_norm": 0.24231371367318064,
+      "learning_rate": 1.6031838201641517e-06,
+      "loss": 0.4779,
+      "step": 3784
+    },
+    {
+      "epoch": 9.254278728606357,
+      "grad_norm": 0.3166356702014923,
+      "learning_rate": 1.6019889513268867e-06,
+      "loss": 0.5674,
+      "step": 3785
+    },
+    {
+      "epoch": 9.256723716381417,
+      "grad_norm": 0.35368932502481937,
+      "learning_rate": 1.6007943179686752e-06,
+      "loss": 0.475,
+      "step": 3786
+    },
+    {
+      "epoch": 9.25916870415648,
+      "grad_norm": 0.429863933558201,
+      "learning_rate": 1.599599920402778e-06,
+      "loss": 0.3623,
+      "step": 3787
+    },
+    {
+      "epoch": 9.26161369193154,
+      "grad_norm": 0.24244222291930287,
+      "learning_rate": 1.5984057589423923e-06,
+      "loss": 0.6596,
+      "step": 3788
+    },
+    {
+      "epoch": 9.2640586797066,
+      "grad_norm": 0.4125940137785228,
+      "learning_rate": 1.5972118339006554e-06,
+      "loss": 0.3446,
+      "step": 3789
+    },
+    {
+      "epoch": 9.266503667481663,
+      "grad_norm": 0.30449985060980256,
+      "learning_rate": 1.5960181455906421e-06,
+      "loss": 0.8954,
+      "step": 3790
+    },
+    {
+      "epoch": 9.268948655256724,
+      "grad_norm": 0.23175075675837326,
+      "learning_rate": 1.5948246943253646e-06,
+      "loss": 0.6334,
+      "step": 3791
+    },
+    {
+      "epoch": 9.271393643031784,
+      "grad_norm": 0.20916099129299942,
+      "learning_rate": 1.5936314804177716e-06,
+      "loss": 0.694,
+      "step": 3792
+    },
+    {
+      "epoch": 9.273838630806846,
+      "grad_norm": 0.28891342757622707,
+      "learning_rate": 1.5924385041807536e-06,
+      "loss": 0.4279,
+      "step": 3793
+    },
+    {
+      "epoch": 9.276283618581907,
+      "grad_norm": 0.2404900443998669,
+      "learning_rate": 1.5912457659271347e-06,
+      "loss": 0.7967,
+      "step": 3794
+    },
+    {
+      "epoch": 9.278728606356967,
+      "grad_norm": 0.2750882550804834,
+      "learning_rate": 1.5900532659696789e-06,
+      "loss": 0.7529,
+      "step": 3795
+    },
+    {
+      "epoch": 9.28117359413203,
+      "grad_norm": 0.30912787736361075,
+      "learning_rate": 1.5888610046210866e-06,
+      "loss": 0.4239,
+      "step": 3796
+    },
+    {
+      "epoch": 9.28361858190709,
+      "grad_norm": 0.2508255942827207,
+      "learning_rate": 1.5876689821939966e-06,
+      "loss": 0.4374,
+      "step": 3797
+    },
+    {
+      "epoch": 9.28606356968215,
+      "grad_norm": 0.23100608921988958,
+      "learning_rate": 1.5864771990009831e-06,
+      "loss": 0.4373,
+      "step": 3798
+    },
+    {
+      "epoch": 9.288508557457213,
+      "grad_norm": 0.28046214821294163,
+      "learning_rate": 1.5852856553545604e-06,
+      "loss": 0.3865,
+      "step": 3799
+    },
+    {
+      "epoch": 9.290953545232274,
+      "grad_norm": 0.390727020032442,
+      "learning_rate": 1.584094351567178e-06,
+      "loss": 0.3697,
+      "step": 3800
+    },
+    {
+      "epoch": 9.293398533007334,
+      "grad_norm": 0.2185051166879761,
+      "learning_rate": 1.582903287951223e-06,
+      "loss": 0.3471,
+      "step": 3801
+    },
+    {
+      "epoch": 9.295843520782396,
+      "grad_norm": 0.22014036037899523,
+      "learning_rate": 1.5817124648190192e-06,
+      "loss": 0.718,
+      "step": 3802
+    },
+    {
+      "epoch": 9.298288508557457,
+      "grad_norm": 0.27577483119513485,
+      "learning_rate": 1.5805218824828273e-06,
+      "loss": 0.3761,
+      "step": 3803
+    },
+    {
+      "epoch": 9.30073349633252,
+      "grad_norm": 0.2625618545995872,
+      "learning_rate": 1.579331541254846e-06,
+      "loss": 0.5664,
+      "step": 3804
+    },
+    {
+      "epoch": 9.30317848410758,
+      "grad_norm": 0.36177283522201026,
+      "learning_rate": 1.5781414414472108e-06,
+      "loss": 0.4816,
+      "step": 3805
+    },
+    {
+      "epoch": 9.30562347188264,
+      "grad_norm": 0.31061749211889167,
+      "learning_rate": 1.5769515833719922e-06,
+      "loss": 0.6579,
+      "step": 3806
+    },
+    {
+      "epoch": 9.3080684596577,
+      "grad_norm": 0.2036908061958182,
+      "learning_rate": 1.5757619673411984e-06,
+      "loss": 0.8445,
+      "step": 3807
+    },
+    {
+      "epoch": 9.310513447432763,
+      "grad_norm": 0.1590616427309036,
+      "learning_rate": 1.5745725936667728e-06,
+      "loss": 0.5905,
+      "step": 3808
+    },
+    {
+      "epoch": 9.312958435207824,
+      "grad_norm": 0.327381254181539,
+      "learning_rate": 1.5733834626605984e-06,
+      "loss": 0.655,
+      "step": 3809
+    },
+    {
+      "epoch": 9.315403422982886,
+      "grad_norm": 0.26672461852296214,
+      "learning_rate": 1.5721945746344916e-06,
+      "loss": 0.4225,
+      "step": 3810
+    },
+    {
+      "epoch": 9.317848410757946,
+      "grad_norm": 0.24269502735188786,
+      "learning_rate": 1.5710059299002062e-06,
+      "loss": 0.4836,
+      "step": 3811
+    },
+    {
+      "epoch": 9.320293398533007,
+      "grad_norm": 0.340286871284897,
+      "learning_rate": 1.5698175287694314e-06,
+      "loss": 0.7579,
+      "step": 3812
+    },
+    {
+      "epoch": 9.32273838630807,
+      "grad_norm": 0.2805701708215805,
+      "learning_rate": 1.5686293715537931e-06,
+      "loss": 0.4126,
+      "step": 3813
+    },
+    {
+      "epoch": 9.32518337408313,
+      "grad_norm": 0.25583317919518245,
+      "learning_rate": 1.5674414585648546e-06,
+      "loss": 0.5164,
+      "step": 3814
+    },
+    {
+      "epoch": 9.32762836185819,
+      "grad_norm": 0.21931110676576776,
+      "learning_rate": 1.5662537901141134e-06,
+      "loss": 0.3632,
+      "step": 3815
+    },
+    {
+      "epoch": 9.330073349633253,
+      "grad_norm": 0.2300101633693936,
+      "learning_rate": 1.565066366513003e-06,
+      "loss": 0.5609,
+      "step": 3816
+    },
+    {
+      "epoch": 9.332518337408313,
+      "grad_norm": 0.2004595426909433,
+      "learning_rate": 1.5638791880728934e-06,
+      "loss": 0.4431,
+      "step": 3817
+    },
+    {
+      "epoch": 9.334963325183374,
+      "grad_norm": 0.3680087940803977,
+      "learning_rate": 1.5626922551050893e-06,
+      "loss": 0.6512,
+      "step": 3818
+    },
+    {
+      "epoch": 9.337408312958436,
+      "grad_norm": 0.33855738738370084,
+      "learning_rate": 1.5615055679208328e-06,
+      "loss": 1.0626,
+      "step": 3819
+    },
+    {
+      "epoch": 9.339853300733497,
+      "grad_norm": 0.17288329992854268,
+      "learning_rate": 1.5603191268312996e-06,
+      "loss": 0.4605,
+      "step": 3820
+    },
+    {
+      "epoch": 9.342298288508557,
+      "grad_norm": 0.25364415036630994,
+      "learning_rate": 1.5591329321476021e-06,
+      "loss": 0.5384,
+      "step": 3821
+    },
+    {
+      "epoch": 9.34474327628362,
+      "grad_norm": 0.3390621030101342,
+      "learning_rate": 1.5579469841807868e-06,
+      "loss": 0.4693,
+      "step": 3822
+    },
+    {
+      "epoch": 9.34718826405868,
+      "grad_norm": 0.3291154609181924,
+      "learning_rate": 1.5567612832418378e-06,
+      "loss": 0.8564,
+      "step": 3823
+    },
+    {
+      "epoch": 9.34963325183374,
+      "grad_norm": 0.25787013124519487,
+      "learning_rate": 1.5555758296416712e-06,
+      "loss": 0.7168,
+      "step": 3824
+    },
+    {
+      "epoch": 9.352078239608803,
+      "grad_norm": 0.24642638379136683,
+      "learning_rate": 1.5543906236911423e-06,
+      "loss": 0.8619,
+      "step": 3825
+    },
+    {
+      "epoch": 9.354523227383863,
+      "grad_norm": 0.2607012277639971,
+      "learning_rate": 1.5532056657010383e-06,
+      "loss": 0.4176,
+      "step": 3826
+    },
+    {
+      "epoch": 9.356968215158924,
+      "grad_norm": 0.21880186727465328,
+      "learning_rate": 1.552020955982082e-06,
+      "loss": 0.5805,
+      "step": 3827
+    },
+    {
+      "epoch": 9.359413202933986,
+      "grad_norm": 0.24452764654808748,
+      "learning_rate": 1.5508364948449311e-06,
+      "loss": 0.6891,
+      "step": 3828
+    },
+    {
+      "epoch": 9.361858190709047,
+      "grad_norm": 0.261778046619333,
+      "learning_rate": 1.5496522826001795e-06,
+      "loss": 0.4183,
+      "step": 3829
+    },
+    {
+      "epoch": 9.364303178484107,
+      "grad_norm": 0.24457021633396422,
+      "learning_rate": 1.5484683195583542e-06,
+      "loss": 0.6153,
+      "step": 3830
+    },
+    {
+      "epoch": 9.36674816625917,
+      "grad_norm": 0.2788064621134573,
+      "learning_rate": 1.5472846060299175e-06,
+      "loss": 0.7051,
+      "step": 3831
+    },
+    {
+      "epoch": 9.36919315403423,
+      "grad_norm": 0.18139567183952385,
+      "learning_rate": 1.5461011423252651e-06,
+      "loss": 0.3695,
+      "step": 3832
+    },
+    {
+      "epoch": 9.37163814180929,
+      "grad_norm": 0.2343131057488101,
+      "learning_rate": 1.54491792875473e-06,
+      "loss": 0.4752,
+      "step": 3833
+    },
+    {
+      "epoch": 9.374083129584353,
+      "grad_norm": 0.21484676209749104,
+      "learning_rate": 1.5437349656285773e-06,
+      "loss": 1.0612,
+      "step": 3834
+    },
+    {
+      "epoch": 9.376528117359413,
+      "grad_norm": 0.25035329740727935,
+      "learning_rate": 1.5425522532570052e-06,
+      "loss": 0.4031,
+      "step": 3835
+    },
+    {
+      "epoch": 9.378973105134474,
+      "grad_norm": 0.26642890222957083,
+      "learning_rate": 1.5413697919501511e-06,
+      "loss": 0.8965,
+      "step": 3836
+    },
+    {
+      "epoch": 9.381418092909536,
+      "grad_norm": 0.2719855763447448,
+      "learning_rate": 1.5401875820180818e-06,
+      "loss": 0.681,
+      "step": 3837
+    },
+    {
+      "epoch": 9.383863080684597,
+      "grad_norm": 0.23550996583705056,
+      "learning_rate": 1.5390056237707995e-06,
+      "loss": 0.5173,
+      "step": 3838
+    },
+    {
+      "epoch": 9.386308068459657,
+      "grad_norm": 0.2269722291731216,
+      "learning_rate": 1.5378239175182414e-06,
+      "loss": 0.4988,
+      "step": 3839
+    },
+    {
+      "epoch": 9.38875305623472,
+      "grad_norm": 0.2748394123876589,
+      "learning_rate": 1.5366424635702777e-06,
+      "loss": 0.3344,
+      "step": 3840
+    },
+    {
+      "epoch": 9.39119804400978,
+      "grad_norm": 0.34625633943457906,
+      "learning_rate": 1.5354612622367127e-06,
+      "loss": 0.4166,
+      "step": 3841
+    },
+    {
+      "epoch": 9.39364303178484,
+      "grad_norm": 0.41023735900581076,
+      "learning_rate": 1.5342803138272839e-06,
+      "loss": 0.3633,
+      "step": 3842
+    },
+    {
+      "epoch": 9.396088019559903,
+      "grad_norm": 0.27251988760932444,
+      "learning_rate": 1.5330996186516638e-06,
+      "loss": 0.4134,
+      "step": 3843
+    },
+    {
+      "epoch": 9.398533007334963,
+      "grad_norm": 0.26046606941966505,
+      "learning_rate": 1.5319191770194575e-06,
+      "loss": 0.427,
+      "step": 3844
+    },
+    {
+      "epoch": 9.400977995110024,
+      "grad_norm": 0.18528006823525117,
+      "learning_rate": 1.5307389892402036e-06,
+      "loss": 0.4514,
+      "step": 3845
+    },
+    {
+      "epoch": 9.403422982885086,
+      "grad_norm": 0.42183434125366914,
+      "learning_rate": 1.5295590556233736e-06,
+      "loss": 0.6703,
+      "step": 3846
+    },
+    {
+      "epoch": 9.405867970660147,
+      "grad_norm": 0.3267238734885524,
+      "learning_rate": 1.528379376478375e-06,
+      "loss": 0.7672,
+      "step": 3847
+    },
+    {
+      "epoch": 9.408312958435207,
+      "grad_norm": 0.2593569367199505,
+      "learning_rate": 1.5271999521145447e-06,
+      "loss": 0.437,
+      "step": 3848
+    },
+    {
+      "epoch": 9.41075794621027,
+      "grad_norm": 0.3508763176999022,
+      "learning_rate": 1.5260207828411565e-06,
+      "loss": 0.7,
+      "step": 3849
+    },
+    {
+      "epoch": 9.41320293398533,
+      "grad_norm": 0.3925283838131019,
+      "learning_rate": 1.5248418689674146e-06,
+      "loss": 0.3971,
+      "step": 3850
+    },
+    {
+      "epoch": 9.41564792176039,
+      "grad_norm": 0.282276042080178,
+      "learning_rate": 1.5236632108024573e-06,
+      "loss": 0.5469,
+      "step": 3851
+    },
+    {
+      "epoch": 9.418092909535453,
+      "grad_norm": 0.2483119944427397,
+      "learning_rate": 1.5224848086553557e-06,
+      "loss": 0.4509,
+      "step": 3852
+    },
+    {
+      "epoch": 9.420537897310513,
+      "grad_norm": 0.21266501317302733,
+      "learning_rate": 1.521306662835114e-06,
+      "loss": 0.6013,
+      "step": 3853
+    },
+    {
+      "epoch": 9.422982885085574,
+      "grad_norm": 0.18382479990475614,
+      "learning_rate": 1.5201287736506693e-06,
+      "loss": 0.3147,
+      "step": 3854
+    },
+    {
+      "epoch": 9.425427872860636,
+      "grad_norm": 0.1716753842204437,
+      "learning_rate": 1.5189511414108905e-06,
+      "loss": 0.519,
+      "step": 3855
+    },
+    {
+      "epoch": 9.427872860635697,
+      "grad_norm": 0.27258692756150965,
+      "learning_rate": 1.5177737664245789e-06,
+      "loss": 0.3697,
+      "step": 3856
+    },
+    {
+      "epoch": 9.430317848410757,
+      "grad_norm": 0.38881125335082445,
+      "learning_rate": 1.5165966490004714e-06,
+      "loss": 0.5634,
+      "step": 3857
+    },
+    {
+      "epoch": 9.43276283618582,
+      "grad_norm": 0.19525853814771407,
+      "learning_rate": 1.5154197894472334e-06,
+      "loss": 0.4505,
+      "step": 3858
+    },
+    {
+      "epoch": 9.43520782396088,
+      "grad_norm": 0.2580931913442357,
+      "learning_rate": 1.5142431880734653e-06,
+      "loss": 0.3955,
+      "step": 3859
+    },
+    {
+      "epoch": 9.43765281173594,
+      "grad_norm": 0.37381110033515363,
+      "learning_rate": 1.513066845187699e-06,
+      "loss": 0.5465,
+      "step": 3860
+    },
+    {
+      "epoch": 9.440097799511003,
+      "grad_norm": 0.2513291669597155,
+      "learning_rate": 1.511890761098398e-06,
+      "loss": 0.6013,
+      "step": 3861
+    },
+    {
+      "epoch": 9.442542787286063,
+      "grad_norm": 0.2966032921538477,
+      "learning_rate": 1.5107149361139583e-06,
+      "loss": 0.4522,
+      "step": 3862
+    },
+    {
+      "epoch": 9.444987775061124,
+      "grad_norm": 0.26345497385855476,
+      "learning_rate": 1.5095393705427086e-06,
+      "loss": 0.4001,
+      "step": 3863
+    },
+    {
+      "epoch": 9.447432762836186,
+      "grad_norm": 0.20996728950229512,
+      "learning_rate": 1.508364064692909e-06,
+      "loss": 0.7981,
+      "step": 3864
+    },
+    {
+      "epoch": 9.449877750611247,
+      "grad_norm": 0.2173885801355247,
+      "learning_rate": 1.5071890188727517e-06,
+      "loss": 0.523,
+      "step": 3865
+    },
+    {
+      "epoch": 9.452322738386307,
+      "grad_norm": 0.28345858612891267,
+      "learning_rate": 1.50601423339036e-06,
+      "loss": 0.3318,
+      "step": 3866
+    },
+    {
+      "epoch": 9.45476772616137,
+      "grad_norm": 0.34560425686590196,
+      "learning_rate": 1.5048397085537897e-06,
+      "loss": 0.4381,
+      "step": 3867
+    },
+    {
+      "epoch": 9.45721271393643,
+      "grad_norm": 0.19511512400930675,
+      "learning_rate": 1.5036654446710285e-06,
+      "loss": 0.5559,
+      "step": 3868
+    },
+    {
+      "epoch": 9.45965770171149,
+      "grad_norm": 0.28881050558040083,
+      "learning_rate": 1.5024914420499958e-06,
+      "loss": 0.6507,
+      "step": 3869
+    },
+    {
+      "epoch": 9.462102689486553,
+      "grad_norm": 0.22786352830792714,
+      "learning_rate": 1.5013177009985413e-06,
+      "loss": 0.398,
+      "step": 3870
+    },
+    {
+      "epoch": 9.464547677261614,
+      "grad_norm": 0.2496813565935898,
+      "learning_rate": 1.5001442218244467e-06,
+      "loss": 0.4082,
+      "step": 3871
+    },
+    {
+      "epoch": 9.466992665036674,
+      "grad_norm": 0.15235222252555736,
+      "learning_rate": 1.4989710048354248e-06,
+      "loss": 0.4066,
+      "step": 3872
+    },
+    {
+      "epoch": 9.469437652811736,
+      "grad_norm": 0.15865061826644536,
+      "learning_rate": 1.4977980503391207e-06,
+      "loss": 0.3741,
+      "step": 3873
+    },
+    {
+      "epoch": 9.471882640586797,
+      "grad_norm": 0.24645026010520765,
+      "learning_rate": 1.4966253586431096e-06,
+      "loss": 0.7443,
+      "step": 3874
+    },
+    {
+      "epoch": 9.474327628361857,
+      "grad_norm": 0.26070164880621505,
+      "learning_rate": 1.4954529300548982e-06,
+      "loss": 0.6742,
+      "step": 3875
+    },
+    {
+      "epoch": 9.47677261613692,
+      "grad_norm": 0.398964430535094,
+      "learning_rate": 1.4942807648819234e-06,
+      "loss": 0.4894,
+      "step": 3876
+    },
+    {
+      "epoch": 9.47921760391198,
+      "grad_norm": 0.21327871634526044,
+      "learning_rate": 1.4931088634315548e-06,
+      "loss": 0.7871,
+      "step": 3877
+    },
+    {
+      "epoch": 9.48166259168704,
+      "grad_norm": 0.2598754737216216,
+      "learning_rate": 1.4919372260110906e-06,
+      "loss": 0.42,
+      "step": 3878
+    },
+    {
+      "epoch": 9.484107579462103,
+      "grad_norm": 0.26201329993546274,
+      "learning_rate": 1.4907658529277624e-06,
+      "loss": 0.3751,
+      "step": 3879
+    },
+    {
+      "epoch": 9.486552567237164,
+      "grad_norm": 0.20246205288724684,
+      "learning_rate": 1.4895947444887305e-06,
+      "loss": 0.3743,
+      "step": 3880
+    },
+    {
+      "epoch": 9.488997555012224,
+      "grad_norm": 0.1893651462240564,
+      "learning_rate": 1.488423901001086e-06,
+      "loss": 0.3432,
+      "step": 3881
+    },
+    {
+      "epoch": 9.491442542787286,
+      "grad_norm": 0.22569244998743596,
+      "learning_rate": 1.4872533227718505e-06,
+      "loss": 0.4795,
+      "step": 3882
+    },
+    {
+      "epoch": 9.493887530562347,
+      "grad_norm": 0.36169043847349547,
+      "learning_rate": 1.4860830101079777e-06,
+      "loss": 0.6255,
+      "step": 3883
+    },
+    {
+      "epoch": 9.496332518337407,
+      "grad_norm": 0.32117371882228074,
+      "learning_rate": 1.4849129633163495e-06,
+      "loss": 0.4783,
+      "step": 3884
+    },
+    {
+      "epoch": 9.49877750611247,
+      "grad_norm": 0.2728824745205743,
+      "learning_rate": 1.4837431827037787e-06,
+      "loss": 0.8893,
+      "step": 3885
+    },
+    {
+      "epoch": 9.50122249388753,
+      "grad_norm": 0.18820591643508766,
+      "learning_rate": 1.4825736685770088e-06,
+      "loss": 0.8901,
+      "step": 3886
+    },
+    {
+      "epoch": 9.503667481662593,
+      "grad_norm": 0.32183934404663295,
+      "learning_rate": 1.4814044212427137e-06,
+      "loss": 0.4946,
+      "step": 3887
+    },
+    {
+      "epoch": 9.506112469437653,
+      "grad_norm": 0.29151050706296716,
+      "learning_rate": 1.4802354410074962e-06,
+      "loss": 0.4417,
+      "step": 3888
+    },
+    {
+      "epoch": 9.508557457212714,
+      "grad_norm": 0.23114284631504753,
+      "learning_rate": 1.479066728177889e-06,
+      "loss": 0.4494,
+      "step": 3889
+    },
+    {
+      "epoch": 9.511002444987774,
+      "grad_norm": 0.21831575849741888,
+      "learning_rate": 1.4778982830603571e-06,
+      "loss": 0.859,
+      "step": 3890
+    },
+    {
+      "epoch": 9.513447432762836,
+      "grad_norm": 0.20068416926450727,
+      "learning_rate": 1.4767301059612928e-06,
+      "loss": 0.5435,
+      "step": 3891
+    },
+    {
+      "epoch": 9.515892420537897,
+      "grad_norm": 0.28036754127399266,
+      "learning_rate": 1.4755621971870182e-06,
+      "loss": 0.8721,
+      "step": 3892
+    },
+    {
+      "epoch": 9.51833740831296,
+      "grad_norm": 0.388235189390517,
+      "learning_rate": 1.474394557043787e-06,
+      "loss": 0.3815,
+      "step": 3893
+    },
+    {
+      "epoch": 9.52078239608802,
+      "grad_norm": 0.26923631005010584,
+      "learning_rate": 1.4732271858377804e-06,
+      "loss": 0.3218,
+      "step": 3894
+    },
+    {
+      "epoch": 9.52322738386308,
+      "grad_norm": 0.24796812926152387,
+      "learning_rate": 1.4720600838751098e-06,
+      "loss": 0.8166,
+      "step": 3895
+    },
+    {
+      "epoch": 9.525672371638143,
+      "grad_norm": 0.2741246544608282,
+      "learning_rate": 1.4708932514618156e-06,
+      "loss": 0.7367,
+      "step": 3896
+    },
+    {
+      "epoch": 9.528117359413203,
+      "grad_norm": 0.32784599263797454,
+      "learning_rate": 1.4697266889038692e-06,
+      "loss": 0.7006,
+      "step": 3897
+    },
+    {
+      "epoch": 9.530562347188264,
+      "grad_norm": 0.2348504195082149,
+      "learning_rate": 1.4685603965071694e-06,
+      "loss": 0.4815,
+      "step": 3898
+    },
+    {
+      "epoch": 9.533007334963326,
+      "grad_norm": 0.4112360504452804,
+      "learning_rate": 1.4673943745775431e-06,
+      "loss": 0.4292,
+      "step": 3899
+    },
+    {
+      "epoch": 9.535452322738386,
+      "grad_norm": 0.2882630232102204,
+      "learning_rate": 1.4662286234207511e-06,
+      "loss": 0.4322,
+      "step": 3900
+    },
+    {
+      "epoch": 9.537897310513447,
+      "grad_norm": 0.20217662879806014,
+      "learning_rate": 1.4650631433424783e-06,
+      "loss": 0.4758,
+      "step": 3901
+    },
+    {
+      "epoch": 9.54034229828851,
+      "grad_norm": 0.2501333131137029,
+      "learning_rate": 1.46389793464834e-06,
+      "loss": 0.4085,
+      "step": 3902
+    },
+    {
+      "epoch": 9.54278728606357,
+      "grad_norm": 0.2618879827763672,
+      "learning_rate": 1.4627329976438813e-06,
+      "loss": 0.425,
+      "step": 3903
+    },
+    {
+      "epoch": 9.54523227383863,
+      "grad_norm": 0.3232067974029118,
+      "learning_rate": 1.4615683326345755e-06,
+      "loss": 0.3979,
+      "step": 3904
+    },
+    {
+      "epoch": 9.547677261613693,
+      "grad_norm": 0.2781067773006322,
+      "learning_rate": 1.460403939925824e-06,
+      "loss": 0.6261,
+      "step": 3905
+    },
+    {
+      "epoch": 9.550122249388753,
+      "grad_norm": 0.2536830901072346,
+      "learning_rate": 1.4592398198229567e-06,
+      "loss": 1.1174,
+      "step": 3906
+    },
+    {
+      "epoch": 9.552567237163814,
+      "grad_norm": 0.2828240899196044,
+      "learning_rate": 1.4580759726312332e-06,
+      "loss": 1.2102,
+      "step": 3907
+    },
+    {
+      "epoch": 9.555012224938876,
+      "grad_norm": 0.2836336620117512,
+      "learning_rate": 1.4569123986558403e-06,
+      "loss": 0.5564,
+      "step": 3908
+    },
+    {
+      "epoch": 9.557457212713937,
+      "grad_norm": 0.2186569259068507,
+      "learning_rate": 1.455749098201895e-06,
+      "loss": 0.432,
+      "step": 3909
+    },
+    {
+      "epoch": 9.559902200488997,
+      "grad_norm": 0.27671182214937934,
+      "learning_rate": 1.454586071574439e-06,
+      "loss": 0.4719,
+      "step": 3910
+    },
+    {
+      "epoch": 9.56234718826406,
+      "grad_norm": 0.29097522782909824,
+      "learning_rate": 1.453423319078448e-06,
+      "loss": 0.959,
+      "step": 3911
+    },
+    {
+      "epoch": 9.56479217603912,
+      "grad_norm": 0.2516092262333198,
+      "learning_rate": 1.4522608410188187e-06,
+      "loss": 1.0286,
+      "step": 3912
+    },
+    {
+      "epoch": 9.56723716381418,
+      "grad_norm": 0.3167854996866788,
+      "learning_rate": 1.4510986377003827e-06,
+      "loss": 0.4634,
+      "step": 3913
+    },
+    {
+      "epoch": 9.569682151589243,
+      "grad_norm": 0.2956333783163217,
+      "learning_rate": 1.4499367094278933e-06,
+      "loss": 1.2497,
+      "step": 3914
+    },
+    {
+      "epoch": 9.572127139364303,
+      "grad_norm": 0.12452372093432812,
+      "learning_rate": 1.4487750565060363e-06,
+      "loss": 0.3454,
+      "step": 3915
+    },
+    {
+      "epoch": 9.574572127139364,
+      "grad_norm": 0.25758294161491413,
+      "learning_rate": 1.4476136792394243e-06,
+      "loss": 0.9257,
+      "step": 3916
+    },
+    {
+      "epoch": 9.577017114914426,
+      "grad_norm": 0.34424993896770056,
+      "learning_rate": 1.4464525779325952e-06,
+      "loss": 0.464,
+      "step": 3917
+    },
+    {
+      "epoch": 9.579462102689487,
+      "grad_norm": 0.19024653218826845,
+      "learning_rate": 1.4452917528900174e-06,
+      "loss": 0.3934,
+      "step": 3918
+    },
+    {
+      "epoch": 9.581907090464547,
+      "grad_norm": 0.3765922551182458,
+      "learning_rate": 1.4441312044160865e-06,
+      "loss": 1.132,
+      "step": 3919
+    },
+    {
+      "epoch": 9.58435207823961,
+      "grad_norm": 0.26661955382907154,
+      "learning_rate": 1.442970932815123e-06,
+      "loss": 0.8372,
+      "step": 3920
+    },
+    {
+      "epoch": 9.58679706601467,
+      "grad_norm": 0.2263815769966482,
+      "learning_rate": 1.4418109383913781e-06,
+      "loss": 0.7024,
+      "step": 3921
+    },
+    {
+      "epoch": 9.58924205378973,
+      "grad_norm": 0.2409953320817237,
+      "learning_rate": 1.4406512214490287e-06,
+      "loss": 0.7336,
+      "step": 3922
+    },
+    {
+      "epoch": 9.591687041564793,
+      "grad_norm": 0.1855472481546824,
+      "learning_rate": 1.4394917822921796e-06,
+      "loss": 0.5094,
+      "step": 3923
+    },
+    {
+      "epoch": 9.594132029339853,
+      "grad_norm": 0.2428079246696036,
+      "learning_rate": 1.4383326212248607e-06,
+      "loss": 0.4238,
+      "step": 3924
+    },
+    {
+      "epoch": 9.596577017114914,
+      "grad_norm": 0.26581152342214764,
+      "learning_rate": 1.4371737385510316e-06,
+      "loss": 0.4901,
+      "step": 3925
+    },
+    {
+      "epoch": 9.599022004889976,
+      "grad_norm": 0.17244434653124355,
+      "learning_rate": 1.4360151345745788e-06,
+      "loss": 0.6601,
+      "step": 3926
+    },
+    {
+      "epoch": 9.601466992665037,
+      "grad_norm": 0.2793973728887399,
+      "learning_rate": 1.4348568095993128e-06,
+      "loss": 0.5162,
+      "step": 3927
+    },
+    {
+      "epoch": 9.603911980440097,
+      "grad_norm": 0.27592847910560553,
+      "learning_rate": 1.433698763928974e-06,
+      "loss": 0.4485,
+      "step": 3928
+    },
+    {
+      "epoch": 9.60635696821516,
+      "grad_norm": 0.3052432638800618,
+      "learning_rate": 1.4325409978672292e-06,
+      "loss": 0.5086,
+      "step": 3929
+    },
+    {
+      "epoch": 9.60880195599022,
+      "grad_norm": 0.2136516872521334,
+      "learning_rate": 1.4313835117176694e-06,
+      "loss": 0.8255,
+      "step": 3930
+    },
+    {
+      "epoch": 9.61124694376528,
+      "grad_norm": 0.2528188484313757,
+      "learning_rate": 1.430226305783815e-06,
+      "loss": 0.5319,
+      "step": 3931
+    },
+    {
+      "epoch": 9.613691931540343,
+      "grad_norm": 0.16787268151409984,
+      "learning_rate": 1.4290693803691119e-06,
+      "loss": 0.3191,
+      "step": 3932
+    },
+    {
+      "epoch": 9.616136919315403,
+      "grad_norm": 0.2627155902395158,
+      "learning_rate": 1.427912735776933e-06,
+      "loss": 0.467,
+      "step": 3933
+    },
+    {
+      "epoch": 9.618581907090464,
+      "grad_norm": 0.2182565257468506,
+      "learning_rate": 1.4267563723105754e-06,
+      "loss": 0.4459,
+      "step": 3934
+    },
+    {
+      "epoch": 9.621026894865526,
+      "grad_norm": 0.24942486480444884,
+      "learning_rate": 1.4256002902732653e-06,
+      "loss": 0.5695,
+      "step": 3935
+    },
+    {
+      "epoch": 9.623471882640587,
+      "grad_norm": 0.34547539878290207,
+      "learning_rate": 1.4244444899681542e-06,
+      "loss": 0.6361,
+      "step": 3936
+    },
+    {
+      "epoch": 9.625916870415647,
+      "grad_norm": 0.2265519687684426,
+      "learning_rate": 1.4232889716983178e-06,
+      "loss": 0.4542,
+      "step": 3937
+    },
+    {
+      "epoch": 9.62836185819071,
+      "grad_norm": 0.21419583897858963,
+      "learning_rate": 1.4221337357667609e-06,
+      "loss": 0.4851,
+      "step": 3938
+    },
+    {
+      "epoch": 9.63080684596577,
+      "grad_norm": 0.20182049869359583,
+      "learning_rate": 1.4209787824764126e-06,
+      "loss": 0.6504,
+      "step": 3939
+    },
+    {
+      "epoch": 9.63325183374083,
+      "grad_norm": 0.26157328597494645,
+      "learning_rate": 1.4198241121301273e-06,
+      "loss": 0.5001,
+      "step": 3940
+    },
+    {
+      "epoch": 9.635696821515893,
+      "grad_norm": 0.31257514794025604,
+      "learning_rate": 1.4186697250306874e-06,
+      "loss": 0.6232,
+      "step": 3941
+    },
+    {
+      "epoch": 9.638141809290953,
+      "grad_norm": 0.17477144754131954,
+      "learning_rate": 1.4175156214807967e-06,
+      "loss": 0.3635,
+      "step": 3942
+    },
+    {
+      "epoch": 9.640586797066014,
+      "grad_norm": 0.26209661829220765,
+      "learning_rate": 1.4163618017830916e-06,
+      "loss": 1.2046,
+      "step": 3943
+    },
+    {
+      "epoch": 9.643031784841076,
+      "grad_norm": 0.22135618638862203,
+      "learning_rate": 1.415208266240127e-06,
+      "loss": 0.5394,
+      "step": 3944
+    },
+    {
+      "epoch": 9.645476772616137,
+      "grad_norm": 0.18221718699621778,
+      "learning_rate": 1.4140550151543876e-06,
+      "loss": 0.6964,
+      "step": 3945
+    },
+    {
+      "epoch": 9.647921760391197,
+      "grad_norm": 0.30832798801944217,
+      "learning_rate": 1.4129020488282825e-06,
+      "loss": 0.3581,
+      "step": 3946
+    },
+    {
+      "epoch": 9.65036674816626,
+      "grad_norm": 0.21838010494159443,
+      "learning_rate": 1.4117493675641447e-06,
+      "loss": 0.4118,
+      "step": 3947
+    },
+    {
+      "epoch": 9.65281173594132,
+      "grad_norm": 0.17583045315217985,
+      "learning_rate": 1.4105969716642344e-06,
+      "loss": 0.2516,
+      "step": 3948
+    },
+    {
+      "epoch": 9.65525672371638,
+      "grad_norm": 0.2810287788402748,
+      "learning_rate": 1.4094448614307366e-06,
+      "loss": 0.6859,
+      "step": 3949
+    },
+    {
+      "epoch": 9.657701711491443,
+      "grad_norm": 0.32960458590376607,
+      "learning_rate": 1.4082930371657597e-06,
+      "loss": 0.6625,
+      "step": 3950
+    },
+    {
+      "epoch": 9.660146699266503,
+      "grad_norm": 0.3286936825929337,
+      "learning_rate": 1.40714149917134e-06,
+      "loss": 0.5802,
+      "step": 3951
+    },
+    {
+      "epoch": 9.662591687041564,
+      "grad_norm": 0.18696778489621488,
+      "learning_rate": 1.4059902477494353e-06,
+      "loss": 0.3198,
+      "step": 3952
+    },
+    {
+      "epoch": 9.665036674816626,
+      "grad_norm": 0.27252071291482494,
+      "learning_rate": 1.4048392832019311e-06,
+      "loss": 0.9785,
+      "step": 3953
+    },
+    {
+      "epoch": 9.667481662591687,
+      "grad_norm": 0.2170101571231686,
+      "learning_rate": 1.4036886058306366e-06,
+      "loss": 0.5676,
+      "step": 3954
+    },
+    {
+      "epoch": 9.669926650366747,
+      "grad_norm": 0.24876790622604092,
+      "learning_rate": 1.4025382159372857e-06,
+      "loss": 0.6646,
+      "step": 3955
+    },
+    {
+      "epoch": 9.67237163814181,
+      "grad_norm": 0.30067965876125013,
+      "learning_rate": 1.401388113823538e-06,
+      "loss": 0.3403,
+      "step": 3956
+    },
+    {
+      "epoch": 9.67481662591687,
+      "grad_norm": 0.27085007712143144,
+      "learning_rate": 1.4002382997909745e-06,
+      "loss": 0.378,
+      "step": 3957
+    },
+    {
+      "epoch": 9.67726161369193,
+      "grad_norm": 0.3428239285869144,
+      "learning_rate": 1.3990887741411036e-06,
+      "loss": 0.7776,
+      "step": 3958
+    },
+    {
+      "epoch": 9.679706601466993,
+      "grad_norm": 0.34318448564404536,
+      "learning_rate": 1.397939537175359e-06,
+      "loss": 0.4245,
+      "step": 3959
+    },
+    {
+      "epoch": 9.682151589242054,
+      "grad_norm": 0.18813678324033897,
+      "learning_rate": 1.3967905891950939e-06,
+      "loss": 0.4382,
+      "step": 3960
+    },
+    {
+      "epoch": 9.684596577017114,
+      "grad_norm": 0.2481956458728824,
+      "learning_rate": 1.3956419305015916e-06,
+      "loss": 0.541,
+      "step": 3961
+    },
+    {
+      "epoch": 9.687041564792176,
+      "grad_norm": 0.2759394854994073,
+      "learning_rate": 1.3944935613960544e-06,
+      "loss": 0.6456,
+      "step": 3962
+    },
+    {
+      "epoch": 9.689486552567237,
+      "grad_norm": 0.23648839449579018,
+      "learning_rate": 1.393345482179612e-06,
+      "loss": 0.8678,
+      "step": 3963
+    },
+    {
+      "epoch": 9.6919315403423,
+      "grad_norm": 0.2775187916056447,
+      "learning_rate": 1.3921976931533168e-06,
+      "loss": 0.4647,
+      "step": 3964
+    },
+    {
+      "epoch": 9.69437652811736,
+      "grad_norm": 0.212561005425861,
+      "learning_rate": 1.3910501946181457e-06,
+      "loss": 0.6494,
+      "step": 3965
+    },
+    {
+      "epoch": 9.69682151589242,
+      "grad_norm": 0.2393154283233034,
+      "learning_rate": 1.3899029868749997e-06,
+      "loss": 0.5593,
+      "step": 3966
+    },
+    {
+      "epoch": 9.69926650366748,
+      "grad_norm": 0.21942466467555843,
+      "learning_rate": 1.3887560702247016e-06,
+      "loss": 0.6501,
+      "step": 3967
+    },
+    {
+      "epoch": 9.701711491442543,
+      "grad_norm": 0.1543197263424504,
+      "learning_rate": 1.3876094449679994e-06,
+      "loss": 0.4184,
+      "step": 3968
+    },
+    {
+      "epoch": 9.704156479217604,
+      "grad_norm": 0.23833409175962472,
+      "learning_rate": 1.386463111405566e-06,
+      "loss": 0.4069,
+      "step": 3969
+    },
+    {
+      "epoch": 9.706601466992666,
+      "grad_norm": 0.2700969070574167,
+      "learning_rate": 1.3853170698379941e-06,
+      "loss": 0.4828,
+      "step": 3970
+    },
+    {
+      "epoch": 9.709046454767726,
+      "grad_norm": 0.18209115182032812,
+      "learning_rate": 1.3841713205658044e-06,
+      "loss": 0.4128,
+      "step": 3971
+    },
+    {
+      "epoch": 9.711491442542787,
+      "grad_norm": 0.2657672354640742,
+      "learning_rate": 1.383025863889436e-06,
+      "loss": 0.4539,
+      "step": 3972
+    },
+    {
+      "epoch": 9.713936430317847,
+      "grad_norm": 0.18597257338833875,
+      "learning_rate": 1.3818807001092555e-06,
+      "loss": 0.4192,
+      "step": 3973
+    },
+    {
+      "epoch": 9.71638141809291,
+      "grad_norm": 0.2849990895018088,
+      "learning_rate": 1.3807358295255505e-06,
+      "loss": 0.5531,
+      "step": 3974
+    },
+    {
+      "epoch": 9.71882640586797,
+      "grad_norm": 0.2018641079331819,
+      "learning_rate": 1.3795912524385324e-06,
+      "loss": 0.7408,
+      "step": 3975
+    },
+    {
+      "epoch": 9.721271393643033,
+      "grad_norm": 0.18256542688006575,
+      "learning_rate": 1.3784469691483365e-06,
+      "loss": 0.5101,
+      "step": 3976
+    },
+    {
+      "epoch": 9.723716381418093,
+      "grad_norm": 0.4149618047659524,
+      "learning_rate": 1.3773029799550187e-06,
+      "loss": 0.3591,
+      "step": 3977
+    },
+    {
+      "epoch": 9.726161369193154,
+      "grad_norm": 0.35049615587324584,
+      "learning_rate": 1.376159285158559e-06,
+      "loss": 0.2927,
+      "step": 3978
+    },
+    {
+      "epoch": 9.728606356968216,
+      "grad_norm": 0.3019717508892536,
+      "learning_rate": 1.3750158850588621e-06,
+      "loss": 0.6626,
+      "step": 3979
+    },
+    {
+      "epoch": 9.731051344743276,
+      "grad_norm": 0.3290939071048803,
+      "learning_rate": 1.3738727799557517e-06,
+      "loss": 0.665,
+      "step": 3980
+    },
+    {
+      "epoch": 9.733496332518337,
+      "grad_norm": 0.2797276764815221,
+      "learning_rate": 1.3727299701489781e-06,
+      "loss": 0.971,
+      "step": 3981
+    },
+    {
+      "epoch": 9.7359413202934,
+      "grad_norm": 0.19758045797299298,
+      "learning_rate": 1.3715874559382097e-06,
+      "loss": 0.4284,
+      "step": 3982
+    },
+    {
+      "epoch": 9.73838630806846,
+      "grad_norm": 0.3391785181832523,
+      "learning_rate": 1.3704452376230418e-06,
+      "loss": 0.4663,
+      "step": 3983
+    },
+    {
+      "epoch": 9.74083129584352,
+      "grad_norm": 0.28474792783621183,
+      "learning_rate": 1.3693033155029902e-06,
+      "loss": 0.3899,
+      "step": 3984
+    },
+    {
+      "epoch": 9.743276283618583,
+      "grad_norm": 0.2151835339931385,
+      "learning_rate": 1.3681616898774907e-06,
+      "loss": 0.4189,
+      "step": 3985
+    },
+    {
+      "epoch": 9.745721271393643,
+      "grad_norm": 0.20035558222650202,
+      "learning_rate": 1.3670203610459073e-06,
+      "loss": 0.4837,
+      "step": 3986
+    },
+    {
+      "epoch": 9.748166259168704,
+      "grad_norm": 0.3322480947920425,
+      "learning_rate": 1.3658793293075198e-06,
+      "loss": 1.0254,
+      "step": 3987
+    },
+    {
+      "epoch": 9.750611246943766,
+      "grad_norm": 0.31196751197145167,
+      "learning_rate": 1.3647385949615339e-06,
+      "loss": 0.4131,
+      "step": 3988
+    },
+    {
+      "epoch": 9.753056234718827,
+      "grad_norm": 0.34403764525947955,
+      "learning_rate": 1.3635981583070768e-06,
+      "loss": 0.533,
+      "step": 3989
+    },
+    {
+      "epoch": 9.755501222493887,
+      "grad_norm": 0.24855990948278087,
+      "learning_rate": 1.3624580196431953e-06,
+      "loss": 0.6523,
+      "step": 3990
+    },
+    {
+      "epoch": 9.75794621026895,
+      "grad_norm": 0.2448847344317707,
+      "learning_rate": 1.3613181792688624e-06,
+      "loss": 0.8425,
+      "step": 3991
+    },
+    {
+      "epoch": 9.76039119804401,
+      "grad_norm": 0.3983280525019243,
+      "learning_rate": 1.3601786374829672e-06,
+      "loss": 0.7042,
+      "step": 3992
+    },
+    {
+      "epoch": 9.76283618581907,
+      "grad_norm": 0.22289976454558866,
+      "learning_rate": 1.3590393945843259e-06,
+      "loss": 0.7046,
+      "step": 3993
+    },
+    {
+      "epoch": 9.765281173594133,
+      "grad_norm": 0.2806807111791484,
+      "learning_rate": 1.3579004508716743e-06,
+      "loss": 1.2044,
+      "step": 3994
+    },
+    {
+      "epoch": 9.767726161369193,
+      "grad_norm": 0.356165075161078,
+      "learning_rate": 1.3567618066436678e-06,
+      "loss": 0.3877,
+      "step": 3995
+    },
+    {
+      "epoch": 9.770171149144254,
+      "grad_norm": 0.2732755945353237,
+      "learning_rate": 1.3556234621988856e-06,
+      "loss": 0.8471,
+      "step": 3996
+    },
+    {
+      "epoch": 9.772616136919316,
+      "grad_norm": 0.3104435388910349,
+      "learning_rate": 1.354485417835828e-06,
+      "loss": 0.5921,
+      "step": 3997
+    },
+    {
+      "epoch": 9.775061124694377,
+      "grad_norm": 0.33305744081239336,
+      "learning_rate": 1.3533476738529167e-06,
+      "loss": 0.4819,
+      "step": 3998
+    },
+    {
+      "epoch": 9.777506112469437,
+      "grad_norm": 0.2609144342432876,
+      "learning_rate": 1.352210230548494e-06,
+      "loss": 0.7856,
+      "step": 3999
+    },
+    {
+      "epoch": 9.7799511002445,
+      "grad_norm": 0.2346505471202344,
+      "learning_rate": 1.3510730882208228e-06,
+      "loss": 0.3456,
+      "step": 4000
+    },
+    {
+      "epoch": 9.78239608801956,
+      "grad_norm": 0.3187090038197892,
+      "learning_rate": 1.3499362471680893e-06,
+      "loss": 0.4256,
+      "step": 4001
+    },
+    {
+      "epoch": 9.78484107579462,
+      "grad_norm": 0.4188158182956479,
+      "learning_rate": 1.3487997076883974e-06,
+      "loss": 0.4658,
+      "step": 4002
+    },
+    {
+      "epoch": 9.787286063569683,
+      "grad_norm": 0.30772753842958833,
+      "learning_rate": 1.3476634700797747e-06,
+      "loss": 0.9264,
+      "step": 4003
+    },
+    {
+      "epoch": 9.789731051344743,
+      "grad_norm": 0.29770352658904536,
+      "learning_rate": 1.3465275346401696e-06,
+      "loss": 0.434,
+      "step": 4004
+    },
+    {
+      "epoch": 9.792176039119804,
+      "grad_norm": 0.2148016282006695,
+      "learning_rate": 1.3453919016674486e-06,
+      "loss": 0.4837,
+      "step": 4005
+    },
+    {
+      "epoch": 9.794621026894866,
+      "grad_norm": 0.2511468203283835,
+      "learning_rate": 1.3442565714594013e-06,
+      "loss": 0.8167,
+      "step": 4006
+    },
+    {
+      "epoch": 9.797066014669927,
+      "grad_norm": 0.18083728849801017,
+      "learning_rate": 1.343121544313738e-06,
+      "loss": 0.4708,
+      "step": 4007
+    },
+    {
+      "epoch": 9.799511002444987,
+      "grad_norm": 0.4324889402612805,
+      "learning_rate": 1.3419868205280878e-06,
+      "loss": 0.4683,
+      "step": 4008
+    },
+    {
+      "epoch": 9.80195599022005,
+      "grad_norm": 0.21189145286353303,
+      "learning_rate": 1.340852400400003e-06,
+      "loss": 0.4278,
+      "step": 4009
+    },
+    {
+      "epoch": 9.80440097799511,
+      "grad_norm": 0.24058925943763806,
+      "learning_rate": 1.339718284226952e-06,
+      "loss": 0.3901,
+      "step": 4010
+    },
+    {
+      "epoch": 9.80684596577017,
+      "grad_norm": 0.34018629658408606,
+      "learning_rate": 1.3385844723063284e-06,
+      "loss": 0.3652,
+      "step": 4011
+    },
+    {
+      "epoch": 9.809290953545233,
+      "grad_norm": 0.20768885335650156,
+      "learning_rate": 1.337450964935442e-06,
+      "loss": 0.4695,
+      "step": 4012
+    },
+    {
+      "epoch": 9.811735941320293,
+      "grad_norm": 0.27313987056560646,
+      "learning_rate": 1.3363177624115245e-06,
+      "loss": 0.4484,
+      "step": 4013
+    },
+    {
+      "epoch": 9.814180929095354,
+      "grad_norm": 0.22675647169767135,
+      "learning_rate": 1.335184865031729e-06,
+      "loss": 0.5852,
+      "step": 4014
+    },
+    {
+      "epoch": 9.816625916870416,
+      "grad_norm": 0.23221625648773955,
+      "learning_rate": 1.3340522730931254e-06,
+      "loss": 0.5577,
+      "step": 4015
+    },
+    {
+      "epoch": 9.819070904645477,
+      "grad_norm": 0.19306287736802782,
+      "learning_rate": 1.3329199868927057e-06,
+      "loss": 0.5059,
+      "step": 4016
+    },
+    {
+      "epoch": 9.821515892420537,
+      "grad_norm": 0.24331144070605967,
+      "learning_rate": 1.3317880067273813e-06,
+      "loss": 0.4951,
+      "step": 4017
+    },
+    {
+      "epoch": 9.8239608801956,
+      "grad_norm": 0.3584028422872811,
+      "learning_rate": 1.330656332893984e-06,
+      "loss": 0.8374,
+      "step": 4018
+    },
+    {
+      "epoch": 9.82640586797066,
+      "grad_norm": 0.24255779782869188,
+      "learning_rate": 1.3295249656892648e-06,
+      "loss": 1.1743,
+      "step": 4019
+    },
+    {
+      "epoch": 9.82885085574572,
+      "grad_norm": 0.28504017675566035,
+      "learning_rate": 1.328393905409892e-06,
+      "loss": 0.7278,
+      "step": 4020
+    },
+    {
+      "epoch": 9.831295843520783,
+      "grad_norm": 0.23545291624481923,
+      "learning_rate": 1.3272631523524588e-06,
+      "loss": 0.5571,
+      "step": 4021
+    },
+    {
+      "epoch": 9.833740831295843,
+      "grad_norm": 0.27442221546789286,
+      "learning_rate": 1.3261327068134707e-06,
+      "loss": 0.4213,
+      "step": 4022
+    },
+    {
+      "epoch": 9.836185819070904,
+      "grad_norm": 0.26127903365357846,
+      "learning_rate": 1.3250025690893587e-06,
+      "loss": 1.0642,
+      "step": 4023
+    },
+    {
+      "epoch": 9.838630806845966,
+      "grad_norm": 0.27470608252322676,
+      "learning_rate": 1.3238727394764713e-06,
+      "loss": 0.6067,
+      "step": 4024
+    },
+    {
+      "epoch": 9.841075794621027,
+      "grad_norm": 0.2573437619683948,
+      "learning_rate": 1.322743218271074e-06,
+      "loss": 0.4248,
+      "step": 4025
+    },
+    {
+      "epoch": 9.843520782396087,
+      "grad_norm": 0.21265396359805472,
+      "learning_rate": 1.3216140057693535e-06,
+      "loss": 0.5927,
+      "step": 4026
+    },
+    {
+      "epoch": 9.84596577017115,
+      "grad_norm": 0.27504638164778344,
+      "learning_rate": 1.320485102267417e-06,
+      "loss": 0.4278,
+      "step": 4027
+    },
+    {
+      "epoch": 9.84841075794621,
+      "grad_norm": 0.24169818341469193,
+      "learning_rate": 1.319356508061285e-06,
+      "loss": 0.5972,
+      "step": 4028
+    },
+    {
+      "epoch": 9.85085574572127,
+      "grad_norm": 0.20905556195545266,
+      "learning_rate": 1.318228223446905e-06,
+      "loss": 0.4348,
+      "step": 4029
+    },
+    {
+      "epoch": 9.853300733496333,
+      "grad_norm": 0.18281378945650573,
+      "learning_rate": 1.3171002487201367e-06,
+      "loss": 0.4958,
+      "step": 4030
+    },
+    {
+      "epoch": 9.855745721271393,
+      "grad_norm": 0.3652458517947851,
+      "learning_rate": 1.315972584176762e-06,
+      "loss": 1.224,
+      "step": 4031
+    },
+    {
+      "epoch": 9.858190709046454,
+      "grad_norm": 0.2405041016435086,
+      "learning_rate": 1.314845230112479e-06,
+      "loss": 0.4903,
+      "step": 4032
+    },
+    {
+      "epoch": 9.860635696821516,
+      "grad_norm": 0.30627644931498266,
+      "learning_rate": 1.3137181868229062e-06,
+      "loss": 0.5234,
+      "step": 4033
+    },
+    {
+      "epoch": 9.863080684596577,
+      "grad_norm": 0.3803625027981747,
+      "learning_rate": 1.3125914546035818e-06,
+      "loss": 0.4284,
+      "step": 4034
+    },
+    {
+      "epoch": 9.865525672371637,
+      "grad_norm": 0.24174079896188697,
+      "learning_rate": 1.3114650337499578e-06,
+      "loss": 0.6802,
+      "step": 4035
+    },
+    {
+      "epoch": 9.8679706601467,
+      "grad_norm": 0.24785162980587833,
+      "learning_rate": 1.3103389245574099e-06,
+      "loss": 0.4202,
+      "step": 4036
+    },
+    {
+      "epoch": 9.87041564792176,
+      "grad_norm": 0.2300014848512258,
+      "learning_rate": 1.3092131273212294e-06,
+      "loss": 0.4173,
+      "step": 4037
+    },
+    {
+      "epoch": 9.87286063569682,
+      "grad_norm": 0.31246322713184427,
+      "learning_rate": 1.3080876423366247e-06,
+      "loss": 0.3977,
+      "step": 4038
+    },
+    {
+      "epoch": 9.875305623471883,
+      "grad_norm": 0.24769167867706052,
+      "learning_rate": 1.3069624698987249e-06,
+      "loss": 0.8479,
+      "step": 4039
+    },
+    {
+      "epoch": 9.877750611246944,
+      "grad_norm": 0.22506985121027237,
+      "learning_rate": 1.3058376103025757e-06,
+      "loss": 0.699,
+      "step": 4040
+    },
+    {
+      "epoch": 9.880195599022004,
+      "grad_norm": 0.30200295002971855,
+      "learning_rate": 1.3047130638431421e-06,
+      "loss": 0.512,
+      "step": 4041
+    },
+    {
+      "epoch": 9.882640586797066,
+      "grad_norm": 0.2136967120677066,
+      "learning_rate": 1.3035888308153038e-06,
+      "loss": 0.4234,
+      "step": 4042
+    },
+    {
+      "epoch": 9.885085574572127,
+      "grad_norm": 0.22256333342879878,
+      "learning_rate": 1.3024649115138615e-06,
+      "loss": 0.4512,
+      "step": 4043
+    },
+    {
+      "epoch": 9.887530562347187,
+      "grad_norm": 0.36412595704100104,
+      "learning_rate": 1.301341306233534e-06,
+      "loss": 0.7507,
+      "step": 4044
+    },
+    {
+      "epoch": 9.88997555012225,
+      "grad_norm": 0.283661783817512,
+      "learning_rate": 1.3002180152689536e-06,
+      "loss": 0.3735,
+      "step": 4045
+    },
+    {
+      "epoch": 9.89242053789731,
+      "grad_norm": 0.16748915632260827,
+      "learning_rate": 1.2990950389146742e-06,
+      "loss": 0.4292,
+      "step": 4046
+    },
+    {
+      "epoch": 9.894865525672373,
+      "grad_norm": 0.2947914039367817,
+      "learning_rate": 1.297972377465167e-06,
+      "loss": 0.472,
+      "step": 4047
+    },
+    {
+      "epoch": 9.897310513447433,
+      "grad_norm": 0.21103275150947443,
+      "learning_rate": 1.2968500312148174e-06,
+      "loss": 0.8662,
+      "step": 4048
+    },
+    {
+      "epoch": 9.899755501222494,
+      "grad_norm": 0.31252306666777663,
+      "learning_rate": 1.2957280004579317e-06,
+      "loss": 0.7353,
+      "step": 4049
+    },
+    {
+      "epoch": 9.902200488997554,
+      "grad_norm": 0.2923266986094965,
+      "learning_rate": 1.2946062854887315e-06,
+      "loss": 0.666,
+      "step": 4050
+    },
+    {
+      "epoch": 9.904645476772616,
+      "grad_norm": 0.27575854504308867,
+      "learning_rate": 1.2934848866013574e-06,
+      "loss": 0.5038,
+      "step": 4051
+    },
+    {
+      "epoch": 9.907090464547677,
+      "grad_norm": 0.26167143181598956,
+      "learning_rate": 1.292363804089864e-06,
+      "loss": 0.5366,
+      "step": 4052
+    },
+    {
+      "epoch": 9.90953545232274,
+      "grad_norm": 0.21611890609422782,
+      "learning_rate": 1.2912430382482252e-06,
+      "loss": 0.6822,
+      "step": 4053
+    },
+    {
+      "epoch": 9.9119804400978,
+      "grad_norm": 0.22680242164140751,
+      "learning_rate": 1.2901225893703328e-06,
+      "loss": 0.5049,
+      "step": 4054
+    },
+    {
+      "epoch": 9.91442542787286,
+      "grad_norm": 0.25993264570727953,
+      "learning_rate": 1.2890024577499926e-06,
+      "loss": 0.617,
+      "step": 4055
+    },
+    {
+      "epoch": 9.91687041564792,
+      "grad_norm": 0.16596350901869003,
+      "learning_rate": 1.2878826436809294e-06,
+      "loss": 0.5196,
+      "step": 4056
+    },
+    {
+      "epoch": 9.919315403422983,
+      "grad_norm": 0.3077170796631459,
+      "learning_rate": 1.286763147456785e-06,
+      "loss": 0.6832,
+      "step": 4057
+    },
+    {
+      "epoch": 9.921760391198044,
+      "grad_norm": 0.289863672865671,
+      "learning_rate": 1.2856439693711147e-06,
+      "loss": 1.0394,
+      "step": 4058
+    },
+    {
+      "epoch": 9.924205378973106,
+      "grad_norm": 0.22318121335478194,
+      "learning_rate": 1.284525109717395e-06,
+      "loss": 0.3355,
+      "step": 4059
+    },
+    {
+      "epoch": 9.926650366748166,
+      "grad_norm": 0.27983770280342873,
+      "learning_rate": 1.2834065687890137e-06,
+      "loss": 0.6487,
+      "step": 4060
+    },
+    {
+      "epoch": 9.929095354523227,
+      "grad_norm": 0.2516985367876321,
+      "learning_rate": 1.2822883468792813e-06,
+      "loss": 0.716,
+      "step": 4061
+    },
+    {
+      "epoch": 9.93154034229829,
+      "grad_norm": 0.16509823194721587,
+      "learning_rate": 1.2811704442814185e-06,
+      "loss": 0.3455,
+      "step": 4062
+    },
+    {
+      "epoch": 9.93398533007335,
+      "grad_norm": 0.19651460729146775,
+      "learning_rate": 1.280052861288566e-06,
+      "loss": 0.5983,
+      "step": 4063
+    },
+    {
+      "epoch": 9.93643031784841,
+      "grad_norm": 0.19802406598885053,
+      "learning_rate": 1.27893559819378e-06,
+      "loss": 0.5085,
+      "step": 4064
+    },
+    {
+      "epoch": 9.938875305623473,
+      "grad_norm": 0.2714731735601288,
+      "learning_rate": 1.2778186552900318e-06,
+      "loss": 0.381,
+      "step": 4065
+    },
+    {
+      "epoch": 9.941320293398533,
+      "grad_norm": 0.2653730714754877,
+      "learning_rate": 1.276702032870209e-06,
+      "loss": 1.4207,
+      "step": 4066
+    },
+    {
+      "epoch": 9.943765281173594,
+      "grad_norm": 0.20978572867131567,
+      "learning_rate": 1.2755857312271174e-06,
+      "loss": 0.4533,
+      "step": 4067
+    },
+    {
+      "epoch": 9.946210268948656,
+      "grad_norm": 0.35499504720385117,
+      "learning_rate": 1.2744697506534745e-06,
+      "loss": 0.4246,
+      "step": 4068
+    },
+    {
+      "epoch": 9.948655256723717,
+      "grad_norm": 0.217983903720869,
+      "learning_rate": 1.2733540914419178e-06,
+      "loss": 0.4871,
+      "step": 4069
+    },
+    {
+      "epoch": 9.951100244498777,
+      "grad_norm": 0.3619942625932222,
+      "learning_rate": 1.2722387538849973e-06,
+      "loss": 0.5489,
+      "step": 4070
+    },
+    {
+      "epoch": 9.95354523227384,
+      "grad_norm": 0.26996248526651395,
+      "learning_rate": 1.2711237382751806e-06,
+      "loss": 0.8182,
+      "step": 4071
+    },
+    {
+      "epoch": 9.9559902200489,
+      "grad_norm": 0.2683802099011985,
+      "learning_rate": 1.2700090449048503e-06,
+      "loss": 0.5322,
+      "step": 4072
+    },
+    {
+      "epoch": 9.95843520782396,
+      "grad_norm": 0.2791425885932258,
+      "learning_rate": 1.2688946740663047e-06,
+      "loss": 0.7948,
+      "step": 4073
+    },
+    {
+      "epoch": 9.960880195599023,
+      "grad_norm": 0.3167766846215492,
+      "learning_rate": 1.2677806260517581e-06,
+      "loss": 0.4159,
+      "step": 4074
+    },
+    {
+      "epoch": 9.963325183374083,
+      "grad_norm": 0.2512942284505814,
+      "learning_rate": 1.2666669011533375e-06,
+      "loss": 0.3536,
+      "step": 4075
+    },
+    {
+      "epoch": 9.965770171149144,
+      "grad_norm": 0.1922633846371159,
+      "learning_rate": 1.2655534996630883e-06,
+      "loss": 0.4131,
+      "step": 4076
+    },
+    {
+      "epoch": 9.968215158924206,
+      "grad_norm": 0.2829681281489595,
+      "learning_rate": 1.2644404218729705e-06,
+      "loss": 0.6787,
+      "step": 4077
+    },
+    {
+      "epoch": 9.970660146699267,
+      "grad_norm": 0.27886341223811234,
+      "learning_rate": 1.2633276680748568e-06,
+      "loss": 0.4792,
+      "step": 4078
+    },
+    {
+      "epoch": 9.973105134474327,
+      "grad_norm": 0.2884935247227573,
+      "learning_rate": 1.2622152385605384e-06,
+      "loss": 0.4922,
+      "step": 4079
+    },
+    {
+      "epoch": 9.97555012224939,
+      "grad_norm": 0.2810686692717784,
+      "learning_rate": 1.2611031336217183e-06,
+      "loss": 0.491,
+      "step": 4080
+    },
+    {
+      "epoch": 9.97799511002445,
+      "grad_norm": 0.17739240444636853,
+      "learning_rate": 1.259991353550017e-06,
+      "loss": 0.4317,
+      "step": 4081
+    },
+    {
+      "epoch": 9.98044009779951,
+      "grad_norm": 0.14834440006009075,
+      "learning_rate": 1.2588798986369678e-06,
+      "loss": 0.4045,
+      "step": 4082
+    },
+    {
+      "epoch": 9.982885085574573,
+      "grad_norm": 0.38700062986101735,
+      "learning_rate": 1.2577687691740201e-06,
+      "loss": 0.4821,
+      "step": 4083
+    },
+    {
+      "epoch": 9.985330073349633,
+      "grad_norm": 0.2838176558247498,
+      "learning_rate": 1.2566579654525384e-06,
+      "loss": 1.0047,
+      "step": 4084
+    },
+    {
+      "epoch": 9.987775061124694,
+      "grad_norm": 0.20977476004828025,
+      "learning_rate": 1.2555474877637991e-06,
+      "loss": 1.0607,
+      "step": 4085
+    },
+    {
+      "epoch": 9.990220048899756,
+      "grad_norm": 0.2427352226732089,
+      "learning_rate": 1.2544373363989958e-06,
+      "loss": 0.3205,
+      "step": 4086
+    },
+    {
+      "epoch": 9.992665036674817,
+      "grad_norm": 0.24564461294695167,
+      "learning_rate": 1.2533275116492367e-06,
+      "loss": 0.6627,
+      "step": 4087
+    },
+    {
+      "epoch": 9.995110024449877,
+      "grad_norm": 0.2921670297897772,
+      "learning_rate": 1.2522180138055407e-06,
+      "loss": 0.4503,
+      "step": 4088
+    },
+    {
+      "epoch": 9.99755501222494,
+      "grad_norm": 0.2963943919570437,
+      "learning_rate": 1.2511088431588462e-06,
+      "loss": 0.601,
+      "step": 4089
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.3269205605410965,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.4649,
+      "step": 4090
+    },
+    {
+      "epoch": 10.00244498777506,
+      "grad_norm": 0.20843576283697876,
+      "learning_rate": 1.2488914846197694e-06,
+      "loss": 0.4783,
+      "step": 4091
+    },
+    {
+      "epoch": 10.004889975550123,
+      "grad_norm": 0.21228257893915267,
+      "learning_rate": 1.2477832973088306e-06,
+      "loss": 0.6509,
+      "step": 4092
+    },
+    {
+      "epoch": 10.007334963325183,
+      "grad_norm": 0.2572391703630435,
+      "learning_rate": 1.2466754383577758e-06,
+      "loss": 0.3411,
+      "step": 4093
+    },
+    {
+      "epoch": 10.009779951100244,
+      "grad_norm": 0.241545368594539,
+      "learning_rate": 1.2455679080571125e-06,
+      "loss": 0.7343,
+      "step": 4094
+    },
+    {
+      "epoch": 10.012224938875306,
+      "grad_norm": 0.35998029760473993,
+      "learning_rate": 1.2444607066972586e-06,
+      "loss": 0.6575,
+      "step": 4095
+    },
+    {
+      "epoch": 10.014669926650367,
+      "grad_norm": 0.2044765807201473,
+      "learning_rate": 1.2433538345685483e-06,
+      "loss": 0.4346,
+      "step": 4096
+    },
+    {
+      "epoch": 10.017114914425427,
+      "grad_norm": 0.2504161142070067,
+      "learning_rate": 1.24224729196123e-06,
+      "loss": 0.4439,
+      "step": 4097
+    },
+    {
+      "epoch": 10.01955990220049,
+      "grad_norm": 0.2903557236014858,
+      "learning_rate": 1.2411410791654625e-06,
+      "loss": 0.4689,
+      "step": 4098
+    },
+    {
+      "epoch": 10.02200488997555,
+      "grad_norm": 0.2591738393184338,
+      "learning_rate": 1.2400351964713217e-06,
+      "loss": 0.8255,
+      "step": 4099
+    },
+    {
+      "epoch": 10.02444987775061,
+      "grad_norm": 0.2329846759559528,
+      "learning_rate": 1.238929644168794e-06,
+      "loss": 0.9687,
+      "step": 4100
+    },
+    {
+      "epoch": 10.026894865525673,
+      "grad_norm": 0.3081185957900225,
+      "learning_rate": 1.2378244225477812e-06,
+      "loss": 1.0416,
+      "step": 4101
+    },
+    {
+      "epoch": 10.029339853300733,
+      "grad_norm": 0.286622540419071,
+      "learning_rate": 1.2367195318980988e-06,
+      "loss": 0.6481,
+      "step": 4102
+    },
+    {
+      "epoch": 10.031784841075794,
+      "grad_norm": 0.31021730314554313,
+      "learning_rate": 1.235614972509471e-06,
+      "loss": 0.4727,
+      "step": 4103
+    },
+    {
+      "epoch": 10.034229828850856,
+      "grad_norm": 0.162757958499169,
+      "learning_rate": 1.2345107446715432e-06,
+      "loss": 0.3342,
+      "step": 4104
+    },
+    {
+      "epoch": 10.036674816625917,
+      "grad_norm": 0.27846512859041017,
+      "learning_rate": 1.233406848673866e-06,
+      "loss": 0.4379,
+      "step": 4105
+    },
+    {
+      "epoch": 10.039119804400977,
+      "grad_norm": 0.204590491923056,
+      "learning_rate": 1.2323032848059068e-06,
+      "loss": 0.4143,
+      "step": 4106
+    },
+    {
+      "epoch": 10.04156479217604,
+      "grad_norm": 0.2873974055529322,
+      "learning_rate": 1.231200053357047e-06,
+      "loss": 0.5169,
+      "step": 4107
+    },
+    {
+      "epoch": 10.0440097799511,
+      "grad_norm": 0.20276498352866473,
+      "learning_rate": 1.2300971546165766e-06,
+      "loss": 0.4137,
+      "step": 4108
+    },
+    {
+      "epoch": 10.04645476772616,
+      "grad_norm": 0.3146174274331853,
+      "learning_rate": 1.228994588873703e-06,
+      "loss": 0.4003,
+      "step": 4109
+    },
+    {
+      "epoch": 10.048899755501223,
+      "grad_norm": 0.23904682396563412,
+      "learning_rate": 1.227892356417542e-06,
+      "loss": 0.5495,
+      "step": 4110
+    },
+    {
+      "epoch": 10.051344743276283,
+      "grad_norm": 0.20316142027532735,
+      "learning_rate": 1.2267904575371259e-06,
+      "loss": 0.5436,
+      "step": 4111
+    },
+    {
+      "epoch": 10.053789731051344,
+      "grad_norm": 0.2232828036353317,
+      "learning_rate": 1.2256888925213982e-06,
+      "loss": 0.8635,
+      "step": 4112
+    },
+    {
+      "epoch": 10.056234718826406,
+      "grad_norm": 0.24173000139408476,
+      "learning_rate": 1.2245876616592126e-06,
+      "loss": 0.5745,
+      "step": 4113
+    },
+    {
+      "epoch": 10.058679706601467,
+      "grad_norm": 0.2001999328932452,
+      "learning_rate": 1.223486765239338e-06,
+      "loss": 0.5304,
+      "step": 4114
+    },
+    {
+      "epoch": 10.061124694376527,
+      "grad_norm": 0.27570629479538195,
+      "learning_rate": 1.222386203550455e-06,
+      "loss": 0.619,
+      "step": 4115
+    },
+    {
+      "epoch": 10.06356968215159,
+      "grad_norm": 0.2941588126495584,
+      "learning_rate": 1.2212859768811553e-06,
+      "loss": 0.7085,
+      "step": 4116
+    },
+    {
+      "epoch": 10.06601466992665,
+      "grad_norm": 0.24627626972521532,
+      "learning_rate": 1.2201860855199453e-06,
+      "loss": 0.695,
+      "step": 4117
+    },
+    {
+      "epoch": 10.06845965770171,
+      "grad_norm": 0.31838314900410647,
+      "learning_rate": 1.219086529755239e-06,
+      "loss": 0.6182,
+      "step": 4118
+    },
+    {
+      "epoch": 10.070904645476773,
+      "grad_norm": 0.31589738946603907,
+      "learning_rate": 1.2179873098753677e-06,
+      "loss": 0.4961,
+      "step": 4119
+    },
+    {
+      "epoch": 10.073349633251834,
+      "grad_norm": 0.22257795628191507,
+      "learning_rate": 1.2168884261685698e-06,
+      "loss": 0.429,
+      "step": 4120
+    },
+    {
+      "epoch": 10.075794621026894,
+      "grad_norm": 0.28921420264830827,
+      "learning_rate": 1.2157898789229989e-06,
+      "loss": 0.7576,
+      "step": 4121
+    },
+    {
+      "epoch": 10.078239608801956,
+      "grad_norm": 0.29053641275724085,
+      "learning_rate": 1.2146916684267196e-06,
+      "loss": 0.6919,
+      "step": 4122
+    },
+    {
+      "epoch": 10.080684596577017,
+      "grad_norm": 0.2737294086045129,
+      "learning_rate": 1.2135937949677066e-06,
+      "loss": 0.3998,
+      "step": 4123
+    },
+    {
+      "epoch": 10.083129584352077,
+      "grad_norm": 0.3496990866546356,
+      "learning_rate": 1.2124962588338483e-06,
+      "loss": 0.8825,
+      "step": 4124
+    },
+    {
+      "epoch": 10.08557457212714,
+      "grad_norm": 0.3072496222576729,
+      "learning_rate": 1.2113990603129432e-06,
+      "loss": 0.7952,
+      "step": 4125
+    },
+    {
+      "epoch": 10.0880195599022,
+      "grad_norm": 0.330229488909154,
+      "learning_rate": 1.2103021996927028e-06,
+      "loss": 0.527,
+      "step": 4126
+    },
+    {
+      "epoch": 10.09046454767726,
+      "grad_norm": 0.40515907641986176,
+      "learning_rate": 1.2092056772607493e-06,
+      "loss": 0.406,
+      "step": 4127
+    },
+    {
+      "epoch": 10.092909535452323,
+      "grad_norm": 0.23914429622564956,
+      "learning_rate": 1.2081094933046142e-06,
+      "loss": 0.7231,
+      "step": 4128
+    },
+    {
+      "epoch": 10.095354523227384,
+      "grad_norm": 0.2559082460774146,
+      "learning_rate": 1.2070136481117442e-06,
+      "loss": 0.594,
+      "step": 4129
+    },
+    {
+      "epoch": 10.097799511002446,
+      "grad_norm": 0.269734036600957,
+      "learning_rate": 1.2059181419694932e-06,
+      "loss": 0.5171,
+      "step": 4130
+    },
+    {
+      "epoch": 10.100244498777506,
+      "grad_norm": 0.32241860877482476,
+      "learning_rate": 1.2048229751651283e-06,
+      "loss": 0.4156,
+      "step": 4131
+    },
+    {
+      "epoch": 10.102689486552567,
+      "grad_norm": 0.22515215823769727,
+      "learning_rate": 1.2037281479858288e-06,
+      "loss": 0.5137,
+      "step": 4132
+    },
+    {
+      "epoch": 10.10513447432763,
+      "grad_norm": 0.30491056815591205,
+      "learning_rate": 1.2026336607186815e-06,
+      "loss": 0.7891,
+      "step": 4133
+    },
+    {
+      "epoch": 10.10757946210269,
+      "grad_norm": 0.28790410039915304,
+      "learning_rate": 1.2015395136506866e-06,
+      "loss": 0.6664,
+      "step": 4134
+    },
+    {
+      "epoch": 10.11002444987775,
+      "grad_norm": 0.1582398271639738,
+      "learning_rate": 1.2004457070687548e-06,
+      "loss": 0.3299,
+      "step": 4135
+    },
+    {
+      "epoch": 10.112469437652813,
+      "grad_norm": 0.19119891549146686,
+      "learning_rate": 1.1993522412597073e-06,
+      "loss": 0.5576,
+      "step": 4136
+    },
+    {
+      "epoch": 10.114914425427873,
+      "grad_norm": 0.30228616505192535,
+      "learning_rate": 1.1982591165102763e-06,
+      "loss": 0.6854,
+      "step": 4137
+    },
+    {
+      "epoch": 10.117359413202934,
+      "grad_norm": 0.27561708240492117,
+      "learning_rate": 1.1971663331071032e-06,
+      "loss": 0.5825,
+      "step": 4138
+    },
+    {
+      "epoch": 10.119804400977996,
+      "grad_norm": 0.388323690395099,
+      "learning_rate": 1.1960738913367417e-06,
+      "loss": 0.4349,
+      "step": 4139
+    },
+    {
+      "epoch": 10.122249388753056,
+      "grad_norm": 0.17641853111428465,
+      "learning_rate": 1.194981791485654e-06,
+      "loss": 0.3887,
+      "step": 4140
+    },
+    {
+      "epoch": 10.124694376528117,
+      "grad_norm": 0.26054947973364223,
+      "learning_rate": 1.193890033840214e-06,
+      "loss": 0.6125,
+      "step": 4141
+    },
+    {
+      "epoch": 10.12713936430318,
+      "grad_norm": 0.21564783351249864,
+      "learning_rate": 1.1927986186867067e-06,
+      "loss": 0.4714,
+      "step": 4142
+    },
+    {
+      "epoch": 10.12958435207824,
+      "grad_norm": 0.3126582058013417,
+      "learning_rate": 1.1917075463113242e-06,
+      "loss": 0.3956,
+      "step": 4143
+    },
+    {
+      "epoch": 10.1320293398533,
+      "grad_norm": 0.2884270653175563,
+      "learning_rate": 1.190616817000172e-06,
+      "loss": 0.4076,
+      "step": 4144
+    },
+    {
+      "epoch": 10.134474327628363,
+      "grad_norm": 0.3081749580720585,
+      "learning_rate": 1.1895264310392645e-06,
+      "loss": 0.3902,
+      "step": 4145
+    },
+    {
+      "epoch": 10.136919315403423,
+      "grad_norm": 0.4556862954832288,
+      "learning_rate": 1.1884363887145237e-06,
+      "loss": 0.752,
+      "step": 4146
+    },
+    {
+      "epoch": 10.139364303178484,
+      "grad_norm": 0.19790824809069463,
+      "learning_rate": 1.187346690311787e-06,
+      "loss": 0.5941,
+      "step": 4147
+    },
+    {
+      "epoch": 10.141809290953546,
+      "grad_norm": 0.24690976280309146,
+      "learning_rate": 1.1862573361167957e-06,
+      "loss": 0.3208,
+      "step": 4148
+    },
+    {
+      "epoch": 10.144254278728607,
+      "grad_norm": 0.3778130988717183,
+      "learning_rate": 1.185168326415205e-06,
+      "loss": 0.3549,
+      "step": 4149
+    },
+    {
+      "epoch": 10.146699266503667,
+      "grad_norm": 0.20341049699409436,
+      "learning_rate": 1.1840796614925764e-06,
+      "loss": 0.4514,
+      "step": 4150
+    },
+    {
+      "epoch": 10.14914425427873,
+      "grad_norm": 0.21723705533523527,
+      "learning_rate": 1.182991341634384e-06,
+      "loss": 0.644,
+      "step": 4151
+    },
+    {
+      "epoch": 10.15158924205379,
+      "grad_norm": 0.24145819203642227,
+      "learning_rate": 1.181903367126011e-06,
+      "loss": 0.674,
+      "step": 4152
+    },
+    {
+      "epoch": 10.15403422982885,
+      "grad_norm": 0.2160016856989047,
+      "learning_rate": 1.180815738252747e-06,
+      "loss": 0.4534,
+      "step": 4153
+    },
+    {
+      "epoch": 10.156479217603913,
+      "grad_norm": 0.3441669993273657,
+      "learning_rate": 1.1797284552997945e-06,
+      "loss": 0.4201,
+      "step": 4154
+    },
+    {
+      "epoch": 10.158924205378973,
+      "grad_norm": 0.2910476570687918,
+      "learning_rate": 1.1786415185522645e-06,
+      "loss": 0.5914,
+      "step": 4155
+    },
+    {
+      "epoch": 10.161369193154034,
+      "grad_norm": 0.20140729786102404,
+      "learning_rate": 1.1775549282951755e-06,
+      "loss": 0.4155,
+      "step": 4156
+    },
+    {
+      "epoch": 10.163814180929096,
+      "grad_norm": 0.29284486893101835,
+      "learning_rate": 1.1764686848134566e-06,
+      "loss": 0.4007,
+      "step": 4157
+    },
+    {
+      "epoch": 10.166259168704157,
+      "grad_norm": 0.27338124082317167,
+      "learning_rate": 1.1753827883919461e-06,
+      "loss": 0.5415,
+      "step": 4158
+    },
+    {
+      "epoch": 10.168704156479217,
+      "grad_norm": 0.33634477557508075,
+      "learning_rate": 1.1742972393153918e-06,
+      "loss": 0.8392,
+      "step": 4159
+    },
+    {
+      "epoch": 10.17114914425428,
+      "grad_norm": 0.35731306434055854,
+      "learning_rate": 1.1732120378684472e-06,
+      "loss": 0.4808,
+      "step": 4160
+    },
+    {
+      "epoch": 10.17359413202934,
+      "grad_norm": 0.21615973895853108,
+      "learning_rate": 1.1721271843356784e-06,
+      "loss": 0.4477,
+      "step": 4161
+    },
+    {
+      "epoch": 10.1760391198044,
+      "grad_norm": 0.22801464095729906,
+      "learning_rate": 1.17104267900156e-06,
+      "loss": 0.4508,
+      "step": 4162
+    },
+    {
+      "epoch": 10.178484107579463,
+      "grad_norm": 0.3105366872180831,
+      "learning_rate": 1.1699585221504717e-06,
+      "loss": 0.6155,
+      "step": 4163
+    },
+    {
+      "epoch": 10.180929095354523,
+      "grad_norm": 0.21894071649710106,
+      "learning_rate": 1.168874714066705e-06,
+      "loss": 0.8423,
+      "step": 4164
+    },
+    {
+      "epoch": 10.183374083129584,
+      "grad_norm": 0.2569470966766589,
+      "learning_rate": 1.167791255034461e-06,
+      "loss": 0.906,
+      "step": 4165
+    },
+    {
+      "epoch": 10.185819070904646,
+      "grad_norm": 0.23821097890448478,
+      "learning_rate": 1.1667081453378448e-06,
+      "loss": 0.5307,
+      "step": 4166
+    },
+    {
+      "epoch": 10.188264058679707,
+      "grad_norm": 0.3022091400946855,
+      "learning_rate": 1.1656253852608736e-06,
+      "loss": 0.7315,
+      "step": 4167
+    },
+    {
+      "epoch": 10.190709046454767,
+      "grad_norm": 0.15477962231889153,
+      "learning_rate": 1.1645429750874724e-06,
+      "loss": 0.3853,
+      "step": 4168
+    },
+    {
+      "epoch": 10.19315403422983,
+      "grad_norm": 0.30326146145813354,
+      "learning_rate": 1.1634609151014743e-06,
+      "loss": 0.8213,
+      "step": 4169
+    },
+    {
+      "epoch": 10.19559902200489,
+      "grad_norm": 0.1763170486185931,
+      "learning_rate": 1.1623792055866185e-06,
+      "loss": 0.3783,
+      "step": 4170
+    },
+    {
+      "epoch": 10.19804400977995,
+      "grad_norm": 0.2700999073428717,
+      "learning_rate": 1.161297846826555e-06,
+      "loss": 0.5596,
+      "step": 4171
+    },
+    {
+      "epoch": 10.200488997555013,
+      "grad_norm": 0.28348043194799327,
+      "learning_rate": 1.1602168391048418e-06,
+      "loss": 1.4361,
+      "step": 4172
+    },
+    {
+      "epoch": 10.202933985330073,
+      "grad_norm": 0.3812604496964369,
+      "learning_rate": 1.159136182704942e-06,
+      "loss": 0.3998,
+      "step": 4173
+    },
+    {
+      "epoch": 10.205378973105134,
+      "grad_norm": 0.3402163874437233,
+      "learning_rate": 1.1580558779102291e-06,
+      "loss": 0.8034,
+      "step": 4174
+    },
+    {
+      "epoch": 10.207823960880196,
+      "grad_norm": 0.21006144137395466,
+      "learning_rate": 1.156975925003985e-06,
+      "loss": 0.9017,
+      "step": 4175
+    },
+    {
+      "epoch": 10.210268948655257,
+      "grad_norm": 0.1926071896728539,
+      "learning_rate": 1.155896324269396e-06,
+      "loss": 0.5015,
+      "step": 4176
+    },
+    {
+      "epoch": 10.212713936430317,
+      "grad_norm": 0.2224797492015203,
+      "learning_rate": 1.1548170759895602e-06,
+      "loss": 0.7502,
+      "step": 4177
+    },
+    {
+      "epoch": 10.21515892420538,
+      "grad_norm": 0.24848998628121807,
+      "learning_rate": 1.153738180447478e-06,
+      "loss": 0.5773,
+      "step": 4178
+    },
+    {
+      "epoch": 10.21760391198044,
+      "grad_norm": 0.21902288233749048,
+      "learning_rate": 1.152659637926065e-06,
+      "loss": 0.5423,
+      "step": 4179
+    },
+    {
+      "epoch": 10.2200488997555,
+      "grad_norm": 0.3049237722929354,
+      "learning_rate": 1.1515814487081362e-06,
+      "loss": 0.9181,
+      "step": 4180
+    },
+    {
+      "epoch": 10.222493887530563,
+      "grad_norm": 0.2315209866675309,
+      "learning_rate": 1.1505036130764186e-06,
+      "loss": 0.5385,
+      "step": 4181
+    },
+    {
+      "epoch": 10.224938875305623,
+      "grad_norm": 0.2943775324155884,
+      "learning_rate": 1.1494261313135466e-06,
+      "loss": 0.5121,
+      "step": 4182
+    },
+    {
+      "epoch": 10.227383863080684,
+      "grad_norm": 0.4174047381364314,
+      "learning_rate": 1.1483490037020585e-06,
+      "loss": 0.4699,
+      "step": 4183
+    },
+    {
+      "epoch": 10.229828850855746,
+      "grad_norm": 0.3253795652437366,
+      "learning_rate": 1.1472722305244024e-06,
+      "loss": 0.4873,
+      "step": 4184
+    },
+    {
+      "epoch": 10.232273838630807,
+      "grad_norm": 0.2815085250810611,
+      "learning_rate": 1.1461958120629345e-06,
+      "loss": 0.6014,
+      "step": 4185
+    },
+    {
+      "epoch": 10.234718826405867,
+      "grad_norm": 0.2146580038711741,
+      "learning_rate": 1.1451197485999142e-06,
+      "loss": 0.4959,
+      "step": 4186
+    },
+    {
+      "epoch": 10.23716381418093,
+      "grad_norm": 0.2732581361140207,
+      "learning_rate": 1.144044040417512e-06,
+      "loss": 0.7633,
+      "step": 4187
+    },
+    {
+      "epoch": 10.23960880195599,
+      "grad_norm": 0.20660481128824887,
+      "learning_rate": 1.1429686877978006e-06,
+      "loss": 0.6877,
+      "step": 4188
+    },
+    {
+      "epoch": 10.24205378973105,
+      "grad_norm": 0.16109358894117934,
+      "learning_rate": 1.1418936910227629e-06,
+      "loss": 0.6768,
+      "step": 4189
+    },
+    {
+      "epoch": 10.244498777506113,
+      "grad_norm": 0.21063067488195503,
+      "learning_rate": 1.1408190503742903e-06,
+      "loss": 0.4372,
+      "step": 4190
+    },
+    {
+      "epoch": 10.246943765281173,
+      "grad_norm": 0.20266345557795643,
+      "learning_rate": 1.1397447661341751e-06,
+      "loss": 0.3692,
+      "step": 4191
+    },
+    {
+      "epoch": 10.249388753056234,
+      "grad_norm": 0.3409355044102508,
+      "learning_rate": 1.1386708385841214e-06,
+      "loss": 0.4144,
+      "step": 4192
+    },
+    {
+      "epoch": 10.251833740831296,
+      "grad_norm": 0.2135103580797468,
+      "learning_rate": 1.1375972680057356e-06,
+      "loss": 0.4847,
+      "step": 4193
+    },
+    {
+      "epoch": 10.254278728606357,
+      "grad_norm": 0.28838238241776076,
+      "learning_rate": 1.1365240546805332e-06,
+      "loss": 0.4302,
+      "step": 4194
+    },
+    {
+      "epoch": 10.256723716381417,
+      "grad_norm": 0.19391426535568707,
+      "learning_rate": 1.1354511988899366e-06,
+      "loss": 0.6519,
+      "step": 4195
+    },
+    {
+      "epoch": 10.25916870415648,
+      "grad_norm": 0.2516415184994791,
+      "learning_rate": 1.1343787009152709e-06,
+      "loss": 0.6085,
+      "step": 4196
+    },
+    {
+      "epoch": 10.26161369193154,
+      "grad_norm": 0.4803650111397249,
+      "learning_rate": 1.133306561037772e-06,
+      "loss": 0.7518,
+      "step": 4197
+    },
+    {
+      "epoch": 10.2640586797066,
+      "grad_norm": 0.3299564140719025,
+      "learning_rate": 1.1322347795385772e-06,
+      "loss": 0.8661,
+      "step": 4198
+    },
+    {
+      "epoch": 10.266503667481663,
+      "grad_norm": 0.2670071811253802,
+      "learning_rate": 1.1311633566987333e-06,
+      "loss": 0.63,
+      "step": 4199
+    },
+    {
+      "epoch": 10.268948655256724,
+      "grad_norm": 0.26863777400741456,
+      "learning_rate": 1.1300922927991914e-06,
+      "loss": 0.3205,
+      "step": 4200
+    },
+    {
+      "epoch": 10.271393643031784,
+      "grad_norm": 0.25421582705898893,
+      "learning_rate": 1.1290215881208096e-06,
+      "loss": 0.6836,
+      "step": 4201
+    },
+    {
+      "epoch": 10.273838630806846,
+      "grad_norm": 0.25380212123474405,
+      "learning_rate": 1.1279512429443524e-06,
+      "loss": 0.4074,
+      "step": 4202
+    },
+    {
+      "epoch": 10.276283618581907,
+      "grad_norm": 0.24722976283467396,
+      "learning_rate": 1.126881257550486e-06,
+      "loss": 0.4332,
+      "step": 4203
+    },
+    {
+      "epoch": 10.278728606356967,
+      "grad_norm": 0.42665278112790245,
+      "learning_rate": 1.1258116322197861e-06,
+      "loss": 0.6619,
+      "step": 4204
+    },
+    {
+      "epoch": 10.28117359413203,
+      "grad_norm": 0.37727090572341593,
+      "learning_rate": 1.1247423672327349e-06,
+      "loss": 0.3989,
+      "step": 4205
+    },
+    {
+      "epoch": 10.28361858190709,
+      "grad_norm": 0.29823749757789775,
+      "learning_rate": 1.1236734628697152e-06,
+      "loss": 0.6988,
+      "step": 4206
+    },
+    {
+      "epoch": 10.28606356968215,
+      "grad_norm": 0.24808436034893372,
+      "learning_rate": 1.1226049194110206e-06,
+      "loss": 0.7238,
+      "step": 4207
+    },
+    {
+      "epoch": 10.288508557457213,
+      "grad_norm": 0.2505010264085653,
+      "learning_rate": 1.1215367371368455e-06,
+      "loss": 0.4817,
+      "step": 4208
+    },
+    {
+      "epoch": 10.290953545232274,
+      "grad_norm": 0.23820776669465543,
+      "learning_rate": 1.1204689163272931e-06,
+      "loss": 0.4623,
+      "step": 4209
+    },
+    {
+      "epoch": 10.293398533007334,
+      "grad_norm": 0.2448760834274258,
+      "learning_rate": 1.1194014572623699e-06,
+      "loss": 0.4795,
+      "step": 4210
+    },
+    {
+      "epoch": 10.295843520782396,
+      "grad_norm": 0.21855635872086634,
+      "learning_rate": 1.1183343602219884e-06,
+      "loss": 0.4256,
+      "step": 4211
+    },
+    {
+      "epoch": 10.298288508557457,
+      "grad_norm": 0.25934787937900033,
+      "learning_rate": 1.1172676254859672e-06,
+      "loss": 0.3763,
+      "step": 4212
+    },
+    {
+      "epoch": 10.30073349633252,
+      "grad_norm": 0.21593605947712682,
+      "learning_rate": 1.1162012533340257e-06,
+      "loss": 0.8065,
+      "step": 4213
+    },
+    {
+      "epoch": 10.30317848410758,
+      "grad_norm": 0.333071118152502,
+      "learning_rate": 1.115135244045793e-06,
+      "loss": 0.8674,
+      "step": 4214
+    },
+    {
+      "epoch": 10.30562347188264,
+      "grad_norm": 0.21052561234933265,
+      "learning_rate": 1.1140695979008017e-06,
+      "loss": 0.5846,
+      "step": 4215
+    },
+    {
+      "epoch": 10.3080684596577,
+      "grad_norm": 0.21395761980933636,
+      "learning_rate": 1.1130043151784866e-06,
+      "loss": 0.5805,
+      "step": 4216
+    },
+    {
+      "epoch": 10.310513447432763,
+      "grad_norm": 0.2610315453236665,
+      "learning_rate": 1.1119393961581915e-06,
+      "loss": 0.4286,
+      "step": 4217
+    },
+    {
+      "epoch": 10.312958435207824,
+      "grad_norm": 0.284873353549121,
+      "learning_rate": 1.11087484111916e-06,
+      "loss": 0.4579,
+      "step": 4218
+    },
+    {
+      "epoch": 10.315403422982886,
+      "grad_norm": 0.2660496054519845,
+      "learning_rate": 1.1098106503405443e-06,
+      "loss": 0.4459,
+      "step": 4219
+    },
+    {
+      "epoch": 10.317848410757946,
+      "grad_norm": 0.2221084619552763,
+      "learning_rate": 1.1087468241014005e-06,
+      "loss": 0.5036,
+      "step": 4220
+    },
+    {
+      "epoch": 10.320293398533007,
+      "grad_norm": 0.27584253905045814,
+      "learning_rate": 1.1076833626806852e-06,
+      "loss": 0.502,
+      "step": 4221
+    },
+    {
+      "epoch": 10.32273838630807,
+      "grad_norm": 0.2846802268178864,
+      "learning_rate": 1.106620266357266e-06,
+      "loss": 0.4134,
+      "step": 4222
+    },
+    {
+      "epoch": 10.32518337408313,
+      "grad_norm": 0.30832109832740445,
+      "learning_rate": 1.1055575354099083e-06,
+      "loss": 0.6541,
+      "step": 4223
+    },
+    {
+      "epoch": 10.32762836185819,
+      "grad_norm": 0.3670009084712405,
+      "learning_rate": 1.1044951701172857e-06,
+      "loss": 0.5937,
+      "step": 4224
+    },
+    {
+      "epoch": 10.330073349633253,
+      "grad_norm": 0.3260661340177253,
+      "learning_rate": 1.1034331707579752e-06,
+      "loss": 0.3496,
+      "step": 4225
+    },
+    {
+      "epoch": 10.332518337408313,
+      "grad_norm": 0.25570231793931764,
+      "learning_rate": 1.102371537610456e-06,
+      "loss": 0.4066,
+      "step": 4226
+    },
+    {
+      "epoch": 10.334963325183374,
+      "grad_norm": 0.2572165327921113,
+      "learning_rate": 1.101310270953114e-06,
+      "loss": 0.5803,
+      "step": 4227
+    },
+    {
+      "epoch": 10.337408312958436,
+      "grad_norm": 0.35429454241041924,
+      "learning_rate": 1.100249371064236e-06,
+      "loss": 0.4324,
+      "step": 4228
+    },
+    {
+      "epoch": 10.339853300733497,
+      "grad_norm": 0.3752294459315225,
+      "learning_rate": 1.0991888382220151e-06,
+      "loss": 0.5597,
+      "step": 4229
+    },
+    {
+      "epoch": 10.342298288508557,
+      "grad_norm": 0.36426667257913814,
+      "learning_rate": 1.0981286727045484e-06,
+      "loss": 0.5793,
+      "step": 4230
+    },
+    {
+      "epoch": 10.34474327628362,
+      "grad_norm": 0.21342809231005616,
+      "learning_rate": 1.0970688747898335e-06,
+      "loss": 0.4607,
+      "step": 4231
+    },
+    {
+      "epoch": 10.34718826405868,
+      "grad_norm": 0.2758253952506122,
+      "learning_rate": 1.0960094447557742e-06,
+      "loss": 1.2919,
+      "step": 4232
+    },
+    {
+      "epoch": 10.34963325183374,
+      "grad_norm": 0.3018893020271271,
+      "learning_rate": 1.0949503828801783e-06,
+      "loss": 0.7432,
+      "step": 4233
+    },
+    {
+      "epoch": 10.352078239608803,
+      "grad_norm": 0.2589862181460002,
+      "learning_rate": 1.0938916894407553e-06,
+      "loss": 0.4057,
+      "step": 4234
+    },
+    {
+      "epoch": 10.354523227383863,
+      "grad_norm": 0.16981586539910698,
+      "learning_rate": 1.09283336471512e-06,
+      "loss": 0.5141,
+      "step": 4235
+    },
+    {
+      "epoch": 10.356968215158924,
+      "grad_norm": 0.26154933006397024,
+      "learning_rate": 1.0917754089807875e-06,
+      "loss": 0.9263,
+      "step": 4236
+    },
+    {
+      "epoch": 10.359413202933986,
+      "grad_norm": 0.2663676827423522,
+      "learning_rate": 1.0907178225151799e-06,
+      "loss": 0.3686,
+      "step": 4237
+    },
+    {
+      "epoch": 10.361858190709047,
+      "grad_norm": 0.2888314656070139,
+      "learning_rate": 1.0896606055956186e-06,
+      "loss": 0.5378,
+      "step": 4238
+    },
+    {
+      "epoch": 10.364303178484107,
+      "grad_norm": 0.21432782182469964,
+      "learning_rate": 1.0886037584993311e-06,
+      "loss": 0.8412,
+      "step": 4239
+    },
+    {
+      "epoch": 10.36674816625917,
+      "grad_norm": 0.15904350964729666,
+      "learning_rate": 1.0875472815034483e-06,
+      "loss": 0.1969,
+      "step": 4240
+    },
+    {
+      "epoch": 10.36919315403423,
+      "grad_norm": 0.3847422486539824,
+      "learning_rate": 1.086491174885e-06,
+      "loss": 0.335,
+      "step": 4241
+    },
+    {
+      "epoch": 10.37163814180929,
+      "grad_norm": 0.244886244899407,
+      "learning_rate": 1.0854354389209226e-06,
+      "loss": 0.6841,
+      "step": 4242
+    },
+    {
+      "epoch": 10.374083129584353,
+      "grad_norm": 0.2423179358033271,
+      "learning_rate": 1.0843800738880545e-06,
+      "loss": 0.5508,
+      "step": 4243
+    },
+    {
+      "epoch": 10.376528117359413,
+      "grad_norm": 0.24481050768794999,
+      "learning_rate": 1.0833250800631368e-06,
+      "loss": 0.3785,
+      "step": 4244
+    },
+    {
+      "epoch": 10.378973105134474,
+      "grad_norm": 0.41548784648028797,
+      "learning_rate": 1.0822704577228133e-06,
+      "loss": 0.7046,
+      "step": 4245
+    },
+    {
+      "epoch": 10.381418092909536,
+      "grad_norm": 0.376757479458428,
+      "learning_rate": 1.0812162071436288e-06,
+      "loss": 0.5152,
+      "step": 4246
+    },
+    {
+      "epoch": 10.383863080684597,
+      "grad_norm": 0.22139308348613054,
+      "learning_rate": 1.0801623286020327e-06,
+      "loss": 0.3987,
+      "step": 4247
+    },
+    {
+      "epoch": 10.386308068459657,
+      "grad_norm": 0.3161245954958729,
+      "learning_rate": 1.0791088223743771e-06,
+      "loss": 0.4938,
+      "step": 4248
+    },
+    {
+      "epoch": 10.38875305623472,
+      "grad_norm": 0.27960886231599424,
+      "learning_rate": 1.078055688736913e-06,
+      "loss": 0.7321,
+      "step": 4249
+    },
+    {
+      "epoch": 10.39119804400978,
+      "grad_norm": 0.45091483854375275,
+      "learning_rate": 1.077002927965799e-06,
+      "loss": 0.4293,
+      "step": 4250
+    },
+    {
+      "epoch": 10.39364303178484,
+      "grad_norm": 0.24169982156743752,
+      "learning_rate": 1.0759505403370906e-06,
+      "loss": 0.3913,
+      "step": 4251
+    },
+    {
+      "epoch": 10.396088019559903,
+      "grad_norm": 0.33563048052452765,
+      "learning_rate": 1.074898526126749e-06,
+      "loss": 0.8785,
+      "step": 4252
+    },
+    {
+      "epoch": 10.398533007334963,
+      "grad_norm": 0.22860336753584973,
+      "learning_rate": 1.0738468856106363e-06,
+      "loss": 0.4365,
+      "step": 4253
+    },
+    {
+      "epoch": 10.400977995110024,
+      "grad_norm": 0.41847078332940246,
+      "learning_rate": 1.0727956190645166e-06,
+      "loss": 0.4003,
+      "step": 4254
+    },
+    {
+      "epoch": 10.403422982885086,
+      "grad_norm": 0.26386490365276594,
+      "learning_rate": 1.0717447267640575e-06,
+      "loss": 0.5579,
+      "step": 4255
+    },
+    {
+      "epoch": 10.405867970660147,
+      "grad_norm": 0.3315566599584672,
+      "learning_rate": 1.0706942089848246e-06,
+      "loss": 0.3855,
+      "step": 4256
+    },
+    {
+      "epoch": 10.408312958435207,
+      "grad_norm": 0.26863576263656813,
+      "learning_rate": 1.0696440660022892e-06,
+      "loss": 0.4303,
+      "step": 4257
+    },
+    {
+      "epoch": 10.41075794621027,
+      "grad_norm": 0.2921422625151501,
+      "learning_rate": 1.068594298091823e-06,
+      "loss": 0.4861,
+      "step": 4258
+    },
+    {
+      "epoch": 10.41320293398533,
+      "grad_norm": 0.26531808905998455,
+      "learning_rate": 1.067544905528698e-06,
+      "loss": 0.7276,
+      "step": 4259
+    },
+    {
+      "epoch": 10.41564792176039,
+      "grad_norm": 0.2860387956517466,
+      "learning_rate": 1.0664958885880903e-06,
+      "loss": 0.6428,
+      "step": 4260
+    },
+    {
+      "epoch": 10.418092909535453,
+      "grad_norm": 0.36487813397526914,
+      "learning_rate": 1.0654472475450748e-06,
+      "loss": 0.6094,
+      "step": 4261
+    },
+    {
+      "epoch": 10.420537897310513,
+      "grad_norm": 0.19662818281492933,
+      "learning_rate": 1.0643989826746298e-06,
+      "loss": 0.4207,
+      "step": 4262
+    },
+    {
+      "epoch": 10.422982885085574,
+      "grad_norm": 0.3788751583815013,
+      "learning_rate": 1.0633510942516354e-06,
+      "loss": 0.4877,
+      "step": 4263
+    },
+    {
+      "epoch": 10.425427872860636,
+      "grad_norm": 0.26201849194313936,
+      "learning_rate": 1.062303582550869e-06,
+      "loss": 0.61,
+      "step": 4264
+    },
+    {
+      "epoch": 10.427872860635697,
+      "grad_norm": 0.3081976089537839,
+      "learning_rate": 1.0612564478470158e-06,
+      "loss": 0.4374,
+      "step": 4265
+    },
+    {
+      "epoch": 10.430317848410757,
+      "grad_norm": 0.32562078149842,
+      "learning_rate": 1.0602096904146561e-06,
+      "loss": 0.4765,
+      "step": 4266
+    },
+    {
+      "epoch": 10.43276283618582,
+      "grad_norm": 0.31390020641837185,
+      "learning_rate": 1.0591633105282742e-06,
+      "loss": 0.5158,
+      "step": 4267
+    },
+    {
+      "epoch": 10.43520782396088,
+      "grad_norm": 0.21320436571176027,
+      "learning_rate": 1.058117308462256e-06,
+      "loss": 0.6171,
+      "step": 4268
+    },
+    {
+      "epoch": 10.43765281173594,
+      "grad_norm": 0.20694455970272932,
+      "learning_rate": 1.0570716844908854e-06,
+      "loss": 0.7641,
+      "step": 4269
+    },
+    {
+      "epoch": 10.440097799511003,
+      "grad_norm": 0.433356465180769,
+      "learning_rate": 1.0560264388883505e-06,
+      "loss": 0.6787,
+      "step": 4270
+    },
+    {
+      "epoch": 10.442542787286063,
+      "grad_norm": 0.19627752136973292,
+      "learning_rate": 1.0549815719287367e-06,
+      "loss": 0.5985,
+      "step": 4271
+    },
+    {
+      "epoch": 10.444987775061124,
+      "grad_norm": 0.3532235142292159,
+      "learning_rate": 1.0539370838860334e-06,
+      "loss": 0.6049,
+      "step": 4272
+    },
+    {
+      "epoch": 10.447432762836186,
+      "grad_norm": 0.18998209868258623,
+      "learning_rate": 1.0528929750341303e-06,
+      "loss": 0.4402,
+      "step": 4273
+    },
+    {
+      "epoch": 10.449877750611247,
+      "grad_norm": 0.2167927347712745,
+      "learning_rate": 1.051849245646814e-06,
+      "loss": 0.3878,
+      "step": 4274
+    },
+    {
+      "epoch": 10.452322738386307,
+      "grad_norm": 0.2235549317041821,
+      "learning_rate": 1.0508058959977757e-06,
+      "loss": 0.8425,
+      "step": 4275
+    },
+    {
+      "epoch": 10.45476772616137,
+      "grad_norm": 0.2556532345971111,
+      "learning_rate": 1.0497629263606057e-06,
+      "loss": 0.4573,
+      "step": 4276
+    },
+    {
+      "epoch": 10.45721271393643,
+      "grad_norm": 0.26422450912041473,
+      "learning_rate": 1.0487203370087943e-06,
+      "loss": 0.5472,
+      "step": 4277
+    },
+    {
+      "epoch": 10.45965770171149,
+      "grad_norm": 0.20641761692696617,
+      "learning_rate": 1.0476781282157332e-06,
+      "loss": 0.452,
+      "step": 4278
+    },
+    {
+      "epoch": 10.462102689486553,
+      "grad_norm": 0.2308920442199556,
+      "learning_rate": 1.0466363002547117e-06,
+      "loss": 0.7833,
+      "step": 4279
+    },
+    {
+      "epoch": 10.464547677261614,
+      "grad_norm": 0.2818830072693171,
+      "learning_rate": 1.0455948533989225e-06,
+      "loss": 0.4071,
+      "step": 4280
+    },
+    {
+      "epoch": 10.466992665036674,
+      "grad_norm": 0.20014125332339097,
+      "learning_rate": 1.044553787921455e-06,
+      "loss": 0.3706,
+      "step": 4281
+    },
+    {
+      "epoch": 10.469437652811736,
+      "grad_norm": 0.29893637195780276,
+      "learning_rate": 1.0435131040953017e-06,
+      "loss": 0.4334,
+      "step": 4282
+    },
+    {
+      "epoch": 10.471882640586797,
+      "grad_norm": 0.2184268115591645,
+      "learning_rate": 1.0424728021933545e-06,
+      "loss": 0.5834,
+      "step": 4283
+    },
+    {
+      "epoch": 10.474327628361857,
+      "grad_norm": 0.2789675432324769,
+      "learning_rate": 1.0414328824884021e-06,
+      "loss": 0.7017,
+      "step": 4284
+    },
+    {
+      "epoch": 10.47677261613692,
+      "grad_norm": 0.2848807353058191,
+      "learning_rate": 1.0403933452531368e-06,
+      "loss": 0.4981,
+      "step": 4285
+    },
+    {
+      "epoch": 10.47921760391198,
+      "grad_norm": 0.30646649759686806,
+      "learning_rate": 1.0393541907601488e-06,
+      "loss": 0.718,
+      "step": 4286
+    },
+    {
+      "epoch": 10.48166259168704,
+      "grad_norm": 0.2670537262739069,
+      "learning_rate": 1.0383154192819277e-06,
+      "loss": 0.6994,
+      "step": 4287
+    },
+    {
+      "epoch": 10.484107579462103,
+      "grad_norm": 0.278557544019373,
+      "learning_rate": 1.037277031090865e-06,
+      "loss": 0.6547,
+      "step": 4288
+    },
+    {
+      "epoch": 10.486552567237164,
+      "grad_norm": 0.32952705090290973,
+      "learning_rate": 1.0362390264592475e-06,
+      "loss": 0.3579,
+      "step": 4289
+    },
+    {
+      "epoch": 10.488997555012224,
+      "grad_norm": 0.22306291470480558,
+      "learning_rate": 1.0352014056592654e-06,
+      "loss": 0.8264,
+      "step": 4290
+    },
+    {
+      "epoch": 10.491442542787286,
+      "grad_norm": 0.2790616778153501,
+      "learning_rate": 1.0341641689630053e-06,
+      "loss": 0.5048,
+      "step": 4291
+    },
+    {
+      "epoch": 10.493887530562347,
+      "grad_norm": 0.20380835201691247,
+      "learning_rate": 1.0331273166424551e-06,
+      "loss": 0.9128,
+      "step": 4292
+    },
+    {
+      "epoch": 10.496332518337407,
+      "grad_norm": 0.19651500785895423,
+      "learning_rate": 1.032090848969502e-06,
+      "loss": 0.6676,
+      "step": 4293
+    },
+    {
+      "epoch": 10.49877750611247,
+      "grad_norm": 0.26744570732099743,
+      "learning_rate": 1.0310547662159299e-06,
+      "loss": 0.3443,
+      "step": 4294
+    },
+    {
+      "epoch": 10.50122249388753,
+      "grad_norm": 0.32483429652313195,
+      "learning_rate": 1.030019068653425e-06,
+      "loss": 0.4567,
+      "step": 4295
+    },
+    {
+      "epoch": 10.503667481662593,
+      "grad_norm": 0.2442696538945737,
+      "learning_rate": 1.0289837565535688e-06,
+      "loss": 0.4345,
+      "step": 4296
+    },
+    {
+      "epoch": 10.506112469437653,
+      "grad_norm": 0.195352710619373,
+      "learning_rate": 1.0279488301878457e-06,
+      "loss": 0.4908,
+      "step": 4297
+    },
+    {
+      "epoch": 10.508557457212714,
+      "grad_norm": 0.33427791501532744,
+      "learning_rate": 1.0269142898276377e-06,
+      "loss": 0.4971,
+      "step": 4298
+    },
+    {
+      "epoch": 10.511002444987774,
+      "grad_norm": 0.20892654797180402,
+      "learning_rate": 1.0258801357442227e-06,
+      "loss": 0.3987,
+      "step": 4299
+    },
+    {
+      "epoch": 10.513447432762836,
+      "grad_norm": 0.2122906776839145,
+      "learning_rate": 1.024846368208782e-06,
+      "loss": 0.533,
+      "step": 4300
+    },
+    {
+      "epoch": 10.515892420537897,
+      "grad_norm": 0.3228820210091164,
+      "learning_rate": 1.0238129874923905e-06,
+      "loss": 0.7088,
+      "step": 4301
+    },
+    {
+      "epoch": 10.51833740831296,
+      "grad_norm": 0.3013798127110642,
+      "learning_rate": 1.022779993866026e-06,
+      "loss": 0.8274,
+      "step": 4302
+    },
+    {
+      "epoch": 10.52078239608802,
+      "grad_norm": 0.26692594965715,
+      "learning_rate": 1.0217473876005635e-06,
+      "loss": 0.8578,
+      "step": 4303
+    },
+    {
+      "epoch": 10.52322738386308,
+      "grad_norm": 0.24590370222879326,
+      "learning_rate": 1.0207151689667742e-06,
+      "loss": 0.5524,
+      "step": 4304
+    },
+    {
+      "epoch": 10.525672371638143,
+      "grad_norm": 0.38640332647289244,
+      "learning_rate": 1.0196833382353306e-06,
+      "loss": 0.7099,
+      "step": 4305
+    },
+    {
+      "epoch": 10.528117359413203,
+      "grad_norm": 0.30321522549706703,
+      "learning_rate": 1.0186518956768026e-06,
+      "loss": 0.5987,
+      "step": 4306
+    },
+    {
+      "epoch": 10.530562347188264,
+      "grad_norm": 0.28960180745559505,
+      "learning_rate": 1.0176208415616564e-06,
+      "loss": 0.776,
+      "step": 4307
+    },
+    {
+      "epoch": 10.533007334963326,
+      "grad_norm": 0.20923029078323754,
+      "learning_rate": 1.0165901761602608e-06,
+      "loss": 0.583,
+      "step": 4308
+    },
+    {
+      "epoch": 10.535452322738386,
+      "grad_norm": 0.25635600129741176,
+      "learning_rate": 1.0155598997428773e-06,
+      "loss": 0.4153,
+      "step": 4309
+    },
+    {
+      "epoch": 10.537897310513447,
+      "grad_norm": 0.2948867049894463,
+      "learning_rate": 1.01453001257967e-06,
+      "loss": 0.5548,
+      "step": 4310
+    },
+    {
+      "epoch": 10.54034229828851,
+      "grad_norm": 0.22432599513869883,
+      "learning_rate": 1.013500514940697e-06,
+      "loss": 0.4419,
+      "step": 4311
+    },
+    {
+      "epoch": 10.54278728606357,
+      "grad_norm": 0.2570326252242165,
+      "learning_rate": 1.012471407095917e-06,
+      "loss": 0.7046,
+      "step": 4312
+    },
+    {
+      "epoch": 10.54523227383863,
+      "grad_norm": 0.19928909167986347,
+      "learning_rate": 1.0114426893151864e-06,
+      "loss": 0.5154,
+      "step": 4313
+    },
+    {
+      "epoch": 10.547677261613693,
+      "grad_norm": 0.21777738208008943,
+      "learning_rate": 1.0104143618682568e-06,
+      "loss": 0.3773,
+      "step": 4314
+    },
+    {
+      "epoch": 10.550122249388753,
+      "grad_norm": 0.1593210421520442,
+      "learning_rate": 1.0093864250247804e-06,
+      "loss": 0.3999,
+      "step": 4315
+    },
+    {
+      "epoch": 10.552567237163814,
+      "grad_norm": 0.3586817935567945,
+      "learning_rate": 1.0083588790543062e-06,
+      "loss": 0.5295,
+      "step": 4316
+    },
+    {
+      "epoch": 10.555012224938876,
+      "grad_norm": 0.29243419840995577,
+      "learning_rate": 1.0073317242262787e-06,
+      "loss": 0.4194,
+      "step": 4317
+    },
+    {
+      "epoch": 10.557457212713937,
+      "grad_norm": 0.26275739205581367,
+      "learning_rate": 1.0063049608100426e-06,
+      "loss": 0.6337,
+      "step": 4318
+    },
+    {
+      "epoch": 10.559902200488997,
+      "grad_norm": 0.2756060460916107,
+      "learning_rate": 1.0052785890748384e-06,
+      "loss": 0.9403,
+      "step": 4319
+    },
+    {
+      "epoch": 10.56234718826406,
+      "grad_norm": 0.20829443309216508,
+      "learning_rate": 1.004252609289805e-06,
+      "loss": 0.9553,
+      "step": 4320
+    },
+    {
+      "epoch": 10.56479217603912,
+      "grad_norm": 0.27059316195803773,
+      "learning_rate": 1.0032270217239765e-06,
+      "loss": 0.8934,
+      "step": 4321
+    },
+    {
+      "epoch": 10.56723716381418,
+      "grad_norm": 0.3626143064646002,
+      "learning_rate": 1.0022018266462861e-06,
+      "loss": 0.4049,
+      "step": 4322
+    },
+    {
+      "epoch": 10.569682151589243,
+      "grad_norm": 0.1927649523245912,
+      "learning_rate": 1.0011770243255642e-06,
+      "loss": 0.4423,
+      "step": 4323
+    },
+    {
+      "epoch": 10.572127139364303,
+      "grad_norm": 0.21369774068766934,
+      "learning_rate": 1.0001526150305357e-06,
+      "loss": 0.7241,
+      "step": 4324
+    },
+    {
+      "epoch": 10.574572127139364,
+      "grad_norm": 0.3759232350000817,
+      "learning_rate": 9.991285990298253e-07,
+      "loss": 0.435,
+      "step": 4325
+    },
+    {
+      "epoch": 10.577017114914426,
+      "grad_norm": 0.17815386883869375,
+      "learning_rate": 9.98104976591954e-07,
+      "loss": 0.6977,
+      "step": 4326
+    },
+    {
+      "epoch": 10.579462102689487,
+      "grad_norm": 0.17530810875953795,
+      "learning_rate": 9.970817479853372e-07,
+      "loss": 0.3524,
+      "step": 4327
+    },
+    {
+      "epoch": 10.581907090464547,
+      "grad_norm": 0.23459776835546303,
+      "learning_rate": 9.960589134782903e-07,
+      "loss": 0.966,
+      "step": 4328
+    },
+    {
+      "epoch": 10.58435207823961,
+      "grad_norm": 0.18033747891273672,
+      "learning_rate": 9.950364733390233e-07,
+      "loss": 0.398,
+      "step": 4329
+    },
+    {
+      "epoch": 10.58679706601467,
+      "grad_norm": 0.21279943215915337,
+      "learning_rate": 9.940144278356444e-07,
+      "loss": 0.5577,
+      "step": 4330
+    },
+    {
+      "epoch": 10.58924205378973,
+      "grad_norm": 0.23986507967631324,
+      "learning_rate": 9.929927772361562e-07,
+      "loss": 0.4536,
+      "step": 4331
+    },
+    {
+      "epoch": 10.591687041564793,
+      "grad_norm": 0.23190694368363074,
+      "learning_rate": 9.919715218084588e-07,
+      "loss": 0.5099,
+      "step": 4332
+    },
+    {
+      "epoch": 10.594132029339853,
+      "grad_norm": 0.2181617359119374,
+      "learning_rate": 9.909506618203501e-07,
+      "loss": 0.4519,
+      "step": 4333
+    },
+    {
+      "epoch": 10.596577017114914,
+      "grad_norm": 0.2620375780093581,
+      "learning_rate": 9.899301975395214e-07,
+      "loss": 0.6629,
+      "step": 4334
+    },
+    {
+      "epoch": 10.599022004889976,
+      "grad_norm": 0.2544811026050055,
+      "learning_rate": 9.889101292335625e-07,
+      "loss": 0.5131,
+      "step": 4335
+    },
+    {
+      "epoch": 10.601466992665037,
+      "grad_norm": 0.24143413021218876,
+      "learning_rate": 9.878904571699596e-07,
+      "loss": 0.8713,
+      "step": 4336
+    },
+    {
+      "epoch": 10.603911980440097,
+      "grad_norm": 0.25564385214669744,
+      "learning_rate": 9.86871181616092e-07,
+      "loss": 0.5029,
+      "step": 4337
+    },
+    {
+      "epoch": 10.60635696821516,
+      "grad_norm": 0.38465103858679484,
+      "learning_rate": 9.85852302839239e-07,
+      "loss": 0.5773,
+      "step": 4338
+    },
+    {
+      "epoch": 10.60880195599022,
+      "grad_norm": 0.22444021703376024,
+      "learning_rate": 9.848338211065714e-07,
+      "loss": 0.8214,
+      "step": 4339
+    },
+    {
+      "epoch": 10.61124694376528,
+      "grad_norm": 0.2792213262790953,
+      "learning_rate": 9.83815736685162e-07,
+      "loss": 0.6426,
+      "step": 4340
+    },
+    {
+      "epoch": 10.613691931540343,
+      "grad_norm": 0.28763017776667527,
+      "learning_rate": 9.82798049841973e-07,
+      "loss": 0.4784,
+      "step": 4341
+    },
+    {
+      "epoch": 10.616136919315403,
+      "grad_norm": 0.29670939092781656,
+      "learning_rate": 9.817807608438663e-07,
+      "loss": 0.4279,
+      "step": 4342
+    },
+    {
+      "epoch": 10.618581907090464,
+      "grad_norm": 0.330271466928609,
+      "learning_rate": 9.807638699575988e-07,
+      "loss": 0.5066,
+      "step": 4343
+    },
+    {
+      "epoch": 10.621026894865526,
+      "grad_norm": 0.23499697602317826,
+      "learning_rate": 9.797473774498212e-07,
+      "loss": 0.8341,
+      "step": 4344
+    },
+    {
+      "epoch": 10.623471882640587,
+      "grad_norm": 0.24688716661319512,
+      "learning_rate": 9.78731283587082e-07,
+      "loss": 0.4716,
+      "step": 4345
+    },
+    {
+      "epoch": 10.625916870415647,
+      "grad_norm": 0.20704999293221601,
+      "learning_rate": 9.77715588635825e-07,
+      "loss": 0.4567,
+      "step": 4346
+    },
+    {
+      "epoch": 10.62836185819071,
+      "grad_norm": 0.20415077914130392,
+      "learning_rate": 9.76700292862387e-07,
+      "loss": 0.4772,
+      "step": 4347
+    },
+    {
+      "epoch": 10.63080684596577,
+      "grad_norm": 0.33140905873083754,
+      "learning_rate": 9.756853965330035e-07,
+      "loss": 0.6902,
+      "step": 4348
+    },
+    {
+      "epoch": 10.63325183374083,
+      "grad_norm": 0.1818198379603264,
+      "learning_rate": 9.74670899913802e-07,
+      "loss": 0.3833,
+      "step": 4349
+    },
+    {
+      "epoch": 10.635696821515893,
+      "grad_norm": 0.3250422855406044,
+      "learning_rate": 9.736568032708068e-07,
+      "loss": 0.4677,
+      "step": 4350
+    },
+    {
+      "epoch": 10.638141809290953,
+      "grad_norm": 0.21018869986651117,
+      "learning_rate": 9.726431068699385e-07,
+      "loss": 0.3753,
+      "step": 4351
+    },
+    {
+      "epoch": 10.640586797066014,
+      "grad_norm": 0.325376689959012,
+      "learning_rate": 9.716298109770107e-07,
+      "loss": 0.5235,
+      "step": 4352
+    },
+    {
+      "epoch": 10.643031784841076,
+      "grad_norm": 0.3099870718506511,
+      "learning_rate": 9.706169158577336e-07,
+      "loss": 0.7004,
+      "step": 4353
+    },
+    {
+      "epoch": 10.645476772616137,
+      "grad_norm": 0.18464272261388961,
+      "learning_rate": 9.696044217777101e-07,
+      "loss": 0.6622,
+      "step": 4354
+    },
+    {
+      "epoch": 10.647921760391197,
+      "grad_norm": 0.26758599303735314,
+      "learning_rate": 9.685923290024398e-07,
+      "loss": 0.5943,
+      "step": 4355
+    },
+    {
+      "epoch": 10.65036674816626,
+      "grad_norm": 0.22330967176009905,
+      "learning_rate": 9.675806377973183e-07,
+      "loss": 0.3678,
+      "step": 4356
+    },
+    {
+      "epoch": 10.65281173594132,
+      "grad_norm": 0.2241055381192565,
+      "learning_rate": 9.665693484276315e-07,
+      "loss": 0.9309,
+      "step": 4357
+    },
+    {
+      "epoch": 10.65525672371638,
+      "grad_norm": 0.20088571868784003,
+      "learning_rate": 9.655584611585647e-07,
+      "loss": 0.8415,
+      "step": 4358
+    },
+    {
+      "epoch": 10.657701711491443,
+      "grad_norm": 0.2737798661349978,
+      "learning_rate": 9.64547976255194e-07,
+      "loss": 0.4643,
+      "step": 4359
+    },
+    {
+      "epoch": 10.660146699266503,
+      "grad_norm": 0.2092178363833883,
+      "learning_rate": 9.635378939824925e-07,
+      "loss": 0.7475,
+      "step": 4360
+    },
+    {
+      "epoch": 10.662591687041564,
+      "grad_norm": 0.24399175982209115,
+      "learning_rate": 9.62528214605327e-07,
+      "loss": 0.8379,
+      "step": 4361
+    },
+    {
+      "epoch": 10.665036674816626,
+      "grad_norm": 0.19122089848163124,
+      "learning_rate": 9.615189383884585e-07,
+      "loss": 0.4814,
+      "step": 4362
+    },
+    {
+      "epoch": 10.667481662591687,
+      "grad_norm": 0.19134590066254722,
+      "learning_rate": 9.60510065596543e-07,
+      "loss": 0.4581,
+      "step": 4363
+    },
+    {
+      "epoch": 10.669926650366747,
+      "grad_norm": 0.25128884113975064,
+      "learning_rate": 9.595015964941287e-07,
+      "loss": 1.0461,
+      "step": 4364
+    },
+    {
+      "epoch": 10.67237163814181,
+      "grad_norm": 0.3547101290265935,
+      "learning_rate": 9.584935313456597e-07,
+      "loss": 0.3851,
+      "step": 4365
+    },
+    {
+      "epoch": 10.67481662591687,
+      "grad_norm": 0.3610904754610384,
+      "learning_rate": 9.574858704154749e-07,
+      "loss": 0.6249,
+      "step": 4366
+    },
+    {
+      "epoch": 10.67726161369193,
+      "grad_norm": 0.3503368125276897,
+      "learning_rate": 9.564786139678043e-07,
+      "loss": 0.632,
+      "step": 4367
+    },
+    {
+      "epoch": 10.679706601466993,
+      "grad_norm": 0.26365466941111376,
+      "learning_rate": 9.554717622667752e-07,
+      "loss": 0.7467,
+      "step": 4368
+    },
+    {
+      "epoch": 10.682151589242054,
+      "grad_norm": 0.3095614530979863,
+      "learning_rate": 9.54465315576406e-07,
+      "loss": 0.5197,
+      "step": 4369
+    },
+    {
+      "epoch": 10.684596577017114,
+      "grad_norm": 0.2847381720183174,
+      "learning_rate": 9.534592741606097e-07,
+      "loss": 1.3271,
+      "step": 4370
+    },
+    {
+      "epoch": 10.687041564792176,
+      "grad_norm": 0.25389374066124937,
+      "learning_rate": 9.524536382831947e-07,
+      "loss": 0.6757,
+      "step": 4371
+    },
+    {
+      "epoch": 10.689486552567237,
+      "grad_norm": 0.21911826613625454,
+      "learning_rate": 9.514484082078609e-07,
+      "loss": 0.4052,
+      "step": 4372
+    },
+    {
+      "epoch": 10.6919315403423,
+      "grad_norm": 0.1764902828990768,
+      "learning_rate": 9.504435841982041e-07,
+      "loss": 0.6313,
+      "step": 4373
+    },
+    {
+      "epoch": 10.69437652811736,
+      "grad_norm": 0.331708896072561,
+      "learning_rate": 9.494391665177097e-07,
+      "loss": 0.3464,
+      "step": 4374
+    },
+    {
+      "epoch": 10.69682151589242,
+      "grad_norm": 0.19659752072918085,
+      "learning_rate": 9.484351554297605e-07,
+      "loss": 0.5552,
+      "step": 4375
+    },
+    {
+      "epoch": 10.69926650366748,
+      "grad_norm": 0.22942094415746575,
+      "learning_rate": 9.474315511976321e-07,
+      "loss": 0.5713,
+      "step": 4376
+    },
+    {
+      "epoch": 10.701711491442543,
+      "grad_norm": 0.2687075149302389,
+      "learning_rate": 9.464283540844902e-07,
+      "loss": 0.5856,
+      "step": 4377
+    },
+    {
+      "epoch": 10.704156479217604,
+      "grad_norm": 0.1921975744919534,
+      "learning_rate": 9.454255643533983e-07,
+      "loss": 0.484,
+      "step": 4378
+    },
+    {
+      "epoch": 10.706601466992666,
+      "grad_norm": 0.2869895532461907,
+      "learning_rate": 9.444231822673086e-07,
+      "loss": 0.4356,
+      "step": 4379
+    },
+    {
+      "epoch": 10.709046454767726,
+      "grad_norm": 0.15579817052682807,
+      "learning_rate": 9.434212080890701e-07,
+      "loss": 0.4563,
+      "step": 4380
+    },
+    {
+      "epoch": 10.711491442542787,
+      "grad_norm": 0.32465613513550284,
+      "learning_rate": 9.424196420814239e-07,
+      "loss": 0.3911,
+      "step": 4381
+    },
+    {
+      "epoch": 10.713936430317847,
+      "grad_norm": 0.31794639758301396,
+      "learning_rate": 9.414184845070007e-07,
+      "loss": 0.681,
+      "step": 4382
+    },
+    {
+      "epoch": 10.71638141809291,
+      "grad_norm": 0.2972437232345382,
+      "learning_rate": 9.404177356283309e-07,
+      "loss": 0.3986,
+      "step": 4383
+    },
+    {
+      "epoch": 10.71882640586797,
+      "grad_norm": 0.3199035978556867,
+      "learning_rate": 9.394173957078309e-07,
+      "loss": 0.3613,
+      "step": 4384
+    },
+    {
+      "epoch": 10.721271393643033,
+      "grad_norm": 0.3287698075179986,
+      "learning_rate": 9.384174650078135e-07,
+      "loss": 0.5073,
+      "step": 4385
+    },
+    {
+      "epoch": 10.723716381418093,
+      "grad_norm": 0.1995913938730153,
+      "learning_rate": 9.374179437904846e-07,
+      "loss": 0.6236,
+      "step": 4386
+    },
+    {
+      "epoch": 10.726161369193154,
+      "grad_norm": 0.2637770153765454,
+      "learning_rate": 9.364188323179394e-07,
+      "loss": 0.7355,
+      "step": 4387
+    },
+    {
+      "epoch": 10.728606356968216,
+      "grad_norm": 0.2292466546258323,
+      "learning_rate": 9.354201308521698e-07,
+      "loss": 0.7841,
+      "step": 4388
+    },
+    {
+      "epoch": 10.731051344743276,
+      "grad_norm": 0.2826226884825545,
+      "learning_rate": 9.344218396550563e-07,
+      "loss": 0.4379,
+      "step": 4389
+    },
+    {
+      "epoch": 10.733496332518337,
+      "grad_norm": 0.2230679421584428,
+      "learning_rate": 9.334239589883748e-07,
+      "loss": 0.3922,
+      "step": 4390
+    },
+    {
+      "epoch": 10.7359413202934,
+      "grad_norm": 0.3298159686786096,
+      "learning_rate": 9.324264891137933e-07,
+      "loss": 0.7186,
+      "step": 4391
+    },
+    {
+      "epoch": 10.73838630806846,
+      "grad_norm": 0.22234638548153243,
+      "learning_rate": 9.314294302928692e-07,
+      "loss": 0.8327,
+      "step": 4392
+    },
+    {
+      "epoch": 10.74083129584352,
+      "grad_norm": 0.38196027450070896,
+      "learning_rate": 9.304327827870552e-07,
+      "loss": 0.7486,
+      "step": 4393
+    },
+    {
+      "epoch": 10.743276283618583,
+      "grad_norm": 0.2634863222326317,
+      "learning_rate": 9.294365468576951e-07,
+      "loss": 0.3903,
+      "step": 4394
+    },
+    {
+      "epoch": 10.745721271393643,
+      "grad_norm": 0.2991866041113629,
+      "learning_rate": 9.284407227660249e-07,
+      "loss": 0.548,
+      "step": 4395
+    },
+    {
+      "epoch": 10.748166259168704,
+      "grad_norm": 0.3673166958037705,
+      "learning_rate": 9.274453107731732e-07,
+      "loss": 0.7222,
+      "step": 4396
+    },
+    {
+      "epoch": 10.750611246943766,
+      "grad_norm": 0.1851462104213581,
+      "learning_rate": 9.264503111401579e-07,
+      "loss": 0.6613,
+      "step": 4397
+    },
+    {
+      "epoch": 10.753056234718827,
+      "grad_norm": 0.39527664798921636,
+      "learning_rate": 9.254557241278927e-07,
+      "loss": 0.3825,
+      "step": 4398
+    },
+    {
+      "epoch": 10.755501222493887,
+      "grad_norm": 0.20968926398614893,
+      "learning_rate": 9.24461549997179e-07,
+      "loss": 0.4655,
+      "step": 4399
+    },
+    {
+      "epoch": 10.75794621026895,
+      "grad_norm": 0.2664069933205161,
+      "learning_rate": 9.234677890087135e-07,
+      "loss": 0.4979,
+      "step": 4400
+    },
+    {
+      "epoch": 10.76039119804401,
+      "grad_norm": 0.20330502942507978,
+      "learning_rate": 9.224744414230832e-07,
+      "loss": 0.3705,
+      "step": 4401
+    },
+    {
+      "epoch": 10.76283618581907,
+      "grad_norm": 0.34511026553869817,
+      "learning_rate": 9.214815075007655e-07,
+      "loss": 0.4957,
+      "step": 4402
+    },
+    {
+      "epoch": 10.765281173594133,
+      "grad_norm": 0.2422112152014097,
+      "learning_rate": 9.204889875021309e-07,
+      "loss": 0.7323,
+      "step": 4403
+    },
+    {
+      "epoch": 10.767726161369193,
+      "grad_norm": 0.25647415461081396,
+      "learning_rate": 9.194968816874411e-07,
+      "loss": 0.7358,
+      "step": 4404
+    },
+    {
+      "epoch": 10.770171149144254,
+      "grad_norm": 0.3964545172489735,
+      "learning_rate": 9.185051903168485e-07,
+      "loss": 0.5988,
+      "step": 4405
+    },
+    {
+      "epoch": 10.772616136919316,
+      "grad_norm": 0.27539749723367507,
+      "learning_rate": 9.175139136503986e-07,
+      "loss": 0.544,
+      "step": 4406
+    },
+    {
+      "epoch": 10.775061124694377,
+      "grad_norm": 0.2820327413058313,
+      "learning_rate": 9.165230519480248e-07,
+      "loss": 0.5782,
+      "step": 4407
+    },
+    {
+      "epoch": 10.777506112469437,
+      "grad_norm": 0.3138077772670629,
+      "learning_rate": 9.155326054695556e-07,
+      "loss": 0.8416,
+      "step": 4408
+    },
+    {
+      "epoch": 10.7799511002445,
+      "grad_norm": 0.2269564467882219,
+      "learning_rate": 9.145425744747072e-07,
+      "loss": 0.4036,
+      "step": 4409
+    },
+    {
+      "epoch": 10.78239608801956,
+      "grad_norm": 0.24756087356975026,
+      "learning_rate": 9.135529592230891e-07,
+      "loss": 0.5553,
+      "step": 4410
+    },
+    {
+      "epoch": 10.78484107579462,
+      "grad_norm": 0.3236770829473707,
+      "learning_rate": 9.125637599742018e-07,
+      "loss": 0.4281,
+      "step": 4411
+    },
+    {
+      "epoch": 10.787286063569683,
+      "grad_norm": 0.3872565800486209,
+      "learning_rate": 9.115749769874343e-07,
+      "loss": 0.587,
+      "step": 4412
+    },
+    {
+      "epoch": 10.789731051344743,
+      "grad_norm": 0.2324783179340739,
+      "learning_rate": 9.10586610522069e-07,
+      "loss": 0.7913,
+      "step": 4413
+    },
+    {
+      "epoch": 10.792176039119804,
+      "grad_norm": 0.28654113015923355,
+      "learning_rate": 9.095986608372784e-07,
+      "loss": 0.7215,
+      "step": 4414
+    },
+    {
+      "epoch": 10.794621026894866,
+      "grad_norm": 0.2993010275184039,
+      "learning_rate": 9.086111281921253e-07,
+      "loss": 0.6842,
+      "step": 4415
+    },
+    {
+      "epoch": 10.797066014669927,
+      "grad_norm": 0.34360704609554243,
+      "learning_rate": 9.076240128455644e-07,
+      "loss": 0.5557,
+      "step": 4416
+    },
+    {
+      "epoch": 10.799511002444987,
+      "grad_norm": 0.2192671455149053,
+      "learning_rate": 9.066373150564381e-07,
+      "loss": 0.3772,
+      "step": 4417
+    },
+    {
+      "epoch": 10.80195599022005,
+      "grad_norm": 0.20585252900141074,
+      "learning_rate": 9.056510350834829e-07,
+      "loss": 0.491,
+      "step": 4418
+    },
+    {
+      "epoch": 10.80440097799511,
+      "grad_norm": 0.2699705529613797,
+      "learning_rate": 9.046651731853223e-07,
+      "loss": 0.3854,
+      "step": 4419
+    },
+    {
+      "epoch": 10.80684596577017,
+      "grad_norm": 0.16863328327896884,
+      "learning_rate": 9.036797296204725e-07,
+      "loss": 0.3168,
+      "step": 4420
+    },
+    {
+      "epoch": 10.809290953545233,
+      "grad_norm": 0.21309466421484127,
+      "learning_rate": 9.026947046473406e-07,
+      "loss": 0.6027,
+      "step": 4421
+    },
+    {
+      "epoch": 10.811735941320293,
+      "grad_norm": 0.23946532233739712,
+      "learning_rate": 9.017100985242208e-07,
+      "loss": 0.998,
+      "step": 4422
+    },
+    {
+      "epoch": 10.814180929095354,
+      "grad_norm": 0.21603804495006582,
+      "learning_rate": 9.007259115093e-07,
+      "loss": 0.538,
+      "step": 4423
+    },
+    {
+      "epoch": 10.816625916870416,
+      "grad_norm": 0.29546548245082865,
+      "learning_rate": 8.997421438606557e-07,
+      "loss": 0.5188,
+      "step": 4424
+    },
+    {
+      "epoch": 10.819070904645477,
+      "grad_norm": 0.24162164792096535,
+      "learning_rate": 8.987587958362517e-07,
+      "loss": 0.4876,
+      "step": 4425
+    },
+    {
+      "epoch": 10.821515892420537,
+      "grad_norm": 0.25630302694170654,
+      "learning_rate": 8.97775867693948e-07,
+      "loss": 0.442,
+      "step": 4426
+    },
+    {
+      "epoch": 10.8239608801956,
+      "grad_norm": 0.19120876699089995,
+      "learning_rate": 8.967933596914877e-07,
+      "loss": 0.4678,
+      "step": 4427
+    },
+    {
+      "epoch": 10.82640586797066,
+      "grad_norm": 0.3636273796371703,
+      "learning_rate": 8.958112720865089e-07,
+      "loss": 0.4169,
+      "step": 4428
+    },
+    {
+      "epoch": 10.82885085574572,
+      "grad_norm": 0.1749989420365941,
+      "learning_rate": 8.94829605136536e-07,
+      "loss": 0.3598,
+      "step": 4429
+    },
+    {
+      "epoch": 10.831295843520783,
+      "grad_norm": 0.2470284400608675,
+      "learning_rate": 8.938483590989852e-07,
+      "loss": 1.2318,
+      "step": 4430
+    },
+    {
+      "epoch": 10.833740831295843,
+      "grad_norm": 0.29891761018647545,
+      "learning_rate": 8.928675342311624e-07,
+      "loss": 0.5585,
+      "step": 4431
+    },
+    {
+      "epoch": 10.836185819070904,
+      "grad_norm": 0.280447994022004,
+      "learning_rate": 8.918871307902608e-07,
+      "loss": 0.5534,
+      "step": 4432
+    },
+    {
+      "epoch": 10.838630806845966,
+      "grad_norm": 0.21079968755899087,
+      "learning_rate": 8.909071490333652e-07,
+      "loss": 0.636,
+      "step": 4433
+    },
+    {
+      "epoch": 10.841075794621027,
+      "grad_norm": 0.3117880234247859,
+      "learning_rate": 8.899275892174506e-07,
+      "loss": 0.3534,
+      "step": 4434
+    },
+    {
+      "epoch": 10.843520782396087,
+      "grad_norm": 0.22603199067615662,
+      "learning_rate": 8.889484515993776e-07,
+      "loss": 0.6899,
+      "step": 4435
+    },
+    {
+      "epoch": 10.84596577017115,
+      "grad_norm": 0.2302250341189511,
+      "learning_rate": 8.879697364359002e-07,
+      "loss": 0.7505,
+      "step": 4436
+    },
+    {
+      "epoch": 10.84841075794621,
+      "grad_norm": 0.3216128410875158,
+      "learning_rate": 8.869914439836594e-07,
+      "loss": 0.4419,
+      "step": 4437
+    },
+    {
+      "epoch": 10.85085574572127,
+      "grad_norm": 0.22257863716300735,
+      "learning_rate": 8.860135744991866e-07,
+      "loss": 0.5931,
+      "step": 4438
+    },
+    {
+      "epoch": 10.853300733496333,
+      "grad_norm": 0.18158574391719975,
+      "learning_rate": 8.850361282389003e-07,
+      "loss": 0.4167,
+      "step": 4439
+    },
+    {
+      "epoch": 10.855745721271393,
+      "grad_norm": 0.32205543383096275,
+      "learning_rate": 8.840591054591097e-07,
+      "loss": 0.7763,
+      "step": 4440
+    },
+    {
+      "epoch": 10.858190709046454,
+      "grad_norm": 0.29840217919148176,
+      "learning_rate": 8.830825064160137e-07,
+      "loss": 0.5101,
+      "step": 4441
+    },
+    {
+      "epoch": 10.860635696821516,
+      "grad_norm": 0.2756364356076906,
+      "learning_rate": 8.821063313656971e-07,
+      "loss": 0.492,
+      "step": 4442
+    },
+    {
+      "epoch": 10.863080684596577,
+      "grad_norm": 0.30828226502492734,
+      "learning_rate": 8.811305805641362e-07,
+      "loss": 0.4277,
+      "step": 4443
+    },
+    {
+      "epoch": 10.865525672371637,
+      "grad_norm": 0.3405026748914933,
+      "learning_rate": 8.80155254267196e-07,
+      "loss": 0.669,
+      "step": 4444
+    },
+    {
+      "epoch": 10.8679706601467,
+      "grad_norm": 0.2880078286270374,
+      "learning_rate": 8.791803527306278e-07,
+      "loss": 0.6152,
+      "step": 4445
+    },
+    {
+      "epoch": 10.87041564792176,
+      "grad_norm": 0.2572118825561372,
+      "learning_rate": 8.78205876210074e-07,
+      "loss": 0.4644,
+      "step": 4446
+    },
+    {
+      "epoch": 10.87286063569682,
+      "grad_norm": 0.38648626433353456,
+      "learning_rate": 8.772318249610645e-07,
+      "loss": 0.3167,
+      "step": 4447
+    },
+    {
+      "epoch": 10.875305623471883,
+      "grad_norm": 0.21856404549859335,
+      "learning_rate": 8.762581992390187e-07,
+      "loss": 0.8888,
+      "step": 4448
+    },
+    {
+      "epoch": 10.877750611246944,
+      "grad_norm": 0.2625702537584893,
+      "learning_rate": 8.752849992992418e-07,
+      "loss": 0.9432,
+      "step": 4449
+    },
+    {
+      "epoch": 10.880195599022004,
+      "grad_norm": 0.28104352573269825,
+      "learning_rate": 8.743122253969302e-07,
+      "loss": 0.5355,
+      "step": 4450
+    },
+    {
+      "epoch": 10.882640586797066,
+      "grad_norm": 0.2559239747828476,
+      "learning_rate": 8.733398777871679e-07,
+      "loss": 1.0392,
+      "step": 4451
+    },
+    {
+      "epoch": 10.885085574572127,
+      "grad_norm": 0.22730916278844424,
+      "learning_rate": 8.723679567249254e-07,
+      "loss": 0.373,
+      "step": 4452
+    },
+    {
+      "epoch": 10.887530562347187,
+      "grad_norm": 0.20767087684718552,
+      "learning_rate": 8.713964624650633e-07,
+      "loss": 0.7799,
+      "step": 4453
+    },
+    {
+      "epoch": 10.88997555012225,
+      "grad_norm": 0.20657583949687602,
+      "learning_rate": 8.704253952623309e-07,
+      "loss": 0.6948,
+      "step": 4454
+    },
+    {
+      "epoch": 10.89242053789731,
+      "grad_norm": 0.30545276658983533,
+      "learning_rate": 8.69454755371362e-07,
+      "loss": 0.4878,
+      "step": 4455
+    },
+    {
+      "epoch": 10.894865525672373,
+      "grad_norm": 0.15837107102478948,
+      "learning_rate": 8.684845430466824e-07,
+      "loss": 0.4047,
+      "step": 4456
+    },
+    {
+      "epoch": 10.897310513447433,
+      "grad_norm": 0.2648991868962773,
+      "learning_rate": 8.675147585427015e-07,
+      "loss": 0.4929,
+      "step": 4457
+    },
+    {
+      "epoch": 10.899755501222494,
+      "grad_norm": 0.20973692875051925,
+      "learning_rate": 8.665454021137226e-07,
+      "loss": 0.3618,
+      "step": 4458
+    },
+    {
+      "epoch": 10.902200488997554,
+      "grad_norm": 0.26653300908609395,
+      "learning_rate": 8.655764740139302e-07,
+      "loss": 0.6734,
+      "step": 4459
+    },
+    {
+      "epoch": 10.904645476772616,
+      "grad_norm": 0.22430595797584799,
+      "learning_rate": 8.646079744974007e-07,
+      "loss": 0.4768,
+      "step": 4460
+    },
+    {
+      "epoch": 10.907090464547677,
+      "grad_norm": 0.2262324485591408,
+      "learning_rate": 8.636399038180975e-07,
+      "loss": 1.1212,
+      "step": 4461
+    },
+    {
+      "epoch": 10.90953545232274,
+      "grad_norm": 0.23896216109058896,
+      "learning_rate": 8.626722622298692e-07,
+      "loss": 0.9531,
+      "step": 4462
+    },
+    {
+      "epoch": 10.9119804400978,
+      "grad_norm": 0.28232467773131853,
+      "learning_rate": 8.617050499864543e-07,
+      "loss": 0.7083,
+      "step": 4463
+    },
+    {
+      "epoch": 10.91442542787286,
+      "grad_norm": 0.12834222325869268,
+      "learning_rate": 8.607382673414788e-07,
+      "loss": 0.2334,
+      "step": 4464
+    },
+    {
+      "epoch": 10.91687041564792,
+      "grad_norm": 0.3180468946084402,
+      "learning_rate": 8.59771914548454e-07,
+      "loss": 0.4638,
+      "step": 4465
+    },
+    {
+      "epoch": 10.919315403422983,
+      "grad_norm": 0.2450976404145126,
+      "learning_rate": 8.588059918607811e-07,
+      "loss": 0.448,
+      "step": 4466
+    },
+    {
+      "epoch": 10.921760391198044,
+      "grad_norm": 0.17675474003336317,
+      "learning_rate": 8.578404995317452e-07,
+      "loss": 0.7308,
+      "step": 4467
+    },
+    {
+      "epoch": 10.924205378973106,
+      "grad_norm": 0.32324907896928085,
+      "learning_rate": 8.56875437814522e-07,
+      "loss": 0.8168,
+      "step": 4468
+    },
+    {
+      "epoch": 10.926650366748166,
+      "grad_norm": 0.2890846951076616,
+      "learning_rate": 8.559108069621719e-07,
+      "loss": 0.8603,
+      "step": 4469
+    },
+    {
+      "epoch": 10.929095354523227,
+      "grad_norm": 0.28180787028995546,
+      "learning_rate": 8.549466072276441e-07,
+      "loss": 0.3787,
+      "step": 4470
+    },
+    {
+      "epoch": 10.93154034229829,
+      "grad_norm": 0.2699585089295434,
+      "learning_rate": 8.539828388637739e-07,
+      "loss": 0.2415,
+      "step": 4471
+    },
+    {
+      "epoch": 10.93398533007335,
+      "grad_norm": 0.2061375638823393,
+      "learning_rate": 8.530195021232823e-07,
+      "loss": 0.5574,
+      "step": 4472
+    },
+    {
+      "epoch": 10.93643031784841,
+      "grad_norm": 0.22817389604323257,
+      "learning_rate": 8.520565972587788e-07,
+      "loss": 0.3478,
+      "step": 4473
+    },
+    {
+      "epoch": 10.938875305623473,
+      "grad_norm": 0.21006231657407667,
+      "learning_rate": 8.5109412452276e-07,
+      "loss": 0.544,
+      "step": 4474
+    },
+    {
+      "epoch": 10.941320293398533,
+      "grad_norm": 0.21186009207457654,
+      "learning_rate": 8.501320841676067e-07,
+      "loss": 0.4297,
+      "step": 4475
+    },
+    {
+      "epoch": 10.943765281173594,
+      "grad_norm": 0.268039698428206,
+      "learning_rate": 8.491704764455894e-07,
+      "loss": 0.7772,
+      "step": 4476
+    },
+    {
+      "epoch": 10.946210268948656,
+      "grad_norm": 0.3150596163643315,
+      "learning_rate": 8.482093016088622e-07,
+      "loss": 0.8611,
+      "step": 4477
+    },
+    {
+      "epoch": 10.948655256723717,
+      "grad_norm": 0.2261951522324787,
+      "learning_rate": 8.472485599094679e-07,
+      "loss": 0.6551,
+      "step": 4478
+    },
+    {
+      "epoch": 10.951100244498777,
+      "grad_norm": 0.26293016278688186,
+      "learning_rate": 8.462882515993348e-07,
+      "loss": 0.5416,
+      "step": 4479
+    },
+    {
+      "epoch": 10.95354523227384,
+      "grad_norm": 0.2504390444504082,
+      "learning_rate": 8.453283769302778e-07,
+      "loss": 0.7449,
+      "step": 4480
+    },
+    {
+      "epoch": 10.9559902200489,
+      "grad_norm": 0.3306988849884077,
+      "learning_rate": 8.443689361539989e-07,
+      "loss": 0.4165,
+      "step": 4481
+    },
+    {
+      "epoch": 10.95843520782396,
+      "grad_norm": 0.22036648806258632,
+      "learning_rate": 8.434099295220841e-07,
+      "loss": 0.699,
+      "step": 4482
+    },
+    {
+      "epoch": 10.960880195599023,
+      "grad_norm": 0.27981473203785723,
+      "learning_rate": 8.424513572860068e-07,
+      "loss": 0.5788,
+      "step": 4483
+    },
+    {
+      "epoch": 10.963325183374083,
+      "grad_norm": 0.25730042240918105,
+      "learning_rate": 8.414932196971284e-07,
+      "loss": 0.7325,
+      "step": 4484
+    },
+    {
+      "epoch": 10.965770171149144,
+      "grad_norm": 0.2463058403824498,
+      "learning_rate": 8.405355170066925e-07,
+      "loss": 0.7869,
+      "step": 4485
+    },
+    {
+      "epoch": 10.968215158924206,
+      "grad_norm": 0.25286765477335593,
+      "learning_rate": 8.395782494658322e-07,
+      "loss": 0.5196,
+      "step": 4486
+    },
+    {
+      "epoch": 10.970660146699267,
+      "grad_norm": 0.31734230187658513,
+      "learning_rate": 8.386214173255633e-07,
+      "loss": 0.7287,
+      "step": 4487
+    },
+    {
+      "epoch": 10.973105134474327,
+      "grad_norm": 0.17246194453919222,
+      "learning_rate": 8.376650208367901e-07,
+      "loss": 0.8018,
+      "step": 4488
+    },
+    {
+      "epoch": 10.97555012224939,
+      "grad_norm": 0.2407156080368671,
+      "learning_rate": 8.367090602503016e-07,
+      "loss": 0.5047,
+      "step": 4489
+    },
+    {
+      "epoch": 10.97799511002445,
+      "grad_norm": 0.2922035398633921,
+      "learning_rate": 8.357535358167723e-07,
+      "loss": 0.4366,
+      "step": 4490
+    },
+    {
+      "epoch": 10.98044009779951,
+      "grad_norm": 0.2437845624651476,
+      "learning_rate": 8.347984477867638e-07,
+      "loss": 0.726,
+      "step": 4491
+    },
+    {
+      "epoch": 10.982885085574573,
+      "grad_norm": 0.3905562708560972,
+      "learning_rate": 8.338437964107201e-07,
+      "loss": 0.4928,
+      "step": 4492
+    },
+    {
+      "epoch": 10.985330073349633,
+      "grad_norm": 0.310411980402308,
+      "learning_rate": 8.328895819389737e-07,
+      "loss": 0.627,
+      "step": 4493
+    },
+    {
+      "epoch": 10.987775061124694,
+      "grad_norm": 0.27887099268850135,
+      "learning_rate": 8.319358046217421e-07,
+      "loss": 0.8326,
+      "step": 4494
+    },
+    {
+      "epoch": 10.990220048899756,
+      "grad_norm": 0.35830875990279687,
+      "learning_rate": 8.309824647091261e-07,
+      "loss": 0.4339,
+      "step": 4495
+    },
+    {
+      "epoch": 10.992665036674817,
+      "grad_norm": 0.34317206265544303,
+      "learning_rate": 8.300295624511146e-07,
+      "loss": 0.8138,
+      "step": 4496
+    },
+    {
+      "epoch": 10.995110024449877,
+      "grad_norm": 0.24199401731588846,
+      "learning_rate": 8.29077098097579e-07,
+      "loss": 0.7366,
+      "step": 4497
+    },
+    {
+      "epoch": 10.99755501222494,
+      "grad_norm": 0.3603846066956584,
+      "learning_rate": 8.281250718982778e-07,
+      "loss": 0.5646,
+      "step": 4498
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.19075556512986527,
+      "learning_rate": 8.271734841028553e-07,
+      "loss": 0.4423,
+      "step": 4499
+    },
+    {
+      "epoch": 11.00244498777506,
+      "grad_norm": 0.1831902131501335,
+      "learning_rate": 8.262223349608367e-07,
+      "loss": 0.871,
+      "step": 4500
+    },
+    {
+      "epoch": 11.004889975550123,
+      "grad_norm": 0.22648632453837284,
+      "learning_rate": 8.252716247216389e-07,
+      "loss": 0.4024,
+      "step": 4501
+    },
+    {
+      "epoch": 11.007334963325183,
+      "grad_norm": 0.2518849476096333,
+      "learning_rate": 8.243213536345568e-07,
+      "loss": 0.49,
+      "step": 4502
+    },
+    {
+      "epoch": 11.009779951100244,
+      "grad_norm": 0.27280179848624325,
+      "learning_rate": 8.233715219487745e-07,
+      "loss": 0.7326,
+      "step": 4503
+    },
+    {
+      "epoch": 11.012224938875306,
+      "grad_norm": 0.22351049791665253,
+      "learning_rate": 8.2242212991336e-07,
+      "loss": 0.4238,
+      "step": 4504
+    },
+    {
+      "epoch": 11.014669926650367,
+      "grad_norm": 0.31340815027961766,
+      "learning_rate": 8.214731777772647e-07,
+      "loss": 0.4889,
+      "step": 4505
+    },
+    {
+      "epoch": 11.017114914425427,
+      "grad_norm": 0.26237060656233774,
+      "learning_rate": 8.205246657893265e-07,
+      "loss": 0.368,
+      "step": 4506
+    },
+    {
+      "epoch": 11.01955990220049,
+      "grad_norm": 0.3124421927923915,
+      "learning_rate": 8.195765941982656e-07,
+      "loss": 0.4683,
+      "step": 4507
+    },
+    {
+      "epoch": 11.02200488997555,
+      "grad_norm": 0.3269462688986921,
+      "learning_rate": 8.186289632526889e-07,
+      "loss": 0.611,
+      "step": 4508
+    },
+    {
+      "epoch": 11.02444987775061,
+      "grad_norm": 0.19219501516140977,
+      "learning_rate": 8.17681773201088e-07,
+      "loss": 0.5032,
+      "step": 4509
+    },
+    {
+      "epoch": 11.026894865525673,
+      "grad_norm": 0.3722903018568003,
+      "learning_rate": 8.167350242918356e-07,
+      "loss": 0.6307,
+      "step": 4510
+    },
+    {
+      "epoch": 11.029339853300733,
+      "grad_norm": 0.24390355541875292,
+      "learning_rate": 8.157887167731923e-07,
+      "loss": 0.5244,
+      "step": 4511
+    },
+    {
+      "epoch": 11.031784841075794,
+      "grad_norm": 0.21709289368777013,
+      "learning_rate": 8.148428508933012e-07,
+      "loss": 0.4133,
+      "step": 4512
+    },
+    {
+      "epoch": 11.034229828850856,
+      "grad_norm": 0.25212275623418967,
+      "learning_rate": 8.138974269001901e-07,
+      "loss": 0.3299,
+      "step": 4513
+    },
+    {
+      "epoch": 11.036674816625917,
+      "grad_norm": 0.3200662022443012,
+      "learning_rate": 8.129524450417719e-07,
+      "loss": 0.4433,
+      "step": 4514
+    },
+    {
+      "epoch": 11.039119804400977,
+      "grad_norm": 0.26501731998419176,
+      "learning_rate": 8.120079055658403e-07,
+      "loss": 0.7023,
+      "step": 4515
+    },
+    {
+      "epoch": 11.04156479217604,
+      "grad_norm": 0.27023465279150916,
+      "learning_rate": 8.110638087200773e-07,
+      "loss": 0.5309,
+      "step": 4516
+    },
+    {
+      "epoch": 11.0440097799511,
+      "grad_norm": 0.27427600503907607,
+      "learning_rate": 8.101201547520451e-07,
+      "loss": 0.4105,
+      "step": 4517
+    },
+    {
+      "epoch": 11.04645476772616,
+      "grad_norm": 0.4488457267980421,
+      "learning_rate": 8.091769439091917e-07,
+      "loss": 0.6137,
+      "step": 4518
+    },
+    {
+      "epoch": 11.048899755501223,
+      "grad_norm": 0.29174034757312417,
+      "learning_rate": 8.082341764388499e-07,
+      "loss": 0.6363,
+      "step": 4519
+    },
+    {
+      "epoch": 11.051344743276283,
+      "grad_norm": 0.32632110113196966,
+      "learning_rate": 8.072918525882328e-07,
+      "loss": 0.5493,
+      "step": 4520
+    },
+    {
+      "epoch": 11.053789731051344,
+      "grad_norm": 0.2498051767537811,
+      "learning_rate": 8.063499726044405e-07,
+      "loss": 0.5249,
+      "step": 4521
+    },
+    {
+      "epoch": 11.056234718826406,
+      "grad_norm": 0.21472642349457857,
+      "learning_rate": 8.054085367344553e-07,
+      "loss": 0.7333,
+      "step": 4522
+    },
+    {
+      "epoch": 11.058679706601467,
+      "grad_norm": 0.2802799571310624,
+      "learning_rate": 8.044675452251433e-07,
+      "loss": 0.4129,
+      "step": 4523
+    },
+    {
+      "epoch": 11.061124694376527,
+      "grad_norm": 0.20538076074997694,
+      "learning_rate": 8.035269983232547e-07,
+      "loss": 0.6974,
+      "step": 4524
+    },
+    {
+      "epoch": 11.06356968215159,
+      "grad_norm": 0.23606579013577947,
+      "learning_rate": 8.02586896275421e-07,
+      "loss": 0.6632,
+      "step": 4525
+    },
+    {
+      "epoch": 11.06601466992665,
+      "grad_norm": 0.23537633148482667,
+      "learning_rate": 8.016472393281602e-07,
+      "loss": 0.7241,
+      "step": 4526
+    },
+    {
+      "epoch": 11.06845965770171,
+      "grad_norm": 0.24498023749857345,
+      "learning_rate": 8.007080277278697e-07,
+      "loss": 0.4311,
+      "step": 4527
+    },
+    {
+      "epoch": 11.070904645476773,
+      "grad_norm": 0.19327076895010875,
+      "learning_rate": 7.997692617208339e-07,
+      "loss": 0.4271,
+      "step": 4528
+    },
+    {
+      "epoch": 11.073349633251834,
+      "grad_norm": 0.32176051550500834,
+      "learning_rate": 7.988309415532186e-07,
+      "loss": 0.3714,
+      "step": 4529
+    },
+    {
+      "epoch": 11.075794621026894,
+      "grad_norm": 0.2934864243299337,
+      "learning_rate": 7.978930674710719e-07,
+      "loss": 0.7543,
+      "step": 4530
+    },
+    {
+      "epoch": 11.078239608801956,
+      "grad_norm": 0.39792587179515393,
+      "learning_rate": 7.969556397203265e-07,
+      "loss": 0.3317,
+      "step": 4531
+    },
+    {
+      "epoch": 11.080684596577017,
+      "grad_norm": 0.2581968232529534,
+      "learning_rate": 7.960186585467975e-07,
+      "loss": 0.4091,
+      "step": 4532
+    },
+    {
+      "epoch": 11.083129584352077,
+      "grad_norm": 0.27645371154528486,
+      "learning_rate": 7.950821241961825e-07,
+      "loss": 0.4343,
+      "step": 4533
+    },
+    {
+      "epoch": 11.08557457212714,
+      "grad_norm": 0.2635637708911598,
+      "learning_rate": 7.941460369140633e-07,
+      "loss": 0.5715,
+      "step": 4534
+    },
+    {
+      "epoch": 11.0880195599022,
+      "grad_norm": 0.1580199215370195,
+      "learning_rate": 7.932103969459018e-07,
+      "loss": 0.3829,
+      "step": 4535
+    },
+    {
+      "epoch": 11.09046454767726,
+      "grad_norm": 0.3242491876617342,
+      "learning_rate": 7.922752045370457e-07,
+      "loss": 0.4256,
+      "step": 4536
+    },
+    {
+      "epoch": 11.092909535452323,
+      "grad_norm": 0.311899012830726,
+      "learning_rate": 7.913404599327224e-07,
+      "loss": 0.4203,
+      "step": 4537
+    },
+    {
+      "epoch": 11.095354523227384,
+      "grad_norm": 0.28490314705409886,
+      "learning_rate": 7.904061633780438e-07,
+      "loss": 0.3744,
+      "step": 4538
+    },
+    {
+      "epoch": 11.097799511002446,
+      "grad_norm": 0.2663349059105018,
+      "learning_rate": 7.894723151180053e-07,
+      "loss": 0.6135,
+      "step": 4539
+    },
+    {
+      "epoch": 11.100244498777506,
+      "grad_norm": 0.189176489976191,
+      "learning_rate": 7.88538915397481e-07,
+      "loss": 0.5604,
+      "step": 4540
+    },
+    {
+      "epoch": 11.102689486552567,
+      "grad_norm": 0.4524506732992623,
+      "learning_rate": 7.876059644612307e-07,
+      "loss": 0.3515,
+      "step": 4541
+    },
+    {
+      "epoch": 11.10513447432763,
+      "grad_norm": 0.1709543263532299,
+      "learning_rate": 7.866734625538966e-07,
+      "loss": 0.6789,
+      "step": 4542
+    },
+    {
+      "epoch": 11.10757946210269,
+      "grad_norm": 0.18420078170250767,
+      "learning_rate": 7.857414099199995e-07,
+      "loss": 0.5494,
+      "step": 4543
+    },
+    {
+      "epoch": 11.11002444987775,
+      "grad_norm": 0.2640621849292725,
+      "learning_rate": 7.848098068039478e-07,
+      "loss": 0.5343,
+      "step": 4544
+    },
+    {
+      "epoch": 11.112469437652813,
+      "grad_norm": 0.2338697513582221,
+      "learning_rate": 7.83878653450027e-07,
+      "loss": 0.5428,
+      "step": 4545
+    },
+    {
+      "epoch": 11.114914425427873,
+      "grad_norm": 0.2335547888613717,
+      "learning_rate": 7.829479501024088e-07,
+      "loss": 0.8721,
+      "step": 4546
+    },
+    {
+      "epoch": 11.117359413202934,
+      "grad_norm": 0.1675602926231691,
+      "learning_rate": 7.820176970051424e-07,
+      "loss": 0.5443,
+      "step": 4547
+    },
+    {
+      "epoch": 11.119804400977996,
+      "grad_norm": 0.23251292590474068,
+      "learning_rate": 7.810878944021629e-07,
+      "loss": 0.3919,
+      "step": 4548
+    },
+    {
+      "epoch": 11.122249388753056,
+      "grad_norm": 0.27223448323065885,
+      "learning_rate": 7.801585425372868e-07,
+      "loss": 0.4393,
+      "step": 4549
+    },
+    {
+      "epoch": 11.124694376528117,
+      "grad_norm": 0.27283738069206576,
+      "learning_rate": 7.792296416542094e-07,
+      "loss": 0.5953,
+      "step": 4550
+    },
+    {
+      "epoch": 11.12713936430318,
+      "grad_norm": 0.20335304915968613,
+      "learning_rate": 7.783011919965109e-07,
+      "loss": 0.4991,
+      "step": 4551
+    },
+    {
+      "epoch": 11.12958435207824,
+      "grad_norm": 0.4183843136386102,
+      "learning_rate": 7.773731938076526e-07,
+      "loss": 0.4945,
+      "step": 4552
+    },
+    {
+      "epoch": 11.1320293398533,
+      "grad_norm": 0.2660458440355159,
+      "learning_rate": 7.764456473309745e-07,
+      "loss": 0.6157,
+      "step": 4553
+    },
+    {
+      "epoch": 11.134474327628363,
+      "grad_norm": 0.3907633924130284,
+      "learning_rate": 7.75518552809704e-07,
+      "loss": 0.3833,
+      "step": 4554
+    },
+    {
+      "epoch": 11.136919315403423,
+      "grad_norm": 0.19566180334377198,
+      "learning_rate": 7.745919104869437e-07,
+      "loss": 0.7381,
+      "step": 4555
+    },
+    {
+      "epoch": 11.139364303178484,
+      "grad_norm": 0.22549769504219344,
+      "learning_rate": 7.736657206056822e-07,
+      "loss": 0.6883,
+      "step": 4556
+    },
+    {
+      "epoch": 11.141809290953546,
+      "grad_norm": 0.2620485949303981,
+      "learning_rate": 7.727399834087862e-07,
+      "loss": 0.6712,
+      "step": 4557
+    },
+    {
+      "epoch": 11.144254278728607,
+      "grad_norm": 0.3061620301986605,
+      "learning_rate": 7.718146991390061e-07,
+      "loss": 0.4059,
+      "step": 4558
+    },
+    {
+      "epoch": 11.146699266503667,
+      "grad_norm": 0.2795302981015312,
+      "learning_rate": 7.708898680389728e-07,
+      "loss": 0.4525,
+      "step": 4559
+    },
+    {
+      "epoch": 11.14914425427873,
+      "grad_norm": 0.3161021901526408,
+      "learning_rate": 7.699654903511971e-07,
+      "loss": 0.4968,
+      "step": 4560
+    },
+    {
+      "epoch": 11.15158924205379,
+      "grad_norm": 0.2590750742302967,
+      "learning_rate": 7.690415663180728e-07,
+      "loss": 0.3749,
+      "step": 4561
+    },
+    {
+      "epoch": 11.15403422982885,
+      "grad_norm": 0.3045676287388493,
+      "learning_rate": 7.681180961818746e-07,
+      "loss": 0.6718,
+      "step": 4562
+    },
+    {
+      "epoch": 11.156479217603913,
+      "grad_norm": 0.18804819154730004,
+      "learning_rate": 7.671950801847558e-07,
+      "loss": 0.4959,
+      "step": 4563
+    },
+    {
+      "epoch": 11.158924205378973,
+      "grad_norm": 0.41019432574228276,
+      "learning_rate": 7.662725185687536e-07,
+      "loss": 0.8298,
+      "step": 4564
+    },
+    {
+      "epoch": 11.161369193154034,
+      "grad_norm": 0.1746370030024984,
+      "learning_rate": 7.65350411575784e-07,
+      "loss": 0.3971,
+      "step": 4565
+    },
+    {
+      "epoch": 11.163814180929096,
+      "grad_norm": 0.2810986144694296,
+      "learning_rate": 7.64428759447646e-07,
+      "loss": 0.5553,
+      "step": 4566
+    },
+    {
+      "epoch": 11.166259168704157,
+      "grad_norm": 0.19058700712436713,
+      "learning_rate": 7.63507562426016e-07,
+      "loss": 0.8493,
+      "step": 4567
+    },
+    {
+      "epoch": 11.168704156479217,
+      "grad_norm": 0.21990978648688894,
+      "learning_rate": 7.625868207524539e-07,
+      "loss": 0.8012,
+      "step": 4568
+    },
+    {
+      "epoch": 11.17114914425428,
+      "grad_norm": 0.3712391035753783,
+      "learning_rate": 7.616665346683999e-07,
+      "loss": 0.4756,
+      "step": 4569
+    },
+    {
+      "epoch": 11.17359413202934,
+      "grad_norm": 0.21110902137737841,
+      "learning_rate": 7.607467044151726e-07,
+      "loss": 0.5216,
+      "step": 4570
+    },
+    {
+      "epoch": 11.1760391198044,
+      "grad_norm": 0.2148089662719569,
+      "learning_rate": 7.598273302339731e-07,
+      "loss": 0.45,
+      "step": 4571
+    },
+    {
+      "epoch": 11.178484107579463,
+      "grad_norm": 0.18759823932888134,
+      "learning_rate": 7.589084123658836e-07,
+      "loss": 0.4838,
+      "step": 4572
+    },
+    {
+      "epoch": 11.180929095354523,
+      "grad_norm": 0.23198150183659141,
+      "learning_rate": 7.579899510518635e-07,
+      "loss": 0.4661,
+      "step": 4573
+    },
+    {
+      "epoch": 11.183374083129584,
+      "grad_norm": 0.3157374918821066,
+      "learning_rate": 7.570719465327559e-07,
+      "loss": 0.4142,
+      "step": 4574
+    },
+    {
+      "epoch": 11.185819070904646,
+      "grad_norm": 0.287021901746644,
+      "learning_rate": 7.561543990492803e-07,
+      "loss": 0.36,
+      "step": 4575
+    },
+    {
+      "epoch": 11.188264058679707,
+      "grad_norm": 0.3097922937251926,
+      "learning_rate": 7.55237308842042e-07,
+      "loss": 0.5866,
+      "step": 4576
+    },
+    {
+      "epoch": 11.190709046454767,
+      "grad_norm": 0.27634681812380657,
+      "learning_rate": 7.543206761515204e-07,
+      "loss": 0.4172,
+      "step": 4577
+    },
+    {
+      "epoch": 11.19315403422983,
+      "grad_norm": 0.22754321989399273,
+      "learning_rate": 7.534045012180785e-07,
+      "loss": 0.3399,
+      "step": 4578
+    },
+    {
+      "epoch": 11.19559902200489,
+      "grad_norm": 0.36061130376252126,
+      "learning_rate": 7.524887842819592e-07,
+      "loss": 0.6007,
+      "step": 4579
+    },
+    {
+      "epoch": 11.19804400977995,
+      "grad_norm": 0.25900358161474,
+      "learning_rate": 7.515735255832829e-07,
+      "loss": 0.5833,
+      "step": 4580
+    },
+    {
+      "epoch": 11.200488997555013,
+      "grad_norm": 0.2408380938861335,
+      "learning_rate": 7.50658725362052e-07,
+      "loss": 0.3851,
+      "step": 4581
+    },
+    {
+      "epoch": 11.202933985330073,
+      "grad_norm": 0.1824988661988685,
+      "learning_rate": 7.497443838581489e-07,
+      "loss": 0.4622,
+      "step": 4582
+    },
+    {
+      "epoch": 11.205378973105134,
+      "grad_norm": 0.22641269223992344,
+      "learning_rate": 7.488305013113331e-07,
+      "loss": 0.4177,
+      "step": 4583
+    },
+    {
+      "epoch": 11.207823960880196,
+      "grad_norm": 0.366831924966499,
+      "learning_rate": 7.479170779612477e-07,
+      "loss": 0.5357,
+      "step": 4584
+    },
+    {
+      "epoch": 11.210268948655257,
+      "grad_norm": 0.23934598266249307,
+      "learning_rate": 7.47004114047411e-07,
+      "loss": 0.451,
+      "step": 4585
+    },
+    {
+      "epoch": 11.212713936430317,
+      "grad_norm": 0.3614203778800607,
+      "learning_rate": 7.460916098092241e-07,
+      "loss": 0.6694,
+      "step": 4586
+    },
+    {
+      "epoch": 11.21515892420538,
+      "grad_norm": 0.23226587465727236,
+      "learning_rate": 7.451795654859664e-07,
+      "loss": 0.6513,
+      "step": 4587
+    },
+    {
+      "epoch": 11.21760391198044,
+      "grad_norm": 0.4845579284545479,
+      "learning_rate": 7.442679813167969e-07,
+      "loss": 0.5452,
+      "step": 4588
+    },
+    {
+      "epoch": 11.2200488997555,
+      "grad_norm": 0.4496858535445958,
+      "learning_rate": 7.433568575407546e-07,
+      "loss": 0.8544,
+      "step": 4589
+    },
+    {
+      "epoch": 11.222493887530563,
+      "grad_norm": 0.2972638589664434,
+      "learning_rate": 7.424461943967557e-07,
+      "loss": 0.7148,
+      "step": 4590
+    },
+    {
+      "epoch": 11.224938875305623,
+      "grad_norm": 0.3927954481194539,
+      "learning_rate": 7.415359921235971e-07,
+      "loss": 0.5038,
+      "step": 4591
+    },
+    {
+      "epoch": 11.227383863080684,
+      "grad_norm": 0.37161589837202824,
+      "learning_rate": 7.40626250959956e-07,
+      "loss": 0.5381,
+      "step": 4592
+    },
+    {
+      "epoch": 11.229828850855746,
+      "grad_norm": 0.22740660239650776,
+      "learning_rate": 7.397169711443853e-07,
+      "loss": 0.3818,
+      "step": 4593
+    },
+    {
+      "epoch": 11.232273838630807,
+      "grad_norm": 0.2667679299181016,
+      "learning_rate": 7.38808152915321e-07,
+      "loss": 0.8382,
+      "step": 4594
+    },
+    {
+      "epoch": 11.234718826405867,
+      "grad_norm": 0.16583011643121837,
+      "learning_rate": 7.378997965110743e-07,
+      "loss": 0.6774,
+      "step": 4595
+    },
+    {
+      "epoch": 11.23716381418093,
+      "grad_norm": 0.3230631245366838,
+      "learning_rate": 7.36991902169838e-07,
+      "loss": 0.71,
+      "step": 4596
+    },
+    {
+      "epoch": 11.23960880195599,
+      "grad_norm": 0.2062874812711294,
+      "learning_rate": 7.360844701296824e-07,
+      "loss": 0.8562,
+      "step": 4597
+    },
+    {
+      "epoch": 11.24205378973105,
+      "grad_norm": 0.23546310601048928,
+      "learning_rate": 7.351775006285572e-07,
+      "loss": 0.5319,
+      "step": 4598
+    },
+    {
+      "epoch": 11.244498777506113,
+      "grad_norm": 0.2568574051148602,
+      "learning_rate": 7.342709939042913e-07,
+      "loss": 0.4844,
+      "step": 4599
+    },
+    {
+      "epoch": 11.246943765281173,
+      "grad_norm": 0.2543677213404928,
+      "learning_rate": 7.333649501945902e-07,
+      "loss": 1.1877,
+      "step": 4600
+    },
+    {
+      "epoch": 11.249388753056234,
+      "grad_norm": 0.22554058228087212,
+      "learning_rate": 7.324593697370397e-07,
+      "loss": 0.4854,
+      "step": 4601
+    },
+    {
+      "epoch": 11.251833740831296,
+      "grad_norm": 0.19813204554163338,
+      "learning_rate": 7.315542527691047e-07,
+      "loss": 0.6132,
+      "step": 4602
+    },
+    {
+      "epoch": 11.254278728606357,
+      "grad_norm": 0.24707319146079448,
+      "learning_rate": 7.306495995281256e-07,
+      "loss": 0.5403,
+      "step": 4603
+    },
+    {
+      "epoch": 11.256723716381417,
+      "grad_norm": 0.2126566510122896,
+      "learning_rate": 7.297454102513257e-07,
+      "loss": 0.6464,
+      "step": 4604
+    },
+    {
+      "epoch": 11.25916870415648,
+      "grad_norm": 0.23557256285424386,
+      "learning_rate": 7.288416851758018e-07,
+      "loss": 0.4614,
+      "step": 4605
+    },
+    {
+      "epoch": 11.26161369193154,
+      "grad_norm": 0.25545575532357595,
+      "learning_rate": 7.279384245385321e-07,
+      "loss": 1.0356,
+      "step": 4606
+    },
+    {
+      "epoch": 11.2640586797066,
+      "grad_norm": 0.2558023841087448,
+      "learning_rate": 7.270356285763724e-07,
+      "loss": 0.6308,
+      "step": 4607
+    },
+    {
+      "epoch": 11.266503667481663,
+      "grad_norm": 0.3241763710407807,
+      "learning_rate": 7.261332975260568e-07,
+      "loss": 0.7364,
+      "step": 4608
+    },
+    {
+      "epoch": 11.268948655256724,
+      "grad_norm": 0.3152978795533111,
+      "learning_rate": 7.252314316241974e-07,
+      "loss": 0.4471,
+      "step": 4609
+    },
+    {
+      "epoch": 11.271393643031784,
+      "grad_norm": 0.34502368141596973,
+      "learning_rate": 7.243300311072826e-07,
+      "loss": 0.6018,
+      "step": 4610
+    },
+    {
+      "epoch": 11.273838630806846,
+      "grad_norm": 0.27893116550623065,
+      "learning_rate": 7.234290962116814e-07,
+      "loss": 0.3969,
+      "step": 4611
+    },
+    {
+      "epoch": 11.276283618581907,
+      "grad_norm": 0.22216832200851147,
+      "learning_rate": 7.225286271736398e-07,
+      "loss": 0.4138,
+      "step": 4612
+    },
+    {
+      "epoch": 11.278728606356967,
+      "grad_norm": 0.33614711576825074,
+      "learning_rate": 7.216286242292806e-07,
+      "loss": 0.4885,
+      "step": 4613
+    },
+    {
+      "epoch": 11.28117359413203,
+      "grad_norm": 0.40238486488548003,
+      "learning_rate": 7.207290876146058e-07,
+      "loss": 0.5064,
+      "step": 4614
+    },
+    {
+      "epoch": 11.28361858190709,
+      "grad_norm": 0.28448362392298293,
+      "learning_rate": 7.198300175654935e-07,
+      "loss": 0.9145,
+      "step": 4615
+    },
+    {
+      "epoch": 11.28606356968215,
+      "grad_norm": 0.18925727140961157,
+      "learning_rate": 7.189314143177012e-07,
+      "loss": 0.5401,
+      "step": 4616
+    },
+    {
+      "epoch": 11.288508557457213,
+      "grad_norm": 0.2836482339860002,
+      "learning_rate": 7.180332781068639e-07,
+      "loss": 1.3362,
+      "step": 4617
+    },
+    {
+      "epoch": 11.290953545232274,
+      "grad_norm": 0.2548240971359618,
+      "learning_rate": 7.171356091684911e-07,
+      "loss": 0.3998,
+      "step": 4618
+    },
+    {
+      "epoch": 11.293398533007334,
+      "grad_norm": 0.21889221573432732,
+      "learning_rate": 7.162384077379756e-07,
+      "loss": 0.8033,
+      "step": 4619
+    },
+    {
+      "epoch": 11.295843520782396,
+      "grad_norm": 0.23813587884859602,
+      "learning_rate": 7.153416740505815e-07,
+      "loss": 0.4348,
+      "step": 4620
+    },
+    {
+      "epoch": 11.298288508557457,
+      "grad_norm": 0.20288273627423753,
+      "learning_rate": 7.144454083414534e-07,
+      "loss": 0.4553,
+      "step": 4621
+    },
+    {
+      "epoch": 11.30073349633252,
+      "grad_norm": 0.16320536097192173,
+      "learning_rate": 7.135496108456142e-07,
+      "loss": 0.6769,
+      "step": 4622
+    },
+    {
+      "epoch": 11.30317848410758,
+      "grad_norm": 0.2541503652407237,
+      "learning_rate": 7.126542817979601e-07,
+      "loss": 0.403,
+      "step": 4623
+    },
+    {
+      "epoch": 11.30562347188264,
+      "grad_norm": 0.337534172544725,
+      "learning_rate": 7.117594214332693e-07,
+      "loss": 0.3935,
+      "step": 4624
+    },
+    {
+      "epoch": 11.3080684596577,
+      "grad_norm": 0.3431093853038867,
+      "learning_rate": 7.108650299861924e-07,
+      "loss": 0.3915,
+      "step": 4625
+    },
+    {
+      "epoch": 11.310513447432763,
+      "grad_norm": 0.22907634597480162,
+      "learning_rate": 7.099711076912605e-07,
+      "loss": 0.4055,
+      "step": 4626
+    },
+    {
+      "epoch": 11.312958435207824,
+      "grad_norm": 0.29437520128522804,
+      "learning_rate": 7.090776547828812e-07,
+      "loss": 0.4124,
+      "step": 4627
+    },
+    {
+      "epoch": 11.315403422982886,
+      "grad_norm": 0.23962019086723044,
+      "learning_rate": 7.081846714953366e-07,
+      "loss": 0.9317,
+      "step": 4628
+    },
+    {
+      "epoch": 11.317848410757946,
+      "grad_norm": 0.2455589771839322,
+      "learning_rate": 7.072921580627884e-07,
+      "loss": 0.519,
+      "step": 4629
+    },
+    {
+      "epoch": 11.320293398533007,
+      "grad_norm": 0.2324623865475341,
+      "learning_rate": 7.064001147192737e-07,
+      "loss": 0.4346,
+      "step": 4630
+    },
+    {
+      "epoch": 11.32273838630807,
+      "grad_norm": 0.21736370872577693,
+      "learning_rate": 7.05508541698707e-07,
+      "loss": 0.9281,
+      "step": 4631
+    },
+    {
+      "epoch": 11.32518337408313,
+      "grad_norm": 0.21464955540580807,
+      "learning_rate": 7.046174392348798e-07,
+      "loss": 0.4252,
+      "step": 4632
+    },
+    {
+      "epoch": 11.32762836185819,
+      "grad_norm": 0.4279262118760031,
+      "learning_rate": 7.037268075614579e-07,
+      "loss": 0.3621,
+      "step": 4633
+    },
+    {
+      "epoch": 11.330073349633253,
+      "grad_norm": 0.21827093917612006,
+      "learning_rate": 7.028366469119872e-07,
+      "loss": 0.5207,
+      "step": 4634
+    },
+    {
+      "epoch": 11.332518337408313,
+      "grad_norm": 0.2302413733245278,
+      "learning_rate": 7.019469575198861e-07,
+      "loss": 0.4252,
+      "step": 4635
+    },
+    {
+      "epoch": 11.334963325183374,
+      "grad_norm": 0.22742970545840108,
+      "learning_rate": 7.010577396184529e-07,
+      "loss": 0.4324,
+      "step": 4636
+    },
+    {
+      "epoch": 11.337408312958436,
+      "grad_norm": 0.22898035712822712,
+      "learning_rate": 7.001689934408612e-07,
+      "loss": 0.7259,
+      "step": 4637
+    },
+    {
+      "epoch": 11.339853300733497,
+      "grad_norm": 0.3442097042119856,
+      "learning_rate": 6.992807192201595e-07,
+      "loss": 0.4214,
+      "step": 4638
+    },
+    {
+      "epoch": 11.342298288508557,
+      "grad_norm": 0.24909817448397711,
+      "learning_rate": 6.98392917189274e-07,
+      "loss": 0.7714,
+      "step": 4639
+    },
+    {
+      "epoch": 11.34474327628362,
+      "grad_norm": 0.3062824989924266,
+      "learning_rate": 6.975055875810066e-07,
+      "loss": 0.3721,
+      "step": 4640
+    },
+    {
+      "epoch": 11.34718826405868,
+      "grad_norm": 0.296249605204892,
+      "learning_rate": 6.966187306280362e-07,
+      "loss": 1.3165,
+      "step": 4641
+    },
+    {
+      "epoch": 11.34963325183374,
+      "grad_norm": 0.2004623556421978,
+      "learning_rate": 6.95732346562917e-07,
+      "loss": 0.511,
+      "step": 4642
+    },
+    {
+      "epoch": 11.352078239608803,
+      "grad_norm": 0.24809297969950717,
+      "learning_rate": 6.948464356180778e-07,
+      "loss": 0.445,
+      "step": 4643
+    },
+    {
+      "epoch": 11.354523227383863,
+      "grad_norm": 0.2776184966045638,
+      "learning_rate": 6.939609980258266e-07,
+      "loss": 0.9459,
+      "step": 4644
+    },
+    {
+      "epoch": 11.356968215158924,
+      "grad_norm": 0.2947671130421909,
+      "learning_rate": 6.930760340183434e-07,
+      "loss": 0.407,
+      "step": 4645
+    },
+    {
+      "epoch": 11.359413202933986,
+      "grad_norm": 0.19982728558159243,
+      "learning_rate": 6.921915438276874e-07,
+      "loss": 0.4048,
+      "step": 4646
+    },
+    {
+      "epoch": 11.361858190709047,
+      "grad_norm": 0.36442212008201647,
+      "learning_rate": 6.913075276857922e-07,
+      "loss": 0.438,
+      "step": 4647
+    },
+    {
+      "epoch": 11.364303178484107,
+      "grad_norm": 0.21591582410523855,
+      "learning_rate": 6.904239858244663e-07,
+      "loss": 0.9395,
+      "step": 4648
+    },
+    {
+      "epoch": 11.36674816625917,
+      "grad_norm": 0.27053122103327965,
+      "learning_rate": 6.895409184753948e-07,
+      "loss": 0.3216,
+      "step": 4649
+    },
+    {
+      "epoch": 11.36919315403423,
+      "grad_norm": 0.24705880093966814,
+      "learning_rate": 6.886583258701382e-07,
+      "loss": 0.4687,
+      "step": 4650
+    },
+    {
+      "epoch": 11.37163814180929,
+      "grad_norm": 0.24431842255738964,
+      "learning_rate": 6.877762082401329e-07,
+      "loss": 0.5262,
+      "step": 4651
+    },
+    {
+      "epoch": 11.374083129584353,
+      "grad_norm": 0.16442323277314677,
+      "learning_rate": 6.868945658166907e-07,
+      "loss": 0.2299,
+      "step": 4652
+    },
+    {
+      "epoch": 11.376528117359413,
+      "grad_norm": 0.2774048376497834,
+      "learning_rate": 6.860133988309972e-07,
+      "loss": 0.3688,
+      "step": 4653
+    },
+    {
+      "epoch": 11.378973105134474,
+      "grad_norm": 0.3216572885423358,
+      "learning_rate": 6.851327075141156e-07,
+      "loss": 0.3753,
+      "step": 4654
+    },
+    {
+      "epoch": 11.381418092909536,
+      "grad_norm": 0.3619002507760992,
+      "learning_rate": 6.842524920969823e-07,
+      "loss": 0.3418,
+      "step": 4655
+    },
+    {
+      "epoch": 11.383863080684597,
+      "grad_norm": 0.2097951453983877,
+      "learning_rate": 6.833727528104106e-07,
+      "loss": 0.3837,
+      "step": 4656
+    },
+    {
+      "epoch": 11.386308068459657,
+      "grad_norm": 0.30306011045090875,
+      "learning_rate": 6.824934898850888e-07,
+      "loss": 0.3994,
+      "step": 4657
+    },
+    {
+      "epoch": 11.38875305623472,
+      "grad_norm": 0.2832519211499832,
+      "learning_rate": 6.816147035515786e-07,
+      "loss": 0.7829,
+      "step": 4658
+    },
+    {
+      "epoch": 11.39119804400978,
+      "grad_norm": 0.3410385734261241,
+      "learning_rate": 6.807363940403183e-07,
+      "loss": 1.1079,
+      "step": 4659
+    },
+    {
+      "epoch": 11.39364303178484,
+      "grad_norm": 0.2533995062679686,
+      "learning_rate": 6.798585615816217e-07,
+      "loss": 0.5504,
+      "step": 4660
+    },
+    {
+      "epoch": 11.396088019559903,
+      "grad_norm": 0.24589591089168408,
+      "learning_rate": 6.789812064056745e-07,
+      "loss": 0.7333,
+      "step": 4661
+    },
+    {
+      "epoch": 11.398533007334963,
+      "grad_norm": 0.30631469737648204,
+      "learning_rate": 6.781043287425417e-07,
+      "loss": 0.4507,
+      "step": 4662
+    },
+    {
+      "epoch": 11.400977995110024,
+      "grad_norm": 0.35631632759489473,
+      "learning_rate": 6.772279288221589e-07,
+      "loss": 0.5311,
+      "step": 4663
+    },
+    {
+      "epoch": 11.403422982885086,
+      "grad_norm": 0.20801036673709464,
+      "learning_rate": 6.763520068743399e-07,
+      "loss": 0.3681,
+      "step": 4664
+    },
+    {
+      "epoch": 11.405867970660147,
+      "grad_norm": 0.43311386587404976,
+      "learning_rate": 6.754765631287696e-07,
+      "loss": 0.6555,
+      "step": 4665
+    },
+    {
+      "epoch": 11.408312958435207,
+      "grad_norm": 0.31063428953038436,
+      "learning_rate": 6.746015978150103e-07,
+      "loss": 0.7819,
+      "step": 4666
+    },
+    {
+      "epoch": 11.41075794621027,
+      "grad_norm": 0.26905037438559565,
+      "learning_rate": 6.737271111624988e-07,
+      "loss": 0.7104,
+      "step": 4667
+    },
+    {
+      "epoch": 11.41320293398533,
+      "grad_norm": 0.20222588616903778,
+      "learning_rate": 6.728531034005436e-07,
+      "loss": 0.6638,
+      "step": 4668
+    },
+    {
+      "epoch": 11.41564792176039,
+      "grad_norm": 0.35169031899850667,
+      "learning_rate": 6.719795747583307e-07,
+      "loss": 0.6985,
+      "step": 4669
+    },
+    {
+      "epoch": 11.418092909535453,
+      "grad_norm": 0.30234314040246196,
+      "learning_rate": 6.711065254649202e-07,
+      "loss": 0.4908,
+      "step": 4670
+    },
+    {
+      "epoch": 11.420537897310513,
+      "grad_norm": 0.3508136750093273,
+      "learning_rate": 6.702339557492427e-07,
+      "loss": 0.6284,
+      "step": 4671
+    },
+    {
+      "epoch": 11.422982885085574,
+      "grad_norm": 0.22186617079664653,
+      "learning_rate": 6.693618658401097e-07,
+      "loss": 0.4307,
+      "step": 4672
+    },
+    {
+      "epoch": 11.425427872860636,
+      "grad_norm": 0.1945600963069221,
+      "learning_rate": 6.684902559662002e-07,
+      "loss": 0.4137,
+      "step": 4673
+    },
+    {
+      "epoch": 11.427872860635697,
+      "grad_norm": 0.35684660982244654,
+      "learning_rate": 6.676191263560722e-07,
+      "loss": 0.8241,
+      "step": 4674
+    },
+    {
+      "epoch": 11.430317848410757,
+      "grad_norm": 0.21395545755212267,
+      "learning_rate": 6.667484772381544e-07,
+      "loss": 0.8457,
+      "step": 4675
+    },
+    {
+      "epoch": 11.43276283618582,
+      "grad_norm": 0.3403706276315063,
+      "learning_rate": 6.658783088407511e-07,
+      "loss": 0.5911,
+      "step": 4676
+    },
+    {
+      "epoch": 11.43520782396088,
+      "grad_norm": 0.255455419115446,
+      "learning_rate": 6.65008621392042e-07,
+      "loss": 0.5204,
+      "step": 4677
+    },
+    {
+      "epoch": 11.43765281173594,
+      "grad_norm": 0.1782976606358615,
+      "learning_rate": 6.641394151200767e-07,
+      "loss": 0.4158,
+      "step": 4678
+    },
+    {
+      "epoch": 11.440097799511003,
+      "grad_norm": 0.3358488155050002,
+      "learning_rate": 6.632706902527824e-07,
+      "loss": 0.4971,
+      "step": 4679
+    },
+    {
+      "epoch": 11.442542787286063,
+      "grad_norm": 0.36234215206696213,
+      "learning_rate": 6.624024470179591e-07,
+      "loss": 0.5798,
+      "step": 4680
+    },
+    {
+      "epoch": 11.444987775061124,
+      "grad_norm": 0.21868132752420108,
+      "learning_rate": 6.615346856432783e-07,
+      "loss": 0.5659,
+      "step": 4681
+    },
+    {
+      "epoch": 11.447432762836186,
+      "grad_norm": 0.28577584058753497,
+      "learning_rate": 6.606674063562882e-07,
+      "loss": 0.4,
+      "step": 4682
+    },
+    {
+      "epoch": 11.449877750611247,
+      "grad_norm": 0.32667491174364294,
+      "learning_rate": 6.598006093844086e-07,
+      "loss": 0.5238,
+      "step": 4683
+    },
+    {
+      "epoch": 11.452322738386307,
+      "grad_norm": 0.2761867026202508,
+      "learning_rate": 6.589342949549343e-07,
+      "loss": 0.3843,
+      "step": 4684
+    },
+    {
+      "epoch": 11.45476772616137,
+      "grad_norm": 0.35971044627052867,
+      "learning_rate": 6.58068463295033e-07,
+      "loss": 0.5836,
+      "step": 4685
+    },
+    {
+      "epoch": 11.45721271393643,
+      "grad_norm": 0.2441430549653803,
+      "learning_rate": 6.572031146317443e-07,
+      "loss": 0.4677,
+      "step": 4686
+    },
+    {
+      "epoch": 11.45965770171149,
+      "grad_norm": 0.2723809117908462,
+      "learning_rate": 6.563382491919837e-07,
+      "loss": 0.8788,
+      "step": 4687
+    },
+    {
+      "epoch": 11.462102689486553,
+      "grad_norm": 0.24840652022406123,
+      "learning_rate": 6.554738672025374e-07,
+      "loss": 0.379,
+      "step": 4688
+    },
+    {
+      "epoch": 11.464547677261614,
+      "grad_norm": 0.2684965237373586,
+      "learning_rate": 6.546099688900668e-07,
+      "loss": 0.4071,
+      "step": 4689
+    },
+    {
+      "epoch": 11.466992665036674,
+      "grad_norm": 0.2924512040150079,
+      "learning_rate": 6.537465544811064e-07,
+      "loss": 0.7088,
+      "step": 4690
+    },
+    {
+      "epoch": 11.469437652811736,
+      "grad_norm": 0.2759319722698551,
+      "learning_rate": 6.528836242020623e-07,
+      "loss": 0.4256,
+      "step": 4691
+    },
+    {
+      "epoch": 11.471882640586797,
+      "grad_norm": 0.22524713931410068,
+      "learning_rate": 6.520211782792154e-07,
+      "loss": 0.4561,
+      "step": 4692
+    },
+    {
+      "epoch": 11.474327628361857,
+      "grad_norm": 0.33794632080087766,
+      "learning_rate": 6.51159216938717e-07,
+      "loss": 0.6215,
+      "step": 4693
+    },
+    {
+      "epoch": 11.47677261613692,
+      "grad_norm": 0.24424887888280425,
+      "learning_rate": 6.502977404065952e-07,
+      "loss": 0.5103,
+      "step": 4694
+    },
+    {
+      "epoch": 11.47921760391198,
+      "grad_norm": 0.26569260831678365,
+      "learning_rate": 6.494367489087489e-07,
+      "loss": 0.7476,
+      "step": 4695
+    },
+    {
+      "epoch": 11.48166259168704,
+      "grad_norm": 0.2080644996543301,
+      "learning_rate": 6.48576242670948e-07,
+      "loss": 0.7307,
+      "step": 4696
+    },
+    {
+      "epoch": 11.484107579462103,
+      "grad_norm": 0.25665579223775065,
+      "learning_rate": 6.477162219188385e-07,
+      "loss": 0.6652,
+      "step": 4697
+    },
+    {
+      "epoch": 11.486552567237164,
+      "grad_norm": 0.23934934641447048,
+      "learning_rate": 6.468566868779366e-07,
+      "loss": 0.3041,
+      "step": 4698
+    },
+    {
+      "epoch": 11.488997555012224,
+      "grad_norm": 0.30332472227421503,
+      "learning_rate": 6.459976377736321e-07,
+      "loss": 0.3868,
+      "step": 4699
+    },
+    {
+      "epoch": 11.491442542787286,
+      "grad_norm": 0.19257547416021592,
+      "learning_rate": 6.451390748311884e-07,
+      "loss": 0.4148,
+      "step": 4700
+    },
+    {
+      "epoch": 11.493887530562347,
+      "grad_norm": 0.33737341908192064,
+      "learning_rate": 6.442809982757389e-07,
+      "loss": 0.4345,
+      "step": 4701
+    },
+    {
+      "epoch": 11.496332518337407,
+      "grad_norm": 0.19277165576829927,
+      "learning_rate": 6.434234083322924e-07,
+      "loss": 0.5699,
+      "step": 4702
+    },
+    {
+      "epoch": 11.49877750611247,
+      "grad_norm": 0.343635346378364,
+      "learning_rate": 6.425663052257269e-07,
+      "loss": 0.7501,
+      "step": 4703
+    },
+    {
+      "epoch": 11.50122249388753,
+      "grad_norm": 0.2581899058828507,
+      "learning_rate": 6.417096891807948e-07,
+      "loss": 0.703,
+      "step": 4704
+    },
+    {
+      "epoch": 11.503667481662593,
+      "grad_norm": 0.20641806504531549,
+      "learning_rate": 6.408535604221222e-07,
+      "loss": 0.6638,
+      "step": 4705
+    },
+    {
+      "epoch": 11.506112469437653,
+      "grad_norm": 0.24382034768639388,
+      "learning_rate": 6.39997919174204e-07,
+      "loss": 0.4442,
+      "step": 4706
+    },
+    {
+      "epoch": 11.508557457212714,
+      "grad_norm": 0.3992288951871595,
+      "learning_rate": 6.391427656614099e-07,
+      "loss": 0.4068,
+      "step": 4707
+    },
+    {
+      "epoch": 11.511002444987774,
+      "grad_norm": 0.34332395298503443,
+      "learning_rate": 6.382881001079797e-07,
+      "loss": 0.3232,
+      "step": 4708
+    },
+    {
+      "epoch": 11.513447432762836,
+      "grad_norm": 0.30555615997850294,
+      "learning_rate": 6.374339227380266e-07,
+      "loss": 1.1958,
+      "step": 4709
+    },
+    {
+      "epoch": 11.515892420537897,
+      "grad_norm": 0.2774514108268633,
+      "learning_rate": 6.365802337755364e-07,
+      "loss": 0.3904,
+      "step": 4710
+    },
+    {
+      "epoch": 11.51833740831296,
+      "grad_norm": 0.2361460885277631,
+      "learning_rate": 6.357270334443646e-07,
+      "loss": 1.021,
+      "step": 4711
+    },
+    {
+      "epoch": 11.52078239608802,
+      "grad_norm": 0.2137875360675618,
+      "learning_rate": 6.34874321968241e-07,
+      "loss": 0.6281,
+      "step": 4712
+    },
+    {
+      "epoch": 11.52322738386308,
+      "grad_norm": 0.20344562746111486,
+      "learning_rate": 6.340220995707647e-07,
+      "loss": 0.4476,
+      "step": 4713
+    },
+    {
+      "epoch": 11.525672371638143,
+      "grad_norm": 0.2024586152036081,
+      "learning_rate": 6.331703664754079e-07,
+      "loss": 0.3174,
+      "step": 4714
+    },
+    {
+      "epoch": 11.528117359413203,
+      "grad_norm": 0.30410610014554074,
+      "learning_rate": 6.323191229055167e-07,
+      "loss": 0.5601,
+      "step": 4715
+    },
+    {
+      "epoch": 11.530562347188264,
+      "grad_norm": 0.3075859124888534,
+      "learning_rate": 6.314683690843049e-07,
+      "loss": 0.472,
+      "step": 4716
+    },
+    {
+      "epoch": 11.533007334963326,
+      "grad_norm": 0.29315517799447915,
+      "learning_rate": 6.306181052348603e-07,
+      "loss": 0.6899,
+      "step": 4717
+    },
+    {
+      "epoch": 11.535452322738386,
+      "grad_norm": 0.24745814130760596,
+      "learning_rate": 6.297683315801406e-07,
+      "loss": 0.8657,
+      "step": 4718
+    },
+    {
+      "epoch": 11.537897310513447,
+      "grad_norm": 0.3148533671283106,
+      "learning_rate": 6.289190483429763e-07,
+      "loss": 1.0306,
+      "step": 4719
+    },
+    {
+      "epoch": 11.54034229828851,
+      "grad_norm": 0.24841765765600288,
+      "learning_rate": 6.280702557460702e-07,
+      "loss": 0.428,
+      "step": 4720
+    },
+    {
+      "epoch": 11.54278728606357,
+      "grad_norm": 0.2936211577818367,
+      "learning_rate": 6.272219540119934e-07,
+      "loss": 0.8074,
+      "step": 4721
+    },
+    {
+      "epoch": 11.54523227383863,
+      "grad_norm": 0.3959761879911899,
+      "learning_rate": 6.263741433631912e-07,
+      "loss": 0.3624,
+      "step": 4722
+    },
+    {
+      "epoch": 11.547677261613693,
+      "grad_norm": 0.22297246706541055,
+      "learning_rate": 6.255268240219777e-07,
+      "loss": 0.3807,
+      "step": 4723
+    },
+    {
+      "epoch": 11.550122249388753,
+      "grad_norm": 0.20407099823515282,
+      "learning_rate": 6.246799962105402e-07,
+      "loss": 0.6452,
+      "step": 4724
+    },
+    {
+      "epoch": 11.552567237163814,
+      "grad_norm": 0.3351603541579998,
+      "learning_rate": 6.238336601509365e-07,
+      "loss": 1.0374,
+      "step": 4725
+    },
+    {
+      "epoch": 11.555012224938876,
+      "grad_norm": 0.2793961725256349,
+      "learning_rate": 6.229878160650951e-07,
+      "loss": 0.8236,
+      "step": 4726
+    },
+    {
+      "epoch": 11.557457212713937,
+      "grad_norm": 0.270311613639359,
+      "learning_rate": 6.221424641748164e-07,
+      "loss": 0.6993,
+      "step": 4727
+    },
+    {
+      "epoch": 11.559902200488997,
+      "grad_norm": 0.28054294301474886,
+      "learning_rate": 6.212976047017693e-07,
+      "loss": 0.6052,
+      "step": 4728
+    },
+    {
+      "epoch": 11.56234718826406,
+      "grad_norm": 0.21911666361189305,
+      "learning_rate": 6.204532378674969e-07,
+      "loss": 0.4869,
+      "step": 4729
+    },
+    {
+      "epoch": 11.56479217603912,
+      "grad_norm": 0.21644075770739324,
+      "learning_rate": 6.196093638934111e-07,
+      "loss": 0.4392,
+      "step": 4730
+    },
+    {
+      "epoch": 11.56723716381418,
+      "grad_norm": 0.22882539990886566,
+      "learning_rate": 6.187659830007945e-07,
+      "loss": 0.4234,
+      "step": 4731
+    },
+    {
+      "epoch": 11.569682151589243,
+      "grad_norm": 0.21863418496022324,
+      "learning_rate": 6.179230954108015e-07,
+      "loss": 0.6179,
+      "step": 4732
+    },
+    {
+      "epoch": 11.572127139364303,
+      "grad_norm": 0.19647658553054734,
+      "learning_rate": 6.170807013444558e-07,
+      "loss": 0.6399,
+      "step": 4733
+    },
+    {
+      "epoch": 11.574572127139364,
+      "grad_norm": 0.34102751466775194,
+      "learning_rate": 6.162388010226525e-07,
+      "loss": 0.7511,
+      "step": 4734
+    },
+    {
+      "epoch": 11.577017114914426,
+      "grad_norm": 0.2980400506514369,
+      "learning_rate": 6.153973946661584e-07,
+      "loss": 0.5991,
+      "step": 4735
+    },
+    {
+      "epoch": 11.579462102689487,
+      "grad_norm": 0.316686584047014,
+      "learning_rate": 6.145564824956071e-07,
+      "loss": 1.173,
+      "step": 4736
+    },
+    {
+      "epoch": 11.581907090464547,
+      "grad_norm": 0.191132278577256,
+      "learning_rate": 6.13716064731508e-07,
+      "loss": 0.3644,
+      "step": 4737
+    },
+    {
+      "epoch": 11.58435207823961,
+      "grad_norm": 0.25096145767759465,
+      "learning_rate": 6.128761415942353e-07,
+      "loss": 0.6492,
+      "step": 4738
+    },
+    {
+      "epoch": 11.58679706601467,
+      "grad_norm": 0.2940977956718163,
+      "learning_rate": 6.120367133040373e-07,
+      "loss": 0.3848,
+      "step": 4739
+    },
+    {
+      "epoch": 11.58924205378973,
+      "grad_norm": 0.3474400954389469,
+      "learning_rate": 6.111977800810318e-07,
+      "loss": 0.3744,
+      "step": 4740
+    },
+    {
+      "epoch": 11.591687041564793,
+      "grad_norm": 0.25241867220980474,
+      "learning_rate": 6.103593421452051e-07,
+      "loss": 0.8455,
+      "step": 4741
+    },
+    {
+      "epoch": 11.594132029339853,
+      "grad_norm": 0.2587972321140265,
+      "learning_rate": 6.095213997164151e-07,
+      "loss": 0.7991,
+      "step": 4742
+    },
+    {
+      "epoch": 11.596577017114914,
+      "grad_norm": 0.28798146351949966,
+      "learning_rate": 6.086839530143907e-07,
+      "loss": 0.8349,
+      "step": 4743
+    },
+    {
+      "epoch": 11.599022004889976,
+      "grad_norm": 0.15825453467876194,
+      "learning_rate": 6.078470022587277e-07,
+      "loss": 0.4509,
+      "step": 4744
+    },
+    {
+      "epoch": 11.601466992665037,
+      "grad_norm": 0.2879051956179973,
+      "learning_rate": 6.070105476688954e-07,
+      "loss": 0.5089,
+      "step": 4745
+    },
+    {
+      "epoch": 11.603911980440097,
+      "grad_norm": 0.25053076704622995,
+      "learning_rate": 6.061745894642298e-07,
+      "loss": 0.5871,
+      "step": 4746
+    },
+    {
+      "epoch": 11.60635696821516,
+      "grad_norm": 0.2720979781311066,
+      "learning_rate": 6.053391278639395e-07,
+      "loss": 0.4148,
+      "step": 4747
+    },
+    {
+      "epoch": 11.60880195599022,
+      "grad_norm": 0.16004316352268858,
+      "learning_rate": 6.04504163087101e-07,
+      "loss": 0.3253,
+      "step": 4748
+    },
+    {
+      "epoch": 11.61124694376528,
+      "grad_norm": 0.16391112069988056,
+      "learning_rate": 6.036696953526613e-07,
+      "loss": 0.3976,
+      "step": 4749
+    },
+    {
+      "epoch": 11.613691931540343,
+      "grad_norm": 0.24942136166954784,
+      "learning_rate": 6.028357248794381e-07,
+      "loss": 0.6889,
+      "step": 4750
+    },
+    {
+      "epoch": 11.616136919315403,
+      "grad_norm": 0.33673629652739545,
+      "learning_rate": 6.020022518861154e-07,
+      "loss": 0.6074,
+      "step": 4751
+    },
+    {
+      "epoch": 11.618581907090464,
+      "grad_norm": 0.28953643553549907,
+      "learning_rate": 6.011692765912503e-07,
+      "loss": 0.4941,
+      "step": 4752
+    },
+    {
+      "epoch": 11.621026894865526,
+      "grad_norm": 0.32859882309217103,
+      "learning_rate": 6.003367992132684e-07,
+      "loss": 0.4491,
+      "step": 4753
+    },
+    {
+      "epoch": 11.623471882640587,
+      "grad_norm": 0.28923243667055515,
+      "learning_rate": 5.995048199704629e-07,
+      "loss": 0.6522,
+      "step": 4754
+    },
+    {
+      "epoch": 11.625916870415647,
+      "grad_norm": 0.28878511620955377,
+      "learning_rate": 5.986733390809993e-07,
+      "loss": 0.5271,
+      "step": 4755
+    },
+    {
+      "epoch": 11.62836185819071,
+      "grad_norm": 0.30909941772357113,
+      "learning_rate": 5.978423567629094e-07,
+      "loss": 0.4535,
+      "step": 4756
+    },
+    {
+      "epoch": 11.63080684596577,
+      "grad_norm": 0.2820366569868769,
+      "learning_rate": 5.970118732340965e-07,
+      "loss": 0.4451,
+      "step": 4757
+    },
+    {
+      "epoch": 11.63325183374083,
+      "grad_norm": 0.37431597036083847,
+      "learning_rate": 5.961818887123324e-07,
+      "loss": 0.3818,
+      "step": 4758
+    },
+    {
+      "epoch": 11.635696821515893,
+      "grad_norm": 0.2587991279452256,
+      "learning_rate": 5.953524034152583e-07,
+      "loss": 0.8937,
+      "step": 4759
+    },
+    {
+      "epoch": 11.638141809290953,
+      "grad_norm": 0.39980339103324153,
+      "learning_rate": 5.945234175603845e-07,
+      "loss": 0.542,
+      "step": 4760
+    },
+    {
+      "epoch": 11.640586797066014,
+      "grad_norm": 0.29578140274044096,
+      "learning_rate": 5.936949313650888e-07,
+      "loss": 0.5412,
+      "step": 4761
+    },
+    {
+      "epoch": 11.643031784841076,
+      "grad_norm": 0.2630602395048049,
+      "learning_rate": 5.928669450466204e-07,
+      "loss": 0.7017,
+      "step": 4762
+    },
+    {
+      "epoch": 11.645476772616137,
+      "grad_norm": 0.25776778687360297,
+      "learning_rate": 5.920394588220962e-07,
+      "loss": 0.3736,
+      "step": 4763
+    },
+    {
+      "epoch": 11.647921760391197,
+      "grad_norm": 0.22098979537874971,
+      "learning_rate": 5.912124729085012e-07,
+      "loss": 0.848,
+      "step": 4764
+    },
+    {
+      "epoch": 11.65036674816626,
+      "grad_norm": 0.23471283258667477,
+      "learning_rate": 5.903859875226914e-07,
+      "loss": 0.4224,
+      "step": 4765
+    },
+    {
+      "epoch": 11.65281173594132,
+      "grad_norm": 0.24627112474030158,
+      "learning_rate": 5.895600028813883e-07,
+      "loss": 0.8884,
+      "step": 4766
+    },
+    {
+      "epoch": 11.65525672371638,
+      "grad_norm": 0.3231945069244839,
+      "learning_rate": 5.887345192011854e-07,
+      "loss": 0.9674,
+      "step": 4767
+    },
+    {
+      "epoch": 11.657701711491443,
+      "grad_norm": 0.28419382779795216,
+      "learning_rate": 5.879095366985427e-07,
+      "loss": 0.4535,
+      "step": 4768
+    },
+    {
+      "epoch": 11.660146699266503,
+      "grad_norm": 0.29642344504321216,
+      "learning_rate": 5.870850555897901e-07,
+      "loss": 0.5306,
+      "step": 4769
+    },
+    {
+      "epoch": 11.662591687041564,
+      "grad_norm": 0.34104366660124297,
+      "learning_rate": 5.862610760911259e-07,
+      "loss": 0.4235,
+      "step": 4770
+    },
+    {
+      "epoch": 11.665036674816626,
+      "grad_norm": 0.3018306727180531,
+      "learning_rate": 5.854375984186144e-07,
+      "loss": 0.5957,
+      "step": 4771
+    },
+    {
+      "epoch": 11.667481662591687,
+      "grad_norm": 0.26606144339349935,
+      "learning_rate": 5.846146227881922e-07,
+      "loss": 0.9826,
+      "step": 4772
+    },
+    {
+      "epoch": 11.669926650366747,
+      "grad_norm": 0.24320805515066607,
+      "learning_rate": 5.837921494156621e-07,
+      "loss": 0.3986,
+      "step": 4773
+    },
+    {
+      "epoch": 11.67237163814181,
+      "grad_norm": 0.25254769082832673,
+      "learning_rate": 5.829701785166944e-07,
+      "loss": 0.4094,
+      "step": 4774
+    },
+    {
+      "epoch": 11.67481662591687,
+      "grad_norm": 0.28233453752064297,
+      "learning_rate": 5.821487103068299e-07,
+      "loss": 0.6972,
+      "step": 4775
+    },
+    {
+      "epoch": 11.67726161369193,
+      "grad_norm": 0.2085014090624554,
+      "learning_rate": 5.813277450014753e-07,
+      "loss": 0.621,
+      "step": 4776
+    },
+    {
+      "epoch": 11.679706601466993,
+      "grad_norm": 0.2643448475375848,
+      "learning_rate": 5.80507282815907e-07,
+      "loss": 0.3342,
+      "step": 4777
+    },
+    {
+      "epoch": 11.682151589242054,
+      "grad_norm": 0.25277522450542955,
+      "learning_rate": 5.796873239652695e-07,
+      "loss": 0.4812,
+      "step": 4778
+    },
+    {
+      "epoch": 11.684596577017114,
+      "grad_norm": 0.3323776488522753,
+      "learning_rate": 5.788678686645729e-07,
+      "loss": 0.4583,
+      "step": 4779
+    },
+    {
+      "epoch": 11.687041564792176,
+      "grad_norm": 0.22873508873163986,
+      "learning_rate": 5.780489171286999e-07,
+      "loss": 0.8051,
+      "step": 4780
+    },
+    {
+      "epoch": 11.689486552567237,
+      "grad_norm": 0.21825315811171078,
+      "learning_rate": 5.772304695723962e-07,
+      "loss": 0.7503,
+      "step": 4781
+    },
+    {
+      "epoch": 11.6919315403423,
+      "grad_norm": 0.26811151089103613,
+      "learning_rate": 5.764125262102782e-07,
+      "loss": 0.4823,
+      "step": 4782
+    },
+    {
+      "epoch": 11.69437652811736,
+      "grad_norm": 0.2798585541690888,
+      "learning_rate": 5.7559508725683e-07,
+      "loss": 0.9482,
+      "step": 4783
+    },
+    {
+      "epoch": 11.69682151589242,
+      "grad_norm": 0.29651023612598965,
+      "learning_rate": 5.747781529264016e-07,
+      "loss": 0.4462,
+      "step": 4784
+    },
+    {
+      "epoch": 11.69926650366748,
+      "grad_norm": 0.3061182296606879,
+      "learning_rate": 5.739617234332132e-07,
+      "loss": 0.4017,
+      "step": 4785
+    },
+    {
+      "epoch": 11.701711491442543,
+      "grad_norm": 0.23505361088022259,
+      "learning_rate": 5.731457989913497e-07,
+      "loss": 0.7383,
+      "step": 4786
+    },
+    {
+      "epoch": 11.704156479217604,
+      "grad_norm": 0.2510445089502352,
+      "learning_rate": 5.723303798147662e-07,
+      "loss": 0.5046,
+      "step": 4787
+    },
+    {
+      "epoch": 11.706601466992666,
+      "grad_norm": 0.26416495690564706,
+      "learning_rate": 5.715154661172845e-07,
+      "loss": 0.6689,
+      "step": 4788
+    },
+    {
+      "epoch": 11.709046454767726,
+      "grad_norm": 0.2989582467744812,
+      "learning_rate": 5.707010581125921e-07,
+      "loss": 0.4811,
+      "step": 4789
+    },
+    {
+      "epoch": 11.711491442542787,
+      "grad_norm": 0.24730358018933668,
+      "learning_rate": 5.69887156014248e-07,
+      "loss": 0.7859,
+      "step": 4790
+    },
+    {
+      "epoch": 11.713936430317847,
+      "grad_norm": 0.2535765138389413,
+      "learning_rate": 5.690737600356735e-07,
+      "loss": 0.7102,
+      "step": 4791
+    },
+    {
+      "epoch": 11.71638141809291,
+      "grad_norm": 0.20786411735110066,
+      "learning_rate": 5.68260870390161e-07,
+      "loss": 0.5841,
+      "step": 4792
+    },
+    {
+      "epoch": 11.71882640586797,
+      "grad_norm": 0.33021637082921645,
+      "learning_rate": 5.67448487290869e-07,
+      "loss": 0.74,
+      "step": 4793
+    },
+    {
+      "epoch": 11.721271393643033,
+      "grad_norm": 0.3230352463070714,
+      "learning_rate": 5.666366109508217e-07,
+      "loss": 0.944,
+      "step": 4794
+    },
+    {
+      "epoch": 11.723716381418093,
+      "grad_norm": 0.32951892122308185,
+      "learning_rate": 5.65825241582913e-07,
+      "loss": 0.6878,
+      "step": 4795
+    },
+    {
+      "epoch": 11.726161369193154,
+      "grad_norm": 0.2660983270359221,
+      "learning_rate": 5.650143793999014e-07,
+      "loss": 0.8063,
+      "step": 4796
+    },
+    {
+      "epoch": 11.728606356968216,
+      "grad_norm": 0.21419916204887143,
+      "learning_rate": 5.642040246144137e-07,
+      "loss": 0.4945,
+      "step": 4797
+    },
+    {
+      "epoch": 11.731051344743276,
+      "grad_norm": 0.2159669160446207,
+      "learning_rate": 5.633941774389446e-07,
+      "loss": 0.3959,
+      "step": 4798
+    },
+    {
+      "epoch": 11.733496332518337,
+      "grad_norm": 0.308496561972242,
+      "learning_rate": 5.625848380858528e-07,
+      "loss": 0.7344,
+      "step": 4799
+    },
+    {
+      "epoch": 11.7359413202934,
+      "grad_norm": 0.3036914977884245,
+      "learning_rate": 5.617760067673667e-07,
+      "loss": 0.3874,
+      "step": 4800
+    },
+    {
+      "epoch": 11.73838630806846,
+      "grad_norm": 0.3093700342797868,
+      "learning_rate": 5.609676836955802e-07,
+      "loss": 0.577,
+      "step": 4801
+    },
+    {
+      "epoch": 11.74083129584352,
+      "grad_norm": 0.29741496865470646,
+      "learning_rate": 5.601598690824542e-07,
+      "loss": 0.506,
+      "step": 4802
+    },
+    {
+      "epoch": 11.743276283618583,
+      "grad_norm": 0.21422546636304476,
+      "learning_rate": 5.593525631398166e-07,
+      "loss": 0.713,
+      "step": 4803
+    },
+    {
+      "epoch": 11.745721271393643,
+      "grad_norm": 0.27688218833680933,
+      "learning_rate": 5.585457660793606e-07,
+      "loss": 0.5534,
+      "step": 4804
+    },
+    {
+      "epoch": 11.748166259168704,
+      "grad_norm": 0.25090102609950776,
+      "learning_rate": 5.57739478112648e-07,
+      "loss": 0.4965,
+      "step": 4805
+    },
+    {
+      "epoch": 11.750611246943766,
+      "grad_norm": 0.34169634769330615,
+      "learning_rate": 5.569336994511043e-07,
+      "loss": 0.3478,
+      "step": 4806
+    },
+    {
+      "epoch": 11.753056234718827,
+      "grad_norm": 0.2391864794593845,
+      "learning_rate": 5.561284303060241e-07,
+      "loss": 0.6459,
+      "step": 4807
+    },
+    {
+      "epoch": 11.755501222493887,
+      "grad_norm": 0.17319199062869503,
+      "learning_rate": 5.553236708885682e-07,
+      "loss": 0.5093,
+      "step": 4808
+    },
+    {
+      "epoch": 11.75794621026895,
+      "grad_norm": 0.22751971750924863,
+      "learning_rate": 5.545194214097613e-07,
+      "loss": 0.6816,
+      "step": 4809
+    },
+    {
+      "epoch": 11.76039119804401,
+      "grad_norm": 0.22590029508442067,
+      "learning_rate": 5.537156820804967e-07,
+      "loss": 0.4387,
+      "step": 4810
+    },
+    {
+      "epoch": 11.76283618581907,
+      "grad_norm": 0.3519924473774395,
+      "learning_rate": 5.529124531115335e-07,
+      "loss": 0.7404,
+      "step": 4811
+    },
+    {
+      "epoch": 11.765281173594133,
+      "grad_norm": 0.27736442192831967,
+      "learning_rate": 5.521097347134965e-07,
+      "loss": 0.3935,
+      "step": 4812
+    },
+    {
+      "epoch": 11.767726161369193,
+      "grad_norm": 0.49202588078919873,
+      "learning_rate": 5.513075270968779e-07,
+      "loss": 0.5001,
+      "step": 4813
+    },
+    {
+      "epoch": 11.770171149144254,
+      "grad_norm": 0.2547540689193078,
+      "learning_rate": 5.505058304720329e-07,
+      "loss": 0.5766,
+      "step": 4814
+    },
+    {
+      "epoch": 11.772616136919316,
+      "grad_norm": 0.4144701496806104,
+      "learning_rate": 5.497046450491866e-07,
+      "loss": 0.6988,
+      "step": 4815
+    },
+    {
+      "epoch": 11.775061124694377,
+      "grad_norm": 0.21293997537949624,
+      "learning_rate": 5.489039710384264e-07,
+      "loss": 0.6032,
+      "step": 4816
+    },
+    {
+      "epoch": 11.777506112469437,
+      "grad_norm": 0.2870500018532013,
+      "learning_rate": 5.481038086497084e-07,
+      "loss": 0.4799,
+      "step": 4817
+    },
+    {
+      "epoch": 11.7799511002445,
+      "grad_norm": 0.17725150164693032,
+      "learning_rate": 5.473041580928543e-07,
+      "loss": 0.6955,
+      "step": 4818
+    },
+    {
+      "epoch": 11.78239608801956,
+      "grad_norm": 0.2820840681507961,
+      "learning_rate": 5.465050195775488e-07,
+      "loss": 0.9522,
+      "step": 4819
+    },
+    {
+      "epoch": 11.78484107579462,
+      "grad_norm": 0.2731751228113001,
+      "learning_rate": 5.457063933133455e-07,
+      "loss": 0.9021,
+      "step": 4820
+    },
+    {
+      "epoch": 11.787286063569683,
+      "grad_norm": 0.3451576462579562,
+      "learning_rate": 5.449082795096627e-07,
+      "loss": 0.4912,
+      "step": 4821
+    },
+    {
+      "epoch": 11.789731051344743,
+      "grad_norm": 0.20933647964075164,
+      "learning_rate": 5.441106783757827e-07,
+      "loss": 0.4948,
+      "step": 4822
+    },
+    {
+      "epoch": 11.792176039119804,
+      "grad_norm": 0.1720602283934634,
+      "learning_rate": 5.433135901208572e-07,
+      "loss": 0.3986,
+      "step": 4823
+    },
+    {
+      "epoch": 11.794621026894866,
+      "grad_norm": 0.41036927141509577,
+      "learning_rate": 5.425170149538986e-07,
+      "loss": 0.6739,
+      "step": 4824
+    },
+    {
+      "epoch": 11.797066014669927,
+      "grad_norm": 0.232071814916651,
+      "learning_rate": 5.417209530837892e-07,
+      "loss": 0.5197,
+      "step": 4825
+    },
+    {
+      "epoch": 11.799511002444987,
+      "grad_norm": 0.28839594399022617,
+      "learning_rate": 5.409254047192724e-07,
+      "loss": 0.5995,
+      "step": 4826
+    },
+    {
+      "epoch": 11.80195599022005,
+      "grad_norm": 0.30384088496321743,
+      "learning_rate": 5.401303700689608e-07,
+      "loss": 0.6156,
+      "step": 4827
+    },
+    {
+      "epoch": 11.80440097799511,
+      "grad_norm": 0.2179882472303422,
+      "learning_rate": 5.393358493413309e-07,
+      "loss": 0.7632,
+      "step": 4828
+    },
+    {
+      "epoch": 11.80684596577017,
+      "grad_norm": 0.23018990538877515,
+      "learning_rate": 5.385418427447229e-07,
+      "loss": 1.0303,
+      "step": 4829
+    },
+    {
+      "epoch": 11.809290953545233,
+      "grad_norm": 0.28913188972653914,
+      "learning_rate": 5.37748350487344e-07,
+      "loss": 0.3733,
+      "step": 4830
+    },
+    {
+      "epoch": 11.811735941320293,
+      "grad_norm": 0.223322779047922,
+      "learning_rate": 5.369553727772672e-07,
+      "loss": 0.342,
+      "step": 4831
+    },
+    {
+      "epoch": 11.814180929095354,
+      "grad_norm": 0.21348720063181859,
+      "learning_rate": 5.361629098224269e-07,
+      "loss": 0.6159,
+      "step": 4832
+    },
+    {
+      "epoch": 11.816625916870416,
+      "grad_norm": 0.20844972795574548,
+      "learning_rate": 5.353709618306277e-07,
+      "loss": 0.8031,
+      "step": 4833
+    },
+    {
+      "epoch": 11.819070904645477,
+      "grad_norm": 0.22148415188121776,
+      "learning_rate": 5.345795290095352e-07,
+      "loss": 0.4855,
+      "step": 4834
+    },
+    {
+      "epoch": 11.821515892420537,
+      "grad_norm": 0.2308077788141017,
+      "learning_rate": 5.337886115666815e-07,
+      "loss": 0.4505,
+      "step": 4835
+    },
+    {
+      "epoch": 11.8239608801956,
+      "grad_norm": 0.275696085278996,
+      "learning_rate": 5.329982097094627e-07,
+      "loss": 0.5977,
+      "step": 4836
+    },
+    {
+      "epoch": 11.82640586797066,
+      "grad_norm": 0.17879173104254753,
+      "learning_rate": 5.322083236451406e-07,
+      "loss": 0.5651,
+      "step": 4837
+    },
+    {
+      "epoch": 11.82885085574572,
+      "grad_norm": 0.2528832636655907,
+      "learning_rate": 5.314189535808423e-07,
+      "loss": 0.406,
+      "step": 4838
+    },
+    {
+      "epoch": 11.831295843520783,
+      "grad_norm": 0.3128629887857883,
+      "learning_rate": 5.306300997235569e-07,
+      "loss": 1.089,
+      "step": 4839
+    },
+    {
+      "epoch": 11.833740831295843,
+      "grad_norm": 0.26680183813521247,
+      "learning_rate": 5.298417622801413e-07,
+      "loss": 0.6939,
+      "step": 4840
+    },
+    {
+      "epoch": 11.836185819070904,
+      "grad_norm": 0.23178691690139192,
+      "learning_rate": 5.290539414573157e-07,
+      "loss": 0.4246,
+      "step": 4841
+    },
+    {
+      "epoch": 11.838630806845966,
+      "grad_norm": 0.24012038258847387,
+      "learning_rate": 5.282666374616637e-07,
+      "loss": 0.679,
+      "step": 4842
+    },
+    {
+      "epoch": 11.841075794621027,
+      "grad_norm": 0.3283111358702929,
+      "learning_rate": 5.274798504996351e-07,
+      "loss": 0.473,
+      "step": 4843
+    },
+    {
+      "epoch": 11.843520782396087,
+      "grad_norm": 0.34528407815708956,
+      "learning_rate": 5.266935807775433e-07,
+      "loss": 0.4242,
+      "step": 4844
+    },
+    {
+      "epoch": 11.84596577017115,
+      "grad_norm": 0.2066340649402097,
+      "learning_rate": 5.259078285015668e-07,
+      "loss": 0.8594,
+      "step": 4845
+    },
+    {
+      "epoch": 11.84841075794621,
+      "grad_norm": 0.3025201122937535,
+      "learning_rate": 5.251225938777469e-07,
+      "loss": 0.7965,
+      "step": 4846
+    },
+    {
+      "epoch": 11.85085574572127,
+      "grad_norm": 0.2988188090146297,
+      "learning_rate": 5.243378771119903e-07,
+      "loss": 0.784,
+      "step": 4847
+    },
+    {
+      "epoch": 11.853300733496333,
+      "grad_norm": 0.28583556733591914,
+      "learning_rate": 5.235536784100689e-07,
+      "loss": 0.724,
+      "step": 4848
+    },
+    {
+      "epoch": 11.855745721271393,
+      "grad_norm": 0.30604816458491546,
+      "learning_rate": 5.227699979776158e-07,
+      "loss": 0.6654,
+      "step": 4849
+    },
+    {
+      "epoch": 11.858190709046454,
+      "grad_norm": 0.41579295242521735,
+      "learning_rate": 5.219868360201305e-07,
+      "loss": 0.471,
+      "step": 4850
+    },
+    {
+      "epoch": 11.860635696821516,
+      "grad_norm": 0.17441021745116003,
+      "learning_rate": 5.212041927429773e-07,
+      "loss": 0.6786,
+      "step": 4851
+    },
+    {
+      "epoch": 11.863080684596577,
+      "grad_norm": 0.22141277171609103,
+      "learning_rate": 5.204220683513811e-07,
+      "loss": 0.4054,
+      "step": 4852
+    },
+    {
+      "epoch": 11.865525672371637,
+      "grad_norm": 0.23043113893081826,
+      "learning_rate": 5.196404630504346e-07,
+      "loss": 0.325,
+      "step": 4853
+    },
+    {
+      "epoch": 11.8679706601467,
+      "grad_norm": 0.2376559306229159,
+      "learning_rate": 5.188593770450906e-07,
+      "loss": 0.6998,
+      "step": 4854
+    },
+    {
+      "epoch": 11.87041564792176,
+      "grad_norm": 0.23466943129903348,
+      "learning_rate": 5.180788105401701e-07,
+      "loss": 0.3305,
+      "step": 4855
+    },
+    {
+      "epoch": 11.87286063569682,
+      "grad_norm": 0.31841479536036726,
+      "learning_rate": 5.172987637403537e-07,
+      "loss": 1.2162,
+      "step": 4856
+    },
+    {
+      "epoch": 11.875305623471883,
+      "grad_norm": 0.20818810144016406,
+      "learning_rate": 5.165192368501884e-07,
+      "loss": 0.6377,
+      "step": 4857
+    },
+    {
+      "epoch": 11.877750611246944,
+      "grad_norm": 0.17997492481757107,
+      "learning_rate": 5.157402300740844e-07,
+      "loss": 0.4752,
+      "step": 4858
+    },
+    {
+      "epoch": 11.880195599022004,
+      "grad_norm": 0.2685022301990982,
+      "learning_rate": 5.149617436163134e-07,
+      "loss": 0.7146,
+      "step": 4859
+    },
+    {
+      "epoch": 11.882640586797066,
+      "grad_norm": 0.23387838356721757,
+      "learning_rate": 5.141837776810141e-07,
+      "loss": 0.688,
+      "step": 4860
+    },
+    {
+      "epoch": 11.885085574572127,
+      "grad_norm": 0.22647858829098802,
+      "learning_rate": 5.134063324721866e-07,
+      "loss": 0.6888,
+      "step": 4861
+    },
+    {
+      "epoch": 11.887530562347187,
+      "grad_norm": 0.24845238553438467,
+      "learning_rate": 5.126294081936939e-07,
+      "loss": 0.4501,
+      "step": 4862
+    },
+    {
+      "epoch": 11.88997555012225,
+      "grad_norm": 0.23427135951762612,
+      "learning_rate": 5.118530050492648e-07,
+      "loss": 0.4487,
+      "step": 4863
+    },
+    {
+      "epoch": 11.89242053789731,
+      "grad_norm": 0.3942334515894535,
+      "learning_rate": 5.110771232424886e-07,
+      "loss": 0.4053,
+      "step": 4864
+    },
+    {
+      "epoch": 11.894865525672373,
+      "grad_norm": 0.36981272366594314,
+      "learning_rate": 5.103017629768197e-07,
+      "loss": 0.4541,
+      "step": 4865
+    },
+    {
+      "epoch": 11.897310513447433,
+      "grad_norm": 0.24032147882402957,
+      "learning_rate": 5.095269244555756e-07,
+      "loss": 0.8607,
+      "step": 4866
+    },
+    {
+      "epoch": 11.899755501222494,
+      "grad_norm": 0.24857342772203783,
+      "learning_rate": 5.087526078819363e-07,
+      "loss": 0.5002,
+      "step": 4867
+    },
+    {
+      "epoch": 11.902200488997554,
+      "grad_norm": 0.26163835010011494,
+      "learning_rate": 5.079788134589464e-07,
+      "loss": 0.6311,
+      "step": 4868
+    },
+    {
+      "epoch": 11.904645476772616,
+      "grad_norm": 0.32514313211445794,
+      "learning_rate": 5.07205541389511e-07,
+      "loss": 0.3993,
+      "step": 4869
+    },
+    {
+      "epoch": 11.907090464547677,
+      "grad_norm": 0.2690740141006805,
+      "learning_rate": 5.064327918764006e-07,
+      "loss": 0.7586,
+      "step": 4870
+    },
+    {
+      "epoch": 11.90953545232274,
+      "grad_norm": 0.13632903066672414,
+      "learning_rate": 5.056605651222482e-07,
+      "loss": 0.3095,
+      "step": 4871
+    },
+    {
+      "epoch": 11.9119804400978,
+      "grad_norm": 0.22379803708199883,
+      "learning_rate": 5.048888613295483e-07,
+      "loss": 0.6278,
+      "step": 4872
+    },
+    {
+      "epoch": 11.91442542787286,
+      "grad_norm": 0.37995192602830485,
+      "learning_rate": 5.041176807006601e-07,
+      "loss": 0.3703,
+      "step": 4873
+    },
+    {
+      "epoch": 11.91687041564792,
+      "grad_norm": 0.23375625996587635,
+      "learning_rate": 5.03347023437804e-07,
+      "loss": 0.4041,
+      "step": 4874
+    },
+    {
+      "epoch": 11.919315403422983,
+      "grad_norm": 0.2795084420195988,
+      "learning_rate": 5.025768897430644e-07,
+      "loss": 0.5918,
+      "step": 4875
+    },
+    {
+      "epoch": 11.921760391198044,
+      "grad_norm": 0.2364884976463528,
+      "learning_rate": 5.018072798183881e-07,
+      "loss": 0.4875,
+      "step": 4876
+    },
+    {
+      "epoch": 11.924205378973106,
+      "grad_norm": 0.19458714156769039,
+      "learning_rate": 5.010381938655845e-07,
+      "loss": 0.738,
+      "step": 4877
+    },
+    {
+      "epoch": 11.926650366748166,
+      "grad_norm": 0.24615643969780854,
+      "learning_rate": 5.00269632086326e-07,
+      "loss": 0.4108,
+      "step": 4878
+    },
+    {
+      "epoch": 11.929095354523227,
+      "grad_norm": 0.2660258900138615,
+      "learning_rate": 4.995015946821458e-07,
+      "loss": 0.5286,
+      "step": 4879
+    },
+    {
+      "epoch": 11.93154034229829,
+      "grad_norm": 0.5732059739086813,
+      "learning_rate": 4.987340818544417e-07,
+      "loss": 0.6345,
+      "step": 4880
+    },
+    {
+      "epoch": 11.93398533007335,
+      "grad_norm": 0.1964365808213967,
+      "learning_rate": 4.979670938044737e-07,
+      "loss": 0.6577,
+      "step": 4881
+    },
+    {
+      "epoch": 11.93643031784841,
+      "grad_norm": 0.3024584516938132,
+      "learning_rate": 4.972006307333624e-07,
+      "loss": 0.4503,
+      "step": 4882
+    },
+    {
+      "epoch": 11.938875305623473,
+      "grad_norm": 0.24161055695260578,
+      "learning_rate": 4.964346928420932e-07,
+      "loss": 0.5782,
+      "step": 4883
+    },
+    {
+      "epoch": 11.941320293398533,
+      "grad_norm": 0.27759733566002004,
+      "learning_rate": 4.95669280331511e-07,
+      "loss": 0.6111,
+      "step": 4884
+    },
+    {
+      "epoch": 11.943765281173594,
+      "grad_norm": 0.39818775129009115,
+      "learning_rate": 4.949043934023257e-07,
+      "loss": 0.4773,
+      "step": 4885
+    },
+    {
+      "epoch": 11.946210268948656,
+      "grad_norm": 0.269868387692756,
+      "learning_rate": 4.941400322551074e-07,
+      "loss": 0.468,
+      "step": 4886
+    },
+    {
+      "epoch": 11.948655256723717,
+      "grad_norm": 0.2692333049835374,
+      "learning_rate": 4.933761970902898e-07,
+      "loss": 0.7696,
+      "step": 4887
+    },
+    {
+      "epoch": 11.951100244498777,
+      "grad_norm": 0.30646921726311277,
+      "learning_rate": 4.926128881081684e-07,
+      "loss": 0.542,
+      "step": 4888
+    },
+    {
+      "epoch": 11.95354523227384,
+      "grad_norm": 0.1910270625763932,
+      "learning_rate": 4.918501055088986e-07,
+      "loss": 0.3754,
+      "step": 4889
+    },
+    {
+      "epoch": 11.9559902200489,
+      "grad_norm": 0.2634752764386183,
+      "learning_rate": 4.910878494925009e-07,
+      "loss": 0.7436,
+      "step": 4890
+    },
+    {
+      "epoch": 11.95843520782396,
+      "grad_norm": 0.2691613176999981,
+      "learning_rate": 4.903261202588561e-07,
+      "loss": 0.6693,
+      "step": 4891
+    },
+    {
+      "epoch": 11.960880195599023,
+      "grad_norm": 0.25267810163609183,
+      "learning_rate": 4.895649180077061e-07,
+      "loss": 1.3046,
+      "step": 4892
+    },
+    {
+      "epoch": 11.963325183374083,
+      "grad_norm": 0.23324289861531824,
+      "learning_rate": 4.888042429386572e-07,
+      "loss": 0.7026,
+      "step": 4893
+    },
+    {
+      "epoch": 11.965770171149144,
+      "grad_norm": 0.29108133442388834,
+      "learning_rate": 4.88044095251174e-07,
+      "loss": 0.4995,
+      "step": 4894
+    },
+    {
+      "epoch": 11.968215158924206,
+      "grad_norm": 0.3410673393714446,
+      "learning_rate": 4.872844751445851e-07,
+      "loss": 0.5615,
+      "step": 4895
+    },
+    {
+      "epoch": 11.970660146699267,
+      "grad_norm": 0.24941392874561313,
+      "learning_rate": 4.865253828180816e-07,
+      "loss": 0.9909,
+      "step": 4896
+    },
+    {
+      "epoch": 11.973105134474327,
+      "grad_norm": 0.25637840700090175,
+      "learning_rate": 4.857668184707126e-07,
+      "loss": 0.6614,
+      "step": 4897
+    },
+    {
+      "epoch": 11.97555012224939,
+      "grad_norm": 0.28223123795727095,
+      "learning_rate": 4.850087823013935e-07,
+      "loss": 0.811,
+      "step": 4898
+    },
+    {
+      "epoch": 11.97799511002445,
+      "grad_norm": 0.2518362812370028,
+      "learning_rate": 4.842512745088973e-07,
+      "loss": 0.4578,
+      "step": 4899
+    },
+    {
+      "epoch": 11.98044009779951,
+      "grad_norm": 0.2937450438739105,
+      "learning_rate": 4.834942952918598e-07,
+      "loss": 0.4915,
+      "step": 4900
+    },
+    {
+      "epoch": 11.982885085574573,
+      "grad_norm": 0.2965872293139258,
+      "learning_rate": 4.827378448487793e-07,
+      "loss": 0.5215,
+      "step": 4901
+    },
+    {
+      "epoch": 11.985330073349633,
+      "grad_norm": 0.29556829670653845,
+      "learning_rate": 4.819819233780135e-07,
+      "loss": 0.4673,
+      "step": 4902
+    },
+    {
+      "epoch": 11.987775061124694,
+      "grad_norm": 0.21410853275804778,
+      "learning_rate": 4.812265310777828e-07,
+      "loss": 0.5493,
+      "step": 4903
+    },
+    {
+      "epoch": 11.990220048899756,
+      "grad_norm": 0.2993202197685635,
+      "learning_rate": 4.804716681461677e-07,
+      "loss": 0.6656,
+      "step": 4904
+    },
+    {
+      "epoch": 11.992665036674817,
+      "grad_norm": 0.39750040402899234,
+      "learning_rate": 4.797173347811109e-07,
+      "loss": 0.3661,
+      "step": 4905
+    },
+    {
+      "epoch": 11.995110024449877,
+      "grad_norm": 0.26321518557157253,
+      "learning_rate": 4.789635311804167e-07,
+      "loss": 0.4174,
+      "step": 4906
+    },
+    {
+      "epoch": 11.99755501222494,
+      "grad_norm": 0.3585518388642963,
+      "learning_rate": 4.782102575417474e-07,
+      "loss": 0.4389,
+      "step": 4907
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.20999734913997922,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.9346,
+      "step": 4908
+    },
+    {
+      "epoch": 12.00244498777506,
+      "grad_norm": 0.14974658037788552,
+      "learning_rate": 4.7670530094045373e-07,
+      "loss": 0.3464,
+      "step": 4909
+    },
+    {
+      "epoch": 12.004889975550123,
+      "grad_norm": 0.2973883503464538,
+      "learning_rate": 4.7595361837246195e-07,
+      "loss": 0.8563,
+      "step": 4910
+    },
+    {
+      "epoch": 12.007334963325183,
+      "grad_norm": 0.4240554816834499,
+      "learning_rate": 4.75202466555765e-07,
+      "loss": 0.3456,
+      "step": 4911
+    },
+    {
+      "epoch": 12.009779951100244,
+      "grad_norm": 0.22702086800957727,
+      "learning_rate": 4.744518456873312e-07,
+      "loss": 0.4787,
+      "step": 4912
+    },
+    {
+      "epoch": 12.012224938875306,
+      "grad_norm": 0.2052304755907729,
+      "learning_rate": 4.7370175596399197e-07,
+      "loss": 0.6494,
+      "step": 4913
+    },
+    {
+      "epoch": 12.014669926650367,
+      "grad_norm": 0.3580177799502247,
+      "learning_rate": 4.729521975824361e-07,
+      "loss": 0.8716,
+      "step": 4914
+    },
+    {
+      "epoch": 12.017114914425427,
+      "grad_norm": 0.4893259899611762,
+      "learning_rate": 4.7220317073921605e-07,
+      "loss": 0.8095,
+      "step": 4915
+    },
+    {
+      "epoch": 12.01955990220049,
+      "grad_norm": 0.3881443725575305,
+      "learning_rate": 4.714546756307442e-07,
+      "loss": 0.49,
+      "step": 4916
+    },
+    {
+      "epoch": 12.02200488997555,
+      "grad_norm": 0.2728848214819317,
+      "learning_rate": 4.7070671245329185e-07,
+      "loss": 0.4269,
+      "step": 4917
+    },
+    {
+      "epoch": 12.02444987775061,
+      "grad_norm": 0.28924819765520177,
+      "learning_rate": 4.6995928140299296e-07,
+      "loss": 0.5143,
+      "step": 4918
+    },
+    {
+      "epoch": 12.026894865525673,
+      "grad_norm": 0.3424151853744166,
+      "learning_rate": 4.6921238267584104e-07,
+      "loss": 0.4133,
+      "step": 4919
+    },
+    {
+      "epoch": 12.029339853300733,
+      "grad_norm": 0.24767702876112604,
+      "learning_rate": 4.684660164676896e-07,
+      "loss": 0.5702,
+      "step": 4920
+    },
+    {
+      "epoch": 12.031784841075794,
+      "grad_norm": 0.23999566822622512,
+      "learning_rate": 4.6772018297425394e-07,
+      "loss": 0.6688,
+      "step": 4921
+    },
+    {
+      "epoch": 12.034229828850856,
+      "grad_norm": 0.22883631460327247,
+      "learning_rate": 4.6697488239110744e-07,
+      "loss": 0.3411,
+      "step": 4922
+    },
+    {
+      "epoch": 12.036674816625917,
+      "grad_norm": 0.45427945552661875,
+      "learning_rate": 4.662301149136861e-07,
+      "loss": 0.7177,
+      "step": 4923
+    },
+    {
+      "epoch": 12.039119804400977,
+      "grad_norm": 0.2528791992440227,
+      "learning_rate": 4.6548588073728363e-07,
+      "loss": 0.6214,
+      "step": 4924
+    },
+    {
+      "epoch": 12.04156479217604,
+      "grad_norm": 0.30876164010648816,
+      "learning_rate": 4.647421800570562e-07,
+      "loss": 0.7487,
+      "step": 4925
+    },
+    {
+      "epoch": 12.0440097799511,
+      "grad_norm": 0.4290919427922011,
+      "learning_rate": 4.639990130680197e-07,
+      "loss": 0.5046,
+      "step": 4926
+    },
+    {
+      "epoch": 12.04645476772616,
+      "grad_norm": 0.24842678931905046,
+      "learning_rate": 4.63256379965048e-07,
+      "loss": 0.4291,
+      "step": 4927
+    },
+    {
+      "epoch": 12.048899755501223,
+      "grad_norm": 0.26996187161922786,
+      "learning_rate": 4.625142809428773e-07,
+      "loss": 0.7915,
+      "step": 4928
+    },
+    {
+      "epoch": 12.051344743276283,
+      "grad_norm": 0.3780335014734925,
+      "learning_rate": 4.6177271619610305e-07,
+      "loss": 0.3806,
+      "step": 4929
+    },
+    {
+      "epoch": 12.053789731051344,
+      "grad_norm": 0.3572646532543196,
+      "learning_rate": 4.610316859191805e-07,
+      "loss": 0.4199,
+      "step": 4930
+    },
+    {
+      "epoch": 12.056234718826406,
+      "grad_norm": 0.2725853249922248,
+      "learning_rate": 4.60291190306425e-07,
+      "loss": 0.695,
+      "step": 4931
+    },
+    {
+      "epoch": 12.058679706601467,
+      "grad_norm": 0.24430421053166984,
+      "learning_rate": 4.5955122955201076e-07,
+      "loss": 0.3543,
+      "step": 4932
+    },
+    {
+      "epoch": 12.061124694376527,
+      "grad_norm": 0.26127812398454425,
+      "learning_rate": 4.5881180384997336e-07,
+      "loss": 0.6491,
+      "step": 4933
+    },
+    {
+      "epoch": 12.06356968215159,
+      "grad_norm": 0.2148919858733965,
+      "learning_rate": 4.5807291339420607e-07,
+      "loss": 0.3953,
+      "step": 4934
+    },
+    {
+      "epoch": 12.06601466992665,
+      "grad_norm": 0.42044662904601915,
+      "learning_rate": 4.573345583784633e-07,
+      "loss": 0.4603,
+      "step": 4935
+    },
+    {
+      "epoch": 12.06845965770171,
+      "grad_norm": 0.28203693741480496,
+      "learning_rate": 4.565967389963594e-07,
+      "loss": 0.6747,
+      "step": 4936
+    },
+    {
+      "epoch": 12.070904645476773,
+      "grad_norm": 0.18577593335260578,
+      "learning_rate": 4.5585945544136615e-07,
+      "loss": 0.3818,
+      "step": 4937
+    },
+    {
+      "epoch": 12.073349633251834,
+      "grad_norm": 0.26607732766170883,
+      "learning_rate": 4.5512270790681684e-07,
+      "loss": 0.5725,
+      "step": 4938
+    },
+    {
+      "epoch": 12.075794621026894,
+      "grad_norm": 0.2823956957916674,
+      "learning_rate": 4.5438649658590405e-07,
+      "loss": 0.9092,
+      "step": 4939
+    },
+    {
+      "epoch": 12.078239608801956,
+      "grad_norm": 0.29256557996793753,
+      "learning_rate": 4.5365082167167786e-07,
+      "loss": 0.5123,
+      "step": 4940
+    },
+    {
+      "epoch": 12.080684596577017,
+      "grad_norm": 0.21145596410616554,
+      "learning_rate": 4.529156833570508e-07,
+      "loss": 0.402,
+      "step": 4941
+    },
+    {
+      "epoch": 12.083129584352077,
+      "grad_norm": 0.2647221025831137,
+      "learning_rate": 4.521810818347916e-07,
+      "loss": 0.5432,
+      "step": 4942
+    },
+    {
+      "epoch": 12.08557457212714,
+      "grad_norm": 0.267127167065398,
+      "learning_rate": 4.5144701729753103e-07,
+      "loss": 0.3557,
+      "step": 4943
+    },
+    {
+      "epoch": 12.0880195599022,
+      "grad_norm": 0.3764503047214616,
+      "learning_rate": 4.507134899377558e-07,
+      "loss": 0.5194,
+      "step": 4944
+    },
+    {
+      "epoch": 12.09046454767726,
+      "grad_norm": 0.19465521547983414,
+      "learning_rate": 4.499804999478144e-07,
+      "loss": 0.4199,
+      "step": 4945
+    },
+    {
+      "epoch": 12.092909535452323,
+      "grad_norm": 0.2828227441637348,
+      "learning_rate": 4.492480475199146e-07,
+      "loss": 1.2859,
+      "step": 4946
+    },
+    {
+      "epoch": 12.095354523227384,
+      "grad_norm": 0.24407840282206383,
+      "learning_rate": 4.485161328461204e-07,
+      "loss": 0.895,
+      "step": 4947
+    },
+    {
+      "epoch": 12.097799511002446,
+      "grad_norm": 0.31045936327212736,
+      "learning_rate": 4.4778475611835766e-07,
+      "loss": 0.431,
+      "step": 4948
+    },
+    {
+      "epoch": 12.100244498777506,
+      "grad_norm": 0.2098592875810141,
+      "learning_rate": 4.4705391752841027e-07,
+      "loss": 0.3738,
+      "step": 4949
+    },
+    {
+      "epoch": 12.102689486552567,
+      "grad_norm": 0.24051221187610344,
+      "learning_rate": 4.4632361726791915e-07,
+      "loss": 1.037,
+      "step": 4950
+    },
+    {
+      "epoch": 12.10513447432763,
+      "grad_norm": 0.254559882601482,
+      "learning_rate": 4.4559385552838843e-07,
+      "loss": 0.4883,
+      "step": 4951
+    },
+    {
+      "epoch": 12.10757946210269,
+      "grad_norm": 0.3127148214794706,
+      "learning_rate": 4.4486463250117626e-07,
+      "loss": 0.4505,
+      "step": 4952
+    },
+    {
+      "epoch": 12.11002444987775,
+      "grad_norm": 0.23420130253635035,
+      "learning_rate": 4.441359483775032e-07,
+      "loss": 0.3181,
+      "step": 4953
+    },
+    {
+      "epoch": 12.112469437652813,
+      "grad_norm": 0.2736165661244627,
+      "learning_rate": 4.4340780334844516e-07,
+      "loss": 0.5839,
+      "step": 4954
+    },
+    {
+      "epoch": 12.114914425427873,
+      "grad_norm": 0.2753630071857551,
+      "learning_rate": 4.4268019760493914e-07,
+      "loss": 0.5044,
+      "step": 4955
+    },
+    {
+      "epoch": 12.117359413202934,
+      "grad_norm": 0.24233135615121223,
+      "learning_rate": 4.419531313377809e-07,
+      "loss": 0.5101,
+      "step": 4956
+    },
+    {
+      "epoch": 12.119804400977996,
+      "grad_norm": 0.2231071059003483,
+      "learning_rate": 4.4122660473762266e-07,
+      "loss": 0.4862,
+      "step": 4957
+    },
+    {
+      "epoch": 12.122249388753056,
+      "grad_norm": 0.1892774326366706,
+      "learning_rate": 4.405006179949767e-07,
+      "loss": 0.4213,
+      "step": 4958
+    },
+    {
+      "epoch": 12.124694376528117,
+      "grad_norm": 0.2279861699639557,
+      "learning_rate": 4.397751713002141e-07,
+      "loss": 0.4399,
+      "step": 4959
+    },
+    {
+      "epoch": 12.12713936430318,
+      "grad_norm": 0.316489340569337,
+      "learning_rate": 4.390502648435624e-07,
+      "loss": 0.4053,
+      "step": 4960
+    },
+    {
+      "epoch": 12.12958435207824,
+      "grad_norm": 0.2143646131539836,
+      "learning_rate": 4.38325898815109e-07,
+      "loss": 0.4888,
+      "step": 4961
+    },
+    {
+      "epoch": 12.1320293398533,
+      "grad_norm": 0.24333042106489217,
+      "learning_rate": 4.376020734047995e-07,
+      "loss": 0.4852,
+      "step": 4962
+    },
+    {
+      "epoch": 12.134474327628363,
+      "grad_norm": 0.26319118922886414,
+      "learning_rate": 4.368787888024384e-07,
+      "loss": 0.6872,
+      "step": 4963
+    },
+    {
+      "epoch": 12.136919315403423,
+      "grad_norm": 0.3009017983293089,
+      "learning_rate": 4.361560451976857e-07,
+      "loss": 0.4274,
+      "step": 4964
+    },
+    {
+      "epoch": 12.139364303178484,
+      "grad_norm": 0.19805722915349858,
+      "learning_rate": 4.3543384278006196e-07,
+      "loss": 0.5214,
+      "step": 4965
+    },
+    {
+      "epoch": 12.141809290953546,
+      "grad_norm": 0.2775371159143536,
+      "learning_rate": 4.347121817389463e-07,
+      "loss": 0.4986,
+      "step": 4966
+    },
+    {
+      "epoch": 12.144254278728607,
+      "grad_norm": 0.29440794914667917,
+      "learning_rate": 4.3399106226357317e-07,
+      "loss": 0.6688,
+      "step": 4967
+    },
+    {
+      "epoch": 12.146699266503667,
+      "grad_norm": 0.25714384591712175,
+      "learning_rate": 4.332704845430372e-07,
+      "loss": 0.9102,
+      "step": 4968
+    },
+    {
+      "epoch": 12.14914425427873,
+      "grad_norm": 0.26350374191122117,
+      "learning_rate": 4.32550448766291e-07,
+      "loss": 0.6438,
+      "step": 4969
+    },
+    {
+      "epoch": 12.15158924205379,
+      "grad_norm": 0.22984760882582447,
+      "learning_rate": 4.3183095512214336e-07,
+      "loss": 0.661,
+      "step": 4970
+    },
+    {
+      "epoch": 12.15403422982885,
+      "grad_norm": 0.23714512029480686,
+      "learning_rate": 4.3111200379926306e-07,
+      "loss": 0.8397,
+      "step": 4971
+    },
+    {
+      "epoch": 12.156479217603913,
+      "grad_norm": 0.28635306255729304,
+      "learning_rate": 4.30393594986174e-07,
+      "loss": 0.616,
+      "step": 4972
+    },
+    {
+      "epoch": 12.158924205378973,
+      "grad_norm": 0.2813402344600723,
+      "learning_rate": 4.296757288712616e-07,
+      "loss": 0.546,
+      "step": 4973
+    },
+    {
+      "epoch": 12.161369193154034,
+      "grad_norm": 0.26632731963454787,
+      "learning_rate": 4.289584056427648e-07,
+      "loss": 0.541,
+      "step": 4974
+    },
+    {
+      "epoch": 12.163814180929096,
+      "grad_norm": 0.19002905742819837,
+      "learning_rate": 4.282416254887833e-07,
+      "loss": 0.8745,
+      "step": 4975
+    },
+    {
+      "epoch": 12.166259168704157,
+      "grad_norm": 0.24860555945941368,
+      "learning_rate": 4.2752538859727355e-07,
+      "loss": 0.4566,
+      "step": 4976
+    },
+    {
+      "epoch": 12.168704156479217,
+      "grad_norm": 0.23039239212238033,
+      "learning_rate": 4.26809695156048e-07,
+      "loss": 0.6884,
+      "step": 4977
+    },
+    {
+      "epoch": 12.17114914425428,
+      "grad_norm": 0.18796083473835296,
+      "learning_rate": 4.260945453527784e-07,
+      "loss": 0.4938,
+      "step": 4978
+    },
+    {
+      "epoch": 12.17359413202934,
+      "grad_norm": 0.31057752004240696,
+      "learning_rate": 4.253799393749944e-07,
+      "loss": 0.5371,
+      "step": 4979
+    },
+    {
+      "epoch": 12.1760391198044,
+      "grad_norm": 0.226832703208997,
+      "learning_rate": 4.2466587741008034e-07,
+      "loss": 0.6169,
+      "step": 4980
+    },
+    {
+      "epoch": 12.178484107579463,
+      "grad_norm": 0.3004969549899206,
+      "learning_rate": 4.239523596452813e-07,
+      "loss": 0.7041,
+      "step": 4981
+    },
+    {
+      "epoch": 12.180929095354523,
+      "grad_norm": 0.2580912284484904,
+      "learning_rate": 4.2323938626769647e-07,
+      "loss": 0.422,
+      "step": 4982
+    },
+    {
+      "epoch": 12.183374083129584,
+      "grad_norm": 0.3738895179260783,
+      "learning_rate": 4.225269574642843e-07,
+      "loss": 0.687,
+      "step": 4983
+    },
+    {
+      "epoch": 12.185819070904646,
+      "grad_norm": 0.3097757571593831,
+      "learning_rate": 4.218150734218604e-07,
+      "loss": 0.4423,
+      "step": 4984
+    },
+    {
+      "epoch": 12.188264058679707,
+      "grad_norm": 0.22358035129012138,
+      "learning_rate": 4.211037343270963e-07,
+      "loss": 0.6642,
+      "step": 4985
+    },
+    {
+      "epoch": 12.190709046454767,
+      "grad_norm": 0.36784878029512597,
+      "learning_rate": 4.2039294036652274e-07,
+      "loss": 0.564,
+      "step": 4986
+    },
+    {
+      "epoch": 12.19315403422983,
+      "grad_norm": 0.33766412668757234,
+      "learning_rate": 4.1968269172652454e-07,
+      "loss": 0.6897,
+      "step": 4987
+    },
+    {
+      "epoch": 12.19559902200489,
+      "grad_norm": 0.24811916424188032,
+      "learning_rate": 4.189729885933461e-07,
+      "loss": 0.722,
+      "step": 4988
+    },
+    {
+      "epoch": 12.19804400977995,
+      "grad_norm": 0.26817919242988325,
+      "learning_rate": 4.182638311530879e-07,
+      "loss": 0.7672,
+      "step": 4989
+    },
+    {
+      "epoch": 12.200488997555013,
+      "grad_norm": 0.31152358691259346,
+      "learning_rate": 4.175552195917065e-07,
+      "loss": 0.4398,
+      "step": 4990
+    },
+    {
+      "epoch": 12.202933985330073,
+      "grad_norm": 0.18347197414672434,
+      "learning_rate": 4.168471540950173e-07,
+      "loss": 0.8236,
+      "step": 4991
+    },
+    {
+      "epoch": 12.205378973105134,
+      "grad_norm": 0.3233397036690673,
+      "learning_rate": 4.161396348486896e-07,
+      "loss": 0.4922,
+      "step": 4992
+    },
+    {
+      "epoch": 12.207823960880196,
+      "grad_norm": 0.2261237803673091,
+      "learning_rate": 4.1543266203825204e-07,
+      "loss": 0.7712,
+      "step": 4993
+    },
+    {
+      "epoch": 12.210268948655257,
+      "grad_norm": 0.2247211866231037,
+      "learning_rate": 4.1472623584908945e-07,
+      "loss": 0.46,
+      "step": 4994
+    },
+    {
+      "epoch": 12.212713936430317,
+      "grad_norm": 0.2848126151299553,
+      "learning_rate": 4.1402035646644216e-07,
+      "loss": 0.8908,
+      "step": 4995
+    },
+    {
+      "epoch": 12.21515892420538,
+      "grad_norm": 0.19839595979620234,
+      "learning_rate": 4.133150240754094e-07,
+      "loss": 0.6037,
+      "step": 4996
+    },
+    {
+      "epoch": 12.21760391198044,
+      "grad_norm": 0.2876660121754633,
+      "learning_rate": 4.1261023886094345e-07,
+      "loss": 0.5639,
+      "step": 4997
+    },
+    {
+      "epoch": 12.2200488997555,
+      "grad_norm": 0.276443185488912,
+      "learning_rate": 4.1190600100785627e-07,
+      "loss": 0.4118,
+      "step": 4998
+    },
+    {
+      "epoch": 12.222493887530563,
+      "grad_norm": 0.18855166535058215,
+      "learning_rate": 4.1120231070081536e-07,
+      "loss": 0.4222,
+      "step": 4999
+    },
+    {
+      "epoch": 12.224938875305623,
+      "grad_norm": 0.15577853072870454,
+      "learning_rate": 4.1049916812434335e-07,
+      "loss": 0.4215,
+      "step": 5000
+    },
+    {
+      "epoch": 12.227383863080684,
+      "grad_norm": 0.2208354021479963,
+      "learning_rate": 4.0979657346282175e-07,
+      "loss": 0.5355,
+      "step": 5001
+    },
+    {
+      "epoch": 12.229828850855746,
+      "grad_norm": 0.19732094560503924,
+      "learning_rate": 4.090945269004856e-07,
+      "loss": 0.4084,
+      "step": 5002
+    },
+    {
+      "epoch": 12.232273838630807,
+      "grad_norm": 0.2865513880507124,
+      "learning_rate": 4.0839302862142825e-07,
+      "loss": 0.4367,
+      "step": 5003
+    },
+    {
+      "epoch": 12.234718826405867,
+      "grad_norm": 0.23650126395243395,
+      "learning_rate": 4.0769207880959837e-07,
+      "loss": 0.6435,
+      "step": 5004
+    },
+    {
+      "epoch": 12.23716381418093,
+      "grad_norm": 0.2837359383769163,
+      "learning_rate": 4.069916776488014e-07,
+      "loss": 0.6187,
+      "step": 5005
+    },
+    {
+      "epoch": 12.23960880195599,
+      "grad_norm": 0.30455379321359616,
+      "learning_rate": 4.062918253226988e-07,
+      "loss": 0.7203,
+      "step": 5006
+    },
+    {
+      "epoch": 12.24205378973105,
+      "grad_norm": 0.37021966736430867,
+      "learning_rate": 4.0559252201480715e-07,
+      "loss": 0.3214,
+      "step": 5007
+    },
+    {
+      "epoch": 12.244498777506113,
+      "grad_norm": 0.3171274598994411,
+      "learning_rate": 4.0489376790849967e-07,
+      "loss": 0.7396,
+      "step": 5008
+    },
+    {
+      "epoch": 12.246943765281173,
+      "grad_norm": 0.2852895153459756,
+      "learning_rate": 4.04195563187007e-07,
+      "loss": 0.7068,
+      "step": 5009
+    },
+    {
+      "epoch": 12.249388753056234,
+      "grad_norm": 0.28382637494332086,
+      "learning_rate": 4.034979080334128e-07,
+      "loss": 0.8304,
+      "step": 5010
+    },
+    {
+      "epoch": 12.251833740831296,
+      "grad_norm": 0.42378223522649555,
+      "learning_rate": 4.028008026306593e-07,
+      "loss": 0.4572,
+      "step": 5011
+    },
+    {
+      "epoch": 12.254278728606357,
+      "grad_norm": 0.19430629363737095,
+      "learning_rate": 4.021042471615427e-07,
+      "loss": 0.6891,
+      "step": 5012
+    },
+    {
+      "epoch": 12.256723716381417,
+      "grad_norm": 0.39115272511553995,
+      "learning_rate": 4.0140824180871616e-07,
+      "loss": 0.4623,
+      "step": 5013
+    },
+    {
+      "epoch": 12.25916870415648,
+      "grad_norm": 0.2603520388334594,
+      "learning_rate": 4.0071278675468893e-07,
+      "loss": 0.4589,
+      "step": 5014
+    },
+    {
+      "epoch": 12.26161369193154,
+      "grad_norm": 0.19412919349615174,
+      "learning_rate": 4.0001788218182345e-07,
+      "loss": 0.6218,
+      "step": 5015
+    },
+    {
+      "epoch": 12.2640586797066,
+      "grad_norm": 0.24701566988964282,
+      "learning_rate": 3.993235282723418e-07,
+      "loss": 0.7555,
+      "step": 5016
+    },
+    {
+      "epoch": 12.266503667481663,
+      "grad_norm": 0.37896591986455336,
+      "learning_rate": 3.9862972520831767e-07,
+      "loss": 0.6261,
+      "step": 5017
+    },
+    {
+      "epoch": 12.268948655256724,
+      "grad_norm": 0.28678038675444295,
+      "learning_rate": 3.979364731716828e-07,
+      "loss": 0.6443,
+      "step": 5018
+    },
+    {
+      "epoch": 12.271393643031784,
+      "grad_norm": 0.2124750689291723,
+      "learning_rate": 3.972437723442246e-07,
+      "loss": 0.7061,
+      "step": 5019
+    },
+    {
+      "epoch": 12.273838630806846,
+      "grad_norm": 0.2171672028171427,
+      "learning_rate": 3.9655162290758357e-07,
+      "loss": 0.6889,
+      "step": 5020
+    },
+    {
+      "epoch": 12.276283618581907,
+      "grad_norm": 0.2606095098441035,
+      "learning_rate": 3.9586002504325843e-07,
+      "loss": 0.6538,
+      "step": 5021
+    },
+    {
+      "epoch": 12.278728606356967,
+      "grad_norm": 0.1885873915613051,
+      "learning_rate": 3.9516897893260075e-07,
+      "loss": 0.4518,
+      "step": 5022
+    },
+    {
+      "epoch": 12.28117359413203,
+      "grad_norm": 0.31624293739335085,
+      "learning_rate": 3.944784847568192e-07,
+      "loss": 0.8837,
+      "step": 5023
+    },
+    {
+      "epoch": 12.28361858190709,
+      "grad_norm": 0.22936654256411337,
+      "learning_rate": 3.9378854269697787e-07,
+      "loss": 0.5566,
+      "step": 5024
+    },
+    {
+      "epoch": 12.28606356968215,
+      "grad_norm": 0.33992117882109785,
+      "learning_rate": 3.930991529339936e-07,
+      "loss": 0.5945,
+      "step": 5025
+    },
+    {
+      "epoch": 12.288508557457213,
+      "grad_norm": 0.2455890228239286,
+      "learning_rate": 3.9241031564864225e-07,
+      "loss": 0.7229,
+      "step": 5026
+    },
+    {
+      "epoch": 12.290953545232274,
+      "grad_norm": 0.2926308229577295,
+      "learning_rate": 3.9172203102155135e-07,
+      "loss": 0.465,
+      "step": 5027
+    },
+    {
+      "epoch": 12.293398533007334,
+      "grad_norm": 0.18619148394811125,
+      "learning_rate": 3.91034299233205e-07,
+      "loss": 0.3258,
+      "step": 5028
+    },
+    {
+      "epoch": 12.295843520782396,
+      "grad_norm": 0.33312540021219555,
+      "learning_rate": 3.903471204639434e-07,
+      "loss": 0.44,
+      "step": 5029
+    },
+    {
+      "epoch": 12.298288508557457,
+      "grad_norm": 0.19905025703442614,
+      "learning_rate": 3.896604948939589e-07,
+      "loss": 0.5781,
+      "step": 5030
+    },
+    {
+      "epoch": 12.30073349633252,
+      "grad_norm": 0.2597890286787564,
+      "learning_rate": 3.889744227033018e-07,
+      "loss": 0.4012,
+      "step": 5031
+    },
+    {
+      "epoch": 12.30317848410758,
+      "grad_norm": 0.26191417147920854,
+      "learning_rate": 3.8828890407187486e-07,
+      "loss": 0.4122,
+      "step": 5032
+    },
+    {
+      "epoch": 12.30562347188264,
+      "grad_norm": 0.2799213915498337,
+      "learning_rate": 3.8760393917943726e-07,
+      "loss": 0.512,
+      "step": 5033
+    },
+    {
+      "epoch": 12.3080684596577,
+      "grad_norm": 0.30207005310324064,
+      "learning_rate": 3.869195282056029e-07,
+      "loss": 0.6469,
+      "step": 5034
+    },
+    {
+      "epoch": 12.310513447432763,
+      "grad_norm": 0.722534654454409,
+      "learning_rate": 3.862356713298387e-07,
+      "loss": 0.4412,
+      "step": 5035
+    },
+    {
+      "epoch": 12.312958435207824,
+      "grad_norm": 0.23657852516259256,
+      "learning_rate": 3.855523687314688e-07,
+      "loss": 0.55,
+      "step": 5036
+    },
+    {
+      "epoch": 12.315403422982886,
+      "grad_norm": 0.2690300598345504,
+      "learning_rate": 3.8486962058967043e-07,
+      "loss": 0.5133,
+      "step": 5037
+    },
+    {
+      "epoch": 12.317848410757946,
+      "grad_norm": 0.23956015325747637,
+      "learning_rate": 3.8418742708347556e-07,
+      "loss": 0.5295,
+      "step": 5038
+    },
+    {
+      "epoch": 12.320293398533007,
+      "grad_norm": 0.2660879263387045,
+      "learning_rate": 3.83505788391772e-07,
+      "loss": 0.3062,
+      "step": 5039
+    },
+    {
+      "epoch": 12.32273838630807,
+      "grad_norm": 0.33229498982937933,
+      "learning_rate": 3.828247046932992e-07,
+      "loss": 0.5089,
+      "step": 5040
+    },
+    {
+      "epoch": 12.32518337408313,
+      "grad_norm": 0.18521643108560396,
+      "learning_rate": 3.821441761666547e-07,
+      "loss": 0.8748,
+      "step": 5041
+    },
+    {
+      "epoch": 12.32762836185819,
+      "grad_norm": 0.38734838474459843,
+      "learning_rate": 3.8146420299028737e-07,
+      "loss": 0.4731,
+      "step": 5042
+    },
+    {
+      "epoch": 12.330073349633253,
+      "grad_norm": 0.29821763679064184,
+      "learning_rate": 3.80784785342502e-07,
+      "loss": 0.4644,
+      "step": 5043
+    },
+    {
+      "epoch": 12.332518337408313,
+      "grad_norm": 0.21019307637097884,
+      "learning_rate": 3.801059234014581e-07,
+      "loss": 0.4445,
+      "step": 5044
+    },
+    {
+      "epoch": 12.334963325183374,
+      "grad_norm": 0.20709640539114063,
+      "learning_rate": 3.7942761734516803e-07,
+      "loss": 0.5038,
+      "step": 5045
+    },
+    {
+      "epoch": 12.337408312958436,
+      "grad_norm": 0.20676618286550388,
+      "learning_rate": 3.787498673514994e-07,
+      "loss": 0.4157,
+      "step": 5046
+    },
+    {
+      "epoch": 12.339853300733497,
+      "grad_norm": 0.1960462683718853,
+      "learning_rate": 3.7807267359817377e-07,
+      "loss": 0.2934,
+      "step": 5047
+    },
+    {
+      "epoch": 12.342298288508557,
+      "grad_norm": 0.3804553137532328,
+      "learning_rate": 3.773960362627671e-07,
+      "loss": 0.5244,
+      "step": 5048
+    },
+    {
+      "epoch": 12.34474327628362,
+      "grad_norm": 0.2984267497164799,
+      "learning_rate": 3.7671995552270943e-07,
+      "loss": 0.6049,
+      "step": 5049
+    },
+    {
+      "epoch": 12.34718826405868,
+      "grad_norm": 0.3254535330350803,
+      "learning_rate": 3.760444315552839e-07,
+      "loss": 0.6505,
+      "step": 5050
+    },
+    {
+      "epoch": 12.34963325183374,
+      "grad_norm": 0.31588070257782663,
+      "learning_rate": 3.753694645376291e-07,
+      "loss": 0.4292,
+      "step": 5051
+    },
+    {
+      "epoch": 12.352078239608803,
+      "grad_norm": 0.2450116203833375,
+      "learning_rate": 3.746950546467357e-07,
+      "loss": 0.4816,
+      "step": 5052
+    },
+    {
+      "epoch": 12.354523227383863,
+      "grad_norm": 0.2990960459893679,
+      "learning_rate": 3.7402120205945025e-07,
+      "loss": 0.799,
+      "step": 5053
+    },
+    {
+      "epoch": 12.356968215158924,
+      "grad_norm": 0.28017150965072696,
+      "learning_rate": 3.733479069524726e-07,
+      "loss": 0.8368,
+      "step": 5054
+    },
+    {
+      "epoch": 12.359413202933986,
+      "grad_norm": 0.3536025582885215,
+      "learning_rate": 3.726751695023553e-07,
+      "loss": 0.596,
+      "step": 5055
+    },
+    {
+      "epoch": 12.361858190709047,
+      "grad_norm": 0.16164615708845306,
+      "learning_rate": 3.720029898855057e-07,
+      "loss": 0.3734,
+      "step": 5056
+    },
+    {
+      "epoch": 12.364303178484107,
+      "grad_norm": 0.18055497733668746,
+      "learning_rate": 3.7133136827818575e-07,
+      "loss": 0.3936,
+      "step": 5057
+    },
+    {
+      "epoch": 12.36674816625917,
+      "grad_norm": 0.26896393565677973,
+      "learning_rate": 3.70660304856508e-07,
+      "loss": 0.7527,
+      "step": 5058
+    },
+    {
+      "epoch": 12.36919315403423,
+      "grad_norm": 0.2098587620020675,
+      "learning_rate": 3.69989799796443e-07,
+      "loss": 0.8906,
+      "step": 5059
+    },
+    {
+      "epoch": 12.37163814180929,
+      "grad_norm": 0.3212849828689608,
+      "learning_rate": 3.693198532738104e-07,
+      "loss": 0.6486,
+      "step": 5060
+    },
+    {
+      "epoch": 12.374083129584353,
+      "grad_norm": 0.32110300602575653,
+      "learning_rate": 3.686504654642875e-07,
+      "loss": 0.8406,
+      "step": 5061
+    },
+    {
+      "epoch": 12.376528117359413,
+      "grad_norm": 0.3150234896490052,
+      "learning_rate": 3.6798163654340107e-07,
+      "loss": 0.4141,
+      "step": 5062
+    },
+    {
+      "epoch": 12.378973105134474,
+      "grad_norm": 0.26603118025847994,
+      "learning_rate": 3.673133666865347e-07,
+      "loss": 0.8156,
+      "step": 5063
+    },
+    {
+      "epoch": 12.381418092909536,
+      "grad_norm": 0.21901387549515933,
+      "learning_rate": 3.666456560689241e-07,
+      "loss": 0.6184,
+      "step": 5064
+    },
+    {
+      "epoch": 12.383863080684597,
+      "grad_norm": 0.2419440298392444,
+      "learning_rate": 3.659785048656572e-07,
+      "loss": 0.4941,
+      "step": 5065
+    },
+    {
+      "epoch": 12.386308068459657,
+      "grad_norm": 0.2408103361825238,
+      "learning_rate": 3.653119132516772e-07,
+      "loss": 1.0229,
+      "step": 5066
+    },
+    {
+      "epoch": 12.38875305623472,
+      "grad_norm": 0.41925097861802874,
+      "learning_rate": 3.6464588140178003e-07,
+      "loss": 0.6872,
+      "step": 5067
+    },
+    {
+      "epoch": 12.39119804400978,
+      "grad_norm": 0.19283537359871333,
+      "learning_rate": 3.639804094906127e-07,
+      "loss": 0.4835,
+      "step": 5068
+    },
+    {
+      "epoch": 12.39364303178484,
+      "grad_norm": 0.3331807967759304,
+      "learning_rate": 3.6331549769267975e-07,
+      "loss": 0.5127,
+      "step": 5069
+    },
+    {
+      "epoch": 12.396088019559903,
+      "grad_norm": 0.2591735292648509,
+      "learning_rate": 3.62651146182334e-07,
+      "loss": 0.7028,
+      "step": 5070
+    },
+    {
+      "epoch": 12.398533007334963,
+      "grad_norm": 0.30381603877381846,
+      "learning_rate": 3.619873551337855e-07,
+      "loss": 0.3901,
+      "step": 5071
+    },
+    {
+      "epoch": 12.400977995110024,
+      "grad_norm": 0.23270857688987157,
+      "learning_rate": 3.613241247210936e-07,
+      "loss": 0.4022,
+      "step": 5072
+    },
+    {
+      "epoch": 12.403422982885086,
+      "grad_norm": 0.25348171035688427,
+      "learning_rate": 3.6066145511817363e-07,
+      "loss": 0.9455,
+      "step": 5073
+    },
+    {
+      "epoch": 12.405867970660147,
+      "grad_norm": 0.22923200733163546,
+      "learning_rate": 3.5999934649879304e-07,
+      "loss": 0.8498,
+      "step": 5074
+    },
+    {
+      "epoch": 12.408312958435207,
+      "grad_norm": 0.35376038981238406,
+      "learning_rate": 3.593377990365707e-07,
+      "loss": 0.9471,
+      "step": 5075
+    },
+    {
+      "epoch": 12.41075794621027,
+      "grad_norm": 0.2550170360615306,
+      "learning_rate": 3.586768129049803e-07,
+      "loss": 0.4988,
+      "step": 5076
+    },
+    {
+      "epoch": 12.41320293398533,
+      "grad_norm": 0.364700913099084,
+      "learning_rate": 3.580163882773477e-07,
+      "loss": 0.435,
+      "step": 5077
+    },
+    {
+      "epoch": 12.41564792176039,
+      "grad_norm": 0.2027159631723815,
+      "learning_rate": 3.573565253268507e-07,
+      "loss": 0.9242,
+      "step": 5078
+    },
+    {
+      "epoch": 12.418092909535453,
+      "grad_norm": 0.2663820848105799,
+      "learning_rate": 3.566972242265207e-07,
+      "loss": 0.9055,
+      "step": 5079
+    },
+    {
+      "epoch": 12.420537897310513,
+      "grad_norm": 0.2643749357542909,
+      "learning_rate": 3.5603848514924155e-07,
+      "loss": 0.7919,
+      "step": 5080
+    },
+    {
+      "epoch": 12.422982885085574,
+      "grad_norm": 0.2611858885328245,
+      "learning_rate": 3.5538030826775066e-07,
+      "loss": 0.9995,
+      "step": 5081
+    },
+    {
+      "epoch": 12.425427872860636,
+      "grad_norm": 0.26408003760376697,
+      "learning_rate": 3.547226937546353e-07,
+      "loss": 0.4547,
+      "step": 5082
+    },
+    {
+      "epoch": 12.427872860635697,
+      "grad_norm": 0.23373015151998647,
+      "learning_rate": 3.540656417823382e-07,
+      "loss": 0.43,
+      "step": 5083
+    },
+    {
+      "epoch": 12.430317848410757,
+      "grad_norm": 0.24577071343016127,
+      "learning_rate": 3.534091525231537e-07,
+      "loss": 0.3435,
+      "step": 5084
+    },
+    {
+      "epoch": 12.43276283618582,
+      "grad_norm": 0.19613120809263992,
+      "learning_rate": 3.5275322614922723e-07,
+      "loss": 0.8463,
+      "step": 5085
+    },
+    {
+      "epoch": 12.43520782396088,
+      "grad_norm": 0.22819306523183447,
+      "learning_rate": 3.52097862832558e-07,
+      "loss": 0.6876,
+      "step": 5086
+    },
+    {
+      "epoch": 12.43765281173594,
+      "grad_norm": 0.20694783884664794,
+      "learning_rate": 3.514430627449986e-07,
+      "loss": 0.5156,
+      "step": 5087
+    },
+    {
+      "epoch": 12.440097799511003,
+      "grad_norm": 0.41915172349507196,
+      "learning_rate": 3.507888260582507e-07,
+      "loss": 0.8944,
+      "step": 5088
+    },
+    {
+      "epoch": 12.442542787286063,
+      "grad_norm": 0.49521445385576895,
+      "learning_rate": 3.501351529438715e-07,
+      "loss": 0.8466,
+      "step": 5089
+    },
+    {
+      "epoch": 12.444987775061124,
+      "grad_norm": 0.20868128342074022,
+      "learning_rate": 3.4948204357326764e-07,
+      "loss": 0.7251,
+      "step": 5090
+    },
+    {
+      "epoch": 12.447432762836186,
+      "grad_norm": 0.2233425243957032,
+      "learning_rate": 3.488294981177012e-07,
+      "loss": 0.5372,
+      "step": 5091
+    },
+    {
+      "epoch": 12.449877750611247,
+      "grad_norm": 0.30600792833956014,
+      "learning_rate": 3.4817751674828317e-07,
+      "loss": 0.4954,
+      "step": 5092
+    },
+    {
+      "epoch": 12.452322738386307,
+      "grad_norm": 0.317578132034094,
+      "learning_rate": 3.4752609963597836e-07,
+      "loss": 0.6282,
+      "step": 5093
+    },
+    {
+      "epoch": 12.45476772616137,
+      "grad_norm": 0.22783869795385514,
+      "learning_rate": 3.468752469516037e-07,
+      "loss": 0.6363,
+      "step": 5094
+    },
+    {
+      "epoch": 12.45721271393643,
+      "grad_norm": 0.2829340669269438,
+      "learning_rate": 3.4622495886582704e-07,
+      "loss": 0.4799,
+      "step": 5095
+    },
+    {
+      "epoch": 12.45965770171149,
+      "grad_norm": 0.198926052813237,
+      "learning_rate": 3.455752355491687e-07,
+      "loss": 0.5273,
+      "step": 5096
+    },
+    {
+      "epoch": 12.462102689486553,
+      "grad_norm": 0.24207039400289684,
+      "learning_rate": 3.44926077172002e-07,
+      "loss": 0.828,
+      "step": 5097
+    },
+    {
+      "epoch": 12.464547677261614,
+      "grad_norm": 0.17038547861866635,
+      "learning_rate": 3.4427748390454986e-07,
+      "loss": 0.6051,
+      "step": 5098
+    },
+    {
+      "epoch": 12.466992665036674,
+      "grad_norm": 0.20057716884790783,
+      "learning_rate": 3.4362945591688955e-07,
+      "loss": 0.5921,
+      "step": 5099
+    },
+    {
+      "epoch": 12.469437652811736,
+      "grad_norm": 0.32037509799051955,
+      "learning_rate": 3.4298199337894687e-07,
+      "loss": 0.4386,
+      "step": 5100
+    },
+    {
+      "epoch": 12.471882640586797,
+      "grad_norm": 0.3088605730807667,
+      "learning_rate": 3.423350964605038e-07,
+      "loss": 0.3344,
+      "step": 5101
+    },
+    {
+      "epoch": 12.474327628361857,
+      "grad_norm": 0.3510904693480172,
+      "learning_rate": 3.4168876533118983e-07,
+      "loss": 0.8158,
+      "step": 5102
+    },
+    {
+      "epoch": 12.47677261613692,
+      "grad_norm": 0.2535272273176726,
+      "learning_rate": 3.4104300016048836e-07,
+      "loss": 0.4056,
+      "step": 5103
+    },
+    {
+      "epoch": 12.47921760391198,
+      "grad_norm": 0.31680296525846063,
+      "learning_rate": 3.4039780111773416e-07,
+      "loss": 0.4245,
+      "step": 5104
+    },
+    {
+      "epoch": 12.48166259168704,
+      "grad_norm": 0.2626988341174822,
+      "learning_rate": 3.3975316837211223e-07,
+      "loss": 0.528,
+      "step": 5105
+    },
+    {
+      "epoch": 12.484107579462103,
+      "grad_norm": 0.2548743953785793,
+      "learning_rate": 3.391091020926604e-07,
+      "loss": 1.2104,
+      "step": 5106
+    },
+    {
+      "epoch": 12.486552567237164,
+      "grad_norm": 0.27267075450532235,
+      "learning_rate": 3.3846560244826854e-07,
+      "loss": 0.6169,
+      "step": 5107
+    },
+    {
+      "epoch": 12.488997555012224,
+      "grad_norm": 0.2593378595025868,
+      "learning_rate": 3.3782266960767556e-07,
+      "loss": 1.075,
+      "step": 5108
+    },
+    {
+      "epoch": 12.491442542787286,
+      "grad_norm": 0.30975157856150604,
+      "learning_rate": 3.371803037394744e-07,
+      "loss": 0.8853,
+      "step": 5109
+    },
+    {
+      "epoch": 12.493887530562347,
+      "grad_norm": 0.3462420209111442,
+      "learning_rate": 3.36538505012107e-07,
+      "loss": 0.3526,
+      "step": 5110
+    },
+    {
+      "epoch": 12.496332518337407,
+      "grad_norm": 0.2823304314844471,
+      "learning_rate": 3.3589727359386787e-07,
+      "loss": 0.4869,
+      "step": 5111
+    },
+    {
+      "epoch": 12.49877750611247,
+      "grad_norm": 0.3963325732300786,
+      "learning_rate": 3.352566096529031e-07,
+      "loss": 0.7501,
+      "step": 5112
+    },
+    {
+      "epoch": 12.50122249388753,
+      "grad_norm": 0.3266685945116019,
+      "learning_rate": 3.346165133572093e-07,
+      "loss": 0.388,
+      "step": 5113
+    },
+    {
+      "epoch": 12.503667481662593,
+      "grad_norm": 0.20047925116983273,
+      "learning_rate": 3.339769848746346e-07,
+      "loss": 0.3427,
+      "step": 5114
+    },
+    {
+      "epoch": 12.506112469437653,
+      "grad_norm": 0.31561336095688586,
+      "learning_rate": 3.333380243728773e-07,
+      "loss": 0.6593,
+      "step": 5115
+    },
+    {
+      "epoch": 12.508557457212714,
+      "grad_norm": 0.21036725355614108,
+      "learning_rate": 3.326996320194878e-07,
+      "loss": 0.7012,
+      "step": 5116
+    },
+    {
+      "epoch": 12.511002444987774,
+      "grad_norm": 0.23328705171523254,
+      "learning_rate": 3.32061807981868e-07,
+      "loss": 0.3479,
+      "step": 5117
+    },
+    {
+      "epoch": 12.513447432762836,
+      "grad_norm": 0.3292262381453281,
+      "learning_rate": 3.314245524272683e-07,
+      "loss": 0.4543,
+      "step": 5118
+    },
+    {
+      "epoch": 12.515892420537897,
+      "grad_norm": 0.2987884360875227,
+      "learning_rate": 3.3078786552279334e-07,
+      "loss": 0.4598,
+      "step": 5119
+    },
+    {
+      "epoch": 12.51833740831296,
+      "grad_norm": 0.2641287563723492,
+      "learning_rate": 3.3015174743539575e-07,
+      "loss": 0.9034,
+      "step": 5120
+    },
+    {
+      "epoch": 12.52078239608802,
+      "grad_norm": 0.38346876829145204,
+      "learning_rate": 3.2951619833188093e-07,
+      "loss": 0.3847,
+      "step": 5121
+    },
+    {
+      "epoch": 12.52322738386308,
+      "grad_norm": 0.21425120542352002,
+      "learning_rate": 3.288812183789039e-07,
+      "loss": 0.8103,
+      "step": 5122
+    },
+    {
+      "epoch": 12.525672371638143,
+      "grad_norm": 0.3610706255456914,
+      "learning_rate": 3.282468077429715e-07,
+      "loss": 0.5151,
+      "step": 5123
+    },
+    {
+      "epoch": 12.528117359413203,
+      "grad_norm": 0.21380993238556853,
+      "learning_rate": 3.276129665904412e-07,
+      "loss": 0.552,
+      "step": 5124
+    },
+    {
+      "epoch": 12.530562347188264,
+      "grad_norm": 0.278544477590112,
+      "learning_rate": 3.2697969508751924e-07,
+      "loss": 0.5829,
+      "step": 5125
+    },
+    {
+      "epoch": 12.533007334963326,
+      "grad_norm": 0.19278899466571883,
+      "learning_rate": 3.2634699340026474e-07,
+      "loss": 0.4843,
+      "step": 5126
+    },
+    {
+      "epoch": 12.535452322738386,
+      "grad_norm": 0.23775038342976557,
+      "learning_rate": 3.2571486169458684e-07,
+      "loss": 0.5256,
+      "step": 5127
+    },
+    {
+      "epoch": 12.537897310513447,
+      "grad_norm": 0.3682607147593088,
+      "learning_rate": 3.2508330013624414e-07,
+      "loss": 0.4077,
+      "step": 5128
+    },
+    {
+      "epoch": 12.54034229828851,
+      "grad_norm": 0.2780970695879525,
+      "learning_rate": 3.2445230889084745e-07,
+      "loss": 0.4224,
+      "step": 5129
+    },
+    {
+      "epoch": 12.54278728606357,
+      "grad_norm": 0.19279153416205014,
+      "learning_rate": 3.238218881238558e-07,
+      "loss": 0.3991,
+      "step": 5130
+    },
+    {
+      "epoch": 12.54523227383863,
+      "grad_norm": 0.18314314776183935,
+      "learning_rate": 3.2319203800058123e-07,
+      "loss": 0.3054,
+      "step": 5131
+    },
+    {
+      "epoch": 12.547677261613693,
+      "grad_norm": 0.34802404020243805,
+      "learning_rate": 3.225627586861849e-07,
+      "loss": 0.7133,
+      "step": 5132
+    },
+    {
+      "epoch": 12.550122249388753,
+      "grad_norm": 0.28021584397509935,
+      "learning_rate": 3.2193405034567637e-07,
+      "loss": 0.4132,
+      "step": 5133
+    },
+    {
+      "epoch": 12.552567237163814,
+      "grad_norm": 0.18543097674854112,
+      "learning_rate": 3.2130591314392e-07,
+      "loss": 0.594,
+      "step": 5134
+    },
+    {
+      "epoch": 12.555012224938876,
+      "grad_norm": 0.3405426231227696,
+      "learning_rate": 3.2067834724562597e-07,
+      "loss": 0.4504,
+      "step": 5135
+    },
+    {
+      "epoch": 12.557457212713937,
+      "grad_norm": 0.24331450462542334,
+      "learning_rate": 3.2005135281535703e-07,
+      "loss": 0.8422,
+      "step": 5136
+    },
+    {
+      "epoch": 12.559902200488997,
+      "grad_norm": 0.2810885896140483,
+      "learning_rate": 3.194249300175259e-07,
+      "loss": 0.4946,
+      "step": 5137
+    },
+    {
+      "epoch": 12.56234718826406,
+      "grad_norm": 0.21835220042502648,
+      "learning_rate": 3.187990790163936e-07,
+      "loss": 0.5129,
+      "step": 5138
+    },
+    {
+      "epoch": 12.56479217603912,
+      "grad_norm": 0.2389457234154542,
+      "learning_rate": 3.181737999760745e-07,
+      "loss": 0.453,
+      "step": 5139
+    },
+    {
+      "epoch": 12.56723716381418,
+      "grad_norm": 0.15684544631973282,
+      "learning_rate": 3.1754909306052905e-07,
+      "loss": 0.4833,
+      "step": 5140
+    },
+    {
+      "epoch": 12.569682151589243,
+      "grad_norm": 0.21536582613224417,
+      "learning_rate": 3.1692495843357096e-07,
+      "loss": 0.6389,
+      "step": 5141
+    },
+    {
+      "epoch": 12.572127139364303,
+      "grad_norm": 0.19929910359836367,
+      "learning_rate": 3.1630139625886293e-07,
+      "loss": 0.4154,
+      "step": 5142
+    },
+    {
+      "epoch": 12.574572127139364,
+      "grad_norm": 0.27472475571797134,
+      "learning_rate": 3.1567840669991526e-07,
+      "loss": 0.9067,
+      "step": 5143
+    },
+    {
+      "epoch": 12.577017114914426,
+      "grad_norm": 0.30476957817864464,
+      "learning_rate": 3.150559899200928e-07,
+      "loss": 0.3928,
+      "step": 5144
+    },
+    {
+      "epoch": 12.579462102689487,
+      "grad_norm": 0.19640029645154206,
+      "learning_rate": 3.144341460826053e-07,
+      "loss": 0.3893,
+      "step": 5145
+    },
+    {
+      "epoch": 12.581907090464547,
+      "grad_norm": 0.31917464931799105,
+      "learning_rate": 3.138128753505157e-07,
+      "loss": 0.3029,
+      "step": 5146
+    },
+    {
+      "epoch": 12.58435207823961,
+      "grad_norm": 0.37799947166298475,
+      "learning_rate": 3.1319217788673526e-07,
+      "loss": 0.4446,
+      "step": 5147
+    },
+    {
+      "epoch": 12.58679706601467,
+      "grad_norm": 0.37987250764314784,
+      "learning_rate": 3.1257205385402444e-07,
+      "loss": 0.6995,
+      "step": 5148
+    },
+    {
+      "epoch": 12.58924205378973,
+      "grad_norm": 0.25492651560285007,
+      "learning_rate": 3.119525034149945e-07,
+      "loss": 0.3702,
+      "step": 5149
+    },
+    {
+      "epoch": 12.591687041564793,
+      "grad_norm": 0.28663233699812285,
+      "learning_rate": 3.113335267321055e-07,
+      "loss": 0.5644,
+      "step": 5150
+    },
+    {
+      "epoch": 12.594132029339853,
+      "grad_norm": 0.17811696660540566,
+      "learning_rate": 3.10715123967667e-07,
+      "loss": 0.3584,
+      "step": 5151
+    },
+    {
+      "epoch": 12.596577017114914,
+      "grad_norm": 0.2625880400575715,
+      "learning_rate": 3.1009729528383935e-07,
+      "loss": 0.6949,
+      "step": 5152
+    },
+    {
+      "epoch": 12.599022004889976,
+      "grad_norm": 0.19809048724931394,
+      "learning_rate": 3.0948004084263015e-07,
+      "loss": 0.6932,
+      "step": 5153
+    },
+    {
+      "epoch": 12.601466992665037,
+      "grad_norm": 0.26855607012489074,
+      "learning_rate": 3.088633608058986e-07,
+      "loss": 1.1185,
+      "step": 5154
+    },
+    {
+      "epoch": 12.603911980440097,
+      "grad_norm": 0.24440314110278524,
+      "learning_rate": 3.082472553353516e-07,
+      "loss": 0.8806,
+      "step": 5155
+    },
+    {
+      "epoch": 12.60635696821516,
+      "grad_norm": 0.20145101056708536,
+      "learning_rate": 3.076317245925464e-07,
+      "loss": 0.6365,
+      "step": 5156
+    },
+    {
+      "epoch": 12.60880195599022,
+      "grad_norm": 0.4212754967308469,
+      "learning_rate": 3.0701676873889e-07,
+      "loss": 0.7532,
+      "step": 5157
+    },
+    {
+      "epoch": 12.61124694376528,
+      "grad_norm": 0.24099640417528337,
+      "learning_rate": 3.0640238793563674e-07,
+      "loss": 0.3521,
+      "step": 5158
+    },
+    {
+      "epoch": 12.613691931540343,
+      "grad_norm": 0.25297117513686623,
+      "learning_rate": 3.057885823438925e-07,
+      "loss": 0.5874,
+      "step": 5159
+    },
+    {
+      "epoch": 12.616136919315403,
+      "grad_norm": 0.14923154617562195,
+      "learning_rate": 3.051753521246095e-07,
+      "loss": 0.5964,
+      "step": 5160
+    },
+    {
+      "epoch": 12.618581907090464,
+      "grad_norm": 0.1941740101197978,
+      "learning_rate": 3.045626974385921e-07,
+      "loss": 0.4657,
+      "step": 5161
+    },
+    {
+      "epoch": 12.621026894865526,
+      "grad_norm": 0.2878315654533411,
+      "learning_rate": 3.0395061844649244e-07,
+      "loss": 0.4447,
+      "step": 5162
+    },
+    {
+      "epoch": 12.623471882640587,
+      "grad_norm": 0.26458991346153937,
+      "learning_rate": 3.0333911530881056e-07,
+      "loss": 0.4358,
+      "step": 5163
+    },
+    {
+      "epoch": 12.625916870415647,
+      "grad_norm": 0.2603447637238707,
+      "learning_rate": 3.0272818818589705e-07,
+      "loss": 0.8532,
+      "step": 5164
+    },
+    {
+      "epoch": 12.62836185819071,
+      "grad_norm": 0.3580342947442217,
+      "learning_rate": 3.0211783723795156e-07,
+      "loss": 0.4814,
+      "step": 5165
+    },
+    {
+      "epoch": 12.63080684596577,
+      "grad_norm": 0.2704413722845608,
+      "learning_rate": 3.0150806262502134e-07,
+      "loss": 0.5383,
+      "step": 5166
+    },
+    {
+      "epoch": 12.63325183374083,
+      "grad_norm": 0.4006365137459578,
+      "learning_rate": 3.0089886450700434e-07,
+      "loss": 0.6651,
+      "step": 5167
+    },
+    {
+      "epoch": 12.635696821515893,
+      "grad_norm": 0.29950743493312165,
+      "learning_rate": 3.002902430436447e-07,
+      "loss": 0.7257,
+      "step": 5168
+    },
+    {
+      "epoch": 12.638141809290953,
+      "grad_norm": 0.2799889912585528,
+      "learning_rate": 2.9968219839453834e-07,
+      "loss": 0.5379,
+      "step": 5169
+    },
+    {
+      "epoch": 12.640586797066014,
+      "grad_norm": 0.28243778253276686,
+      "learning_rate": 2.990747307191275e-07,
+      "loss": 0.6974,
+      "step": 5170
+    },
+    {
+      "epoch": 12.643031784841076,
+      "grad_norm": 0.2509942892720632,
+      "learning_rate": 2.9846784017670436e-07,
+      "loss": 0.6443,
+      "step": 5171
+    },
+    {
+      "epoch": 12.645476772616137,
+      "grad_norm": 0.26492601259450355,
+      "learning_rate": 2.978615269264101e-07,
+      "loss": 0.6146,
+      "step": 5172
+    },
+    {
+      "epoch": 12.647921760391197,
+      "grad_norm": 0.2922459343076512,
+      "learning_rate": 2.972557911272328e-07,
+      "loss": 0.5994,
+      "step": 5173
+    },
+    {
+      "epoch": 12.65036674816626,
+      "grad_norm": 0.3375489669821644,
+      "learning_rate": 2.966506329380112e-07,
+      "loss": 0.4566,
+      "step": 5174
+    },
+    {
+      "epoch": 12.65281173594132,
+      "grad_norm": 0.28047494143061924,
+      "learning_rate": 2.9604605251743134e-07,
+      "loss": 0.4858,
+      "step": 5175
+    },
+    {
+      "epoch": 12.65525672371638,
+      "grad_norm": 0.21266152121398801,
+      "learning_rate": 2.954420500240274e-07,
+      "loss": 0.5052,
+      "step": 5176
+    },
+    {
+      "epoch": 12.657701711491443,
+      "grad_norm": 0.24093700460948447,
+      "learning_rate": 2.9483862561618423e-07,
+      "loss": 0.4744,
+      "step": 5177
+    },
+    {
+      "epoch": 12.660146699266503,
+      "grad_norm": 0.2679128287762201,
+      "learning_rate": 2.9423577945213195e-07,
+      "loss": 0.4327,
+      "step": 5178
+    },
+    {
+      "epoch": 12.662591687041564,
+      "grad_norm": 0.30895518904870495,
+      "learning_rate": 2.9363351168995137e-07,
+      "loss": 0.4151,
+      "step": 5179
+    },
+    {
+      "epoch": 12.665036674816626,
+      "grad_norm": 0.18931732182159922,
+      "learning_rate": 2.930318224875714e-07,
+      "loss": 0.6778,
+      "step": 5180
+    },
+    {
+      "epoch": 12.667481662591687,
+      "grad_norm": 0.232150789310097,
+      "learning_rate": 2.924307120027675e-07,
+      "loss": 0.5085,
+      "step": 5181
+    },
+    {
+      "epoch": 12.669926650366747,
+      "grad_norm": 0.3806923443868992,
+      "learning_rate": 2.918301803931656e-07,
+      "loss": 0.494,
+      "step": 5182
+    },
+    {
+      "epoch": 12.67237163814181,
+      "grad_norm": 0.32740764702296093,
+      "learning_rate": 2.912302278162379e-07,
+      "loss": 0.4642,
+      "step": 5183
+    },
+    {
+      "epoch": 12.67481662591687,
+      "grad_norm": 0.23302610622078251,
+      "learning_rate": 2.906308544293063e-07,
+      "loss": 0.6467,
+      "step": 5184
+    },
+    {
+      "epoch": 12.67726161369193,
+      "grad_norm": 0.22754600499941294,
+      "learning_rate": 2.900320603895407e-07,
+      "loss": 0.4914,
+      "step": 5185
+    },
+    {
+      "epoch": 12.679706601466993,
+      "grad_norm": 0.320721160588856,
+      "learning_rate": 2.8943384585395695e-07,
+      "loss": 0.3906,
+      "step": 5186
+    },
+    {
+      "epoch": 12.682151589242054,
+      "grad_norm": 0.24620231175735657,
+      "learning_rate": 2.8883621097942293e-07,
+      "loss": 0.5993,
+      "step": 5187
+    },
+    {
+      "epoch": 12.684596577017114,
+      "grad_norm": 0.3011897073637522,
+      "learning_rate": 2.882391559226502e-07,
+      "loss": 0.3867,
+      "step": 5188
+    },
+    {
+      "epoch": 12.687041564792176,
+      "grad_norm": 0.2222029109860588,
+      "learning_rate": 2.876426808402011e-07,
+      "loss": 0.4039,
+      "step": 5189
+    },
+    {
+      "epoch": 12.689486552567237,
+      "grad_norm": 0.3329278989083132,
+      "learning_rate": 2.8704678588848537e-07,
+      "loss": 0.3435,
+      "step": 5190
+    },
+    {
+      "epoch": 12.6919315403423,
+      "grad_norm": 0.2002252995695997,
+      "learning_rate": 2.8645147122375955e-07,
+      "loss": 0.4659,
+      "step": 5191
+    },
+    {
+      "epoch": 12.69437652811736,
+      "grad_norm": 0.33216250466344543,
+      "learning_rate": 2.858567370021295e-07,
+      "loss": 0.6221,
+      "step": 5192
+    },
+    {
+      "epoch": 12.69682151589242,
+      "grad_norm": 0.2409126908139336,
+      "learning_rate": 2.8526258337954747e-07,
+      "loss": 0.7108,
+      "step": 5193
+    },
+    {
+      "epoch": 12.69926650366748,
+      "grad_norm": 0.2650223443963618,
+      "learning_rate": 2.8466901051181413e-07,
+      "loss": 0.4551,
+      "step": 5194
+    },
+    {
+      "epoch": 12.701711491442543,
+      "grad_norm": 0.2949734096850486,
+      "learning_rate": 2.8407601855457904e-07,
+      "loss": 0.3564,
+      "step": 5195
+    },
+    {
+      "epoch": 12.704156479217604,
+      "grad_norm": 0.2649792369120882,
+      "learning_rate": 2.8348360766333654e-07,
+      "loss": 0.4216,
+      "step": 5196
+    },
+    {
+      "epoch": 12.706601466992666,
+      "grad_norm": 0.22943497394287962,
+      "learning_rate": 2.828917779934315e-07,
+      "loss": 0.4269,
+      "step": 5197
+    },
+    {
+      "epoch": 12.709046454767726,
+      "grad_norm": 0.32464884793854537,
+      "learning_rate": 2.823005297000547e-07,
+      "loss": 0.3871,
+      "step": 5198
+    },
+    {
+      "epoch": 12.711491442542787,
+      "grad_norm": 0.35896404103761553,
+      "learning_rate": 2.817098629382453e-07,
+      "loss": 0.3829,
+      "step": 5199
+    },
+    {
+      "epoch": 12.713936430317847,
+      "grad_norm": 0.27432890475642124,
+      "learning_rate": 2.811197778628902e-07,
+      "loss": 0.6644,
+      "step": 5200
+    },
+    {
+      "epoch": 12.71638141809291,
+      "grad_norm": 0.25066415365345374,
+      "learning_rate": 2.8053027462872193e-07,
+      "loss": 0.3341,
+      "step": 5201
+    },
+    {
+      "epoch": 12.71882640586797,
+      "grad_norm": 0.28049607441557894,
+      "learning_rate": 2.799413533903228e-07,
+      "loss": 0.4482,
+      "step": 5202
+    },
+    {
+      "epoch": 12.721271393643033,
+      "grad_norm": 0.2146342635734247,
+      "learning_rate": 2.793530143021203e-07,
+      "loss": 0.4462,
+      "step": 5203
+    },
+    {
+      "epoch": 12.723716381418093,
+      "grad_norm": 0.21652588508581377,
+      "learning_rate": 2.7876525751839136e-07,
+      "loss": 0.4361,
+      "step": 5204
+    },
+    {
+      "epoch": 12.726161369193154,
+      "grad_norm": 0.2823082625689088,
+      "learning_rate": 2.7817808319325957e-07,
+      "loss": 0.416,
+      "step": 5205
+    },
+    {
+      "epoch": 12.728606356968216,
+      "grad_norm": 0.3781358207524691,
+      "learning_rate": 2.7759149148069405e-07,
+      "loss": 0.4373,
+      "step": 5206
+    },
+    {
+      "epoch": 12.731051344743276,
+      "grad_norm": 0.2116006880177325,
+      "learning_rate": 2.770054825345142e-07,
+      "loss": 0.3855,
+      "step": 5207
+    },
+    {
+      "epoch": 12.733496332518337,
+      "grad_norm": 0.3615709300162841,
+      "learning_rate": 2.7642005650838327e-07,
+      "loss": 0.4363,
+      "step": 5208
+    },
+    {
+      "epoch": 12.7359413202934,
+      "grad_norm": 0.1852833977046477,
+      "learning_rate": 2.7583521355581456e-07,
+      "loss": 0.4038,
+      "step": 5209
+    },
+    {
+      "epoch": 12.73838630806846,
+      "grad_norm": 0.16976770063026442,
+      "learning_rate": 2.7525095383016756e-07,
+      "loss": 0.3218,
+      "step": 5210
+    },
+    {
+      "epoch": 12.74083129584352,
+      "grad_norm": 0.32496451054913517,
+      "learning_rate": 2.746672774846476e-07,
+      "loss": 0.7784,
+      "step": 5211
+    },
+    {
+      "epoch": 12.743276283618583,
+      "grad_norm": 0.21009765279073134,
+      "learning_rate": 2.740841846723091e-07,
+      "loss": 0.595,
+      "step": 5212
+    },
+    {
+      "epoch": 12.745721271393643,
+      "grad_norm": 0.25883741429036217,
+      "learning_rate": 2.7350167554605095e-07,
+      "loss": 0.4179,
+      "step": 5213
+    },
+    {
+      "epoch": 12.748166259168704,
+      "grad_norm": 0.25094756444846045,
+      "learning_rate": 2.729197502586209e-07,
+      "loss": 0.8024,
+      "step": 5214
+    },
+    {
+      "epoch": 12.750611246943766,
+      "grad_norm": 0.2780265350247338,
+      "learning_rate": 2.7233840896261383e-07,
+      "loss": 0.4074,
+      "step": 5215
+    },
+    {
+      "epoch": 12.753056234718827,
+      "grad_norm": 0.2514472285340264,
+      "learning_rate": 2.7175765181046976e-07,
+      "loss": 0.5315,
+      "step": 5216
+    },
+    {
+      "epoch": 12.755501222493887,
+      "grad_norm": 0.22773834296021908,
+      "learning_rate": 2.711774789544774e-07,
+      "loss": 0.4848,
+      "step": 5217
+    },
+    {
+      "epoch": 12.75794621026895,
+      "grad_norm": 0.30949999053136407,
+      "learning_rate": 2.705978905467699e-07,
+      "loss": 0.6573,
+      "step": 5218
+    },
+    {
+      "epoch": 12.76039119804401,
+      "grad_norm": 0.2688742350389799,
+      "learning_rate": 2.7001888673932997e-07,
+      "loss": 0.4817,
+      "step": 5219
+    },
+    {
+      "epoch": 12.76283618581907,
+      "grad_norm": 0.26348197540349444,
+      "learning_rate": 2.694404676839857e-07,
+      "loss": 0.496,
+      "step": 5220
+    },
+    {
+      "epoch": 12.765281173594133,
+      "grad_norm": 0.3018462759952364,
+      "learning_rate": 2.688626335324107e-07,
+      "loss": 0.6579,
+      "step": 5221
+    },
+    {
+      "epoch": 12.767726161369193,
+      "grad_norm": 0.24513101512945973,
+      "learning_rate": 2.6828538443612775e-07,
+      "loss": 0.4866,
+      "step": 5222
+    },
+    {
+      "epoch": 12.770171149144254,
+      "grad_norm": 0.3155327842275826,
+      "learning_rate": 2.6770872054650346e-07,
+      "loss": 0.676,
+      "step": 5223
+    },
+    {
+      "epoch": 12.772616136919316,
+      "grad_norm": 0.2754944703839549,
+      "learning_rate": 2.671326420147527e-07,
+      "loss": 1.4302,
+      "step": 5224
+    },
+    {
+      "epoch": 12.775061124694377,
+      "grad_norm": 0.2482185391421997,
+      "learning_rate": 2.6655714899193747e-07,
+      "loss": 0.5609,
+      "step": 5225
+    },
+    {
+      "epoch": 12.777506112469437,
+      "grad_norm": 0.2767043381750146,
+      "learning_rate": 2.659822416289634e-07,
+      "loss": 0.5948,
+      "step": 5226
+    },
+    {
+      "epoch": 12.7799511002445,
+      "grad_norm": 0.2739869438031292,
+      "learning_rate": 2.6540792007658604e-07,
+      "loss": 1.2608,
+      "step": 5227
+    },
+    {
+      "epoch": 12.78239608801956,
+      "grad_norm": 0.24464727717917417,
+      "learning_rate": 2.6483418448540436e-07,
+      "loss": 0.5985,
+      "step": 5228
+    },
+    {
+      "epoch": 12.78484107579462,
+      "grad_norm": 0.1362553830402673,
+      "learning_rate": 2.6426103500586497e-07,
+      "loss": 0.3043,
+      "step": 5229
+    },
+    {
+      "epoch": 12.787286063569683,
+      "grad_norm": 0.22163382716444335,
+      "learning_rate": 2.6368847178826206e-07,
+      "loss": 0.5105,
+      "step": 5230
+    },
+    {
+      "epoch": 12.789731051344743,
+      "grad_norm": 0.3266005319060546,
+      "learning_rate": 2.631164949827339e-07,
+      "loss": 0.4125,
+      "step": 5231
+    },
+    {
+      "epoch": 12.792176039119804,
+      "grad_norm": 0.24664218267627727,
+      "learning_rate": 2.625451047392666e-07,
+      "loss": 0.4946,
+      "step": 5232
+    },
+    {
+      "epoch": 12.794621026894866,
+      "grad_norm": 0.37974253690985504,
+      "learning_rate": 2.6197430120769057e-07,
+      "loss": 0.7891,
+      "step": 5233
+    },
+    {
+      "epoch": 12.797066014669927,
+      "grad_norm": 0.24601464662923175,
+      "learning_rate": 2.6140408453768443e-07,
+      "loss": 0.3651,
+      "step": 5234
+    },
+    {
+      "epoch": 12.799511002444987,
+      "grad_norm": 0.23614770743011776,
+      "learning_rate": 2.608344548787722e-07,
+      "loss": 0.7393,
+      "step": 5235
+    },
+    {
+      "epoch": 12.80195599022005,
+      "grad_norm": 0.2105610735170176,
+      "learning_rate": 2.6026541238032305e-07,
+      "loss": 0.4872,
+      "step": 5236
+    },
+    {
+      "epoch": 12.80440097799511,
+      "grad_norm": 0.3014705821190384,
+      "learning_rate": 2.5969695719155306e-07,
+      "loss": 0.7052,
+      "step": 5237
+    },
+    {
+      "epoch": 12.80684596577017,
+      "grad_norm": 0.2993352531596114,
+      "learning_rate": 2.591290894615253e-07,
+      "loss": 0.6453,
+      "step": 5238
+    },
+    {
+      "epoch": 12.809290953545233,
+      "grad_norm": 0.29063579728885197,
+      "learning_rate": 2.585618093391465e-07,
+      "loss": 0.4788,
+      "step": 5239
+    },
+    {
+      "epoch": 12.811735941320293,
+      "grad_norm": 0.18878750735181637,
+      "learning_rate": 2.5799511697317063e-07,
+      "loss": 0.5009,
+      "step": 5240
+    },
+    {
+      "epoch": 12.814180929095354,
+      "grad_norm": 0.19710273983925064,
+      "learning_rate": 2.5742901251219755e-07,
+      "loss": 0.6211,
+      "step": 5241
+    },
+    {
+      "epoch": 12.816625916870416,
+      "grad_norm": 0.3069539859022823,
+      "learning_rate": 2.568634961046737e-07,
+      "loss": 0.5093,
+      "step": 5242
+    },
+    {
+      "epoch": 12.819070904645477,
+      "grad_norm": 0.23243765010478865,
+      "learning_rate": 2.5629856789888886e-07,
+      "loss": 0.4658,
+      "step": 5243
+    },
+    {
+      "epoch": 12.821515892420537,
+      "grad_norm": 0.18214907839241887,
+      "learning_rate": 2.5573422804298116e-07,
+      "loss": 0.5782,
+      "step": 5244
+    },
+    {
+      "epoch": 12.8239608801956,
+      "grad_norm": 0.279185238796717,
+      "learning_rate": 2.551704766849336e-07,
+      "loss": 0.3215,
+      "step": 5245
+    },
+    {
+      "epoch": 12.82640586797066,
+      "grad_norm": 0.438622213658891,
+      "learning_rate": 2.546073139725735e-07,
+      "loss": 0.8295,
+      "step": 5246
+    },
+    {
+      "epoch": 12.82885085574572,
+      "grad_norm": 0.319508634353802,
+      "learning_rate": 2.5404474005357616e-07,
+      "loss": 0.7422,
+      "step": 5247
+    },
+    {
+      "epoch": 12.831295843520783,
+      "grad_norm": 0.18144734016536387,
+      "learning_rate": 2.534827550754612e-07,
+      "loss": 0.6788,
+      "step": 5248
+    },
+    {
+      "epoch": 12.833740831295843,
+      "grad_norm": 0.3149596625076904,
+      "learning_rate": 2.5292135918559336e-07,
+      "loss": 0.4859,
+      "step": 5249
+    },
+    {
+      "epoch": 12.836185819070904,
+      "grad_norm": 0.18640102926270288,
+      "learning_rate": 2.5236055253118427e-07,
+      "loss": 0.4601,
+      "step": 5250
+    },
+    {
+      "epoch": 12.838630806845966,
+      "grad_norm": 0.19884349237845794,
+      "learning_rate": 2.5180033525928903e-07,
+      "loss": 0.8453,
+      "step": 5251
+    },
+    {
+      "epoch": 12.841075794621027,
+      "grad_norm": 0.30900307263420557,
+      "learning_rate": 2.5124070751681105e-07,
+      "loss": 0.4911,
+      "step": 5252
+    },
+    {
+      "epoch": 12.843520782396087,
+      "grad_norm": 0.260310064411079,
+      "learning_rate": 2.5068166945049626e-07,
+      "loss": 0.4059,
+      "step": 5253
+    },
+    {
+      "epoch": 12.84596577017115,
+      "grad_norm": 0.2758246031019974,
+      "learning_rate": 2.5012322120693767e-07,
+      "loss": 1.0052,
+      "step": 5254
+    },
+    {
+      "epoch": 12.84841075794621,
+      "grad_norm": 0.30640059668481695,
+      "learning_rate": 2.4956536293257395e-07,
+      "loss": 0.4693,
+      "step": 5255
+    },
+    {
+      "epoch": 12.85085574572127,
+      "grad_norm": 0.21853809598914828,
+      "learning_rate": 2.49008094773687e-07,
+      "loss": 0.5638,
+      "step": 5256
+    },
+    {
+      "epoch": 12.853300733496333,
+      "grad_norm": 0.19853819714831356,
+      "learning_rate": 2.4845141687640575e-07,
+      "loss": 0.3486,
+      "step": 5257
+    },
+    {
+      "epoch": 12.855745721271393,
+      "grad_norm": 0.4405351296806944,
+      "learning_rate": 2.478953293867045e-07,
+      "loss": 0.5163,
+      "step": 5258
+    },
+    {
+      "epoch": 12.858190709046454,
+      "grad_norm": 0.22562241260470897,
+      "learning_rate": 2.473398324504014e-07,
+      "loss": 0.3981,
+      "step": 5259
+    },
+    {
+      "epoch": 12.860635696821516,
+      "grad_norm": 0.22175161090902956,
+      "learning_rate": 2.467849262131608e-07,
+      "loss": 0.4365,
+      "step": 5260
+    },
+    {
+      "epoch": 12.863080684596577,
+      "grad_norm": 0.24053610884563748,
+      "learning_rate": 2.462306108204912e-07,
+      "loss": 0.7823,
+      "step": 5261
+    },
+    {
+      "epoch": 12.865525672371637,
+      "grad_norm": 0.30698792317480383,
+      "learning_rate": 2.4567688641774807e-07,
+      "loss": 0.5416,
+      "step": 5262
+    },
+    {
+      "epoch": 12.8679706601467,
+      "grad_norm": 0.27238465007926205,
+      "learning_rate": 2.4512375315012936e-07,
+      "loss": 0.9008,
+      "step": 5263
+    },
+    {
+      "epoch": 12.87041564792176,
+      "grad_norm": 0.16045767618149717,
+      "learning_rate": 2.445712111626797e-07,
+      "loss": 0.4154,
+      "step": 5264
+    },
+    {
+      "epoch": 12.87286063569682,
+      "grad_norm": 0.23513368548845895,
+      "learning_rate": 2.440192606002889e-07,
+      "loss": 0.6774,
+      "step": 5265
+    },
+    {
+      "epoch": 12.875305623471883,
+      "grad_norm": 0.29438278935875534,
+      "learning_rate": 2.434679016076899e-07,
+      "loss": 0.5803,
+      "step": 5266
+    },
+    {
+      "epoch": 12.877750611246944,
+      "grad_norm": 0.346545047447368,
+      "learning_rate": 2.429171343294623e-07,
+      "loss": 0.4551,
+      "step": 5267
+    },
+    {
+      "epoch": 12.880195599022004,
+      "grad_norm": 0.3661193023855829,
+      "learning_rate": 2.423669589100303e-07,
+      "loss": 0.4447,
+      "step": 5268
+    },
+    {
+      "epoch": 12.882640586797066,
+      "grad_norm": 0.18746366701342593,
+      "learning_rate": 2.4181737549366116e-07,
+      "loss": 0.4685,
+      "step": 5269
+    },
+    {
+      "epoch": 12.885085574572127,
+      "grad_norm": 0.4406280828186168,
+      "learning_rate": 2.412683842244698e-07,
+      "loss": 0.7355,
+      "step": 5270
+    },
+    {
+      "epoch": 12.887530562347187,
+      "grad_norm": 0.22495259508090046,
+      "learning_rate": 2.4071998524641316e-07,
+      "loss": 0.5261,
+      "step": 5271
+    },
+    {
+      "epoch": 12.88997555012225,
+      "grad_norm": 0.2764002953090943,
+      "learning_rate": 2.401721787032943e-07,
+      "loss": 0.4324,
+      "step": 5272
+    },
+    {
+      "epoch": 12.89242053789731,
+      "grad_norm": 0.3244219316132107,
+      "learning_rate": 2.396249647387608e-07,
+      "loss": 0.5819,
+      "step": 5273
+    },
+    {
+      "epoch": 12.894865525672373,
+      "grad_norm": 0.29823243135004235,
+      "learning_rate": 2.390783434963045e-07,
+      "loss": 0.3971,
+      "step": 5274
+    },
+    {
+      "epoch": 12.897310513447433,
+      "grad_norm": 0.38379828998583276,
+      "learning_rate": 2.3853231511926254e-07,
+      "loss": 0.4241,
+      "step": 5275
+    },
+    {
+      "epoch": 12.899755501222494,
+      "grad_norm": 0.2449225238525846,
+      "learning_rate": 2.379868797508153e-07,
+      "loss": 0.4387,
+      "step": 5276
+    },
+    {
+      "epoch": 12.902200488997554,
+      "grad_norm": 0.257087109060615,
+      "learning_rate": 2.374420375339889e-07,
+      "loss": 0.5999,
+      "step": 5277
+    },
+    {
+      "epoch": 12.904645476772616,
+      "grad_norm": 0.21015208319102757,
+      "learning_rate": 2.368977886116536e-07,
+      "loss": 0.8812,
+      "step": 5278
+    },
+    {
+      "epoch": 12.907090464547677,
+      "grad_norm": 0.24517387618174827,
+      "learning_rate": 2.3635413312652317e-07,
+      "loss": 0.329,
+      "step": 5279
+    },
+    {
+      "epoch": 12.90953545232274,
+      "grad_norm": 0.19854802905360885,
+      "learning_rate": 2.3581107122115727e-07,
+      "loss": 0.5393,
+      "step": 5280
+    },
+    {
+      "epoch": 12.9119804400978,
+      "grad_norm": 0.22126762132374855,
+      "learning_rate": 2.3526860303795862e-07,
+      "loss": 1.0191,
+      "step": 5281
+    },
+    {
+      "epoch": 12.91442542787286,
+      "grad_norm": 0.29150296524215424,
+      "learning_rate": 2.3472672871917512e-07,
+      "loss": 0.7109,
+      "step": 5282
+    },
+    {
+      "epoch": 12.91687041564792,
+      "grad_norm": 0.1676333631320654,
+      "learning_rate": 2.341854484068984e-07,
+      "loss": 0.3989,
+      "step": 5283
+    },
+    {
+      "epoch": 12.919315403422983,
+      "grad_norm": 0.26081201047253366,
+      "learning_rate": 2.3364476224306487e-07,
+      "loss": 0.3281,
+      "step": 5284
+    },
+    {
+      "epoch": 12.921760391198044,
+      "grad_norm": 0.3476032310703982,
+      "learning_rate": 2.3310467036945506e-07,
+      "loss": 0.4529,
+      "step": 5285
+    },
+    {
+      "epoch": 12.924205378973106,
+      "grad_norm": 0.3573125308026871,
+      "learning_rate": 2.3256517292769233e-07,
+      "loss": 0.95,
+      "step": 5286
+    },
+    {
+      "epoch": 12.926650366748166,
+      "grad_norm": 0.24438454382884092,
+      "learning_rate": 2.32026270059246e-07,
+      "loss": 0.3649,
+      "step": 5287
+    },
+    {
+      "epoch": 12.929095354523227,
+      "grad_norm": 0.2834501584720449,
+      "learning_rate": 2.3148796190542944e-07,
+      "loss": 0.5657,
+      "step": 5288
+    },
+    {
+      "epoch": 12.93154034229829,
+      "grad_norm": 0.37949541056886554,
+      "learning_rate": 2.3095024860739778e-07,
+      "loss": 0.5565,
+      "step": 5289
+    },
+    {
+      "epoch": 12.93398533007335,
+      "grad_norm": 0.3100830799379428,
+      "learning_rate": 2.3041313030615326e-07,
+      "loss": 0.4378,
+      "step": 5290
+    },
+    {
+      "epoch": 12.93643031784841,
+      "grad_norm": 0.3318476031877716,
+      "learning_rate": 2.2987660714253957e-07,
+      "loss": 0.7293,
+      "step": 5291
+    },
+    {
+      "epoch": 12.938875305623473,
+      "grad_norm": 0.2922565759173333,
+      "learning_rate": 2.2934067925724556e-07,
+      "loss": 0.4255,
+      "step": 5292
+    },
+    {
+      "epoch": 12.941320293398533,
+      "grad_norm": 0.32286458237381965,
+      "learning_rate": 2.288053467908044e-07,
+      "loss": 0.8685,
+      "step": 5293
+    },
+    {
+      "epoch": 12.943765281173594,
+      "grad_norm": 0.21848679340543833,
+      "learning_rate": 2.28270609883591e-07,
+      "loss": 0.6735,
+      "step": 5294
+    },
+    {
+      "epoch": 12.946210268948656,
+      "grad_norm": 0.3104114709910301,
+      "learning_rate": 2.2773646867582766e-07,
+      "loss": 0.3946,
+      "step": 5295
+    },
+    {
+      "epoch": 12.948655256723717,
+      "grad_norm": 0.3432079540998316,
+      "learning_rate": 2.2720292330757732e-07,
+      "loss": 0.4689,
+      "step": 5296
+    },
+    {
+      "epoch": 12.951100244498777,
+      "grad_norm": 0.35220692407940674,
+      "learning_rate": 2.266699739187475e-07,
+      "loss": 0.3511,
+      "step": 5297
+    },
+    {
+      "epoch": 12.95354523227384,
+      "grad_norm": 0.2123614177686676,
+      "learning_rate": 2.2613762064909088e-07,
+      "loss": 0.935,
+      "step": 5298
+    },
+    {
+      "epoch": 12.9559902200489,
+      "grad_norm": 0.22134645456079946,
+      "learning_rate": 2.2560586363820136e-07,
+      "loss": 0.8635,
+      "step": 5299
+    },
+    {
+      "epoch": 12.95843520782396,
+      "grad_norm": 0.26908719672239767,
+      "learning_rate": 2.2507470302551875e-07,
+      "loss": 0.8641,
+      "step": 5300
+    },
+    {
+      "epoch": 12.960880195599023,
+      "grad_norm": 0.2664152416064593,
+      "learning_rate": 2.2454413895032502e-07,
+      "loss": 0.4228,
+      "step": 5301
+    },
+    {
+      "epoch": 12.963325183374083,
+      "grad_norm": 0.24451647326137868,
+      "learning_rate": 2.2401417155174603e-07,
+      "loss": 0.8422,
+      "step": 5302
+    },
+    {
+      "epoch": 12.965770171149144,
+      "grad_norm": 0.2735374076193863,
+      "learning_rate": 2.2348480096875236e-07,
+      "loss": 0.739,
+      "step": 5303
+    },
+    {
+      "epoch": 12.968215158924206,
+      "grad_norm": 0.26235278067517126,
+      "learning_rate": 2.2295602734015543e-07,
+      "loss": 1.2821,
+      "step": 5304
+    },
+    {
+      "epoch": 12.970660146699267,
+      "grad_norm": 0.20280595631051063,
+      "learning_rate": 2.2242785080461348e-07,
+      "loss": 0.4569,
+      "step": 5305
+    },
+    {
+      "epoch": 12.973105134474327,
+      "grad_norm": 0.20909248654981144,
+      "learning_rate": 2.2190027150062577e-07,
+      "loss": 0.4857,
+      "step": 5306
+    },
+    {
+      "epoch": 12.97555012224939,
+      "grad_norm": 0.16615060105985366,
+      "learning_rate": 2.2137328956653547e-07,
+      "loss": 0.3971,
+      "step": 5307
+    },
+    {
+      "epoch": 12.97799511002445,
+      "grad_norm": 0.2905958825547304,
+      "learning_rate": 2.2084690514052988e-07,
+      "loss": 0.49,
+      "step": 5308
+    },
+    {
+      "epoch": 12.98044009779951,
+      "grad_norm": 0.2515527370419866,
+      "learning_rate": 2.2032111836063824e-07,
+      "loss": 0.3889,
+      "step": 5309
+    },
+    {
+      "epoch": 12.982885085574573,
+      "grad_norm": 0.29437477432934656,
+      "learning_rate": 2.1979592936473503e-07,
+      "loss": 1.102,
+      "step": 5310
+    },
+    {
+      "epoch": 12.985330073349633,
+      "grad_norm": 0.3444615176684247,
+      "learning_rate": 2.1927133829053539e-07,
+      "loss": 0.8434,
+      "step": 5311
+    },
+    {
+      "epoch": 12.987775061124694,
+      "grad_norm": 0.22372589132459014,
+      "learning_rate": 2.1874734527560004e-07,
+      "loss": 0.4278,
+      "step": 5312
+    },
+    {
+      "epoch": 12.990220048899756,
+      "grad_norm": 0.31864551462775814,
+      "learning_rate": 2.1822395045733219e-07,
+      "loss": 0.6109,
+      "step": 5313
+    },
+    {
+      "epoch": 12.992665036674817,
+      "grad_norm": 0.18661989727995962,
+      "learning_rate": 2.1770115397297676e-07,
+      "loss": 0.7362,
+      "step": 5314
+    },
+    {
+      "epoch": 12.995110024449877,
+      "grad_norm": 0.3008802839931786,
+      "learning_rate": 2.1717895595962407e-07,
+      "loss": 0.3169,
+      "step": 5315
+    },
+    {
+      "epoch": 12.99755501222494,
+      "grad_norm": 0.21947199878524767,
+      "learning_rate": 2.1665735655420572e-07,
+      "loss": 0.7924,
+      "step": 5316
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.19176785869123447,
+      "learning_rate": 2.1613635589349756e-07,
+      "loss": 0.7732,
+      "step": 5317
+    },
+    {
+      "epoch": 13.00244498777506,
+      "grad_norm": 0.2507008404251588,
+      "learning_rate": 2.156159541141181e-07,
+      "loss": 0.4056,
+      "step": 5318
+    },
+    {
+      "epoch": 13.004889975550123,
+      "grad_norm": 0.23214940599424377,
+      "learning_rate": 2.1509615135252765e-07,
+      "loss": 0.7119,
+      "step": 5319
+    },
+    {
+      "epoch": 13.007334963325183,
+      "grad_norm": 0.3312498623122705,
+      "learning_rate": 2.145769477450313e-07,
+      "loss": 0.6496,
+      "step": 5320
+    },
+    {
+      "epoch": 13.009779951100244,
+      "grad_norm": 0.30629841105680783,
+      "learning_rate": 2.1405834342777514e-07,
+      "loss": 0.4458,
+      "step": 5321
+    },
+    {
+      "epoch": 13.012224938875306,
+      "grad_norm": 0.274052827589616,
+      "learning_rate": 2.1354033853674988e-07,
+      "loss": 0.4463,
+      "step": 5322
+    },
+    {
+      "epoch": 13.014669926650367,
+      "grad_norm": 0.2445954854397924,
+      "learning_rate": 2.1302293320778823e-07,
+      "loss": 0.4283,
+      "step": 5323
+    },
+    {
+      "epoch": 13.017114914425427,
+      "grad_norm": 0.20344198660713717,
+      "learning_rate": 2.1250612757656524e-07,
+      "loss": 0.7197,
+      "step": 5324
+    },
+    {
+      "epoch": 13.01955990220049,
+      "grad_norm": 0.21298510893567818,
+      "learning_rate": 2.119899217785995e-07,
+      "loss": 0.9394,
+      "step": 5325
+    },
+    {
+      "epoch": 13.02200488997555,
+      "grad_norm": 0.2399602456370534,
+      "learning_rate": 2.114743159492519e-07,
+      "loss": 0.4625,
+      "step": 5326
+    },
+    {
+      "epoch": 13.02444987775061,
+      "grad_norm": 0.3172765932094714,
+      "learning_rate": 2.1095931022372617e-07,
+      "loss": 0.8134,
+      "step": 5327
+    },
+    {
+      "epoch": 13.026894865525673,
+      "grad_norm": 0.45600522058629817,
+      "learning_rate": 2.104449047370691e-07,
+      "loss": 0.3351,
+      "step": 5328
+    },
+    {
+      "epoch": 13.029339853300733,
+      "grad_norm": 0.22104308683857282,
+      "learning_rate": 2.0993109962416825e-07,
+      "loss": 0.6739,
+      "step": 5329
+    },
+    {
+      "epoch": 13.031784841075794,
+      "grad_norm": 0.3645849599527093,
+      "learning_rate": 2.0941789501975674e-07,
+      "loss": 0.5293,
+      "step": 5330
+    },
+    {
+      "epoch": 13.034229828850856,
+      "grad_norm": 0.2010478825834421,
+      "learning_rate": 2.0890529105840718e-07,
+      "loss": 0.5359,
+      "step": 5331
+    },
+    {
+      "epoch": 13.036674816625917,
+      "grad_norm": 0.32287510369430916,
+      "learning_rate": 2.083932878745365e-07,
+      "loss": 0.4544,
+      "step": 5332
+    },
+    {
+      "epoch": 13.039119804400977,
+      "grad_norm": 0.2751944598769133,
+      "learning_rate": 2.0788188560240424e-07,
+      "loss": 0.4229,
+      "step": 5333
+    },
+    {
+      "epoch": 13.04156479217604,
+      "grad_norm": 0.3539152542983435,
+      "learning_rate": 2.0737108437611063e-07,
+      "loss": 0.7066,
+      "step": 5334
+    },
+    {
+      "epoch": 13.0440097799511,
+      "grad_norm": 0.18101064267533062,
+      "learning_rate": 2.0686088432960023e-07,
+      "loss": 0.4297,
+      "step": 5335
+    },
+    {
+      "epoch": 13.04645476772616,
+      "grad_norm": 0.23542239064179407,
+      "learning_rate": 2.063512855966593e-07,
+      "loss": 0.5085,
+      "step": 5336
+    },
+    {
+      "epoch": 13.048899755501223,
+      "grad_norm": 0.28634667190884916,
+      "learning_rate": 2.0584228831091568e-07,
+      "loss": 1.0158,
+      "step": 5337
+    },
+    {
+      "epoch": 13.051344743276283,
+      "grad_norm": 0.238126699886235,
+      "learning_rate": 2.0533389260584125e-07,
+      "loss": 0.6384,
+      "step": 5338
+    },
+    {
+      "epoch": 13.053789731051344,
+      "grad_norm": 0.23313225719223077,
+      "learning_rate": 2.0482609861474768e-07,
+      "loss": 0.4673,
+      "step": 5339
+    },
+    {
+      "epoch": 13.056234718826406,
+      "grad_norm": 0.2861646552161169,
+      "learning_rate": 2.0431890647079094e-07,
+      "loss": 0.3791,
+      "step": 5340
+    },
+    {
+      "epoch": 13.058679706601467,
+      "grad_norm": 0.4015048471701256,
+      "learning_rate": 2.0381231630696803e-07,
+      "loss": 0.3763,
+      "step": 5341
+    },
+    {
+      "epoch": 13.061124694376527,
+      "grad_norm": 0.3157175129837912,
+      "learning_rate": 2.0330632825611847e-07,
+      "loss": 0.4153,
+      "step": 5342
+    },
+    {
+      "epoch": 13.06356968215159,
+      "grad_norm": 0.4269116108957552,
+      "learning_rate": 2.0280094245092473e-07,
+      "loss": 0.4999,
+      "step": 5343
+    },
+    {
+      "epoch": 13.06601466992665,
+      "grad_norm": 0.43000251455104793,
+      "learning_rate": 2.022961590239092e-07,
+      "loss": 0.3838,
+      "step": 5344
+    },
+    {
+      "epoch": 13.06845965770171,
+      "grad_norm": 0.27999802326962814,
+      "learning_rate": 2.0179197810743844e-07,
+      "loss": 0.4955,
+      "step": 5345
+    },
+    {
+      "epoch": 13.070904645476773,
+      "grad_norm": 0.18722740493649925,
+      "learning_rate": 2.0128839983372089e-07,
+      "loss": 0.4067,
+      "step": 5346
+    },
+    {
+      "epoch": 13.073349633251834,
+      "grad_norm": 0.3547853253482003,
+      "learning_rate": 2.0078542433480424e-07,
+      "loss": 0.3308,
+      "step": 5347
+    },
+    {
+      "epoch": 13.075794621026894,
+      "grad_norm": 0.30296289033957463,
+      "learning_rate": 2.0028305174258271e-07,
+      "loss": 1.0912,
+      "step": 5348
+    },
+    {
+      "epoch": 13.078239608801956,
+      "grad_norm": 0.2684021167343446,
+      "learning_rate": 1.9978128218878818e-07,
+      "loss": 0.3196,
+      "step": 5349
+    },
+    {
+      "epoch": 13.080684596577017,
+      "grad_norm": 0.18875785564764416,
+      "learning_rate": 1.9928011580499673e-07,
+      "loss": 0.6256,
+      "step": 5350
+    },
+    {
+      "epoch": 13.083129584352077,
+      "grad_norm": 0.27504183176476815,
+      "learning_rate": 1.987795527226255e-07,
+      "loss": 0.5669,
+      "step": 5351
+    },
+    {
+      "epoch": 13.08557457212714,
+      "grad_norm": 0.3561250770631697,
+      "learning_rate": 1.9827959307293338e-07,
+      "loss": 0.3427,
+      "step": 5352
+    },
+    {
+      "epoch": 13.0880195599022,
+      "grad_norm": 0.1853786784129405,
+      "learning_rate": 1.9778023698702213e-07,
+      "loss": 0.7762,
+      "step": 5353
+    },
+    {
+      "epoch": 13.09046454767726,
+      "grad_norm": 0.2876215987968845,
+      "learning_rate": 1.972814845958332e-07,
+      "loss": 1.1142,
+      "step": 5354
+    },
+    {
+      "epoch": 13.092909535452323,
+      "grad_norm": 0.28513341283044574,
+      "learning_rate": 1.9678333603015133e-07,
+      "loss": 0.702,
+      "step": 5355
+    },
+    {
+      "epoch": 13.095354523227384,
+      "grad_norm": 0.42181986000768024,
+      "learning_rate": 1.962857914206029e-07,
+      "loss": 0.3485,
+      "step": 5356
+    },
+    {
+      "epoch": 13.097799511002446,
+      "grad_norm": 0.43440265389411786,
+      "learning_rate": 1.9578885089765465e-07,
+      "loss": 0.9741,
+      "step": 5357
+    },
+    {
+      "epoch": 13.100244498777506,
+      "grad_norm": 0.2794655809888336,
+      "learning_rate": 1.9529251459161653e-07,
+      "loss": 0.6586,
+      "step": 5358
+    },
+    {
+      "epoch": 13.102689486552567,
+      "grad_norm": 0.17775960780612599,
+      "learning_rate": 1.9479678263263858e-07,
+      "loss": 0.4277,
+      "step": 5359
+    },
+    {
+      "epoch": 13.10513447432763,
+      "grad_norm": 0.2680588259617518,
+      "learning_rate": 1.943016551507143e-07,
+      "loss": 0.4925,
+      "step": 5360
+    },
+    {
+      "epoch": 13.10757946210269,
+      "grad_norm": 0.39906458816742285,
+      "learning_rate": 1.93807132275676e-07,
+      "loss": 0.4112,
+      "step": 5361
+    },
+    {
+      "epoch": 13.11002444987775,
+      "grad_norm": 0.30471016300707354,
+      "learning_rate": 1.933132141371996e-07,
+      "loss": 0.4946,
+      "step": 5362
+    },
+    {
+      "epoch": 13.112469437652813,
+      "grad_norm": 0.2098381639596224,
+      "learning_rate": 1.9281990086480186e-07,
+      "loss": 0.4547,
+      "step": 5363
+    },
+    {
+      "epoch": 13.114914425427873,
+      "grad_norm": 0.2104823333765495,
+      "learning_rate": 1.9232719258784032e-07,
+      "loss": 0.3506,
+      "step": 5364
+    },
+    {
+      "epoch": 13.117359413202934,
+      "grad_norm": 0.3558425152259091,
+      "learning_rate": 1.9183508943551478e-07,
+      "loss": 0.6394,
+      "step": 5365
+    },
+    {
+      "epoch": 13.119804400977996,
+      "grad_norm": 0.13938805273559338,
+      "learning_rate": 1.913435915368664e-07,
+      "loss": 0.3192,
+      "step": 5366
+    },
+    {
+      "epoch": 13.122249388753056,
+      "grad_norm": 0.26666424210265416,
+      "learning_rate": 1.9085269902077596e-07,
+      "loss": 0.5769,
+      "step": 5367
+    },
+    {
+      "epoch": 13.124694376528117,
+      "grad_norm": 0.2485025781861825,
+      "learning_rate": 1.9036241201596777e-07,
+      "loss": 1.1123,
+      "step": 5368
+    },
+    {
+      "epoch": 13.12713936430318,
+      "grad_norm": 0.2848209180720596,
+      "learning_rate": 1.898727306510051e-07,
+      "loss": 0.4232,
+      "step": 5369
+    },
+    {
+      "epoch": 13.12958435207824,
+      "grad_norm": 0.22310877988146,
+      "learning_rate": 1.8938365505429546e-07,
+      "loss": 0.4586,
+      "step": 5370
+    },
+    {
+      "epoch": 13.1320293398533,
+      "grad_norm": 0.2781800836351792,
+      "learning_rate": 1.8889518535408408e-07,
+      "loss": 0.4173,
+      "step": 5371
+    },
+    {
+      "epoch": 13.134474327628363,
+      "grad_norm": 0.1858373991184152,
+      "learning_rate": 1.884073216784593e-07,
+      "loss": 0.421,
+      "step": 5372
+    },
+    {
+      "epoch": 13.136919315403423,
+      "grad_norm": 0.26055126789028776,
+      "learning_rate": 1.879200641553508e-07,
+      "loss": 0.4762,
+      "step": 5373
+    },
+    {
+      "epoch": 13.139364303178484,
+      "grad_norm": 0.2586943160775621,
+      "learning_rate": 1.874334129125277e-07,
+      "loss": 0.8741,
+      "step": 5374
+    },
+    {
+      "epoch": 13.141809290953546,
+      "grad_norm": 0.27012592786365863,
+      "learning_rate": 1.8694736807760182e-07,
+      "loss": 1.148,
+      "step": 5375
+    },
+    {
+      "epoch": 13.144254278728607,
+      "grad_norm": 0.24833506044983888,
+      "learning_rate": 1.8646192977802512e-07,
+      "loss": 1.0364,
+      "step": 5376
+    },
+    {
+      "epoch": 13.146699266503667,
+      "grad_norm": 0.2627093372286388,
+      "learning_rate": 1.8597709814108994e-07,
+      "loss": 0.441,
+      "step": 5377
+    },
+    {
+      "epoch": 13.14914425427873,
+      "grad_norm": 0.33038045302867636,
+      "learning_rate": 1.8549287329393152e-07,
+      "loss": 0.3233,
+      "step": 5378
+    },
+    {
+      "epoch": 13.15158924205379,
+      "grad_norm": 0.2294171361742112,
+      "learning_rate": 1.85009255363523e-07,
+      "loss": 0.5905,
+      "step": 5379
+    },
+    {
+      "epoch": 13.15403422982885,
+      "grad_norm": 0.3288735135583903,
+      "learning_rate": 1.8452624447668183e-07,
+      "loss": 0.6526,
+      "step": 5380
+    },
+    {
+      "epoch": 13.156479217603913,
+      "grad_norm": 0.33241961435341855,
+      "learning_rate": 1.8404384076006337e-07,
+      "loss": 0.5513,
+      "step": 5381
+    },
+    {
+      "epoch": 13.158924205378973,
+      "grad_norm": 0.3128579067905457,
+      "learning_rate": 1.8356204434016506e-07,
+      "loss": 0.6678,
+      "step": 5382
+    },
+    {
+      "epoch": 13.161369193154034,
+      "grad_norm": 0.30575572484832775,
+      "learning_rate": 1.8308085534332583e-07,
+      "loss": 0.9598,
+      "step": 5383
+    },
+    {
+      "epoch": 13.163814180929096,
+      "grad_norm": 0.30994106215091816,
+      "learning_rate": 1.8260027389572304e-07,
+      "loss": 0.496,
+      "step": 5384
+    },
+    {
+      "epoch": 13.166259168704157,
+      "grad_norm": 0.25375397853172155,
+      "learning_rate": 1.8212030012337707e-07,
+      "loss": 0.4537,
+      "step": 5385
+    },
+    {
+      "epoch": 13.168704156479217,
+      "grad_norm": 0.23529172021408684,
+      "learning_rate": 1.816409341521483e-07,
+      "loss": 0.5188,
+      "step": 5386
+    },
+    {
+      "epoch": 13.17114914425428,
+      "grad_norm": 0.37632655701366396,
+      "learning_rate": 1.8116217610773674e-07,
+      "loss": 0.5368,
+      "step": 5387
+    },
+    {
+      "epoch": 13.17359413202934,
+      "grad_norm": 0.28815332515151965,
+      "learning_rate": 1.806840261156842e-07,
+      "loss": 0.4294,
+      "step": 5388
+    },
+    {
+      "epoch": 13.1760391198044,
+      "grad_norm": 0.3420186153916135,
+      "learning_rate": 1.8020648430137178e-07,
+      "loss": 0.4324,
+      "step": 5389
+    },
+    {
+      "epoch": 13.178484107579463,
+      "grad_norm": 0.23125020863378182,
+      "learning_rate": 1.7972955079002264e-07,
+      "loss": 0.6136,
+      "step": 5390
+    },
+    {
+      "epoch": 13.180929095354523,
+      "grad_norm": 0.2108950475877583,
+      "learning_rate": 1.7925322570669922e-07,
+      "loss": 0.3598,
+      "step": 5391
+    },
+    {
+      "epoch": 13.183374083129584,
+      "grad_norm": 0.23539782124721473,
+      "learning_rate": 1.7877750917630526e-07,
+      "loss": 0.4782,
+      "step": 5392
+    },
+    {
+      "epoch": 13.185819070904646,
+      "grad_norm": 0.2570534941175301,
+      "learning_rate": 1.7830240132358483e-07,
+      "loss": 0.3959,
+      "step": 5393
+    },
+    {
+      "epoch": 13.188264058679707,
+      "grad_norm": 0.19278439001622985,
+      "learning_rate": 1.7782790227312107e-07,
+      "loss": 0.4183,
+      "step": 5394
+    },
+    {
+      "epoch": 13.190709046454767,
+      "grad_norm": 0.22205381240663455,
+      "learning_rate": 1.7735401214933917e-07,
+      "loss": 0.5167,
+      "step": 5395
+    },
+    {
+      "epoch": 13.19315403422983,
+      "grad_norm": 0.3077070213800318,
+      "learning_rate": 1.768807310765039e-07,
+      "loss": 0.9318,
+      "step": 5396
+    },
+    {
+      "epoch": 13.19559902200489,
+      "grad_norm": 0.24152133875360682,
+      "learning_rate": 1.7640805917872012e-07,
+      "loss": 0.5515,
+      "step": 5397
+    },
+    {
+      "epoch": 13.19804400977995,
+      "grad_norm": 0.3141975585054514,
+      "learning_rate": 1.7593599657993398e-07,
+      "loss": 0.4818,
+      "step": 5398
+    },
+    {
+      "epoch": 13.200488997555013,
+      "grad_norm": 0.18260588498724392,
+      "learning_rate": 1.7546454340392982e-07,
+      "loss": 0.4238,
+      "step": 5399
+    },
+    {
+      "epoch": 13.202933985330073,
+      "grad_norm": 0.34983530438151883,
+      "learning_rate": 1.7499369977433455e-07,
+      "loss": 0.5123,
+      "step": 5400
+    },
+    {
+      "epoch": 13.205378973105134,
+      "grad_norm": 0.3478786911167567,
+      "learning_rate": 1.7452346581461332e-07,
+      "loss": 0.6526,
+      "step": 5401
+    },
+    {
+      "epoch": 13.207823960880196,
+      "grad_norm": 0.23796683240605365,
+      "learning_rate": 1.7405384164807304e-07,
+      "loss": 0.6726,
+      "step": 5402
+    },
+    {
+      "epoch": 13.210268948655257,
+      "grad_norm": 0.3156136031891151,
+      "learning_rate": 1.7358482739785994e-07,
+      "loss": 0.3996,
+      "step": 5403
+    },
+    {
+      "epoch": 13.212713936430317,
+      "grad_norm": 0.2829726352384672,
+      "learning_rate": 1.731164231869595e-07,
+      "loss": 0.4556,
+      "step": 5404
+    },
+    {
+      "epoch": 13.21515892420538,
+      "grad_norm": 0.289498340957733,
+      "learning_rate": 1.726486291381982e-07,
+      "loss": 0.7063,
+      "step": 5405
+    },
+    {
+      "epoch": 13.21760391198044,
+      "grad_norm": 0.2864270363966545,
+      "learning_rate": 1.721814453742432e-07,
+      "loss": 0.5701,
+      "step": 5406
+    },
+    {
+      "epoch": 13.2200488997555,
+      "grad_norm": 0.2604235273711713,
+      "learning_rate": 1.7171487201759985e-07,
+      "loss": 0.4336,
+      "step": 5407
+    },
+    {
+      "epoch": 13.222493887530563,
+      "grad_norm": 0.2901454280928388,
+      "learning_rate": 1.7124890919061499e-07,
+      "loss": 0.3852,
+      "step": 5408
+    },
+    {
+      "epoch": 13.224938875305623,
+      "grad_norm": 0.2739360748475561,
+      "learning_rate": 1.7078355701547395e-07,
+      "loss": 0.5694,
+      "step": 5409
+    },
+    {
+      "epoch": 13.227383863080684,
+      "grad_norm": 0.29789762494675154,
+      "learning_rate": 1.7031881561420328e-07,
+      "loss": 0.8256,
+      "step": 5410
+    },
+    {
+      "epoch": 13.229828850855746,
+      "grad_norm": 0.2996639185880644,
+      "learning_rate": 1.698546851086691e-07,
+      "loss": 0.5712,
+      "step": 5411
+    },
+    {
+      "epoch": 13.232273838630807,
+      "grad_norm": 0.2114121850825437,
+      "learning_rate": 1.6939116562057602e-07,
+      "loss": 0.5365,
+      "step": 5412
+    },
+    {
+      "epoch": 13.234718826405867,
+      "grad_norm": 0.24672784008966578,
+      "learning_rate": 1.6892825727147093e-07,
+      "loss": 0.8515,
+      "step": 5413
+    },
+    {
+      "epoch": 13.23716381418093,
+      "grad_norm": 0.22342188053126433,
+      "learning_rate": 1.6846596018273765e-07,
+      "loss": 0.727,
+      "step": 5414
+    },
+    {
+      "epoch": 13.23960880195599,
+      "grad_norm": 0.19140015142261502,
+      "learning_rate": 1.6800427447560163e-07,
+      "loss": 0.5274,
+      "step": 5415
+    },
+    {
+      "epoch": 13.24205378973105,
+      "grad_norm": 0.26030001943765374,
+      "learning_rate": 1.6754320027112804e-07,
+      "loss": 0.5158,
+      "step": 5416
+    },
+    {
+      "epoch": 13.244498777506113,
+      "grad_norm": 0.31538007506023785,
+      "learning_rate": 1.6708273769021982e-07,
+      "loss": 0.4878,
+      "step": 5417
+    },
+    {
+      "epoch": 13.246943765281173,
+      "grad_norm": 0.26249968645003663,
+      "learning_rate": 1.666228868536218e-07,
+      "loss": 0.9716,
+      "step": 5418
+    },
+    {
+      "epoch": 13.249388753056234,
+      "grad_norm": 0.29411296807861387,
+      "learning_rate": 1.6616364788191701e-07,
+      "loss": 0.5538,
+      "step": 5419
+    },
+    {
+      "epoch": 13.251833740831296,
+      "grad_norm": 0.3100028316246551,
+      "learning_rate": 1.6570502089552797e-07,
+      "loss": 0.3993,
+      "step": 5420
+    },
+    {
+      "epoch": 13.254278728606357,
+      "grad_norm": 0.28717284408981897,
+      "learning_rate": 1.652470060147185e-07,
+      "loss": 0.3507,
+      "step": 5421
+    },
+    {
+      "epoch": 13.256723716381417,
+      "grad_norm": 0.2728850107273424,
+      "learning_rate": 1.647896033595886e-07,
+      "loss": 0.3709,
+      "step": 5422
+    },
+    {
+      "epoch": 13.25916870415648,
+      "grad_norm": 0.24961662446534846,
+      "learning_rate": 1.643328130500818e-07,
+      "loss": 0.8345,
+      "step": 5423
+    },
+    {
+      "epoch": 13.26161369193154,
+      "grad_norm": 0.23503284287766496,
+      "learning_rate": 1.6387663520597759e-07,
+      "loss": 0.8496,
+      "step": 5424
+    },
+    {
+      "epoch": 13.2640586797066,
+      "grad_norm": 0.20369153136981225,
+      "learning_rate": 1.634210699468969e-07,
+      "loss": 0.3983,
+      "step": 5425
+    },
+    {
+      "epoch": 13.266503667481663,
+      "grad_norm": 0.35033394533814466,
+      "learning_rate": 1.6296611739229916e-07,
+      "loss": 0.6519,
+      "step": 5426
+    },
+    {
+      "epoch": 13.268948655256724,
+      "grad_norm": 0.3200013353050572,
+      "learning_rate": 1.6251177766148319e-07,
+      "loss": 0.6848,
+      "step": 5427
+    },
+    {
+      "epoch": 13.271393643031784,
+      "grad_norm": 0.16736278824731854,
+      "learning_rate": 1.6205805087358778e-07,
+      "loss": 0.4779,
+      "step": 5428
+    },
+    {
+      "epoch": 13.273838630806846,
+      "grad_norm": 0.21544050507727647,
+      "learning_rate": 1.6160493714758973e-07,
+      "loss": 0.363,
+      "step": 5429
+    },
+    {
+      "epoch": 13.276283618581907,
+      "grad_norm": 0.3471500597252846,
+      "learning_rate": 1.6115243660230624e-07,
+      "loss": 0.4872,
+      "step": 5430
+    },
+    {
+      "epoch": 13.278728606356967,
+      "grad_norm": 0.2169835441942447,
+      "learning_rate": 1.6070054935639373e-07,
+      "loss": 0.4654,
+      "step": 5431
+    },
+    {
+      "epoch": 13.28117359413203,
+      "grad_norm": 0.31038049465885664,
+      "learning_rate": 1.6024927552834634e-07,
+      "loss": 0.4282,
+      "step": 5432
+    },
+    {
+      "epoch": 13.28361858190709,
+      "grad_norm": 0.2591053837566953,
+      "learning_rate": 1.5979861523649914e-07,
+      "loss": 0.3958,
+      "step": 5433
+    },
+    {
+      "epoch": 13.28606356968215,
+      "grad_norm": 0.3058535559941627,
+      "learning_rate": 1.5934856859902535e-07,
+      "loss": 0.663,
+      "step": 5434
+    },
+    {
+      "epoch": 13.288508557457213,
+      "grad_norm": 0.2505403970219653,
+      "learning_rate": 1.5889913573393784e-07,
+      "loss": 0.7135,
+      "step": 5435
+    },
+    {
+      "epoch": 13.290953545232274,
+      "grad_norm": 0.2753032421602159,
+      "learning_rate": 1.584503167590881e-07,
+      "loss": 0.9208,
+      "step": 5436
+    },
+    {
+      "epoch": 13.293398533007334,
+      "grad_norm": 0.2441708607650385,
+      "learning_rate": 1.5800211179216622e-07,
+      "loss": 0.475,
+      "step": 5437
+    },
+    {
+      "epoch": 13.295843520782396,
+      "grad_norm": 0.26648614781931507,
+      "learning_rate": 1.5755452095070262e-07,
+      "loss": 0.5796,
+      "step": 5438
+    },
+    {
+      "epoch": 13.298288508557457,
+      "grad_norm": 0.23103480182826988,
+      "learning_rate": 1.5710754435206476e-07,
+      "loss": 0.5013,
+      "step": 5439
+    },
+    {
+      "epoch": 13.30073349633252,
+      "grad_norm": 0.42689724115997363,
+      "learning_rate": 1.5666118211346115e-07,
+      "loss": 0.4449,
+      "step": 5440
+    },
+    {
+      "epoch": 13.30317848410758,
+      "grad_norm": 0.20371610254782882,
+      "learning_rate": 1.5621543435193815e-07,
+      "loss": 1.0173,
+      "step": 5441
+    },
+    {
+      "epoch": 13.30562347188264,
+      "grad_norm": 0.3845821748326833,
+      "learning_rate": 1.5577030118438002e-07,
+      "loss": 0.5033,
+      "step": 5442
+    },
+    {
+      "epoch": 13.3080684596577,
+      "grad_norm": 0.26362582317648026,
+      "learning_rate": 1.5532578272751176e-07,
+      "loss": 0.5584,
+      "step": 5443
+    },
+    {
+      "epoch": 13.310513447432763,
+      "grad_norm": 0.4045725625627985,
+      "learning_rate": 1.5488187909789592e-07,
+      "loss": 0.4591,
+      "step": 5444
+    },
+    {
+      "epoch": 13.312958435207824,
+      "grad_norm": 0.24794199515767687,
+      "learning_rate": 1.5443859041193443e-07,
+      "loss": 1.285,
+      "step": 5445
+    },
+    {
+      "epoch": 13.315403422982886,
+      "grad_norm": 0.2700159370747307,
+      "learning_rate": 1.5399591678586818e-07,
+      "loss": 0.4318,
+      "step": 5446
+    },
+    {
+      "epoch": 13.317848410757946,
+      "grad_norm": 0.288098459304985,
+      "learning_rate": 1.5355385833577542e-07,
+      "loss": 0.6339,
+      "step": 5447
+    },
+    {
+      "epoch": 13.320293398533007,
+      "grad_norm": 0.2733879970621616,
+      "learning_rate": 1.5311241517757453e-07,
+      "loss": 0.3282,
+      "step": 5448
+    },
+    {
+      "epoch": 13.32273838630807,
+      "grad_norm": 0.24433091524197376,
+      "learning_rate": 1.5267158742702154e-07,
+      "loss": 0.7697,
+      "step": 5449
+    },
+    {
+      "epoch": 13.32518337408313,
+      "grad_norm": 0.2552396165325618,
+      "learning_rate": 1.5223137519971175e-07,
+      "loss": 0.6787,
+      "step": 5450
+    },
+    {
+      "epoch": 13.32762836185819,
+      "grad_norm": 0.19813482855934522,
+      "learning_rate": 1.5179177861107953e-07,
+      "loss": 0.6919,
+      "step": 5451
+    },
+    {
+      "epoch": 13.330073349633253,
+      "grad_norm": 0.32427617012288973,
+      "learning_rate": 1.5135279777639622e-07,
+      "loss": 0.8107,
+      "step": 5452
+    },
+    {
+      "epoch": 13.332518337408313,
+      "grad_norm": 0.2844620406411681,
+      "learning_rate": 1.509144328107731e-07,
+      "loss": 0.4028,
+      "step": 5453
+    },
+    {
+      "epoch": 13.334963325183374,
+      "grad_norm": 0.40510371931792327,
+      "learning_rate": 1.5047668382915965e-07,
+      "loss": 0.4015,
+      "step": 5454
+    },
+    {
+      "epoch": 13.337408312958436,
+      "grad_norm": 0.23784702829753648,
+      "learning_rate": 1.5003955094634316e-07,
+      "loss": 0.5732,
+      "step": 5455
+    },
+    {
+      "epoch": 13.339853300733497,
+      "grad_norm": 0.3776919732760176,
+      "learning_rate": 1.4960303427695083e-07,
+      "loss": 0.7404,
+      "step": 5456
+    },
+    {
+      "epoch": 13.342298288508557,
+      "grad_norm": 0.2629738149289269,
+      "learning_rate": 1.491671339354464e-07,
+      "loss": 0.6322,
+      "step": 5457
+    },
+    {
+      "epoch": 13.34474327628362,
+      "grad_norm": 0.4048695479502515,
+      "learning_rate": 1.4873185003613344e-07,
+      "loss": 0.5214,
+      "step": 5458
+    },
+    {
+      "epoch": 13.34718826405868,
+      "grad_norm": 0.16627672781917818,
+      "learning_rate": 1.4829718269315286e-07,
+      "loss": 0.3935,
+      "step": 5459
+    },
+    {
+      "epoch": 13.34963325183374,
+      "grad_norm": 0.18038169680147564,
+      "learning_rate": 1.4786313202048457e-07,
+      "loss": 0.3242,
+      "step": 5460
+    },
+    {
+      "epoch": 13.352078239608803,
+      "grad_norm": 0.25426090928832334,
+      "learning_rate": 1.474296981319473e-07,
+      "loss": 0.6927,
+      "step": 5461
+    },
+    {
+      "epoch": 13.354523227383863,
+      "grad_norm": 0.27156593608876545,
+      "learning_rate": 1.4699688114119591e-07,
+      "loss": 1.21,
+      "step": 5462
+    },
+    {
+      "epoch": 13.356968215158924,
+      "grad_norm": 0.26794482778588635,
+      "learning_rate": 1.4656468116172606e-07,
+      "loss": 0.5235,
+      "step": 5463
+    },
+    {
+      "epoch": 13.359413202933986,
+      "grad_norm": 0.262087658132917,
+      "learning_rate": 1.4613309830687068e-07,
+      "loss": 0.7544,
+      "step": 5464
+    },
+    {
+      "epoch": 13.361858190709047,
+      "grad_norm": 0.22971049088929713,
+      "learning_rate": 1.4570213268979922e-07,
+      "loss": 0.6002,
+      "step": 5465
+    },
+    {
+      "epoch": 13.364303178484107,
+      "grad_norm": 0.2545606803547083,
+      "learning_rate": 1.4527178442352242e-07,
+      "loss": 0.5226,
+      "step": 5466
+    },
+    {
+      "epoch": 13.36674816625917,
+      "grad_norm": 0.2798879064892046,
+      "learning_rate": 1.448420536208861e-07,
+      "loss": 0.6584,
+      "step": 5467
+    },
+    {
+      "epoch": 13.36919315403423,
+      "grad_norm": 0.31484292006273545,
+      "learning_rate": 1.4441294039457683e-07,
+      "loss": 0.4756,
+      "step": 5468
+    },
+    {
+      "epoch": 13.37163814180929,
+      "grad_norm": 0.22179126495474688,
+      "learning_rate": 1.4398444485711677e-07,
+      "loss": 0.4373,
+      "step": 5469
+    },
+    {
+      "epoch": 13.374083129584353,
+      "grad_norm": 0.48191616107992935,
+      "learning_rate": 1.4355656712086746e-07,
+      "loss": 0.3548,
+      "step": 5470
+    },
+    {
+      "epoch": 13.376528117359413,
+      "grad_norm": 0.2245549080198216,
+      "learning_rate": 1.4312930729802887e-07,
+      "loss": 0.7357,
+      "step": 5471
+    },
+    {
+      "epoch": 13.378973105134474,
+      "grad_norm": 0.26151041781822554,
+      "learning_rate": 1.4270266550063777e-07,
+      "loss": 1.2821,
+      "step": 5472
+    },
+    {
+      "epoch": 13.381418092909536,
+      "grad_norm": 0.21174856190448904,
+      "learning_rate": 1.422766418405694e-07,
+      "loss": 0.5342,
+      "step": 5473
+    },
+    {
+      "epoch": 13.383863080684597,
+      "grad_norm": 0.3366622616466138,
+      "learning_rate": 1.4185123642953742e-07,
+      "loss": 0.8511,
+      "step": 5474
+    },
+    {
+      "epoch": 13.386308068459657,
+      "grad_norm": 0.32167258081763,
+      "learning_rate": 1.4142644937909206e-07,
+      "loss": 0.6384,
+      "step": 5475
+    },
+    {
+      "epoch": 13.38875305623472,
+      "grad_norm": 0.34986141558194417,
+      "learning_rate": 1.4100228080062284e-07,
+      "loss": 0.7247,
+      "step": 5476
+    },
+    {
+      "epoch": 13.39119804400978,
+      "grad_norm": 0.3124050914274767,
+      "learning_rate": 1.4057873080535605e-07,
+      "loss": 0.7543,
+      "step": 5477
+    },
+    {
+      "epoch": 13.39364303178484,
+      "grad_norm": 0.342714201933395,
+      "learning_rate": 1.4015579950435676e-07,
+      "loss": 0.5818,
+      "step": 5478
+    },
+    {
+      "epoch": 13.396088019559903,
+      "grad_norm": 0.34807431439368264,
+      "learning_rate": 1.3973348700852678e-07,
+      "loss": 0.4377,
+      "step": 5479
+    },
+    {
+      "epoch": 13.398533007334963,
+      "grad_norm": 0.33408715591217425,
+      "learning_rate": 1.3931179342860618e-07,
+      "loss": 0.4069,
+      "step": 5480
+    },
+    {
+      "epoch": 13.400977995110024,
+      "grad_norm": 0.2824110238634206,
+      "learning_rate": 1.3889071887517287e-07,
+      "loss": 0.8315,
+      "step": 5481
+    },
+    {
+      "epoch": 13.403422982885086,
+      "grad_norm": 0.17699753469013663,
+      "learning_rate": 1.3847026345864184e-07,
+      "loss": 0.5688,
+      "step": 5482
+    },
+    {
+      "epoch": 13.405867970660147,
+      "grad_norm": 0.3802514115381719,
+      "learning_rate": 1.380504272892666e-07,
+      "loss": 0.4387,
+      "step": 5483
+    },
+    {
+      "epoch": 13.408312958435207,
+      "grad_norm": 0.29446160182884545,
+      "learning_rate": 1.3763121047713767e-07,
+      "loss": 0.4712,
+      "step": 5484
+    },
+    {
+      "epoch": 13.41075794621027,
+      "grad_norm": 0.18475740564409524,
+      "learning_rate": 1.3721261313218298e-07,
+      "loss": 0.7188,
+      "step": 5485
+    },
+    {
+      "epoch": 13.41320293398533,
+      "grad_norm": 0.2833861629393022,
+      "learning_rate": 1.3679463536416887e-07,
+      "loss": 0.796,
+      "step": 5486
+    },
+    {
+      "epoch": 13.41564792176039,
+      "grad_norm": 0.267642139077156,
+      "learning_rate": 1.363772772826974e-07,
+      "loss": 0.4476,
+      "step": 5487
+    },
+    {
+      "epoch": 13.418092909535453,
+      "grad_norm": 0.28880346818983627,
+      "learning_rate": 1.3596053899721157e-07,
+      "loss": 0.4699,
+      "step": 5488
+    },
+    {
+      "epoch": 13.420537897310513,
+      "grad_norm": 0.22870223697992778,
+      "learning_rate": 1.3554442061698785e-07,
+      "loss": 0.9342,
+      "step": 5489
+    },
+    {
+      "epoch": 13.422982885085574,
+      "grad_norm": 0.2816303351311041,
+      "learning_rate": 1.351289222511426e-07,
+      "loss": 0.4656,
+      "step": 5490
+    },
+    {
+      "epoch": 13.425427872860636,
+      "grad_norm": 0.4071698967253718,
+      "learning_rate": 1.3471404400862976e-07,
+      "loss": 0.7067,
+      "step": 5491
+    },
+    {
+      "epoch": 13.427872860635697,
+      "grad_norm": 0.1893772907896548,
+      "learning_rate": 1.342997859982384e-07,
+      "loss": 0.6452,
+      "step": 5492
+    },
+    {
+      "epoch": 13.430317848410757,
+      "grad_norm": 0.26997642285480666,
+      "learning_rate": 1.3388614832859774e-07,
+      "loss": 0.6897,
+      "step": 5493
+    },
+    {
+      "epoch": 13.43276283618582,
+      "grad_norm": 0.24348747411767688,
+      "learning_rate": 1.3347313110817272e-07,
+      "loss": 0.8249,
+      "step": 5494
+    },
+    {
+      "epoch": 13.43520782396088,
+      "grad_norm": 0.32386295527794956,
+      "learning_rate": 1.330607344452653e-07,
+      "loss": 0.4036,
+      "step": 5495
+    },
+    {
+      "epoch": 13.43765281173594,
+      "grad_norm": 0.256221278276785,
+      "learning_rate": 1.3264895844801618e-07,
+      "loss": 0.7382,
+      "step": 5496
+    },
+    {
+      "epoch": 13.440097799511003,
+      "grad_norm": 0.31016210968045405,
+      "learning_rate": 1.322378032244015e-07,
+      "loss": 0.5195,
+      "step": 5497
+    },
+    {
+      "epoch": 13.442542787286063,
+      "grad_norm": 0.25013645072378415,
+      "learning_rate": 1.3182726888223667e-07,
+      "loss": 0.7298,
+      "step": 5498
+    },
+    {
+      "epoch": 13.444987775061124,
+      "grad_norm": 0.1846426342626662,
+      "learning_rate": 1.3141735552917228e-07,
+      "loss": 0.435,
+      "step": 5499
+    },
+    {
+      "epoch": 13.447432762836186,
+      "grad_norm": 0.32572540090788976,
+      "learning_rate": 1.3100806327269733e-07,
+      "loss": 0.4677,
+      "step": 5500
+    },
+    {
+      "epoch": 13.449877750611247,
+      "grad_norm": 0.22450410603233825,
+      "learning_rate": 1.3059939222013818e-07,
+      "loss": 0.6829,
+      "step": 5501
+    },
+    {
+      "epoch": 13.452322738386307,
+      "grad_norm": 0.23357462477144836,
+      "learning_rate": 1.301913424786569e-07,
+      "loss": 0.4415,
+      "step": 5502
+    },
+    {
+      "epoch": 13.45476772616137,
+      "grad_norm": 0.2727292281769691,
+      "learning_rate": 1.297839141552537e-07,
+      "loss": 0.4654,
+      "step": 5503
+    },
+    {
+      "epoch": 13.45721271393643,
+      "grad_norm": 0.24544774538442626,
+      "learning_rate": 1.2937710735676596e-07,
+      "loss": 0.8823,
+      "step": 5504
+    },
+    {
+      "epoch": 13.45965770171149,
+      "grad_norm": 0.34006390174217604,
+      "learning_rate": 1.2897092218986718e-07,
+      "loss": 0.4274,
+      "step": 5505
+    },
+    {
+      "epoch": 13.462102689486553,
+      "grad_norm": 0.26371114119596045,
+      "learning_rate": 1.2856535876106886e-07,
+      "loss": 0.9728,
+      "step": 5506
+    },
+    {
+      "epoch": 13.464547677261614,
+      "grad_norm": 0.2151636074054762,
+      "learning_rate": 1.281604171767187e-07,
+      "loss": 0.4613,
+      "step": 5507
+    },
+    {
+      "epoch": 13.466992665036674,
+      "grad_norm": 0.38804716176156856,
+      "learning_rate": 1.27756097543002e-07,
+      "loss": 0.8472,
+      "step": 5508
+    },
+    {
+      "epoch": 13.469437652811736,
+      "grad_norm": 0.2536928639402621,
+      "learning_rate": 1.2735239996594013e-07,
+      "loss": 0.37,
+      "step": 5509
+    },
+    {
+      "epoch": 13.471882640586797,
+      "grad_norm": 0.20325778689076104,
+      "learning_rate": 1.2694932455139226e-07,
+      "loss": 0.4524,
+      "step": 5510
+    },
+    {
+      "epoch": 13.474327628361857,
+      "grad_norm": 0.28878343008149404,
+      "learning_rate": 1.2654687140505413e-07,
+      "loss": 0.3618,
+      "step": 5511
+    },
+    {
+      "epoch": 13.47677261613692,
+      "grad_norm": 0.30093682869050103,
+      "learning_rate": 1.261450406324577e-07,
+      "loss": 0.4545,
+      "step": 5512
+    },
+    {
+      "epoch": 13.47921760391198,
+      "grad_norm": 0.19270717204212476,
+      "learning_rate": 1.257438323389723e-07,
+      "loss": 0.3434,
+      "step": 5513
+    },
+    {
+      "epoch": 13.48166259168704,
+      "grad_norm": 0.2420352155070497,
+      "learning_rate": 1.2534324662980434e-07,
+      "loss": 0.4591,
+      "step": 5514
+    },
+    {
+      "epoch": 13.484107579462103,
+      "grad_norm": 0.40176290576989615,
+      "learning_rate": 1.2494328360999592e-07,
+      "loss": 0.3896,
+      "step": 5515
+    },
+    {
+      "epoch": 13.486552567237164,
+      "grad_norm": 0.2998968031832042,
+      "learning_rate": 1.2454394338442726e-07,
+      "loss": 0.5822,
+      "step": 5516
+    },
+    {
+      "epoch": 13.488997555012224,
+      "grad_norm": 0.3503253198656912,
+      "learning_rate": 1.2414522605781355e-07,
+      "loss": 0.6744,
+      "step": 5517
+    },
+    {
+      "epoch": 13.491442542787286,
+      "grad_norm": 0.23791028559403393,
+      "learning_rate": 1.237471317347083e-07,
+      "loss": 0.3759,
+      "step": 5518
+    },
+    {
+      "epoch": 13.493887530562347,
+      "grad_norm": 0.27728368245119683,
+      "learning_rate": 1.2334966051950086e-07,
+      "loss": 0.4714,
+      "step": 5519
+    },
+    {
+      "epoch": 13.496332518337407,
+      "grad_norm": 0.24732886272790217,
+      "learning_rate": 1.22952812516417e-07,
+      "loss": 0.3861,
+      "step": 5520
+    },
+    {
+      "epoch": 13.49877750611247,
+      "grad_norm": 0.20110125569955892,
+      "learning_rate": 1.2255658782952013e-07,
+      "loss": 0.6863,
+      "step": 5521
+    },
+    {
+      "epoch": 13.50122249388753,
+      "grad_norm": 0.20074626203303475,
+      "learning_rate": 1.2216098656270824e-07,
+      "loss": 0.4338,
+      "step": 5522
+    },
+    {
+      "epoch": 13.503667481662593,
+      "grad_norm": 0.169916250263693,
+      "learning_rate": 1.2176600881971784e-07,
+      "loss": 0.4296,
+      "step": 5523
+    },
+    {
+      "epoch": 13.506112469437653,
+      "grad_norm": 0.24214297617312927,
+      "learning_rate": 1.2137165470412134e-07,
+      "loss": 0.4364,
+      "step": 5524
+    },
+    {
+      "epoch": 13.508557457212714,
+      "grad_norm": 0.3218353993077113,
+      "learning_rate": 1.2097792431932653e-07,
+      "loss": 0.591,
+      "step": 5525
+    },
+    {
+      "epoch": 13.511002444987774,
+      "grad_norm": 0.2882503381878655,
+      "learning_rate": 1.205848177685795e-07,
+      "loss": 0.3309,
+      "step": 5526
+    },
+    {
+      "epoch": 13.513447432762836,
+      "grad_norm": 0.32826182968853174,
+      "learning_rate": 1.201923351549608e-07,
+      "loss": 0.6806,
+      "step": 5527
+    },
+    {
+      "epoch": 13.515892420537897,
+      "grad_norm": 0.3016361991693213,
+      "learning_rate": 1.1980047658138926e-07,
+      "loss": 0.3373,
+      "step": 5528
+    },
+    {
+      "epoch": 13.51833740831296,
+      "grad_norm": 0.23492412532991028,
+      "learning_rate": 1.1940924215061878e-07,
+      "loss": 0.9251,
+      "step": 5529
+    },
+    {
+      "epoch": 13.52078239608802,
+      "grad_norm": 0.35464190228328957,
+      "learning_rate": 1.1901863196523921e-07,
+      "loss": 0.6157,
+      "step": 5530
+    },
+    {
+      "epoch": 13.52322738386308,
+      "grad_norm": 0.27884521536947593,
+      "learning_rate": 1.1862864612767921e-07,
+      "loss": 0.4315,
+      "step": 5531
+    },
+    {
+      "epoch": 13.525672371638143,
+      "grad_norm": 0.3335432288201711,
+      "learning_rate": 1.1823928474020057e-07,
+      "loss": 0.3951,
+      "step": 5532
+    },
+    {
+      "epoch": 13.528117359413203,
+      "grad_norm": 0.20952527514343117,
+      "learning_rate": 1.178505479049033e-07,
+      "loss": 0.4901,
+      "step": 5533
+    },
+    {
+      "epoch": 13.530562347188264,
+      "grad_norm": 0.20416209376344643,
+      "learning_rate": 1.1746243572372335e-07,
+      "loss": 0.6897,
+      "step": 5534
+    },
+    {
+      "epoch": 13.533007334963326,
+      "grad_norm": 0.37810003170379586,
+      "learning_rate": 1.1707494829843208e-07,
+      "loss": 0.7884,
+      "step": 5535
+    },
+    {
+      "epoch": 13.535452322738386,
+      "grad_norm": 0.2757127167778145,
+      "learning_rate": 1.1668808573063795e-07,
+      "loss": 0.6171,
+      "step": 5536
+    },
+    {
+      "epoch": 13.537897310513447,
+      "grad_norm": 0.281349043444956,
+      "learning_rate": 1.1630184812178453e-07,
+      "loss": 0.5106,
+      "step": 5537
+    },
+    {
+      "epoch": 13.54034229828851,
+      "grad_norm": 0.27364883590576317,
+      "learning_rate": 1.1591623557315302e-07,
+      "loss": 0.3692,
+      "step": 5538
+    },
+    {
+      "epoch": 13.54278728606357,
+      "grad_norm": 0.21645327035397532,
+      "learning_rate": 1.1553124818585948e-07,
+      "loss": 0.7495,
+      "step": 5539
+    },
+    {
+      "epoch": 13.54523227383863,
+      "grad_norm": 0.22542112697996644,
+      "learning_rate": 1.1514688606085594e-07,
+      "loss": 0.43,
+      "step": 5540
+    },
+    {
+      "epoch": 13.547677261613693,
+      "grad_norm": 0.2595556855489822,
+      "learning_rate": 1.1476314929893207e-07,
+      "loss": 0.3733,
+      "step": 5541
+    },
+    {
+      "epoch": 13.550122249388753,
+      "grad_norm": 0.20784251358452624,
+      "learning_rate": 1.1438003800071123e-07,
+      "loss": 0.7624,
+      "step": 5542
+    },
+    {
+      "epoch": 13.552567237163814,
+      "grad_norm": 0.3736623667846649,
+      "learning_rate": 1.1399755226665476e-07,
+      "loss": 0.4928,
+      "step": 5543
+    },
+    {
+      "epoch": 13.555012224938876,
+      "grad_norm": 0.22931721144500014,
+      "learning_rate": 1.1361569219705909e-07,
+      "loss": 0.4382,
+      "step": 5544
+    },
+    {
+      "epoch": 13.557457212713937,
+      "grad_norm": 0.27195582440396815,
+      "learning_rate": 1.1323445789205633e-07,
+      "loss": 0.5064,
+      "step": 5545
+    },
+    {
+      "epoch": 13.559902200488997,
+      "grad_norm": 0.15860479836460462,
+      "learning_rate": 1.1285384945161542e-07,
+      "loss": 0.4119,
+      "step": 5546
+    },
+    {
+      "epoch": 13.56234718826406,
+      "grad_norm": 0.3413168392662795,
+      "learning_rate": 1.1247386697553986e-07,
+      "loss": 0.3358,
+      "step": 5547
+    },
+    {
+      "epoch": 13.56479217603912,
+      "grad_norm": 0.2540496734017776,
+      "learning_rate": 1.120945105634702e-07,
+      "loss": 0.8551,
+      "step": 5548
+    },
+    {
+      "epoch": 13.56723716381418,
+      "grad_norm": 0.2834539217337119,
+      "learning_rate": 1.1171578031488245e-07,
+      "loss": 0.6504,
+      "step": 5549
+    },
+    {
+      "epoch": 13.569682151589243,
+      "grad_norm": 0.16213642286649776,
+      "learning_rate": 1.1133767632908799e-07,
+      "loss": 0.3996,
+      "step": 5550
+    },
+    {
+      "epoch": 13.572127139364303,
+      "grad_norm": 0.2401672763681036,
+      "learning_rate": 1.1096019870523444e-07,
+      "loss": 0.7987,
+      "step": 5551
+    },
+    {
+      "epoch": 13.574572127139364,
+      "grad_norm": 0.22753802013516647,
+      "learning_rate": 1.1058334754230515e-07,
+      "loss": 0.3622,
+      "step": 5552
+    },
+    {
+      "epoch": 13.577017114914426,
+      "grad_norm": 0.19888001714476328,
+      "learning_rate": 1.1020712293911912e-07,
+      "loss": 0.5267,
+      "step": 5553
+    },
+    {
+      "epoch": 13.579462102689487,
+      "grad_norm": 0.4040812928179129,
+      "learning_rate": 1.0983152499433159e-07,
+      "loss": 0.6383,
+      "step": 5554
+    },
+    {
+      "epoch": 13.581907090464547,
+      "grad_norm": 0.4122965750207208,
+      "learning_rate": 1.0945655380643183e-07,
+      "loss": 0.7265,
+      "step": 5555
+    },
+    {
+      "epoch": 13.58435207823961,
+      "grad_norm": 0.24515409049586184,
+      "learning_rate": 1.0908220947374704e-07,
+      "loss": 0.6433,
+      "step": 5556
+    },
+    {
+      "epoch": 13.58679706601467,
+      "grad_norm": 0.3209985402865371,
+      "learning_rate": 1.0870849209443784e-07,
+      "loss": 0.424,
+      "step": 5557
+    },
+    {
+      "epoch": 13.58924205378973,
+      "grad_norm": 0.3491010832431343,
+      "learning_rate": 1.0833540176650197e-07,
+      "loss": 0.386,
+      "step": 5558
+    },
+    {
+      "epoch": 13.591687041564793,
+      "grad_norm": 0.26611475726174433,
+      "learning_rate": 1.0796293858777285e-07,
+      "loss": 0.8779,
+      "step": 5559
+    },
+    {
+      "epoch": 13.594132029339853,
+      "grad_norm": 0.26785842469949483,
+      "learning_rate": 1.075911026559176e-07,
+      "loss": 0.7852,
+      "step": 5560
+    },
+    {
+      "epoch": 13.596577017114914,
+      "grad_norm": 0.3196723739359268,
+      "learning_rate": 1.0721989406844074e-07,
+      "loss": 0.7188,
+      "step": 5561
+    },
+    {
+      "epoch": 13.599022004889976,
+      "grad_norm": 0.3377698865709029,
+      "learning_rate": 1.0684931292268163e-07,
+      "loss": 0.3626,
+      "step": 5562
+    },
+    {
+      "epoch": 13.601466992665037,
+      "grad_norm": 0.37309932370865584,
+      "learning_rate": 1.0647935931581532e-07,
+      "loss": 1.1618,
+      "step": 5563
+    },
+    {
+      "epoch": 13.603911980440097,
+      "grad_norm": 0.3178922148360617,
+      "learning_rate": 1.0611003334485226e-07,
+      "loss": 0.7084,
+      "step": 5564
+    },
+    {
+      "epoch": 13.60635696821516,
+      "grad_norm": 0.3275162778672461,
+      "learning_rate": 1.0574133510663748e-07,
+      "loss": 0.7152,
+      "step": 5565
+    },
+    {
+      "epoch": 13.60880195599022,
+      "grad_norm": 0.3201526475463653,
+      "learning_rate": 1.0537326469785281e-07,
+      "loss": 0.5989,
+      "step": 5566
+    },
+    {
+      "epoch": 13.61124694376528,
+      "grad_norm": 0.21027932118077458,
+      "learning_rate": 1.050058222150141e-07,
+      "loss": 0.7291,
+      "step": 5567
+    },
+    {
+      "epoch": 13.613691931540343,
+      "grad_norm": 0.29295512146782515,
+      "learning_rate": 1.0463900775447345e-07,
+      "loss": 0.6523,
+      "step": 5568
+    },
+    {
+      "epoch": 13.616136919315403,
+      "grad_norm": 0.22941105746364399,
+      "learning_rate": 1.0427282141241862e-07,
+      "loss": 0.7076,
+      "step": 5569
+    },
+    {
+      "epoch": 13.618581907090464,
+      "grad_norm": 0.2277477790862266,
+      "learning_rate": 1.0390726328487088e-07,
+      "loss": 0.9967,
+      "step": 5570
+    },
+    {
+      "epoch": 13.621026894865526,
+      "grad_norm": 0.3537175954855979,
+      "learning_rate": 1.0354233346768855e-07,
+      "loss": 0.6843,
+      "step": 5571
+    },
+    {
+      "epoch": 13.623471882640587,
+      "grad_norm": 0.214071385457428,
+      "learning_rate": 1.0317803205656479e-07,
+      "loss": 0.6996,
+      "step": 5572
+    },
+    {
+      "epoch": 13.625916870415647,
+      "grad_norm": 0.24350129312577748,
+      "learning_rate": 1.0281435914702736e-07,
+      "loss": 0.7302,
+      "step": 5573
+    },
+    {
+      "epoch": 13.62836185819071,
+      "grad_norm": 0.35748596914359504,
+      "learning_rate": 1.0245131483444026e-07,
+      "loss": 0.7052,
+      "step": 5574
+    },
+    {
+      "epoch": 13.63080684596577,
+      "grad_norm": 0.2464543130820322,
+      "learning_rate": 1.0208889921400122e-07,
+      "loss": 0.3268,
+      "step": 5575
+    },
+    {
+      "epoch": 13.63325183374083,
+      "grad_norm": 0.2530724922034853,
+      "learning_rate": 1.017271123807445e-07,
+      "loss": 0.3802,
+      "step": 5576
+    },
+    {
+      "epoch": 13.635696821515893,
+      "grad_norm": 0.2547182461894175,
+      "learning_rate": 1.0136595442953839e-07,
+      "loss": 0.3817,
+      "step": 5577
+    },
+    {
+      "epoch": 13.638141809290953,
+      "grad_norm": 0.2347048562609496,
+      "learning_rate": 1.0100542545508713e-07,
+      "loss": 0.6289,
+      "step": 5578
+    },
+    {
+      "epoch": 13.640586797066014,
+      "grad_norm": 0.40699033551254127,
+      "learning_rate": 1.006455255519298e-07,
+      "loss": 0.4444,
+      "step": 5579
+    },
+    {
+      "epoch": 13.643031784841076,
+      "grad_norm": 0.253408994334587,
+      "learning_rate": 1.0028625481443982e-07,
+      "loss": 0.372,
+      "step": 5580
+    },
+    {
+      "epoch": 13.645476772616137,
+      "grad_norm": 0.2700962927274832,
+      "learning_rate": 9.992761333682683e-08,
+      "loss": 0.6272,
+      "step": 5581
+    },
+    {
+      "epoch": 13.647921760391197,
+      "grad_norm": 0.2586830769537017,
+      "learning_rate": 9.956960121313447e-08,
+      "loss": 0.5185,
+      "step": 5582
+    },
+    {
+      "epoch": 13.65036674816626,
+      "grad_norm": 0.3032139142960979,
+      "learning_rate": 9.921221853724155e-08,
+      "loss": 0.6208,
+      "step": 5583
+    },
+    {
+      "epoch": 13.65281173594132,
+      "grad_norm": 0.2910797918523374,
+      "learning_rate": 9.885546540286255e-08,
+      "loss": 0.4145,
+      "step": 5584
+    },
+    {
+      "epoch": 13.65525672371638,
+      "grad_norm": 0.3131412803201063,
+      "learning_rate": 9.849934190354599e-08,
+      "loss": 0.4628,
+      "step": 5585
+    },
+    {
+      "epoch": 13.657701711491443,
+      "grad_norm": 0.18618772916662946,
+      "learning_rate": 9.814384813267575e-08,
+      "loss": 0.7337,
+      "step": 5586
+    },
+    {
+      "epoch": 13.660146699266503,
+      "grad_norm": 0.3064789270112802,
+      "learning_rate": 9.778898418347005e-08,
+      "loss": 0.8502,
+      "step": 5587
+    },
+    {
+      "epoch": 13.662591687041564,
+      "grad_norm": 0.2457187182585951,
+      "learning_rate": 9.743475014898251e-08,
+      "loss": 0.5088,
+      "step": 5588
+    },
+    {
+      "epoch": 13.665036674816626,
+      "grad_norm": 0.2747033736230123,
+      "learning_rate": 9.70811461221019e-08,
+      "loss": 1.0551,
+      "step": 5589
+    },
+    {
+      "epoch": 13.667481662591687,
+      "grad_norm": 0.18071413014314702,
+      "learning_rate": 9.672817219555041e-08,
+      "loss": 0.3956,
+      "step": 5590
+    },
+    {
+      "epoch": 13.669926650366747,
+      "grad_norm": 0.2002353052001889,
+      "learning_rate": 9.637582846188654e-08,
+      "loss": 0.41,
+      "step": 5591
+    },
+    {
+      "epoch": 13.67237163814181,
+      "grad_norm": 0.2655963786380672,
+      "learning_rate": 9.602411501350273e-08,
+      "loss": 0.5863,
+      "step": 5592
+    },
+    {
+      "epoch": 13.67481662591687,
+      "grad_norm": 0.26722078536378874,
+      "learning_rate": 9.567303194262634e-08,
+      "loss": 0.5135,
+      "step": 5593
+    },
+    {
+      "epoch": 13.67726161369193,
+      "grad_norm": 0.44084598116593093,
+      "learning_rate": 9.5322579341319e-08,
+      "loss": 0.5545,
+      "step": 5594
+    },
+    {
+      "epoch": 13.679706601466993,
+      "grad_norm": 0.22793067029525646,
+      "learning_rate": 9.497275730147776e-08,
+      "loss": 0.4857,
+      "step": 5595
+    },
+    {
+      "epoch": 13.682151589242054,
+      "grad_norm": 0.3366078425956773,
+      "learning_rate": 9.462356591483424e-08,
+      "loss": 0.692,
+      "step": 5596
+    },
+    {
+      "epoch": 13.684596577017114,
+      "grad_norm": 0.3016154508879463,
+      "learning_rate": 9.427500527295381e-08,
+      "loss": 0.4885,
+      "step": 5597
+    },
+    {
+      "epoch": 13.687041564792176,
+      "grad_norm": 0.2949481875266221,
+      "learning_rate": 9.392707546723751e-08,
+      "loss": 0.5066,
+      "step": 5598
+    },
+    {
+      "epoch": 13.689486552567237,
+      "grad_norm": 0.2743711463609389,
+      "learning_rate": 9.357977658892043e-08,
+      "loss": 0.7031,
+      "step": 5599
+    },
+    {
+      "epoch": 13.6919315403423,
+      "grad_norm": 0.3975870256365274,
+      "learning_rate": 9.323310872907165e-08,
+      "loss": 0.6523,
+      "step": 5600
+    },
+    {
+      "epoch": 13.69437652811736,
+      "grad_norm": 0.3006610452872104,
+      "learning_rate": 9.288707197859626e-08,
+      "loss": 0.6622,
+      "step": 5601
+    },
+    {
+      "epoch": 13.69682151589242,
+      "grad_norm": 0.28910606408622425,
+      "learning_rate": 9.254166642823303e-08,
+      "loss": 0.3731,
+      "step": 5602
+    },
+    {
+      "epoch": 13.69926650366748,
+      "grad_norm": 0.24739031345940038,
+      "learning_rate": 9.219689216855427e-08,
+      "loss": 0.5051,
+      "step": 5603
+    },
+    {
+      "epoch": 13.701711491442543,
+      "grad_norm": 0.24729099273927424,
+      "learning_rate": 9.185274928996901e-08,
+      "loss": 1.0197,
+      "step": 5604
+    },
+    {
+      "epoch": 13.704156479217604,
+      "grad_norm": 0.2586187907145074,
+      "learning_rate": 9.150923788271788e-08,
+      "loss": 0.599,
+      "step": 5605
+    },
+    {
+      "epoch": 13.706601466992666,
+      "grad_norm": 0.26733276592245103,
+      "learning_rate": 9.11663580368788e-08,
+      "loss": 0.6108,
+      "step": 5606
+    },
+    {
+      "epoch": 13.709046454767726,
+      "grad_norm": 0.16962873030933348,
+      "learning_rate": 9.082410984236179e-08,
+      "loss": 0.4192,
+      "step": 5607
+    },
+    {
+      "epoch": 13.711491442542787,
+      "grad_norm": 0.19109742641524302,
+      "learning_rate": 9.048249338891257e-08,
+      "loss": 0.5796,
+      "step": 5608
+    },
+    {
+      "epoch": 13.713936430317847,
+      "grad_norm": 0.2575927257921119,
+      "learning_rate": 9.014150876611089e-08,
+      "loss": 0.6919,
+      "step": 5609
+    },
+    {
+      "epoch": 13.71638141809291,
+      "grad_norm": 0.21605336601919534,
+      "learning_rate": 8.980115606337047e-08,
+      "loss": 0.5767,
+      "step": 5610
+    },
+    {
+      "epoch": 13.71882640586797,
+      "grad_norm": 0.27651660457344274,
+      "learning_rate": 8.94614353699394e-08,
+      "loss": 0.651,
+      "step": 5611
+    },
+    {
+      "epoch": 13.721271393643033,
+      "grad_norm": 0.37645635866485083,
+      "learning_rate": 8.912234677490111e-08,
+      "loss": 0.4514,
+      "step": 5612
+    },
+    {
+      "epoch": 13.723716381418093,
+      "grad_norm": 0.2070683807925209,
+      "learning_rate": 8.878389036717144e-08,
+      "loss": 0.552,
+      "step": 5613
+    },
+    {
+      "epoch": 13.726161369193154,
+      "grad_norm": 0.3094229333854138,
+      "learning_rate": 8.844606623550189e-08,
+      "loss": 0.4939,
+      "step": 5614
+    },
+    {
+      "epoch": 13.728606356968216,
+      "grad_norm": 0.17967611499043526,
+      "learning_rate": 8.810887446847688e-08,
+      "loss": 0.4057,
+      "step": 5615
+    },
+    {
+      "epoch": 13.731051344743276,
+      "grad_norm": 0.26738429494402016,
+      "learning_rate": 8.777231515451707e-08,
+      "loss": 0.6242,
+      "step": 5616
+    },
+    {
+      "epoch": 13.733496332518337,
+      "grad_norm": 0.232749654494077,
+      "learning_rate": 8.743638838187545e-08,
+      "loss": 0.6963,
+      "step": 5617
+    },
+    {
+      "epoch": 13.7359413202934,
+      "grad_norm": 0.23981390645506026,
+      "learning_rate": 8.710109423863966e-08,
+      "loss": 0.4963,
+      "step": 5618
+    },
+    {
+      "epoch": 13.73838630806846,
+      "grad_norm": 0.1939563806979734,
+      "learning_rate": 8.676643281273156e-08,
+      "loss": 0.4204,
+      "step": 5619
+    },
+    {
+      "epoch": 13.74083129584352,
+      "grad_norm": 0.27966211203387115,
+      "learning_rate": 8.643240419190679e-08,
+      "loss": 0.6107,
+      "step": 5620
+    },
+    {
+      "epoch": 13.743276283618583,
+      "grad_norm": 0.2558240384902012,
+      "learning_rate": 8.60990084637553e-08,
+      "loss": 0.9455,
+      "step": 5621
+    },
+    {
+      "epoch": 13.745721271393643,
+      "grad_norm": 0.2561290626797861,
+      "learning_rate": 8.576624571570186e-08,
+      "loss": 0.352,
+      "step": 5622
+    },
+    {
+      "epoch": 13.748166259168704,
+      "grad_norm": 0.29951124682271113,
+      "learning_rate": 8.543411603500362e-08,
+      "loss": 0.4775,
+      "step": 5623
+    },
+    {
+      "epoch": 13.750611246943766,
+      "grad_norm": 0.2008154712772762,
+      "learning_rate": 8.510261950875287e-08,
+      "loss": 0.706,
+      "step": 5624
+    },
+    {
+      "epoch": 13.753056234718827,
+      "grad_norm": 0.18568914027581912,
+      "learning_rate": 8.477175622387563e-08,
+      "loss": 0.5576,
+      "step": 5625
+    },
+    {
+      "epoch": 13.755501222493887,
+      "grad_norm": 0.26797044014583915,
+      "learning_rate": 8.444152626713137e-08,
+      "loss": 0.4845,
+      "step": 5626
+    },
+    {
+      "epoch": 13.75794621026895,
+      "grad_norm": 0.1876005338849139,
+      "learning_rate": 8.411192972511473e-08,
+      "loss": 0.4242,
+      "step": 5627
+    },
+    {
+      "epoch": 13.76039119804401,
+      "grad_norm": 0.2607193734334943,
+      "learning_rate": 8.378296668425295e-08,
+      "loss": 0.4002,
+      "step": 5628
+    },
+    {
+      "epoch": 13.76283618581907,
+      "grad_norm": 0.3065621075125215,
+      "learning_rate": 8.345463723080816e-08,
+      "loss": 0.4738,
+      "step": 5629
+    },
+    {
+      "epoch": 13.765281173594133,
+      "grad_norm": 0.2567670482454851,
+      "learning_rate": 8.312694145087536e-08,
+      "loss": 0.394,
+      "step": 5630
+    },
+    {
+      "epoch": 13.767726161369193,
+      "grad_norm": 0.3002023303836253,
+      "learning_rate": 8.279987943038387e-08,
+      "loss": 1.0003,
+      "step": 5631
+    },
+    {
+      "epoch": 13.770171149144254,
+      "grad_norm": 0.29251310064018576,
+      "learning_rate": 8.247345125509731e-08,
+      "loss": 0.4632,
+      "step": 5632
+    },
+    {
+      "epoch": 13.772616136919316,
+      "grad_norm": 0.38975783466576047,
+      "learning_rate": 8.214765701061195e-08,
+      "loss": 0.493,
+      "step": 5633
+    },
+    {
+      "epoch": 13.775061124694377,
+      "grad_norm": 0.3813235569340884,
+      "learning_rate": 8.182249678235915e-08,
+      "loss": 0.7381,
+      "step": 5634
+    },
+    {
+      "epoch": 13.777506112469437,
+      "grad_norm": 0.19287406815628588,
+      "learning_rate": 8.149797065560295e-08,
+      "loss": 0.3958,
+      "step": 5635
+    },
+    {
+      "epoch": 13.7799511002445,
+      "grad_norm": 0.3013083609596784,
+      "learning_rate": 8.117407871544164e-08,
+      "loss": 0.3714,
+      "step": 5636
+    },
+    {
+      "epoch": 13.78239608801956,
+      "grad_norm": 0.1922372539946887,
+      "learning_rate": 8.085082104680703e-08,
+      "loss": 0.6339,
+      "step": 5637
+    },
+    {
+      "epoch": 13.78484107579462,
+      "grad_norm": 0.2560376188748304,
+      "learning_rate": 8.052819773446491e-08,
+      "loss": 0.3944,
+      "step": 5638
+    },
+    {
+      "epoch": 13.787286063569683,
+      "grad_norm": 0.2649681491755055,
+      "learning_rate": 8.020620886301455e-08,
+      "loss": 1.0834,
+      "step": 5639
+    },
+    {
+      "epoch": 13.789731051344743,
+      "grad_norm": 0.23791159543852178,
+      "learning_rate": 7.988485451688816e-08,
+      "loss": 0.6581,
+      "step": 5640
+    },
+    {
+      "epoch": 13.792176039119804,
+      "grad_norm": 0.2576480953647651,
+      "learning_rate": 7.956413478035302e-08,
+      "loss": 0.3729,
+      "step": 5641
+    },
+    {
+      "epoch": 13.794621026894866,
+      "grad_norm": 0.19345502612854307,
+      "learning_rate": 7.924404973750882e-08,
+      "loss": 0.4756,
+      "step": 5642
+    },
+    {
+      "epoch": 13.797066014669927,
+      "grad_norm": 0.33689254297536514,
+      "learning_rate": 7.892459947228898e-08,
+      "loss": 0.4492,
+      "step": 5643
+    },
+    {
+      "epoch": 13.799511002444987,
+      "grad_norm": 0.2661417476710071,
+      "learning_rate": 7.860578406846119e-08,
+      "loss": 0.402,
+      "step": 5644
+    },
+    {
+      "epoch": 13.80195599022005,
+      "grad_norm": 0.21273356311174985,
+      "learning_rate": 7.828760360962557e-08,
+      "loss": 0.4415,
+      "step": 5645
+    },
+    {
+      "epoch": 13.80440097799511,
+      "grad_norm": 0.16459861997507153,
+      "learning_rate": 7.797005817921671e-08,
+      "loss": 0.5038,
+      "step": 5646
+    },
+    {
+      "epoch": 13.80684596577017,
+      "grad_norm": 0.28082257240417136,
+      "learning_rate": 7.765314786050193e-08,
+      "loss": 0.5834,
+      "step": 5647
+    },
+    {
+      "epoch": 13.809290953545233,
+      "grad_norm": 0.20134198057690336,
+      "learning_rate": 7.73368727365828e-08,
+      "loss": 0.3513,
+      "step": 5648
+    },
+    {
+      "epoch": 13.811735941320293,
+      "grad_norm": 0.22177385520647788,
+      "learning_rate": 7.702123289039381e-08,
+      "loss": 0.4342,
+      "step": 5649
+    },
+    {
+      "epoch": 13.814180929095354,
+      "grad_norm": 0.258314550031699,
+      "learning_rate": 7.670622840470266e-08,
+      "loss": 0.9537,
+      "step": 5650
+    },
+    {
+      "epoch": 13.816625916870416,
+      "grad_norm": 0.275645109158862,
+      "learning_rate": 7.639185936211102e-08,
+      "loss": 0.7908,
+      "step": 5651
+    },
+    {
+      "epoch": 13.819070904645477,
+      "grad_norm": 0.3100151195823987,
+      "learning_rate": 7.607812584505381e-08,
+      "loss": 0.3885,
+      "step": 5652
+    },
+    {
+      "epoch": 13.821515892420537,
+      "grad_norm": 0.2501411160924779,
+      "learning_rate": 7.576502793579854e-08,
+      "loss": 0.3672,
+      "step": 5653
+    },
+    {
+      "epoch": 13.8239608801956,
+      "grad_norm": 0.26189423340120893,
+      "learning_rate": 7.545256571644732e-08,
+      "loss": 0.4584,
+      "step": 5654
+    },
+    {
+      "epoch": 13.82640586797066,
+      "grad_norm": 0.24923541158079357,
+      "learning_rate": 7.514073926893434e-08,
+      "loss": 0.7607,
+      "step": 5655
+    },
+    {
+      "epoch": 13.82885085574572,
+      "grad_norm": 0.22822968553625902,
+      "learning_rate": 7.482954867502778e-08,
+      "loss": 0.3692,
+      "step": 5656
+    },
+    {
+      "epoch": 13.831295843520783,
+      "grad_norm": 0.31626538773344065,
+      "learning_rate": 7.451899401632934e-08,
+      "loss": 0.4656,
+      "step": 5657
+    },
+    {
+      "epoch": 13.833740831295843,
+      "grad_norm": 0.37669617118209225,
+      "learning_rate": 7.420907537427274e-08,
+      "loss": 0.5186,
+      "step": 5658
+    },
+    {
+      "epoch": 13.836185819070904,
+      "grad_norm": 0.2725636513235219,
+      "learning_rate": 7.38997928301266e-08,
+      "loss": 0.71,
+      "step": 5659
+    },
+    {
+      "epoch": 13.838630806845966,
+      "grad_norm": 0.19050397559028576,
+      "learning_rate": 7.359114646499132e-08,
+      "loss": 0.4655,
+      "step": 5660
+    },
+    {
+      "epoch": 13.841075794621027,
+      "grad_norm": 0.29809880828616875,
+      "learning_rate": 7.328313635980133e-08,
+      "loss": 0.3895,
+      "step": 5661
+    },
+    {
+      "epoch": 13.843520782396087,
+      "grad_norm": 0.26127137286759944,
+      "learning_rate": 7.297576259532424e-08,
+      "loss": 0.5182,
+      "step": 5662
+    },
+    {
+      "epoch": 13.84596577017115,
+      "grad_norm": 0.2221476818756455,
+      "learning_rate": 7.266902525215975e-08,
+      "loss": 0.3993,
+      "step": 5663
+    },
+    {
+      "epoch": 13.84841075794621,
+      "grad_norm": 0.24812101016135327,
+      "learning_rate": 7.236292441074183e-08,
+      "loss": 1.3562,
+      "step": 5664
+    },
+    {
+      "epoch": 13.85085574572127,
+      "grad_norm": 0.23690908231347022,
+      "learning_rate": 7.205746015133714e-08,
+      "loss": 0.5625,
+      "step": 5665
+    },
+    {
+      "epoch": 13.853300733496333,
+      "grad_norm": 0.2254165759273135,
+      "learning_rate": 7.175263255404519e-08,
+      "loss": 0.4106,
+      "step": 5666
+    },
+    {
+      "epoch": 13.855745721271393,
+      "grad_norm": 0.27894230635927325,
+      "learning_rate": 7.1448441698799e-08,
+      "loss": 0.4381,
+      "step": 5667
+    },
+    {
+      "epoch": 13.858190709046454,
+      "grad_norm": 0.2970428587120057,
+      "learning_rate": 7.11448876653642e-08,
+      "loss": 0.4672,
+      "step": 5668
+    },
+    {
+      "epoch": 13.860635696821516,
+      "grad_norm": 0.2188016856271119,
+      "learning_rate": 7.084197053333935e-08,
+      "loss": 0.4248,
+      "step": 5669
+    },
+    {
+      "epoch": 13.863080684596577,
+      "grad_norm": 0.22295590564714088,
+      "learning_rate": 7.053969038215675e-08,
+      "loss": 0.3437,
+      "step": 5670
+    },
+    {
+      "epoch": 13.865525672371637,
+      "grad_norm": 0.2658431040469962,
+      "learning_rate": 7.02380472910813e-08,
+      "loss": 0.7699,
+      "step": 5671
+    },
+    {
+      "epoch": 13.8679706601467,
+      "grad_norm": 0.24060295044355764,
+      "learning_rate": 6.993704133921031e-08,
+      "loss": 0.456,
+      "step": 5672
+    },
+    {
+      "epoch": 13.87041564792176,
+      "grad_norm": 0.2504773059857715,
+      "learning_rate": 6.963667260547424e-08,
+      "loss": 0.3673,
+      "step": 5673
+    },
+    {
+      "epoch": 13.87286063569682,
+      "grad_norm": 0.2320020000388462,
+      "learning_rate": 6.933694116863732e-08,
+      "loss": 0.768,
+      "step": 5674
+    },
+    {
+      "epoch": 13.875305623471883,
+      "grad_norm": 0.35343926487640603,
+      "learning_rate": 6.903784710729556e-08,
+      "loss": 0.9156,
+      "step": 5675
+    },
+    {
+      "epoch": 13.877750611246944,
+      "grad_norm": 0.14781233743998162,
+      "learning_rate": 6.873939049987815e-08,
+      "loss": 0.4237,
+      "step": 5676
+    },
+    {
+      "epoch": 13.880195599022004,
+      "grad_norm": 0.2405914601348876,
+      "learning_rate": 6.84415714246478e-08,
+      "loss": 1.1149,
+      "step": 5677
+    },
+    {
+      "epoch": 13.882640586797066,
+      "grad_norm": 0.19366228739159255,
+      "learning_rate": 6.814438995969869e-08,
+      "loss": 0.7345,
+      "step": 5678
+    },
+    {
+      "epoch": 13.885085574572127,
+      "grad_norm": 0.3126684624303472,
+      "learning_rate": 6.784784618295931e-08,
+      "loss": 0.3926,
+      "step": 5679
+    },
+    {
+      "epoch": 13.887530562347187,
+      "grad_norm": 0.19913135887303235,
+      "learning_rate": 6.755194017218969e-08,
+      "loss": 0.6636,
+      "step": 5680
+    },
+    {
+      "epoch": 13.88997555012225,
+      "grad_norm": 0.2764871408509584,
+      "learning_rate": 6.725667200498359e-08,
+      "loss": 0.424,
+      "step": 5681
+    },
+    {
+      "epoch": 13.89242053789731,
+      "grad_norm": 0.27110874004422386,
+      "learning_rate": 6.696204175876686e-08,
+      "loss": 0.8935,
+      "step": 5682
+    },
+    {
+      "epoch": 13.894865525672373,
+      "grad_norm": 0.36660469013234076,
+      "learning_rate": 6.666804951079825e-08,
+      "loss": 0.442,
+      "step": 5683
+    },
+    {
+      "epoch": 13.897310513447433,
+      "grad_norm": 0.23605112407126835,
+      "learning_rate": 6.637469533816942e-08,
+      "loss": 0.5302,
+      "step": 5684
+    },
+    {
+      "epoch": 13.899755501222494,
+      "grad_norm": 0.19421516023246746,
+      "learning_rate": 6.608197931780496e-08,
+      "loss": 0.3436,
+      "step": 5685
+    },
+    {
+      "epoch": 13.902200488997554,
+      "grad_norm": 0.2557163795818338,
+      "learning_rate": 6.57899015264607e-08,
+      "loss": 0.4836,
+      "step": 5686
+    },
+    {
+      "epoch": 13.904645476772616,
+      "grad_norm": 0.24336404126836905,
+      "learning_rate": 6.549846204072674e-08,
+      "loss": 0.8938,
+      "step": 5687
+    },
+    {
+      "epoch": 13.907090464547677,
+      "grad_norm": 0.21311180509306366,
+      "learning_rate": 6.52076609370253e-08,
+      "loss": 0.7128,
+      "step": 5688
+    },
+    {
+      "epoch": 13.90953545232274,
+      "grad_norm": 0.3119674325525338,
+      "learning_rate": 6.491749829161064e-08,
+      "loss": 0.6151,
+      "step": 5689
+    },
+    {
+      "epoch": 13.9119804400978,
+      "grad_norm": 0.22805670842899523,
+      "learning_rate": 6.462797418057026e-08,
+      "loss": 0.555,
+      "step": 5690
+    },
+    {
+      "epoch": 13.91442542787286,
+      "grad_norm": 0.19712025178994977,
+      "learning_rate": 6.433908867982396e-08,
+      "loss": 0.7656,
+      "step": 5691
+    },
+    {
+      "epoch": 13.91687041564792,
+      "grad_norm": 0.2783258649754196,
+      "learning_rate": 6.40508418651245e-08,
+      "loss": 0.4223,
+      "step": 5692
+    },
+    {
+      "epoch": 13.919315403422983,
+      "grad_norm": 0.30330294826347237,
+      "learning_rate": 6.37632338120564e-08,
+      "loss": 0.4482,
+      "step": 5693
+    },
+    {
+      "epoch": 13.921760391198044,
+      "grad_norm": 0.23712160942872165,
+      "learning_rate": 6.347626459603712e-08,
+      "loss": 0.4084,
+      "step": 5694
+    },
+    {
+      "epoch": 13.924205378973106,
+      "grad_norm": 0.30705814404516824,
+      "learning_rate": 6.318993429231701e-08,
+      "loss": 0.3655,
+      "step": 5695
+    },
+    {
+      "epoch": 13.926650366748166,
+      "grad_norm": 0.3847960084801866,
+      "learning_rate": 6.29042429759777e-08,
+      "loss": 0.5673,
+      "step": 5696
+    },
+    {
+      "epoch": 13.929095354523227,
+      "grad_norm": 0.3173446863891032,
+      "learning_rate": 6.26191907219345e-08,
+      "loss": 0.4031,
+      "step": 5697
+    },
+    {
+      "epoch": 13.93154034229829,
+      "grad_norm": 0.28772152516507377,
+      "learning_rate": 6.233477760493434e-08,
+      "loss": 0.325,
+      "step": 5698
+    },
+    {
+      "epoch": 13.93398533007335,
+      "grad_norm": 0.4109835014227078,
+      "learning_rate": 6.205100369955697e-08,
+      "loss": 0.3943,
+      "step": 5699
+    },
+    {
+      "epoch": 13.93643031784841,
+      "grad_norm": 0.1813254542501591,
+      "learning_rate": 6.176786908021454e-08,
+      "loss": 0.7648,
+      "step": 5700
+    },
+    {
+      "epoch": 13.938875305623473,
+      "grad_norm": 0.19124120902598046,
+      "learning_rate": 6.148537382115072e-08,
+      "loss": 0.6457,
+      "step": 5701
+    },
+    {
+      "epoch": 13.941320293398533,
+      "grad_norm": 0.289905686606393,
+      "learning_rate": 6.120351799644348e-08,
+      "loss": 0.7174,
+      "step": 5702
+    },
+    {
+      "epoch": 13.943765281173594,
+      "grad_norm": 0.1961814892727553,
+      "learning_rate": 6.092230168000091e-08,
+      "loss": 0.4523,
+      "step": 5703
+    },
+    {
+      "epoch": 13.946210268948656,
+      "grad_norm": 0.1814368265122953,
+      "learning_rate": 6.064172494556486e-08,
+      "loss": 0.3739,
+      "step": 5704
+    },
+    {
+      "epoch": 13.948655256723717,
+      "grad_norm": 0.25653401971813755,
+      "learning_rate": 6.03617878667087e-08,
+      "loss": 1.2176,
+      "step": 5705
+    },
+    {
+      "epoch": 13.951100244498777,
+      "grad_norm": 0.26187902538561575,
+      "learning_rate": 6.008249051683846e-08,
+      "loss": 0.7135,
+      "step": 5706
+    },
+    {
+      "epoch": 13.95354523227384,
+      "grad_norm": 0.2238991789221987,
+      "learning_rate": 5.980383296919246e-08,
+      "loss": 0.4338,
+      "step": 5707
+    },
+    {
+      "epoch": 13.9559902200489,
+      "grad_norm": 0.2924162212545847,
+      "learning_rate": 5.952581529684032e-08,
+      "loss": 0.4481,
+      "step": 5708
+    },
+    {
+      "epoch": 13.95843520782396,
+      "grad_norm": 0.25905106056044247,
+      "learning_rate": 5.9248437572685394e-08,
+      "loss": 0.572,
+      "step": 5709
+    },
+    {
+      "epoch": 13.960880195599023,
+      "grad_norm": 0.3037916468831237,
+      "learning_rate": 5.897169986946255e-08,
+      "loss": 0.6279,
+      "step": 5710
+    },
+    {
+      "epoch": 13.963325183374083,
+      "grad_norm": 0.3758677418363692,
+      "learning_rate": 5.869560225973847e-08,
+      "loss": 0.6665,
+      "step": 5711
+    },
+    {
+      "epoch": 13.965770171149144,
+      "grad_norm": 0.3517825453987885,
+      "learning_rate": 5.842014481591191e-08,
+      "loss": 0.3757,
+      "step": 5712
+    },
+    {
+      "epoch": 13.968215158924206,
+      "grad_norm": 0.24231342837645276,
+      "learning_rate": 5.8145327610214517e-08,
+      "loss": 0.4617,
+      "step": 5713
+    },
+    {
+      "epoch": 13.970660146699267,
+      "grad_norm": 0.42769162915816133,
+      "learning_rate": 5.7871150714709775e-08,
+      "loss": 0.6231,
+      "step": 5714
+    },
+    {
+      "epoch": 13.973105134474327,
+      "grad_norm": 0.38546248623982654,
+      "learning_rate": 5.759761420129323e-08,
+      "loss": 0.3858,
+      "step": 5715
+    },
+    {
+      "epoch": 13.97555012224939,
+      "grad_norm": 0.22900211024519998,
+      "learning_rate": 5.732471814169194e-08,
+      "loss": 0.4308,
+      "step": 5716
+    },
+    {
+      "epoch": 13.97799511002445,
+      "grad_norm": 0.22073803768835076,
+      "learning_rate": 5.705246260746561e-08,
+      "loss": 0.8944,
+      "step": 5717
+    },
+    {
+      "epoch": 13.98044009779951,
+      "grad_norm": 0.24392192024486242,
+      "learning_rate": 5.6780847670006024e-08,
+      "loss": 0.4954,
+      "step": 5718
+    },
+    {
+      "epoch": 13.982885085574573,
+      "grad_norm": 0.2282555014573532,
+      "learning_rate": 5.650987340053704e-08,
+      "loss": 0.805,
+      "step": 5719
+    },
+    {
+      "epoch": 13.985330073349633,
+      "grad_norm": 0.3478591864448769,
+      "learning_rate": 5.623953987011405e-08,
+      "loss": 0.5645,
+      "step": 5720
+    },
+    {
+      "epoch": 13.987775061124694,
+      "grad_norm": 0.19278412553455818,
+      "learning_rate": 5.5969847149624504e-08,
+      "loss": 0.4398,
+      "step": 5721
+    },
+    {
+      "epoch": 13.990220048899756,
+      "grad_norm": 0.18703972728150048,
+      "learning_rate": 5.5700795309788504e-08,
+      "loss": 0.5281,
+      "step": 5722
+    },
+    {
+      "epoch": 13.992665036674817,
+      "grad_norm": 0.30102840640489964,
+      "learning_rate": 5.543238442115684e-08,
+      "loss": 1.2257,
+      "step": 5723
+    },
+    {
+      "epoch": 13.995110024449877,
+      "grad_norm": 0.23766909768482902,
+      "learning_rate": 5.516461455411348e-08,
+      "loss": 0.6255,
+      "step": 5724
+    },
+    {
+      "epoch": 13.99755501222494,
+      "grad_norm": 0.26714910233036704,
+      "learning_rate": 5.489748577887449e-08,
+      "loss": 0.8751,
+      "step": 5725
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.3298800115212367,
+      "learning_rate": 5.463099816548578e-08,
+      "loss": 0.3495,
+      "step": 5726
+    },
+    {
+      "epoch": 14.00244498777506,
+      "grad_norm": 0.22192010782224986,
+      "learning_rate": 5.436515178382784e-08,
+      "loss": 0.6872,
+      "step": 5727
+    },
+    {
+      "epoch": 14.004889975550123,
+      "grad_norm": 0.2879347368121287,
+      "learning_rate": 5.409994670361046e-08,
+      "loss": 0.551,
+      "step": 5728
+    },
+    {
+      "epoch": 14.007334963325183,
+      "grad_norm": 0.2504483246657384,
+      "learning_rate": 5.383538299437718e-08,
+      "loss": 0.5138,
+      "step": 5729
+    },
+    {
+      "epoch": 14.009779951100244,
+      "grad_norm": 0.35599323991963117,
+      "learning_rate": 5.357146072550279e-08,
+      "loss": 0.3974,
+      "step": 5730
+    },
+    {
+      "epoch": 14.012224938875306,
+      "grad_norm": 0.23943331782735083,
+      "learning_rate": 5.330817996619331e-08,
+      "loss": 0.5203,
+      "step": 5731
+    },
+    {
+      "epoch": 14.014669926650367,
+      "grad_norm": 0.22062892944837517,
+      "learning_rate": 5.3045540785487126e-08,
+      "loss": 0.8056,
+      "step": 5732
+    },
+    {
+      "epoch": 14.017114914425427,
+      "grad_norm": 0.2828544134546111,
+      "learning_rate": 5.2783543252254155e-08,
+      "loss": 0.7433,
+      "step": 5733
+    },
+    {
+      "epoch": 14.01955990220049,
+      "grad_norm": 0.21081910692921516,
+      "learning_rate": 5.252218743519638e-08,
+      "loss": 0.5019,
+      "step": 5734
+    },
+    {
+      "epoch": 14.02200488997555,
+      "grad_norm": 0.2669155030292534,
+      "learning_rate": 5.226147340284732e-08,
+      "loss": 0.8479,
+      "step": 5735
+    },
+    {
+      "epoch": 14.02444987775061,
+      "grad_norm": 0.2790013268955985,
+      "learning_rate": 5.2001401223572e-08,
+      "loss": 0.6977,
+      "step": 5736
+    },
+    {
+      "epoch": 14.026894865525673,
+      "grad_norm": 0.32816481984915846,
+      "learning_rate": 5.1741970965567265e-08,
+      "loss": 0.5888,
+      "step": 5737
+    },
+    {
+      "epoch": 14.029339853300733,
+      "grad_norm": 0.28714087407627525,
+      "learning_rate": 5.148318269686176e-08,
+      "loss": 0.5297,
+      "step": 5738
+    },
+    {
+      "epoch": 14.031784841075794,
+      "grad_norm": 0.18498835416486617,
+      "learning_rate": 5.122503648531535e-08,
+      "loss": 0.3846,
+      "step": 5739
+    },
+    {
+      "epoch": 14.034229828850856,
+      "grad_norm": 0.3183783486535158,
+      "learning_rate": 5.096753239862057e-08,
+      "loss": 0.955,
+      "step": 5740
+    },
+    {
+      "epoch": 14.036674816625917,
+      "grad_norm": 0.34311010252291824,
+      "learning_rate": 5.071067050430034e-08,
+      "loss": 0.6715,
+      "step": 5741
+    },
+    {
+      "epoch": 14.039119804400977,
+      "grad_norm": 0.24539892462208537,
+      "learning_rate": 5.045445086970996e-08,
+      "loss": 0.9961,
+      "step": 5742
+    },
+    {
+      "epoch": 14.04156479217604,
+      "grad_norm": 0.2516239191165433,
+      "learning_rate": 5.0198873562035955e-08,
+      "loss": 0.4897,
+      "step": 5743
+    },
+    {
+      "epoch": 14.0440097799511,
+      "grad_norm": 0.3866157286017683,
+      "learning_rate": 4.9943938648296387e-08,
+      "loss": 0.8461,
+      "step": 5744
+    },
+    {
+      "epoch": 14.04645476772616,
+      "grad_norm": 0.2361703214114367,
+      "learning_rate": 4.968964619534139e-08,
+      "loss": 0.4465,
+      "step": 5745
+    },
+    {
+      "epoch": 14.048899755501223,
+      "grad_norm": 0.238516751797963,
+      "learning_rate": 4.943599626985207e-08,
+      "loss": 0.4396,
+      "step": 5746
+    },
+    {
+      "epoch": 14.051344743276283,
+      "grad_norm": 0.307040114652679,
+      "learning_rate": 4.918298893834161e-08,
+      "loss": 0.4036,
+      "step": 5747
+    },
+    {
+      "epoch": 14.053789731051344,
+      "grad_norm": 0.176317695365481,
+      "learning_rate": 4.893062426715361e-08,
+      "loss": 0.573,
+      "step": 5748
+    },
+    {
+      "epoch": 14.056234718826406,
+      "grad_norm": 0.25214111050787685,
+      "learning_rate": 4.867890232246431e-08,
+      "loss": 0.3647,
+      "step": 5749
+    },
+    {
+      "epoch": 14.058679706601467,
+      "grad_norm": 0.22568228292484724,
+      "learning_rate": 4.8427823170281166e-08,
+      "loss": 0.6943,
+      "step": 5750
+    },
+    {
+      "epoch": 14.061124694376527,
+      "grad_norm": 0.25393174608989827,
+      "learning_rate": 4.8177386876442354e-08,
+      "loss": 0.837,
+      "step": 5751
+    },
+    {
+      "epoch": 14.06356968215159,
+      "grad_norm": 0.16866394044734753,
+      "learning_rate": 4.792759350661841e-08,
+      "loss": 0.5495,
+      "step": 5752
+    },
+    {
+      "epoch": 14.06601466992665,
+      "grad_norm": 0.31340198118609325,
+      "learning_rate": 4.7678443126310814e-08,
+      "loss": 0.4447,
+      "step": 5753
+    },
+    {
+      "epoch": 14.06845965770171,
+      "grad_norm": 0.218632707539735,
+      "learning_rate": 4.742993580085231e-08,
+      "loss": 0.696,
+      "step": 5754
+    },
+    {
+      "epoch": 14.070904645476773,
+      "grad_norm": 0.2582786896053266,
+      "learning_rate": 4.7182071595407164e-08,
+      "loss": 0.6007,
+      "step": 5755
+    },
+    {
+      "epoch": 14.073349633251834,
+      "grad_norm": 0.22280107086787024,
+      "learning_rate": 4.693485057497144e-08,
+      "loss": 0.7417,
+      "step": 5756
+    },
+    {
+      "epoch": 14.075794621026894,
+      "grad_norm": 0.2979382944684062,
+      "learning_rate": 4.668827280437188e-08,
+      "loss": 0.7841,
+      "step": 5757
+    },
+    {
+      "epoch": 14.078239608801956,
+      "grad_norm": 0.2839969661401148,
+      "learning_rate": 4.644233834826678e-08,
+      "loss": 0.5157,
+      "step": 5758
+    },
+    {
+      "epoch": 14.080684596577017,
+      "grad_norm": 0.19479949942606686,
+      "learning_rate": 4.619704727114593e-08,
+      "loss": 0.4837,
+      "step": 5759
+    },
+    {
+      "epoch": 14.083129584352077,
+      "grad_norm": 0.20972120580148954,
+      "learning_rate": 4.5952399637330114e-08,
+      "loss": 0.7799,
+      "step": 5760
+    },
+    {
+      "epoch": 14.08557457212714,
+      "grad_norm": 0.26987748208919193,
+      "learning_rate": 4.570839551097162e-08,
+      "loss": 0.5006,
+      "step": 5761
+    },
+    {
+      "epoch": 14.0880195599022,
+      "grad_norm": 0.27305830778720164,
+      "learning_rate": 4.546503495605398e-08,
+      "loss": 0.8407,
+      "step": 5762
+    },
+    {
+      "epoch": 14.09046454767726,
+      "grad_norm": 0.35038717884448706,
+      "learning_rate": 4.5222318036391985e-08,
+      "loss": 0.4351,
+      "step": 5763
+    },
+    {
+      "epoch": 14.092909535452323,
+      "grad_norm": 0.32793521890272304,
+      "learning_rate": 4.498024481563112e-08,
+      "loss": 0.3932,
+      "step": 5764
+    },
+    {
+      "epoch": 14.095354523227384,
+      "grad_norm": 0.222753086369839,
+      "learning_rate": 4.4738815357248664e-08,
+      "loss": 0.544,
+      "step": 5765
+    },
+    {
+      "epoch": 14.097799511002446,
+      "grad_norm": 0.4282622543133458,
+      "learning_rate": 4.449802972455342e-08,
+      "loss": 0.5178,
+      "step": 5766
+    },
+    {
+      "epoch": 14.100244498777506,
+      "grad_norm": 0.3569361387002512,
+      "learning_rate": 4.425788798068459e-08,
+      "loss": 0.663,
+      "step": 5767
+    },
+    {
+      "epoch": 14.102689486552567,
+      "grad_norm": 0.2548868236955468,
+      "learning_rate": 4.401839018861265e-08,
+      "loss": 0.7352,
+      "step": 5768
+    },
+    {
+      "epoch": 14.10513447432763,
+      "grad_norm": 0.3939596972060168,
+      "learning_rate": 4.3779536411139856e-08,
+      "loss": 0.8855,
+      "step": 5769
+    },
+    {
+      "epoch": 14.10757946210269,
+      "grad_norm": 0.24702342476808542,
+      "learning_rate": 4.354132671089861e-08,
+      "loss": 0.72,
+      "step": 5770
+    },
+    {
+      "epoch": 14.11002444987775,
+      "grad_norm": 0.32146872624906675,
+      "learning_rate": 4.33037611503534e-08,
+      "loss": 0.4705,
+      "step": 5771
+    },
+    {
+      "epoch": 14.112469437652813,
+      "grad_norm": 0.22284450075668427,
+      "learning_rate": 4.306683979179882e-08,
+      "loss": 0.378,
+      "step": 5772
+    },
+    {
+      "epoch": 14.114914425427873,
+      "grad_norm": 0.25090653870013874,
+      "learning_rate": 4.2830562697361856e-08,
+      "loss": 0.3851,
+      "step": 5773
+    },
+    {
+      "epoch": 14.117359413202934,
+      "grad_norm": 0.33274588254089627,
+      "learning_rate": 4.2594929928999054e-08,
+      "loss": 0.5454,
+      "step": 5774
+    },
+    {
+      "epoch": 14.119804400977996,
+      "grad_norm": 0.3419712093315197,
+      "learning_rate": 4.235994154849904e-08,
+      "loss": 0.4075,
+      "step": 5775
+    },
+    {
+      "epoch": 14.122249388753056,
+      "grad_norm": 0.28474404662421293,
+      "learning_rate": 4.212559761748086e-08,
+      "loss": 0.522,
+      "step": 5776
+    },
+    {
+      "epoch": 14.124694376528117,
+      "grad_norm": 0.21447609684088564,
+      "learning_rate": 4.1891898197395367e-08,
+      "loss": 0.3952,
+      "step": 5777
+    },
+    {
+      "epoch": 14.12713936430318,
+      "grad_norm": 0.24304734405654124,
+      "learning_rate": 4.165884334952352e-08,
+      "loss": 0.8457,
+      "step": 5778
+    },
+    {
+      "epoch": 14.12958435207824,
+      "grad_norm": 0.26982376550642384,
+      "learning_rate": 4.142643313497785e-08,
+      "loss": 0.3908,
+      "step": 5779
+    },
+    {
+      "epoch": 14.1320293398533,
+      "grad_norm": 0.17896362938785879,
+      "learning_rate": 4.119466761470153e-08,
+      "loss": 0.598,
+      "step": 5780
+    },
+    {
+      "epoch": 14.134474327628363,
+      "grad_norm": 0.18992077251970388,
+      "learning_rate": 4.096354684946846e-08,
+      "loss": 0.5713,
+      "step": 5781
+    },
+    {
+      "epoch": 14.136919315403423,
+      "grad_norm": 0.26972108640436354,
+      "learning_rate": 4.0733070899884316e-08,
+      "loss": 0.5261,
+      "step": 5782
+    },
+    {
+      "epoch": 14.139364303178484,
+      "grad_norm": 0.3468339580445499,
+      "learning_rate": 4.050323982638521e-08,
+      "loss": 0.9662,
+      "step": 5783
+    },
+    {
+      "epoch": 14.141809290953546,
+      "grad_norm": 0.1732534930877171,
+      "learning_rate": 4.0274053689237645e-08,
+      "loss": 0.5134,
+      "step": 5784
+    },
+    {
+      "epoch": 14.144254278728607,
+      "grad_norm": 0.28928495557866196,
+      "learning_rate": 4.0045512548539665e-08,
+      "loss": 0.843,
+      "step": 5785
+    },
+    {
+      "epoch": 14.146699266503667,
+      "grad_norm": 0.3186053722928193,
+      "learning_rate": 3.9817616464219996e-08,
+      "loss": 0.6506,
+      "step": 5786
+    },
+    {
+      "epoch": 14.14914425427873,
+      "grad_norm": 0.25442512894081876,
+      "learning_rate": 3.959036549603834e-08,
+      "loss": 0.7672,
+      "step": 5787
+    },
+    {
+      "epoch": 14.15158924205379,
+      "grad_norm": 0.2339412941506138,
+      "learning_rate": 3.93637597035848e-08,
+      "loss": 0.5266,
+      "step": 5788
+    },
+    {
+      "epoch": 14.15403422982885,
+      "grad_norm": 0.2474792418936597,
+      "learning_rate": 3.913779914628074e-08,
+      "loss": 0.3655,
+      "step": 5789
+    },
+    {
+      "epoch": 14.156479217603913,
+      "grad_norm": 0.2871799981087508,
+      "learning_rate": 3.8912483883378486e-08,
+      "loss": 0.8848,
+      "step": 5790
+    },
+    {
+      "epoch": 14.158924205378973,
+      "grad_norm": 0.19100529412154701,
+      "learning_rate": 3.868781397396049e-08,
+      "loss": 0.4238,
+      "step": 5791
+    },
+    {
+      "epoch": 14.161369193154034,
+      "grad_norm": 0.28444728316033757,
+      "learning_rate": 3.846378947694046e-08,
+      "loss": 0.7894,
+      "step": 5792
+    },
+    {
+      "epoch": 14.163814180929096,
+      "grad_norm": 0.17480592619873836,
+      "learning_rate": 3.824041045106308e-08,
+      "loss": 0.6868,
+      "step": 5793
+    },
+    {
+      "epoch": 14.166259168704157,
+      "grad_norm": 0.2916003025169568,
+      "learning_rate": 3.8017676954902884e-08,
+      "loss": 1.0485,
+      "step": 5794
+    },
+    {
+      "epoch": 14.168704156479217,
+      "grad_norm": 0.24219757604746475,
+      "learning_rate": 3.7795589046865924e-08,
+      "loss": 0.607,
+      "step": 5795
+    },
+    {
+      "epoch": 14.17114914425428,
+      "grad_norm": 0.282070639694405,
+      "learning_rate": 3.757414678518867e-08,
+      "loss": 0.5482,
+      "step": 5796
+    },
+    {
+      "epoch": 14.17359413202934,
+      "grad_norm": 0.32637063388131127,
+      "learning_rate": 3.735335022793857e-08,
+      "loss": 0.4665,
+      "step": 5797
+    },
+    {
+      "epoch": 14.1760391198044,
+      "grad_norm": 0.32643344181320255,
+      "learning_rate": 3.713319943301319e-08,
+      "loss": 0.7588,
+      "step": 5798
+    },
+    {
+      "epoch": 14.178484107579463,
+      "grad_norm": 0.2772353649858115,
+      "learning_rate": 3.691369445814136e-08,
+      "loss": 0.643,
+      "step": 5799
+    },
+    {
+      "epoch": 14.180929095354523,
+      "grad_norm": 0.3227630973400612,
+      "learning_rate": 3.669483536088231e-08,
+      "loss": 1.5477,
+      "step": 5800
+    },
+    {
+      "epoch": 14.183374083129584,
+      "grad_norm": 0.272374224596172,
+      "learning_rate": 3.6476622198625965e-08,
+      "loss": 0.4688,
+      "step": 5801
+    },
+    {
+      "epoch": 14.185819070904646,
+      "grad_norm": 0.2921798931564791,
+      "learning_rate": 3.625905502859239e-08,
+      "loss": 0.4324,
+      "step": 5802
+    },
+    {
+      "epoch": 14.188264058679707,
+      "grad_norm": 0.24301959083423358,
+      "learning_rate": 3.6042133907833176e-08,
+      "loss": 0.4456,
+      "step": 5803
+    },
+    {
+      "epoch": 14.190709046454767,
+      "grad_norm": 0.2509397771046098,
+      "learning_rate": 3.582585889322976e-08,
+      "loss": 0.8093,
+      "step": 5804
+    },
+    {
+      "epoch": 14.19315403422983,
+      "grad_norm": 0.21047896067486083,
+      "learning_rate": 3.5610230041494834e-08,
+      "loss": 0.9347,
+      "step": 5805
+    },
+    {
+      "epoch": 14.19559902200489,
+      "grad_norm": 0.2997670530108449,
+      "learning_rate": 3.5395247409170385e-08,
+      "loss": 1.1111,
+      "step": 5806
+    },
+    {
+      "epoch": 14.19804400977995,
+      "grad_norm": 0.22629703693309292,
+      "learning_rate": 3.518091105263022e-08,
+      "loss": 0.4266,
+      "step": 5807
+    },
+    {
+      "epoch": 14.200488997555013,
+      "grad_norm": 0.2430767257480146,
+      "learning_rate": 3.496722102807853e-08,
+      "loss": 0.9914,
+      "step": 5808
+    },
+    {
+      "epoch": 14.202933985330073,
+      "grad_norm": 0.2543193292456342,
+      "learning_rate": 3.475417739154913e-08,
+      "loss": 0.4999,
+      "step": 5809
+    },
+    {
+      "epoch": 14.205378973105134,
+      "grad_norm": 0.31183675571629743,
+      "learning_rate": 3.4541780198907584e-08,
+      "loss": 0.7114,
+      "step": 5810
+    },
+    {
+      "epoch": 14.207823960880196,
+      "grad_norm": 0.31855643673328343,
+      "learning_rate": 3.4330029505848794e-08,
+      "loss": 0.6549,
+      "step": 5811
+    },
+    {
+      "epoch": 14.210268948655257,
+      "grad_norm": 0.32168105736284275,
+      "learning_rate": 3.41189253678989e-08,
+      "loss": 0.4952,
+      "step": 5812
+    },
+    {
+      "epoch": 14.212713936430317,
+      "grad_norm": 0.17470943079997792,
+      "learning_rate": 3.390846784041418e-08,
+      "loss": 0.4032,
+      "step": 5813
+    },
+    {
+      "epoch": 14.21515892420538,
+      "grad_norm": 0.1983161791480925,
+      "learning_rate": 3.3698656978581325e-08,
+      "loss": 0.4699,
+      "step": 5814
+    },
+    {
+      "epoch": 14.21760391198044,
+      "grad_norm": 0.2887570403266946,
+      "learning_rate": 3.348949283741798e-08,
+      "loss": 0.5431,
+      "step": 5815
+    },
+    {
+      "epoch": 14.2200488997555,
+      "grad_norm": 0.23896624940573702,
+      "learning_rate": 3.3280975471771114e-08,
+      "loss": 0.4592,
+      "step": 5816
+    },
+    {
+      "epoch": 14.222493887530563,
+      "grad_norm": 0.2542155307981523,
+      "learning_rate": 3.3073104936318936e-08,
+      "loss": 0.4584,
+      "step": 5817
+    },
+    {
+      "epoch": 14.224938875305623,
+      "grad_norm": 0.2481052644371504,
+      "learning_rate": 3.2865881285570054e-08,
+      "loss": 0.401,
+      "step": 5818
+    },
+    {
+      "epoch": 14.227383863080684,
+      "grad_norm": 0.20676258494954353,
+      "learning_rate": 3.265930457386296e-08,
+      "loss": 0.6551,
+      "step": 5819
+    },
+    {
+      "epoch": 14.229828850855746,
+      "grad_norm": 0.2849613847337719,
+      "learning_rate": 3.245337485536737e-08,
+      "loss": 0.5869,
+      "step": 5820
+    },
+    {
+      "epoch": 14.232273838630807,
+      "grad_norm": 0.3626632418726902,
+      "learning_rate": 3.224809218408231e-08,
+      "loss": 0.7387,
+      "step": 5821
+    },
+    {
+      "epoch": 14.234718826405867,
+      "grad_norm": 0.19636986069558876,
+      "learning_rate": 3.204345661383751e-08,
+      "loss": 0.4553,
+      "step": 5822
+    },
+    {
+      "epoch": 14.23716381418093,
+      "grad_norm": 0.2222285707980301,
+      "learning_rate": 3.1839468198293644e-08,
+      "loss": 0.3726,
+      "step": 5823
+    },
+    {
+      "epoch": 14.23960880195599,
+      "grad_norm": 0.28211746781872943,
+      "learning_rate": 3.163612699094043e-08,
+      "loss": 0.6344,
+      "step": 5824
+    },
+    {
+      "epoch": 14.24205378973105,
+      "grad_norm": 0.3583086516114762,
+      "learning_rate": 3.14334330450991e-08,
+      "loss": 0.6171,
+      "step": 5825
+    },
+    {
+      "epoch": 14.244498777506113,
+      "grad_norm": 0.33559944457654917,
+      "learning_rate": 3.1231386413920485e-08,
+      "loss": 0.7111,
+      "step": 5826
+    },
+    {
+      "epoch": 14.246943765281173,
+      "grad_norm": 0.23220668053602828,
+      "learning_rate": 3.1029987150385796e-08,
+      "loss": 0.4841,
+      "step": 5827
+    },
+    {
+      "epoch": 14.249388753056234,
+      "grad_norm": 0.22459419682948784,
+      "learning_rate": 3.082923530730669e-08,
+      "loss": 0.7535,
+      "step": 5828
+    },
+    {
+      "epoch": 14.251833740831296,
+      "grad_norm": 0.267867488759652,
+      "learning_rate": 3.062913093732467e-08,
+      "loss": 0.5828,
+      "step": 5829
+    },
+    {
+      "epoch": 14.254278728606357,
+      "grad_norm": 0.19388296447291925,
+      "learning_rate": 3.042967409291192e-08,
+      "loss": 0.4412,
+      "step": 5830
+    },
+    {
+      "epoch": 14.256723716381417,
+      "grad_norm": 0.28650977167044434,
+      "learning_rate": 3.023086482637022e-08,
+      "loss": 0.4488,
+      "step": 5831
+    },
+    {
+      "epoch": 14.25916870415648,
+      "grad_norm": 0.3235625557495168,
+      "learning_rate": 3.0032703189832315e-08,
+      "loss": 0.5247,
+      "step": 5832
+    },
+    {
+      "epoch": 14.26161369193154,
+      "grad_norm": 1.6108017619387143,
+      "learning_rate": 2.983518923526052e-08,
+      "loss": 0.805,
+      "step": 5833
+    },
+    {
+      "epoch": 14.2640586797066,
+      "grad_norm": 0.31446525579947765,
+      "learning_rate": 2.9638323014447577e-08,
+      "loss": 0.7721,
+      "step": 5834
+    },
+    {
+      "epoch": 14.266503667481663,
+      "grad_norm": 0.24807680695131962,
+      "learning_rate": 2.944210457901636e-08,
+      "loss": 0.5272,
+      "step": 5835
+    },
+    {
+      "epoch": 14.268948655256724,
+      "grad_norm": 0.2745691982755478,
+      "learning_rate": 2.9246533980419878e-08,
+      "loss": 0.5274,
+      "step": 5836
+    },
+    {
+      "epoch": 14.271393643031784,
+      "grad_norm": 0.3368195697390879,
+      "learning_rate": 2.9051611269940718e-08,
+      "loss": 0.6726,
+      "step": 5837
+    },
+    {
+      "epoch": 14.273838630806846,
+      "grad_norm": 0.1933341907948257,
+      "learning_rate": 2.8857336498692712e-08,
+      "loss": 0.3729,
+      "step": 5838
+    },
+    {
+      "epoch": 14.276283618581907,
+      "grad_norm": 0.2594667788203325,
+      "learning_rate": 2.8663709717619003e-08,
+      "loss": 0.4295,
+      "step": 5839
+    },
+    {
+      "epoch": 14.278728606356967,
+      "grad_norm": 0.2639111743020132,
+      "learning_rate": 2.847073097749259e-08,
+      "loss": 0.383,
+      "step": 5840
+    },
+    {
+      "epoch": 14.28117359413203,
+      "grad_norm": 0.26811570536981266,
+      "learning_rate": 2.8278400328917434e-08,
+      "loss": 0.4062,
+      "step": 5841
+    },
+    {
+      "epoch": 14.28361858190709,
+      "grad_norm": 0.256179869294639,
+      "learning_rate": 2.8086717822326526e-08,
+      "loss": 0.6226,
+      "step": 5842
+    },
+    {
+      "epoch": 14.28606356968215,
+      "grad_norm": 0.21704994850977963,
+      "learning_rate": 2.7895683507984107e-08,
+      "loss": 0.4964,
+      "step": 5843
+    },
+    {
+      "epoch": 14.288508557457213,
+      "grad_norm": 0.16805759807252935,
+      "learning_rate": 2.770529743598316e-08,
+      "loss": 0.321,
+      "step": 5844
+    },
+    {
+      "epoch": 14.290953545232274,
+      "grad_norm": 0.25032701111503786,
+      "learning_rate": 2.751555965624736e-08,
+      "loss": 0.5232,
+      "step": 5845
+    },
+    {
+      "epoch": 14.293398533007334,
+      "grad_norm": 0.3030503931391864,
+      "learning_rate": 2.7326470218530242e-08,
+      "loss": 0.3471,
+      "step": 5846
+    },
+    {
+      "epoch": 14.295843520782396,
+      "grad_norm": 0.2639972823139374,
+      "learning_rate": 2.7138029172415757e-08,
+      "loss": 0.3927,
+      "step": 5847
+    },
+    {
+      "epoch": 14.298288508557457,
+      "grad_norm": 0.31790158040174044,
+      "learning_rate": 2.6950236567317157e-08,
+      "loss": 0.3144,
+      "step": 5848
+    },
+    {
+      "epoch": 14.30073349633252,
+      "grad_norm": 0.28752766629175236,
+      "learning_rate": 2.6763092452477823e-08,
+      "loss": 0.5689,
+      "step": 5849
+    },
+    {
+      "epoch": 14.30317848410758,
+      "grad_norm": 0.20613251497739601,
+      "learning_rate": 2.6576596876971562e-08,
+      "loss": 0.9234,
+      "step": 5850
+    },
+    {
+      "epoch": 14.30562347188264,
+      "grad_norm": 0.2960427100855193,
+      "learning_rate": 2.639074988970175e-08,
+      "loss": 0.3269,
+      "step": 5851
+    },
+    {
+      "epoch": 14.3080684596577,
+      "grad_norm": 0.25671179313754044,
+      "learning_rate": 2.6205551539401353e-08,
+      "loss": 0.4295,
+      "step": 5852
+    },
+    {
+      "epoch": 14.310513447432763,
+      "grad_norm": 0.2496847246034851,
+      "learning_rate": 2.602100187463402e-08,
+      "loss": 0.6114,
+      "step": 5853
+    },
+    {
+      "epoch": 14.312958435207824,
+      "grad_norm": 0.41377589381240654,
+      "learning_rate": 2.5837100943792714e-08,
+      "loss": 0.9258,
+      "step": 5854
+    },
+    {
+      "epoch": 14.315403422982886,
+      "grad_norm": 0.24918178144569356,
+      "learning_rate": 2.56538487951008e-08,
+      "loss": 0.4429,
+      "step": 5855
+    },
+    {
+      "epoch": 14.317848410757946,
+      "grad_norm": 0.3600099652689131,
+      "learning_rate": 2.547124547661067e-08,
+      "loss": 0.4993,
+      "step": 5856
+    },
+    {
+      "epoch": 14.320293398533007,
+      "grad_norm": 0.317314336296521,
+      "learning_rate": 2.5289291036205142e-08,
+      "loss": 0.329,
+      "step": 5857
+    },
+    {
+      "epoch": 14.32273838630807,
+      "grad_norm": 0.24780148755521586,
+      "learning_rate": 2.5107985521597434e-08,
+      "loss": 0.8495,
+      "step": 5858
+    },
+    {
+      "epoch": 14.32518337408313,
+      "grad_norm": 0.19761235090892323,
+      "learning_rate": 2.4927328980329236e-08,
+      "loss": 0.6029,
+      "step": 5859
+    },
+    {
+      "epoch": 14.32762836185819,
+      "grad_norm": 0.24276681499798555,
+      "learning_rate": 2.474732145977321e-08,
+      "loss": 0.6108,
+      "step": 5860
+    },
+    {
+      "epoch": 14.330073349633253,
+      "grad_norm": 0.4499166168837541,
+      "learning_rate": 2.45679630071316e-08,
+      "loss": 0.7908,
+      "step": 5861
+    },
+    {
+      "epoch": 14.332518337408313,
+      "grad_norm": 0.23121864585343233,
+      "learning_rate": 2.4389253669435674e-08,
+      "loss": 0.4192,
+      "step": 5862
+    },
+    {
+      "epoch": 14.334963325183374,
+      "grad_norm": 0.21463803044609225,
+      "learning_rate": 2.4211193493547946e-08,
+      "loss": 0.8216,
+      "step": 5863
+    },
+    {
+      "epoch": 14.337408312958436,
+      "grad_norm": 0.32413040044665375,
+      "learning_rate": 2.40337825261594e-08,
+      "loss": 0.4791,
+      "step": 5864
+    },
+    {
+      "epoch": 14.339853300733497,
+      "grad_norm": 0.23079069001757216,
+      "learning_rate": 2.3857020813791433e-08,
+      "loss": 0.4953,
+      "step": 5865
+    },
+    {
+      "epoch": 14.342298288508557,
+      "grad_norm": 0.2447484706211273,
+      "learning_rate": 2.3680908402794745e-08,
+      "loss": 0.8017,
+      "step": 5866
+    },
+    {
+      "epoch": 14.34474327628362,
+      "grad_norm": 0.27288029653247675,
+      "learning_rate": 2.3505445339350175e-08,
+      "loss": 0.6673,
+      "step": 5867
+    },
+    {
+      "epoch": 14.34718826405868,
+      "grad_norm": 0.24957128080161445,
+      "learning_rate": 2.333063166946842e-08,
+      "loss": 0.6841,
+      "step": 5868
+    },
+    {
+      "epoch": 14.34963325183374,
+      "grad_norm": 0.2174832054818101,
+      "learning_rate": 2.3156467438989194e-08,
+      "loss": 0.6158,
+      "step": 5869
+    },
+    {
+      "epoch": 14.352078239608803,
+      "grad_norm": 0.32288480461399,
+      "learning_rate": 2.2982952693582638e-08,
+      "loss": 0.4772,
+      "step": 5870
+    },
+    {
+      "epoch": 14.354523227383863,
+      "grad_norm": 0.23771156325053922,
+      "learning_rate": 2.281008747874819e-08,
+      "loss": 0.4998,
+      "step": 5871
+    },
+    {
+      "epoch": 14.356968215158924,
+      "grad_norm": 0.24868999418083254,
+      "learning_rate": 2.2637871839814863e-08,
+      "loss": 1.0037,
+      "step": 5872
+    },
+    {
+      "epoch": 14.359413202933986,
+      "grad_norm": 0.2281428282696578,
+      "learning_rate": 2.2466305821941815e-08,
+      "loss": 0.4357,
+      "step": 5873
+    },
+    {
+      "epoch": 14.361858190709047,
+      "grad_norm": 0.3040265399616088,
+      "learning_rate": 2.22953894701175e-08,
+      "loss": 0.3662,
+      "step": 5874
+    },
+    {
+      "epoch": 14.364303178484107,
+      "grad_norm": 0.22099601567100147,
+      "learning_rate": 2.2125122829160516e-08,
+      "loss": 0.4357,
+      "step": 5875
+    },
+    {
+      "epoch": 14.36674816625917,
+      "grad_norm": 0.3326749355981543,
+      "learning_rate": 2.1955505943717926e-08,
+      "loss": 0.5771,
+      "step": 5876
+    },
+    {
+      "epoch": 14.36919315403423,
+      "grad_norm": 0.28093276359216685,
+      "learning_rate": 2.178653885826748e-08,
+      "loss": 0.5564,
+      "step": 5877
+    },
+    {
+      "epoch": 14.37163814180929,
+      "grad_norm": 0.19460678368847734,
+      "learning_rate": 2.1618221617116243e-08,
+      "loss": 0.4142,
+      "step": 5878
+    },
+    {
+      "epoch": 14.374083129584353,
+      "grad_norm": 0.1666924558980049,
+      "learning_rate": 2.1450554264400846e-08,
+      "loss": 0.5622,
+      "step": 5879
+    },
+    {
+      "epoch": 14.376528117359413,
+      "grad_norm": 0.4720935220821703,
+      "learning_rate": 2.1283536844087514e-08,
+      "loss": 0.4856,
+      "step": 5880
+    },
+    {
+      "epoch": 14.378973105134474,
+      "grad_norm": 0.18700294343653207,
+      "learning_rate": 2.1117169399972327e-08,
+      "loss": 0.6129,
+      "step": 5881
+    },
+    {
+      "epoch": 14.381418092909536,
+      "grad_norm": 0.36678118975137125,
+      "learning_rate": 2.0951451975679827e-08,
+      "loss": 0.4568,
+      "step": 5882
+    },
+    {
+      "epoch": 14.383863080684597,
+      "grad_norm": 0.28995097553404914,
+      "learning_rate": 2.078638461466581e-08,
+      "loss": 0.756,
+      "step": 5883
+    },
+    {
+      "epoch": 14.386308068459657,
+      "grad_norm": 0.19667664652681321,
+      "learning_rate": 2.062196736021399e-08,
+      "loss": 0.421,
+      "step": 5884
+    },
+    {
+      "epoch": 14.38875305623472,
+      "grad_norm": 0.2908773173423205,
+      "learning_rate": 2.0458200255439042e-08,
+      "loss": 0.6971,
+      "step": 5885
+    },
+    {
+      "epoch": 14.39119804400978,
+      "grad_norm": 0.22692559188191663,
+      "learning_rate": 2.029508334328384e-08,
+      "loss": 0.388,
+      "step": 5886
+    },
+    {
+      "epoch": 14.39364303178484,
+      "grad_norm": 0.21265110816814112,
+      "learning_rate": 2.0132616666521666e-08,
+      "loss": 0.4936,
+      "step": 5887
+    },
+    {
+      "epoch": 14.396088019559903,
+      "grad_norm": 0.21045318807801613,
+      "learning_rate": 1.9970800267755108e-08,
+      "loss": 0.6917,
+      "step": 5888
+    },
+    {
+      "epoch": 14.398533007334963,
+      "grad_norm": 0.21180267442415515,
+      "learning_rate": 1.980963418941606e-08,
+      "loss": 0.3705,
+      "step": 5889
+    },
+    {
+      "epoch": 14.400977995110024,
+      "grad_norm": 0.3685732318156828,
+      "learning_rate": 1.964911847376544e-08,
+      "loss": 0.859,
+      "step": 5890
+    },
+    {
+      "epoch": 14.403422982885086,
+      "grad_norm": 0.33655485843873795,
+      "learning_rate": 1.9489253162895127e-08,
+      "loss": 0.3365,
+      "step": 5891
+    },
+    {
+      "epoch": 14.405867970660147,
+      "grad_norm": 0.26800385093450046,
+      "learning_rate": 1.9330038298724652e-08,
+      "loss": 0.9645,
+      "step": 5892
+    },
+    {
+      "epoch": 14.408312958435207,
+      "grad_norm": 0.23234158208899475,
+      "learning_rate": 1.9171473923003946e-08,
+      "loss": 0.6413,
+      "step": 5893
+    },
+    {
+      "epoch": 14.41075794621027,
+      "grad_norm": 0.28703713000881426,
+      "learning_rate": 1.9013560077312253e-08,
+      "loss": 0.3595,
+      "step": 5894
+    },
+    {
+      "epoch": 14.41320293398533,
+      "grad_norm": 0.3136894419786609,
+      "learning_rate": 1.885629680305867e-08,
+      "loss": 0.4776,
+      "step": 5895
+    },
+    {
+      "epoch": 14.41564792176039,
+      "grad_norm": 0.19211267149853545,
+      "learning_rate": 1.869968414148049e-08,
+      "loss": 0.4327,
+      "step": 5896
+    },
+    {
+      "epoch": 14.418092909535453,
+      "grad_norm": 0.2236724758160594,
+      "learning_rate": 1.8543722133645693e-08,
+      "loss": 0.4481,
+      "step": 5897
+    },
+    {
+      "epoch": 14.420537897310513,
+      "grad_norm": 0.23997687927456546,
+      "learning_rate": 1.838841082045073e-08,
+      "loss": 0.7161,
+      "step": 5898
+    },
+    {
+      "epoch": 14.422982885085574,
+      "grad_norm": 0.2470977475350639,
+      "learning_rate": 1.8233750242621917e-08,
+      "loss": 0.5961,
+      "step": 5899
+    },
+    {
+      "epoch": 14.425427872860636,
+      "grad_norm": 0.3582095420789469,
+      "learning_rate": 1.8079740440714866e-08,
+      "loss": 0.7068,
+      "step": 5900
+    },
+    {
+      "epoch": 14.427872860635697,
+      "grad_norm": 0.2367564665371195,
+      "learning_rate": 1.7926381455114495e-08,
+      "loss": 0.8264,
+      "step": 5901
+    },
+    {
+      "epoch": 14.430317848410757,
+      "grad_norm": 0.20600421315518097,
+      "learning_rate": 1.7773673326035024e-08,
+      "loss": 0.644,
+      "step": 5902
+    },
+    {
+      "epoch": 14.43276283618582,
+      "grad_norm": 0.3341196720073081,
+      "learning_rate": 1.762161609351998e-08,
+      "loss": 0.3405,
+      "step": 5903
+    },
+    {
+      "epoch": 14.43520782396088,
+      "grad_norm": 0.1995247221169888,
+      "learning_rate": 1.7470209797441905e-08,
+      "loss": 0.4356,
+      "step": 5904
+    },
+    {
+      "epoch": 14.43765281173594,
+      "grad_norm": 0.24691762455068278,
+      "learning_rate": 1.7319454477503495e-08,
+      "loss": 0.8613,
+      "step": 5905
+    },
+    {
+      "epoch": 14.440097799511003,
+      "grad_norm": 0.23092519815290935,
+      "learning_rate": 1.7169350173236176e-08,
+      "loss": 0.7153,
+      "step": 5906
+    },
+    {
+      "epoch": 14.442542787286063,
+      "grad_norm": 0.264879717034351,
+      "learning_rate": 1.7019896924000412e-08,
+      "loss": 0.3733,
+      "step": 5907
+    },
+    {
+      "epoch": 14.444987775061124,
+      "grad_norm": 0.23277367272197552,
+      "learning_rate": 1.6871094768986795e-08,
+      "loss": 0.5023,
+      "step": 5908
+    },
+    {
+      "epoch": 14.447432762836186,
+      "grad_norm": 0.3917808881196965,
+      "learning_rate": 1.6722943747214117e-08,
+      "loss": 0.4251,
+      "step": 5909
+    },
+    {
+      "epoch": 14.449877750611247,
+      "grad_norm": 0.28663015315239915,
+      "learning_rate": 1.6575443897531295e-08,
+      "loss": 0.4217,
+      "step": 5910
+    },
+    {
+      "epoch": 14.452322738386307,
+      "grad_norm": 0.21802247534454908,
+      "learning_rate": 1.6428595258616008e-08,
+      "loss": 0.4989,
+      "step": 5911
+    },
+    {
+      "epoch": 14.45476772616137,
+      "grad_norm": 0.30452590965971515,
+      "learning_rate": 1.628239786897551e-08,
+      "loss": 0.6905,
+      "step": 5912
+    },
+    {
+      "epoch": 14.45721271393643,
+      "grad_norm": 0.23974505361225917,
+      "learning_rate": 1.6136851766945793e-08,
+      "loss": 0.5973,
+      "step": 5913
+    },
+    {
+      "epoch": 14.45965770171149,
+      "grad_norm": 0.3599449784250477,
+      "learning_rate": 1.5991956990692724e-08,
+      "loss": 0.4503,
+      "step": 5914
+    },
+    {
+      "epoch": 14.462102689486553,
+      "grad_norm": 0.40809627687685907,
+      "learning_rate": 1.5847713578210632e-08,
+      "loss": 0.4662,
+      "step": 5915
+    },
+    {
+      "epoch": 14.464547677261614,
+      "grad_norm": 0.17465008783350863,
+      "learning_rate": 1.5704121567323704e-08,
+      "loss": 0.4139,
+      "step": 5916
+    },
+    {
+      "epoch": 14.466992665036674,
+      "grad_norm": 0.42271568944935967,
+      "learning_rate": 1.5561180995685156e-08,
+      "loss": 0.3809,
+      "step": 5917
+    },
+    {
+      "epoch": 14.469437652811736,
+      "grad_norm": 0.24442652435875195,
+      "learning_rate": 1.5418891900777223e-08,
+      "loss": 0.325,
+      "step": 5918
+    },
+    {
+      "epoch": 14.471882640586797,
+      "grad_norm": 0.14954539407623077,
+      "learning_rate": 1.527725431991145e-08,
+      "loss": 0.479,
+      "step": 5919
+    },
+    {
+      "epoch": 14.474327628361857,
+      "grad_norm": 0.27725361238971163,
+      "learning_rate": 1.5136268290227852e-08,
+      "loss": 0.5152,
+      "step": 5920
+    },
+    {
+      "epoch": 14.47677261613692,
+      "grad_norm": 0.28664971567029324,
+      "learning_rate": 1.499593384869713e-08,
+      "loss": 0.3667,
+      "step": 5921
+    },
+    {
+      "epoch": 14.47921760391198,
+      "grad_norm": 0.20328984180663284,
+      "learning_rate": 1.4856251032117353e-08,
+      "loss": 0.8143,
+      "step": 5922
+    },
+    {
+      "epoch": 14.48166259168704,
+      "grad_norm": 0.3158681674724048,
+      "learning_rate": 1.4717219877116996e-08,
+      "loss": 0.3009,
+      "step": 5923
+    },
+    {
+      "epoch": 14.484107579462103,
+      "grad_norm": 0.5917968853268613,
+      "learning_rate": 1.4578840420153006e-08,
+      "loss": 0.831,
+      "step": 5924
+    },
+    {
+      "epoch": 14.486552567237164,
+      "grad_norm": 0.2422789309488887,
+      "learning_rate": 1.4441112697511638e-08,
+      "loss": 0.5739,
+      "step": 5925
+    },
+    {
+      "epoch": 14.488997555012224,
+      "grad_norm": 0.1697282485955006,
+      "learning_rate": 1.4304036745308724e-08,
+      "loss": 0.4268,
+      "step": 5926
+    },
+    {
+      "epoch": 14.491442542787286,
+      "grad_norm": 0.22723014484916346,
+      "learning_rate": 1.4167612599488012e-08,
+      "loss": 0.6064,
+      "step": 5927
+    },
+    {
+      "epoch": 14.493887530562347,
+      "grad_norm": 0.3569433971771532,
+      "learning_rate": 1.4031840295823663e-08,
+      "loss": 0.566,
+      "step": 5928
+    },
+    {
+      "epoch": 14.496332518337407,
+      "grad_norm": 0.20195702490881626,
+      "learning_rate": 1.3896719869917753e-08,
+      "loss": 0.4743,
+      "step": 5929
+    },
+    {
+      "epoch": 14.49877750611247,
+      "grad_norm": 0.2197320089290472,
+      "learning_rate": 1.3762251357202216e-08,
+      "loss": 0.5578,
+      "step": 5930
+    },
+    {
+      "epoch": 14.50122249388753,
+      "grad_norm": 0.20256403039679488,
+      "learning_rate": 1.3628434792938016e-08,
+      "loss": 0.3571,
+      "step": 5931
+    },
+    {
+      "epoch": 14.503667481662593,
+      "grad_norm": 0.2059181372928567,
+      "learning_rate": 1.3495270212214584e-08,
+      "loss": 0.9066,
+      "step": 5932
+    },
+    {
+      "epoch": 14.506112469437653,
+      "grad_norm": 0.2903535547845707,
+      "learning_rate": 1.3362757649950652e-08,
+      "loss": 0.8555,
+      "step": 5933
+    },
+    {
+      "epoch": 14.508557457212714,
+      "grad_norm": 0.22822853438766613,
+      "learning_rate": 1.323089714089426e-08,
+      "loss": 0.4462,
+      "step": 5934
+    },
+    {
+      "epoch": 14.511002444987774,
+      "grad_norm": 0.18205427653832412,
+      "learning_rate": 1.3099688719622195e-08,
+      "loss": 0.4886,
+      "step": 5935
+    },
+    {
+      "epoch": 14.513447432762836,
+      "grad_norm": 0.28864277506262753,
+      "learning_rate": 1.2969132420540265e-08,
+      "loss": 0.4744,
+      "step": 5936
+    },
+    {
+      "epoch": 14.515892420537897,
+      "grad_norm": 0.3287538193052468,
+      "learning_rate": 1.2839228277883032e-08,
+      "loss": 0.3179,
+      "step": 5937
+    },
+    {
+      "epoch": 14.51833740831296,
+      "grad_norm": 0.2629643344453778,
+      "learning_rate": 1.2709976325715189e-08,
+      "loss": 1.4419,
+      "step": 5938
+    },
+    {
+      "epoch": 14.52078239608802,
+      "grad_norm": 0.24686352942596163,
+      "learning_rate": 1.2581376597928796e-08,
+      "loss": 0.5284,
+      "step": 5939
+    },
+    {
+      "epoch": 14.52322738386308,
+      "grad_norm": 0.23014420988141793,
+      "learning_rate": 1.2453429128245765e-08,
+      "loss": 0.6586,
+      "step": 5940
+    },
+    {
+      "epoch": 14.525672371638143,
+      "grad_norm": 0.3216063361305703,
+      "learning_rate": 1.2326133950217034e-08,
+      "loss": 0.655,
+      "step": 5941
+    },
+    {
+      "epoch": 14.528117359413203,
+      "grad_norm": 0.18130755911801785,
+      "learning_rate": 1.2199491097222293e-08,
+      "loss": 0.6401,
+      "step": 5942
+    },
+    {
+      "epoch": 14.530562347188264,
+      "grad_norm": 0.3403314851621884,
+      "learning_rate": 1.2073500602470257e-08,
+      "loss": 0.458,
+      "step": 5943
+    },
+    {
+      "epoch": 14.533007334963326,
+      "grad_norm": 0.23245002733190495,
+      "learning_rate": 1.1948162498998107e-08,
+      "loss": 0.4228,
+      "step": 5944
+    },
+    {
+      "epoch": 14.535452322738386,
+      "grad_norm": 0.24536593026426867,
+      "learning_rate": 1.1823476819672885e-08,
+      "loss": 0.8562,
+      "step": 5945
+    },
+    {
+      "epoch": 14.537897310513447,
+      "grad_norm": 0.3546474440078184,
+      "learning_rate": 1.169944359718983e-08,
+      "loss": 0.3927,
+      "step": 5946
+    },
+    {
+      "epoch": 14.54034229828851,
+      "grad_norm": 0.3157771654121839,
+      "learning_rate": 1.1576062864073201e-08,
+      "loss": 0.4022,
+      "step": 5947
+    },
+    {
+      "epoch": 14.54278728606357,
+      "grad_norm": 0.43788749389038056,
+      "learning_rate": 1.1453334652676285e-08,
+      "loss": 0.4188,
+      "step": 5948
+    },
+    {
+      "epoch": 14.54523227383863,
+      "grad_norm": 0.24826591774214785,
+      "learning_rate": 1.1331258995181116e-08,
+      "loss": 0.741,
+      "step": 5949
+    },
+    {
+      "epoch": 14.547677261613693,
+      "grad_norm": 0.2724595057026709,
+      "learning_rate": 1.1209835923599033e-08,
+      "loss": 0.6084,
+      "step": 5950
+    },
+    {
+      "epoch": 14.550122249388753,
+      "grad_norm": 0.3328890855915936,
+      "learning_rate": 1.1089065469769844e-08,
+      "loss": 0.4137,
+      "step": 5951
+    },
+    {
+      "epoch": 14.552567237163814,
+      "grad_norm": 0.2815242918215007,
+      "learning_rate": 1.0968947665362107e-08,
+      "loss": 0.7544,
+      "step": 5952
+    },
+    {
+      "epoch": 14.555012224938876,
+      "grad_norm": 0.3134859748126106,
+      "learning_rate": 1.084948254187368e-08,
+      "loss": 0.8459,
+      "step": 5953
+    },
+    {
+      "epoch": 14.557457212713937,
+      "grad_norm": 0.2871337806962544,
+      "learning_rate": 1.0730670130630893e-08,
+      "loss": 0.6503,
+      "step": 5954
+    },
+    {
+      "epoch": 14.559902200488997,
+      "grad_norm": 0.4192824698489224,
+      "learning_rate": 1.0612510462789383e-08,
+      "loss": 0.3846,
+      "step": 5955
+    },
+    {
+      "epoch": 14.56234718826406,
+      "grad_norm": 0.44153812040528045,
+      "learning_rate": 1.0495003569332974e-08,
+      "loss": 0.5928,
+      "step": 5956
+    },
+    {
+      "epoch": 14.56479217603912,
+      "grad_norm": 0.33955729192970996,
+      "learning_rate": 1.0378149481074796e-08,
+      "loss": 0.6242,
+      "step": 5957
+    },
+    {
+      "epoch": 14.56723716381418,
+      "grad_norm": 0.22637009065739055,
+      "learning_rate": 1.0261948228656726e-08,
+      "loss": 0.8407,
+      "step": 5958
+    },
+    {
+      "epoch": 14.569682151589243,
+      "grad_norm": 0.314303991489773,
+      "learning_rate": 1.0146399842549116e-08,
+      "loss": 0.5836,
+      "step": 5959
+    },
+    {
+      "epoch": 14.572127139364303,
+      "grad_norm": 0.2849968244545076,
+      "learning_rate": 1.0031504353051891e-08,
+      "loss": 0.5092,
+      "step": 5960
+    },
+    {
+      "epoch": 14.574572127139364,
+      "grad_norm": 0.31104698716324186,
+      "learning_rate": 9.917261790292898e-09,
+      "loss": 1.1197,
+      "step": 5961
+    },
+    {
+      "epoch": 14.577017114914426,
+      "grad_norm": 0.24736896285181478,
+      "learning_rate": 9.803672184229285e-09,
+      "loss": 0.7574,
+      "step": 5962
+    },
+    {
+      "epoch": 14.579462102689487,
+      "grad_norm": 0.2938954362423323,
+      "learning_rate": 9.690735564646946e-09,
+      "loss": 0.5853,
+      "step": 5963
+    },
+    {
+      "epoch": 14.581907090464547,
+      "grad_norm": 0.2550767577779828,
+      "learning_rate": 9.57845196116025e-09,
+      "loss": 0.4612,
+      "step": 5964
+    },
+    {
+      "epoch": 14.58435207823961,
+      "grad_norm": 0.1789110933635988,
+      "learning_rate": 9.46682140321259e-09,
+      "loss": 0.7292,
+      "step": 5965
+    },
+    {
+      "epoch": 14.58679706601467,
+      "grad_norm": 0.27553399752375046,
+      "learning_rate": 9.355843920076101e-09,
+      "loss": 0.6208,
+      "step": 5966
+    },
+    {
+      "epoch": 14.58924205378973,
+      "grad_norm": 0.24773899640474598,
+      "learning_rate": 9.2455195408514e-09,
+      "loss": 0.8385,
+      "step": 5967
+    },
+    {
+      "epoch": 14.591687041564793,
+      "grad_norm": 0.3733501862142067,
+      "learning_rate": 9.135848294468396e-09,
+      "loss": 0.8394,
+      "step": 5968
+    },
+    {
+      "epoch": 14.594132029339853,
+      "grad_norm": 0.3578934411237306,
+      "learning_rate": 9.0268302096852e-09,
+      "loss": 0.7831,
+      "step": 5969
+    },
+    {
+      "epoch": 14.596577017114914,
+      "grad_norm": 0.27935597065238066,
+      "learning_rate": 8.918465315088943e-09,
+      "loss": 0.4194,
+      "step": 5970
+    },
+    {
+      "epoch": 14.599022004889976,
+      "grad_norm": 0.34961499210108704,
+      "learning_rate": 8.810753639095227e-09,
+      "loss": 0.8387,
+      "step": 5971
+    },
+    {
+      "epoch": 14.601466992665037,
+      "grad_norm": 0.26205186074461606,
+      "learning_rate": 8.70369520994896e-09,
+      "loss": 0.3376,
+      "step": 5972
+    },
+    {
+      "epoch": 14.603911980440097,
+      "grad_norm": 0.3454017371791375,
+      "learning_rate": 8.597290055722685e-09,
+      "loss": 0.4305,
+      "step": 5973
+    },
+    {
+      "epoch": 14.60635696821516,
+      "grad_norm": 0.2654511539156003,
+      "learning_rate": 8.491538204318805e-09,
+      "loss": 0.605,
+      "step": 5974
+    },
+    {
+      "epoch": 14.60880195599022,
+      "grad_norm": 0.22827230052113323,
+      "learning_rate": 8.386439683467362e-09,
+      "loss": 0.5083,
+      "step": 5975
+    },
+    {
+      "epoch": 14.61124694376528,
+      "grad_norm": 0.21119746496374842,
+      "learning_rate": 8.281994520727976e-09,
+      "loss": 0.6664,
+      "step": 5976
+    },
+    {
+      "epoch": 14.613691931540343,
+      "grad_norm": 0.15489263894510524,
+      "learning_rate": 8.178202743488462e-09,
+      "loss": 0.4218,
+      "step": 5977
+    },
+    {
+      "epoch": 14.616136919315403,
+      "grad_norm": 0.19392786515917243,
+      "learning_rate": 8.075064378965381e-09,
+      "loss": 0.5175,
+      "step": 5978
+    },
+    {
+      "epoch": 14.618581907090464,
+      "grad_norm": 0.18273882377754652,
+      "learning_rate": 7.972579454204043e-09,
+      "loss": 0.4118,
+      "step": 5979
+    },
+    {
+      "epoch": 14.621026894865526,
+      "grad_norm": 0.22618739440532434,
+      "learning_rate": 7.870747996078231e-09,
+      "loss": 0.7905,
+      "step": 5980
+    },
+    {
+      "epoch": 14.623471882640587,
+      "grad_norm": 0.23847204498152716,
+      "learning_rate": 7.769570031290752e-09,
+      "loss": 0.4172,
+      "step": 5981
+    },
+    {
+      "epoch": 14.625916870415647,
+      "grad_norm": 0.3629398630770936,
+      "learning_rate": 7.669045586372325e-09,
+      "loss": 0.3507,
+      "step": 5982
+    },
+    {
+      "epoch": 14.62836185819071,
+      "grad_norm": 0.21479144620517518,
+      "learning_rate": 7.569174687683257e-09,
+      "loss": 0.3935,
+      "step": 5983
+    },
+    {
+      "epoch": 14.63080684596577,
+      "grad_norm": 0.23743563284813463,
+      "learning_rate": 7.469957361411483e-09,
+      "loss": 0.4452,
+      "step": 5984
+    },
+    {
+      "epoch": 14.63325183374083,
+      "grad_norm": 0.20815162666555456,
+      "learning_rate": 7.371393633574253e-09,
+      "loss": 0.4598,
+      "step": 5985
+    },
+    {
+      "epoch": 14.635696821515893,
+      "grad_norm": 0.2942837952267298,
+      "learning_rate": 7.273483530017556e-09,
+      "loss": 0.7701,
+      "step": 5986
+    },
+    {
+      "epoch": 14.638141809290953,
+      "grad_norm": 0.24847949017980506,
+      "learning_rate": 7.176227076415299e-09,
+      "loss": 0.9498,
+      "step": 5987
+    },
+    {
+      "epoch": 14.640586797066014,
+      "grad_norm": 0.2886962492048405,
+      "learning_rate": 7.0796242982704155e-09,
+      "loss": 1.057,
+      "step": 5988
+    },
+    {
+      "epoch": 14.643031784841076,
+      "grad_norm": 0.2407338622916156,
+      "learning_rate": 6.98367522091431e-09,
+      "loss": 0.7743,
+      "step": 5989
+    },
+    {
+      "epoch": 14.645476772616137,
+      "grad_norm": 0.35345018489614627,
+      "learning_rate": 6.888379869507134e-09,
+      "loss": 0.4915,
+      "step": 5990
+    },
+    {
+      "epoch": 14.647921760391197,
+      "grad_norm": 0.2588222064637483,
+      "learning_rate": 6.7937382690375085e-09,
+      "loss": 0.6053,
+      "step": 5991
+    },
+    {
+      "epoch": 14.65036674816626,
+      "grad_norm": 0.19407202420345154,
+      "learning_rate": 6.699750444322806e-09,
+      "loss": 0.7861,
+      "step": 5992
+    },
+    {
+      "epoch": 14.65281173594132,
+      "grad_norm": 0.38056007830827043,
+      "learning_rate": 6.6064164200083105e-09,
+      "loss": 0.4325,
+      "step": 5993
+    },
+    {
+      "epoch": 14.65525672371638,
+      "grad_norm": 0.324321971744957,
+      "learning_rate": 6.513736220568889e-09,
+      "loss": 0.9293,
+      "step": 5994
+    },
+    {
+      "epoch": 14.657701711491443,
+      "grad_norm": 0.3170154986244328,
+      "learning_rate": 6.421709870306769e-09,
+      "loss": 0.4786,
+      "step": 5995
+    },
+    {
+      "epoch": 14.660146699266503,
+      "grad_norm": 0.24505145820311475,
+      "learning_rate": 6.330337393354036e-09,
+      "loss": 0.5915,
+      "step": 5996
+    },
+    {
+      "epoch": 14.662591687041564,
+      "grad_norm": 0.2775314860635019,
+      "learning_rate": 6.239618813670412e-09,
+      "loss": 0.5127,
+      "step": 5997
+    },
+    {
+      "epoch": 14.665036674816626,
+      "grad_norm": 0.22930119755436948,
+      "learning_rate": 6.149554155044091e-09,
+      "loss": 0.8323,
+      "step": 5998
+    },
+    {
+      "epoch": 14.667481662591687,
+      "grad_norm": 0.2309457425243145,
+      "learning_rate": 6.060143441092292e-09,
+      "loss": 0.4913,
+      "step": 5999
+    },
+    {
+      "epoch": 14.669926650366747,
+      "grad_norm": 0.2421342613995642,
+      "learning_rate": 5.971386695260706e-09,
+      "loss": 0.9249,
+      "step": 6000
+    },
+    {
+      "epoch": 14.67237163814181,
+      "grad_norm": 0.2290265428212619,
+      "learning_rate": 5.8832839408232144e-09,
+      "loss": 0.3409,
+      "step": 6001
+    },
+    {
+      "epoch": 14.67481662591687,
+      "grad_norm": 0.19172885839247258,
+      "learning_rate": 5.795835200882449e-09,
+      "loss": 0.6352,
+      "step": 6002
+    },
+    {
+      "epoch": 14.67726161369193,
+      "grad_norm": 0.2390363370720583,
+      "learning_rate": 5.709040498369512e-09,
+      "loss": 0.8822,
+      "step": 6003
+    },
+    {
+      "epoch": 14.679706601466993,
+      "grad_norm": 0.21259312329093594,
+      "learning_rate": 5.6228998560436955e-09,
+      "loss": 0.6894,
+      "step": 6004
+    },
+    {
+      "epoch": 14.682151589242054,
+      "grad_norm": 0.3373322950250815,
+      "learning_rate": 5.537413296493321e-09,
+      "loss": 0.6568,
+      "step": 6005
+    },
+    {
+      "epoch": 14.684596577017114,
+      "grad_norm": 0.24686164502993627,
+      "learning_rate": 5.4525808421349025e-09,
+      "loss": 0.5377,
+      "step": 6006
+    },
+    {
+      "epoch": 14.687041564792176,
+      "grad_norm": 0.3449344032626107,
+      "learning_rate": 5.368402515213422e-09,
+      "loss": 0.4257,
+      "step": 6007
+    },
+    {
+      "epoch": 14.689486552567237,
+      "grad_norm": 0.2896496601183963,
+      "learning_rate": 5.284878337802057e-09,
+      "loss": 0.3719,
+      "step": 6008
+    },
+    {
+      "epoch": 14.6919315403423,
+      "grad_norm": 0.22205103984039853,
+      "learning_rate": 5.202008331803287e-09,
+      "loss": 0.3607,
+      "step": 6009
+    },
+    {
+      "epoch": 14.69437652811736,
+      "grad_norm": 0.21567865837402447,
+      "learning_rate": 5.119792518947231e-09,
+      "loss": 0.3643,
+      "step": 6010
+    },
+    {
+      "epoch": 14.69682151589242,
+      "grad_norm": 0.2768312071846796,
+      "learning_rate": 5.038230920792752e-09,
+      "loss": 0.7869,
+      "step": 6011
+    },
+    {
+      "epoch": 14.69926650366748,
+      "grad_norm": 0.24095109061649272,
+      "learning_rate": 4.95732355872719e-09,
+      "loss": 0.6822,
+      "step": 6012
+    },
+    {
+      "epoch": 14.701711491442543,
+      "grad_norm": 0.31400601391504207,
+      "learning_rate": 4.877070453966348e-09,
+      "loss": 0.6928,
+      "step": 6013
+    },
+    {
+      "epoch": 14.704156479217604,
+      "grad_norm": 0.2530379183141433,
+      "learning_rate": 4.797471627554507e-09,
+      "loss": 0.3742,
+      "step": 6014
+    },
+    {
+      "epoch": 14.706601466992666,
+      "grad_norm": 0.34061381163783283,
+      "learning_rate": 4.7185271003641346e-09,
+      "loss": 0.7075,
+      "step": 6015
+    },
+    {
+      "epoch": 14.709046454767726,
+      "grad_norm": 0.3105218567832549,
+      "learning_rate": 4.640236893096173e-09,
+      "loss": 0.4934,
+      "step": 6016
+    },
+    {
+      "epoch": 14.711491442542787,
+      "grad_norm": 0.15443151716892953,
+      "learning_rate": 4.562601026280589e-09,
+      "loss": 0.3425,
+      "step": 6017
+    },
+    {
+      "epoch": 14.713936430317847,
+      "grad_norm": 0.1744557264533319,
+      "learning_rate": 4.485619520274709e-09,
+      "loss": 0.4387,
+      "step": 6018
+    },
+    {
+      "epoch": 14.71638141809291,
+      "grad_norm": 0.24883032921484904,
+      "learning_rate": 4.409292395265441e-09,
+      "loss": 0.6523,
+      "step": 6019
+    },
+    {
+      "epoch": 14.71882640586797,
+      "grad_norm": 0.15763629075769425,
+      "learning_rate": 4.333619671267053e-09,
+      "loss": 0.4129,
+      "step": 6020
+    },
+    {
+      "epoch": 14.721271393643033,
+      "grad_norm": 0.2573036895064007,
+      "learning_rate": 4.258601368122839e-09,
+      "loss": 0.4781,
+      "step": 6021
+    },
+    {
+      "epoch": 14.723716381418093,
+      "grad_norm": 0.2235907691225696,
+      "learning_rate": 4.184237505504008e-09,
+      "loss": 0.5225,
+      "step": 6022
+    },
+    {
+      "epoch": 14.726161369193154,
+      "grad_norm": 0.2521904329130875,
+      "learning_rate": 4.110528102911071e-09,
+      "loss": 0.4445,
+      "step": 6023
+    },
+    {
+      "epoch": 14.728606356968216,
+      "grad_norm": 0.3550970617349348,
+      "learning_rate": 4.037473179671625e-09,
+      "loss": 0.5881,
+      "step": 6024
+    },
+    {
+      "epoch": 14.731051344743276,
+      "grad_norm": 0.262785714811522,
+      "learning_rate": 3.9650727549431224e-09,
+      "loss": 0.4914,
+      "step": 6025
+    },
+    {
+      "epoch": 14.733496332518337,
+      "grad_norm": 0.19715560985937872,
+      "learning_rate": 3.893326847710099e-09,
+      "loss": 0.5852,
+      "step": 6026
+    },
+    {
+      "epoch": 14.7359413202934,
+      "grad_norm": 0.19421441157057798,
+      "learning_rate": 3.82223547678584e-09,
+      "loss": 0.3984,
+      "step": 6027
+    },
+    {
+      "epoch": 14.73838630806846,
+      "grad_norm": 0.2331120945539375,
+      "learning_rate": 3.751798660812378e-09,
+      "loss": 0.4183,
+      "step": 6028
+    },
+    {
+      "epoch": 14.74083129584352,
+      "grad_norm": 0.2167657792224364,
+      "learning_rate": 3.682016418260215e-09,
+      "loss": 0.6365,
+      "step": 6029
+    },
+    {
+      "epoch": 14.743276283618583,
+      "grad_norm": 0.22012217801081127,
+      "learning_rate": 3.612888767427214e-09,
+      "loss": 0.5316,
+      "step": 6030
+    },
+    {
+      "epoch": 14.745721271393643,
+      "grad_norm": 0.33133075114151056,
+      "learning_rate": 3.544415726440542e-09,
+      "loss": 0.5028,
+      "step": 6031
+    },
+    {
+      "epoch": 14.748166259168704,
+      "grad_norm": 0.2052462827748625,
+      "learning_rate": 3.476597313255281e-09,
+      "loss": 0.5279,
+      "step": 6032
+    },
+    {
+      "epoch": 14.750611246943766,
+      "grad_norm": 0.29823741499673084,
+      "learning_rate": 3.4094335456549834e-09,
+      "loss": 0.5983,
+      "step": 6033
+    },
+    {
+      "epoch": 14.753056234718827,
+      "grad_norm": 0.23820070159138937,
+      "learning_rate": 3.3429244412516713e-09,
+      "loss": 0.3798,
+      "step": 6034
+    },
+    {
+      "epoch": 14.755501222493887,
+      "grad_norm": 0.29170791712317556,
+      "learning_rate": 3.2770700174852843e-09,
+      "loss": 0.6791,
+      "step": 6035
+    },
+    {
+      "epoch": 14.75794621026895,
+      "grad_norm": 0.2825423108756363,
+      "learning_rate": 3.2118702916247856e-09,
+      "loss": 0.7253,
+      "step": 6036
+    },
+    {
+      "epoch": 14.76039119804401,
+      "grad_norm": 0.2890878537413693,
+      "learning_rate": 3.1473252807667776e-09,
+      "loss": 0.4895,
+      "step": 6037
+    },
+    {
+      "epoch": 14.76283618581907,
+      "grad_norm": 0.29379187125581036,
+      "learning_rate": 3.083435001836332e-09,
+      "loss": 0.3434,
+      "step": 6038
+    },
+    {
+      "epoch": 14.765281173594133,
+      "grad_norm": 0.2579358018145425,
+      "learning_rate": 3.0201994715872706e-09,
+      "loss": 0.3923,
+      "step": 6039
+    },
+    {
+      "epoch": 14.767726161369193,
+      "grad_norm": 0.2391035060441323,
+      "learning_rate": 2.957618706601051e-09,
+      "loss": 0.4424,
+      "step": 6040
+    },
+    {
+      "epoch": 14.770171149144254,
+      "grad_norm": 0.19576696855516323,
+      "learning_rate": 2.8956927232878796e-09,
+      "loss": 0.4877,
+      "step": 6041
+    },
+    {
+      "epoch": 14.772616136919316,
+      "grad_norm": 0.240593856185608,
+      "learning_rate": 2.834421537886156e-09,
+      "loss": 0.6814,
+      "step": 6042
+    },
+    {
+      "epoch": 14.775061124694377,
+      "grad_norm": 0.2772221945451591,
+      "learning_rate": 2.773805166462751e-09,
+      "loss": 0.8086,
+      "step": 6043
+    },
+    {
+      "epoch": 14.777506112469437,
+      "grad_norm": 0.2100573816261956,
+      "learning_rate": 2.713843624912449e-09,
+      "loss": 0.507,
+      "step": 6044
+    },
+    {
+      "epoch": 14.7799511002445,
+      "grad_norm": 0.34050871174936265,
+      "learning_rate": 2.6545369289587843e-09,
+      "loss": 0.487,
+      "step": 6045
+    },
+    {
+      "epoch": 14.78239608801956,
+      "grad_norm": 0.27647510912626844,
+      "learning_rate": 2.5958850941532053e-09,
+      "loss": 1.0872,
+      "step": 6046
+    },
+    {
+      "epoch": 14.78484107579462,
+      "grad_norm": 0.4111356064594576,
+      "learning_rate": 2.537888135875355e-09,
+      "loss": 0.8439,
+      "step": 6047
+    },
+    {
+      "epoch": 14.787286063569683,
+      "grad_norm": 0.1906839892483114,
+      "learning_rate": 2.4805460693336226e-09,
+      "loss": 0.4373,
+      "step": 6048
+    },
+    {
+      "epoch": 14.789731051344743,
+      "grad_norm": 0.24808603343107538,
+      "learning_rate": 2.423858909564314e-09,
+      "loss": 0.406,
+      "step": 6049
+    },
+    {
+      "epoch": 14.792176039119804,
+      "grad_norm": 0.37771878750139837,
+      "learning_rate": 2.3678266714322053e-09,
+      "loss": 0.3198,
+      "step": 6050
+    },
+    {
+      "epoch": 14.794621026894866,
+      "grad_norm": 0.2160390329535855,
+      "learning_rate": 2.3124493696302653e-09,
+      "loss": 0.4314,
+      "step": 6051
+    },
+    {
+      "epoch": 14.797066014669927,
+      "grad_norm": 0.3051128816311609,
+      "learning_rate": 2.2577270186793786e-09,
+      "loss": 0.4283,
+      "step": 6052
+    },
+    {
+      "epoch": 14.799511002444987,
+      "grad_norm": 0.3763204378554294,
+      "learning_rate": 2.2036596329294556e-09,
+      "loss": 0.631,
+      "step": 6053
+    },
+    {
+      "epoch": 14.80195599022005,
+      "grad_norm": 0.35005814685659103,
+      "learning_rate": 2.150247226557767e-09,
+      "loss": 0.4814,
+      "step": 6054
+    },
+    {
+      "epoch": 14.80440097799511,
+      "grad_norm": 0.2929301930980776,
+      "learning_rate": 2.0974898135703305e-09,
+      "loss": 0.7775,
+      "step": 6055
+    },
+    {
+      "epoch": 14.80684596577017,
+      "grad_norm": 0.2049185745490653,
+      "learning_rate": 2.0453874078016355e-09,
+      "loss": 0.4771,
+      "step": 6056
+    },
+    {
+      "epoch": 14.809290953545233,
+      "grad_norm": 0.1649797453771252,
+      "learning_rate": 1.993940022914087e-09,
+      "loss": 0.6443,
+      "step": 6057
+    },
+    {
+      "epoch": 14.811735941320293,
+      "grad_norm": 0.16624824826202417,
+      "learning_rate": 1.943147672398005e-09,
+      "loss": 0.4902,
+      "step": 6058
+    },
+    {
+      "epoch": 14.814180929095354,
+      "grad_norm": 0.28322847142945035,
+      "learning_rate": 1.8930103695727344e-09,
+      "loss": 0.5741,
+      "step": 6059
+    },
+    {
+      "epoch": 14.816625916870416,
+      "grad_norm": 0.3099886287964264,
+      "learning_rate": 1.8435281275849814e-09,
+      "loss": 0.4538,
+      "step": 6060
+    },
+    {
+      "epoch": 14.819070904645477,
+      "grad_norm": 0.29517200474282757,
+      "learning_rate": 1.794700959410478e-09,
+      "loss": 0.4519,
+      "step": 6061
+    },
+    {
+      "epoch": 14.821515892420537,
+      "grad_norm": 0.24404313300535843,
+      "learning_rate": 1.746528877852871e-09,
+      "loss": 0.3422,
+      "step": 6062
+    },
+    {
+      "epoch": 14.8239608801956,
+      "grad_norm": 0.22328362712086952,
+      "learning_rate": 1.699011895543723e-09,
+      "loss": 0.7422,
+      "step": 6063
+    },
+    {
+      "epoch": 14.82640586797066,
+      "grad_norm": 0.19718418400246815,
+      "learning_rate": 1.6521500249430667e-09,
+      "loss": 0.5134,
+      "step": 6064
+    },
+    {
+      "epoch": 14.82885085574572,
+      "grad_norm": 0.1940561726468138,
+      "learning_rate": 1.6059432783394058e-09,
+      "loss": 0.5373,
+      "step": 6065
+    },
+    {
+      "epoch": 14.831295843520783,
+      "grad_norm": 0.26077652921015565,
+      "learning_rate": 1.5603916678488817e-09,
+      "loss": 0.8625,
+      "step": 6066
+    },
+    {
+      "epoch": 14.833740831295843,
+      "grad_norm": 0.31417874742777985,
+      "learning_rate": 1.5154952054166617e-09,
+      "loss": 0.5852,
+      "step": 6067
+    },
+    {
+      "epoch": 14.836185819070904,
+      "grad_norm": 0.3209894253250868,
+      "learning_rate": 1.4712539028149952e-09,
+      "loss": 0.5541,
+      "step": 6068
+    },
+    {
+      "epoch": 14.838630806845966,
+      "grad_norm": 0.21047144209334923,
+      "learning_rate": 1.4276677716457131e-09,
+      "loss": 0.3689,
+      "step": 6069
+    },
+    {
+      "epoch": 14.841075794621027,
+      "grad_norm": 0.21693428824831038,
+      "learning_rate": 1.3847368233374515e-09,
+      "loss": 0.9993,
+      "step": 6070
+    },
+    {
+      "epoch": 14.843520782396087,
+      "grad_norm": 0.4590276494781457,
+      "learning_rate": 1.3424610691478711e-09,
+      "loss": 0.4425,
+      "step": 6071
+    },
+    {
+      "epoch": 14.84596577017115,
+      "grad_norm": 0.2578953441546614,
+      "learning_rate": 1.3008405201625495e-09,
+      "loss": 0.4734,
+      "step": 6072
+    },
+    {
+      "epoch": 14.84841075794621,
+      "grad_norm": 0.33917636190556705,
+      "learning_rate": 1.2598751872958114e-09,
+      "loss": 0.6688,
+      "step": 6073
+    },
+    {
+      "epoch": 14.85085574572127,
+      "grad_norm": 0.27530160112989455,
+      "learning_rate": 1.2195650812890648e-09,
+      "loss": 0.5289,
+      "step": 6074
+    },
+    {
+      "epoch": 14.853300733496333,
+      "grad_norm": 0.2154966682765608,
+      "learning_rate": 1.1799102127130203e-09,
+      "loss": 0.417,
+      "step": 6075
+    },
+    {
+      "epoch": 14.855745721271393,
+      "grad_norm": 0.30745640063522156,
+      "learning_rate": 1.1409105919660268e-09,
+      "loss": 0.3396,
+      "step": 6076
+    },
+    {
+      "epoch": 14.858190709046454,
+      "grad_norm": 0.3223174564683795,
+      "learning_rate": 1.1025662292743488e-09,
+      "loss": 0.4648,
+      "step": 6077
+    },
+    {
+      "epoch": 14.860635696821516,
+      "grad_norm": 0.22441185036931252,
+      "learning_rate": 1.0648771346929986e-09,
+      "loss": 0.4567,
+      "step": 6078
+    },
+    {
+      "epoch": 14.863080684596577,
+      "grad_norm": 0.35753099593101356,
+      "learning_rate": 1.0278433181051817e-09,
+      "loss": 0.5927,
+      "step": 6079
+    },
+    {
+      "epoch": 14.865525672371637,
+      "grad_norm": 0.3359838507410903,
+      "learning_rate": 9.914647892214634e-10,
+      "loss": 0.5264,
+      "step": 6080
+    },
+    {
+      "epoch": 14.8679706601467,
+      "grad_norm": 0.20156700931920143,
+      "learning_rate": 9.557415575814355e-10,
+      "loss": 0.6415,
+      "step": 6081
+    },
+    {
+      "epoch": 14.87041564792176,
+      "grad_norm": 0.14308754578036692,
+      "learning_rate": 9.206736325526044e-10,
+      "loss": 0.2489,
+      "step": 6082
+    },
+    {
+      "epoch": 14.87286063569682,
+      "grad_norm": 0.278511609295102,
+      "learning_rate": 8.862610233303925e-10,
+      "loss": 0.9581,
+      "step": 6083
+    },
+    {
+      "epoch": 14.875305623471883,
+      "grad_norm": 0.24077294849432818,
+      "learning_rate": 8.525037389386925e-10,
+      "loss": 0.35,
+      "step": 6084
+    },
+    {
+      "epoch": 14.877750611246944,
+      "grad_norm": 0.2193195410966309,
+      "learning_rate": 8.194017882295902e-10,
+      "loss": 0.4227,
+      "step": 6085
+    },
+    {
+      "epoch": 14.880195599022004,
+      "grad_norm": 0.1941903427998383,
+      "learning_rate": 7.869551798828091e-10,
+      "loss": 0.6199,
+      "step": 6086
+    },
+    {
+      "epoch": 14.882640586797066,
+      "grad_norm": 0.3156006195839875,
+      "learning_rate": 7.551639224068207e-10,
+      "loss": 0.4067,
+      "step": 6087
+    },
+    {
+      "epoch": 14.885085574572127,
+      "grad_norm": 0.28527074687708986,
+      "learning_rate": 7.240280241380127e-10,
+      "loss": 0.7832,
+      "step": 6088
+    },
+    {
+      "epoch": 14.887530562347187,
+      "grad_norm": 0.3521589981447914,
+      "learning_rate": 6.935474932406872e-10,
+      "loss": 0.4085,
+      "step": 6089
+    },
+    {
+      "epoch": 14.88997555012225,
+      "grad_norm": 0.37967704995307855,
+      "learning_rate": 6.63722337707895e-10,
+      "loss": 0.3486,
+      "step": 6090
+    },
+    {
+      "epoch": 14.89242053789731,
+      "grad_norm": 0.24573600595054293,
+      "learning_rate": 6.345525653606021e-10,
+      "loss": 0.8806,
+      "step": 6091
+    },
+    {
+      "epoch": 14.894865525672373,
+      "grad_norm": 0.27177497279807844,
+      "learning_rate": 6.060381838471352e-10,
+      "loss": 0.6773,
+      "step": 6092
+    },
+    {
+      "epoch": 14.897310513447433,
+      "grad_norm": 0.3140673910233011,
+      "learning_rate": 5.781792006451237e-10,
+      "loss": 0.807,
+      "step": 6093
+    },
+    {
+      "epoch": 14.899755501222494,
+      "grad_norm": 0.20959808951064923,
+      "learning_rate": 5.509756230598351e-10,
+      "loss": 0.4013,
+      "step": 6094
+    },
+    {
+      "epoch": 14.902200488997554,
+      "grad_norm": 0.25672289390834235,
+      "learning_rate": 5.244274582244524e-10,
+      "loss": 0.4963,
+      "step": 6095
+    },
+    {
+      "epoch": 14.904645476772616,
+      "grad_norm": 0.17681462007478535,
+      "learning_rate": 4.985347131006291e-10,
+      "loss": 0.447,
+      "step": 6096
+    },
+    {
+      "epoch": 14.907090464547677,
+      "grad_norm": 0.2779137446642931,
+      "learning_rate": 4.73297394477934e-10,
+      "loss": 0.6167,
+      "step": 6097
+    },
+    {
+      "epoch": 14.90953545232274,
+      "grad_norm": 0.35764417329776804,
+      "learning_rate": 4.487155089741291e-10,
+      "loss": 0.4435,
+      "step": 6098
+    },
+    {
+      "epoch": 14.9119804400978,
+      "grad_norm": 0.39071684985120936,
+      "learning_rate": 4.247890630354468e-10,
+      "loss": 0.6944,
+      "step": 6099
+    },
+    {
+      "epoch": 14.91442542787286,
+      "grad_norm": 0.24672364232590013,
+      "learning_rate": 4.0151806293575733e-10,
+      "loss": 0.5014,
+      "step": 6100
+    },
+    {
+      "epoch": 14.91687041564792,
+      "grad_norm": 0.3305020093427622,
+      "learning_rate": 3.789025147774017e-10,
+      "loss": 0.4721,
+      "step": 6101
+    },
+    {
+      "epoch": 14.919315403422983,
+      "grad_norm": 0.21254176497695207,
+      "learning_rate": 3.569424244903585e-10,
+      "loss": 0.5032,
+      "step": 6102
+    },
+    {
+      "epoch": 14.921760391198044,
+      "grad_norm": 0.3303295121425075,
+      "learning_rate": 3.3563779783307715e-10,
+      "loss": 0.5623,
+      "step": 6103
+    },
+    {
+      "epoch": 14.924205378973106,
+      "grad_norm": 0.25570965246055927,
+      "learning_rate": 3.1498864039275487e-10,
+      "loss": 0.4422,
+      "step": 6104
+    },
+    {
+      "epoch": 14.926650366748166,
+      "grad_norm": 0.21943883857846808,
+      "learning_rate": 2.949949575833944e-10,
+      "loss": 0.7702,
+      "step": 6105
+    },
+    {
+      "epoch": 14.929095354523227,
+      "grad_norm": 0.35506875983878006,
+      "learning_rate": 2.756567546477462e-10,
+      "loss": 0.6513,
+      "step": 6106
+    },
+    {
+      "epoch": 14.93154034229829,
+      "grad_norm": 0.43100986641863454,
+      "learning_rate": 2.569740366573092e-10,
+      "loss": 0.6179,
+      "step": 6107
+    },
+    {
+      "epoch": 14.93398533007335,
+      "grad_norm": 0.33852455104193246,
+      "learning_rate": 2.3894680851094254e-10,
+      "loss": 0.4555,
+      "step": 6108
+    },
+    {
+      "epoch": 14.93643031784841,
+      "grad_norm": 0.2997287660461397,
+      "learning_rate": 2.2157507493542062e-10,
+      "loss": 0.4912,
+      "step": 6109
+    },
+    {
+      "epoch": 14.938875305623473,
+      "grad_norm": 0.36365554388015014,
+      "learning_rate": 2.0485884048626614e-10,
+      "loss": 0.4247,
+      "step": 6110
+    },
+    {
+      "epoch": 14.941320293398533,
+      "grad_norm": 0.21761490860903127,
+      "learning_rate": 1.8879810954691714e-10,
+      "loss": 0.726,
+      "step": 6111
+    },
+    {
+      "epoch": 14.943765281173594,
+      "grad_norm": 0.26260422366110636,
+      "learning_rate": 1.733928863287271e-10,
+      "loss": 0.6425,
+      "step": 6112
+    },
+    {
+      "epoch": 14.946210268948656,
+      "grad_norm": 0.33083469470162913,
+      "learning_rate": 1.5864317487152004e-10,
+      "loss": 0.464,
+      "step": 6113
+    },
+    {
+      "epoch": 14.948655256723717,
+      "grad_norm": 0.39038244878432726,
+      "learning_rate": 1.445489790427579e-10,
+      "loss": 0.4409,
+      "step": 6114
+    },
+    {
+      "epoch": 14.951100244498777,
+      "grad_norm": 0.29973846111304453,
+      "learning_rate": 1.311103025383731e-10,
+      "loss": 0.3724,
+      "step": 6115
+    },
+    {
+      "epoch": 14.95354523227384,
+      "grad_norm": 0.13253995018446735,
+      "learning_rate": 1.1832714888221353e-10,
+      "loss": 0.547,
+      "step": 6116
+    },
+    {
+      "epoch": 14.9559902200489,
+      "grad_norm": 0.34153815086086503,
+      "learning_rate": 1.0619952142659761e-10,
+      "loss": 0.4168,
+      "step": 6117
+    },
+    {
+      "epoch": 14.95843520782396,
+      "grad_norm": 0.20630446108208855,
+      "learning_rate": 9.472742335120411e-11,
+      "loss": 0.5538,
+      "step": 6118
+    },
+    {
+      "epoch": 14.960880195599023,
+      "grad_norm": 0.28672511549056245,
+      "learning_rate": 8.391085766473739e-11,
+      "loss": 0.3125,
+      "step": 6119
+    },
+    {
+      "epoch": 14.963325183374083,
+      "grad_norm": 0.6869203370630521,
+      "learning_rate": 7.374982720326218e-11,
+      "loss": 0.6003,
+      "step": 6120
+    },
+    {
+      "epoch": 14.965770171149144,
+      "grad_norm": 0.362939994202542,
+      "learning_rate": 6.424433463131374e-11,
+      "loss": 0.4988,
+      "step": 6121
+    },
+    {
+      "epoch": 14.968215158924206,
+      "grad_norm": 0.4673443001120997,
+      "learning_rate": 5.5394382441342764e-11,
+      "loss": 0.4976,
+      "step": 6122
+    },
+    {
+      "epoch": 14.970660146699267,
+      "grad_norm": 0.24007421082167638,
+      "learning_rate": 4.7199972954270476e-11,
+      "loss": 0.8173,
+      "step": 6123
+    },
+    {
+      "epoch": 14.973105134474327,
+      "grad_norm": 0.29326816384038545,
+      "learning_rate": 3.966110831865599e-11,
+      "loss": 0.429,
+      "step": 6124
+    },
+    {
+      "epoch": 14.97555012224939,
+      "grad_norm": 0.2177326378161841,
+      "learning_rate": 3.2777790511251404e-11,
+      "loss": 0.7403,
+      "step": 6125
+    },
+    {
+      "epoch": 14.97799511002445,
+      "grad_norm": 0.27485976870018675,
+      "learning_rate": 2.6550021337279354e-11,
+      "loss": 0.5434,
+      "step": 6126
+    },
+    {
+      "epoch": 14.98044009779951,
+      "grad_norm": 0.21829330167434846,
+      "learning_rate": 2.0977802429877902e-11,
+      "loss": 0.5768,
+      "step": 6127
+    },
+    {
+      "epoch": 14.982885085574573,
+      "grad_norm": 0.17184937449556248,
+      "learning_rate": 1.6061135249823e-11,
+      "loss": 0.5274,
+      "step": 6128
+    },
+    {
+      "epoch": 14.985330073349633,
+      "grad_norm": 0.2751475564433603,
+      "learning_rate": 1.180002108663869e-11,
+      "loss": 0.6488,
+      "step": 6129
+    },
+    {
+      "epoch": 14.987775061124694,
+      "grad_norm": 0.2576255837945489,
+      "learning_rate": 8.194461057764447e-12,
+      "loss": 0.846,
+      "step": 6130
+    },
+    {
+      "epoch": 14.990220048899756,
+      "grad_norm": 0.273057326128053,
+      "learning_rate": 5.244456108277618e-12,
+      "loss": 0.5486,
+      "step": 6131
+    },
+    {
+      "epoch": 14.992665036674817,
+      "grad_norm": 0.45194041858522627,
+      "learning_rate": 2.9500070122812085e-12,
+      "loss": 0.3828,
+      "step": 6132
+    },
+    {
+      "epoch": 14.995110024449877,
+      "grad_norm": 0.3662560694585631,
+      "learning_rate": 1.3111143709609863e-12,
+      "loss": 0.6831,
+      "step": 6133
+    },
+    {
+      "epoch": 14.99755501222494,
+      "grad_norm": 0.19005317783635392,
+      "learning_rate": 3.277786142508177e-13,
+      "loss": 0.5808,
+      "step": 6134
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.2616668832139272,
+      "learning_rate": 0.0,
+      "loss": 0.4434,
+      "step": 6135
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6135,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9138056050737152.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}