diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,32076 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9992357661444402,
+  "eval_steps": 500,
+  "global_step": 4578,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.927839542379235,
+      "learning_rate": 7.246376811594204e-08,
+      "loss": 1.4187,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.897891843901256,
+      "learning_rate": 1.4492753623188408e-07,
+      "loss": 1.4457,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.74515914875457,
+      "learning_rate": 2.173913043478261e-07,
+      "loss": 1.4833,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.976101368135662,
+      "learning_rate": 2.8985507246376816e-07,
+      "loss": 1.4363,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.379096829434,
+      "learning_rate": 3.623188405797102e-07,
+      "loss": 1.4435,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.085518432548206,
+      "learning_rate": 4.347826086956522e-07,
+      "loss": 1.4404,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.71633108948502,
+      "learning_rate": 5.072463768115942e-07,
+      "loss": 1.4255,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.341063051891314,
+      "learning_rate": 5.797101449275363e-07,
+      "loss": 1.4392,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.48220282379503,
+      "learning_rate": 6.521739130434783e-07,
+      "loss": 1.3336,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.17784222128118,
+      "learning_rate": 7.246376811594204e-07,
+      "loss": 1.3491,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.803270402693713,
+      "learning_rate": 7.971014492753623e-07,
+      "loss": 1.3097,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.346917637475407,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 1.1231,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.204438255171711,
+      "learning_rate": 9.420289855072465e-07,
+      "loss": 1.1275,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.811579961673939,
+      "learning_rate": 1.0144927536231885e-06,
+      "loss": 1.0892,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.696726752928813,
+      "learning_rate": 1.0869565217391306e-06,
+      "loss": 1.0825,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.282027045139074,
+      "learning_rate": 1.1594202898550726e-06,
+      "loss": 0.8599,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.887050978929367,
+      "learning_rate": 1.2318840579710147e-06,
+      "loss": 0.7142,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.407414255186723,
+      "learning_rate": 1.3043478260869566e-06,
+      "loss": 0.7291,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.267238791898411,
+      "learning_rate": 1.3768115942028987e-06,
+      "loss": 0.7227,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.317662631710455,
+      "learning_rate": 1.4492753623188408e-06,
+      "loss": 0.6645,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.572185584800502,
+      "learning_rate": 1.521739130434783e-06,
+      "loss": 0.6602,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.462535778352174,
+      "learning_rate": 1.5942028985507246e-06,
+      "loss": 0.5697,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.032898930458917,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.5431,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2470963695937933,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 0.529,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.43917854923453,
+      "learning_rate": 1.8115942028985508e-06,
+      "loss": 0.5044,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.270851750309605,
+      "learning_rate": 1.884057971014493e-06,
+      "loss": 0.4761,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.344259465725712,
+      "learning_rate": 1.956521739130435e-06,
+      "loss": 0.4551,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.397616949094848,
+      "learning_rate": 2.028985507246377e-06,
+      "loss": 0.4766,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9686829827396912,
+      "learning_rate": 2.101449275362319e-06,
+      "loss": 0.443,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6593284003291986,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 0.4391,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.364644300511352,
+      "learning_rate": 2.246376811594203e-06,
+      "loss": 0.4426,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3445575113660295,
+      "learning_rate": 2.3188405797101453e-06,
+      "loss": 0.4385,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9185278355815096,
+      "learning_rate": 2.391304347826087e-06,
+      "loss": 0.4299,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9485054698266127,
+      "learning_rate": 2.4637681159420295e-06,
+      "loss": 0.4169,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.147518241626594,
+      "learning_rate": 2.5362318840579714e-06,
+      "loss": 0.4169,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7935905242088153,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 0.413,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6661577876299598,
+      "learning_rate": 2.6811594202898555e-06,
+      "loss": 0.4002,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5787306696325698,
+      "learning_rate": 2.7536231884057974e-06,
+      "loss": 0.3924,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4630238170509346,
+      "learning_rate": 2.8260869565217393e-06,
+      "loss": 0.379,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5365561468314406,
+      "learning_rate": 2.8985507246376816e-06,
+      "loss": 0.3774,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3775350441634724,
+      "learning_rate": 2.9710144927536235e-06,
+      "loss": 0.3973,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3835980297643549,
+      "learning_rate": 3.043478260869566e-06,
+      "loss": 0.3645,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3752273513908346,
+      "learning_rate": 3.1159420289855073e-06,
+      "loss": 0.3656,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5101361925977579,
+      "learning_rate": 3.188405797101449e-06,
+      "loss": 0.3968,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.569269254320879,
+      "learning_rate": 3.2608695652173914e-06,
+      "loss": 0.3366,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.446657039577689,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.3978,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.606021832721281,
+      "learning_rate": 3.4057971014492756e-06,
+      "loss": 0.3687,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8700058888355593,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 0.3831,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.545529367498492,
+      "learning_rate": 3.55072463768116e-06,
+      "loss": 0.3691,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3404934046178592,
+      "learning_rate": 3.6231884057971017e-06,
+      "loss": 0.3579,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.503087435698558,
+      "learning_rate": 3.6956521739130436e-06,
+      "loss": 0.3491,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3947786844631256,
+      "learning_rate": 3.768115942028986e-06,
+      "loss": 0.3699,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3695875001860502,
+      "learning_rate": 3.840579710144928e-06,
+      "loss": 0.3632,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.704807685689475,
+      "learning_rate": 3.91304347826087e-06,
+      "loss": 0.3589,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5190608963245247,
+      "learning_rate": 3.9855072463768115e-06,
+      "loss": 0.3498,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8935311578022014,
+      "learning_rate": 4.057971014492754e-06,
+      "loss": 0.3605,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5508729063331417,
+      "learning_rate": 4.130434782608696e-06,
+      "loss": 0.3406,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5607986151497846,
+      "learning_rate": 4.202898550724638e-06,
+      "loss": 0.3728,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6155142674682406,
+      "learning_rate": 4.27536231884058e-06,
+      "loss": 0.3764,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6365566820277735,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.3427,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3888711431610992,
+      "learning_rate": 4.4202898550724645e-06,
+      "loss": 0.3632,
+      "step": 61
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.140544756823672,
+      "learning_rate": 4.492753623188406e-06,
+      "loss": 0.345,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3867545355964541,
+      "learning_rate": 4.565217391304348e-06,
+      "loss": 0.3436,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4919849435985038,
+      "learning_rate": 4.637681159420291e-06,
+      "loss": 0.3624,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4899992811817997,
+      "learning_rate": 4.710144927536232e-06,
+      "loss": 0.3608,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2761537340329068,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 0.3469,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.186101427118252,
+      "learning_rate": 4.855072463768117e-06,
+      "loss": 0.3322,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1966087376084877,
+      "learning_rate": 4.927536231884059e-06,
+      "loss": 0.3258,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4643000385390108,
+      "learning_rate": 5e-06,
+      "loss": 0.34,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2990708533125066,
+      "learning_rate": 5.072463768115943e-06,
+      "loss": 0.3448,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3418923534244593,
+      "learning_rate": 5.144927536231884e-06,
+      "loss": 0.3423,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1923145064850034,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 0.3482,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5004721527816665,
+      "learning_rate": 5.289855072463769e-06,
+      "loss": 0.3506,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.362576915554368,
+      "learning_rate": 5.362318840579711e-06,
+      "loss": 0.3282,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3736958938611148,
+      "learning_rate": 5.4347826086956525e-06,
+      "loss": 0.3308,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.257887980252355,
+      "learning_rate": 5.507246376811595e-06,
+      "loss": 0.3625,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4146222196894767,
+      "learning_rate": 5.579710144927537e-06,
+      "loss": 0.3263,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4161504913561358,
+      "learning_rate": 5.652173913043479e-06,
+      "loss": 0.3621,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4864831743937728,
+      "learning_rate": 5.724637681159421e-06,
+      "loss": 0.3346,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2955067548994326,
+      "learning_rate": 5.797101449275363e-06,
+      "loss": 0.3569,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2840029718046646,
+      "learning_rate": 5.8695652173913055e-06,
+      "loss": 0.3304,
+      "step": 81
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3835861216321874,
+      "learning_rate": 5.942028985507247e-06,
+      "loss": 0.3003,
+      "step": 82
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4575819550876519,
+      "learning_rate": 6.014492753623189e-06,
+      "loss": 0.3263,
+      "step": 83
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4319296123588712,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 0.3467,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2714564585991774,
+      "learning_rate": 6.159420289855072e-06,
+      "loss": 0.3475,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3144374166914212,
+      "learning_rate": 6.2318840579710145e-06,
+      "loss": 0.3399,
+      "step": 86
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2772149228478709,
+      "learning_rate": 6.304347826086958e-06,
+      "loss": 0.3008,
+      "step": 87
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2644670174707007,
+      "learning_rate": 6.376811594202898e-06,
+      "loss": 0.3158,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.535970858256948,
+      "learning_rate": 6.449275362318841e-06,
+      "loss": 0.3515,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4149071596883427,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 0.3392,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7250936274767752,
+      "learning_rate": 6.594202898550725e-06,
+      "loss": 0.3489,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3473945058421855,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3324,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.514662948671623,
+      "learning_rate": 6.739130434782609e-06,
+      "loss": 0.352,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4441343203629755,
+      "learning_rate": 6.811594202898551e-06,
+      "loss": 0.3254,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.838608451706039,
+      "learning_rate": 6.884057971014493e-06,
+      "loss": 0.3574,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3969014727069928,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 0.3692,
+      "step": 96
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3303035353638768,
+      "learning_rate": 7.028985507246377e-06,
+      "loss": 0.359,
+      "step": 97
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.750243316159876,
+      "learning_rate": 7.10144927536232e-06,
+      "loss": 0.3302,
+      "step": 98
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.361636711771327,
+      "learning_rate": 7.173913043478261e-06,
+      "loss": 0.3431,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1126922900428122,
+      "learning_rate": 7.246376811594203e-06,
+      "loss": 0.3255,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1754432545245468,
+      "learning_rate": 7.318840579710146e-06,
+      "loss": 0.3345,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2409071955652502,
+      "learning_rate": 7.391304347826087e-06,
+      "loss": 0.3244,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5041586322277132,
+      "learning_rate": 7.4637681159420295e-06,
+      "loss": 0.3331,
+      "step": 103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3372881209168015,
+      "learning_rate": 7.536231884057972e-06,
+      "loss": 0.3323,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5385686463709698,
+      "learning_rate": 7.608695652173914e-06,
+      "loss": 0.3393,
+      "step": 105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3528175815017602,
+      "learning_rate": 7.681159420289856e-06,
+      "loss": 0.3249,
+      "step": 106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2399929083163284,
+      "learning_rate": 7.753623188405797e-06,
+      "loss": 0.3155,
+      "step": 107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1469860609073483,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 0.3227,
+      "step": 108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2711197230802913,
+      "learning_rate": 7.898550724637682e-06,
+      "loss": 0.3203,
+      "step": 109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9945739153416782,
+      "learning_rate": 7.971014492753623e-06,
+      "loss": 0.3452,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2478469039054814,
+      "learning_rate": 8.043478260869566e-06,
+      "loss": 0.2981,
+      "step": 111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2700532912928022,
+      "learning_rate": 8.115942028985508e-06,
+      "loss": 0.3533,
+      "step": 112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0830502471303005,
+      "learning_rate": 8.188405797101449e-06,
+      "loss": 0.3243,
+      "step": 113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3608801998237683,
+      "learning_rate": 8.260869565217392e-06,
+      "loss": 0.3365,
+      "step": 114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.231900930108884,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.3087,
+      "step": 115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.293008016411097,
+      "learning_rate": 8.405797101449275e-06,
+      "loss": 0.3175,
+      "step": 116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.283843870401509,
+      "learning_rate": 8.478260869565218e-06,
+      "loss": 0.3204,
+      "step": 117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6897089606858247,
+      "learning_rate": 8.55072463768116e-06,
+      "loss": 0.3506,
+      "step": 118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.188548431948901,
+      "learning_rate": 8.623188405797103e-06,
+      "loss": 0.3263,
+      "step": 119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5620895526855842,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.3415,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2745794068876093,
+      "learning_rate": 8.768115942028986e-06,
+      "loss": 0.3536,
+      "step": 121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4585134771229842,
+      "learning_rate": 8.840579710144929e-06,
+      "loss": 0.298,
+      "step": 122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4233596771645736,
+      "learning_rate": 8.91304347826087e-06,
+      "loss": 0.3406,
+      "step": 123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4199007704429607,
+      "learning_rate": 8.985507246376812e-06,
+      "loss": 0.3506,
+      "step": 124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6043601815381534,
+      "learning_rate": 9.057971014492755e-06,
+      "loss": 0.3133,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2451195314887067,
+      "learning_rate": 9.130434782608697e-06,
+      "loss": 0.3428,
+      "step": 126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.204331737952428,
+      "learning_rate": 9.202898550724638e-06,
+      "loss": 0.3265,
+      "step": 127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2341532300638969,
+      "learning_rate": 9.275362318840581e-06,
+      "loss": 0.3118,
+      "step": 128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.450498748864925,
+      "learning_rate": 9.347826086956523e-06,
+      "loss": 0.3018,
+      "step": 129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6776537224206935,
+      "learning_rate": 9.420289855072464e-06,
+      "loss": 0.3402,
+      "step": 130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2582932033840875,
+      "learning_rate": 9.492753623188407e-06,
+      "loss": 0.3293,
+      "step": 131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2419539309748644,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 0.3475,
+      "step": 132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4626840319753827,
+      "learning_rate": 9.63768115942029e-06,
+      "loss": 0.3218,
+      "step": 133
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4086470688151649,
+      "learning_rate": 9.710144927536233e-06,
+      "loss": 0.3326,
+      "step": 134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1832342660059814,
+      "learning_rate": 9.782608695652175e-06,
+      "loss": 0.3495,
+      "step": 135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1662022582525668,
+      "learning_rate": 9.855072463768118e-06,
+      "loss": 0.3136,
+      "step": 136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1253275220462284,
+      "learning_rate": 9.927536231884058e-06,
+      "loss": 0.296,
+      "step": 137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3009747311477502,
+      "learning_rate": 1e-05,
+      "loss": 0.3365,
+      "step": 138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7819305568542165,
+      "learning_rate": 9.999998748376249e-06,
+      "loss": 0.3326,
+      "step": 139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2452433535210794,
+      "learning_rate": 9.999994993505624e-06,
+      "loss": 0.3387,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1798645618755854,
+      "learning_rate": 9.999988735390003e-06,
+      "loss": 0.3296,
+      "step": 141
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0364633257418387,
+      "learning_rate": 9.99997997403252e-06,
+      "loss": 0.3234,
+      "step": 142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.222980813314042,
+      "learning_rate": 9.999968709437563e-06,
+      "loss": 0.2901,
+      "step": 143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3402161753914412,
+      "learning_rate": 9.999954941610767e-06,
+      "loss": 0.3351,
+      "step": 144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1703805900778905,
+      "learning_rate": 9.999938670559031e-06,
+      "loss": 0.3282,
+      "step": 145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.159562511113094,
+      "learning_rate": 9.999919896290497e-06,
+      "loss": 0.3674,
+      "step": 146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2702607770193408,
+      "learning_rate": 9.999898618814565e-06,
+      "loss": 0.3059,
+      "step": 147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5372321983315806,
+      "learning_rate": 9.999874838141888e-06,
+      "loss": 0.3441,
+      "step": 148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4851303151582709,
+      "learning_rate": 9.999848554284371e-06,
+      "loss": 0.3488,
+      "step": 149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6337369564669073,
+      "learning_rate": 9.999819767255175e-06,
+      "loss": 0.3132,
+      "step": 150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3186991719581065,
+      "learning_rate": 9.99978847706871e-06,
+      "loss": 0.3263,
+      "step": 151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3474722275489617,
+      "learning_rate": 9.999754683740643e-06,
+      "loss": 0.3403,
+      "step": 152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0264026768297818,
+      "learning_rate": 9.999718387287891e-06,
+      "loss": 0.3093,
+      "step": 153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2170433089079558,
+      "learning_rate": 9.999679587728626e-06,
+      "loss": 0.3295,
+      "step": 154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1398537161072468,
+      "learning_rate": 9.999638285082276e-06,
+      "loss": 0.332,
+      "step": 155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0755992523122655,
+      "learning_rate": 9.999594479369515e-06,
+      "loss": 0.2995,
+      "step": 156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2964917321364673,
+      "learning_rate": 9.999548170612276e-06,
+      "loss": 0.3302,
+      "step": 157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.327032953106811,
+      "learning_rate": 9.999499358833745e-06,
+      "loss": 0.2932,
+      "step": 158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2988599669992225,
+      "learning_rate": 9.999448044058357e-06,
+      "loss": 0.3443,
+      "step": 159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.198581420427164,
+      "learning_rate": 9.999394226311804e-06,
+      "loss": 0.333,
+      "step": 160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.655566411858971,
+      "learning_rate": 9.999337905621031e-06,
+      "loss": 0.3239,
+      "step": 161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3684526271739883,
+      "learning_rate": 9.999279082014233e-06,
+      "loss": 0.3073,
+      "step": 162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0633167758802433,
+      "learning_rate": 9.99921775552086e-06,
+      "loss": 0.3116,
+      "step": 163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4370535135131788,
+      "learning_rate": 9.999153926171617e-06,
+      "loss": 0.3173,
+      "step": 164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1748888904195218,
+      "learning_rate": 9.999087593998458e-06,
+      "loss": 0.317,
+      "step": 165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2451025772117574,
+      "learning_rate": 9.999018759034594e-06,
+      "loss": 0.3164,
+      "step": 166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4790308949815747,
+      "learning_rate": 9.998947421314485e-06,
+      "loss": 0.3337,
+      "step": 167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3060764078076734,
+      "learning_rate": 9.998873580873848e-06,
+      "loss": 0.3024,
+      "step": 168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5314388960605791,
+      "learning_rate": 9.998797237749651e-06,
+      "loss": 0.3516,
+      "step": 169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0037962291925078,
+      "learning_rate": 9.998718391980113e-06,
+      "loss": 0.3041,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7960525739501374,
+      "learning_rate": 9.998637043604711e-06,
+      "loss": 0.3309,
+      "step": 171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1472716037873594,
+      "learning_rate": 9.998553192664172e-06,
+      "loss": 0.3181,
+      "step": 172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1105713736869427,
+      "learning_rate": 9.998466839200474e-06,
+      "loss": 0.3035,
+      "step": 173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3183773619437402,
+      "learning_rate": 9.998377983256851e-06,
+      "loss": 0.3275,
+      "step": 174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0755897306493971,
+      "learning_rate": 9.998286624877786e-06,
+      "loss": 0.3359,
+      "step": 175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0754923899735205,
+      "learning_rate": 9.998192764109023e-06,
+      "loss": 0.3065,
+      "step": 176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1209623389677426,
+      "learning_rate": 9.99809640099755e-06,
+      "loss": 0.3091,
+      "step": 177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.117836589100619,
+      "learning_rate": 9.99799753559161e-06,
+      "loss": 0.3206,
+      "step": 178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.083984183294291,
+      "learning_rate": 9.997896167940701e-06,
+      "loss": 0.3371,
+      "step": 179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2346954931836587,
+      "learning_rate": 9.997792298095572e-06,
+      "loss": 0.321,
+      "step": 180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.301643342929843,
+      "learning_rate": 9.99768592610823e-06,
+      "loss": 0.3305,
+      "step": 181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3291529863778404,
+      "learning_rate": 9.997577052031922e-06,
+      "loss": 0.3223,
+      "step": 182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1397111534686903,
+      "learning_rate": 9.997465675921163e-06,
+      "loss": 0.336,
+      "step": 183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.109101633720081,
+      "learning_rate": 9.99735179783171e-06,
+      "loss": 0.3066,
+      "step": 184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.066947895585489,
+      "learning_rate": 9.997235417820577e-06,
+      "loss": 0.3175,
+      "step": 185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5842360661975408,
+      "learning_rate": 9.997116535946028e-06,
+      "loss": 0.2934,
+      "step": 186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.003938102595649,
+      "learning_rate": 9.996995152267581e-06,
+      "loss": 0.3311,
+      "step": 187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0918705999012022,
+      "learning_rate": 9.99687126684601e-06,
+      "loss": 0.3262,
+      "step": 188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1370584685899559,
+      "learning_rate": 9.996744879743336e-06,
+      "loss": 0.3037,
+      "step": 189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2428123444521229,
+      "learning_rate": 9.996615991022835e-06,
+      "loss": 0.3315,
+      "step": 190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.030065043561401,
+      "learning_rate": 9.996484600749035e-06,
+      "loss": 0.3346,
+      "step": 191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1742115699895361,
+      "learning_rate": 9.996350708987715e-06,
+      "loss": 0.3342,
+      "step": 192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.014158772709524,
+      "learning_rate": 9.99621431580591e-06,
+      "loss": 0.3227,
+      "step": 193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2023129093779927,
+      "learning_rate": 9.996075421271905e-06,
+      "loss": 0.3323,
+      "step": 194
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1471162556215335,
+      "learning_rate": 9.995934025455236e-06,
+      "loss": 0.3224,
+      "step": 195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3048355390588604,
+      "learning_rate": 9.995790128426693e-06,
+      "loss": 0.3129,
+      "step": 196
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1856480756519412,
+      "learning_rate": 9.99564373025832e-06,
+      "loss": 0.2874,
+      "step": 197
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.15460665723546,
+      "learning_rate": 9.99549483102341e-06,
+      "loss": 0.3123,
+      "step": 198
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0720813629502337,
+      "learning_rate": 9.995343430796507e-06,
+      "loss": 0.2996,
+      "step": 199
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0381783301789584,
+      "learning_rate": 9.995189529653412e-06,
+      "loss": 0.2997,
+      "step": 200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0691423674105218,
+      "learning_rate": 9.995033127671174e-06,
+      "loss": 0.3114,
+      "step": 201
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2247336163527505,
+      "learning_rate": 9.9948742249281e-06,
+      "loss": 0.3301,
+      "step": 202
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1169216569925846,
+      "learning_rate": 9.994712821503737e-06,
+      "loss": 0.3099,
+      "step": 203
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0786223302051947,
+      "learning_rate": 9.994548917478899e-06,
+      "loss": 0.3029,
+      "step": 204
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9742186819710661,
+      "learning_rate": 9.99438251293564e-06,
+      "loss": 0.3104,
+      "step": 205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4947679915532137,
+      "learning_rate": 9.994213607957273e-06,
+      "loss": 0.314,
+      "step": 206
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1266330251621477,
+      "learning_rate": 9.994042202628356e-06,
+      "loss": 0.3122,
+      "step": 207
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2133447111179074,
+      "learning_rate": 9.993868297034709e-06,
+      "loss": 0.3348,
+      "step": 208
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1884809282444446,
+      "learning_rate": 9.993691891263393e-06,
+      "loss": 0.3535,
+      "step": 209
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3461809634713862,
+      "learning_rate": 9.993512985402724e-06,
+      "loss": 0.3384,
+      "step": 210
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0919195285934127,
+      "learning_rate": 9.993331579542278e-06,
+      "loss": 0.3156,
+      "step": 211
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5281963800242617,
+      "learning_rate": 9.993147673772869e-06,
+      "loss": 0.3121,
+      "step": 212
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0783489087310267,
+      "learning_rate": 9.992961268186575e-06,
+      "loss": 0.2956,
+      "step": 213
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0523108176910188,
+      "learning_rate": 9.992772362876714e-06,
+      "loss": 0.3045,
+      "step": 214
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0574734307457236,
+      "learning_rate": 9.992580957937868e-06,
+      "loss": 0.3073,
+      "step": 215
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0779839236415683,
+      "learning_rate": 9.992387053465857e-06,
+      "loss": 0.3122,
+      "step": 216
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1115362852594475,
+      "learning_rate": 9.992190649557765e-06,
+      "loss": 0.3297,
+      "step": 217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9826547704108827,
+      "learning_rate": 9.991991746311916e-06,
+      "loss": 0.2954,
+      "step": 218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0228229869982102,
+      "learning_rate": 9.991790343827896e-06,
+      "loss": 0.3226,
+      "step": 219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0237401183378125,
+      "learning_rate": 9.991586442206533e-06,
+      "loss": 0.3104,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0561665964241984,
+      "learning_rate": 9.991380041549911e-06,
+      "loss": 0.3181,
+      "step": 221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.094247937702541,
+      "learning_rate": 9.99117114196137e-06,
+      "loss": 0.3291,
+      "step": 222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9636152414117666,
+      "learning_rate": 9.990959743545487e-06,
+      "loss": 0.298,
+      "step": 223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1018451060945151,
+      "learning_rate": 9.990745846408102e-06,
+      "loss": 0.3096,
+      "step": 224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.073469403865194,
+      "learning_rate": 9.990529450656303e-06,
+      "loss": 0.3194,
+      "step": 225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4133791194620402,
+      "learning_rate": 9.99031055639843e-06,
+      "loss": 0.3035,
+      "step": 226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1007667882807946,
+      "learning_rate": 9.99008916374407e-06,
+      "loss": 0.3509,
+      "step": 227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1433193922988405,
+      "learning_rate": 9.989865272804064e-06,
+      "loss": 0.2989,
+      "step": 228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9938673845852148,
+      "learning_rate": 9.989638883690501e-06,
+      "loss": 0.3166,
+      "step": 229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1863879277806728,
+      "learning_rate": 9.989409996516726e-06,
+      "loss": 0.2999,
+      "step": 230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.300733432556779,
+      "learning_rate": 9.989178611397328e-06,
+      "loss": 0.3374,
+      "step": 231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1092396972572642,
+      "learning_rate": 9.98894472844815e-06,
+      "loss": 0.3454,
+      "step": 232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1082305505618948,
+      "learning_rate": 9.988708347786288e-06,
+      "loss": 0.3187,
+      "step": 233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.084537611910107,
+      "learning_rate": 9.988469469530086e-06,
+      "loss": 0.3168,
+      "step": 234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5504425386821095,
+      "learning_rate": 9.988228093799136e-06,
+      "loss": 0.3044,
+      "step": 235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0472422083626418,
+      "learning_rate": 9.987984220714283e-06,
+      "loss": 0.2997,
+      "step": 236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1526004025002554,
+      "learning_rate": 9.987737850397623e-06,
+      "loss": 0.3229,
+      "step": 237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.995597418812663,
+      "learning_rate": 9.9874889829725e-06,
+      "loss": 0.2903,
+      "step": 238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.180901658981736,
+      "learning_rate": 9.987237618563509e-06,
+      "loss": 0.2961,
+      "step": 239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0988091649633387,
+      "learning_rate": 9.986983757296498e-06,
+      "loss": 0.3209,
+      "step": 240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9633388976798277,
+      "learning_rate": 9.98672739929856e-06,
+      "loss": 0.3178,
+      "step": 241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1107617475488154,
+      "learning_rate": 9.986468544698041e-06,
+      "loss": 0.3202,
+      "step": 242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1581880095530737,
+      "learning_rate": 9.986207193624537e-06,
+      "loss": 0.3274,
+      "step": 243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2466228317717647,
+      "learning_rate": 9.985943346208893e-06,
+      "loss": 0.3182,
+      "step": 244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2918403696043619,
+      "learning_rate": 9.985677002583204e-06,
+      "loss": 0.3312,
+      "step": 245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0776335120754421,
+      "learning_rate": 9.985408162880813e-06,
+      "loss": 0.3207,
+      "step": 246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.270331455397705,
+      "learning_rate": 9.98513682723632e-06,
+      "loss": 0.2932,
+      "step": 247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9357055456401338,
+      "learning_rate": 9.984862995785564e-06,
+      "loss": 0.2993,
+      "step": 248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0503760574363121,
+      "learning_rate": 9.984586668665641e-06,
+      "loss": 0.3098,
+      "step": 249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1405732809908076,
+      "learning_rate": 9.984307846014891e-06,
+      "loss": 0.3212,
+      "step": 250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.105262949385254,
+      "learning_rate": 9.98402652797291e-06,
+      "loss": 0.317,
+      "step": 251
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0905158046101258,
+      "learning_rate": 9.983742714680538e-06,
+      "loss": 0.336,
+      "step": 252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0167542082315735,
+      "learning_rate": 9.983456406279866e-06,
+      "loss": 0.308,
+      "step": 253
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.110569696781936,
+      "learning_rate": 9.983167602914236e-06,
+      "loss": 0.2811,
+      "step": 254
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.982958052866081,
+      "learning_rate": 9.982876304728233e-06,
+      "loss": 0.2926,
+      "step": 255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.159873341950794,
+      "learning_rate": 9.9825825118677e-06,
+      "loss": 0.3335,
+      "step": 256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1730771598354455,
+      "learning_rate": 9.98228622447972e-06,
+      "loss": 0.3185,
+      "step": 257
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0815016170858163,
+      "learning_rate": 9.981987442712634e-06,
+      "loss": 0.3273,
+      "step": 258
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.186622352221272,
+      "learning_rate": 9.981686166716023e-06,
+      "loss": 0.3136,
+      "step": 259
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1749236789950672,
+      "learning_rate": 9.981382396640721e-06,
+      "loss": 0.3426,
+      "step": 260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1014043047613271,
+      "learning_rate": 9.981076132638811e-06,
+      "loss": 0.3224,
+      "step": 261
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.94525537729492,
+      "learning_rate": 9.980767374863627e-06,
+      "loss": 0.3416,
+      "step": 262
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0683968749589203,
+      "learning_rate": 9.980456123469743e-06,
+      "loss": 0.334,
+      "step": 263
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0994231872111615,
+      "learning_rate": 9.98014237861299e-06,
+      "loss": 0.3134,
+      "step": 264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.065805927674682,
+      "learning_rate": 9.979826140450443e-06,
+      "loss": 0.3224,
+      "step": 265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.073814763103337,
+      "learning_rate": 9.97950740914043e-06,
+      "loss": 0.3157,
+      "step": 266
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0523476836246886,
+      "learning_rate": 9.979186184842517e-06,
+      "loss": 0.3354,
+      "step": 267
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.963122444612501,
+      "learning_rate": 9.978862467717532e-06,
+      "loss": 0.3124,
+      "step": 268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9586042838287766,
+      "learning_rate": 9.97853625792754e-06,
+      "loss": 0.2932,
+      "step": 269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0967662375613878,
+      "learning_rate": 9.978207555635856e-06,
+      "loss": 0.3034,
+      "step": 270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9349318369936347,
+      "learning_rate": 9.977876361007049e-06,
+      "loss": 0.3086,
+      "step": 271
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.046613402235858,
+      "learning_rate": 9.977542674206927e-06,
+      "loss": 0.3081,
+      "step": 272
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.01733626908822,
+      "learning_rate": 9.977206495402554e-06,
+      "loss": 0.338,
+      "step": 273
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0447159708171132,
+      "learning_rate": 9.976867824762236e-06,
+      "loss": 0.3269,
+      "step": 274
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9199517916239215,
+      "learning_rate": 9.976526662455528e-06,
+      "loss": 0.3118,
+      "step": 275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9968529262830496,
+      "learning_rate": 9.976183008653232e-06,
+      "loss": 0.305,
+      "step": 276
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5508491711556633,
+      "learning_rate": 9.975836863527402e-06,
+      "loss": 0.3192,
+      "step": 277
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9621911060205496,
+      "learning_rate": 9.97548822725133e-06,
+      "loss": 0.3081,
+      "step": 278
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.124969825190591,
+      "learning_rate": 9.975137099999566e-06,
+      "loss": 0.3267,
+      "step": 279
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.040978423519749,
+      "learning_rate": 9.974783481947896e-06,
+      "loss": 0.3109,
+      "step": 280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9775265468701901,
+      "learning_rate": 9.974427373273364e-06,
+      "loss": 0.3095,
+      "step": 281
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9560296959648664,
+      "learning_rate": 9.974068774154252e-06,
+      "loss": 0.315,
+      "step": 282
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9178767765535655,
+      "learning_rate": 9.973707684770095e-06,
+      "loss": 0.3246,
+      "step": 283
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9903216929660608,
+      "learning_rate": 9.97334410530167e-06,
+      "loss": 0.2988,
+      "step": 284
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0472066533751248,
+      "learning_rate": 9.972978035931003e-06,
+      "loss": 0.3535,
+      "step": 285
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0657222670639235,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 0.2986,
+      "step": 286
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0651977174794531,
+      "learning_rate": 9.972238428217283e-06,
+      "loss": 0.3317,
+      "step": 287
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2806287162676435,
+      "learning_rate": 9.971864890244514e-06,
+      "loss": 0.3182,
+      "step": 288
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1708511386388298,
+      "learning_rate": 9.971488863110071e-06,
+      "loss": 0.3316,
+      "step": 289
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9542484995807301,
+      "learning_rate": 9.971110347002213e-06,
+      "loss": 0.2973,
+      "step": 290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8599122437234269,
+      "learning_rate": 9.970729342110446e-06,
+      "loss": 0.2831,
+      "step": 291
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9023120893983787,
+      "learning_rate": 9.970345848625516e-06,
+      "loss": 0.3197,
+      "step": 292
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9509171249603806,
+      "learning_rate": 9.96995986673942e-06,
+      "loss": 0.3199,
+      "step": 293
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0023145862867555,
+      "learning_rate": 9.9695713966454e-06,
+      "loss": 0.3099,
+      "step": 294
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9976221452260706,
+      "learning_rate": 9.969180438537944e-06,
+      "loss": 0.3258,
+      "step": 295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.021394507281699,
+      "learning_rate": 9.968786992612785e-06,
+      "loss": 0.3151,
+      "step": 296
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.996597980135569,
+      "learning_rate": 9.9683910590669e-06,
+      "loss": 0.3058,
+      "step": 297
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0116560872951126,
+      "learning_rate": 9.967992638098517e-06,
+      "loss": 0.2971,
+      "step": 298
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9387362317513296,
+      "learning_rate": 9.967591729907098e-06,
+      "loss": 0.3295,
+      "step": 299
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0055588358821534,
+      "learning_rate": 9.967188334693364e-06,
+      "loss": 0.3279,
+      "step": 300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0740687755465586,
+      "learning_rate": 9.966782452659272e-06,
+      "loss": 0.2887,
+      "step": 301
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.078696627638513,
+      "learning_rate": 9.966374084008025e-06,
+      "loss": 0.2979,
+      "step": 302
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0671969167739481,
+      "learning_rate": 9.965963228944077e-06,
+      "loss": 0.318,
+      "step": 303
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.251030207096568,
+      "learning_rate": 9.965549887673121e-06,
+      "loss": 0.3306,
+      "step": 304
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3745468869874284,
+      "learning_rate": 9.965134060402092e-06,
+      "loss": 0.3203,
+      "step": 305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9380305295224106,
+      "learning_rate": 9.964715747339178e-06,
+      "loss": 0.2963,
+      "step": 306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9433067732381588,
+      "learning_rate": 9.964294948693805e-06,
+      "loss": 0.3308,
+      "step": 307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8916306843891029,
+      "learning_rate": 9.963871664676647e-06,
+      "loss": 0.318,
+      "step": 308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.901071424915498,
+      "learning_rate": 9.963445895499622e-06,
+      "loss": 0.2913,
+      "step": 309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0774026053517727,
+      "learning_rate": 9.963017641375888e-06,
+      "loss": 0.2945,
+      "step": 310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9448531473952472,
+      "learning_rate": 9.962586902519851e-06,
+      "loss": 0.3312,
+      "step": 311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1065637407516429,
+      "learning_rate": 9.962153679147162e-06,
+      "loss": 0.3072,
+      "step": 312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8886810248773547,
+      "learning_rate": 9.961717971474714e-06,
+      "loss": 0.2916,
+      "step": 313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9202353889527449,
+      "learning_rate": 9.961279779720641e-06,
+      "loss": 0.3026,
+      "step": 314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8948687502332382,
+      "learning_rate": 9.960839104104326e-06,
+      "loss": 0.3147,
+      "step": 315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9247523546770249,
+      "learning_rate": 9.960395944846394e-06,
+      "loss": 0.3324,
+      "step": 316
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9372170838558451,
+      "learning_rate": 9.95995030216871e-06,
+      "loss": 0.2989,
+      "step": 317
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9153428237908091,
+      "learning_rate": 9.959502176294384e-06,
+      "loss": 0.3003,
+      "step": 318
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0423926238896777,
+      "learning_rate": 9.959051567447775e-06,
+      "loss": 0.3285,
+      "step": 319
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8832880745515919,
+      "learning_rate": 9.958598475854474e-06,
+      "loss": 0.3064,
+      "step": 320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0312637729058367,
+      "learning_rate": 9.958142901741325e-06,
+      "loss": 0.3061,
+      "step": 321
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9240811725076746,
+      "learning_rate": 9.957684845336409e-06,
+      "loss": 0.3143,
+      "step": 322
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.999152113776159,
+      "learning_rate": 9.957224306869053e-06,
+      "loss": 0.2859,
+      "step": 323
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0193682871626268,
+      "learning_rate": 9.956761286569824e-06,
+      "loss": 0.2806,
+      "step": 324
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9298802461272866,
+      "learning_rate": 9.956295784670536e-06,
+      "loss": 0.309,
+      "step": 325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2198396876716135,
+      "learning_rate": 9.955827801404237e-06,
+      "loss": 0.3315,
+      "step": 326
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1338065413828167,
+      "learning_rate": 9.955357337005226e-06,
+      "loss": 0.3198,
+      "step": 327
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0473921783074103,
+      "learning_rate": 9.954884391709043e-06,
+      "loss": 0.3044,
+      "step": 328
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9303503913405896,
+      "learning_rate": 9.954408965752463e-06,
+      "loss": 0.3477,
+      "step": 329
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9485470055321111,
+      "learning_rate": 9.95393105937351e-06,
+      "loss": 0.303,
+      "step": 330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9952262995292349,
+      "learning_rate": 9.953450672811449e-06,
+      "loss": 0.2898,
+      "step": 331
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9986212771675842,
+      "learning_rate": 9.952967806306784e-06,
+      "loss": 0.3003,
+      "step": 332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0939978520542877,
+      "learning_rate": 9.95248246010126e-06,
+      "loss": 0.3162,
+      "step": 333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0275941580907522,
+      "learning_rate": 9.951994634437866e-06,
+      "loss": 0.3082,
+      "step": 334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9859442833998504,
+      "learning_rate": 9.951504329560835e-06,
+      "loss": 0.3024,
+      "step": 335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9897009307894633,
+      "learning_rate": 9.951011545715637e-06,
+      "loss": 0.3262,
+      "step": 336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9084229349571925,
+      "learning_rate": 9.95051628314898e-06,
+      "loss": 0.3123,
+      "step": 337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0364876270795529,
+      "learning_rate": 9.950018542108818e-06,
+      "loss": 0.2941,
+      "step": 338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0608789586393688,
+      "learning_rate": 9.94951832284435e-06,
+      "loss": 0.3364,
+      "step": 339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0108927742181069,
+      "learning_rate": 9.949015625606007e-06,
+      "loss": 0.3177,
+      "step": 340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.043605085592431,
+      "learning_rate": 9.94851045064546e-06,
+      "loss": 0.3129,
+      "step": 341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8573830134527594,
+      "learning_rate": 9.948002798215632e-06,
+      "loss": 0.3053,
+      "step": 342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1135324469682708,
+      "learning_rate": 9.947492668570675e-06,
+      "loss": 0.3288,
+      "step": 343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9385091670226149,
+      "learning_rate": 9.946980061965985e-06,
+      "loss": 0.2919,
+      "step": 344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9222855276245867,
+      "learning_rate": 9.9464649786582e-06,
+      "loss": 0.2984,
+      "step": 345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0394853455705548,
+      "learning_rate": 9.945947418905193e-06,
+      "loss": 0.3109,
+      "step": 346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9893976115424427,
+      "learning_rate": 9.945427382966084e-06,
+      "loss": 0.309,
+      "step": 347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0727380540111127,
+      "learning_rate": 9.944904871101227e-06,
+      "loss": 0.3152,
+      "step": 348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0614228225830642,
+      "learning_rate": 9.944379883572217e-06,
+      "loss": 0.2902,
+      "step": 349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2559430180872728,
+      "learning_rate": 9.943852420641888e-06,
+      "loss": 0.3044,
+      "step": 350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0021945015143596,
+      "learning_rate": 9.943322482574316e-06,
+      "loss": 0.3169,
+      "step": 351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1735242976605886,
+      "learning_rate": 9.942790069634814e-06,
+      "loss": 0.3175,
+      "step": 352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9373631283414116,
+      "learning_rate": 9.94225518208993e-06,
+      "loss": 0.2979,
+      "step": 353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8510471799848879,
+      "learning_rate": 9.941717820207461e-06,
+      "loss": 0.2944,
+      "step": 354
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8607333297375305,
+      "learning_rate": 9.941177984256436e-06,
+      "loss": 0.2844,
+      "step": 355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8555979469501821,
+      "learning_rate": 9.94063567450712e-06,
+      "loss": 0.321,
+      "step": 356
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9946504089457169,
+      "learning_rate": 9.940090891231024e-06,
+      "loss": 0.3025,
+      "step": 357
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9757285803750265,
+      "learning_rate": 9.939543634700891e-06,
+      "loss": 0.2916,
+      "step": 358
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0054518067802982,
+      "learning_rate": 9.938993905190706e-06,
+      "loss": 0.2957,
+      "step": 359
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9837209394924226,
+      "learning_rate": 9.938441702975689e-06,
+      "loss": 0.3031,
+      "step": 360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0013485770654755,
+      "learning_rate": 9.937887028332303e-06,
+      "loss": 0.3251,
+      "step": 361
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8534314101989456,
+      "learning_rate": 9.937329881538242e-06,
+      "loss": 0.2961,
+      "step": 362
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0111835333263233,
+      "learning_rate": 9.936770262872444e-06,
+      "loss": 0.2896,
+      "step": 363
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1019257704184062,
+      "learning_rate": 9.936208172615081e-06,
+      "loss": 0.2877,
+      "step": 364
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8957236672935278,
+      "learning_rate": 9.93564361104756e-06,
+      "loss": 0.2887,
+      "step": 365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.066423831975303,
+      "learning_rate": 9.935076578452535e-06,
+      "loss": 0.2815,
+      "step": 366
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9330818953448664,
+      "learning_rate": 9.934507075113886e-06,
+      "loss": 0.3098,
+      "step": 367
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9359387921811626,
+      "learning_rate": 9.933935101316735e-06,
+      "loss": 0.2848,
+      "step": 368
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0036203294046473,
+      "learning_rate": 9.933360657347441e-06,
+      "loss": 0.3012,
+      "step": 369
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1517258280605915,
+      "learning_rate": 9.932783743493599e-06,
+      "loss": 0.3183,
+      "step": 370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2179303555628678,
+      "learning_rate": 9.932204360044042e-06,
+      "loss": 0.3099,
+      "step": 371
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9622580536974517,
+      "learning_rate": 9.931622507288835e-06,
+      "loss": 0.3056,
+      "step": 372
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8848572248852374,
+      "learning_rate": 9.931038185519285e-06,
+      "loss": 0.3082,
+      "step": 373
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9282076251105636,
+      "learning_rate": 9.930451395027932e-06,
+      "loss": 0.2989,
+      "step": 374
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.993512626449498,
+      "learning_rate": 9.92986213610855e-06,
+      "loss": 0.3188,
+      "step": 375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1632964177794034,
+      "learning_rate": 9.929270409056154e-06,
+      "loss": 0.3083,
+      "step": 376
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9096574647380296,
+      "learning_rate": 9.92867621416699e-06,
+      "loss": 0.3112,
+      "step": 377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0037922126946068,
+      "learning_rate": 9.928079551738542e-06,
+      "loss": 0.3322,
+      "step": 378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9169449498850912,
+      "learning_rate": 9.927480422069531e-06,
+      "loss": 0.3061,
+      "step": 379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8554853448022172,
+      "learning_rate": 9.926878825459907e-06,
+      "loss": 0.305,
+      "step": 380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1684483774222083,
+      "learning_rate": 9.926274762210862e-06,
+      "loss": 0.3334,
+      "step": 381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9665816576560077,
+      "learning_rate": 9.925668232624819e-06,
+      "loss": 0.2771,
+      "step": 382
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9021152468617711,
+      "learning_rate": 9.925059237005437e-06,
+      "loss": 0.3261,
+      "step": 383
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9981236315596377,
+      "learning_rate": 9.924447775657607e-06,
+      "loss": 0.3004,
+      "step": 384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9890262642639216,
+      "learning_rate": 9.92383384888746e-06,
+      "loss": 0.3102,
+      "step": 385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2643205276341294,
+      "learning_rate": 9.923217457002356e-06,
+      "loss": 0.3275,
+      "step": 386
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0018660127956895,
+      "learning_rate": 9.922598600310893e-06,
+      "loss": 0.3179,
+      "step": 387
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1842429821810712,
+      "learning_rate": 9.9219772791229e-06,
+      "loss": 0.2959,
+      "step": 388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9495617188462666,
+      "learning_rate": 9.921353493749443e-06,
+      "loss": 0.3208,
+      "step": 389
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0181793720679035,
+      "learning_rate": 9.92072724450282e-06,
+      "loss": 0.2995,
+      "step": 390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.183244420602268,
+      "learning_rate": 9.920098531696558e-06,
+      "loss": 0.3414,
+      "step": 391
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9810838624708067,
+      "learning_rate": 9.919467355645425e-06,
+      "loss": 0.3107,
+      "step": 392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1963752641284957,
+      "learning_rate": 9.91883371666542e-06,
+      "loss": 0.2857,
+      "step": 393
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.96111312858274,
+      "learning_rate": 9.91819761507377e-06,
+      "loss": 0.3299,
+      "step": 394
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.941573701537557,
+      "learning_rate": 9.917559051188943e-06,
+      "loss": 0.31,
+      "step": 395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0846185692878827,
+      "learning_rate": 9.916918025330635e-06,
+      "loss": 0.2914,
+      "step": 396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9532262466225393,
+      "learning_rate": 9.916274537819774e-06,
+      "loss": 0.3144,
+      "step": 397
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8779321643121134,
+      "learning_rate": 9.915628588978522e-06,
+      "loss": 0.3015,
+      "step": 398
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9136168698292808,
+      "learning_rate": 9.914980179130273e-06,
+      "loss": 0.3086,
+      "step": 399
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9614224023650695,
+      "learning_rate": 9.914329308599655e-06,
+      "loss": 0.3212,
+      "step": 400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0627507968644327,
+      "learning_rate": 9.913675977712522e-06,
+      "loss": 0.3195,
+      "step": 401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.996284369770512,
+      "learning_rate": 9.913020186795967e-06,
+      "loss": 0.2963,
+      "step": 402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.544107608251565,
+      "learning_rate": 9.912361936178312e-06,
+      "loss": 0.3108,
+      "step": 403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9418944496892594,
+      "learning_rate": 9.911701226189105e-06,
+      "loss": 0.3186,
+      "step": 404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8861445030797086,
+      "learning_rate": 9.911038057159136e-06,
+      "loss": 0.2807,
+      "step": 405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.009366987777871,
+      "learning_rate": 9.910372429420416e-06,
+      "loss": 0.3192,
+      "step": 406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2726152801960393,
+      "learning_rate": 9.909704343306194e-06,
+      "loss": 0.2786,
+      "step": 407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.120314362504939,
+      "learning_rate": 9.909033799150947e-06,
+      "loss": 0.2925,
+      "step": 408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1497713595302208,
+      "learning_rate": 9.908360797290379e-06,
+      "loss": 0.3049,
+      "step": 409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.032079824579938,
+      "learning_rate": 9.907685338061433e-06,
+      "loss": 0.3024,
+      "step": 410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0079738722128897,
+      "learning_rate": 9.907007421802273e-06,
+      "loss": 0.2911,
+      "step": 411
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9383727811528801,
+      "learning_rate": 9.906327048852301e-06,
+      "loss": 0.3125,
+      "step": 412
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0083699434279851,
+      "learning_rate": 9.90564421955214e-06,
+      "loss": 0.2851,
+      "step": 413
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9386567240724545,
+      "learning_rate": 9.904958934243655e-06,
+      "loss": 0.3109,
+      "step": 414
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8811309936336299,
+      "learning_rate": 9.90427119326993e-06,
+      "loss": 0.3076,
+      "step": 415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.105063219517801,
+      "learning_rate": 9.903580996975282e-06,
+      "loss": 0.3021,
+      "step": 416
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2613805596613774,
+      "learning_rate": 9.902888345705257e-06,
+      "loss": 0.3202,
+      "step": 417
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9488354681120358,
+      "learning_rate": 9.902193239806634e-06,
+      "loss": 0.3342,
+      "step": 418
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.927622142743386,
+      "learning_rate": 9.901495679627415e-06,
+      "loss": 0.314,
+      "step": 419
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8859498901659816,
+      "learning_rate": 9.90079566551683e-06,
+      "loss": 0.269,
+      "step": 420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0748757815637442,
+      "learning_rate": 9.900093197825349e-06,
+      "loss": 0.3249,
+      "step": 421
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9194292091909244,
+      "learning_rate": 9.899388276904653e-06,
+      "loss": 0.3306,
+      "step": 422
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0150030958300658,
+      "learning_rate": 9.898680903107668e-06,
+      "loss": 0.3157,
+      "step": 423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.891373489557512,
+      "learning_rate": 9.897971076788532e-06,
+      "loss": 0.3197,
+      "step": 424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1981184304700172,
+      "learning_rate": 9.897258798302626e-06,
+      "loss": 0.3073,
+      "step": 425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.953349834482229,
+      "learning_rate": 9.89654406800655e-06,
+      "loss": 0.3344,
+      "step": 426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9298548668425412,
+      "learning_rate": 9.895826886258134e-06,
+      "loss": 0.2893,
+      "step": 427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8924569349922167,
+      "learning_rate": 9.895107253416434e-06,
+      "loss": 0.3168,
+      "step": 428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8883197338440372,
+      "learning_rate": 9.894385169841731e-06,
+      "loss": 0.3238,
+      "step": 429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9219843900420306,
+      "learning_rate": 9.893660635895542e-06,
+      "loss": 0.2987,
+      "step": 430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.920032675790607,
+      "learning_rate": 9.892933651940597e-06,
+      "loss": 0.2735,
+      "step": 431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2872533451297792,
+      "learning_rate": 9.892204218340866e-06,
+      "loss": 0.3223,
+      "step": 432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1201481350825793,
+      "learning_rate": 9.891472335461537e-06,
+      "loss": 0.2833,
+      "step": 433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1018946245134453,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.3121,
+      "step": 434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9371832132221567,
+      "learning_rate": 9.890001223330983e-06,
+      "loss": 0.2834,
+      "step": 435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8837263611188634,
+      "learning_rate": 9.889261994816266e-06,
+      "loss": 0.3261,
+      "step": 436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.999209350760891,
+      "learning_rate": 9.888520318494976e-06,
+      "loss": 0.3164,
+      "step": 437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9399745877296535,
+      "learning_rate": 9.887776194738433e-06,
+      "loss": 0.3073,
+      "step": 438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9307878862909978,
+      "learning_rate": 9.887029623919178e-06,
+      "loss": 0.305,
+      "step": 439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9359830420515445,
+      "learning_rate": 9.886280606410984e-06,
+      "loss": 0.3241,
+      "step": 440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0779149422452345,
+      "learning_rate": 9.885529142588845e-06,
+      "loss": 0.3345,
+      "step": 441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0445050089103123,
+      "learning_rate": 9.884775232828985e-06,
+      "loss": 0.2985,
+      "step": 442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.588180393332987,
+      "learning_rate": 9.884018877508844e-06,
+      "loss": 0.2869,
+      "step": 443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8225385251085251,
+      "learning_rate": 9.883260077007092e-06,
+      "loss": 0.2973,
+      "step": 444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1843572762386145,
+      "learning_rate": 9.882498831703623e-06,
+      "loss": 0.3103,
+      "step": 445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9672531829215212,
+      "learning_rate": 9.881735141979554e-06,
+      "loss": 0.3028,
+      "step": 446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.983255598703798,
+      "learning_rate": 9.880969008217225e-06,
+      "loss": 0.3049,
+      "step": 447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1501883652980007,
+      "learning_rate": 9.8802004308002e-06,
+      "loss": 0.3015,
+      "step": 448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9415845537765902,
+      "learning_rate": 9.87942941011327e-06,
+      "loss": 0.3026,
+      "step": 449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0104901966109014,
+      "learning_rate": 9.878655946542443e-06,
+      "loss": 0.2978,
+      "step": 450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.969030724128093,
+      "learning_rate": 9.877880040474954e-06,
+      "loss": 0.2883,
+      "step": 451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9947073123499418,
+      "learning_rate": 9.87710169229926e-06,
+      "loss": 0.2938,
+      "step": 452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8895897020735741,
+      "learning_rate": 9.876320902405041e-06,
+      "loss": 0.3097,
+      "step": 453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9610495755236163,
+      "learning_rate": 9.8755376711832e-06,
+      "loss": 0.3166,
+      "step": 454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1935566339241421,
+      "learning_rate": 9.87475199902586e-06,
+      "loss": 0.3132,
+      "step": 455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2426237065397119,
+      "learning_rate": 9.873963886326365e-06,
+      "loss": 0.3092,
+      "step": 456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9905885265177632,
+      "learning_rate": 9.873173333479289e-06,
+      "loss": 0.3294,
+      "step": 457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9576118765319891,
+      "learning_rate": 9.872380340880416e-06,
+      "loss": 0.3051,
+      "step": 458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9928402712465237,
+      "learning_rate": 9.871584908926763e-06,
+      "loss": 0.3004,
+      "step": 459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9407881228479146,
+      "learning_rate": 9.870787038016556e-06,
+      "loss": 0.3029,
+      "step": 460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0026895855306448,
+      "learning_rate": 9.869986728549254e-06,
+      "loss": 0.2894,
+      "step": 461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9043135230663318,
+      "learning_rate": 9.869183980925531e-06,
+      "loss": 0.2834,
+      "step": 462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0219548605370647,
+      "learning_rate": 9.86837879554728e-06,
+      "loss": 0.3194,
+      "step": 463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2469620942123085,
+      "learning_rate": 9.867571172817617e-06,
+      "loss": 0.3013,
+      "step": 464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9821388257202688,
+      "learning_rate": 9.866761113140881e-06,
+      "loss": 0.2876,
+      "step": 465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0545764677633238,
+      "learning_rate": 9.865948616922623e-06,
+      "loss": 0.2968,
+      "step": 466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9553222249805321,
+      "learning_rate": 9.865133684569622e-06,
+      "loss": 0.3087,
+      "step": 467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8765110502370476,
+      "learning_rate": 9.864316316489873e-06,
+      "loss": 0.311,
+      "step": 468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.974331466027113,
+      "learning_rate": 9.863496513092592e-06,
+      "loss": 0.2938,
+      "step": 469
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9037719876901005,
+      "learning_rate": 9.862674274788212e-06,
+      "loss": 0.2889,
+      "step": 470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9030537938374069,
+      "learning_rate": 9.861849601988384e-06,
+      "loss": 0.3131,
+      "step": 471
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.924433010186973,
+      "learning_rate": 9.861022495105983e-06,
+      "loss": 0.2932,
+      "step": 472
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8598486456425402,
+      "learning_rate": 9.860192954555099e-06,
+      "loss": 0.3052,
+      "step": 473
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1012129440763454,
+      "learning_rate": 9.85936098075104e-06,
+      "loss": 0.3148,
+      "step": 474
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0438212793884623,
+      "learning_rate": 9.858526574110336e-06,
+      "loss": 0.2883,
+      "step": 475
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0354742602397748,
+      "learning_rate": 9.85768973505073e-06,
+      "loss": 0.2966,
+      "step": 476
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8961968246385683,
+      "learning_rate": 9.856850463991186e-06,
+      "loss": 0.2876,
+      "step": 477
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9473193344207916,
+      "learning_rate": 9.856008761351882e-06,
+      "loss": 0.2803,
+      "step": 478
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8120752955006466,
+      "learning_rate": 9.855164627554219e-06,
+      "loss": 0.2977,
+      "step": 479
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9343358167444801,
+      "learning_rate": 9.85431806302081e-06,
+      "loss": 0.3044,
+      "step": 480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8908135564409262,
+      "learning_rate": 9.853469068175491e-06,
+      "loss": 0.3027,
+      "step": 481
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9968800241149782,
+      "learning_rate": 9.85261764344331e-06,
+      "loss": 0.2998,
+      "step": 482
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8904358131743887,
+      "learning_rate": 9.851763789250526e-06,
+      "loss": 0.3019,
+      "step": 483
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.884826983355429,
+      "learning_rate": 9.850907506024627e-06,
+      "loss": 0.322,
+      "step": 484
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0732848075010377,
+      "learning_rate": 9.850048794194309e-06,
+      "loss": 0.3069,
+      "step": 485
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8631327126132019,
+      "learning_rate": 9.849187654189486e-06,
+      "loss": 0.2904,
+      "step": 486
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8540044301533286,
+      "learning_rate": 9.848324086441287e-06,
+      "loss": 0.3093,
+      "step": 487
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0364126301381997,
+      "learning_rate": 9.847458091382057e-06,
+      "loss": 0.3103,
+      "step": 488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9485365070763563,
+      "learning_rate": 9.846589669445356e-06,
+      "loss": 0.3075,
+      "step": 489
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8584777962319268,
+      "learning_rate": 9.845718821065957e-06,
+      "loss": 0.2959,
+      "step": 490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9725953795738541,
+      "learning_rate": 9.844845546679852e-06,
+      "loss": 0.2965,
+      "step": 491
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8662201700850226,
+      "learning_rate": 9.843969846724247e-06,
+      "loss": 0.2919,
+      "step": 492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3582394111911273,
+      "learning_rate": 9.843091721637559e-06,
+      "loss": 0.3049,
+      "step": 493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8650887566489152,
+      "learning_rate": 9.842211171859418e-06,
+      "loss": 0.2768,
+      "step": 494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9886288435999849,
+      "learning_rate": 9.841328197830676e-06,
+      "loss": 0.3299,
+      "step": 495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9388394282386096,
+      "learning_rate": 9.840442799993389e-06,
+      "loss": 0.2911,
+      "step": 496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.259831967013768,
+      "learning_rate": 9.839554978790832e-06,
+      "loss": 0.3082,
+      "step": 497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9609202597825957,
+      "learning_rate": 9.838664734667496e-06,
+      "loss": 0.2949,
+      "step": 498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0648801710123923,
+      "learning_rate": 9.837772068069076e-06,
+      "loss": 0.3208,
+      "step": 499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9477395251417614,
+      "learning_rate": 9.83687697944249e-06,
+      "loss": 0.2889,
+      "step": 500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9624440901113315,
+      "learning_rate": 9.835979469235857e-06,
+      "loss": 0.3024,
+      "step": 501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9381445243278739,
+      "learning_rate": 9.835079537898523e-06,
+      "loss": 0.3179,
+      "step": 502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9224759984807771,
+      "learning_rate": 9.834177185881033e-06,
+      "loss": 0.308,
+      "step": 503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0800267552905547,
+      "learning_rate": 9.833272413635151e-06,
+      "loss": 0.312,
+      "step": 504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8773068155581771,
+      "learning_rate": 9.83236522161385e-06,
+      "loss": 0.2711,
+      "step": 505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9518720728299088,
+      "learning_rate": 9.831455610271315e-06,
+      "loss": 0.2831,
+      "step": 506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8884171714438645,
+      "learning_rate": 9.830543580062944e-06,
+      "loss": 0.2812,
+      "step": 507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1470782193176783,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.3134,
+      "step": 508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8725103065917922,
+      "learning_rate": 9.82871226487633e-06,
+      "loss": 0.2857,
+      "step": 509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9904072726584443,
+      "learning_rate": 9.827792980814934e-06,
+      "loss": 0.3074,
+      "step": 510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1469364627469665,
+      "learning_rate": 9.826871279721396e-06,
+      "loss": 0.3081,
+      "step": 511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7636030921005252,
+      "learning_rate": 9.825947162057163e-06,
+      "loss": 0.3003,
+      "step": 512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0281465874084414,
+      "learning_rate": 9.825020628284896e-06,
+      "loss": 0.3112,
+      "step": 513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8553056892343849,
+      "learning_rate": 9.82409167886846e-06,
+      "loss": 0.2863,
+      "step": 514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9975534160095696,
+      "learning_rate": 9.823160314272937e-06,
+      "loss": 0.2991,
+      "step": 515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0782810223611714,
+      "learning_rate": 9.822226534964613e-06,
+      "loss": 0.2894,
+      "step": 516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8848460437006963,
+      "learning_rate": 9.821290341410985e-06,
+      "loss": 0.2731,
+      "step": 517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0017147949159553,
+      "learning_rate": 9.820351734080754e-06,
+      "loss": 0.3204,
+      "step": 518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0669214482330378,
+      "learning_rate": 9.819410713443837e-06,
+      "loss": 0.2989,
+      "step": 519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9631547885989498,
+      "learning_rate": 9.818467279971356e-06,
+      "loss": 0.2954,
+      "step": 520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.882952923212777,
+      "learning_rate": 9.817521434135637e-06,
+      "loss": 0.3128,
+      "step": 521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9768413848602449,
+      "learning_rate": 9.81657317641022e-06,
+      "loss": 0.271,
+      "step": 522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0020760828909483,
+      "learning_rate": 9.81562250726985e-06,
+      "loss": 0.3041,
+      "step": 523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9135231162804728,
+      "learning_rate": 9.814669427190477e-06,
+      "loss": 0.3045,
+      "step": 524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1774868746422986,
+      "learning_rate": 9.813713936649261e-06,
+      "loss": 0.3172,
+      "step": 525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0484621367811402,
+      "learning_rate": 9.81275603612457e-06,
+      "loss": 0.3176,
+      "step": 526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9114473955847042,
+      "learning_rate": 9.811795726095971e-06,
+      "loss": 0.3129,
+      "step": 527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8981537188779963,
+      "learning_rate": 9.810833007044247e-06,
+      "loss": 0.3111,
+      "step": 528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9759011880539448,
+      "learning_rate": 9.809867879451384e-06,
+      "loss": 0.3011,
+      "step": 529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8052896939529108,
+      "learning_rate": 9.808900343800568e-06,
+      "loss": 0.2815,
+      "step": 530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8787689902655803,
+      "learning_rate": 9.807930400576199e-06,
+      "loss": 0.2933,
+      "step": 531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6453539329518752,
+      "learning_rate": 9.80695805026388e-06,
+      "loss": 0.3021,
+      "step": 532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.121002758472313,
+      "learning_rate": 9.805983293350413e-06,
+      "loss": 0.2905,
+      "step": 533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0758930236601458,
+      "learning_rate": 9.80500613032381e-06,
+      "loss": 0.2716,
+      "step": 534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9724800036918171,
+      "learning_rate": 9.80402656167329e-06,
+      "loss": 0.3013,
+      "step": 535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0125368244958404,
+      "learning_rate": 9.803044587889273e-06,
+      "loss": 0.2937,
+      "step": 536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0481100683194176,
+      "learning_rate": 9.802060209463382e-06,
+      "loss": 0.2742,
+      "step": 537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8703136252909133,
+      "learning_rate": 9.801073426888447e-06,
+      "loss": 0.2964,
+      "step": 538
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9239807966202227,
+      "learning_rate": 9.800084240658502e-06,
+      "loss": 0.3042,
+      "step": 539
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9925109422509448,
+      "learning_rate": 9.799092651268778e-06,
+      "loss": 0.3075,
+      "step": 540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0161329147102809,
+      "learning_rate": 9.798098659215719e-06,
+      "loss": 0.3232,
+      "step": 541
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.046265948741348,
+      "learning_rate": 9.79710226499696e-06,
+      "loss": 0.3246,
+      "step": 542
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4215390069758445,
+      "learning_rate": 9.796103469111352e-06,
+      "loss": 0.265,
+      "step": 543
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8339579025100371,
+      "learning_rate": 9.795102272058937e-06,
+      "loss": 0.3156,
+      "step": 544
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9764614590089565,
+      "learning_rate": 9.794098674340966e-06,
+      "loss": 0.3054,
+      "step": 545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9935612091419888,
+      "learning_rate": 9.79309267645989e-06,
+      "loss": 0.2903,
+      "step": 546
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.170913773336367,
+      "learning_rate": 9.79208427891936e-06,
+      "loss": 0.2941,
+      "step": 547
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8365415684098837,
+      "learning_rate": 9.791073482224229e-06,
+      "loss": 0.3102,
+      "step": 548
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0832622307310933,
+      "learning_rate": 9.790060286880556e-06,
+      "loss": 0.2954,
+      "step": 549
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0423024877253086,
+      "learning_rate": 9.789044693395593e-06,
+      "loss": 0.3199,
+      "step": 550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8769318598310154,
+      "learning_rate": 9.788026702277796e-06,
+      "loss": 0.2956,
+      "step": 551
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9477871443147515,
+      "learning_rate": 9.787006314036824e-06,
+      "loss": 0.273,
+      "step": 552
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.877756554141733,
+      "learning_rate": 9.785983529183533e-06,
+      "loss": 0.291,
+      "step": 553
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0172888893635066,
+      "learning_rate": 9.784958348229978e-06,
+      "loss": 0.307,
+      "step": 554
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3667181767171672,
+      "learning_rate": 9.783930771689419e-06,
+      "loss": 0.3097,
+      "step": 555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.964902749389785,
+      "learning_rate": 9.78290080007631e-06,
+      "loss": 0.2933,
+      "step": 556
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8763483712190122,
+      "learning_rate": 9.781868433906303e-06,
+      "loss": 0.3175,
+      "step": 557
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0546514197073489,
+      "learning_rate": 9.780833673696255e-06,
+      "loss": 0.3007,
+      "step": 558
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1023796451685313,
+      "learning_rate": 9.779796519964217e-06,
+      "loss": 0.3155,
+      "step": 559
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0535296587935272,
+      "learning_rate": 9.778756973229442e-06,
+      "loss": 0.3195,
+      "step": 560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.967123340394309,
+      "learning_rate": 9.777715034012373e-06,
+      "loss": 0.3031,
+      "step": 561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0440413330741956,
+      "learning_rate": 9.776670702834662e-06,
+      "loss": 0.309,
+      "step": 562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9052509665253674,
+      "learning_rate": 9.775623980219149e-06,
+      "loss": 0.333,
+      "step": 563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.973279935686728,
+      "learning_rate": 9.774574866689878e-06,
+      "loss": 0.3155,
+      "step": 564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1048302619894568,
+      "learning_rate": 9.773523362772084e-06,
+      "loss": 0.311,
+      "step": 565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0332644493195144,
+      "learning_rate": 9.772469468992206e-06,
+      "loss": 0.3143,
+      "step": 566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.976712671603609,
+      "learning_rate": 9.771413185877872e-06,
+      "loss": 0.34,
+      "step": 567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9181770722710126,
+      "learning_rate": 9.77035451395791e-06,
+      "loss": 0.2985,
+      "step": 568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9682326576581887,
+      "learning_rate": 9.769293453762344e-06,
+      "loss": 0.3136,
+      "step": 569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9136731116071773,
+      "learning_rate": 9.768230005822394e-06,
+      "loss": 0.2964,
+      "step": 570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8404271108757277,
+      "learning_rate": 9.767164170670476e-06,
+      "loss": 0.2902,
+      "step": 571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.923397886196809,
+      "learning_rate": 9.766095948840195e-06,
+      "loss": 0.3013,
+      "step": 572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9322805817338072,
+      "learning_rate": 9.76502534086636e-06,
+      "loss": 0.3403,
+      "step": 573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9075126296750671,
+      "learning_rate": 9.76395234728497e-06,
+      "loss": 0.2818,
+      "step": 574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8574775032645842,
+      "learning_rate": 9.762876968633218e-06,
+      "loss": 0.3108,
+      "step": 575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7369860634276577,
+      "learning_rate": 9.761799205449492e-06,
+      "loss": 0.3228,
+      "step": 576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9593302982192452,
+      "learning_rate": 9.760719058273371e-06,
+      "loss": 0.3032,
+      "step": 577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8922726988489721,
+      "learning_rate": 9.759636527645633e-06,
+      "loss": 0.3055,
+      "step": 578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0006484141754068,
+      "learning_rate": 9.758551614108247e-06,
+      "loss": 0.3008,
+      "step": 579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8041568833969451,
+      "learning_rate": 9.757464318204374e-06,
+      "loss": 0.2817,
+      "step": 580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9613395051807275,
+      "learning_rate": 9.756374640478363e-06,
+      "loss": 0.2805,
+      "step": 581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8485459884609585,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.3114,
+      "step": 582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8549259656777752,
+      "learning_rate": 9.754188141743326e-06,
+      "loss": 0.3089,
+      "step": 583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2355100330032374,
+      "learning_rate": 9.753091321828964e-06,
+      "loss": 0.3053,
+      "step": 584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0191645217094323,
+      "learning_rate": 9.751992122281809e-06,
+      "loss": 0.3284,
+      "step": 585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9707766118947296,
+      "learning_rate": 9.750890543652169e-06,
+      "loss": 0.2889,
+      "step": 586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.950710545968049,
+      "learning_rate": 9.749786586491557e-06,
+      "loss": 0.3119,
+      "step": 587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8655791219723055,
+      "learning_rate": 9.74868025135266e-06,
+      "loss": 0.2904,
+      "step": 588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0568062319972356,
+      "learning_rate": 9.74757153878937e-06,
+      "loss": 0.2906,
+      "step": 589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.946021853259604,
+      "learning_rate": 9.74646044935676e-06,
+      "loss": 0.3106,
+      "step": 590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9728983075774222,
+      "learning_rate": 9.7453469836111e-06,
+      "loss": 0.3005,
+      "step": 591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9050443488045111,
+      "learning_rate": 9.744231142109842e-06,
+      "loss": 0.2902,
+      "step": 592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8223666615587953,
+      "learning_rate": 9.743112925411633e-06,
+      "loss": 0.2843,
+      "step": 593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8805870113770274,
+      "learning_rate": 9.741992334076309e-06,
+      "loss": 0.2838,
+      "step": 594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8893081830529975,
+      "learning_rate": 9.740869368664894e-06,
+      "loss": 0.2895,
+      "step": 595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1083381172582616,
+      "learning_rate": 9.739744029739596e-06,
+      "loss": 0.3198,
+      "step": 596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8252944308864874,
+      "learning_rate": 9.738616317863819e-06,
+      "loss": 0.2739,
+      "step": 597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.910346876442311,
+      "learning_rate": 9.737486233602149e-06,
+      "loss": 0.3122,
+      "step": 598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0623813164181861,
+      "learning_rate": 9.736353777520365e-06,
+      "loss": 0.3357,
+      "step": 599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9368389317545601,
+      "learning_rate": 9.735218950185428e-06,
+      "loss": 0.2797,
+      "step": 600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9057332370596921,
+      "learning_rate": 9.734081752165491e-06,
+      "loss": 0.2793,
+      "step": 601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9330021133139659,
+      "learning_rate": 9.73294218402989e-06,
+      "loss": 0.2855,
+      "step": 602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.93406625705914,
+      "learning_rate": 9.731800246349149e-06,
+      "loss": 0.306,
+      "step": 603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9866297850121047,
+      "learning_rate": 9.730655939694979e-06,
+      "loss": 0.2969,
+      "step": 604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.855797406975609,
+      "learning_rate": 9.729509264640276e-06,
+      "loss": 0.2707,
+      "step": 605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9259359822020421,
+      "learning_rate": 9.728360221759125e-06,
+      "loss": 0.318,
+      "step": 606
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9843062495753839,
+      "learning_rate": 9.72720881162679e-06,
+      "loss": 0.3113,
+      "step": 607
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8712872772165575,
+      "learning_rate": 9.726055034819726e-06,
+      "loss": 0.3088,
+      "step": 608
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1652673424479871,
+      "learning_rate": 9.72489889191557e-06,
+      "loss": 0.2933,
+      "step": 609
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.889763928298685,
+      "learning_rate": 9.723740383493146e-06,
+      "loss": 0.324,
+      "step": 610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9693066295018677,
+      "learning_rate": 9.722579510132458e-06,
+      "loss": 0.2773,
+      "step": 611
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.94352356096985,
+      "learning_rate": 9.721416272414698e-06,
+      "loss": 0.2799,
+      "step": 612
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8395430964353808,
+      "learning_rate": 9.720250670922242e-06,
+      "loss": 0.297,
+      "step": 613
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8271406673369176,
+      "learning_rate": 9.719082706238645e-06,
+      "loss": 0.2792,
+      "step": 614
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8063478691053482,
+      "learning_rate": 9.717912378948651e-06,
+      "loss": 0.2851,
+      "step": 615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0093377519106843,
+      "learning_rate": 9.716739689638179e-06,
+      "loss": 0.2784,
+      "step": 616
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.927212703260063,
+      "learning_rate": 9.715564638894342e-06,
+      "loss": 0.28,
+      "step": 617
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8267541782109383,
+      "learning_rate": 9.714387227305422e-06,
+      "loss": 0.2546,
+      "step": 618
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1296527040194584,
+      "learning_rate": 9.713207455460893e-06,
+      "loss": 0.2875,
+      "step": 619
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9313213458871407,
+      "learning_rate": 9.712025323951406e-06,
+      "loss": 0.2751,
+      "step": 620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9853105279648214,
+      "learning_rate": 9.710840833368796e-06,
+      "loss": 0.308,
+      "step": 621
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9156716250569875,
+      "learning_rate": 9.709653984306077e-06,
+      "loss": 0.3261,
+      "step": 622
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8431599297556726,
+      "learning_rate": 9.708464777357444e-06,
+      "loss": 0.2898,
+      "step": 623
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.132237394681003,
+      "learning_rate": 9.70727321311827e-06,
+      "loss": 0.3104,
+      "step": 624
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9423691406507311,
+      "learning_rate": 9.706079292185119e-06,
+      "loss": 0.3006,
+      "step": 625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0983579145598148,
+      "learning_rate": 9.70488301515572e-06,
+      "loss": 0.2989,
+      "step": 626
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8998177096940331,
+      "learning_rate": 9.70368438262899e-06,
+      "loss": 0.3192,
+      "step": 627
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1386240140069672,
+      "learning_rate": 9.702483395205023e-06,
+      "loss": 0.286,
+      "step": 628
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8142790230193068,
+      "learning_rate": 9.701280053485096e-06,
+      "loss": 0.297,
+      "step": 629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9023270716354581,
+      "learning_rate": 9.700074358071658e-06,
+      "loss": 0.2858,
+      "step": 630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9922476143996078,
+      "learning_rate": 9.698866309568341e-06,
+      "loss": 0.3095,
+      "step": 631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9149521395449867,
+      "learning_rate": 9.697655908579958e-06,
+      "loss": 0.3079,
+      "step": 632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1280932468584632,
+      "learning_rate": 9.696443155712488e-06,
+      "loss": 0.3153,
+      "step": 633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7896000543471028,
+      "learning_rate": 9.695228051573098e-06,
+      "loss": 0.3108,
+      "step": 634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.220428219049545,
+      "learning_rate": 9.694010596770132e-06,
+      "loss": 0.2845,
+      "step": 635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.818710503381198,
+      "learning_rate": 9.692790791913107e-06,
+      "loss": 0.3112,
+      "step": 636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9070209417219086,
+      "learning_rate": 9.691568637612714e-06,
+      "loss": 0.2953,
+      "step": 637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9449885462757757,
+      "learning_rate": 9.69034413448083e-06,
+      "loss": 0.3218,
+      "step": 638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9576673965270387,
+      "learning_rate": 9.689117283130498e-06,
+      "loss": 0.3229,
+      "step": 639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6672732575758267,
+      "learning_rate": 9.68788808417594e-06,
+      "loss": 0.3137,
+      "step": 640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8519965507184188,
+      "learning_rate": 9.686656538232558e-06,
+      "loss": 0.2986,
+      "step": 641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9676119904326532,
+      "learning_rate": 9.68542264591692e-06,
+      "loss": 0.2844,
+      "step": 642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8002113864844775,
+      "learning_rate": 9.684186407846774e-06,
+      "loss": 0.2906,
+      "step": 643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2119730303401748,
+      "learning_rate": 9.682947824641045e-06,
+      "loss": 0.3023,
+      "step": 644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8901389650239466,
+      "learning_rate": 9.681706896919829e-06,
+      "loss": 0.2953,
+      "step": 645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9649623312043676,
+      "learning_rate": 9.680463625304393e-06,
+      "loss": 0.2828,
+      "step": 646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9129312793284708,
+      "learning_rate": 9.679218010417182e-06,
+      "loss": 0.2928,
+      "step": 647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0238088155099039,
+      "learning_rate": 9.677970052881811e-06,
+      "loss": 0.2916,
+      "step": 648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.866038206274319,
+      "learning_rate": 9.676719753323074e-06,
+      "loss": 0.2915,
+      "step": 649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8778102942627326,
+      "learning_rate": 9.675467112366925e-06,
+      "loss": 0.2876,
+      "step": 650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9881881137281102,
+      "learning_rate": 9.674212130640506e-06,
+      "loss": 0.2965,
+      "step": 651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.759277122849023,
+      "learning_rate": 9.672954808772118e-06,
+      "loss": 0.2818,
+      "step": 652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9948402891617754,
+      "learning_rate": 9.67169514739124e-06,
+      "loss": 0.3181,
+      "step": 653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8895369812295857,
+      "learning_rate": 9.670433147128522e-06,
+      "loss": 0.2703,
+      "step": 654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8317094516473278,
+      "learning_rate": 9.669168808615781e-06,
+      "loss": 0.3081,
+      "step": 655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.950061324081722,
+      "learning_rate": 9.667902132486009e-06,
+      "loss": 0.322,
+      "step": 656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1894689143161226,
+      "learning_rate": 9.666633119373368e-06,
+      "loss": 0.3162,
+      "step": 657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9560706799199761,
+      "learning_rate": 9.665361769913187e-06,
+      "loss": 0.3121,
+      "step": 658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7948346227386711,
+      "learning_rate": 9.664088084741969e-06,
+      "loss": 0.2991,
+      "step": 659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.862651134085211,
+      "learning_rate": 9.66281206449738e-06,
+      "loss": 0.2879,
+      "step": 660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9193310293015627,
+      "learning_rate": 9.661533709818261e-06,
+      "loss": 0.2849,
+      "step": 661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9330421809709497,
+      "learning_rate": 9.660253021344618e-06,
+      "loss": 0.2802,
+      "step": 662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9459597821561858,
+      "learning_rate": 9.658969999717631e-06,
+      "loss": 0.2996,
+      "step": 663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9544040307813255,
+      "learning_rate": 9.65768464557964e-06,
+      "loss": 0.3171,
+      "step": 664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8873610370227537,
+      "learning_rate": 9.656396959574161e-06,
+      "loss": 0.3015,
+      "step": 665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8293776088091516,
+      "learning_rate": 9.65510694234587e-06,
+      "loss": 0.2816,
+      "step": 666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.88308222135162,
+      "learning_rate": 9.653814594540614e-06,
+      "loss": 0.2908,
+      "step": 667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7999958407905239,
+      "learning_rate": 9.652519916805406e-06,
+      "loss": 0.3032,
+      "step": 668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9032115114358377,
+      "learning_rate": 9.651222909788428e-06,
+      "loss": 0.2686,
+      "step": 669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9176528437006064,
+      "learning_rate": 9.649923574139023e-06,
+      "loss": 0.3307,
+      "step": 670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.844595868096892,
+      "learning_rate": 9.648621910507704e-06,
+      "loss": 0.3237,
+      "step": 671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8585065215342441,
+      "learning_rate": 9.64731791954615e-06,
+      "loss": 0.2893,
+      "step": 672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9496696670397138,
+      "learning_rate": 9.6460116019072e-06,
+      "loss": 0.2753,
+      "step": 673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0415257078416809,
+      "learning_rate": 9.644702958244865e-06,
+      "loss": 0.3085,
+      "step": 674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9863671526811983,
+      "learning_rate": 9.643391989214312e-06,
+      "loss": 0.2811,
+      "step": 675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0446323658993957,
+      "learning_rate": 9.642078695471882e-06,
+      "loss": 0.2959,
+      "step": 676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8513157160058349,
+      "learning_rate": 9.640763077675072e-06,
+      "loss": 0.2854,
+      "step": 677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9477176596113701,
+      "learning_rate": 9.639445136482549e-06,
+      "loss": 0.2857,
+      "step": 678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9771156868298045,
+      "learning_rate": 9.638124872554134e-06,
+      "loss": 0.3329,
+      "step": 679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9164110642896016,
+      "learning_rate": 9.636802286550817e-06,
+      "loss": 0.2983,
+      "step": 680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9024826688924966,
+      "learning_rate": 9.635477379134756e-06,
+      "loss": 0.2858,
+      "step": 681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0052359824398576,
+      "learning_rate": 9.634150150969259e-06,
+      "loss": 0.3005,
+      "step": 682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7974233616887504,
+      "learning_rate": 9.632820602718806e-06,
+      "loss": 0.2783,
+      "step": 683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8329901959242912,
+      "learning_rate": 9.631488735049033e-06,
+      "loss": 0.2863,
+      "step": 684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9367452659262151,
+      "learning_rate": 9.630154548626737e-06,
+      "loss": 0.3371,
+      "step": 685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9022618990635041,
+      "learning_rate": 9.628818044119884e-06,
+      "loss": 0.2773,
+      "step": 686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8804690230410054,
+      "learning_rate": 9.627479222197587e-06,
+      "loss": 0.302,
+      "step": 687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8390542210976147,
+      "learning_rate": 9.62613808353013e-06,
+      "loss": 0.3183,
+      "step": 688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9873940902078265,
+      "learning_rate": 9.624794628788954e-06,
+      "loss": 0.3005,
+      "step": 689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0365239854805255,
+      "learning_rate": 9.623448858646658e-06,
+      "loss": 0.3029,
+      "step": 690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0258518869904039,
+      "learning_rate": 9.622100773777002e-06,
+      "loss": 0.3141,
+      "step": 691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8108340378353379,
+      "learning_rate": 9.620750374854901e-06,
+      "loss": 0.2857,
+      "step": 692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8630521885962367,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.3472,
+      "step": 693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7868594147794321,
+      "learning_rate": 9.618042637558838e-06,
+      "loss": 0.2817,
+      "step": 694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9077661955692031,
+      "learning_rate": 9.6166853005405e-06,
+      "loss": 0.3054,
+      "step": 695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8999993518591055,
+      "learning_rate": 9.615325652180976e-06,
+      "loss": 0.3117,
+      "step": 696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0509183216696258,
+      "learning_rate": 9.613963693160968e-06,
+      "loss": 0.2945,
+      "step": 697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.95836945889705,
+      "learning_rate": 9.612599424162344e-06,
+      "loss": 0.2937,
+      "step": 698
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9234555148978588,
+      "learning_rate": 9.611232845868124e-06,
+      "loss": 0.2865,
+      "step": 699
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8753483111488297,
+      "learning_rate": 9.609863958962483e-06,
+      "loss": 0.3219,
+      "step": 700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8081746095513056,
+      "learning_rate": 9.608492764130752e-06,
+      "loss": 0.2939,
+      "step": 701
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9207292707288537,
+      "learning_rate": 9.607119262059426e-06,
+      "loss": 0.296,
+      "step": 702
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7812644951644561,
+      "learning_rate": 9.60574345343614e-06,
+      "loss": 0.2625,
+      "step": 703
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9468602491980164,
+      "learning_rate": 9.604365338949697e-06,
+      "loss": 0.3013,
+      "step": 704
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.947059896279348,
+      "learning_rate": 9.602984919290049e-06,
+      "loss": 0.293,
+      "step": 705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.826502737587624,
+      "learning_rate": 9.601602195148299e-06,
+      "loss": 0.2643,
+      "step": 706
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8919635611198461,
+      "learning_rate": 9.60021716721671e-06,
+      "loss": 0.3165,
+      "step": 707
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8055951917951212,
+      "learning_rate": 9.598829836188694e-06,
+      "loss": 0.294,
+      "step": 708
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.771723143987034,
+      "learning_rate": 9.59744020275882e-06,
+      "loss": 0.2851,
+      "step": 709
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8957481700588616,
+      "learning_rate": 9.596048267622806e-06,
+      "loss": 0.3044,
+      "step": 710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9236132410869837,
+      "learning_rate": 9.594654031477522e-06,
+      "loss": 0.2999,
+      "step": 711
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8902966867809415,
+      "learning_rate": 9.593257495020993e-06,
+      "loss": 0.2978,
+      "step": 712
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8764602008824675,
+      "learning_rate": 9.591858658952396e-06,
+      "loss": 0.2752,
+      "step": 713
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0386851329448878,
+      "learning_rate": 9.590457523972055e-06,
+      "loss": 0.2871,
+      "step": 714
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8220291917345657,
+      "learning_rate": 9.58905409078145e-06,
+      "loss": 0.2816,
+      "step": 715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8962975277017898,
+      "learning_rate": 9.587648360083205e-06,
+      "loss": 0.3074,
+      "step": 716
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7641058626161074,
+      "learning_rate": 9.586240332581101e-06,
+      "loss": 0.2814,
+      "step": 717
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9065961729219398,
+      "learning_rate": 9.584830008980068e-06,
+      "loss": 0.3061,
+      "step": 718
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9268868156532888,
+      "learning_rate": 9.58341738998618e-06,
+      "loss": 0.3064,
+      "step": 719
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8276543330322088,
+      "learning_rate": 9.582002476306668e-06,
+      "loss": 0.2835,
+      "step": 720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8566470288678341,
+      "learning_rate": 9.580585268649904e-06,
+      "loss": 0.2947,
+      "step": 721
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7816808212185605,
+      "learning_rate": 9.579165767725416e-06,
+      "loss": 0.265,
+      "step": 722
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2815140404823824,
+      "learning_rate": 9.577743974243875e-06,
+      "loss": 0.2863,
+      "step": 723
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8961117268975057,
+      "learning_rate": 9.576319888917099e-06,
+      "loss": 0.2941,
+      "step": 724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8311446875520786,
+      "learning_rate": 9.574893512458057e-06,
+      "loss": 0.2731,
+      "step": 725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9217774839156472,
+      "learning_rate": 9.573464845580864e-06,
+      "loss": 0.2982,
+      "step": 726
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.382631793541333,
+      "learning_rate": 9.572033889000783e-06,
+      "loss": 0.2888,
+      "step": 727
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8055070039890403,
+      "learning_rate": 9.570600643434217e-06,
+      "loss": 0.3048,
+      "step": 728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9053991240354922,
+      "learning_rate": 9.569165109598725e-06,
+      "loss": 0.3069,
+      "step": 729
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.812840353450397,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.2784,
+      "step": 730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.994332597233364,
+      "learning_rate": 9.5662871799969e-06,
+      "loss": 0.2835,
+      "step": 731
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9027150945530164,
+      "learning_rate": 9.5648447856714e-06,
+      "loss": 0.3054,
+      "step": 732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9246932807272121,
+      "learning_rate": 9.563400105958638e-06,
+      "loss": 0.2963,
+      "step": 733
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.868785686528086,
+      "learning_rate": 9.561953141581895e-06,
+      "loss": 0.3009,
+      "step": 734
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.386900609138038,
+      "learning_rate": 9.560503893265589e-06,
+      "loss": 0.2889,
+      "step": 735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8627433618153931,
+      "learning_rate": 9.559052361735288e-06,
+      "loss": 0.2517,
+      "step": 736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8270106246020236,
+      "learning_rate": 9.557598547717702e-06,
+      "loss": 0.2928,
+      "step": 737
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8460535943090868,
+      "learning_rate": 9.55614245194068e-06,
+      "loss": 0.278,
+      "step": 738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1629114812098298,
+      "learning_rate": 9.554684075133215e-06,
+      "loss": 0.3129,
+      "step": 739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8749492014985071,
+      "learning_rate": 9.553223418025443e-06,
+      "loss": 0.3246,
+      "step": 740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.846833773574965,
+      "learning_rate": 9.551760481348644e-06,
+      "loss": 0.2816,
+      "step": 741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9115728614276448,
+      "learning_rate": 9.550295265835235e-06,
+      "loss": 0.3169,
+      "step": 742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9012912645202439,
+      "learning_rate": 9.548827772218772e-06,
+      "loss": 0.2692,
+      "step": 743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1016464847749186,
+      "learning_rate": 9.547358001233959e-06,
+      "loss": 0.3046,
+      "step": 744
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8675482567699598,
+      "learning_rate": 9.545885953616636e-06,
+      "loss": 0.2934,
+      "step": 745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8392136341929028,
+      "learning_rate": 9.544411630103782e-06,
+      "loss": 0.2826,
+      "step": 746
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.832551575312471,
+      "learning_rate": 9.542935031433515e-06,
+      "loss": 0.2862,
+      "step": 747
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1072435928053113,
+      "learning_rate": 9.541456158345094e-06,
+      "loss": 0.2858,
+      "step": 748
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8964473221797549,
+      "learning_rate": 9.539975011578918e-06,
+      "loss": 0.3083,
+      "step": 749
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8533476040520513,
+      "learning_rate": 9.538491591876522e-06,
+      "loss": 0.3031,
+      "step": 750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.815297306886387,
+      "learning_rate": 9.537005899980576e-06,
+      "loss": 0.2681,
+      "step": 751
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0340992986105655,
+      "learning_rate": 9.535517936634895e-06,
+      "loss": 0.2929,
+      "step": 752
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8762757559683118,
+      "learning_rate": 9.534027702584425e-06,
+      "loss": 0.2813,
+      "step": 753
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9421108573335424,
+      "learning_rate": 9.53253519857525e-06,
+      "loss": 0.2991,
+      "step": 754
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.93153220785374,
+      "learning_rate": 9.531040425354595e-06,
+      "loss": 0.2908,
+      "step": 755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8899012104473397,
+      "learning_rate": 9.529543383670814e-06,
+      "loss": 0.3025,
+      "step": 756
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9557365407459275,
+      "learning_rate": 9.5280440742734e-06,
+      "loss": 0.3218,
+      "step": 757
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8836501227260621,
+      "learning_rate": 9.526542497912984e-06,
+      "loss": 0.2946,
+      "step": 758
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9817280471929559,
+      "learning_rate": 9.52503865534133e-06,
+      "loss": 0.2937,
+      "step": 759
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8830283018119094,
+      "learning_rate": 9.523532547311332e-06,
+      "loss": 0.3065,
+      "step": 760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8547516485281283,
+      "learning_rate": 9.522024174577025e-06,
+      "loss": 0.3058,
+      "step": 761
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9310786222048866,
+      "learning_rate": 9.520513537893574e-06,
+      "loss": 0.2864,
+      "step": 762
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0304945369344525,
+      "learning_rate": 9.51900063801728e-06,
+      "loss": 0.3345,
+      "step": 763
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1850223793393164,
+      "learning_rate": 9.517485475705574e-06,
+      "loss": 0.3246,
+      "step": 764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8640504507677242,
+      "learning_rate": 9.515968051717022e-06,
+      "loss": 0.2797,
+      "step": 765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.869084111916035,
+      "learning_rate": 9.514448366811322e-06,
+      "loss": 0.2838,
+      "step": 766
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9093853808189806,
+      "learning_rate": 9.512926421749305e-06,
+      "loss": 0.2892,
+      "step": 767
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9739521427294343,
+      "learning_rate": 9.511402217292927e-06,
+      "loss": 0.3006,
+      "step": 768
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9232466856850685,
+      "learning_rate": 9.509875754205284e-06,
+      "loss": 0.2919,
+      "step": 769
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.85588203731113,
+      "learning_rate": 9.508347033250597e-06,
+      "loss": 0.2905,
+      "step": 770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8745743332021642,
+      "learning_rate": 9.506816055194223e-06,
+      "loss": 0.2911,
+      "step": 771
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.835473066400697,
+      "learning_rate": 9.505282820802642e-06,
+      "loss": 0.3186,
+      "step": 772
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8427523172217285,
+      "learning_rate": 9.503747330843468e-06,
+      "loss": 0.2843,
+      "step": 773
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9500553640628205,
+      "learning_rate": 9.502209586085444e-06,
+      "loss": 0.3155,
+      "step": 774
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8490024678002577,
+      "learning_rate": 9.50066958729844e-06,
+      "loss": 0.306,
+      "step": 775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8529436037835251,
+      "learning_rate": 9.499127335253456e-06,
+      "loss": 0.2919,
+      "step": 776
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9776410977803515,
+      "learning_rate": 9.497582830722617e-06,
+      "loss": 0.2887,
+      "step": 777
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8326172933341096,
+      "learning_rate": 9.496036074479184e-06,
+      "loss": 0.3071,
+      "step": 778
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8607797446328205,
+      "learning_rate": 9.494487067297538e-06,
+      "loss": 0.2882,
+      "step": 779
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8223517285520908,
+      "learning_rate": 9.492935809953185e-06,
+      "loss": 0.2817,
+      "step": 780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8889380101757274,
+      "learning_rate": 9.491382303222765e-06,
+      "loss": 0.3256,
+      "step": 781
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9425346283242363,
+      "learning_rate": 9.48982654788404e-06,
+      "loss": 0.2771,
+      "step": 782
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8748677552027798,
+      "learning_rate": 9.488268544715897e-06,
+      "loss": 0.2999,
+      "step": 783
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8003037015727706,
+      "learning_rate": 9.486708294498348e-06,
+      "loss": 0.2971,
+      "step": 784
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8095276680253128,
+      "learning_rate": 9.485145798012536e-06,
+      "loss": 0.2909,
+      "step": 785
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8624529608768572,
+      "learning_rate": 9.48358105604072e-06,
+      "loss": 0.2744,
+      "step": 786
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1412198401157028,
+      "learning_rate": 9.482014069366288e-06,
+      "loss": 0.2708,
+      "step": 787
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9101694949436693,
+      "learning_rate": 9.480444838773753e-06,
+      "loss": 0.2901,
+      "step": 788
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8811110008656867,
+      "learning_rate": 9.478873365048748e-06,
+      "loss": 0.2936,
+      "step": 789
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4554141283565372,
+      "learning_rate": 9.47729964897803e-06,
+      "loss": 0.3089,
+      "step": 790
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.514712507779276,
+      "learning_rate": 9.475723691349482e-06,
+      "loss": 0.3138,
+      "step": 791
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8574162267200894,
+      "learning_rate": 9.474145492952102e-06,
+      "loss": 0.2936,
+      "step": 792
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8569157255520903,
+      "learning_rate": 9.472565054576017e-06,
+      "loss": 0.3177,
+      "step": 793
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9821106685453473,
+      "learning_rate": 9.470982377012475e-06,
+      "loss": 0.3035,
+      "step": 794
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1199242252153978,
+      "learning_rate": 9.469397461053838e-06,
+      "loss": 0.2883,
+      "step": 795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8580907423407629,
+      "learning_rate": 9.467810307493595e-06,
+      "loss": 0.2975,
+      "step": 796
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8375440607848597,
+      "learning_rate": 9.466220917126354e-06,
+      "loss": 0.2941,
+      "step": 797
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8468336542456532,
+      "learning_rate": 9.464629290747844e-06,
+      "loss": 0.3253,
+      "step": 798
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.879595109823932,
+      "learning_rate": 9.463035429154909e-06,
+      "loss": 0.3197,
+      "step": 799
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8635724151945764,
+      "learning_rate": 9.461439333145516e-06,
+      "loss": 0.298,
+      "step": 800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8496217879997817,
+      "learning_rate": 9.459841003518752e-06,
+      "loss": 0.2958,
+      "step": 801
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9276593770028081,
+      "learning_rate": 9.458240441074818e-06,
+      "loss": 0.3085,
+      "step": 802
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8267069588839523,
+      "learning_rate": 9.456637646615035e-06,
+      "loss": 0.2919,
+      "step": 803
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.827125855770537,
+      "learning_rate": 9.45503262094184e-06,
+      "loss": 0.2888,
+      "step": 804
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0265267923686927,
+      "learning_rate": 9.45342536485879e-06,
+      "loss": 0.3087,
+      "step": 805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8866287270030091,
+      "learning_rate": 9.451815879170556e-06,
+      "loss": 0.3234,
+      "step": 806
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8295159739509618,
+      "learning_rate": 9.450204164682928e-06,
+      "loss": 0.29,
+      "step": 807
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9703818540674096,
+      "learning_rate": 9.448590222202808e-06,
+      "loss": 0.3104,
+      "step": 808
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8601549274070033,
+      "learning_rate": 9.446974052538216e-06,
+      "loss": 0.3084,
+      "step": 809
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9423700887755752,
+      "learning_rate": 9.445355656498284e-06,
+      "loss": 0.3048,
+      "step": 810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9817564521280995,
+      "learning_rate": 9.44373503489327e-06,
+      "loss": 0.3078,
+      "step": 811
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7769804086677088,
+      "learning_rate": 9.442112188534526e-06,
+      "loss": 0.2355,
+      "step": 812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.881848437694567,
+      "learning_rate": 9.440487118234536e-06,
+      "loss": 0.2874,
+      "step": 813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8617868018432439,
+      "learning_rate": 9.438859824806889e-06,
+      "loss": 0.2782,
+      "step": 814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9436202330361685,
+      "learning_rate": 9.43723030906629e-06,
+      "loss": 0.3191,
+      "step": 815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9754255598331768,
+      "learning_rate": 9.435598571828552e-06,
+      "loss": 0.2988,
+      "step": 816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.006718843643427,
+      "learning_rate": 9.433964613910608e-06,
+      "loss": 0.3144,
+      "step": 817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7993367523878032,
+      "learning_rate": 9.432328436130493e-06,
+      "loss": 0.2814,
+      "step": 818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8129920568084246,
+      "learning_rate": 9.430690039307364e-06,
+      "loss": 0.2724,
+      "step": 819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8445597813148775,
+      "learning_rate": 9.429049424261478e-06,
+      "loss": 0.302,
+      "step": 820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8403820212793752,
+      "learning_rate": 9.427406591814213e-06,
+      "loss": 0.2957,
+      "step": 821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0549137539759934,
+      "learning_rate": 9.425761542788049e-06,
+      "loss": 0.2919,
+      "step": 822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9119646739844718,
+      "learning_rate": 9.42411427800658e-06,
+      "loss": 0.2895,
+      "step": 823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8498197868809308,
+      "learning_rate": 9.422464798294509e-06,
+      "loss": 0.3103,
+      "step": 824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1306017918115772,
+      "learning_rate": 9.420813104477647e-06,
+      "loss": 0.2892,
+      "step": 825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8404172049983765,
+      "learning_rate": 9.419159197382913e-06,
+      "loss": 0.2899,
+      "step": 826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9417816756340168,
+      "learning_rate": 9.417503077838334e-06,
+      "loss": 0.2891,
+      "step": 827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1083078366884342,
+      "learning_rate": 9.415844746673047e-06,
+      "loss": 0.3075,
+      "step": 828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.788700631608647,
+      "learning_rate": 9.414184204717294e-06,
+      "loss": 0.3024,
+      "step": 829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8667490511634363,
+      "learning_rate": 9.412521452802425e-06,
+      "loss": 0.3192,
+      "step": 830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1510014429541953,
+      "learning_rate": 9.410856491760896e-06,
+      "loss": 0.2789,
+      "step": 831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9954647866014833,
+      "learning_rate": 9.409189322426267e-06,
+      "loss": 0.3096,
+      "step": 832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.879293310588163,
+      "learning_rate": 9.40751994563321e-06,
+      "loss": 0.3098,
+      "step": 833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7936275652231053,
+      "learning_rate": 9.40584836221749e-06,
+      "loss": 0.2932,
+      "step": 834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8756176556636531,
+      "learning_rate": 9.404174573015992e-06,
+      "loss": 0.3135,
+      "step": 835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8794873668643259,
+      "learning_rate": 9.402498578866695e-06,
+      "loss": 0.2944,
+      "step": 836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9363131272865827,
+      "learning_rate": 9.400820380608684e-06,
+      "loss": 0.2835,
+      "step": 837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8178980342092357,
+      "learning_rate": 9.399139979082148e-06,
+      "loss": 0.28,
+      "step": 838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7911781136432933,
+      "learning_rate": 9.39745737512838e-06,
+      "loss": 0.2861,
+      "step": 839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9408977244855932,
+      "learning_rate": 9.395772569589774e-06,
+      "loss": 0.3016,
+      "step": 840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0066970511655928,
+      "learning_rate": 9.394085563309827e-06,
+      "loss": 0.2959,
+      "step": 841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1391486004581102,
+      "learning_rate": 9.39239635713314e-06,
+      "loss": 0.2975,
+      "step": 842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9133076562062654,
+      "learning_rate": 9.390704951905412e-06,
+      "loss": 0.297,
+      "step": 843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4350617948612325,
+      "learning_rate": 9.389011348473442e-06,
+      "loss": 0.2932,
+      "step": 844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8637651524495079,
+      "learning_rate": 9.387315547685134e-06,
+      "loss": 0.3021,
+      "step": 845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8206195901938079,
+      "learning_rate": 9.38561755038949e-06,
+      "loss": 0.2783,
+      "step": 846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9241366288222048,
+      "learning_rate": 9.38391735743661e-06,
+      "loss": 0.2939,
+      "step": 847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.007855998115931,
+      "learning_rate": 9.382214969677697e-06,
+      "loss": 0.288,
+      "step": 848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9479551755252988,
+      "learning_rate": 9.380510387965047e-06,
+      "loss": 0.2948,
+      "step": 849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8535875952247718,
+      "learning_rate": 9.37880361315206e-06,
+      "loss": 0.2778,
+      "step": 850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8996679630642753,
+      "learning_rate": 9.377094646093234e-06,
+      "loss": 0.2773,
+      "step": 851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9251984270552472,
+      "learning_rate": 9.375383487644161e-06,
+      "loss": 0.3002,
+      "step": 852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7395122150650661,
+      "learning_rate": 9.37367013866153e-06,
+      "loss": 0.294,
+      "step": 853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0472099321350454,
+      "learning_rate": 9.37195460000313e-06,
+      "loss": 0.3006,
+      "step": 854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8642069120573578,
+      "learning_rate": 9.370236872527845e-06,
+      "loss": 0.2769,
+      "step": 855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8559694174735498,
+      "learning_rate": 9.368516957095652e-06,
+      "loss": 0.3093,
+      "step": 856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8612321727748472,
+      "learning_rate": 9.366794854567627e-06,
+      "loss": 0.2884,
+      "step": 857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0863386432240258,
+      "learning_rate": 9.365070565805941e-06,
+      "loss": 0.3104,
+      "step": 858
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.112951063205447,
+      "learning_rate": 9.363344091673858e-06,
+      "loss": 0.2678,
+      "step": 859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8823645836067258,
+      "learning_rate": 9.361615433035736e-06,
+      "loss": 0.2811,
+      "step": 860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9235285374199348,
+      "learning_rate": 9.359884590757027e-06,
+      "loss": 0.2632,
+      "step": 861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7983614848143838,
+      "learning_rate": 9.358151565704275e-06,
+      "loss": 0.2834,
+      "step": 862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.856827149851681,
+      "learning_rate": 9.356416358745119e-06,
+      "loss": 0.3141,
+      "step": 863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8905458911925971,
+      "learning_rate": 9.35467897074829e-06,
+      "loss": 0.2798,
+      "step": 864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9698675393488211,
+      "learning_rate": 9.352939402583611e-06,
+      "loss": 0.2908,
+      "step": 865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8188670357459928,
+      "learning_rate": 9.351197655121995e-06,
+      "loss": 0.2753,
+      "step": 866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9106485894367246,
+      "learning_rate": 9.349453729235447e-06,
+      "loss": 0.2879,
+      "step": 867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.208090277741334,
+      "learning_rate": 9.347707625797062e-06,
+      "loss": 0.2863,
+      "step": 868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0117640779174069,
+      "learning_rate": 9.345959345681028e-06,
+      "loss": 0.3034,
+      "step": 869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8141283293195836,
+      "learning_rate": 9.34420888976262e-06,
+      "loss": 0.286,
+      "step": 870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8372364039035375,
+      "learning_rate": 9.342456258918201e-06,
+      "loss": 0.3039,
+      "step": 871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8739777615316039,
+      "learning_rate": 9.340701454025224e-06,
+      "loss": 0.2898,
+      "step": 872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8239691094721886,
+      "learning_rate": 9.338944475962236e-06,
+      "loss": 0.2787,
+      "step": 873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9987895656065597,
+      "learning_rate": 9.337185325608864e-06,
+      "loss": 0.303,
+      "step": 874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0804814286975546,
+      "learning_rate": 9.335424003845825e-06,
+      "loss": 0.2768,
+      "step": 875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1319364473857263,
+      "learning_rate": 9.333660511554926e-06,
+      "loss": 0.2848,
+      "step": 876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8658988272911661,
+      "learning_rate": 9.331894849619056e-06,
+      "loss": 0.3088,
+      "step": 877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9438749832023814,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.321,
+      "step": 878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.840665540305858,
+      "learning_rate": 9.328357020349405e-06,
+      "loss": 0.2874,
+      "step": 879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8271554155733746,
+      "learning_rate": 9.326584854786835e-06,
+      "loss": 0.2789,
+      "step": 880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2021579753945852,
+      "learning_rate": 9.32481052312172e-06,
+      "loss": 0.2908,
+      "step": 881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7888163203165288,
+      "learning_rate": 9.323034026242378e-06,
+      "loss": 0.2763,
+      "step": 882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8522269043982951,
+      "learning_rate": 9.32125536503821e-06,
+      "loss": 0.3002,
+      "step": 883
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8588408449227071,
+      "learning_rate": 9.319474540399704e-06,
+      "loss": 0.2735,
+      "step": 884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.133892243896553,
+      "learning_rate": 9.317691553218427e-06,
+      "loss": 0.2919,
+      "step": 885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8791057698335092,
+      "learning_rate": 9.315906404387032e-06,
+      "loss": 0.2878,
+      "step": 886
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.861489171633895,
+      "learning_rate": 9.314119094799251e-06,
+      "loss": 0.2983,
+      "step": 887
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7934938771798711,
+      "learning_rate": 9.312329625349903e-06,
+      "loss": 0.2684,
+      "step": 888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9624171927688504,
+      "learning_rate": 9.310537996934882e-06,
+      "loss": 0.251,
+      "step": 889
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2584251579775378,
+      "learning_rate": 9.308744210451167e-06,
+      "loss": 0.3022,
+      "step": 890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8271332801415037,
+      "learning_rate": 9.306948266796815e-06,
+      "loss": 0.2911,
+      "step": 891
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8907893777014692,
+      "learning_rate": 9.305150166870966e-06,
+      "loss": 0.2964,
+      "step": 892
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9171560623290392,
+      "learning_rate": 9.303349911573838e-06,
+      "loss": 0.2842,
+      "step": 893
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0288184481953773,
+      "learning_rate": 9.301547501806725e-06,
+      "loss": 0.2969,
+      "step": 894
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8927724341856474,
+      "learning_rate": 9.299742938472007e-06,
+      "loss": 0.287,
+      "step": 895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.03188444587739,
+      "learning_rate": 9.297936222473134e-06,
+      "loss": 0.2688,
+      "step": 896
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1869682135140314,
+      "learning_rate": 9.29612735471464e-06,
+      "loss": 0.2971,
+      "step": 897
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9343284898125006,
+      "learning_rate": 9.294316336102132e-06,
+      "loss": 0.2977,
+      "step": 898
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7907442557908259,
+      "learning_rate": 9.292503167542295e-06,
+      "loss": 0.2864,
+      "step": 899
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8441807065173209,
+      "learning_rate": 9.290687849942893e-06,
+      "loss": 0.2768,
+      "step": 900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8237539795796035,
+      "learning_rate": 9.288870384212762e-06,
+      "loss": 0.3128,
+      "step": 901
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9176402634288503,
+      "learning_rate": 9.287050771261818e-06,
+      "loss": 0.3154,
+      "step": 902
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9272810254414143,
+      "learning_rate": 9.285229012001047e-06,
+      "loss": 0.3221,
+      "step": 903
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8738736773008821,
+      "learning_rate": 9.283405107342512e-06,
+      "loss": 0.2682,
+      "step": 904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.836619249517823,
+      "learning_rate": 9.281579058199351e-06,
+      "loss": 0.2871,
+      "step": 905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.809822976198804,
+      "learning_rate": 9.279750865485772e-06,
+      "loss": 0.286,
+      "step": 906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.842454673927645,
+      "learning_rate": 9.277920530117064e-06,
+      "loss": 0.2897,
+      "step": 907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8260756568612249,
+      "learning_rate": 9.276088053009578e-06,
+      "loss": 0.2565,
+      "step": 908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0653636269892635,
+      "learning_rate": 9.274253435080745e-06,
+      "loss": 0.257,
+      "step": 909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9247413603541864,
+      "learning_rate": 9.272416677249067e-06,
+      "loss": 0.2967,
+      "step": 910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7652095895514494,
+      "learning_rate": 9.270577780434114e-06,
+      "loss": 0.2926,
+      "step": 911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7928325652663306,
+      "learning_rate": 9.268736745556528e-06,
+      "loss": 0.2935,
+      "step": 912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8414198271734192,
+      "learning_rate": 9.266893573538023e-06,
+      "loss": 0.2885,
+      "step": 913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8841818833281408,
+      "learning_rate": 9.265048265301385e-06,
+      "loss": 0.2981,
+      "step": 914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8621901295739697,
+      "learning_rate": 9.263200821770462e-06,
+      "loss": 0.2838,
+      "step": 915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9155666472422365,
+      "learning_rate": 9.261351243870178e-06,
+      "loss": 0.304,
+      "step": 916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9057242515208089,
+      "learning_rate": 9.259499532526523e-06,
+      "loss": 0.2947,
+      "step": 917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.805898206380013,
+      "learning_rate": 9.257645688666557e-06,
+      "loss": 0.2901,
+      "step": 918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8971062737458916,
+      "learning_rate": 9.255789713218403e-06,
+      "loss": 0.296,
+      "step": 919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5294850608289163,
+      "learning_rate": 9.253931607111257e-06,
+      "loss": 0.307,
+      "step": 920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9455596605923509,
+      "learning_rate": 9.252071371275378e-06,
+      "loss": 0.2859,
+      "step": 921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8105685785411889,
+      "learning_rate": 9.25020900664209e-06,
+      "loss": 0.2736,
+      "step": 922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7547930060677561,
+      "learning_rate": 9.248344514143786e-06,
+      "loss": 0.2723,
+      "step": 923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7749129888115395,
+      "learning_rate": 9.246477894713925e-06,
+      "loss": 0.278,
+      "step": 924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8443104476088372,
+      "learning_rate": 9.244609149287029e-06,
+      "loss": 0.292,
+      "step": 925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8390719511123592,
+      "learning_rate": 9.242738278798682e-06,
+      "loss": 0.2805,
+      "step": 926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8588557599051301,
+      "learning_rate": 9.240865284185536e-06,
+      "loss": 0.2728,
+      "step": 927
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8889623649672607,
+      "learning_rate": 9.238990166385304e-06,
+      "loss": 0.3227,
+      "step": 928
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9470865592584471,
+      "learning_rate": 9.237112926336764e-06,
+      "loss": 0.2837,
+      "step": 929
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6457691478333254,
+      "learning_rate": 9.235233564979756e-06,
+      "loss": 0.3164,
+      "step": 930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8730017954417162,
+      "learning_rate": 9.233352083255177e-06,
+      "loss": 0.2887,
+      "step": 931
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8716854761486493,
+      "learning_rate": 9.231468482104994e-06,
+      "loss": 0.2905,
+      "step": 932
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0048511989159452,
+      "learning_rate": 9.229582762472232e-06,
+      "loss": 0.2909,
+      "step": 933
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8639730517068004,
+      "learning_rate": 9.227694925300972e-06,
+      "loss": 0.2935,
+      "step": 934
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8334136598245802,
+      "learning_rate": 9.22580497153636e-06,
+      "loss": 0.2777,
+      "step": 935
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.987469072968897,
+      "learning_rate": 9.223912902124601e-06,
+      "loss": 0.286,
+      "step": 936
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3335301355601916,
+      "learning_rate": 9.222018718012957e-06,
+      "loss": 0.294,
+      "step": 937
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9220187368338754,
+      "learning_rate": 9.220122420149753e-06,
+      "loss": 0.2707,
+      "step": 938
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9039730370393728,
+      "learning_rate": 9.218224009484367e-06,
+      "loss": 0.2883,
+      "step": 939
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9064246984508019,
+      "learning_rate": 9.216323486967239e-06,
+      "loss": 0.2867,
+      "step": 940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9059479645639341,
+      "learning_rate": 9.214420853549864e-06,
+      "loss": 0.3076,
+      "step": 941
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8888306189273443,
+      "learning_rate": 9.212516110184794e-06,
+      "loss": 0.2737,
+      "step": 942
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9161748117283569,
+      "learning_rate": 9.21060925782564e-06,
+      "loss": 0.277,
+      "step": 943
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9115634238559038,
+      "learning_rate": 9.208700297427063e-06,
+      "loss": 0.2789,
+      "step": 944
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7686916043297766,
+      "learning_rate": 9.206789229944786e-06,
+      "loss": 0.267,
+      "step": 945
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8951883722594175,
+      "learning_rate": 9.204876056335584e-06,
+      "loss": 0.2547,
+      "step": 946
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0024122268972289,
+      "learning_rate": 9.202960777557284e-06,
+      "loss": 0.2801,
+      "step": 947
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9022469909255946,
+      "learning_rate": 9.201043394568773e-06,
+      "loss": 0.2762,
+      "step": 948
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9733034405444227,
+      "learning_rate": 9.199123908329983e-06,
+      "loss": 0.2924,
+      "step": 949
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9244326099149883,
+      "learning_rate": 9.197202319801908e-06,
+      "loss": 0.2904,
+      "step": 950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0105693560329896,
+      "learning_rate": 9.195278629946588e-06,
+      "loss": 0.301,
+      "step": 951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7746234682867039,
+      "learning_rate": 9.193352839727122e-06,
+      "loss": 0.2858,
+      "step": 952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0079645508826307,
+      "learning_rate": 9.191424950107648e-06,
+      "loss": 0.2974,
+      "step": 953
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9805152294259997,
+      "learning_rate": 9.18949496205337e-06,
+      "loss": 0.2732,
+      "step": 954
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8195889030065746,
+      "learning_rate": 9.187562876530528e-06,
+      "loss": 0.2981,
+      "step": 955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9695615865285837,
+      "learning_rate": 9.18562869450643e-06,
+      "loss": 0.3109,
+      "step": 956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7454310122346272,
+      "learning_rate": 9.183692416949415e-06,
+      "loss": 0.2714,
+      "step": 957
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8699230757839601,
+      "learning_rate": 9.181754044828882e-06,
+      "loss": 0.293,
+      "step": 958
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8901418360035612,
+      "learning_rate": 9.179813579115276e-06,
+      "loss": 0.2955,
+      "step": 959
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8608628701516943,
+      "learning_rate": 9.17787102078009e-06,
+      "loss": 0.2962,
+      "step": 960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7602337176888495,
+      "learning_rate": 9.175926370795865e-06,
+      "loss": 0.2697,
+      "step": 961
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8449209493589672,
+      "learning_rate": 9.17397963013619e-06,
+      "loss": 0.2982,
+      "step": 962
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8891599246884826,
+      "learning_rate": 9.172030799775698e-06,
+      "loss": 0.3059,
+      "step": 963
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9788305793057432,
+      "learning_rate": 9.170079880690072e-06,
+      "loss": 0.2933,
+      "step": 964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8103577587985094,
+      "learning_rate": 9.168126873856035e-06,
+      "loss": 0.2798,
+      "step": 965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9103869960639215,
+      "learning_rate": 9.166171780251365e-06,
+      "loss": 0.2722,
+      "step": 966
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9462851974322191,
+      "learning_rate": 9.164214600854872e-06,
+      "loss": 0.2841,
+      "step": 967
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9555127768586833,
+      "learning_rate": 9.162255336646422e-06,
+      "loss": 0.3086,
+      "step": 968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8182879632710671,
+      "learning_rate": 9.160293988606917e-06,
+      "loss": 0.2994,
+      "step": 969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9292008762485428,
+      "learning_rate": 9.158330557718304e-06,
+      "loss": 0.2783,
+      "step": 970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9916483039052528,
+      "learning_rate": 9.156365044963575e-06,
+      "loss": 0.3017,
+      "step": 971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8838836882951397,
+      "learning_rate": 9.154397451326766e-06,
+      "loss": 0.2881,
+      "step": 972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8490184628272415,
+      "learning_rate": 9.152427777792947e-06,
+      "loss": 0.2875,
+      "step": 973
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8192366258441256,
+      "learning_rate": 9.150456025348236e-06,
+      "loss": 0.2963,
+      "step": 974
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.838280836209056,
+      "learning_rate": 9.14848219497979e-06,
+      "loss": 0.2958,
+      "step": 975
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0961167814796302,
+      "learning_rate": 9.146506287675805e-06,
+      "loss": 0.2815,
+      "step": 976
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.375022989014107,
+      "learning_rate": 9.14452830442552e-06,
+      "loss": 0.2823,
+      "step": 977
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7954546264936178,
+      "learning_rate": 9.142548246219212e-06,
+      "loss": 0.294,
+      "step": 978
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8734827877444283,
+      "learning_rate": 9.140566114048193e-06,
+      "loss": 0.2836,
+      "step": 979
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.033163355479958,
+      "learning_rate": 9.138581908904818e-06,
+      "loss": 0.2891,
+      "step": 980
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9135088467563042,
+      "learning_rate": 9.136595631782478e-06,
+      "loss": 0.2954,
+      "step": 981
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9798779332543218,
+      "learning_rate": 9.134607283675601e-06,
+      "loss": 0.2962,
+      "step": 982
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9664336962860643,
+      "learning_rate": 9.132616865579655e-06,
+      "loss": 0.3,
+      "step": 983
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8115093487712638,
+      "learning_rate": 9.13062437849114e-06,
+      "loss": 0.2742,
+      "step": 984
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7841301258334026,
+      "learning_rate": 9.128629823407594e-06,
+      "loss": 0.3014,
+      "step": 985
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6002127823398822,
+      "learning_rate": 9.12663320132759e-06,
+      "loss": 0.2855,
+      "step": 986
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9713283360927657,
+      "learning_rate": 9.124634513250736e-06,
+      "loss": 0.2914,
+      "step": 987
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8374461089399972,
+      "learning_rate": 9.122633760177674e-06,
+      "loss": 0.2983,
+      "step": 988
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9188668304361733,
+      "learning_rate": 9.120630943110078e-06,
+      "loss": 0.3115,
+      "step": 989
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0817317025761692,
+      "learning_rate": 9.118626063050661e-06,
+      "loss": 0.2959,
+      "step": 990
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8619428086454621,
+      "learning_rate": 9.116619121003165e-06,
+      "loss": 0.2935,
+      "step": 991
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.904498528553428,
+      "learning_rate": 9.114610117972362e-06,
+      "loss": 0.3078,
+      "step": 992
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8041674919968415,
+      "learning_rate": 9.112599054964058e-06,
+      "loss": 0.2822,
+      "step": 993
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8889369034755443,
+      "learning_rate": 9.110585932985095e-06,
+      "loss": 0.292,
+      "step": 994
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0561467318207969,
+      "learning_rate": 9.108570753043335e-06,
+      "loss": 0.313,
+      "step": 995
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0159916308656372,
+      "learning_rate": 9.106553516147683e-06,
+      "loss": 0.3024,
+      "step": 996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.830007720403381,
+      "learning_rate": 9.104534223308065e-06,
+      "loss": 0.2895,
+      "step": 997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8804184502104871,
+      "learning_rate": 9.102512875535439e-06,
+      "loss": 0.2557,
+      "step": 998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0398977461914503,
+      "learning_rate": 9.100489473841791e-06,
+      "loss": 0.2908,
+      "step": 999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7824993856074117,
+      "learning_rate": 9.098464019240137e-06,
+      "loss": 0.282,
+      "step": 1000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8159324924784345,
+      "learning_rate": 9.09643651274452e-06,
+      "loss": 0.2931,
+      "step": 1001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1172152633650416,
+      "learning_rate": 9.09440695537001e-06,
+      "loss": 0.2777,
+      "step": 1002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.103811373539539,
+      "learning_rate": 9.092375348132704e-06,
+      "loss": 0.3395,
+      "step": 1003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.86587502415063,
+      "learning_rate": 9.090341692049725e-06,
+      "loss": 0.2714,
+      "step": 1004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8885149982155749,
+      "learning_rate": 9.088305988139222e-06,
+      "loss": 0.2911,
+      "step": 1005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8932463894714175,
+      "learning_rate": 9.086268237420368e-06,
+      "loss": 0.3178,
+      "step": 1006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9338911625359034,
+      "learning_rate": 9.084228440913363e-06,
+      "loss": 0.3046,
+      "step": 1007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.997530517449528,
+      "learning_rate": 9.082186599639429e-06,
+      "loss": 0.3032,
+      "step": 1008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8169852836234923,
+      "learning_rate": 9.080142714620814e-06,
+      "loss": 0.2771,
+      "step": 1009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7732684123956964,
+      "learning_rate": 9.07809678688079e-06,
+      "loss": 0.2744,
+      "step": 1010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7476602652034093,
+      "learning_rate": 9.076048817443646e-06,
+      "loss": 0.2943,
+      "step": 1011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7710341701723532,
+      "learning_rate": 9.073998807334698e-06,
+      "loss": 0.2879,
+      "step": 1012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0359140334673587,
+      "learning_rate": 9.071946757580282e-06,
+      "loss": 0.2844,
+      "step": 1013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7977060070847706,
+      "learning_rate": 9.069892669207757e-06,
+      "loss": 0.2632,
+      "step": 1014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0037873275668951,
+      "learning_rate": 9.067836543245502e-06,
+      "loss": 0.3025,
+      "step": 1015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.247996871210174,
+      "learning_rate": 9.065778380722914e-06,
+      "loss": 0.281,
+      "step": 1016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8094315533230731,
+      "learning_rate": 9.06371818267041e-06,
+      "loss": 0.2926,
+      "step": 1017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9053074750771563,
+      "learning_rate": 9.06165595011943e-06,
+      "loss": 0.2926,
+      "step": 1018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8452997012762646,
+      "learning_rate": 9.059591684102429e-06,
+      "loss": 0.2847,
+      "step": 1019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9620077743321217,
+      "learning_rate": 9.057525385652877e-06,
+      "loss": 0.308,
+      "step": 1020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0193944422777836,
+      "learning_rate": 9.05545705580527e-06,
+      "loss": 0.2925,
+      "step": 1021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.937147454779043,
+      "learning_rate": 9.053386695595114e-06,
+      "loss": 0.2779,
+      "step": 1022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6094269441762086,
+      "learning_rate": 9.051314306058934e-06,
+      "loss": 0.2963,
+      "step": 1023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0494665729766557,
+      "learning_rate": 9.049239888234271e-06,
+      "loss": 0.2872,
+      "step": 1024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0196514581815597,
+      "learning_rate": 9.047163443159682e-06,
+      "loss": 0.3081,
+      "step": 1025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8863919369391932,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.285,
+      "step": 1026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8664742617392321,
+      "learning_rate": 9.043004475420022e-06,
+      "loss": 0.2955,
+      "step": 1027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9824909805555038,
+      "learning_rate": 9.040921954837139e-06,
+      "loss": 0.3114,
+      "step": 1028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9518123022620132,
+      "learning_rate": 9.038837411168697e-06,
+      "loss": 0.2881,
+      "step": 1029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8267584755937749,
+      "learning_rate": 9.036750845458323e-06,
+      "loss": 0.2875,
+      "step": 1030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8844559547080485,
+      "learning_rate": 9.034662258750653e-06,
+      "loss": 0.3008,
+      "step": 1031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8351947530168294,
+      "learning_rate": 9.032571652091341e-06,
+      "loss": 0.3062,
+      "step": 1032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0880652744785866,
+      "learning_rate": 9.030479026527048e-06,
+      "loss": 0.3076,
+      "step": 1033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8473102742901676,
+      "learning_rate": 9.028384383105444e-06,
+      "loss": 0.2933,
+      "step": 1034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8219507136186361,
+      "learning_rate": 9.02628772287521e-06,
+      "loss": 0.2803,
+      "step": 1035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8460712276742769,
+      "learning_rate": 9.02418904688604e-06,
+      "loss": 0.2743,
+      "step": 1036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8422823202910152,
+      "learning_rate": 9.022088356188633e-06,
+      "loss": 0.2775,
+      "step": 1037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8381047290718223,
+      "learning_rate": 9.019985651834703e-06,
+      "loss": 0.2845,
+      "step": 1038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9135194591238742,
+      "learning_rate": 9.017880934876965e-06,
+      "loss": 0.2796,
+      "step": 1039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9360856811500936,
+      "learning_rate": 9.015774206369144e-06,
+      "loss": 0.2774,
+      "step": 1040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8194040317548508,
+      "learning_rate": 9.013665467365972e-06,
+      "loss": 0.2828,
+      "step": 1041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7880795349709719,
+      "learning_rate": 9.01155471892319e-06,
+      "loss": 0.2889,
+      "step": 1042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.906734725191205,
+      "learning_rate": 9.009441962097543e-06,
+      "loss": 0.2962,
+      "step": 1043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9232825277848354,
+      "learning_rate": 9.007327197946782e-06,
+      "loss": 0.325,
+      "step": 1044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8010562422381178,
+      "learning_rate": 9.00521042752966e-06,
+      "loss": 0.2514,
+      "step": 1045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8236606984156546,
+      "learning_rate": 9.00309165190594e-06,
+      "loss": 0.2825,
+      "step": 1046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3474950556415726,
+      "learning_rate": 9.000970872136383e-06,
+      "loss": 0.2852,
+      "step": 1047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8688324967841917,
+      "learning_rate": 8.99884808928276e-06,
+      "loss": 0.312,
+      "step": 1048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8929479934354532,
+      "learning_rate": 8.99672330440784e-06,
+      "loss": 0.2944,
+      "step": 1049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7820459979927873,
+      "learning_rate": 8.994596518575393e-06,
+      "loss": 0.2714,
+      "step": 1050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7991506529059146,
+      "learning_rate": 8.992467732850194e-06,
+      "loss": 0.295,
+      "step": 1051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9909761072464127,
+      "learning_rate": 8.990336948298023e-06,
+      "loss": 0.2785,
+      "step": 1052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8725892095273134,
+      "learning_rate": 8.98820416598565e-06,
+      "loss": 0.2978,
+      "step": 1053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7927038504199657,
+      "learning_rate": 8.986069386980855e-06,
+      "loss": 0.2827,
+      "step": 1054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.83169859465051,
+      "learning_rate": 8.983932612352413e-06,
+      "loss": 0.2615,
+      "step": 1055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8830762786429017,
+      "learning_rate": 8.9817938431701e-06,
+      "loss": 0.3001,
+      "step": 1056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9074635863011504,
+      "learning_rate": 8.979653080504689e-06,
+      "loss": 0.2822,
+      "step": 1057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.850970029482275,
+      "learning_rate": 8.97751032542795e-06,
+      "loss": 0.2982,
+      "step": 1058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8193835727884411,
+      "learning_rate": 8.975365579012656e-06,
+      "loss": 0.2767,
+      "step": 1059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8154449349969328,
+      "learning_rate": 8.97321884233257e-06,
+      "loss": 0.2898,
+      "step": 1060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7426376466974205,
+      "learning_rate": 8.971070116462457e-06,
+      "loss": 0.2626,
+      "step": 1061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8045561986467159,
+      "learning_rate": 8.968919402478076e-06,
+      "loss": 0.2853,
+      "step": 1062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9068189578607857,
+      "learning_rate": 8.966766701456177e-06,
+      "loss": 0.3003,
+      "step": 1063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9147589257612875,
+      "learning_rate": 8.964612014474512e-06,
+      "loss": 0.3095,
+      "step": 1064
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8563694530235656,
+      "learning_rate": 8.962455342611821e-06,
+      "loss": 0.3033,
+      "step": 1065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9138835527924555,
+      "learning_rate": 8.960296686947846e-06,
+      "loss": 0.2763,
+      "step": 1066
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8997101581078247,
+      "learning_rate": 8.958136048563311e-06,
+      "loss": 0.2794,
+      "step": 1067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8725391463053324,
+      "learning_rate": 8.955973428539943e-06,
+      "loss": 0.293,
+      "step": 1068
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9544842673638455,
+      "learning_rate": 8.953808827960455e-06,
+      "loss": 0.2913,
+      "step": 1069
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.870782802550556,
+      "learning_rate": 8.95164224790855e-06,
+      "loss": 0.3116,
+      "step": 1070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8809810155298883,
+      "learning_rate": 8.949473689468931e-06,
+      "loss": 0.2729,
+      "step": 1071
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0212388181582708,
+      "learning_rate": 8.947303153727283e-06,
+      "loss": 0.2853,
+      "step": 1072
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9317327931129179,
+      "learning_rate": 8.945130641770281e-06,
+      "loss": 0.2795,
+      "step": 1073
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0273489008721577,
+      "learning_rate": 8.942956154685596e-06,
+      "loss": 0.3005,
+      "step": 1074
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.090328731130555,
+      "learning_rate": 8.940779693561883e-06,
+      "loss": 0.2687,
+      "step": 1075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8656158369517427,
+      "learning_rate": 8.938601259488784e-06,
+      "loss": 0.3012,
+      "step": 1076
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8064437968876219,
+      "learning_rate": 8.936420853556935e-06,
+      "loss": 0.2925,
+      "step": 1077
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.808406775359356,
+      "learning_rate": 8.93423847685795e-06,
+      "loss": 0.2907,
+      "step": 1078
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8248430087662719,
+      "learning_rate": 8.932054130484439e-06,
+      "loss": 0.2812,
+      "step": 1079
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7588494007691389,
+      "learning_rate": 8.929867815529992e-06,
+      "loss": 0.2923,
+      "step": 1080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3126143266947312,
+      "learning_rate": 8.927679533089188e-06,
+      "loss": 0.3131,
+      "step": 1081
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9087660415119041,
+      "learning_rate": 8.925489284257587e-06,
+      "loss": 0.2884,
+      "step": 1082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7923694680863372,
+      "learning_rate": 8.923297070131738e-06,
+      "loss": 0.2874,
+      "step": 1083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8045460183530982,
+      "learning_rate": 8.921102891809171e-06,
+      "loss": 0.2692,
+      "step": 1084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.949242482663008,
+      "learning_rate": 8.9189067503884e-06,
+      "loss": 0.3074,
+      "step": 1085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.221354215533485,
+      "learning_rate": 8.916708646968924e-06,
+      "loss": 0.2891,
+      "step": 1086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8933277540289524,
+      "learning_rate": 8.91450858265122e-06,
+      "loss": 0.2924,
+      "step": 1087
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0916304805310337,
+      "learning_rate": 8.91230655853675e-06,
+      "loss": 0.2792,
+      "step": 1088
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7677412265836807,
+      "learning_rate": 8.910102575727957e-06,
+      "loss": 0.2764,
+      "step": 1089
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.175349613431111,
+      "learning_rate": 8.907896635328262e-06,
+      "loss": 0.3128,
+      "step": 1090
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8223607360245779,
+      "learning_rate": 8.90568873844207e-06,
+      "loss": 0.2982,
+      "step": 1091
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.827850922550598,
+      "learning_rate": 8.903478886174763e-06,
+      "loss": 0.3035,
+      "step": 1092
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8489941969737266,
+      "learning_rate": 8.901267079632703e-06,
+      "loss": 0.2751,
+      "step": 1093
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0134991501930277,
+      "learning_rate": 8.899053319923228e-06,
+      "loss": 0.2784,
+      "step": 1094
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8099475323840574,
+      "learning_rate": 8.896837608154656e-06,
+      "loss": 0.2898,
+      "step": 1095
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0074915289745932,
+      "learning_rate": 8.894619945436285e-06,
+      "loss": 0.293,
+      "step": 1096
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8114890597716427,
+      "learning_rate": 8.892400332878381e-06,
+      "loss": 0.2657,
+      "step": 1097
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.787695138588218,
+      "learning_rate": 8.890178771592198e-06,
+      "loss": 0.2719,
+      "step": 1098
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8364047298252915,
+      "learning_rate": 8.887955262689956e-06,
+      "loss": 0.2677,
+      "step": 1099
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8472662186294005,
+      "learning_rate": 8.885729807284855e-06,
+      "loss": 0.2986,
+      "step": 1100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9555865046293442,
+      "learning_rate": 8.883502406491067e-06,
+      "loss": 0.2845,
+      "step": 1101
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8373871841474976,
+      "learning_rate": 8.881273061423741e-06,
+      "loss": 0.2919,
+      "step": 1102
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8368786737352881,
+      "learning_rate": 8.879041773198996e-06,
+      "loss": 0.2836,
+      "step": 1103
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8679677530352499,
+      "learning_rate": 8.876808542933924e-06,
+      "loss": 0.312,
+      "step": 1104
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8501157797331976,
+      "learning_rate": 8.874573371746593e-06,
+      "loss": 0.2861,
+      "step": 1105
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9862631789176272,
+      "learning_rate": 8.87233626075604e-06,
+      "loss": 0.296,
+      "step": 1106
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7637120701657337,
+      "learning_rate": 8.87009721108227e-06,
+      "loss": 0.2792,
+      "step": 1107
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.023281173055104,
+      "learning_rate": 8.86785622384627e-06,
+      "loss": 0.2931,
+      "step": 1108
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8321063275026105,
+      "learning_rate": 8.865613300169982e-06,
+      "loss": 0.2737,
+      "step": 1109
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8792279173825345,
+      "learning_rate": 8.863368441176326e-06,
+      "loss": 0.2921,
+      "step": 1110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9320009062698632,
+      "learning_rate": 8.861121647989191e-06,
+      "loss": 0.2512,
+      "step": 1111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.826303064267481,
+      "learning_rate": 8.858872921733433e-06,
+      "loss": 0.302,
+      "step": 1112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9198607002714378,
+      "learning_rate": 8.856622263534875e-06,
+      "loss": 0.2831,
+      "step": 1113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8439721431395046,
+      "learning_rate": 8.854369674520306e-06,
+      "loss": 0.2754,
+      "step": 1114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8433783948731434,
+      "learning_rate": 8.852115155817488e-06,
+      "loss": 0.2728,
+      "step": 1115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5982458863993252,
+      "learning_rate": 8.84985870855514e-06,
+      "loss": 0.3151,
+      "step": 1116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9614956155356379,
+      "learning_rate": 8.847600333862956e-06,
+      "loss": 0.2956,
+      "step": 1117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9265677464179773,
+      "learning_rate": 8.845340032871584e-06,
+      "loss": 0.2961,
+      "step": 1118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1445808709206706,
+      "learning_rate": 8.843077806712648e-06,
+      "loss": 0.2971,
+      "step": 1119
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0448494481511088,
+      "learning_rate": 8.840813656518728e-06,
+      "loss": 0.307,
+      "step": 1120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.847687297637253,
+      "learning_rate": 8.838547583423372e-06,
+      "loss": 0.2863,
+      "step": 1121
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8610002929547346,
+      "learning_rate": 8.836279588561084e-06,
+      "loss": 0.2793,
+      "step": 1122
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8390680741475408,
+      "learning_rate": 8.834009673067337e-06,
+      "loss": 0.2856,
+      "step": 1123
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9221712809455795,
+      "learning_rate": 8.831737838078562e-06,
+      "loss": 0.2765,
+      "step": 1124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8674532403057623,
+      "learning_rate": 8.829464084732155e-06,
+      "loss": 0.2959,
+      "step": 1125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8310737644748386,
+      "learning_rate": 8.827188414166468e-06,
+      "loss": 0.2772,
+      "step": 1126
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7918844349956276,
+      "learning_rate": 8.824910827520813e-06,
+      "loss": 0.2634,
+      "step": 1127
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7751697191286269,
+      "learning_rate": 8.822631325935463e-06,
+      "loss": 0.2673,
+      "step": 1128
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8639354069820713,
+      "learning_rate": 8.820349910551651e-06,
+      "loss": 0.2744,
+      "step": 1129
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9072048913545605,
+      "learning_rate": 8.818066582511564e-06,
+      "loss": 0.2699,
+      "step": 1130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9003864669807441,
+      "learning_rate": 8.815781342958352e-06,
+      "loss": 0.2953,
+      "step": 1131
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7752917275196763,
+      "learning_rate": 8.813494193036115e-06,
+      "loss": 0.286,
+      "step": 1132
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9087877219475374,
+      "learning_rate": 8.811205133889917e-06,
+      "loss": 0.3035,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8412112344289798,
+      "learning_rate": 8.808914166665773e-06,
+      "loss": 0.2897,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8771459546372602,
+      "learning_rate": 8.806621292510655e-06,
+      "loss": 0.3115,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9518041712711581,
+      "learning_rate": 8.80432651257249e-06,
+      "loss": 0.2947,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8865264366350861,
+      "learning_rate": 8.802029828000157e-06,
+      "loss": 0.3034,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8723299264661493,
+      "learning_rate": 8.799731239943488e-06,
+      "loss": 0.2891,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.783228063036025,
+      "learning_rate": 8.797430749553275e-06,
+      "loss": 0.2607,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9669608679585704,
+      "learning_rate": 8.795128357981253e-06,
+      "loss": 0.3082,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8840227856698054,
+      "learning_rate": 8.792824066380116e-06,
+      "loss": 0.2881,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0925396805096328,
+      "learning_rate": 8.790517875903506e-06,
+      "loss": 0.2936,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9465148848189856,
+      "learning_rate": 8.788209787706014e-06,
+      "loss": 0.2722,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8858093622921612,
+      "learning_rate": 8.785899802943186e-06,
+      "loss": 0.2709,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8268969744308056,
+      "learning_rate": 8.783587922771514e-06,
+      "loss": 0.2781,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8667822715142004,
+      "learning_rate": 8.781274148348438e-06,
+      "loss": 0.2822,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7995015834917721,
+      "learning_rate": 8.778958480832349e-06,
+      "loss": 0.266,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9237046446197497,
+      "learning_rate": 8.776640921382585e-06,
+      "loss": 0.3091,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8277797017893856,
+      "learning_rate": 8.77432147115943e-06,
+      "loss": 0.2779,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9323170465248967,
+      "learning_rate": 8.772000131324117e-06,
+      "loss": 0.2683,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8988679766259542,
+      "learning_rate": 8.769676903038824e-06,
+      "loss": 0.2796,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.841289336882445,
+      "learning_rate": 8.767351787466674e-06,
+      "loss": 0.3033,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8723212119301512,
+      "learning_rate": 8.765024785771732e-06,
+      "loss": 0.271,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8667094695662202,
+      "learning_rate": 8.762695899119012e-06,
+      "loss": 0.3109,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9364367960965306,
+      "learning_rate": 8.760365128674473e-06,
+      "loss": 0.2621,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8637567662977274,
+      "learning_rate": 8.75803247560501e-06,
+      "loss": 0.2682,
+      "step": 1156
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7771228741020489,
+      "learning_rate": 8.755697941078465e-06,
+      "loss": 0.2775,
+      "step": 1157
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8030965913603096,
+      "learning_rate": 8.753361526263622e-06,
+      "loss": 0.2828,
+      "step": 1158
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1147441496834443,
+      "learning_rate": 8.751023232330208e-06,
+      "loss": 0.2702,
+      "step": 1159
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9944681485506074,
+      "learning_rate": 8.748683060448886e-06,
+      "loss": 0.3056,
+      "step": 1160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9085468143918304,
+      "learning_rate": 8.746341011791265e-06,
+      "loss": 0.2924,
+      "step": 1161
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7703033487574767,
+      "learning_rate": 8.743997087529886e-06,
+      "loss": 0.2847,
+      "step": 1162
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9919530242524427,
+      "learning_rate": 8.741651288838237e-06,
+      "loss": 0.2986,
+      "step": 1163
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9214446721397883,
+      "learning_rate": 8.739303616890739e-06,
+      "loss": 0.2923,
+      "step": 1164
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8842708401405422,
+      "learning_rate": 8.736954072862755e-06,
+      "loss": 0.2874,
+      "step": 1165
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8140416158763237,
+      "learning_rate": 8.734602657930582e-06,
+      "loss": 0.2898,
+      "step": 1166
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8875749646514202,
+      "learning_rate": 8.732249373271455e-06,
+      "loss": 0.3095,
+      "step": 1167
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8080814089769658,
+      "learning_rate": 8.729894220063542e-06,
+      "loss": 0.2908,
+      "step": 1168
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8524903864959033,
+      "learning_rate": 8.727537199485953e-06,
+      "loss": 0.2733,
+      "step": 1169
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9535370552374823,
+      "learning_rate": 8.725178312718727e-06,
+      "loss": 0.2845,
+      "step": 1170
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4529851838630312,
+      "learning_rate": 8.72281756094284e-06,
+      "loss": 0.301,
+      "step": 1171
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9456418776878173,
+      "learning_rate": 8.720454945340202e-06,
+      "loss": 0.2684,
+      "step": 1172
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9007570365744721,
+      "learning_rate": 8.718090467093654e-06,
+      "loss": 0.2955,
+      "step": 1173
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.555229104768037,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 0.2645,
+      "step": 1174
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8896888448299884,
+      "learning_rate": 8.713355927404863e-06,
+      "loss": 0.2768,
+      "step": 1175
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9962316510826698,
+      "learning_rate": 8.710985868332964e-06,
+      "loss": 0.3008,
+      "step": 1176
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8514370829514153,
+      "learning_rate": 8.708613951357842e-06,
+      "loss": 0.2814,
+      "step": 1177
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8016959240438326,
+      "learning_rate": 8.706240177667003e-06,
+      "loss": 0.2829,
+      "step": 1178
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7708335363564391,
+      "learning_rate": 8.703864548448868e-06,
+      "loss": 0.2935,
+      "step": 1179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8780509248628652,
+      "learning_rate": 8.701487064892797e-06,
+      "loss": 0.3088,
+      "step": 1180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8019629702107494,
+      "learning_rate": 8.699107728189078e-06,
+      "loss": 0.2759,
+      "step": 1181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9330239378391645,
+      "learning_rate": 8.696726539528924e-06,
+      "loss": 0.2986,
+      "step": 1182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9340577860044905,
+      "learning_rate": 8.694343500104474e-06,
+      "loss": 0.3016,
+      "step": 1183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8703563468942063,
+      "learning_rate": 8.691958611108795e-06,
+      "loss": 0.2985,
+      "step": 1184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8244920683720385,
+      "learning_rate": 8.689571873735884e-06,
+      "loss": 0.2792,
+      "step": 1185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8193852310052947,
+      "learning_rate": 8.687183289180659e-06,
+      "loss": 0.2812,
+      "step": 1186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7698533636763573,
+      "learning_rate": 8.684792858638962e-06,
+      "loss": 0.2886,
+      "step": 1187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8382537980073421,
+      "learning_rate": 8.682400583307562e-06,
+      "loss": 0.2877,
+      "step": 1188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8600744305701581,
+      "learning_rate": 8.68000646438415e-06,
+      "loss": 0.303,
+      "step": 1189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8247332584469913,
+      "learning_rate": 8.67761050306734e-06,
+      "loss": 0.2887,
+      "step": 1190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8319813061410001,
+      "learning_rate": 8.675212700556667e-06,
+      "loss": 0.2848,
+      "step": 1191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8716319961361565,
+      "learning_rate": 8.672813058052596e-06,
+      "loss": 0.277,
+      "step": 1192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.882880742647485,
+      "learning_rate": 8.670411576756502e-06,
+      "loss": 0.2965,
+      "step": 1193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8738899952632294,
+      "learning_rate": 8.668008257870684e-06,
+      "loss": 0.2915,
+      "step": 1194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7569491216471897,
+      "learning_rate": 8.665603102598365e-06,
+      "loss": 0.2863,
+      "step": 1195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0863596070708132,
+      "learning_rate": 8.663196112143685e-06,
+      "loss": 0.2696,
+      "step": 1196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1080331327324784,
+      "learning_rate": 8.660787287711703e-06,
+      "loss": 0.296,
+      "step": 1197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8446631609254158,
+      "learning_rate": 8.658376630508391e-06,
+      "loss": 0.2861,
+      "step": 1198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6281778925456396,
+      "learning_rate": 8.655964141740652e-06,
+      "loss": 0.2922,
+      "step": 1199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.151186331679405,
+      "learning_rate": 8.65354982261629e-06,
+      "loss": 0.2978,
+      "step": 1200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8366507427288172,
+      "learning_rate": 8.651133674344035e-06,
+      "loss": 0.3026,
+      "step": 1201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8188746391794702,
+      "learning_rate": 8.648715698133529e-06,
+      "loss": 0.2643,
+      "step": 1202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8037229978320742,
+      "learning_rate": 8.646295895195334e-06,
+      "loss": 0.2716,
+      "step": 1203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8359704650818838,
+      "learning_rate": 8.64387426674092e-06,
+      "loss": 0.2626,
+      "step": 1204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8883935445351888,
+      "learning_rate": 8.641450813982675e-06,
+      "loss": 0.2746,
+      "step": 1205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7976574242882727,
+      "learning_rate": 8.639025538133899e-06,
+      "loss": 0.2738,
+      "step": 1206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.861133019389586,
+      "learning_rate": 8.636598440408808e-06,
+      "loss": 0.2799,
+      "step": 1207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.860171037735001,
+      "learning_rate": 8.634169522022522e-06,
+      "loss": 0.3017,
+      "step": 1208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7595764243833507,
+      "learning_rate": 8.631738784191083e-06,
+      "loss": 0.2694,
+      "step": 1209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8301465598283961,
+      "learning_rate": 8.629306228131435e-06,
+      "loss": 0.2964,
+      "step": 1210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9148598500058235,
+      "learning_rate": 8.626871855061438e-06,
+      "loss": 0.2984,
+      "step": 1211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8736459359944557,
+      "learning_rate": 8.62443566619986e-06,
+      "loss": 0.2703,
+      "step": 1212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8056474151168246,
+      "learning_rate": 8.621997662766378e-06,
+      "loss": 0.2856,
+      "step": 1213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8808101695392239,
+      "learning_rate": 8.619557845981571e-06,
+      "loss": 0.2913,
+      "step": 1214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8026008397087347,
+      "learning_rate": 8.617116217066943e-06,
+      "loss": 0.2638,
+      "step": 1215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7553667769711315,
+      "learning_rate": 8.614672777244885e-06,
+      "loss": 0.2765,
+      "step": 1216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8027844997266683,
+      "learning_rate": 8.612227527738707e-06,
+      "loss": 0.2756,
+      "step": 1217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8291819409487662,
+      "learning_rate": 8.609780469772623e-06,
+      "loss": 0.3239,
+      "step": 1218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8851835445040951,
+      "learning_rate": 8.607331604571751e-06,
+      "loss": 0.3348,
+      "step": 1219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8570598233480861,
+      "learning_rate": 8.604880933362113e-06,
+      "loss": 0.3264,
+      "step": 1220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8123127570826785,
+      "learning_rate": 8.602428457370638e-06,
+      "loss": 0.2864,
+      "step": 1221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0758112582853112,
+      "learning_rate": 8.599974177825155e-06,
+      "loss": 0.3108,
+      "step": 1222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8679347284891222,
+      "learning_rate": 8.597518095954399e-06,
+      "loss": 0.2968,
+      "step": 1223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9664504711127342,
+      "learning_rate": 8.595060212988006e-06,
+      "loss": 0.3044,
+      "step": 1224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7685550976338744,
+      "learning_rate": 8.592600530156514e-06,
+      "loss": 0.2604,
+      "step": 1225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8121676334964893,
+      "learning_rate": 8.590139048691361e-06,
+      "loss": 0.3187,
+      "step": 1226
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.845705564318107,
+      "learning_rate": 8.587675769824889e-06,
+      "loss": 0.3341,
+      "step": 1227
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9846282729248195,
+      "learning_rate": 8.585210694790333e-06,
+      "loss": 0.2557,
+      "step": 1228
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.847258512846584,
+      "learning_rate": 8.582743824821835e-06,
+      "loss": 0.2847,
+      "step": 1229
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7753771939342485,
+      "learning_rate": 8.580275161154432e-06,
+      "loss": 0.2637,
+      "step": 1230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3932719379879508,
+      "learning_rate": 8.577804705024059e-06,
+      "loss": 0.2697,
+      "step": 1231
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8408573912935949,
+      "learning_rate": 8.575332457667546e-06,
+      "loss": 0.2752,
+      "step": 1232
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.891570192926385,
+      "learning_rate": 8.572858420322626e-06,
+      "loss": 0.292,
+      "step": 1233
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9454231411810442,
+      "learning_rate": 8.570382594227923e-06,
+      "loss": 0.2825,
+      "step": 1234
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7852450797557089,
+      "learning_rate": 8.567904980622961e-06,
+      "loss": 0.2707,
+      "step": 1235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.918931196190175,
+      "learning_rate": 8.56542558074815e-06,
+      "loss": 0.2828,
+      "step": 1236
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.831516318958459,
+      "learning_rate": 8.562944395844806e-06,
+      "loss": 0.2822,
+      "step": 1237
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8726535634662969,
+      "learning_rate": 8.56046142715513e-06,
+      "loss": 0.3159,
+      "step": 1238
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8963666498700373,
+      "learning_rate": 8.557976675922217e-06,
+      "loss": 0.2965,
+      "step": 1239
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7686204560552754,
+      "learning_rate": 8.555490143390061e-06,
+      "loss": 0.2633,
+      "step": 1240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8644905299794845,
+      "learning_rate": 8.55300183080354e-06,
+      "loss": 0.3009,
+      "step": 1241
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8185365261083403,
+      "learning_rate": 8.550511739408428e-06,
+      "loss": 0.2916,
+      "step": 1242
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8349921459531039,
+      "learning_rate": 8.548019870451391e-06,
+      "loss": 0.2517,
+      "step": 1243
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8240118656654619,
+      "learning_rate": 8.545526225179976e-06,
+      "loss": 0.2869,
+      "step": 1244
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7459678817377312,
+      "learning_rate": 8.543030804842629e-06,
+      "loss": 0.2533,
+      "step": 1245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7600592632781346,
+      "learning_rate": 8.540533610688679e-06,
+      "loss": 0.2867,
+      "step": 1246
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8600858542179503,
+      "learning_rate": 8.538034643968347e-06,
+      "loss": 0.3046,
+      "step": 1247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.851026848035097,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.2897,
+      "step": 1248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8903716065044361,
+      "learning_rate": 8.533031397833846e-06,
+      "loss": 0.2768,
+      "step": 1249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.881759797824775,
+      "learning_rate": 8.53052712092455e-06,
+      "loss": 0.3005,
+      "step": 1250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9241945499390929,
+      "learning_rate": 8.528021076458614e-06,
+      "loss": 0.3031,
+      "step": 1251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9065332194826717,
+      "learning_rate": 8.52551326569069e-06,
+      "loss": 0.287,
+      "step": 1252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8466978550963095,
+      "learning_rate": 8.523003689876312e-06,
+      "loss": 0.2766,
+      "step": 1253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7524468126750535,
+      "learning_rate": 8.520492350271895e-06,
+      "loss": 0.2941,
+      "step": 1254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9023541695663685,
+      "learning_rate": 8.517979248134744e-06,
+      "loss": 0.2754,
+      "step": 1255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8592502384921313,
+      "learning_rate": 8.51546438472304e-06,
+      "loss": 0.3024,
+      "step": 1256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7934449858112272,
+      "learning_rate": 8.512947761295847e-06,
+      "loss": 0.2556,
+      "step": 1257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0174762459292628,
+      "learning_rate": 8.510429379113114e-06,
+      "loss": 0.3072,
+      "step": 1258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.783954298628004,
+      "learning_rate": 8.507909239435665e-06,
+      "loss": 0.2859,
+      "step": 1259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9458839318154137,
+      "learning_rate": 8.50538734352521e-06,
+      "loss": 0.3139,
+      "step": 1260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8837747522501123,
+      "learning_rate": 8.502863692644332e-06,
+      "loss": 0.2856,
+      "step": 1261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.139535201985753,
+      "learning_rate": 8.500338288056496e-06,
+      "loss": 0.2978,
+      "step": 1262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8932934068446332,
+      "learning_rate": 8.497811131026046e-06,
+      "loss": 0.2804,
+      "step": 1263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.795069917671534,
+      "learning_rate": 8.495282222818202e-06,
+      "loss": 0.2668,
+      "step": 1264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9748509616943495,
+      "learning_rate": 8.492751564699058e-06,
+      "loss": 0.2838,
+      "step": 1265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7979574710868815,
+      "learning_rate": 8.490219157935589e-06,
+      "loss": 0.2978,
+      "step": 1266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8210793657112446,
+      "learning_rate": 8.487685003795643e-06,
+      "loss": 0.2918,
+      "step": 1267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7717574740174196,
+      "learning_rate": 8.485149103547943e-06,
+      "loss": 0.2774,
+      "step": 1268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8180759475217292,
+      "learning_rate": 8.482611458462084e-06,
+      "loss": 0.2851,
+      "step": 1269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8237607604272335,
+      "learning_rate": 8.480072069808541e-06,
+      "loss": 0.2634,
+      "step": 1270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7888409684440406,
+      "learning_rate": 8.477530938858654e-06,
+      "loss": 0.2523,
+      "step": 1271
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7922294571945844,
+      "learning_rate": 8.47498806688464e-06,
+      "loss": 0.2724,
+      "step": 1272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7939892523018727,
+      "learning_rate": 8.472443455159586e-06,
+      "loss": 0.2813,
+      "step": 1273
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8676353662229221,
+      "learning_rate": 8.469897104957454e-06,
+      "loss": 0.3035,
+      "step": 1274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8276948387075483,
+      "learning_rate": 8.467349017553067e-06,
+      "loss": 0.2938,
+      "step": 1275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8666917133026221,
+      "learning_rate": 8.464799194222129e-06,
+      "loss": 0.2769,
+      "step": 1276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7988336365442387,
+      "learning_rate": 8.462247636241206e-06,
+      "loss": 0.2886,
+      "step": 1277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8280872105702147,
+      "learning_rate": 8.459694344887732e-06,
+      "loss": 0.2815,
+      "step": 1278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8393156002039183,
+      "learning_rate": 8.457139321440014e-06,
+      "loss": 0.2829,
+      "step": 1279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7785306452015923,
+      "learning_rate": 8.454582567177223e-06,
+      "loss": 0.2783,
+      "step": 1280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8875097588266471,
+      "learning_rate": 8.452024083379396e-06,
+      "loss": 0.2742,
+      "step": 1281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9746731450723977,
+      "learning_rate": 8.449463871327434e-06,
+      "loss": 0.2711,
+      "step": 1282
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9526790053739405,
+      "learning_rate": 8.446901932303112e-06,
+      "loss": 0.2759,
+      "step": 1283
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8247188800052124,
+      "learning_rate": 8.444338267589057e-06,
+      "loss": 0.2998,
+      "step": 1284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9411128592844598,
+      "learning_rate": 8.44177287846877e-06,
+      "loss": 0.2821,
+      "step": 1285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8065009969591056,
+      "learning_rate": 8.439205766226613e-06,
+      "loss": 0.2673,
+      "step": 1286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7127429561320459,
+      "learning_rate": 8.436636932147805e-06,
+      "loss": 0.2746,
+      "step": 1287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6067593379570597,
+      "learning_rate": 8.434066377518437e-06,
+      "loss": 0.3054,
+      "step": 1288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7525375659886303,
+      "learning_rate": 8.431494103625452e-06,
+      "loss": 0.285,
+      "step": 1289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8516163536955835,
+      "learning_rate": 8.428920111756658e-06,
+      "loss": 0.3074,
+      "step": 1290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7976757987298863,
+      "learning_rate": 8.426344403200723e-06,
+      "loss": 0.2921,
+      "step": 1291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8391581387392031,
+      "learning_rate": 8.423766979247176e-06,
+      "loss": 0.3,
+      "step": 1292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7523199991968278,
+      "learning_rate": 8.421187841186402e-06,
+      "loss": 0.2791,
+      "step": 1293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0040212977975873,
+      "learning_rate": 8.418606990309645e-06,
+      "loss": 0.281,
+      "step": 1294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9958163689240821,
+      "learning_rate": 8.416024427909005e-06,
+      "loss": 0.283,
+      "step": 1295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8507196072314376,
+      "learning_rate": 8.413440155277444e-06,
+      "loss": 0.2764,
+      "step": 1296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7778099832508532,
+      "learning_rate": 8.410854173708773e-06,
+      "loss": 0.2734,
+      "step": 1297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7266603336836026,
+      "learning_rate": 8.408266484497664e-06,
+      "loss": 0.2654,
+      "step": 1298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7173935072469654,
+      "learning_rate": 8.405677088939644e-06,
+      "loss": 0.2721,
+      "step": 1299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8837426987888988,
+      "learning_rate": 8.403085988331092e-06,
+      "loss": 0.2792,
+      "step": 1300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7375811356152685,
+      "learning_rate": 8.400493183969238e-06,
+      "loss": 0.2811,
+      "step": 1301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8014380631196004,
+      "learning_rate": 8.397898677152173e-06,
+      "loss": 0.2794,
+      "step": 1302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9897518399795819,
+      "learning_rate": 8.395302469178832e-06,
+      "loss": 0.2659,
+      "step": 1303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1853812517243338,
+      "learning_rate": 8.392704561349006e-06,
+      "loss": 0.3036,
+      "step": 1304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8729935837069027,
+      "learning_rate": 8.390104954963338e-06,
+      "loss": 0.2821,
+      "step": 1305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.76140841073224,
+      "learning_rate": 8.387503651323317e-06,
+      "loss": 0.2681,
+      "step": 1306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7750990232639889,
+      "learning_rate": 8.384900651731287e-06,
+      "loss": 0.2924,
+      "step": 1307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.090282208325012,
+      "learning_rate": 8.382295957490435e-06,
+      "loss": 0.2819,
+      "step": 1308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8198798782173169,
+      "learning_rate": 8.379689569904802e-06,
+      "loss": 0.2797,
+      "step": 1309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8996241957532166,
+      "learning_rate": 8.377081490279278e-06,
+      "loss": 0.3165,
+      "step": 1310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7258460675163332,
+      "learning_rate": 8.37447171991959e-06,
+      "loss": 0.2772,
+      "step": 1311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8060455543787077,
+      "learning_rate": 8.371860260132323e-06,
+      "loss": 0.2885,
+      "step": 1312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8127141175360583,
+      "learning_rate": 8.369247112224901e-06,
+      "loss": 0.2754,
+      "step": 1313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7609509460256,
+      "learning_rate": 8.366632277505598e-06,
+      "loss": 0.2612,
+      "step": 1314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8268878660459441,
+      "learning_rate": 8.364015757283524e-06,
+      "loss": 0.3076,
+      "step": 1315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9833889355575077,
+      "learning_rate": 8.361397552868645e-06,
+      "loss": 0.303,
+      "step": 1316
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.787140956753501,
+      "learning_rate": 8.35877766557176e-06,
+      "loss": 0.2934,
+      "step": 1317
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8652751867497934,
+      "learning_rate": 8.356156096704516e-06,
+      "loss": 0.2828,
+      "step": 1318
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.77261869101778,
+      "learning_rate": 8.353532847579397e-06,
+      "loss": 0.2885,
+      "step": 1319
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0666943019818431,
+      "learning_rate": 8.350907919509734e-06,
+      "loss": 0.2792,
+      "step": 1320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1452211051261798,
+      "learning_rate": 8.348281313809697e-06,
+      "loss": 0.2776,
+      "step": 1321
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8653792379502637,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.3011,
+      "step": 1322
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8036603142512203,
+      "learning_rate": 8.343023074779368e-06,
+      "loss": 0.2841,
+      "step": 1323
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8033681284701747,
+      "learning_rate": 8.340391444081615e-06,
+      "loss": 0.2639,
+      "step": 1324
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7960569161017784,
+      "learning_rate": 8.337758141018551e-06,
+      "loss": 0.2797,
+      "step": 1325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9192851956476661,
+      "learning_rate": 8.335123166908544e-06,
+      "loss": 0.2874,
+      "step": 1326
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8192995953124015,
+      "learning_rate": 8.332486523070789e-06,
+      "loss": 0.2766,
+      "step": 1327
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9778904491433598,
+      "learning_rate": 8.329848210825322e-06,
+      "loss": 0.3059,
+      "step": 1328
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8135994713243163,
+      "learning_rate": 8.32720823149301e-06,
+      "loss": 0.2809,
+      "step": 1329
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8191410506420979,
+      "learning_rate": 8.324566586395563e-06,
+      "loss": 0.2864,
+      "step": 1330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9376418618158257,
+      "learning_rate": 8.321923276855514e-06,
+      "loss": 0.2949,
+      "step": 1331
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8215248745641547,
+      "learning_rate": 8.319278304196237e-06,
+      "loss": 0.3093,
+      "step": 1332
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8203808156535026,
+      "learning_rate": 8.316631669741934e-06,
+      "loss": 0.2799,
+      "step": 1333
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9276261986912336,
+      "learning_rate": 8.313983374817643e-06,
+      "loss": 0.275,
+      "step": 1334
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4262294071764974,
+      "learning_rate": 8.311333420749232e-06,
+      "loss": 0.2991,
+      "step": 1335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9126624516043198,
+      "learning_rate": 8.3086818088634e-06,
+      "loss": 0.2864,
+      "step": 1336
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8691479634285798,
+      "learning_rate": 8.306028540487671e-06,
+      "loss": 0.2961,
+      "step": 1337
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8876197950503014,
+      "learning_rate": 8.303373616950408e-06,
+      "loss": 0.2904,
+      "step": 1338
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8142403157347624,
+      "learning_rate": 8.300717039580792e-06,
+      "loss": 0.2986,
+      "step": 1339
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1444451254901453,
+      "learning_rate": 8.298058809708841e-06,
+      "loss": 0.2812,
+      "step": 1340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9188172052885476,
+      "learning_rate": 8.295398928665395e-06,
+      "loss": 0.27,
+      "step": 1341
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8382470430353128,
+      "learning_rate": 8.29273739778212e-06,
+      "loss": 0.2846,
+      "step": 1342
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9116343034611665,
+      "learning_rate": 8.290074218391515e-06,
+      "loss": 0.2747,
+      "step": 1343
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8350136692230988,
+      "learning_rate": 8.287409391826895e-06,
+      "loss": 0.3103,
+      "step": 1344
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7447702150340365,
+      "learning_rate": 8.284742919422406e-06,
+      "loss": 0.2907,
+      "step": 1345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8404721640013486,
+      "learning_rate": 8.282074802513016e-06,
+      "loss": 0.2776,
+      "step": 1346
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3543076539795946,
+      "learning_rate": 8.279405042434515e-06,
+      "loss": 0.2963,
+      "step": 1347
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.988268501337794,
+      "learning_rate": 8.27673364052352e-06,
+      "loss": 0.2873,
+      "step": 1348
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8884311737605137,
+      "learning_rate": 8.274060598117464e-06,
+      "loss": 0.2586,
+      "step": 1349
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7784683459343201,
+      "learning_rate": 8.271385916554605e-06,
+      "loss": 0.2771,
+      "step": 1350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7728697375891662,
+      "learning_rate": 8.268709597174022e-06,
+      "loss": 0.3031,
+      "step": 1351
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7895100218864005,
+      "learning_rate": 8.266031641315612e-06,
+      "loss": 0.2849,
+      "step": 1352
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9709657031768645,
+      "learning_rate": 8.263352050320094e-06,
+      "loss": 0.278,
+      "step": 1353
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7074797549279322,
+      "learning_rate": 8.260670825529002e-06,
+      "loss": 0.2717,
+      "step": 1354
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.806773528064426,
+      "learning_rate": 8.257987968284691e-06,
+      "loss": 0.2834,
+      "step": 1355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7822210898326847,
+      "learning_rate": 8.255303479930332e-06,
+      "loss": 0.2772,
+      "step": 1356
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7973327464995397,
+      "learning_rate": 8.252617361809911e-06,
+      "loss": 0.2922,
+      "step": 1357
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9842652126717448,
+      "learning_rate": 8.249929615268234e-06,
+      "loss": 0.2706,
+      "step": 1358
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9977031628411999,
+      "learning_rate": 8.247240241650918e-06,
+      "loss": 0.2846,
+      "step": 1359
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8262683403062453,
+      "learning_rate": 8.2445492423044e-06,
+      "loss": 0.2603,
+      "step": 1360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1720606705001673,
+      "learning_rate": 8.241856618575923e-06,
+      "loss": 0.2736,
+      "step": 1361
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9504418590982402,
+      "learning_rate": 8.239162371813552e-06,
+      "loss": 0.3024,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8162140939993675,
+      "learning_rate": 8.236466503366155e-06,
+      "loss": 0.2728,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0201111161706065,
+      "learning_rate": 8.233769014583423e-06,
+      "loss": 0.2819,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.216932690525291,
+      "learning_rate": 8.231069906815847e-06,
+      "loss": 0.2963,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8157683589153703,
+      "learning_rate": 8.22836918141474e-06,
+      "loss": 0.2881,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8794614236348264,
+      "learning_rate": 8.225666839732213e-06,
+      "loss": 0.3137,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7651339507843549,
+      "learning_rate": 8.222962883121196e-06,
+      "loss": 0.302,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7778815613343578,
+      "learning_rate": 8.220257312935422e-06,
+      "loss": 0.2882,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.818446073742485,
+      "learning_rate": 8.217550130529433e-06,
+      "loss": 0.2929,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8384037130533977,
+      "learning_rate": 8.214841337258579e-06,
+      "loss": 0.2722,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7724426752338719,
+      "learning_rate": 8.212130934479015e-06,
+      "loss": 0.275,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8198611805294007,
+      "learning_rate": 8.209418923547706e-06,
+      "loss": 0.2791,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7917861076732777,
+      "learning_rate": 8.206705305822414e-06,
+      "loss": 0.2744,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2863406471638896,
+      "learning_rate": 8.203990082661714e-06,
+      "loss": 0.2887,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9046950215080193,
+      "learning_rate": 8.201273255424979e-06,
+      "loss": 0.2772,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8870684226062803,
+      "learning_rate": 8.19855482547239e-06,
+      "loss": 0.2676,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8453222002236846,
+      "learning_rate": 8.195834794164925e-06,
+      "loss": 0.2763,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8397584790298327,
+      "learning_rate": 8.193113162864368e-06,
+      "loss": 0.2933,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8477172821324435,
+      "learning_rate": 8.190389932933301e-06,
+      "loss": 0.3001,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8866099205585865,
+      "learning_rate": 8.187665105735108e-06,
+      "loss": 0.3025,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8793388873059008,
+      "learning_rate": 8.184938682633973e-06,
+      "loss": 0.2872,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7388446681086214,
+      "learning_rate": 8.182210664994879e-06,
+      "loss": 0.2768,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8143817424924138,
+      "learning_rate": 8.179481054183605e-06,
+      "loss": 0.2768,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.376346300984398,
+      "learning_rate": 8.17674985156673e-06,
+      "loss": 0.2695,
+      "step": 1385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8240743815014119,
+      "learning_rate": 8.17401705851163e-06,
+      "loss": 0.2804,
+      "step": 1386
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7738612965462391,
+      "learning_rate": 8.171282676386474e-06,
+      "loss": 0.2621,
+      "step": 1387
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0962874560755291,
+      "learning_rate": 8.168546706560231e-06,
+      "loss": 0.2808,
+      "step": 1388
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.075586286990241,
+      "learning_rate": 8.165809150402664e-06,
+      "loss": 0.3147,
+      "step": 1389
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9048243183647892,
+      "learning_rate": 8.163070009284326e-06,
+      "loss": 0.2868,
+      "step": 1390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8111821440627807,
+      "learning_rate": 8.16032928457657e-06,
+      "loss": 0.2615,
+      "step": 1391
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7160309310336828,
+      "learning_rate": 8.157586977651534e-06,
+      "loss": 0.272,
+      "step": 1392
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8189508561562523,
+      "learning_rate": 8.154843089882159e-06,
+      "loss": 0.2834,
+      "step": 1393
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.811539082838094,
+      "learning_rate": 8.152097622642167e-06,
+      "loss": 0.2901,
+      "step": 1394
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7630734419842238,
+      "learning_rate": 8.149350577306075e-06,
+      "loss": 0.2708,
+      "step": 1395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8142183310629753,
+      "learning_rate": 8.146601955249187e-06,
+      "loss": 0.2954,
+      "step": 1396
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.763566369267271,
+      "learning_rate": 8.143851757847606e-06,
+      "loss": 0.279,
+      "step": 1397
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7877398669124583,
+      "learning_rate": 8.141099986478212e-06,
+      "loss": 0.2746,
+      "step": 1398
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8450667721634565,
+      "learning_rate": 8.138346642518679e-06,
+      "loss": 0.2835,
+      "step": 1399
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8394366730480824,
+      "learning_rate": 8.135591727347469e-06,
+      "loss": 0.2851,
+      "step": 1400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8026743029341092,
+      "learning_rate": 8.132835242343828e-06,
+      "loss": 0.2797,
+      "step": 1401
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7381180169791723,
+      "learning_rate": 8.130077188887786e-06,
+      "loss": 0.286,
+      "step": 1402
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.805669661530859,
+      "learning_rate": 8.127317568360164e-06,
+      "loss": 0.2761,
+      "step": 1403
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8985342460240192,
+      "learning_rate": 8.124556382142566e-06,
+      "loss": 0.3173,
+      "step": 1404
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8702785621705023,
+      "learning_rate": 8.121793631617375e-06,
+      "loss": 0.2907,
+      "step": 1405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9858485288744274,
+      "learning_rate": 8.119029318167763e-06,
+      "loss": 0.3005,
+      "step": 1406
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8180904580134752,
+      "learning_rate": 8.116263443177681e-06,
+      "loss": 0.2782,
+      "step": 1407
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7565062185233945,
+      "learning_rate": 8.113496008031863e-06,
+      "loss": 0.287,
+      "step": 1408
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7600615875326058,
+      "learning_rate": 8.110727014115825e-06,
+      "loss": 0.2883,
+      "step": 1409
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7768284775118554,
+      "learning_rate": 8.107956462815862e-06,
+      "loss": 0.2772,
+      "step": 1410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.777510116317734,
+      "learning_rate": 8.105184355519047e-06,
+      "loss": 0.2846,
+      "step": 1411
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8451757278039651,
+      "learning_rate": 8.102410693613239e-06,
+      "loss": 0.2945,
+      "step": 1412
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8111126593448429,
+      "learning_rate": 8.099635478487064e-06,
+      "loss": 0.278,
+      "step": 1413
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7843839598682596,
+      "learning_rate": 8.096858711529938e-06,
+      "loss": 0.2693,
+      "step": 1414
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8165488192364219,
+      "learning_rate": 8.094080394132045e-06,
+      "loss": 0.2829,
+      "step": 1415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7426116442169254,
+      "learning_rate": 8.09130052768435e-06,
+      "loss": 0.2727,
+      "step": 1416
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8290385982047915,
+      "learning_rate": 8.088519113578588e-06,
+      "loss": 0.2892,
+      "step": 1417
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.762221551213677,
+      "learning_rate": 8.085736153207277e-06,
+      "loss": 0.2945,
+      "step": 1418
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8236431509129253,
+      "learning_rate": 8.082951647963702e-06,
+      "loss": 0.2659,
+      "step": 1419
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.802846495295589,
+      "learning_rate": 8.080165599241924e-06,
+      "loss": 0.2846,
+      "step": 1420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8104480932343064,
+      "learning_rate": 8.07737800843678e-06,
+      "loss": 0.291,
+      "step": 1421
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8008667771523194,
+      "learning_rate": 8.074588876943872e-06,
+      "loss": 0.2909,
+      "step": 1422
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8062546458473153,
+      "learning_rate": 8.07179820615958e-06,
+      "loss": 0.2882,
+      "step": 1423
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8385755238517676,
+      "learning_rate": 8.069005997481053e-06,
+      "loss": 0.3094,
+      "step": 1424
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8190412911844063,
+      "learning_rate": 8.066212252306204e-06,
+      "loss": 0.2718,
+      "step": 1425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7624338252266949,
+      "learning_rate": 8.063416972033725e-06,
+      "loss": 0.2714,
+      "step": 1426
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.792385334866942,
+      "learning_rate": 8.060620158063067e-06,
+      "loss": 0.2975,
+      "step": 1427
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7757579208070389,
+      "learning_rate": 8.057821811794457e-06,
+      "loss": 0.2644,
+      "step": 1428
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8217678331718805,
+      "learning_rate": 8.055021934628886e-06,
+      "loss": 0.2915,
+      "step": 1429
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8168839320184221,
+      "learning_rate": 8.05222052796811e-06,
+      "loss": 0.2807,
+      "step": 1430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8511312170848377,
+      "learning_rate": 8.049417593214652e-06,
+      "loss": 0.2678,
+      "step": 1431
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8146774784691951,
+      "learning_rate": 8.046613131771798e-06,
+      "loss": 0.2662,
+      "step": 1432
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7388976188057744,
+      "learning_rate": 8.043807145043604e-06,
+      "loss": 0.2644,
+      "step": 1433
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7994086201656646,
+      "learning_rate": 8.040999634434883e-06,
+      "loss": 0.3043,
+      "step": 1434
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.864734357145189,
+      "learning_rate": 8.038190601351215e-06,
+      "loss": 0.29,
+      "step": 1435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5231253851639006,
+      "learning_rate": 8.03538004719894e-06,
+      "loss": 0.2549,
+      "step": 1436
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8470702193696265,
+      "learning_rate": 8.032567973385162e-06,
+      "loss": 0.3249,
+      "step": 1437
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7817426330880487,
+      "learning_rate": 8.029754381317741e-06,
+      "loss": 0.2656,
+      "step": 1438
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.64277353379814,
+      "learning_rate": 8.026939272405306e-06,
+      "loss": 0.289,
+      "step": 1439
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7839414858157064,
+      "learning_rate": 8.024122648057234e-06,
+      "loss": 0.2737,
+      "step": 1440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8958354487282021,
+      "learning_rate": 8.021304509683672e-06,
+      "loss": 0.2907,
+      "step": 1441
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8224123580548967,
+      "learning_rate": 8.018484858695514e-06,
+      "loss": 0.2792,
+      "step": 1442
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8263729341281115,
+      "learning_rate": 8.015663696504424e-06,
+      "loss": 0.2736,
+      "step": 1443
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8334456138144333,
+      "learning_rate": 8.012841024522808e-06,
+      "loss": 0.3261,
+      "step": 1444
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.825366798344378,
+      "learning_rate": 8.01001684416384e-06,
+      "loss": 0.2976,
+      "step": 1445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3208258655783218,
+      "learning_rate": 8.00719115684144e-06,
+      "loss": 0.2956,
+      "step": 1446
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7590176017425264,
+      "learning_rate": 8.004363963970294e-06,
+      "loss": 0.2733,
+      "step": 1447
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7082517456318983,
+      "learning_rate": 8.001535266965829e-06,
+      "loss": 0.2785,
+      "step": 1448
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7896128392913527,
+      "learning_rate": 7.998705067244232e-06,
+      "loss": 0.308,
+      "step": 1449
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.783138927164651,
+      "learning_rate": 7.995873366222442e-06,
+      "loss": 0.2637,
+      "step": 1450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7527124233945439,
+      "learning_rate": 7.993040165318148e-06,
+      "loss": 0.2777,
+      "step": 1451
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8266789518326695,
+      "learning_rate": 7.990205465949792e-06,
+      "loss": 0.2869,
+      "step": 1452
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8507969147796602,
+      "learning_rate": 7.987369269536563e-06,
+      "loss": 0.2601,
+      "step": 1453
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7891331227817714,
+      "learning_rate": 7.9845315774984e-06,
+      "loss": 0.3031,
+      "step": 1454
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7928191920311444,
+      "learning_rate": 7.981692391255996e-06,
+      "loss": 0.2608,
+      "step": 1455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9146677574690956,
+      "learning_rate": 7.978851712230786e-06,
+      "loss": 0.2971,
+      "step": 1456
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0575224624015458,
+      "learning_rate": 7.976009541844956e-06,
+      "loss": 0.2756,
+      "step": 1457
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8061447312635872,
+      "learning_rate": 7.973165881521435e-06,
+      "loss": 0.2683,
+      "step": 1458
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7987445715622473,
+      "learning_rate": 7.9703207326839e-06,
+      "loss": 0.2776,
+      "step": 1459
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7730059416926843,
+      "learning_rate": 7.967474096756775e-06,
+      "loss": 0.2575,
+      "step": 1460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7810950271546707,
+      "learning_rate": 7.964625975165225e-06,
+      "loss": 0.2899,
+      "step": 1461
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7734209364869274,
+      "learning_rate": 7.961776369335164e-06,
+      "loss": 0.2722,
+      "step": 1462
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7604011877840355,
+      "learning_rate": 7.958925280693243e-06,
+      "loss": 0.2543,
+      "step": 1463
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8416634163182992,
+      "learning_rate": 7.95607271066686e-06,
+      "loss": 0.3009,
+      "step": 1464
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7507923437396616,
+      "learning_rate": 7.95321866068415e-06,
+      "loss": 0.238,
+      "step": 1465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4595933118496487,
+      "learning_rate": 7.950363132173992e-06,
+      "loss": 0.2987,
+      "step": 1466
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8566131962650662,
+      "learning_rate": 7.947506126566009e-06,
+      "loss": 0.2715,
+      "step": 1467
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8756672429939591,
+      "learning_rate": 7.944647645290555e-06,
+      "loss": 0.2998,
+      "step": 1468
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8205145247201313,
+      "learning_rate": 7.94178768977873e-06,
+      "loss": 0.3145,
+      "step": 1469
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8895574288007237,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.2744,
+      "step": 1470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8272439332420523,
+      "learning_rate": 7.93606336177404e-06,
+      "loss": 0.302,
+      "step": 1471
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1689407833256216,
+      "learning_rate": 7.933198992147057e-06,
+      "loss": 0.2873,
+      "step": 1472
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8621997023924174,
+      "learning_rate": 7.930333154015467e-06,
+      "loss": 0.277,
+      "step": 1473
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8053277256606999,
+      "learning_rate": 7.927465848814045e-06,
+      "loss": 0.2867,
+      "step": 1474
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9750436904081562,
+      "learning_rate": 7.92459707797831e-06,
+      "loss": 0.2768,
+      "step": 1475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8678137288898734,
+      "learning_rate": 7.921726842944507e-06,
+      "loss": 0.2808,
+      "step": 1476
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8761913156392614,
+      "learning_rate": 7.918855145149622e-06,
+      "loss": 0.3115,
+      "step": 1477
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0224831669155336,
+      "learning_rate": 7.915981986031367e-06,
+      "loss": 0.2709,
+      "step": 1478
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8065908108678069,
+      "learning_rate": 7.913107367028188e-06,
+      "loss": 0.2976,
+      "step": 1479
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.760061151276328,
+      "learning_rate": 7.910231289579263e-06,
+      "loss": 0.2643,
+      "step": 1480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8255258864686359,
+      "learning_rate": 7.907353755124493e-06,
+      "loss": 0.2686,
+      "step": 1481
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8117256539253475,
+      "learning_rate": 7.90447476510452e-06,
+      "loss": 0.2866,
+      "step": 1482
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8016547569710417,
+      "learning_rate": 7.901594320960709e-06,
+      "loss": 0.2725,
+      "step": 1483
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8264825856985024,
+      "learning_rate": 7.898712424135149e-06,
+      "loss": 0.2784,
+      "step": 1484
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8399957937574787,
+      "learning_rate": 7.895829076070663e-06,
+      "loss": 0.2997,
+      "step": 1485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8655905659305005,
+      "learning_rate": 7.892944278210796e-06,
+      "loss": 0.2706,
+      "step": 1486
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9622184210061497,
+      "learning_rate": 7.890058031999825e-06,
+      "loss": 0.3016,
+      "step": 1487
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8148534828169289,
+      "learning_rate": 7.887170338882742e-06,
+      "loss": 0.2682,
+      "step": 1488
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0196489729372573,
+      "learning_rate": 7.884281200305272e-06,
+      "loss": 0.2916,
+      "step": 1489
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9817185047241809,
+      "learning_rate": 7.88139061771386e-06,
+      "loss": 0.2653,
+      "step": 1490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7506628457920514,
+      "learning_rate": 7.878498592555675e-06,
+      "loss": 0.273,
+      "step": 1491
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.869447772605467,
+      "learning_rate": 7.875605126278608e-06,
+      "loss": 0.2967,
+      "step": 1492
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8350793868467963,
+      "learning_rate": 7.872710220331271e-06,
+      "loss": 0.3074,
+      "step": 1493
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8202945976683922,
+      "learning_rate": 7.869813876162997e-06,
+      "loss": 0.278,
+      "step": 1494
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7039761669461602,
+      "learning_rate": 7.866916095223842e-06,
+      "loss": 0.248,
+      "step": 1495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8162160546949805,
+      "learning_rate": 7.864016878964575e-06,
+      "loss": 0.3058,
+      "step": 1496
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9997939246850418,
+      "learning_rate": 7.86111622883669e-06,
+      "loss": 0.2688,
+      "step": 1497
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8018011060440153,
+      "learning_rate": 7.858214146292394e-06,
+      "loss": 0.3027,
+      "step": 1498
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.758802486376414,
+      "learning_rate": 7.855310632784615e-06,
+      "loss": 0.2602,
+      "step": 1499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7807864261444986,
+      "learning_rate": 7.852405689766993e-06,
+      "loss": 0.2814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7612167859160681,
+      "learning_rate": 7.849499318693887e-06,
+      "loss": 0.2889,
+      "step": 1501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8833956316200079,
+      "learning_rate": 7.846591521020373e-06,
+      "loss": 0.2691,
+      "step": 1502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9111737420768589,
+      "learning_rate": 7.843682298202235e-06,
+      "loss": 0.29,
+      "step": 1503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7663375222283876,
+      "learning_rate": 7.840771651695976e-06,
+      "loss": 0.2764,
+      "step": 1504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9900929113630423,
+      "learning_rate": 7.83785958295881e-06,
+      "loss": 0.2935,
+      "step": 1505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8203333845598397,
+      "learning_rate": 7.834946093448658e-06,
+      "loss": 0.2708,
+      "step": 1506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7494434407225327,
+      "learning_rate": 7.832031184624165e-06,
+      "loss": 0.2599,
+      "step": 1507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7532456202033322,
+      "learning_rate": 7.829114857944672e-06,
+      "loss": 0.2856,
+      "step": 1508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7993076301771292,
+      "learning_rate": 7.826197114870243e-06,
+      "loss": 0.2959,
+      "step": 1509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8777773704094564,
+      "learning_rate": 7.823277956861635e-06,
+      "loss": 0.2778,
+      "step": 1510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8904845331625537,
+      "learning_rate": 7.820357385380333e-06,
+      "loss": 0.2707,
+      "step": 1511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8491117720642429,
+      "learning_rate": 7.817435401888514e-06,
+      "loss": 0.277,
+      "step": 1512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9094580714161332,
+      "learning_rate": 7.814512007849069e-06,
+      "loss": 0.3003,
+      "step": 1513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9029872386006449,
+      "learning_rate": 7.811587204725593e-06,
+      "loss": 0.2975,
+      "step": 1514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7684030858857039,
+      "learning_rate": 7.808660993982389e-06,
+      "loss": 0.277,
+      "step": 1515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9520674208976132,
+      "learning_rate": 7.80573337708446e-06,
+      "loss": 0.2871,
+      "step": 1516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7915361387619609,
+      "learning_rate": 7.802804355497519e-06,
+      "loss": 0.249,
+      "step": 1517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9101835085033889,
+      "learning_rate": 7.799873930687979e-06,
+      "loss": 0.2896,
+      "step": 1518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9073135009980376,
+      "learning_rate": 7.796942104122953e-06,
+      "loss": 0.2816,
+      "step": 1519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0772603822638585,
+      "learning_rate": 7.79400887727026e-06,
+      "loss": 0.3034,
+      "step": 1520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8309577355124775,
+      "learning_rate": 7.79107425159842e-06,
+      "loss": 0.2769,
+      "step": 1521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0136916881683655,
+      "learning_rate": 7.78813822857665e-06,
+      "loss": 0.2978,
+      "step": 1522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7596398213353267,
+      "learning_rate": 7.785200809674869e-06,
+      "loss": 0.2596,
+      "step": 1523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8635703989738237,
+      "learning_rate": 7.782261996363694e-06,
+      "loss": 0.2885,
+      "step": 1524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7543680368928855,
+      "learning_rate": 7.779321790114442e-06,
+      "loss": 0.2692,
+      "step": 1525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7490333406852111,
+      "learning_rate": 7.776380192399122e-06,
+      "loss": 0.2766,
+      "step": 1526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8035715823625134,
+      "learning_rate": 7.773437204690449e-06,
+      "loss": 0.2801,
+      "step": 1527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7794423542671997,
+      "learning_rate": 7.770492828461824e-06,
+      "loss": 0.3225,
+      "step": 1528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7957705018273902,
+      "learning_rate": 7.767547065187349e-06,
+      "loss": 0.2768,
+      "step": 1529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7451887769847855,
+      "learning_rate": 7.764599916341817e-06,
+      "loss": 0.2703,
+      "step": 1530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7306061417497395,
+      "learning_rate": 7.761651383400722e-06,
+      "loss": 0.2638,
+      "step": 1531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8426790987326729,
+      "learning_rate": 7.75870146784024e-06,
+      "loss": 0.2706,
+      "step": 1532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8669471205873402,
+      "learning_rate": 7.755750171137245e-06,
+      "loss": 0.2834,
+      "step": 1533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8859119963147409,
+      "learning_rate": 7.752797494769307e-06,
+      "loss": 0.2952,
+      "step": 1534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9198805846869554,
+      "learning_rate": 7.749843440214675e-06,
+      "loss": 0.2966,
+      "step": 1535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7858995795565105,
+      "learning_rate": 7.7468880089523e-06,
+      "loss": 0.2663,
+      "step": 1536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8613758361347971,
+      "learning_rate": 7.743931202461818e-06,
+      "loss": 0.2759,
+      "step": 1537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8153975356581273,
+      "learning_rate": 7.74097302222355e-06,
+      "loss": 0.272,
+      "step": 1538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9067433030207944,
+      "learning_rate": 7.738013469718507e-06,
+      "loss": 0.2942,
+      "step": 1539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7927349961640611,
+      "learning_rate": 7.73505254642839e-06,
+      "loss": 0.2763,
+      "step": 1540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8526117936003229,
+      "learning_rate": 7.73209025383558e-06,
+      "loss": 0.2683,
+      "step": 1541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8184396889019998,
+      "learning_rate": 7.729126593423151e-06,
+      "loss": 0.3109,
+      "step": 1542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.921859948614392,
+      "learning_rate": 7.726161566674856e-06,
+      "loss": 0.2831,
+      "step": 1543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8191005497404884,
+      "learning_rate": 7.723195175075136e-06,
+      "loss": 0.2702,
+      "step": 1544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8398560989379976,
+      "learning_rate": 7.720227420109112e-06,
+      "loss": 0.284,
+      "step": 1545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.910722309338154,
+      "learning_rate": 7.71725830326259e-06,
+      "loss": 0.2913,
+      "step": 1546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8661372242637406,
+      "learning_rate": 7.714287826022056e-06,
+      "loss": 0.2918,
+      "step": 1547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7477604385197407,
+      "learning_rate": 7.711315989874677e-06,
+      "loss": 0.2661,
+      "step": 1548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6861445328654948,
+      "learning_rate": 7.708342796308303e-06,
+      "loss": 0.2503,
+      "step": 1549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7572450460995089,
+      "learning_rate": 7.70536824681146e-06,
+      "loss": 0.2773,
+      "step": 1550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7896988791822799,
+      "learning_rate": 7.702392342873358e-06,
+      "loss": 0.3016,
+      "step": 1551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9432504967498507,
+      "learning_rate": 7.69941508598388e-06,
+      "loss": 0.2818,
+      "step": 1552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7803474400954176,
+      "learning_rate": 7.696436477633588e-06,
+      "loss": 0.2838,
+      "step": 1553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8973727804342867,
+      "learning_rate": 7.69345651931372e-06,
+      "loss": 0.2579,
+      "step": 1554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.937541825991562,
+      "learning_rate": 7.690475212516191e-06,
+      "loss": 0.2849,
+      "step": 1555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8825678164825909,
+      "learning_rate": 7.687492558733593e-06,
+      "loss": 0.276,
+      "step": 1556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9848037276340559,
+      "learning_rate": 7.684508559459188e-06,
+      "loss": 0.3123,
+      "step": 1557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8583563414402845,
+      "learning_rate": 7.681523216186912e-06,
+      "loss": 0.2801,
+      "step": 1558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8097379614458422,
+      "learning_rate": 7.67853653041138e-06,
+      "loss": 0.2817,
+      "step": 1559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9111576392487886,
+      "learning_rate": 7.675548503627872e-06,
+      "loss": 0.2803,
+      "step": 1560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8220707320451502,
+      "learning_rate": 7.672559137332343e-06,
+      "loss": 0.3072,
+      "step": 1561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9306000223650084,
+      "learning_rate": 7.669568433021418e-06,
+      "loss": 0.2816,
+      "step": 1562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7174608292406655,
+      "learning_rate": 7.666576392192389e-06,
+      "loss": 0.2804,
+      "step": 1563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0012534628762466,
+      "learning_rate": 7.663583016343224e-06,
+      "loss": 0.2903,
+      "step": 1564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.240447487620713,
+      "learning_rate": 7.66058830697255e-06,
+      "loss": 0.2975,
+      "step": 1565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0786625765868305,
+      "learning_rate": 7.65759226557967e-06,
+      "loss": 0.2948,
+      "step": 1566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7786780069271979,
+      "learning_rate": 7.65459489366455e-06,
+      "loss": 0.2816,
+      "step": 1567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8264703168272057,
+      "learning_rate": 7.651596192727826e-06,
+      "loss": 0.2877,
+      "step": 1568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8662904655391195,
+      "learning_rate": 7.648596164270792e-06,
+      "loss": 0.302,
+      "step": 1569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9516560756373207,
+      "learning_rate": 7.64559480979541e-06,
+      "loss": 0.291,
+      "step": 1570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0861186909548726,
+      "learning_rate": 7.64259213080431e-06,
+      "loss": 0.3113,
+      "step": 1571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7522488545693069,
+      "learning_rate": 7.63958812880078e-06,
+      "loss": 0.2487,
+      "step": 1572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.845409246732738,
+      "learning_rate": 7.636582805288771e-06,
+      "loss": 0.2851,
+      "step": 1573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8052205558727634,
+      "learning_rate": 7.633576161772898e-06,
+      "loss": 0.2593,
+      "step": 1574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7884385031863786,
+      "learning_rate": 7.630568199758436e-06,
+      "loss": 0.2742,
+      "step": 1575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8168990822212525,
+      "learning_rate": 7.627558920751321e-06,
+      "loss": 0.2809,
+      "step": 1576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8100403003699771,
+      "learning_rate": 7.624548326258144e-06,
+      "loss": 0.2818,
+      "step": 1577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.767042231053844,
+      "learning_rate": 7.621536417786159e-06,
+      "loss": 0.2825,
+      "step": 1578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0699438128898473,
+      "learning_rate": 7.618523196843276e-06,
+      "loss": 0.2981,
+      "step": 1579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8983659083902245,
+      "learning_rate": 7.615508664938065e-06,
+      "loss": 0.3009,
+      "step": 1580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2577707518409031,
+      "learning_rate": 7.612492823579744e-06,
+      "loss": 0.2547,
+      "step": 1581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7716673615345931,
+      "learning_rate": 7.6094756742782e-06,
+      "loss": 0.2636,
+      "step": 1582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8661904360368491,
+      "learning_rate": 7.606457218543961e-06,
+      "loss": 0.2702,
+      "step": 1583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8275778756812623,
+      "learning_rate": 7.6034374578882205e-06,
+      "loss": 0.2759,
+      "step": 1584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8770218056516232,
+      "learning_rate": 7.600416393822816e-06,
+      "loss": 0.2856,
+      "step": 1585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7268691060948221,
+      "learning_rate": 7.597394027860243e-06,
+      "loss": 0.263,
+      "step": 1586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.047386577713546,
+      "learning_rate": 7.594370361513648e-06,
+      "loss": 0.2828,
+      "step": 1587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8828905572504454,
+      "learning_rate": 7.5913453962968296e-06,
+      "loss": 0.3103,
+      "step": 1588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8003229453183097,
+      "learning_rate": 7.5883191337242315e-06,
+      "loss": 0.2822,
+      "step": 1589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.825040962872128,
+      "learning_rate": 7.585291575310952e-06,
+      "loss": 0.2733,
+      "step": 1590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7968410638404704,
+      "learning_rate": 7.582262722572739e-06,
+      "loss": 0.2817,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.85092583464101,
+      "learning_rate": 7.579232577025984e-06,
+      "loss": 0.2906,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9181531336983434,
+      "learning_rate": 7.576201140187727e-06,
+      "loss": 0.2734,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8141305959448977,
+      "learning_rate": 7.5731684135756566e-06,
+      "loss": 0.2571,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8011633146266064,
+      "learning_rate": 7.570134398708106e-06,
+      "loss": 0.2729,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.79150655128365,
+      "learning_rate": 7.567099097104053e-06,
+      "loss": 0.2603,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7781238639291225,
+      "learning_rate": 7.5640625102831215e-06,
+      "loss": 0.2919,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8173856144385161,
+      "learning_rate": 7.5610246397655715e-06,
+      "loss": 0.2744,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7836037837390007,
+      "learning_rate": 7.557985487072318e-06,
+      "loss": 0.2671,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.815664441441121,
+      "learning_rate": 7.5549450537249096e-06,
+      "loss": 0.2659,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7357975293039324,
+      "learning_rate": 7.551903341245534e-06,
+      "loss": 0.2769,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6851184168800262,
+      "learning_rate": 7.548860351157028e-06,
+      "loss": 0.2807,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7227158236079662,
+      "learning_rate": 7.54581608498286e-06,
+      "loss": 0.2303,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0268306525905755,
+      "learning_rate": 7.542770544247141e-06,
+      "loss": 0.2875,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7916000740198295,
+      "learning_rate": 7.53972373047462e-06,
+      "loss": 0.2855,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9837154055927362,
+      "learning_rate": 7.536675645190682e-06,
+      "loss": 0.2746,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7952710703708994,
+      "learning_rate": 7.53362628992135e-06,
+      "loss": 0.2796,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.770387907834022,
+      "learning_rate": 7.530575666193283e-06,
+      "loss": 0.2682,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8171998670491365,
+      "learning_rate": 7.527523775533771e-06,
+      "loss": 0.2662,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9169801499072286,
+      "learning_rate": 7.524470619470745e-06,
+      "loss": 0.2983,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.764470198504701,
+      "learning_rate": 7.521416199532765e-06,
+      "loss": 0.2758,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9951494598641277,
+      "learning_rate": 7.518360517249024e-06,
+      "loss": 0.2873,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9120870640886928,
+      "learning_rate": 7.515303574149348e-06,
+      "loss": 0.2663,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.274052503128875,
+      "learning_rate": 7.512245371764197e-06,
+      "loss": 0.2937,
+      "step": 1614
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.747783481430921,
+      "learning_rate": 7.509185911624653e-06,
+      "loss": 0.2855,
+      "step": 1615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.772151061589686,
+      "learning_rate": 7.506125195262437e-06,
+      "loss": 0.2833,
+      "step": 1616
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7948341449262875,
+      "learning_rate": 7.503063224209896e-06,
+      "loss": 0.2571,
+      "step": 1617
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8713675836201565,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2837,
+      "step": 1618
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8362209307252766,
+      "learning_rate": 7.4969355241663554e-06,
+      "loss": 0.2883,
+      "step": 1619
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8130127412507159,
+      "learning_rate": 7.493869798243188e-06,
+      "loss": 0.2678,
+      "step": 1620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0806563559337727,
+      "learning_rate": 7.490802823765352e-06,
+      "loss": 0.2814,
+      "step": 1621
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9337256289914934,
+      "learning_rate": 7.487734602268326e-06,
+      "loss": 0.2847,
+      "step": 1622
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8399135897674889,
+      "learning_rate": 7.484665135288214e-06,
+      "loss": 0.2686,
+      "step": 1623
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7843466982215382,
+      "learning_rate": 7.4815944243617455e-06,
+      "loss": 0.2821,
+      "step": 1624
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8011118160992257,
+      "learning_rate": 7.478522471026267e-06,
+      "loss": 0.2908,
+      "step": 1625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9132488481319146,
+      "learning_rate": 7.475449276819753e-06,
+      "loss": 0.2944,
+      "step": 1626
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8216124753138614,
+      "learning_rate": 7.472374843280794e-06,
+      "loss": 0.2725,
+      "step": 1627
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8813698977909934,
+      "learning_rate": 7.469299171948608e-06,
+      "loss": 0.2997,
+      "step": 1628
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7392913093476957,
+      "learning_rate": 7.466222264363022e-06,
+      "loss": 0.2812,
+      "step": 1629
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7239486780599199,
+      "learning_rate": 7.463144122064494e-06,
+      "loss": 0.2737,
+      "step": 1630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8267124871604024,
+      "learning_rate": 7.460064746594091e-06,
+      "loss": 0.2991,
+      "step": 1631
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7553932298829961,
+      "learning_rate": 7.456984139493502e-06,
+      "loss": 0.2843,
+      "step": 1632
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7879643761698112,
+      "learning_rate": 7.453902302305032e-06,
+      "loss": 0.2702,
+      "step": 1633
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8343723671686345,
+      "learning_rate": 7.4508192365716005e-06,
+      "loss": 0.2837,
+      "step": 1634
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9396347536365841,
+      "learning_rate": 7.447734943836742e-06,
+      "loss": 0.2931,
+      "step": 1635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.834287850775466,
+      "learning_rate": 7.444649425644606e-06,
+      "loss": 0.2865,
+      "step": 1636
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8444505788888228,
+      "learning_rate": 7.441562683539959e-06,
+      "loss": 0.2814,
+      "step": 1637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7607364371461979,
+      "learning_rate": 7.438474719068174e-06,
+      "loss": 0.2986,
+      "step": 1638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.199164349340232,
+      "learning_rate": 7.435385533775239e-06,
+      "loss": 0.246,
+      "step": 1639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9030164623461093,
+      "learning_rate": 7.432295129207753e-06,
+      "loss": 0.2691,
+      "step": 1640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8392033313918988,
+      "learning_rate": 7.429203506912927e-06,
+      "loss": 0.2628,
+      "step": 1641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8331825683703635,
+      "learning_rate": 7.426110668438578e-06,
+      "loss": 0.2881,
+      "step": 1642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8682716829756117,
+      "learning_rate": 7.423016615333135e-06,
+      "loss": 0.2946,
+      "step": 1643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7846371044417214,
+      "learning_rate": 7.419921349145635e-06,
+      "loss": 0.2703,
+      "step": 1644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8569258372547846,
+      "learning_rate": 7.4168248714257204e-06,
+      "loss": 0.268,
+      "step": 1645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8433598831330701,
+      "learning_rate": 7.413727183723642e-06,
+      "loss": 0.2797,
+      "step": 1646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.881977552146873,
+      "learning_rate": 7.410628287590254e-06,
+      "loss": 0.2916,
+      "step": 1647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8503508767835208,
+      "learning_rate": 7.4075281845770196e-06,
+      "loss": 0.2609,
+      "step": 1648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8928328584940355,
+      "learning_rate": 7.404426876236003e-06,
+      "loss": 0.303,
+      "step": 1649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7966186555743715,
+      "learning_rate": 7.401324364119872e-06,
+      "loss": 0.2692,
+      "step": 1650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8064663345815343,
+      "learning_rate": 7.3982206497818965e-06,
+      "loss": 0.2796,
+      "step": 1651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7069879457793663,
+      "learning_rate": 7.395115734775953e-06,
+      "loss": 0.2749,
+      "step": 1652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7803397475328926,
+      "learning_rate": 7.392009620656513e-06,
+      "loss": 0.2544,
+      "step": 1653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8125931424193473,
+      "learning_rate": 7.388902308978652e-06,
+      "loss": 0.2894,
+      "step": 1654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.916897815410437,
+      "learning_rate": 7.3857938012980425e-06,
+      "loss": 0.2791,
+      "step": 1655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7714336142651191,
+      "learning_rate": 7.382684099170959e-06,
+      "loss": 0.2677,
+      "step": 1656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2470085891616345,
+      "learning_rate": 7.379573204154274e-06,
+      "loss": 0.2857,
+      "step": 1657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7170165181206646,
+      "learning_rate": 7.37646111780545e-06,
+      "loss": 0.2706,
+      "step": 1658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8364536161361461,
+      "learning_rate": 7.3733478416825565e-06,
+      "loss": 0.2977,
+      "step": 1659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.727378600192885,
+      "learning_rate": 7.370233377344252e-06,
+      "loss": 0.2806,
+      "step": 1660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7811492703882663,
+      "learning_rate": 7.367117726349793e-06,
+      "loss": 0.2776,
+      "step": 1661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8643726509436449,
+      "learning_rate": 7.364000890259024e-06,
+      "loss": 0.2792,
+      "step": 1662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.897896972949269,
+      "learning_rate": 7.360882870632393e-06,
+      "loss": 0.2822,
+      "step": 1663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7807878940703933,
+      "learning_rate": 7.357763669030931e-06,
+      "loss": 0.2851,
+      "step": 1664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9411271298936754,
+      "learning_rate": 7.354643287016268e-06,
+      "loss": 0.3028,
+      "step": 1665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7641660393510575,
+      "learning_rate": 7.351521726150619e-06,
+      "loss": 0.2838,
+      "step": 1666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0676373867310565,
+      "learning_rate": 7.348398987996792e-06,
+      "loss": 0.2774,
+      "step": 1667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7418115103935384,
+      "learning_rate": 7.3452750741181855e-06,
+      "loss": 0.2743,
+      "step": 1668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.942632614035707,
+      "learning_rate": 7.342149986078786e-06,
+      "loss": 0.2902,
+      "step": 1669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9246088893131054,
+      "learning_rate": 7.339023725443166e-06,
+      "loss": 0.2967,
+      "step": 1670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7250467947384419,
+      "learning_rate": 7.335896293776486e-06,
+      "loss": 0.2784,
+      "step": 1671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8018457095603927,
+      "learning_rate": 7.332767692644493e-06,
+      "loss": 0.2845,
+      "step": 1672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8280964990104027,
+      "learning_rate": 7.329637923613522e-06,
+      "loss": 0.2798,
+      "step": 1673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8648080540228092,
+      "learning_rate": 7.326506988250488e-06,
+      "loss": 0.2861,
+      "step": 1674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8191789627496174,
+      "learning_rate": 7.323374888122893e-06,
+      "loss": 0.2728,
+      "step": 1675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7829570846152243,
+      "learning_rate": 7.320241624798821e-06,
+      "loss": 0.2632,
+      "step": 1676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8967491934676134,
+      "learning_rate": 7.31710719984694e-06,
+      "loss": 0.3122,
+      "step": 1677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.816519380495374,
+      "learning_rate": 7.313971614836496e-06,
+      "loss": 0.2798,
+      "step": 1678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.978492880040161,
+      "learning_rate": 7.310834871337319e-06,
+      "loss": 0.254,
+      "step": 1679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8419503266012873,
+      "learning_rate": 7.307696970919819e-06,
+      "loss": 0.2816,
+      "step": 1680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.778581577377549,
+      "learning_rate": 7.3045579151549836e-06,
+      "loss": 0.2872,
+      "step": 1681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8591120745562594,
+      "learning_rate": 7.3014177056143795e-06,
+      "loss": 0.2785,
+      "step": 1682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7446216286152907,
+      "learning_rate": 7.298276343870152e-06,
+      "loss": 0.2864,
+      "step": 1683
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8220651833463559,
+      "learning_rate": 7.295133831495021e-06,
+      "loss": 0.293,
+      "step": 1684
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8199685390818615,
+      "learning_rate": 7.291990170062283e-06,
+      "loss": 0.2744,
+      "step": 1685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7794325364953729,
+      "learning_rate": 7.288845361145812e-06,
+      "loss": 0.273,
+      "step": 1686
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8261722428966315,
+      "learning_rate": 7.285699406320055e-06,
+      "loss": 0.2692,
+      "step": 1687
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8416378770197799,
+      "learning_rate": 7.282552307160033e-06,
+      "loss": 0.2794,
+      "step": 1688
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8204318811025931,
+      "learning_rate": 7.2794040652413374e-06,
+      "loss": 0.2762,
+      "step": 1689
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8293487024555088,
+      "learning_rate": 7.276254682140137e-06,
+      "loss": 0.285,
+      "step": 1690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7879566019649478,
+      "learning_rate": 7.273104159433166e-06,
+      "loss": 0.2593,
+      "step": 1691
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0407482700848187,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.2888,
+      "step": 1692
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8511955281076429,
+      "learning_rate": 7.26679970151172e-06,
+      "loss": 0.2867,
+      "step": 1693
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4062390110838476,
+      "learning_rate": 7.2636457694535655e-06,
+      "loss": 0.3086,
+      "step": 1694
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8585483976662772,
+      "learning_rate": 7.260490704102287e-06,
+      "loss": 0.268,
+      "step": 1695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8340786866995145,
+      "learning_rate": 7.2573345070374675e-06,
+      "loss": 0.2688,
+      "step": 1696
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.799576358502223,
+      "learning_rate": 7.254177179839257e-06,
+      "loss": 0.2865,
+      "step": 1697
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7288928566458596,
+      "learning_rate": 7.251018724088367e-06,
+      "loss": 0.2608,
+      "step": 1698
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8414397231133561,
+      "learning_rate": 7.247859141366077e-06,
+      "loss": 0.2936,
+      "step": 1699
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7409954401506411,
+      "learning_rate": 7.244698433254231e-06,
+      "loss": 0.2665,
+      "step": 1700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.897029909014281,
+      "learning_rate": 7.241536601335237e-06,
+      "loss": 0.2848,
+      "step": 1701
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1594426306405037,
+      "learning_rate": 7.238373647192063e-06,
+      "loss": 0.2831,
+      "step": 1702
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7248708203970783,
+      "learning_rate": 7.235209572408241e-06,
+      "loss": 0.2807,
+      "step": 1703
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8609044754707597,
+      "learning_rate": 7.232044378567864e-06,
+      "loss": 0.2891,
+      "step": 1704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8295489481261596,
+      "learning_rate": 7.228878067255585e-06,
+      "loss": 0.2912,
+      "step": 1705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.899170974991406,
+      "learning_rate": 7.225710640056614e-06,
+      "loss": 0.2824,
+      "step": 1706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8018562105689494,
+      "learning_rate": 7.222542098556722e-06,
+      "loss": 0.2673,
+      "step": 1707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8790671428906918,
+      "learning_rate": 7.2193724443422405e-06,
+      "loss": 0.2866,
+      "step": 1708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8612987766561174,
+      "learning_rate": 7.2162016790000545e-06,
+      "loss": 0.2684,
+      "step": 1709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7710569583586294,
+      "learning_rate": 7.213029804117604e-06,
+      "loss": 0.298,
+      "step": 1710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7682330591096805,
+      "learning_rate": 7.209856821282888e-06,
+      "loss": 0.262,
+      "step": 1711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7346107610525915,
+      "learning_rate": 7.206682732084458e-06,
+      "loss": 0.2808,
+      "step": 1712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9382003149811986,
+      "learning_rate": 7.203507538111423e-06,
+      "loss": 0.2734,
+      "step": 1713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.764179208445095,
+      "learning_rate": 7.200331240953439e-06,
+      "loss": 0.2757,
+      "step": 1714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8031112657019002,
+      "learning_rate": 7.197153842200718e-06,
+      "loss": 0.2794,
+      "step": 1715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7596060442587482,
+      "learning_rate": 7.193975343444024e-06,
+      "loss": 0.2581,
+      "step": 1716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7158000377134397,
+      "learning_rate": 7.190795746274673e-06,
+      "loss": 0.2757,
+      "step": 1717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.919289639390274,
+      "learning_rate": 7.187615052284522e-06,
+      "loss": 0.2934,
+      "step": 1718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8597917336125098,
+      "learning_rate": 7.184433263065991e-06,
+      "loss": 0.2581,
+      "step": 1719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7726832831268787,
+      "learning_rate": 7.181250380212037e-06,
+      "loss": 0.2974,
+      "step": 1720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7655521991492308,
+      "learning_rate": 7.178066405316172e-06,
+      "loss": 0.2559,
+      "step": 1721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7550386609343781,
+      "learning_rate": 7.174881339972448e-06,
+      "loss": 0.2725,
+      "step": 1722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8629719730988605,
+      "learning_rate": 7.171695185775468e-06,
+      "loss": 0.2643,
+      "step": 1723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8278851397370779,
+      "learning_rate": 7.1685079443203784e-06,
+      "loss": 0.2944,
+      "step": 1724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7695840940385571,
+      "learning_rate": 7.165319617202871e-06,
+      "loss": 0.2791,
+      "step": 1725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7785729755272401,
+      "learning_rate": 7.16213020601918e-06,
+      "loss": 0.2793,
+      "step": 1726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9944704899554991,
+      "learning_rate": 7.158939712366079e-06,
+      "loss": 0.2856,
+      "step": 1727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7393496721507371,
+      "learning_rate": 7.155748137840892e-06,
+      "loss": 0.254,
+      "step": 1728
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8377318386923918,
+      "learning_rate": 7.1525554840414765e-06,
+      "loss": 0.2741,
+      "step": 1729
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9333426971487835,
+      "learning_rate": 7.149361752566233e-06,
+      "loss": 0.2781,
+      "step": 1730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7992912675700437,
+      "learning_rate": 7.146166945014102e-06,
+      "loss": 0.2681,
+      "step": 1731
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7997100711589723,
+      "learning_rate": 7.142971062984563e-06,
+      "loss": 0.2743,
+      "step": 1732
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8452702014425518,
+      "learning_rate": 7.139774108077633e-06,
+      "loss": 0.297,
+      "step": 1733
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8803705593600958,
+      "learning_rate": 7.136576081893863e-06,
+      "loss": 0.2635,
+      "step": 1734
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3381947762535928,
+      "learning_rate": 7.133376986034346e-06,
+      "loss": 0.2699,
+      "step": 1735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8235193598185109,
+      "learning_rate": 7.130176822100707e-06,
+      "loss": 0.2663,
+      "step": 1736
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8329521093086811,
+      "learning_rate": 7.126975591695108e-06,
+      "loss": 0.2747,
+      "step": 1737
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8443968589859789,
+      "learning_rate": 7.12377329642024e-06,
+      "loss": 0.278,
+      "step": 1738
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8015433750335496,
+      "learning_rate": 7.120569937879333e-06,
+      "loss": 0.2847,
+      "step": 1739
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7532719289044695,
+      "learning_rate": 7.117365517676146e-06,
+      "loss": 0.2762,
+      "step": 1740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7816753854682754,
+      "learning_rate": 7.11416003741497e-06,
+      "loss": 0.2654,
+      "step": 1741
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.966097320855663,
+      "learning_rate": 7.110953498700629e-06,
+      "loss": 0.2578,
+      "step": 1742
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8226332052242857,
+      "learning_rate": 7.107745903138472e-06,
+      "loss": 0.3025,
+      "step": 1743
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9126348462186298,
+      "learning_rate": 7.104537252334383e-06,
+      "loss": 0.2811,
+      "step": 1744
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9577656041985475,
+      "learning_rate": 7.101327547894771e-06,
+      "loss": 0.2725,
+      "step": 1745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3082857881291217,
+      "learning_rate": 7.09811679142657e-06,
+      "loss": 0.2572,
+      "step": 1746
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7850170117558714,
+      "learning_rate": 7.0949049845372475e-06,
+      "loss": 0.2709,
+      "step": 1747
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7777703920593871,
+      "learning_rate": 7.09169212883479e-06,
+      "loss": 0.2777,
+      "step": 1748
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7747593866541603,
+      "learning_rate": 7.0884782259277155e-06,
+      "loss": 0.2573,
+      "step": 1749
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7423127884285942,
+      "learning_rate": 7.085263277425058e-06,
+      "loss": 0.2689,
+      "step": 1750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7334847739908479,
+      "learning_rate": 7.082047284936385e-06,
+      "loss": 0.2643,
+      "step": 1751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7724101387532805,
+      "learning_rate": 7.078830250071778e-06,
+      "loss": 0.2685,
+      "step": 1752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7596037394991769,
+      "learning_rate": 7.075612174441846e-06,
+      "loss": 0.2603,
+      "step": 1753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7790606862172679,
+      "learning_rate": 7.072393059657716e-06,
+      "loss": 0.3022,
+      "step": 1754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8417009159575106,
+      "learning_rate": 7.069172907331034e-06,
+      "loss": 0.2594,
+      "step": 1755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7848440208059229,
+      "learning_rate": 7.065951719073973e-06,
+      "loss": 0.2784,
+      "step": 1756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7654708115011382,
+      "learning_rate": 7.0627294964992145e-06,
+      "loss": 0.2758,
+      "step": 1757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9880071503210487,
+      "learning_rate": 7.059506241219964e-06,
+      "loss": 0.2862,
+      "step": 1758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1086682805987622,
+      "learning_rate": 7.056281954849944e-06,
+      "loss": 0.2745,
+      "step": 1759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8650563621151435,
+      "learning_rate": 7.053056639003389e-06,
+      "loss": 0.2863,
+      "step": 1760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7562655672704935,
+      "learning_rate": 7.049830295295057e-06,
+      "loss": 0.2482,
+      "step": 1761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.083168531486479,
+      "learning_rate": 7.046602925340209e-06,
+      "loss": 0.265,
+      "step": 1762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7449063938230934,
+      "learning_rate": 7.04337453075463e-06,
+      "loss": 0.2727,
+      "step": 1763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7634152582000389,
+      "learning_rate": 7.0401451131546125e-06,
+      "loss": 0.2668,
+      "step": 1764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7548787664162825,
+      "learning_rate": 7.036914674156964e-06,
+      "loss": 0.274,
+      "step": 1765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.742632374638571,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 0.2675,
+      "step": 1766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0586895549924933,
+      "learning_rate": 7.0304507384385535e-06,
+      "loss": 0.2521,
+      "step": 1767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8572378947751754,
+      "learning_rate": 7.027217244953958e-06,
+      "loss": 0.2782,
+      "step": 1768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7257051414320815,
+      "learning_rate": 7.023982736544063e-06,
+      "loss": 0.2785,
+      "step": 1769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7839455933959433,
+      "learning_rate": 7.020747214828221e-06,
+      "loss": 0.2882,
+      "step": 1770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8173870359340352,
+      "learning_rate": 7.017510681426296e-06,
+      "loss": 0.297,
+      "step": 1771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8052514618158552,
+      "learning_rate": 7.014273137958656e-06,
+      "loss": 0.2814,
+      "step": 1772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7869281298253242,
+      "learning_rate": 7.011034586046177e-06,
+      "loss": 0.2646,
+      "step": 1773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8595998774693434,
+      "learning_rate": 7.007795027310236e-06,
+      "loss": 0.283,
+      "step": 1774
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8184672550643898,
+      "learning_rate": 7.004554463372719e-06,
+      "loss": 0.2863,
+      "step": 1775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7369062707380489,
+      "learning_rate": 7.001312895856011e-06,
+      "loss": 0.2734,
+      "step": 1776
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9860709238278139,
+      "learning_rate": 6.998070326383002e-06,
+      "loss": 0.2939,
+      "step": 1777
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8513690987541971,
+      "learning_rate": 6.994826756577082e-06,
+      "loss": 0.2691,
+      "step": 1778
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.916520336039129,
+      "learning_rate": 6.991582188062144e-06,
+      "loss": 0.2678,
+      "step": 1779
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7894836855927121,
+      "learning_rate": 6.988336622462578e-06,
+      "loss": 0.275,
+      "step": 1780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8292329254892908,
+      "learning_rate": 6.985090061403277e-06,
+      "loss": 0.2894,
+      "step": 1781
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6812795197912004,
+      "learning_rate": 6.981842506509626e-06,
+      "loss": 0.2612,
+      "step": 1782
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7997102496120534,
+      "learning_rate": 6.978593959407516e-06,
+      "loss": 0.2938,
+      "step": 1783
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7128814375105025,
+      "learning_rate": 6.975344421723329e-06,
+      "loss": 0.2667,
+      "step": 1784
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8668458097932082,
+      "learning_rate": 6.9720938950839445e-06,
+      "loss": 0.2666,
+      "step": 1785
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8228736102644899,
+      "learning_rate": 6.968842381116736e-06,
+      "loss": 0.2749,
+      "step": 1786
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7965857801165129,
+      "learning_rate": 6.965589881449574e-06,
+      "loss": 0.2642,
+      "step": 1787
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9168993268189636,
+      "learning_rate": 6.962336397710819e-06,
+      "loss": 0.2888,
+      "step": 1788
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.830322887198747,
+      "learning_rate": 6.959081931529328e-06,
+      "loss": 0.2668,
+      "step": 1789
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7949806902805615,
+      "learning_rate": 6.955826484534447e-06,
+      "loss": 0.2753,
+      "step": 1790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9214232244930192,
+      "learning_rate": 6.952570058356013e-06,
+      "loss": 0.2652,
+      "step": 1791
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9131276692803154,
+      "learning_rate": 6.949312654624355e-06,
+      "loss": 0.3172,
+      "step": 1792
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8150224385280416,
+      "learning_rate": 6.946054274970292e-06,
+      "loss": 0.3072,
+      "step": 1793
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7749855939811231,
+      "learning_rate": 6.942794921025127e-06,
+      "loss": 0.2821,
+      "step": 1794
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8077931806939087,
+      "learning_rate": 6.939534594420656e-06,
+      "loss": 0.2661,
+      "step": 1795
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7862115364874112,
+      "learning_rate": 6.93627329678916e-06,
+      "loss": 0.2979,
+      "step": 1796
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8619674812365525,
+      "learning_rate": 6.933011029763406e-06,
+      "loss": 0.2837,
+      "step": 1797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7840104909694408,
+      "learning_rate": 6.9297477949766445e-06,
+      "loss": 0.2483,
+      "step": 1798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7977129103627658,
+      "learning_rate": 6.9264835940626145e-06,
+      "loss": 0.2948,
+      "step": 1799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7148961274335996,
+      "learning_rate": 6.9232184286555345e-06,
+      "loss": 0.2529,
+      "step": 1800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8394266226375869,
+      "learning_rate": 6.919952300390111e-06,
+      "loss": 0.252,
+      "step": 1801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8647538642812049,
+      "learning_rate": 6.916685210901527e-06,
+      "loss": 0.2947,
+      "step": 1802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7707564025979443,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.2712,
+      "step": 1803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7622082843572645,
+      "learning_rate": 6.910148154798026e-06,
+      "loss": 0.2686,
+      "step": 1804
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8765003898419722,
+      "learning_rate": 6.906878191455883e-06,
+      "loss": 0.2756,
+      "step": 1805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0124750528256148,
+      "learning_rate": 6.903607273436128e-06,
+      "loss": 0.2767,
+      "step": 1806
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8185629739899599,
+      "learning_rate": 6.900335402376341e-06,
+      "loss": 0.2872,
+      "step": 1807
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8415423484000762,
+      "learning_rate": 6.897062579914587e-06,
+      "loss": 0.2762,
+      "step": 1808
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7860415540739579,
+      "learning_rate": 6.8937888076893964e-06,
+      "loss": 0.2673,
+      "step": 1809
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8385343885368797,
+      "learning_rate": 6.890514087339789e-06,
+      "loss": 0.297,
+      "step": 1810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8033313099479813,
+      "learning_rate": 6.887238420505247e-06,
+      "loss": 0.2837,
+      "step": 1811
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8248889294699653,
+      "learning_rate": 6.883961808825732e-06,
+      "loss": 0.274,
+      "step": 1812
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.719207790825955,
+      "learning_rate": 6.88068425394168e-06,
+      "loss": 0.2734,
+      "step": 1813
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7146497229696308,
+      "learning_rate": 6.8774057574939956e-06,
+      "loss": 0.2651,
+      "step": 1814
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8202034276742195,
+      "learning_rate": 6.874126321124057e-06,
+      "loss": 0.2639,
+      "step": 1815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7829423708533703,
+      "learning_rate": 6.870845946473711e-06,
+      "loss": 0.275,
+      "step": 1816
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0178059787684202,
+      "learning_rate": 6.867564635185279e-06,
+      "loss": 0.2824,
+      "step": 1817
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.745119054195976,
+      "learning_rate": 6.864282388901544e-06,
+      "loss": 0.2977,
+      "step": 1818
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7171465961064905,
+      "learning_rate": 6.860999209265762e-06,
+      "loss": 0.267,
+      "step": 1819
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7966040075301501,
+      "learning_rate": 6.857715097921657e-06,
+      "loss": 0.2814,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4183129589995005,
+      "learning_rate": 6.854430056513417e-06,
+      "loss": 0.284,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9038932031509079,
+      "learning_rate": 6.851144086685695e-06,
+      "loss": 0.2541,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7578467275659945,
+      "learning_rate": 6.847857190083611e-06,
+      "loss": 0.2712,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7892462624222709,
+      "learning_rate": 6.844569368352748e-06,
+      "loss": 0.2801,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8678061796466984,
+      "learning_rate": 6.841280623139153e-06,
+      "loss": 0.2765,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8395573882972761,
+      "learning_rate": 6.837990956089334e-06,
+      "loss": 0.2952,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7454373131926565,
+      "learning_rate": 6.8347003688502614e-06,
+      "loss": 0.2746,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8570184249089252,
+      "learning_rate": 6.831408863069364e-06,
+      "loss": 0.265,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.888339538630694,
+      "learning_rate": 6.828116440394536e-06,
+      "loss": 0.2632,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7788192528439788,
+      "learning_rate": 6.824823102474127e-06,
+      "loss": 0.2717,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8207375292871161,
+      "learning_rate": 6.821528850956942e-06,
+      "loss": 0.2574,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7946084202164586,
+      "learning_rate": 6.818233687492248e-06,
+      "loss": 0.2622,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.788804330204311,
+      "learning_rate": 6.814937613729766e-06,
+      "loss": 0.303,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7988487963833797,
+      "learning_rate": 6.811640631319677e-06,
+      "loss": 0.2726,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7574794517309196,
+      "learning_rate": 6.80834274191261e-06,
+      "loss": 0.2755,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7807050027133744,
+      "learning_rate": 6.805043947159652e-06,
+      "loss": 0.2654,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7529103967494202,
+      "learning_rate": 6.801744248712345e-06,
+      "loss": 0.2719,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7848078153696524,
+      "learning_rate": 6.79844364822268e-06,
+      "loss": 0.2752,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8684446587478822,
+      "learning_rate": 6.795142147343101e-06,
+      "loss": 0.2683,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8442360985237036,
+      "learning_rate": 6.7918397477265e-06,
+      "loss": 0.2847,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.780535626932052,
+      "learning_rate": 6.78853645102623e-06,
+      "loss": 0.2786,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7650445616789762,
+      "learning_rate": 6.785232258896078e-06,
+      "loss": 0.2643,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9419649042200724,
+      "learning_rate": 6.781927172990285e-06,
+      "loss": 0.2892,
+      "step": 1843
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.892508964609761,
+      "learning_rate": 6.778621194963545e-06,
+      "loss": 0.2717,
+      "step": 1844
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9422634188340857,
+      "learning_rate": 6.775314326470992e-06,
+      "loss": 0.3022,
+      "step": 1845
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7521225559892537,
+      "learning_rate": 6.77200656916821e-06,
+      "loss": 0.2664,
+      "step": 1846
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7999258127262665,
+      "learning_rate": 6.768697924711223e-06,
+      "loss": 0.2766,
+      "step": 1847
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6737896483256152,
+      "learning_rate": 6.765388394756504e-06,
+      "loss": 0.2685,
+      "step": 1848
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8303578358852908,
+      "learning_rate": 6.762077980960968e-06,
+      "loss": 0.2945,
+      "step": 1849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8629177938777671,
+      "learning_rate": 6.758766684981972e-06,
+      "loss": 0.2869,
+      "step": 1850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7406970417383174,
+      "learning_rate": 6.755454508477313e-06,
+      "loss": 0.2511,
+      "step": 1851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8677940754508966,
+      "learning_rate": 6.752141453105231e-06,
+      "loss": 0.283,
+      "step": 1852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8269559600582462,
+      "learning_rate": 6.748827520524406e-06,
+      "loss": 0.2683,
+      "step": 1853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9377701714954148,
+      "learning_rate": 6.745512712393958e-06,
+      "loss": 0.2731,
+      "step": 1854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9287764368018571,
+      "learning_rate": 6.742197030373442e-06,
+      "loss": 0.2812,
+      "step": 1855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9164382369601928,
+      "learning_rate": 6.7388804761228535e-06,
+      "loss": 0.2902,
+      "step": 1856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.143153495041861,
+      "learning_rate": 6.735563051302622e-06,
+      "loss": 0.2727,
+      "step": 1857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7868890044582834,
+      "learning_rate": 6.732244757573619e-06,
+      "loss": 0.2481,
+      "step": 1858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8313733613114832,
+      "learning_rate": 6.728925596597142e-06,
+      "loss": 0.2715,
+      "step": 1859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7464116261859559,
+      "learning_rate": 6.725605570034928e-06,
+      "loss": 0.2705,
+      "step": 1860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7666626884522852,
+      "learning_rate": 6.722284679549149e-06,
+      "loss": 0.2849,
+      "step": 1861
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7266600929822588,
+      "learning_rate": 6.718962926802407e-06,
+      "loss": 0.2647,
+      "step": 1862
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8356093445961695,
+      "learning_rate": 6.715640313457733e-06,
+      "loss": 0.286,
+      "step": 1863
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8380893349411164,
+      "learning_rate": 6.712316841178592e-06,
+      "loss": 0.2982,
+      "step": 1864
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.800416001030339,
+      "learning_rate": 6.708992511628882e-06,
+      "loss": 0.2835,
+      "step": 1865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8193761295343053,
+      "learning_rate": 6.705667326472926e-06,
+      "loss": 0.2788,
+      "step": 1866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7828802633508328,
+      "learning_rate": 6.702341287375473e-06,
+      "loss": 0.2633,
+      "step": 1867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8433760463561417,
+      "learning_rate": 6.699014396001707e-06,
+      "loss": 0.263,
+      "step": 1868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8490887170241008,
+      "learning_rate": 6.6956866540172305e-06,
+      "loss": 0.2552,
+      "step": 1869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7549822532217539,
+      "learning_rate": 6.69235806308808e-06,
+      "loss": 0.2892,
+      "step": 1870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8367731564117415,
+      "learning_rate": 6.6890286248807104e-06,
+      "loss": 0.2769,
+      "step": 1871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8381152607081171,
+      "learning_rate": 6.685698341062003e-06,
+      "loss": 0.2962,
+      "step": 1872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9453814822328721,
+      "learning_rate": 6.682367213299264e-06,
+      "loss": 0.2823,
+      "step": 1873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7945405345997448,
+      "learning_rate": 6.67903524326022e-06,
+      "loss": 0.2775,
+      "step": 1874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7836085992046492,
+      "learning_rate": 6.6757024326130204e-06,
+      "loss": 0.273,
+      "step": 1875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5210450070767783,
+      "learning_rate": 6.672368783026235e-06,
+      "loss": 0.2639,
+      "step": 1876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8028067484360977,
+      "learning_rate": 6.669034296168855e-06,
+      "loss": 0.2705,
+      "step": 1877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.827860740063605,
+      "learning_rate": 6.665698973710289e-06,
+      "loss": 0.2989,
+      "step": 1878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8321291306582843,
+      "learning_rate": 6.662362817320364e-06,
+      "loss": 0.2659,
+      "step": 1879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8821140873532841,
+      "learning_rate": 6.659025828669323e-06,
+      "loss": 0.2684,
+      "step": 1880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7957083911264379,
+      "learning_rate": 6.655688009427832e-06,
+      "loss": 0.2806,
+      "step": 1881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7812235608288512,
+      "learning_rate": 6.652349361266967e-06,
+      "loss": 0.2869,
+      "step": 1882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.82021132084765,
+      "learning_rate": 6.6490098858582176e-06,
+      "loss": 0.278,
+      "step": 1883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2872008527272985,
+      "learning_rate": 6.645669584873495e-06,
+      "loss": 0.2835,
+      "step": 1884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7509231169951046,
+      "learning_rate": 6.6423284599851145e-06,
+      "loss": 0.2663,
+      "step": 1885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7360512683589652,
+      "learning_rate": 6.638986512865814e-06,
+      "loss": 0.2577,
+      "step": 1886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8332534236687871,
+      "learning_rate": 6.6356437451887336e-06,
+      "loss": 0.3042,
+      "step": 1887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8068581172775373,
+      "learning_rate": 6.632300158627427e-06,
+      "loss": 0.2887,
+      "step": 1888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7616846681647065,
+      "learning_rate": 6.628955754855865e-06,
+      "loss": 0.3127,
+      "step": 1889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9681257711607067,
+      "learning_rate": 6.625610535548418e-06,
+      "loss": 0.2569,
+      "step": 1890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8059726315439394,
+      "learning_rate": 6.6222645023798675e-06,
+      "loss": 0.2727,
+      "step": 1891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7981124968480723,
+      "learning_rate": 6.6189176570254045e-06,
+      "loss": 0.2511,
+      "step": 1892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7586502928931304,
+      "learning_rate": 6.615570001160626e-06,
+      "loss": 0.2781,
+      "step": 1893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.78167905303485,
+      "learning_rate": 6.612221536461534e-06,
+      "loss": 0.3089,
+      "step": 1894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8497070953330546,
+      "learning_rate": 6.608872264604536e-06,
+      "loss": 0.2795,
+      "step": 1895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1556373297858538,
+      "learning_rate": 6.605522187266441e-06,
+      "loss": 0.2889,
+      "step": 1896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8411605518280044,
+      "learning_rate": 6.6021713061244655e-06,
+      "loss": 0.2939,
+      "step": 1897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7737900951221209,
+      "learning_rate": 6.598819622856227e-06,
+      "loss": 0.265,
+      "step": 1898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7232453558705487,
+      "learning_rate": 6.595467139139743e-06,
+      "loss": 0.2547,
+      "step": 1899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8052265879126966,
+      "learning_rate": 6.592113856653433e-06,
+      "loss": 0.2646,
+      "step": 1900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8728065026014801,
+      "learning_rate": 6.588759777076116e-06,
+      "loss": 0.2509,
+      "step": 1901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8224284336840014,
+      "learning_rate": 6.585404902087011e-06,
+      "loss": 0.2507,
+      "step": 1902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7620686410401272,
+      "learning_rate": 6.582049233365734e-06,
+      "loss": 0.2729,
+      "step": 1903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7735297959874908,
+      "learning_rate": 6.578692772592298e-06,
+      "loss": 0.27,
+      "step": 1904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7349337535756242,
+      "learning_rate": 6.575335521447114e-06,
+      "loss": 0.2881,
+      "step": 1905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8268736935444758,
+      "learning_rate": 6.571977481610989e-06,
+      "loss": 0.2748,
+      "step": 1906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7983209701331083,
+      "learning_rate": 6.568618654765124e-06,
+      "loss": 0.2761,
+      "step": 1907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9615029615894742,
+      "learning_rate": 6.565259042591112e-06,
+      "loss": 0.2739,
+      "step": 1908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8878354806330305,
+      "learning_rate": 6.561898646770943e-06,
+      "loss": 0.2715,
+      "step": 1909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7535030892067545,
+      "learning_rate": 6.558537468986998e-06,
+      "loss": 0.2502,
+      "step": 1910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7635946368354521,
+      "learning_rate": 6.555175510922047e-06,
+      "loss": 0.2745,
+      "step": 1911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9890178138302,
+      "learning_rate": 6.551812774259253e-06,
+      "loss": 0.3,
+      "step": 1912
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8568170141477122,
+      "learning_rate": 6.548449260682169e-06,
+      "loss": 0.2656,
+      "step": 1913
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8906011600657048,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.2728,
+      "step": 1914
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7793111572880544,
+      "learning_rate": 6.541719909521287e-06,
+      "loss": 0.2884,
+      "step": 1915
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.741957953874454,
+      "learning_rate": 6.538354075306532e-06,
+      "loss": 0.2582,
+      "step": 1916
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8296756499941067,
+      "learning_rate": 6.5349874709155805e-06,
+      "loss": 0.2739,
+      "step": 1917
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7520586327671879,
+      "learning_rate": 6.531620098033919e-06,
+      "loss": 0.2583,
+      "step": 1918
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6871387684882602,
+      "learning_rate": 6.528251958347418e-06,
+      "loss": 0.2699,
+      "step": 1919
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8805085669637309,
+      "learning_rate": 6.524883053542339e-06,
+      "loss": 0.2691,
+      "step": 1920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7595934769114278,
+      "learning_rate": 6.521513385305321e-06,
+      "loss": 0.252,
+      "step": 1921
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8397439758201035,
+      "learning_rate": 6.518142955323389e-06,
+      "loss": 0.2582,
+      "step": 1922
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0213150950350063,
+      "learning_rate": 6.514771765283942e-06,
+      "loss": 0.2596,
+      "step": 1923
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4497694153903509,
+      "learning_rate": 6.511399816874769e-06,
+      "loss": 0.2811,
+      "step": 1924
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9758452040494925,
+      "learning_rate": 6.508027111784033e-06,
+      "loss": 0.2809,
+      "step": 1925
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8201366636138401,
+      "learning_rate": 6.504653651700278e-06,
+      "loss": 0.2742,
+      "step": 1926
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8163781034218034,
+      "learning_rate": 6.501279438312422e-06,
+      "loss": 0.2705,
+      "step": 1927
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.442731718961944,
+      "learning_rate": 6.497904473309766e-06,
+      "loss": 0.2775,
+      "step": 1928
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9471225557430727,
+      "learning_rate": 6.494528758381985e-06,
+      "loss": 0.2654,
+      "step": 1929
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.705192764524926,
+      "learning_rate": 6.4911522952191275e-06,
+      "loss": 0.2624,
+      "step": 1930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7806733221788663,
+      "learning_rate": 6.487775085511618e-06,
+      "loss": 0.2797,
+      "step": 1931
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7833539052114034,
+      "learning_rate": 6.4843971309502554e-06,
+      "loss": 0.2768,
+      "step": 1932
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8968618303149174,
+      "learning_rate": 6.481018433226212e-06,
+      "loss": 0.2681,
+      "step": 1933
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.829569739111911,
+      "learning_rate": 6.477638994031028e-06,
+      "loss": 0.2898,
+      "step": 1934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7874610463109872,
+      "learning_rate": 6.474258815056622e-06,
+      "loss": 0.2892,
+      "step": 1935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9204034921041345,
+      "learning_rate": 6.470877897995275e-06,
+      "loss": 0.2894,
+      "step": 1936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7550783124798898,
+      "learning_rate": 6.467496244539644e-06,
+      "loss": 0.2616,
+      "step": 1937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.750579957195619,
+      "learning_rate": 6.464113856382752e-06,
+      "loss": 0.2939,
+      "step": 1938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7634010684442557,
+      "learning_rate": 6.460730735217991e-06,
+      "loss": 0.2641,
+      "step": 1939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9038654570644069,
+      "learning_rate": 6.457346882739114e-06,
+      "loss": 0.2668,
+      "step": 1940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6826281737859484,
+      "learning_rate": 6.45396230064025e-06,
+      "loss": 0.2518,
+      "step": 1941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8847783285124754,
+      "learning_rate": 6.450576990615888e-06,
+      "loss": 0.2942,
+      "step": 1942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8105215054519787,
+      "learning_rate": 6.447190954360878e-06,
+      "loss": 0.2788,
+      "step": 1943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9038720068909176,
+      "learning_rate": 6.443804193570441e-06,
+      "loss": 0.2738,
+      "step": 1944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9133673072925703,
+      "learning_rate": 6.440416709940156e-06,
+      "loss": 0.2841,
+      "step": 1945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8720313388806107,
+      "learning_rate": 6.437028505165965e-06,
+      "loss": 0.2799,
+      "step": 1946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.903474577062172,
+      "learning_rate": 6.433639580944171e-06,
+      "loss": 0.2702,
+      "step": 1947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7745384218858543,
+      "learning_rate": 6.430249938971438e-06,
+      "loss": 0.2601,
+      "step": 1948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7741128471121047,
+      "learning_rate": 6.426859580944786e-06,
+      "loss": 0.2717,
+      "step": 1949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8146960073234596,
+      "learning_rate": 6.423468508561599e-06,
+      "loss": 0.2695,
+      "step": 1950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8504302141590706,
+      "learning_rate": 6.420076723519615e-06,
+      "loss": 0.278,
+      "step": 1951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7393646851106714,
+      "learning_rate": 6.416684227516927e-06,
+      "loss": 0.2405,
+      "step": 1952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8379847208130288,
+      "learning_rate": 6.41329102225199e-06,
+      "loss": 0.249,
+      "step": 1953
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7851261702460455,
+      "learning_rate": 6.409897109423607e-06,
+      "loss": 0.268,
+      "step": 1954
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8585224719390991,
+      "learning_rate": 6.406502490730944e-06,
+      "loss": 0.3158,
+      "step": 1955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8819476655934881,
+      "learning_rate": 6.403107167873509e-06,
+      "loss": 0.2718,
+      "step": 1956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7800179717972927,
+      "learning_rate": 6.399711142551172e-06,
+      "loss": 0.2572,
+      "step": 1957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.800849314929032,
+      "learning_rate": 6.396314416464151e-06,
+      "loss": 0.2697,
+      "step": 1958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7888300070002772,
+      "learning_rate": 6.3929169913130165e-06,
+      "loss": 0.2808,
+      "step": 1959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7888849793498722,
+      "learning_rate": 6.389518868798683e-06,
+      "loss": 0.2859,
+      "step": 1960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7589509868669029,
+      "learning_rate": 6.3861200506224255e-06,
+      "loss": 0.2836,
+      "step": 1961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.816029521595813,
+      "learning_rate": 6.382720538485856e-06,
+      "loss": 0.2909,
+      "step": 1962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8858089203153157,
+      "learning_rate": 6.37932033409094e-06,
+      "loss": 0.2661,
+      "step": 1963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7909958706991513,
+      "learning_rate": 6.375919439139988e-06,
+      "loss": 0.2818,
+      "step": 1964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8079857526147861,
+      "learning_rate": 6.372517855335656e-06,
+      "loss": 0.2779,
+      "step": 1965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7505020906320573,
+      "learning_rate": 6.3691155843809475e-06,
+      "loss": 0.2825,
+      "step": 1966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8138540177520711,
+      "learning_rate": 6.365712627979204e-06,
+      "loss": 0.2897,
+      "step": 1967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9203103141819395,
+      "learning_rate": 6.3623089878341146e-06,
+      "loss": 0.2738,
+      "step": 1968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.803944971551273,
+      "learning_rate": 6.358904665649712e-06,
+      "loss": 0.2702,
+      "step": 1969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7980720641970154,
+      "learning_rate": 6.355499663130369e-06,
+      "loss": 0.285,
+      "step": 1970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7290648570694908,
+      "learning_rate": 6.352093981980796e-06,
+      "loss": 0.2604,
+      "step": 1971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8076229278294601,
+      "learning_rate": 6.348687623906047e-06,
+      "loss": 0.2845,
+      "step": 1972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8678888155320968,
+      "learning_rate": 6.345280590611512e-06,
+      "loss": 0.2439,
+      "step": 1973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7414348869951592,
+      "learning_rate": 6.341872883802923e-06,
+      "loss": 0.2622,
+      "step": 1974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7810241104076934,
+      "learning_rate": 6.338464505186344e-06,
+      "loss": 0.2879,
+      "step": 1975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.929127585108396,
+      "learning_rate": 6.33505545646818e-06,
+      "loss": 0.3007,
+      "step": 1976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9650697465964181,
+      "learning_rate": 6.331645739355168e-06,
+      "loss": 0.255,
+      "step": 1977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9403437826714144,
+      "learning_rate": 6.328235355554382e-06,
+      "loss": 0.2991,
+      "step": 1978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.800704754235717,
+      "learning_rate": 6.324824306773231e-06,
+      "loss": 0.2578,
+      "step": 1979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7687820188212391,
+      "learning_rate": 6.32141259471945e-06,
+      "loss": 0.2684,
+      "step": 1980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8910357906378141,
+      "learning_rate": 6.318000221101115e-06,
+      "loss": 0.2817,
+      "step": 1981
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8650932016350911,
+      "learning_rate": 6.314587187626628e-06,
+      "loss": 0.2622,
+      "step": 1982
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8080663830366237,
+      "learning_rate": 6.311173496004723e-06,
+      "loss": 0.2992,
+      "step": 1983
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8764811820271351,
+      "learning_rate": 6.307759147944461e-06,
+      "loss": 0.2775,
+      "step": 1984
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7517552155888431,
+      "learning_rate": 6.304344145155234e-06,
+      "loss": 0.2748,
+      "step": 1985
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7207423311741115,
+      "learning_rate": 6.3009284893467655e-06,
+      "loss": 0.281,
+      "step": 1986
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8417877405092894,
+      "learning_rate": 6.297512182229098e-06,
+      "loss": 0.2449,
+      "step": 1987
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7894778974805355,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.2732,
+      "step": 1988
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7769946160999486,
+      "learning_rate": 6.290677620907982e-06,
+      "loss": 0.2851,
+      "step": 1989
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8313518201634298,
+      "learning_rate": 6.287259370126255e-06,
+      "loss": 0.2907,
+      "step": 1990
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7679699099273584,
+      "learning_rate": 6.283840474878767e-06,
+      "loss": 0.2551,
+      "step": 1991
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7734633704835453,
+      "learning_rate": 6.280420936877187e-06,
+      "loss": 0.2954,
+      "step": 1992
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8188559057853843,
+      "learning_rate": 6.2770007578335044e-06,
+      "loss": 0.2672,
+      "step": 1993
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7776609075544186,
+      "learning_rate": 6.273579939460032e-06,
+      "loss": 0.2917,
+      "step": 1994
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7199092192397869,
+      "learning_rate": 6.270158483469398e-06,
+      "loss": 0.2599,
+      "step": 1995
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9997410416149503,
+      "learning_rate": 6.266736391574553e-06,
+      "loss": 0.276,
+      "step": 1996
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7687458527372526,
+      "learning_rate": 6.263313665488767e-06,
+      "loss": 0.2547,
+      "step": 1997
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7823333046536625,
+      "learning_rate": 6.259890306925627e-06,
+      "loss": 0.2558,
+      "step": 1998
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.764556849464287,
+      "learning_rate": 6.256466317599033e-06,
+      "loss": 0.2626,
+      "step": 1999
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7985798381255325,
+      "learning_rate": 6.253041699223206e-06,
+      "loss": 0.266,
+      "step": 2000
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9311606505338256,
+      "learning_rate": 6.2496164535126766e-06,
+      "loss": 0.272,
+      "step": 2001
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.789927511259,
+      "learning_rate": 6.246190582182296e-06,
+      "loss": 0.2625,
+      "step": 2002
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7950458545665098,
+      "learning_rate": 6.2427640869472235e-06,
+      "loss": 0.2766,
+      "step": 2003
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6977432468049745,
+      "learning_rate": 6.239336969522931e-06,
+      "loss": 0.2529,
+      "step": 2004
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9225036206313751,
+      "learning_rate": 6.2359092316252035e-06,
+      "loss": 0.2947,
+      "step": 2005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8254783229240075,
+      "learning_rate": 6.232480874970137e-06,
+      "loss": 0.2756,
+      "step": 2006
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7762690309071891,
+      "learning_rate": 6.229051901274137e-06,
+      "loss": 0.2904,
+      "step": 2007
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8858013656831054,
+      "learning_rate": 6.225622312253916e-06,
+      "loss": 0.274,
+      "step": 2008
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8267183095872146,
+      "learning_rate": 6.222192109626496e-06,
+      "loss": 0.2901,
+      "step": 2009
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8433367411362701,
+      "learning_rate": 6.218761295109209e-06,
+      "loss": 0.2677,
+      "step": 2010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8333395085806302,
+      "learning_rate": 6.215329870419688e-06,
+      "loss": 0.2893,
+      "step": 2011
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7944684080500283,
+      "learning_rate": 6.211897837275873e-06,
+      "loss": 0.2935,
+      "step": 2012
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9640872742520753,
+      "learning_rate": 6.208465197396013e-06,
+      "loss": 0.2713,
+      "step": 2013
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7448989363042914,
+      "learning_rate": 6.205031952498656e-06,
+      "loss": 0.2766,
+      "step": 2014
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.84903548285289,
+      "learning_rate": 6.201598104302653e-06,
+      "loss": 0.2905,
+      "step": 2015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2356936403047212,
+      "learning_rate": 6.198163654527161e-06,
+      "loss": 0.2676,
+      "step": 2016
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8463453970233744,
+      "learning_rate": 6.194728604891632e-06,
+      "loss": 0.2465,
+      "step": 2017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.829675786239781,
+      "learning_rate": 6.191292957115825e-06,
+      "loss": 0.2911,
+      "step": 2018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8128035786936424,
+      "learning_rate": 6.187856712919795e-06,
+      "loss": 0.2866,
+      "step": 2019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7303287330221199,
+      "learning_rate": 6.184419874023895e-06,
+      "loss": 0.2518,
+      "step": 2020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7830548906650386,
+      "learning_rate": 6.1809824421487765e-06,
+      "loss": 0.2769,
+      "step": 2021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7693266195584383,
+      "learning_rate": 6.177544419015388e-06,
+      "loss": 0.2902,
+      "step": 2022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9266499570716225,
+      "learning_rate": 6.174105806344975e-06,
+      "loss": 0.2956,
+      "step": 2023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7704802839582955,
+      "learning_rate": 6.170666605859077e-06,
+      "loss": 0.2932,
+      "step": 2024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8960468926506433,
+      "learning_rate": 6.1672268192795285e-06,
+      "loss": 0.2913,
+      "step": 2025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.132350121106818,
+      "learning_rate": 6.163786448328455e-06,
+      "loss": 0.2611,
+      "step": 2026
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7550013502146588,
+      "learning_rate": 6.160345494728278e-06,
+      "loss": 0.2523,
+      "step": 2027
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8783783231372921,
+      "learning_rate": 6.156903960201709e-06,
+      "loss": 0.2765,
+      "step": 2028
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7565325280389755,
+      "learning_rate": 6.15346184647175e-06,
+      "loss": 0.2547,
+      "step": 2029
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.850347238223973,
+      "learning_rate": 6.150019155261694e-06,
+      "loss": 0.2559,
+      "step": 2030
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.840292487530662,
+      "learning_rate": 6.1465758882951236e-06,
+      "loss": 0.2496,
+      "step": 2031
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7466516173120405,
+      "learning_rate": 6.1431320472959054e-06,
+      "loss": 0.2718,
+      "step": 2032
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8256035818167602,
+      "learning_rate": 6.1396876339882e-06,
+      "loss": 0.2689,
+      "step": 2033
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7594720743947756,
+      "learning_rate": 6.136242650096451e-06,
+      "loss": 0.2626,
+      "step": 2034
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7650417878183807,
+      "learning_rate": 6.132797097345387e-06,
+      "loss": 0.2925,
+      "step": 2035
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7939228478963118,
+      "learning_rate": 6.12935097746002e-06,
+      "loss": 0.3029,
+      "step": 2036
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8229450081174955,
+      "learning_rate": 6.125904292165652e-06,
+      "loss": 0.2708,
+      "step": 2037
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.744966687320915,
+      "learning_rate": 6.122457043187863e-06,
+      "loss": 0.2631,
+      "step": 2038
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8823127488271977,
+      "learning_rate": 6.119009232252517e-06,
+      "loss": 0.2656,
+      "step": 2039
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9472936836411072,
+      "learning_rate": 6.115560861085757e-06,
+      "loss": 0.2791,
+      "step": 2040
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9215595680764215,
+      "learning_rate": 6.112111931414009e-06,
+      "loss": 0.2735,
+      "step": 2041
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7185450802882857,
+      "learning_rate": 6.108662444963979e-06,
+      "loss": 0.2568,
+      "step": 2042
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7766986447032911,
+      "learning_rate": 6.10521240346265e-06,
+      "loss": 0.2727,
+      "step": 2043
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9327356004837255,
+      "learning_rate": 6.101761808637283e-06,
+      "loss": 0.2471,
+      "step": 2044
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.786672560218332,
+      "learning_rate": 6.098310662215416e-06,
+      "loss": 0.2755,
+      "step": 2045
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7825601240654209,
+      "learning_rate": 6.094858965924866e-06,
+      "loss": 0.2886,
+      "step": 2046
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8936639701334957,
+      "learning_rate": 6.091406721493724e-06,
+      "loss": 0.2465,
+      "step": 2047
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7935927828799876,
+      "learning_rate": 6.087953930650349e-06,
+      "loss": 0.2652,
+      "step": 2048
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9205406853684917,
+      "learning_rate": 6.084500595123383e-06,
+      "loss": 0.2974,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8621475808950727,
+      "learning_rate": 6.081046716641735e-06,
+      "loss": 0.2662,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8598806785004818,
+      "learning_rate": 6.077592296934592e-06,
+      "loss": 0.269,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.733905735726895,
+      "learning_rate": 6.0741373377314005e-06,
+      "loss": 0.2643,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8835583528135433,
+      "learning_rate": 6.070681840761889e-06,
+      "loss": 0.2735,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7818930932956059,
+      "learning_rate": 6.0672258077560475e-06,
+      "loss": 0.2797,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8887441672767806,
+      "learning_rate": 6.063769240444141e-06,
+      "loss": 0.2562,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8398557374993765,
+      "learning_rate": 6.060312140556694e-06,
+      "loss": 0.2566,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8390185627265141,
+      "learning_rate": 6.056854509824503e-06,
+      "loss": 0.2545,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7203024512347121,
+      "learning_rate": 6.053396349978632e-06,
+      "loss": 0.2728,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8636128786854753,
+      "learning_rate": 6.049937662750405e-06,
+      "loss": 0.2651,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7885933170918943,
+      "learning_rate": 6.046478449871409e-06,
+      "loss": 0.2675,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7690200628450568,
+      "learning_rate": 6.043018713073502e-06,
+      "loss": 0.2651,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7799529059451528,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 0.2671,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8537697998580078,
+      "learning_rate": 6.036097674649672e-06,
+      "loss": 0.3006,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7886127257224356,
+      "learning_rate": 6.032636376488763e-06,
+      "loss": 0.2842,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7032420463662117,
+      "learning_rate": 6.029174561338968e-06,
+      "loss": 0.2387,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7789554981986004,
+      "learning_rate": 6.025712230933443e-06,
+      "loss": 0.2659,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.833253259990681,
+      "learning_rate": 6.022249387005604e-06,
+      "loss": 0.2798,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7839597544367761,
+      "learning_rate": 6.018786031289119e-06,
+      "loss": 0.2702,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0470658416104062,
+      "learning_rate": 6.015322165517916e-06,
+      "loss": 0.2795,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.729156413185845,
+      "learning_rate": 6.0118577914261786e-06,
+      "loss": 0.2574,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7718412746085208,
+      "learning_rate": 6.008392910748345e-06,
+      "loss": 0.2604,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9527956873944406,
+      "learning_rate": 6.004927525219103e-06,
+      "loss": 0.2653,
+      "step": 2072
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7762192531075536,
+      "learning_rate": 6.001461636573397e-06,
+      "loss": 0.2502,
+      "step": 2073
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8289772677280964,
+      "learning_rate": 5.997995246546424e-06,
+      "loss": 0.285,
+      "step": 2074
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7508100644663498,
+      "learning_rate": 5.9945283568736305e-06,
+      "loss": 0.263,
+      "step": 2075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7455328617874042,
+      "learning_rate": 5.9910609692907105e-06,
+      "loss": 0.2547,
+      "step": 2076
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9870420257243928,
+      "learning_rate": 5.9875930855336116e-06,
+      "loss": 0.3175,
+      "step": 2077
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8098462529311664,
+      "learning_rate": 5.984124707338528e-06,
+      "loss": 0.2714,
+      "step": 2078
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8290767586952471,
+      "learning_rate": 5.980655836441902e-06,
+      "loss": 0.2673,
+      "step": 2079
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8643673115499309,
+      "learning_rate": 5.977186474580421e-06,
+      "loss": 0.3146,
+      "step": 2080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7679716865583694,
+      "learning_rate": 5.973716623491019e-06,
+      "loss": 0.2765,
+      "step": 2081
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9527538477783469,
+      "learning_rate": 5.970246284910877e-06,
+      "loss": 0.2675,
+      "step": 2082
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6868079419252217,
+      "learning_rate": 5.966775460577418e-06,
+      "loss": 0.2472,
+      "step": 2083
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7737972877298808,
+      "learning_rate": 5.963304152228307e-06,
+      "loss": 0.2564,
+      "step": 2084
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9853714992985939,
+      "learning_rate": 5.959832361601453e-06,
+      "loss": 0.2722,
+      "step": 2085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3735257957914084,
+      "learning_rate": 5.9563600904350074e-06,
+      "loss": 0.2639,
+      "step": 2086
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8271469139947053,
+      "learning_rate": 5.95288734046736e-06,
+      "loss": 0.2893,
+      "step": 2087
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7890316103301945,
+      "learning_rate": 5.949414113437142e-06,
+      "loss": 0.2599,
+      "step": 2088
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7758322724873867,
+      "learning_rate": 5.945940411083222e-06,
+      "loss": 0.2654,
+      "step": 2089
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7572324449681874,
+      "learning_rate": 5.942466235144708e-06,
+      "loss": 0.2972,
+      "step": 2090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7851221726758572,
+      "learning_rate": 5.9389915873609465e-06,
+      "loss": 0.2715,
+      "step": 2091
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9719941799160116,
+      "learning_rate": 5.935516469471514e-06,
+      "loss": 0.2881,
+      "step": 2092
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.855372903764549,
+      "learning_rate": 5.932040883216228e-06,
+      "loss": 0.2939,
+      "step": 2093
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7740802542655721,
+      "learning_rate": 5.9285648303351404e-06,
+      "loss": 0.2565,
+      "step": 2094
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7734634674892804,
+      "learning_rate": 5.9250883125685355e-06,
+      "loss": 0.2717,
+      "step": 2095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9120433895977362,
+      "learning_rate": 5.921611331656928e-06,
+      "loss": 0.2669,
+      "step": 2096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7554345967763235,
+      "learning_rate": 5.918133889341067e-06,
+      "loss": 0.2888,
+      "step": 2097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8189322493975437,
+      "learning_rate": 5.914655987361934e-06,
+      "loss": 0.2877,
+      "step": 2098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7506114792985991,
+      "learning_rate": 5.911177627460739e-06,
+      "loss": 0.2601,
+      "step": 2099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7915823818144307,
+      "learning_rate": 5.907698811378919e-06,
+      "loss": 0.2754,
+      "step": 2100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6954985031137924,
+      "learning_rate": 5.904219540858143e-06,
+      "loss": 0.2345,
+      "step": 2101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6701060653662853,
+      "learning_rate": 5.900739817640304e-06,
+      "loss": 0.2305,
+      "step": 2102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7516978160680381,
+      "learning_rate": 5.897259643467528e-06,
+      "loss": 0.2477,
+      "step": 2103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8995772524400262,
+      "learning_rate": 5.893779020082159e-06,
+      "loss": 0.3027,
+      "step": 2104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8414452275457881,
+      "learning_rate": 5.890297949226768e-06,
+      "loss": 0.2876,
+      "step": 2105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8104055032679528,
+      "learning_rate": 5.886816432644155e-06,
+      "loss": 0.2706,
+      "step": 2106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.776211744681916,
+      "learning_rate": 5.88333447207734e-06,
+      "loss": 0.2607,
+      "step": 2107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.757570776287936,
+      "learning_rate": 5.8798520692695605e-06,
+      "loss": 0.2665,
+      "step": 2108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7270412697019871,
+      "learning_rate": 5.876369225964283e-06,
+      "loss": 0.2737,
+      "step": 2109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7850741399368211,
+      "learning_rate": 5.872885943905191e-06,
+      "loss": 0.2789,
+      "step": 2110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8155483380300721,
+      "learning_rate": 5.869402224836187e-06,
+      "loss": 0.2607,
+      "step": 2111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7296102050516866,
+      "learning_rate": 5.865918070501394e-06,
+      "loss": 0.2558,
+      "step": 2112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.742934384511311,
+      "learning_rate": 5.862433482645151e-06,
+      "loss": 0.2791,
+      "step": 2113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7853161589909201,
+      "learning_rate": 5.858948463012015e-06,
+      "loss": 0.2772,
+      "step": 2114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7328315779717333,
+      "learning_rate": 5.855463013346763e-06,
+      "loss": 0.2693,
+      "step": 2115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7813581598458009,
+      "learning_rate": 5.851977135394379e-06,
+      "loss": 0.2799,
+      "step": 2116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8078524709701174,
+      "learning_rate": 5.848490830900067e-06,
+      "loss": 0.2761,
+      "step": 2117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.767116554075408,
+      "learning_rate": 5.8450041016092465e-06,
+      "loss": 0.2707,
+      "step": 2118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.719967890151342,
+      "learning_rate": 5.841516949267544e-06,
+      "loss": 0.2766,
+      "step": 2119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8211024663148292,
+      "learning_rate": 5.838029375620801e-06,
+      "loss": 0.2711,
+      "step": 2120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8516051550949817,
+      "learning_rate": 5.83454138241507e-06,
+      "loss": 0.2634,
+      "step": 2121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7181654914649876,
+      "learning_rate": 5.831052971396613e-06,
+      "loss": 0.2629,
+      "step": 2122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7579034860488993,
+      "learning_rate": 5.8275641443119015e-06,
+      "loss": 0.2743,
+      "step": 2123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.728816555978512,
+      "learning_rate": 5.824074902907614e-06,
+      "loss": 0.2743,
+      "step": 2124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7453088975759964,
+      "learning_rate": 5.820585248930638e-06,
+      "loss": 0.2631,
+      "step": 2125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7946848864211161,
+      "learning_rate": 5.817095184128067e-06,
+      "loss": 0.2746,
+      "step": 2126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7517575049612089,
+      "learning_rate": 5.813604710247201e-06,
+      "loss": 0.2513,
+      "step": 2127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.741774229313437,
+      "learning_rate": 5.810113829035544e-06,
+      "loss": 0.2825,
+      "step": 2128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7097743330396422,
+      "learning_rate": 5.806622542240801e-06,
+      "loss": 0.2657,
+      "step": 2129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7800579233203112,
+      "learning_rate": 5.803130851610887e-06,
+      "loss": 0.2643,
+      "step": 2130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9756835439174829,
+      "learning_rate": 5.799638758893914e-06,
+      "loss": 0.2703,
+      "step": 2131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7809444695209218,
+      "learning_rate": 5.796146265838195e-06,
+      "loss": 0.2642,
+      "step": 2132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.860937387545592,
+      "learning_rate": 5.792653374192245e-06,
+      "loss": 0.2948,
+      "step": 2133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8005736799830491,
+      "learning_rate": 5.78916008570478e-06,
+      "loss": 0.2805,
+      "step": 2134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.860258044620671,
+      "learning_rate": 5.785666402124715e-06,
+      "loss": 0.2673,
+      "step": 2135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7977921338913065,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.262,
+      "step": 2136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8015876984295568,
+      "learning_rate": 5.778677856683412e-06,
+      "loss": 0.2764,
+      "step": 2137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6958835658681575,
+      "learning_rate": 5.77518299832099e-06,
+      "loss": 0.2621,
+      "step": 2138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7851400785827746,
+      "learning_rate": 5.771687751863587e-06,
+      "loss": 0.2715,
+      "step": 2139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7914702613906611,
+      "learning_rate": 5.768192119061096e-06,
+      "loss": 0.2689,
+      "step": 2140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.961084436441799,
+      "learning_rate": 5.764696101663605e-06,
+      "loss": 0.2818,
+      "step": 2141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7102163630179585,
+      "learning_rate": 5.761199701421392e-06,
+      "loss": 0.2607,
+      "step": 2142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8228307278110318,
+      "learning_rate": 5.757702920084931e-06,
+      "loss": 0.2619,
+      "step": 2143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7757782104907147,
+      "learning_rate": 5.754205759404879e-06,
+      "loss": 0.2687,
+      "step": 2144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7488466473559799,
+      "learning_rate": 5.750708221132092e-06,
+      "loss": 0.2793,
+      "step": 2145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7433648874958508,
+      "learning_rate": 5.747210307017609e-06,
+      "loss": 0.2421,
+      "step": 2146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7504125311668597,
+      "learning_rate": 5.743712018812659e-06,
+      "loss": 0.273,
+      "step": 2147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7814116775927623,
+      "learning_rate": 5.740213358268658e-06,
+      "loss": 0.2621,
+      "step": 2148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7795843837563964,
+      "learning_rate": 5.73671432713721e-06,
+      "loss": 0.263,
+      "step": 2149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9004200283358184,
+      "learning_rate": 5.733214927170101e-06,
+      "loss": 0.2792,
+      "step": 2150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.745312208838729,
+      "learning_rate": 5.729715160119306e-06,
+      "loss": 0.2568,
+      "step": 2151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7584570240373023,
+      "learning_rate": 5.726215027736981e-06,
+      "loss": 0.2656,
+      "step": 2152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.803548445089954,
+      "learning_rate": 5.722714531775463e-06,
+      "loss": 0.2764,
+      "step": 2153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7943749250184243,
+      "learning_rate": 5.719213673987277e-06,
+      "loss": 0.2603,
+      "step": 2154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7703043004161156,
+      "learning_rate": 5.715712456125127e-06,
+      "loss": 0.2671,
+      "step": 2155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8174186875753501,
+      "learning_rate": 5.712210879941891e-06,
+      "loss": 0.279,
+      "step": 2156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8271193210504887,
+      "learning_rate": 5.7087089471906345e-06,
+      "loss": 0.2878,
+      "step": 2157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7153828429414296,
+      "learning_rate": 5.705206659624597e-06,
+      "loss": 0.2716,
+      "step": 2158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8041051114644618,
+      "learning_rate": 5.701704018997198e-06,
+      "loss": 0.2742,
+      "step": 2159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7905393395119593,
+      "learning_rate": 5.698201027062034e-06,
+      "loss": 0.2549,
+      "step": 2160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8395795138930428,
+      "learning_rate": 5.694697685572873e-06,
+      "loss": 0.2708,
+      "step": 2161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8186524105397563,
+      "learning_rate": 5.6911939962836625e-06,
+      "loss": 0.293,
+      "step": 2162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.789306833790475,
+      "learning_rate": 5.687689960948526e-06,
+      "loss": 0.2864,
+      "step": 2163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8668952682743266,
+      "learning_rate": 5.684185581321752e-06,
+      "loss": 0.2876,
+      "step": 2164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8092117244111303,
+      "learning_rate": 5.680680859157808e-06,
+      "loss": 0.2721,
+      "step": 2165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7803947122222203,
+      "learning_rate": 5.6771757962113325e-06,
+      "loss": 0.2637,
+      "step": 2166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8148673062197249,
+      "learning_rate": 5.673670394237133e-06,
+      "loss": 0.2532,
+      "step": 2167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8043196496707214,
+      "learning_rate": 5.670164654990189e-06,
+      "loss": 0.2692,
+      "step": 2168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6886474935898494,
+      "learning_rate": 5.666658580225643e-06,
+      "loss": 0.2391,
+      "step": 2169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8509489158221799,
+      "learning_rate": 5.663152171698813e-06,
+      "loss": 0.2947,
+      "step": 2170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0229819955924404,
+      "learning_rate": 5.659645431165179e-06,
+      "loss": 0.2701,
+      "step": 2171
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7932935575528063,
+      "learning_rate": 5.656138360380391e-06,
+      "loss": 0.3052,
+      "step": 2172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9683080254449506,
+      "learning_rate": 5.65263096110026e-06,
+      "loss": 0.2869,
+      "step": 2173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7846394762584661,
+      "learning_rate": 5.649123235080762e-06,
+      "loss": 0.289,
+      "step": 2174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0712010302512034,
+      "learning_rate": 5.645615184078044e-06,
+      "loss": 0.2765,
+      "step": 2175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7891480631172199,
+      "learning_rate": 5.642106809848405e-06,
+      "loss": 0.2701,
+      "step": 2176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7405728962355087,
+      "learning_rate": 5.638598114148314e-06,
+      "loss": 0.2707,
+      "step": 2177
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7875568551971973,
+      "learning_rate": 5.635089098734394e-06,
+      "loss": 0.276,
+      "step": 2178
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7889117500785916,
+      "learning_rate": 5.631579765363436e-06,
+      "loss": 0.2701,
+      "step": 2179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7025786393324915,
+      "learning_rate": 5.628070115792384e-06,
+      "loss": 0.2686,
+      "step": 2180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7929988960884666,
+      "learning_rate": 5.624560151778341e-06,
+      "loss": 0.2868,
+      "step": 2181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8284106137567406,
+      "learning_rate": 5.621049875078571e-06,
+      "loss": 0.293,
+      "step": 2182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8218003802360905,
+      "learning_rate": 5.617539287450492e-06,
+      "loss": 0.2574,
+      "step": 2183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7589179744714613,
+      "learning_rate": 5.614028390651676e-06,
+      "loss": 0.2713,
+      "step": 2184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7619510134972965,
+      "learning_rate": 5.610517186439853e-06,
+      "loss": 0.2571,
+      "step": 2185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0859931546240895,
+      "learning_rate": 5.6070056765729065e-06,
+      "loss": 0.2766,
+      "step": 2186
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7441055224321526,
+      "learning_rate": 5.6034938628088705e-06,
+      "loss": 0.2688,
+      "step": 2187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8439427026315117,
+      "learning_rate": 5.599981746905935e-06,
+      "loss": 0.2657,
+      "step": 2188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7680548286866057,
+      "learning_rate": 5.596469330622437e-06,
+      "loss": 0.2639,
+      "step": 2189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7966631007978717,
+      "learning_rate": 5.592956615716867e-06,
+      "loss": 0.2825,
+      "step": 2190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.730134842674534,
+      "learning_rate": 5.589443603947863e-06,
+      "loss": 0.2571,
+      "step": 2191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7730195899820353,
+      "learning_rate": 5.585930297074214e-06,
+      "loss": 0.2606,
+      "step": 2192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8016697458111541,
+      "learning_rate": 5.582416696854853e-06,
+      "loss": 0.2791,
+      "step": 2193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7272803392192837,
+      "learning_rate": 5.578902805048866e-06,
+      "loss": 0.2608,
+      "step": 2194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7061576801196486,
+      "learning_rate": 5.575388623415476e-06,
+      "loss": 0.2333,
+      "step": 2195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6985599085855471,
+      "learning_rate": 5.571874153714063e-06,
+      "loss": 0.2513,
+      "step": 2196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8515309967159153,
+      "learning_rate": 5.568359397704138e-06,
+      "loss": 0.2702,
+      "step": 2197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8194053838639703,
+      "learning_rate": 5.564844357145365e-06,
+      "loss": 0.3005,
+      "step": 2198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8392957835564029,
+      "learning_rate": 5.561329033797547e-06,
+      "loss": 0.2651,
+      "step": 2199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7447457418348077,
+      "learning_rate": 5.55781342942063e-06,
+      "loss": 0.2584,
+      "step": 2200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7278228843386922,
+      "learning_rate": 5.5542975457746966e-06,
+      "loss": 0.2488,
+      "step": 2201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7805754348677372,
+      "learning_rate": 5.550781384619974e-06,
+      "loss": 0.284,
+      "step": 2202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8111324944897336,
+      "learning_rate": 5.5472649477168264e-06,
+      "loss": 0.2697,
+      "step": 2203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0392242947665973,
+      "learning_rate": 5.543748236825756e-06,
+      "loss": 0.2768,
+      "step": 2204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0778940544234314,
+      "learning_rate": 5.5402312537074035e-06,
+      "loss": 0.2972,
+      "step": 2205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8645769261266549,
+      "learning_rate": 5.536714000122543e-06,
+      "loss": 0.2678,
+      "step": 2206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8118362712855375,
+      "learning_rate": 5.5331964778320875e-06,
+      "loss": 0.2948,
+      "step": 2207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8557535211602187,
+      "learning_rate": 5.529678688597081e-06,
+      "loss": 0.2808,
+      "step": 2208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0536611781658196,
+      "learning_rate": 5.5261606341787035e-06,
+      "loss": 0.2564,
+      "step": 2209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8703405802405775,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.2749,
+      "step": 2210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7183998736888241,
+      "learning_rate": 5.5191237368372175e-06,
+      "loss": 0.2643,
+      "step": 2211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9624400514395779,
+      "learning_rate": 5.515604897437129e-06,
+      "loss": 0.2958,
+      "step": 2212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8195865754957742,
+      "learning_rate": 5.512085799899705e-06,
+      "loss": 0.2783,
+      "step": 2213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6639350009599383,
+      "learning_rate": 5.50856644598678e-06,
+      "loss": 0.2693,
+      "step": 2214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7330471012027615,
+      "learning_rate": 5.505046837460319e-06,
+      "loss": 0.2797,
+      "step": 2215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8349513890671368,
+      "learning_rate": 5.501526976082413e-06,
+      "loss": 0.2711,
+      "step": 2216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7816238690785422,
+      "learning_rate": 5.498006863615275e-06,
+      "loss": 0.252,
+      "step": 2217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7477417832231045,
+      "learning_rate": 5.49448650182125e-06,
+      "loss": 0.2508,
+      "step": 2218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7756800267912629,
+      "learning_rate": 5.490965892462804e-06,
+      "loss": 0.2754,
+      "step": 2219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7715780133363578,
+      "learning_rate": 5.487445037302531e-06,
+      "loss": 0.2781,
+      "step": 2220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7707341558515736,
+      "learning_rate": 5.483923938103143e-06,
+      "loss": 0.2506,
+      "step": 2221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7086282717194262,
+      "learning_rate": 5.480402596627477e-06,
+      "loss": 0.2499,
+      "step": 2222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7475247963397312,
+      "learning_rate": 5.476881014638491e-06,
+      "loss": 0.2293,
+      "step": 2223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.806620472218494,
+      "learning_rate": 5.473359193899265e-06,
+      "loss": 0.2727,
+      "step": 2224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.834887537233682,
+      "learning_rate": 5.469837136172993e-06,
+      "loss": 0.274,
+      "step": 2225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0770649868912343,
+      "learning_rate": 5.466314843222993e-06,
+      "loss": 0.2767,
+      "step": 2226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7329066168947298,
+      "learning_rate": 5.4627923168127e-06,
+      "loss": 0.2571,
+      "step": 2227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.748817548162202,
+      "learning_rate": 5.459269558705667e-06,
+      "loss": 0.2653,
+      "step": 2228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8149074537661395,
+      "learning_rate": 5.4557465706655564e-06,
+      "loss": 0.2867,
+      "step": 2229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8141566673080615,
+      "learning_rate": 5.452223354456154e-06,
+      "loss": 0.2578,
+      "step": 2230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7395428461485227,
+      "learning_rate": 5.4486999118413555e-06,
+      "loss": 0.2558,
+      "step": 2231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.802981835594688,
+      "learning_rate": 5.445176244585171e-06,
+      "loss": 0.2619,
+      "step": 2232
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8133974895540018,
+      "learning_rate": 5.441652354451721e-06,
+      "loss": 0.2794,
+      "step": 2233
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7416258096861319,
+      "learning_rate": 5.438128243205241e-06,
+      "loss": 0.28,
+      "step": 2234
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9061114674208737,
+      "learning_rate": 5.434603912610073e-06,
+      "loss": 0.2778,
+      "step": 2235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7853961398223023,
+      "learning_rate": 5.431079364430677e-06,
+      "loss": 0.2837,
+      "step": 2236
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7951983141285898,
+      "learning_rate": 5.427554600431611e-06,
+      "loss": 0.2769,
+      "step": 2237
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8008485584923264,
+      "learning_rate": 5.4240296223775465e-06,
+      "loss": 0.2905,
+      "step": 2238
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8424534218845423,
+      "learning_rate": 5.420504432033265e-06,
+      "loss": 0.2815,
+      "step": 2239
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9015986899369799,
+      "learning_rate": 5.416979031163649e-06,
+      "loss": 0.2979,
+      "step": 2240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7124980364997018,
+      "learning_rate": 5.4134534215336895e-06,
+      "loss": 0.2481,
+      "step": 2241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8379792746828988,
+      "learning_rate": 5.409927604908481e-06,
+      "loss": 0.2444,
+      "step": 2242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7168444930987256,
+      "learning_rate": 5.406401583053222e-06,
+      "loss": 0.2522,
+      "step": 2243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7546389105388905,
+      "learning_rate": 5.402875357733215e-06,
+      "loss": 0.2574,
+      "step": 2244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7564220017084332,
+      "learning_rate": 5.399348930713859e-06,
+      "loss": 0.2786,
+      "step": 2245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.852289272577411,
+      "learning_rate": 5.395822303760662e-06,
+      "loss": 0.2844,
+      "step": 2246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7964142126234829,
+      "learning_rate": 5.392295478639226e-06,
+      "loss": 0.2751,
+      "step": 2247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7503828913298911,
+      "learning_rate": 5.388768457115254e-06,
+      "loss": 0.2734,
+      "step": 2248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8347113025997926,
+      "learning_rate": 5.38524124095455e-06,
+      "loss": 0.2576,
+      "step": 2249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8399292636701047,
+      "learning_rate": 5.381713831923008e-06,
+      "loss": 0.2739,
+      "step": 2250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8541491009721551,
+      "learning_rate": 5.378186231786629e-06,
+      "loss": 0.2591,
+      "step": 2251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8188842088209559,
+      "learning_rate": 5.3746584423115025e-06,
+      "loss": 0.2561,
+      "step": 2252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7171601281742271,
+      "learning_rate": 5.371130465263813e-06,
+      "loss": 0.2757,
+      "step": 2253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7746366417759074,
+      "learning_rate": 5.367602302409841e-06,
+      "loss": 0.2421,
+      "step": 2254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8112925587393744,
+      "learning_rate": 5.364073955515961e-06,
+      "loss": 0.2607,
+      "step": 2255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7614056995838535,
+      "learning_rate": 5.360545426348638e-06,
+      "loss": 0.2754,
+      "step": 2256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7508635498502665,
+      "learning_rate": 5.357016716674426e-06,
+      "loss": 0.2647,
+      "step": 2257
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7624008572981068,
+      "learning_rate": 5.353487828259973e-06,
+      "loss": 0.2603,
+      "step": 2258
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7451709722536544,
+      "learning_rate": 5.349958762872016e-06,
+      "loss": 0.2551,
+      "step": 2259
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7683392249305656,
+      "learning_rate": 5.346429522277381e-06,
+      "loss": 0.2562,
+      "step": 2260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8119477620494221,
+      "learning_rate": 5.3429001082429766e-06,
+      "loss": 0.2708,
+      "step": 2261
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7854870504379622,
+      "learning_rate": 5.339370522535805e-06,
+      "loss": 0.2617,
+      "step": 2262
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7944377808890871,
+      "learning_rate": 5.33584076692295e-06,
+      "loss": 0.2538,
+      "step": 2263
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.838960010157412,
+      "learning_rate": 5.3323108431715845e-06,
+      "loss": 0.2678,
+      "step": 2264
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8818491466096816,
+      "learning_rate": 5.32878075304896e-06,
+      "loss": 0.2723,
+      "step": 2265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7748607812283757,
+      "learning_rate": 5.325250498322417e-06,
+      "loss": 0.3104,
+      "step": 2266
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7533264819431744,
+      "learning_rate": 5.321720080759373e-06,
+      "loss": 0.273,
+      "step": 2267
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7428814564033245,
+      "learning_rate": 5.318189502127332e-06,
+      "loss": 0.2801,
+      "step": 2268
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7345119262409304,
+      "learning_rate": 5.3146587641938766e-06,
+      "loss": 0.2659,
+      "step": 2269
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7729379210063374,
+      "learning_rate": 5.311127868726667e-06,
+      "loss": 0.2574,
+      "step": 2270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8288944571111502,
+      "learning_rate": 5.307596817493445e-06,
+      "loss": 0.2692,
+      "step": 2271
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.797074172914102,
+      "learning_rate": 5.304065612262033e-06,
+      "loss": 0.2559,
+      "step": 2272
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.750360445725693,
+      "learning_rate": 5.300534254800321e-06,
+      "loss": 0.2428,
+      "step": 2273
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7894284877507864,
+      "learning_rate": 5.297002746876284e-06,
+      "loss": 0.2828,
+      "step": 2274
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8961574424545945,
+      "learning_rate": 5.293471090257971e-06,
+      "loss": 0.2887,
+      "step": 2275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7971145895731171,
+      "learning_rate": 5.289939286713505e-06,
+      "loss": 0.2688,
+      "step": 2276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8530771987872031,
+      "learning_rate": 5.286407338011079e-06,
+      "loss": 0.3164,
+      "step": 2277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.035532625296521,
+      "learning_rate": 5.282875245918963e-06,
+      "loss": 0.2745,
+      "step": 2278
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8751585214042478,
+      "learning_rate": 5.279343012205495e-06,
+      "loss": 0.2714,
+      "step": 2279
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7695663555767011,
+      "learning_rate": 5.275810638639088e-06,
+      "loss": 0.266,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8480952550529547,
+      "learning_rate": 5.272278126988224e-06,
+      "loss": 0.2951,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7627599642181915,
+      "learning_rate": 5.26874547902145e-06,
+      "loss": 0.26,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.779762597077081,
+      "learning_rate": 5.265212696507387e-06,
+      "loss": 0.266,
+      "step": 2283
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.856246758782901,
+      "learning_rate": 5.2616797812147205e-06,
+      "loss": 0.2745,
+      "step": 2284
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8230794260624629,
+      "learning_rate": 5.258146734912202e-06,
+      "loss": 0.2573,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8907594770269753,
+      "learning_rate": 5.2546135593686484e-06,
+      "loss": 0.269,
+      "step": 2286
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8358401546039593,
+      "learning_rate": 5.251080256352944e-06,
+      "loss": 0.2471,
+      "step": 2287
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7980074044404522,
+      "learning_rate": 5.247546827634035e-06,
+      "loss": 0.2512,
+      "step": 2288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8254287190948283,
+      "learning_rate": 5.244013274980932e-06,
+      "loss": 0.247,
+      "step": 2289
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7643885468173156,
+      "learning_rate": 5.240479600162704e-06,
+      "loss": 0.2801,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7511904583001524,
+      "learning_rate": 5.2369458049484824e-06,
+      "loss": 0.2394,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7426459820904299,
+      "learning_rate": 5.233411891107464e-06,
+      "loss": 0.2221,
+      "step": 2292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7981614170596104,
+      "learning_rate": 5.229877860408899e-06,
+      "loss": 0.2352,
+      "step": 2293
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6737735698968934,
+      "learning_rate": 5.226343714622096e-06,
+      "loss": 0.2171,
+      "step": 2294
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7102532658531497,
+      "learning_rate": 5.222809455516427e-06,
+      "loss": 0.2104,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6981764683303089,
+      "learning_rate": 5.219275084861315e-06,
+      "loss": 0.2102,
+      "step": 2296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7442353061412942,
+      "learning_rate": 5.2157406044262424e-06,
+      "loss": 0.2245,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.509113459864884,
+      "learning_rate": 5.212206015980742e-06,
+      "loss": 0.2215,
+      "step": 2298
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8217532351221407,
+      "learning_rate": 5.208671321294406e-06,
+      "loss": 0.2346,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7869624424966375,
+      "learning_rate": 5.2051365221368785e-06,
+      "loss": 0.2248,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7445306537410045,
+      "learning_rate": 5.201601620277854e-06,
+      "loss": 0.2228,
+      "step": 2301
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8622877528780682,
+      "learning_rate": 5.198066617487078e-06,
+      "loss": 0.232,
+      "step": 2302
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.74964724220508,
+      "learning_rate": 5.194531515534349e-06,
+      "loss": 0.211,
+      "step": 2303
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7049232215972648,
+      "learning_rate": 5.190996316189515e-06,
+      "loss": 0.2099,
+      "step": 2304
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8484662824378518,
+      "learning_rate": 5.18746102122247e-06,
+      "loss": 0.2214,
+      "step": 2305
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7670687448711863,
+      "learning_rate": 5.1839256324031575e-06,
+      "loss": 0.2063,
+      "step": 2306
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8039060793456593,
+      "learning_rate": 5.18039015150157e-06,
+      "loss": 0.2325,
+      "step": 2307
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7602983782369528,
+      "learning_rate": 5.176854580287744e-06,
+      "loss": 0.2079,
+      "step": 2308
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7256221394535961,
+      "learning_rate": 5.173318920531761e-06,
+      "loss": 0.1913,
+      "step": 2309
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.730584149474201,
+      "learning_rate": 5.169783174003744e-06,
+      "loss": 0.2049,
+      "step": 2310
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7206703649552568,
+      "learning_rate": 5.166247342473868e-06,
+      "loss": 0.2114,
+      "step": 2311
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7914718650822328,
+      "learning_rate": 5.1627114277123435e-06,
+      "loss": 0.2189,
+      "step": 2312
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8124657238084069,
+      "learning_rate": 5.159175431489424e-06,
+      "loss": 0.225,
+      "step": 2313
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7620308061383572,
+      "learning_rate": 5.155639355575402e-06,
+      "loss": 0.2169,
+      "step": 2314
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8029658471647007,
+      "learning_rate": 5.152103201740617e-06,
+      "loss": 0.2096,
+      "step": 2315
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8453006351263708,
+      "learning_rate": 5.14856697175544e-06,
+      "loss": 0.2341,
+      "step": 2316
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.336065583668775,
+      "learning_rate": 5.145030667390281e-06,
+      "loss": 0.1907,
+      "step": 2317
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7779029368562268,
+      "learning_rate": 5.141494290415592e-06,
+      "loss": 0.2193,
+      "step": 2318
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7932416935829201,
+      "learning_rate": 5.137957842601856e-06,
+      "loss": 0.2222,
+      "step": 2319
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.854463863311718,
+      "learning_rate": 5.1344213257195965e-06,
+      "loss": 0.2099,
+      "step": 2320
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7207795438976249,
+      "learning_rate": 5.130884741539367e-06,
+      "loss": 0.2069,
+      "step": 2321
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7642991340284006,
+      "learning_rate": 5.127348091831755e-06,
+      "loss": 0.2218,
+      "step": 2322
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7904034362685738,
+      "learning_rate": 5.123811378367387e-06,
+      "loss": 0.2159,
+      "step": 2323
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7538155763311662,
+      "learning_rate": 5.120274602916915e-06,
+      "loss": 0.2064,
+      "step": 2324
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.1411973467558454,
+      "learning_rate": 5.116737767251021e-06,
+      "loss": 0.2305,
+      "step": 2325
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.785804923031754,
+      "learning_rate": 5.1132008731404225e-06,
+      "loss": 0.2234,
+      "step": 2326
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8089390847296116,
+      "learning_rate": 5.1096639223558645e-06,
+      "loss": 0.2242,
+      "step": 2327
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7564433989938529,
+      "learning_rate": 5.106126916668118e-06,
+      "loss": 0.1994,
+      "step": 2328
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8590192480671469,
+      "learning_rate": 5.102589857847984e-06,
+      "loss": 0.2371,
+      "step": 2329
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7518750101278008,
+      "learning_rate": 5.099052747666289e-06,
+      "loss": 0.2209,
+      "step": 2330
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8503303462629028,
+      "learning_rate": 5.0955155878938845e-06,
+      "loss": 0.2267,
+      "step": 2331
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7561693162317759,
+      "learning_rate": 5.09197838030165e-06,
+      "loss": 0.2173,
+      "step": 2332
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7749955150658469,
+      "learning_rate": 5.088441126660484e-06,
+      "loss": 0.2338,
+      "step": 2333
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8657890463862576,
+      "learning_rate": 5.084903828741312e-06,
+      "loss": 0.2112,
+      "step": 2334
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8321359200956945,
+      "learning_rate": 5.081366488315082e-06,
+      "loss": 0.213,
+      "step": 2335
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8667716407939225,
+      "learning_rate": 5.0778291071527596e-06,
+      "loss": 0.2107,
+      "step": 2336
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7423276101319868,
+      "learning_rate": 5.074291687025334e-06,
+      "loss": 0.2246,
+      "step": 2337
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8080055918589729,
+      "learning_rate": 5.070754229703811e-06,
+      "loss": 0.2376,
+      "step": 2338
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8015538580453513,
+      "learning_rate": 5.067216736959219e-06,
+      "loss": 0.2143,
+      "step": 2339
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8125721735940292,
+      "learning_rate": 5.063679210562602e-06,
+      "loss": 0.2381,
+      "step": 2340
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7252489540508559,
+      "learning_rate": 5.06014165228502e-06,
+      "loss": 0.2072,
+      "step": 2341
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8247727641340207,
+      "learning_rate": 5.056604063897549e-06,
+      "loss": 0.2372,
+      "step": 2342
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7874279028862208,
+      "learning_rate": 5.053066447171282e-06,
+      "loss": 0.1942,
+      "step": 2343
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7989428742107043,
+      "learning_rate": 5.049528803877324e-06,
+      "loss": 0.1986,
+      "step": 2344
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.414169516941604,
+      "learning_rate": 5.045991135786796e-06,
+      "loss": 0.2194,
+      "step": 2345
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7290255250238488,
+      "learning_rate": 5.042453444670829e-06,
+      "loss": 0.2178,
+      "step": 2346
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7983575612785196,
+      "learning_rate": 5.038915732300565e-06,
+      "loss": 0.2089,
+      "step": 2347
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7393768290462234,
+      "learning_rate": 5.0353780004471605e-06,
+      "loss": 0.2147,
+      "step": 2348
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7302389138508927,
+      "learning_rate": 5.0318402508817765e-06,
+      "loss": 0.2004,
+      "step": 2349
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9098830214261502,
+      "learning_rate": 5.028302485375588e-06,
+      "loss": 0.2196,
+      "step": 2350
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7622854288858806,
+      "learning_rate": 5.024764705699773e-06,
+      "loss": 0.2202,
+      "step": 2351
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7894152310419376,
+      "learning_rate": 5.021226913625522e-06,
+      "loss": 0.2249,
+      "step": 2352
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8210119117666117,
+      "learning_rate": 5.0176891109240265e-06,
+      "loss": 0.2287,
+      "step": 2353
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7780721295182953,
+      "learning_rate": 5.0141512993664864e-06,
+      "loss": 0.2301,
+      "step": 2354
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8137529735212614,
+      "learning_rate": 5.0106134807241045e-06,
+      "loss": 0.2209,
+      "step": 2355
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7799121205168418,
+      "learning_rate": 5.007075656768091e-06,
+      "loss": 0.2165,
+      "step": 2356
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7698357427985956,
+      "learning_rate": 5.0035378292696525e-06,
+      "loss": 0.2012,
+      "step": 2357
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7858214165684966,
+      "learning_rate": 5e-06,
+      "loss": 0.2302,
+      "step": 2358
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7910422501896973,
+      "learning_rate": 4.9964621707303475e-06,
+      "loss": 0.2283,
+      "step": 2359
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.87550307004753,
+      "learning_rate": 4.99292434323191e-06,
+      "loss": 0.2091,
+      "step": 2360
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8221068539403279,
+      "learning_rate": 4.989386519275895e-06,
+      "loss": 0.2163,
+      "step": 2361
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7102727532952308,
+      "learning_rate": 4.985848700633515e-06,
+      "loss": 0.2061,
+      "step": 2362
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9594129678659911,
+      "learning_rate": 4.9823108890759735e-06,
+      "loss": 0.2251,
+      "step": 2363
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7434562937094628,
+      "learning_rate": 4.97877308637448e-06,
+      "loss": 0.2009,
+      "step": 2364
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.802722379853632,
+      "learning_rate": 4.975235294300227e-06,
+      "loss": 0.2009,
+      "step": 2365
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7158551391965297,
+      "learning_rate": 4.971697514624414e-06,
+      "loss": 0.2012,
+      "step": 2366
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7564741724644032,
+      "learning_rate": 4.9681597491182235e-06,
+      "loss": 0.2269,
+      "step": 2367
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7535420395139762,
+      "learning_rate": 4.964621999552841e-06,
+      "loss": 0.2063,
+      "step": 2368
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8557401952778546,
+      "learning_rate": 4.961084267699435e-06,
+      "loss": 0.2125,
+      "step": 2369
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9010621547144847,
+      "learning_rate": 4.957546555329174e-06,
+      "loss": 0.2186,
+      "step": 2370
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7926711456676137,
+      "learning_rate": 4.9540088642132044e-06,
+      "loss": 0.1964,
+      "step": 2371
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9203596789659817,
+      "learning_rate": 4.950471196122677e-06,
+      "loss": 0.2068,
+      "step": 2372
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.594562261458365,
+      "learning_rate": 4.94693355282872e-06,
+      "loss": 0.2167,
+      "step": 2373
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8413506756258259,
+      "learning_rate": 4.943395936102453e-06,
+      "loss": 0.2057,
+      "step": 2374
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8127019540860767,
+      "learning_rate": 4.939858347714981e-06,
+      "loss": 0.1981,
+      "step": 2375
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7853928104403218,
+      "learning_rate": 4.9363207894374e-06,
+      "loss": 0.2213,
+      "step": 2376
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7860298092308415,
+      "learning_rate": 4.9327832630407825e-06,
+      "loss": 0.2067,
+      "step": 2377
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.788364968531785,
+      "learning_rate": 4.929245770296191e-06,
+      "loss": 0.2409,
+      "step": 2378
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8049853712323751,
+      "learning_rate": 4.925708312974667e-06,
+      "loss": 0.2136,
+      "step": 2379
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8765179965198429,
+      "learning_rate": 4.922170892847242e-06,
+      "loss": 0.2347,
+      "step": 2380
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5158464723241556,
+      "learning_rate": 4.91863351168492e-06,
+      "loss": 0.225,
+      "step": 2381
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7534045922643255,
+      "learning_rate": 4.91509617125869e-06,
+      "loss": 0.2293,
+      "step": 2382
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7161967653605342,
+      "learning_rate": 4.911558873339517e-06,
+      "loss": 0.1982,
+      "step": 2383
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7729397790809244,
+      "learning_rate": 4.908021619698352e-06,
+      "loss": 0.1975,
+      "step": 2384
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8275747490531485,
+      "learning_rate": 4.904484412106117e-06,
+      "loss": 0.2262,
+      "step": 2385
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8205438200881707,
+      "learning_rate": 4.900947252333712e-06,
+      "loss": 0.2154,
+      "step": 2386
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7623140111123707,
+      "learning_rate": 4.897410142152017e-06,
+      "loss": 0.2202,
+      "step": 2387
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8065882111571177,
+      "learning_rate": 4.8938730833318825e-06,
+      "loss": 0.2274,
+      "step": 2388
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7921337599564683,
+      "learning_rate": 4.890336077644137e-06,
+      "loss": 0.2092,
+      "step": 2389
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7767470361687048,
+      "learning_rate": 4.8867991268595775e-06,
+      "loss": 0.2008,
+      "step": 2390
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8800911204699133,
+      "learning_rate": 4.88326223274898e-06,
+      "loss": 0.2179,
+      "step": 2391
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8193938853133478,
+      "learning_rate": 4.879725397083087e-06,
+      "loss": 0.2107,
+      "step": 2392
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.5000009817069209,
+      "learning_rate": 4.876188621632614e-06,
+      "loss": 0.2195,
+      "step": 2393
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8302698821762354,
+      "learning_rate": 4.872651908168244e-06,
+      "loss": 0.2179,
+      "step": 2394
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8593334228586935,
+      "learning_rate": 4.869115258460636e-06,
+      "loss": 0.2099,
+      "step": 2395
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8057952377647806,
+      "learning_rate": 4.865578674280406e-06,
+      "loss": 0.2125,
+      "step": 2396
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7087968824001948,
+      "learning_rate": 4.862042157398147e-06,
+      "loss": 0.2035,
+      "step": 2397
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7578360107756573,
+      "learning_rate": 4.85850570958441e-06,
+      "loss": 0.218,
+      "step": 2398
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8122280223453303,
+      "learning_rate": 4.854969332609722e-06,
+      "loss": 0.204,
+      "step": 2399
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.734281785431388,
+      "learning_rate": 4.851433028244562e-06,
+      "loss": 0.205,
+      "step": 2400
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7630926654820535,
+      "learning_rate": 4.847896798259386e-06,
+      "loss": 0.2127,
+      "step": 2401
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.049225156696353,
+      "learning_rate": 4.844360644424599e-06,
+      "loss": 0.2034,
+      "step": 2402
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.720526710189086,
+      "learning_rate": 4.840824568510579e-06,
+      "loss": 0.2151,
+      "step": 2403
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3668326381261617,
+      "learning_rate": 4.837288572287658e-06,
+      "loss": 0.2044,
+      "step": 2404
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8088538581769678,
+      "learning_rate": 4.8337526575261336e-06,
+      "loss": 0.2138,
+      "step": 2405
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8231740753193969,
+      "learning_rate": 4.830216825996257e-06,
+      "loss": 0.2102,
+      "step": 2406
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8073635776455979,
+      "learning_rate": 4.82668107946824e-06,
+      "loss": 0.2191,
+      "step": 2407
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.712687053013133,
+      "learning_rate": 4.8231454197122575e-06,
+      "loss": 0.1871,
+      "step": 2408
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8086558571183987,
+      "learning_rate": 4.81960984849843e-06,
+      "loss": 0.2008,
+      "step": 2409
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.796532009706796,
+      "learning_rate": 4.816074367596843e-06,
+      "loss": 0.2227,
+      "step": 2410
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8136993517844047,
+      "learning_rate": 4.81253897877753e-06,
+      "loss": 0.218,
+      "step": 2411
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8601462000942094,
+      "learning_rate": 4.809003683810487e-06,
+      "loss": 0.2333,
+      "step": 2412
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8409812308918455,
+      "learning_rate": 4.805468484465651e-06,
+      "loss": 0.2096,
+      "step": 2413
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9220857541544131,
+      "learning_rate": 4.801933382512923e-06,
+      "loss": 0.2258,
+      "step": 2414
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.803176628281169,
+      "learning_rate": 4.798398379722147e-06,
+      "loss": 0.2066,
+      "step": 2415
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9524532964887351,
+      "learning_rate": 4.794863477863122e-06,
+      "loss": 0.2044,
+      "step": 2416
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8740075606311924,
+      "learning_rate": 4.791328678705594e-06,
+      "loss": 0.2053,
+      "step": 2417
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8173723921734092,
+      "learning_rate": 4.78779398401926e-06,
+      "loss": 0.2292,
+      "step": 2418
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.854319263798769,
+      "learning_rate": 4.784259395573759e-06,
+      "loss": 0.2091,
+      "step": 2419
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8562772817307991,
+      "learning_rate": 4.780724915138686e-06,
+      "loss": 0.2154,
+      "step": 2420
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7564093446123603,
+      "learning_rate": 4.777190544483574e-06,
+      "loss": 0.199,
+      "step": 2421
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8123992129756347,
+      "learning_rate": 4.7736562853779066e-06,
+      "loss": 0.2171,
+      "step": 2422
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8174177451037266,
+      "learning_rate": 4.770122139591103e-06,
+      "loss": 0.2282,
+      "step": 2423
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7237517867039658,
+      "learning_rate": 4.766588108892538e-06,
+      "loss": 0.1972,
+      "step": 2424
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0186673786035758,
+      "learning_rate": 4.763054195051519e-06,
+      "loss": 0.2176,
+      "step": 2425
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7787924819487061,
+      "learning_rate": 4.7595203998373e-06,
+      "loss": 0.2194,
+      "step": 2426
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8745842702334482,
+      "learning_rate": 4.75598672501907e-06,
+      "loss": 0.2336,
+      "step": 2427
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8558141850509695,
+      "learning_rate": 4.752453172365966e-06,
+      "loss": 0.2377,
+      "step": 2428
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0721086587945297,
+      "learning_rate": 4.7489197436470575e-06,
+      "loss": 0.2226,
+      "step": 2429
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7963495714984682,
+      "learning_rate": 4.745386440631354e-06,
+      "loss": 0.2125,
+      "step": 2430
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8301947289397346,
+      "learning_rate": 4.7418532650878005e-06,
+      "loss": 0.2058,
+      "step": 2431
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7766174689633839,
+      "learning_rate": 4.738320218785281e-06,
+      "loss": 0.2018,
+      "step": 2432
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8206855422729938,
+      "learning_rate": 4.734787303492615e-06,
+      "loss": 0.2283,
+      "step": 2433
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.824890015874798,
+      "learning_rate": 4.73125452097855e-06,
+      "loss": 0.221,
+      "step": 2434
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.1851409435978577,
+      "learning_rate": 4.727721873011778e-06,
+      "loss": 0.1967,
+      "step": 2435
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7945109662543894,
+      "learning_rate": 4.7241893613609126e-06,
+      "loss": 0.2175,
+      "step": 2436
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8756014268947288,
+      "learning_rate": 4.720656987794507e-06,
+      "loss": 0.2272,
+      "step": 2437
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7622096788229816,
+      "learning_rate": 4.717124754081038e-06,
+      "loss": 0.228,
+      "step": 2438
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8668603313058094,
+      "learning_rate": 4.713592661988922e-06,
+      "loss": 0.2125,
+      "step": 2439
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8688221999691017,
+      "learning_rate": 4.710060713286496e-06,
+      "loss": 0.2197,
+      "step": 2440
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7144475663224694,
+      "learning_rate": 4.70652890974203e-06,
+      "loss": 0.2047,
+      "step": 2441
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7097760909152804,
+      "learning_rate": 4.702997253123716e-06,
+      "loss": 0.1942,
+      "step": 2442
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8049577534365748,
+      "learning_rate": 4.6994657451996815e-06,
+      "loss": 0.2153,
+      "step": 2443
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7901677277063015,
+      "learning_rate": 4.6959343877379695e-06,
+      "loss": 0.2197,
+      "step": 2444
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8916335353291767,
+      "learning_rate": 4.692403182506557e-06,
+      "loss": 0.2246,
+      "step": 2445
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7762709950889023,
+      "learning_rate": 4.688872131273334e-06,
+      "loss": 0.1977,
+      "step": 2446
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8357740988597325,
+      "learning_rate": 4.685341235806126e-06,
+      "loss": 0.1916,
+      "step": 2447
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.852387638547402,
+      "learning_rate": 4.6818104978726685e-06,
+      "loss": 0.2129,
+      "step": 2448
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8299344758079562,
+      "learning_rate": 4.678279919240629e-06,
+      "loss": 0.2348,
+      "step": 2449
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.739991194052082,
+      "learning_rate": 4.674749501677585e-06,
+      "loss": 0.2046,
+      "step": 2450
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8063429998260216,
+      "learning_rate": 4.671219246951043e-06,
+      "loss": 0.2099,
+      "step": 2451
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8320244900747046,
+      "learning_rate": 4.667689156828417e-06,
+      "loss": 0.2038,
+      "step": 2452
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8812166432031461,
+      "learning_rate": 4.664159233077051e-06,
+      "loss": 0.2283,
+      "step": 2453
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8501520464911608,
+      "learning_rate": 4.660629477464197e-06,
+      "loss": 0.2103,
+      "step": 2454
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7671721262878564,
+      "learning_rate": 4.657099891757023e-06,
+      "loss": 0.2147,
+      "step": 2455
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9178583796556953,
+      "learning_rate": 4.653570477722621e-06,
+      "loss": 0.2153,
+      "step": 2456
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8547656689977728,
+      "learning_rate": 4.650041237127983e-06,
+      "loss": 0.2358,
+      "step": 2457
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.070807582398223,
+      "learning_rate": 4.646512171740028e-06,
+      "loss": 0.2433,
+      "step": 2458
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.1292882870735088,
+      "learning_rate": 4.642983283325574e-06,
+      "loss": 0.2029,
+      "step": 2459
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7634678679368851,
+      "learning_rate": 4.639454573651364e-06,
+      "loss": 0.2198,
+      "step": 2460
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8011404022612336,
+      "learning_rate": 4.6359260444840395e-06,
+      "loss": 0.2129,
+      "step": 2461
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8593120045994026,
+      "learning_rate": 4.6323976975901595e-06,
+      "loss": 0.2129,
+      "step": 2462
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8112514203210169,
+      "learning_rate": 4.628869534736187e-06,
+      "loss": 0.2292,
+      "step": 2463
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7041474029127963,
+      "learning_rate": 4.625341557688499e-06,
+      "loss": 0.1988,
+      "step": 2464
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8829776827435314,
+      "learning_rate": 4.621813768213372e-06,
+      "loss": 0.2344,
+      "step": 2465
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8378296458112924,
+      "learning_rate": 4.618286168076993e-06,
+      "loss": 0.2138,
+      "step": 2466
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8327135741574812,
+      "learning_rate": 4.614758759045452e-06,
+      "loss": 0.2042,
+      "step": 2467
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8582140941589388,
+      "learning_rate": 4.611231542884747e-06,
+      "loss": 0.2184,
+      "step": 2468
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9988831015022456,
+      "learning_rate": 4.6077045213607765e-06,
+      "loss": 0.2,
+      "step": 2469
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8356654630830821,
+      "learning_rate": 4.6041776962393405e-06,
+      "loss": 0.224,
+      "step": 2470
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8221277369816072,
+      "learning_rate": 4.600651069286141e-06,
+      "loss": 0.2199,
+      "step": 2471
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9248367724210906,
+      "learning_rate": 4.597124642266788e-06,
+      "loss": 0.2211,
+      "step": 2472
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7818367732529174,
+      "learning_rate": 4.593598416946779e-06,
+      "loss": 0.2236,
+      "step": 2473
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.6236855109569035,
+      "learning_rate": 4.590072395091521e-06,
+      "loss": 0.2098,
+      "step": 2474
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9134814142494224,
+      "learning_rate": 4.586546578466311e-06,
+      "loss": 0.2325,
+      "step": 2475
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9980219530651077,
+      "learning_rate": 4.583020968836353e-06,
+      "loss": 0.2232,
+      "step": 2476
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7832126599405965,
+      "learning_rate": 4.579495567966738e-06,
+      "loss": 0.2011,
+      "step": 2477
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8606031418651612,
+      "learning_rate": 4.575970377622456e-06,
+      "loss": 0.2007,
+      "step": 2478
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7922045706009951,
+      "learning_rate": 4.572445399568391e-06,
+      "loss": 0.2093,
+      "step": 2479
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7904711446611421,
+      "learning_rate": 4.568920635569325e-06,
+      "loss": 0.2166,
+      "step": 2480
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8708589731453467,
+      "learning_rate": 4.565396087389928e-06,
+      "loss": 0.2343,
+      "step": 2481
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8906454097240946,
+      "learning_rate": 4.56187175679476e-06,
+      "loss": 0.2083,
+      "step": 2482
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8386904061942737,
+      "learning_rate": 4.55834764554828e-06,
+      "loss": 0.2481,
+      "step": 2483
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7661649518909113,
+      "learning_rate": 4.554823755414829e-06,
+      "loss": 0.2114,
+      "step": 2484
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.836550268068223,
+      "learning_rate": 4.551300088158645e-06,
+      "loss": 0.2048,
+      "step": 2485
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.880513650188697,
+      "learning_rate": 4.547776645543846e-06,
+      "loss": 0.2237,
+      "step": 2486
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7289147453228295,
+      "learning_rate": 4.544253429334444e-06,
+      "loss": 0.2102,
+      "step": 2487
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.871295225843014,
+      "learning_rate": 4.540730441294334e-06,
+      "loss": 0.1965,
+      "step": 2488
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.9254085015489751,
+      "learning_rate": 4.5372076831873006e-06,
+      "loss": 0.2109,
+      "step": 2489
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8609964641046355,
+      "learning_rate": 4.533685156777008e-06,
+      "loss": 0.2218,
+      "step": 2490
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7402764396526651,
+      "learning_rate": 4.53016286382701e-06,
+      "loss": 0.1937,
+      "step": 2491
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7729779971622546,
+      "learning_rate": 4.526640806100737e-06,
+      "loss": 0.2106,
+      "step": 2492
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7972882214026777,
+      "learning_rate": 4.523118985361511e-06,
+      "loss": 0.2066,
+      "step": 2493
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7283196268374225,
+      "learning_rate": 4.519597403372524e-06,
+      "loss": 0.2108,
+      "step": 2494
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7671886438836237,
+      "learning_rate": 4.516076061896859e-06,
+      "loss": 0.2267,
+      "step": 2495
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7978630943156916,
+      "learning_rate": 4.5125549626974695e-06,
+      "loss": 0.2055,
+      "step": 2496
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8462359278202543,
+      "learning_rate": 4.5090341075371964e-06,
+      "loss": 0.2314,
+      "step": 2497
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8215667242812845,
+      "learning_rate": 4.505513498178752e-06,
+      "loss": 0.2378,
+      "step": 2498
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7502918207413535,
+      "learning_rate": 4.501993136384727e-06,
+      "loss": 0.2044,
+      "step": 2499
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7894863937279685,
+      "learning_rate": 4.4984730239175885e-06,
+      "loss": 0.2088,
+      "step": 2500
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0229840581933742,
+      "learning_rate": 4.494953162539682e-06,
+      "loss": 0.2159,
+      "step": 2501
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.790992259796568,
+      "learning_rate": 4.491433554013221e-06,
+      "loss": 0.2279,
+      "step": 2502
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8009487040194007,
+      "learning_rate": 4.487914200100296e-06,
+      "loss": 0.2008,
+      "step": 2503
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7370284496245337,
+      "learning_rate": 4.484395102562873e-06,
+      "loss": 0.2086,
+      "step": 2504
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.780302865328674,
+      "learning_rate": 4.4808762631627825e-06,
+      "loss": 0.2258,
+      "step": 2505
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.2132552572721438,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.2105,
+      "step": 2506
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7316723605826713,
+      "learning_rate": 4.4738393658212965e-06,
+      "loss": 0.2086,
+      "step": 2507
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.794285096611385,
+      "learning_rate": 4.47032131140292e-06,
+      "loss": 0.2081,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8104603549367724,
+      "learning_rate": 4.466803522167914e-06,
+      "loss": 0.2233,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7545102178419477,
+      "learning_rate": 4.463285999877458e-06,
+      "loss": 0.2126,
+      "step": 2510
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8301521398593597,
+      "learning_rate": 4.459768746292597e-06,
+      "loss": 0.2101,
+      "step": 2511
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7821821792456858,
+      "learning_rate": 4.456251763174245e-06,
+      "loss": 0.2297,
+      "step": 2512
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.781835242772354,
+      "learning_rate": 4.452735052283175e-06,
+      "loss": 0.2246,
+      "step": 2513
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8034040260340953,
+      "learning_rate": 4.449218615380029e-06,
+      "loss": 0.2235,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7586018423338683,
+      "learning_rate": 4.445702454225305e-06,
+      "loss": 0.195,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8868587055102558,
+      "learning_rate": 4.442186570579373e-06,
+      "loss": 0.2389,
+      "step": 2516
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7672005180419742,
+      "learning_rate": 4.438670966202455e-06,
+      "loss": 0.2,
+      "step": 2517
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8772337154220415,
+      "learning_rate": 4.435155642854637e-06,
+      "loss": 0.2355,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7553565181915971,
+      "learning_rate": 4.431640602295863e-06,
+      "loss": 0.2022,
+      "step": 2519
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.19703490590536,
+      "learning_rate": 4.42812584628594e-06,
+      "loss": 0.2309,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8739272672987183,
+      "learning_rate": 4.4246113765845244e-06,
+      "loss": 0.2321,
+      "step": 2521
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7574698365761353,
+      "learning_rate": 4.421097194951137e-06,
+      "loss": 0.2085,
+      "step": 2522
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7731497100374779,
+      "learning_rate": 4.4175833031451475e-06,
+      "loss": 0.202,
+      "step": 2523
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7886305245113171,
+      "learning_rate": 4.414069702925789e-06,
+      "loss": 0.2199,
+      "step": 2524
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8811570213028409,
+      "learning_rate": 4.4105563960521395e-06,
+      "loss": 0.2256,
+      "step": 2525
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.870375162600956,
+      "learning_rate": 4.407043384283136e-06,
+      "loss": 0.2062,
+      "step": 2526
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7848889908469066,
+      "learning_rate": 4.4035306693775645e-06,
+      "loss": 0.2303,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7954886939931597,
+      "learning_rate": 4.400018253094065e-06,
+      "loss": 0.2207,
+      "step": 2528
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0021897681959684,
+      "learning_rate": 4.396506137191131e-06,
+      "loss": 0.2166,
+      "step": 2529
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7526426112598633,
+      "learning_rate": 4.3929943234270935e-06,
+      "loss": 0.2143,
+      "step": 2530
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7329310323540988,
+      "learning_rate": 4.389482813560148e-06,
+      "loss": 0.214,
+      "step": 2531
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8307057582585351,
+      "learning_rate": 4.385971609348324e-06,
+      "loss": 0.2035,
+      "step": 2532
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7786708023635445,
+      "learning_rate": 4.38246071254951e-06,
+      "loss": 0.1976,
+      "step": 2533
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7936944825981446,
+      "learning_rate": 4.37895012492143e-06,
+      "loss": 0.2248,
+      "step": 2534
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7238156511576821,
+      "learning_rate": 4.375439848221661e-06,
+      "loss": 0.198,
+      "step": 2535
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7788903159046112,
+      "learning_rate": 4.371929884207617e-06,
+      "loss": 0.2159,
+      "step": 2536
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8003797819260319,
+      "learning_rate": 4.3684202346365655e-06,
+      "loss": 0.2169,
+      "step": 2537
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8597189607287592,
+      "learning_rate": 4.364910901265607e-06,
+      "loss": 0.205,
+      "step": 2538
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7782539954362055,
+      "learning_rate": 4.3614018858516885e-06,
+      "loss": 0.2186,
+      "step": 2539
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7680398998061609,
+      "learning_rate": 4.3578931901515955e-06,
+      "loss": 0.1915,
+      "step": 2540
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8361591948753632,
+      "learning_rate": 4.354384815921958e-06,
+      "loss": 0.2222,
+      "step": 2541
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8573688969411207,
+      "learning_rate": 4.350876764919238e-06,
+      "loss": 0.2232,
+      "step": 2542
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8327990901861727,
+      "learning_rate": 4.347369038899744e-06,
+      "loss": 0.2086,
+      "step": 2543
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.2270838229047847,
+      "learning_rate": 4.343861639619611e-06,
+      "loss": 0.2262,
+      "step": 2544
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.0124523114246633,
+      "learning_rate": 4.3403545688348225e-06,
+      "loss": 0.2259,
+      "step": 2545
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7466129292067859,
+      "learning_rate": 4.336847828301189e-06,
+      "loss": 0.2149,
+      "step": 2546
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7415151619574827,
+      "learning_rate": 4.333341419774359e-06,
+      "loss": 0.2036,
+      "step": 2547
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8403315023171162,
+      "learning_rate": 4.329835345009813e-06,
+      "loss": 0.2326,
+      "step": 2548
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8758664265220447,
+      "learning_rate": 4.3263296057628685e-06,
+      "loss": 0.2013,
+      "step": 2549
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.9391956390817267,
+      "learning_rate": 4.322824203788669e-06,
+      "loss": 0.2088,
+      "step": 2550
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7771360172674586,
+      "learning_rate": 4.3193191408421945e-06,
+      "loss": 0.2025,
+      "step": 2551
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7963018476043396,
+      "learning_rate": 4.3158144186782504e-06,
+      "loss": 0.2007,
+      "step": 2552
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8270061857869859,
+      "learning_rate": 4.312310039051476e-06,
+      "loss": 0.2167,
+      "step": 2553
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7944515133014589,
+      "learning_rate": 4.308806003716338e-06,
+      "loss": 0.2173,
+      "step": 2554
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8699419827936057,
+      "learning_rate": 4.305302314427128e-06,
+      "loss": 0.2207,
+      "step": 2555
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7773773160724962,
+      "learning_rate": 4.301798972937968e-06,
+      "loss": 0.1973,
+      "step": 2556
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7618130250589411,
+      "learning_rate": 4.2982959810028025e-06,
+      "loss": 0.2145,
+      "step": 2557
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8090355266716042,
+      "learning_rate": 4.294793340375405e-06,
+      "loss": 0.2066,
+      "step": 2558
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7739907870112505,
+      "learning_rate": 4.291291052809366e-06,
+      "loss": 0.2211,
+      "step": 2559
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7584789428536816,
+      "learning_rate": 4.287789120058111e-06,
+      "loss": 0.2239,
+      "step": 2560
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7706719162241845,
+      "learning_rate": 4.284287543874875e-06,
+      "loss": 0.2144,
+      "step": 2561
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.822506536180516,
+      "learning_rate": 4.280786326012724e-06,
+      "loss": 0.2106,
+      "step": 2562
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7561010461334408,
+      "learning_rate": 4.2772854682245365e-06,
+      "loss": 0.2207,
+      "step": 2563
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9017526126118699,
+      "learning_rate": 4.273784972263021e-06,
+      "loss": 0.2324,
+      "step": 2564
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7888468813412907,
+      "learning_rate": 4.270284839880696e-06,
+      "loss": 0.209,
+      "step": 2565
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0136923994022726,
+      "learning_rate": 4.266785072829901e-06,
+      "loss": 0.214,
+      "step": 2566
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7563404658989648,
+      "learning_rate": 4.263285672862792e-06,
+      "loss": 0.1987,
+      "step": 2567
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7445394674232683,
+      "learning_rate": 4.259786641731344e-06,
+      "loss": 0.1919,
+      "step": 2568
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0680589961218432,
+      "learning_rate": 4.256287981187343e-06,
+      "loss": 0.1964,
+      "step": 2569
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8772641781764245,
+      "learning_rate": 4.252789692982394e-06,
+      "loss": 0.2131,
+      "step": 2570
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7840184448244282,
+      "learning_rate": 4.2492917788679095e-06,
+      "loss": 0.2029,
+      "step": 2571
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7460587791426857,
+      "learning_rate": 4.245794240595123e-06,
+      "loss": 0.2184,
+      "step": 2572
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0106146470347488,
+      "learning_rate": 4.242297079915071e-06,
+      "loss": 0.2072,
+      "step": 2573
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8015871873316319,
+      "learning_rate": 4.23880029857861e-06,
+      "loss": 0.204,
+      "step": 2574
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.389903629723513,
+      "learning_rate": 4.235303898336396e-06,
+      "loss": 0.2339,
+      "step": 2575
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.829002017097399,
+      "learning_rate": 4.231807880938904e-06,
+      "loss": 0.2083,
+      "step": 2576
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.4595933065246967,
+      "learning_rate": 4.228312248136414e-06,
+      "loss": 0.213,
+      "step": 2577
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8663428477544075,
+      "learning_rate": 4.224817001679011e-06,
+      "loss": 0.2179,
+      "step": 2578
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7685139133644665,
+      "learning_rate": 4.221322143316589e-06,
+      "loss": 0.2109,
+      "step": 2579
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0148802259160965,
+      "learning_rate": 4.217827674798845e-06,
+      "loss": 0.2237,
+      "step": 2580
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7355237350935212,
+      "learning_rate": 4.214333597875288e-06,
+      "loss": 0.1928,
+      "step": 2581
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8174339469196688,
+      "learning_rate": 4.210839914295219e-06,
+      "loss": 0.2031,
+      "step": 2582
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.605079532483912,
+      "learning_rate": 4.2073466258077564e-06,
+      "loss": 0.2072,
+      "step": 2583
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8193206842341186,
+      "learning_rate": 4.203853734161806e-06,
+      "loss": 0.2089,
+      "step": 2584
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8838420523815028,
+      "learning_rate": 4.200361241106088e-06,
+      "loss": 0.224,
+      "step": 2585
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8763461868590937,
+      "learning_rate": 4.196869148389114e-06,
+      "loss": 0.2069,
+      "step": 2586
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.779271417953496,
+      "learning_rate": 4.1933774577592e-06,
+      "loss": 0.2219,
+      "step": 2587
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.887178861593821,
+      "learning_rate": 4.189886170964458e-06,
+      "loss": 0.2084,
+      "step": 2588
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.823621401604201,
+      "learning_rate": 4.1863952897528e-06,
+      "loss": 0.2264,
+      "step": 2589
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7748703504453128,
+      "learning_rate": 4.1829048158719344e-06,
+      "loss": 0.2107,
+      "step": 2590
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7587224610332843,
+      "learning_rate": 4.179414751069365e-06,
+      "loss": 0.1962,
+      "step": 2591
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7971348612739539,
+      "learning_rate": 4.1759250970923876e-06,
+      "loss": 0.2319,
+      "step": 2592
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.825450159801212,
+      "learning_rate": 4.172435855688101e-06,
+      "loss": 0.207,
+      "step": 2593
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8763172748185678,
+      "learning_rate": 4.168947028603389e-06,
+      "loss": 0.2251,
+      "step": 2594
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8893618963110642,
+      "learning_rate": 4.165458617584933e-06,
+      "loss": 0.2316,
+      "step": 2595
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9026704127956183,
+      "learning_rate": 4.161970624379201e-06,
+      "loss": 0.2003,
+      "step": 2596
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.849967022815981,
+      "learning_rate": 4.158483050732458e-06,
+      "loss": 0.237,
+      "step": 2597
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7401115749197553,
+      "learning_rate": 4.154995898390756e-06,
+      "loss": 0.1925,
+      "step": 2598
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7450751239051122,
+      "learning_rate": 4.151509169099935e-06,
+      "loss": 0.2206,
+      "step": 2599
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9451239227849199,
+      "learning_rate": 4.1480228646056225e-06,
+      "loss": 0.223,
+      "step": 2600
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8441873613501288,
+      "learning_rate": 4.144536986653239e-06,
+      "loss": 0.2296,
+      "step": 2601
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9735456972972173,
+      "learning_rate": 4.141051536987986e-06,
+      "loss": 0.2147,
+      "step": 2602
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8170365717654773,
+      "learning_rate": 4.13756651735485e-06,
+      "loss": 0.1969,
+      "step": 2603
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8210256548470238,
+      "learning_rate": 4.1340819294986076e-06,
+      "loss": 0.22,
+      "step": 2604
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7890074219255183,
+      "learning_rate": 4.130597775163814e-06,
+      "loss": 0.2234,
+      "step": 2605
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7518548017768676,
+      "learning_rate": 4.127114056094811e-06,
+      "loss": 0.2115,
+      "step": 2606
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.833106039511241,
+      "learning_rate": 4.1236307740357175e-06,
+      "loss": 0.2193,
+      "step": 2607
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.898505879824806,
+      "learning_rate": 4.12014793073044e-06,
+      "loss": 0.2215,
+      "step": 2608
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8407393877302821,
+      "learning_rate": 4.116665527922662e-06,
+      "loss": 0.207,
+      "step": 2609
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7906997366739332,
+      "learning_rate": 4.113183567355846e-06,
+      "loss": 0.2142,
+      "step": 2610
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9530969802880402,
+      "learning_rate": 4.109702050773233e-06,
+      "loss": 0.2192,
+      "step": 2611
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9119217157324918,
+      "learning_rate": 4.106220979917845e-06,
+      "loss": 0.2043,
+      "step": 2612
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7481473929727954,
+      "learning_rate": 4.102740356532474e-06,
+      "loss": 0.1969,
+      "step": 2613
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8175005940485587,
+      "learning_rate": 4.099260182359698e-06,
+      "loss": 0.1985,
+      "step": 2614
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.881868179333485,
+      "learning_rate": 4.0957804591418596e-06,
+      "loss": 0.198,
+      "step": 2615
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7957474487326218,
+      "learning_rate": 4.092301188621084e-06,
+      "loss": 0.1995,
+      "step": 2616
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.998252491214412,
+      "learning_rate": 4.088822372539263e-06,
+      "loss": 0.2144,
+      "step": 2617
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7757537905130248,
+      "learning_rate": 4.085344012638067e-06,
+      "loss": 0.2023,
+      "step": 2618
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7307798858092079,
+      "learning_rate": 4.081866110658934e-06,
+      "loss": 0.214,
+      "step": 2619
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7669248893335564,
+      "learning_rate": 4.078388668343075e-06,
+      "loss": 0.2043,
+      "step": 2620
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8375601985020418,
+      "learning_rate": 4.074911687431466e-06,
+      "loss": 0.2119,
+      "step": 2621
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.258357715591786,
+      "learning_rate": 4.071435169664861e-06,
+      "loss": 0.2136,
+      "step": 2622
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7588154065408534,
+      "learning_rate": 4.0679591167837725e-06,
+      "loss": 0.1929,
+      "step": 2623
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7481297979912713,
+      "learning_rate": 4.064483530528487e-06,
+      "loss": 0.2092,
+      "step": 2624
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7520598696444645,
+      "learning_rate": 4.061008412639055e-06,
+      "loss": 0.2021,
+      "step": 2625
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7897222530702417,
+      "learning_rate": 4.057533764855291e-06,
+      "loss": 0.235,
+      "step": 2626
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9492751554712184,
+      "learning_rate": 4.054059588916779e-06,
+      "loss": 0.2151,
+      "step": 2627
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7688177408280645,
+      "learning_rate": 4.050585886562858e-06,
+      "loss": 0.2044,
+      "step": 2628
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8399712193650434,
+      "learning_rate": 4.047112659532641e-06,
+      "loss": 0.2083,
+      "step": 2629
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7610798149400709,
+      "learning_rate": 4.043639909564993e-06,
+      "loss": 0.2187,
+      "step": 2630
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8234593937783196,
+      "learning_rate": 4.040167638398549e-06,
+      "loss": 0.2204,
+      "step": 2631
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8652832938938475,
+      "learning_rate": 4.036695847771694e-06,
+      "loss": 0.2007,
+      "step": 2632
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7458584667636345,
+      "learning_rate": 4.033224539422584e-06,
+      "loss": 0.2134,
+      "step": 2633
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.755086185209679,
+      "learning_rate": 4.029753715089124e-06,
+      "loss": 0.2037,
+      "step": 2634
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.922871028483147,
+      "learning_rate": 4.026283376508983e-06,
+      "loss": 0.2205,
+      "step": 2635
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.747160418944521,
+      "learning_rate": 4.02281352541958e-06,
+      "loss": 0.2005,
+      "step": 2636
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8998173494443632,
+      "learning_rate": 4.0193441635581e-06,
+      "loss": 0.2234,
+      "step": 2637
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8382448780851479,
+      "learning_rate": 4.015875292661474e-06,
+      "loss": 0.2263,
+      "step": 2638
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7634438814948097,
+      "learning_rate": 4.012406914466391e-06,
+      "loss": 0.2103,
+      "step": 2639
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8324756260480739,
+      "learning_rate": 4.008939030709291e-06,
+      "loss": 0.2107,
+      "step": 2640
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7370840542539753,
+      "learning_rate": 4.005471643126372e-06,
+      "loss": 0.1919,
+      "step": 2641
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7781663176095093,
+      "learning_rate": 4.002004753453577e-06,
+      "loss": 0.2178,
+      "step": 2642
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.811826478410373,
+      "learning_rate": 3.998538363426605e-06,
+      "loss": 0.2074,
+      "step": 2643
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8295274249805757,
+      "learning_rate": 3.995072474780899e-06,
+      "loss": 0.2238,
+      "step": 2644
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9344848502417133,
+      "learning_rate": 3.9916070892516575e-06,
+      "loss": 0.2279,
+      "step": 2645
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7888518829408209,
+      "learning_rate": 3.988142208573822e-06,
+      "loss": 0.2035,
+      "step": 2646
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8081453632062138,
+      "learning_rate": 3.984677834482085e-06,
+      "loss": 0.2253,
+      "step": 2647
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8016944172658942,
+      "learning_rate": 3.981213968710882e-06,
+      "loss": 0.2034,
+      "step": 2648
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.128821975314949,
+      "learning_rate": 3.977750612994397e-06,
+      "loss": 0.2006,
+      "step": 2649
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7588499923058202,
+      "learning_rate": 3.9742877690665575e-06,
+      "loss": 0.1956,
+      "step": 2650
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7669432714579391,
+      "learning_rate": 3.970825438661032e-06,
+      "loss": 0.2091,
+      "step": 2651
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8051302780682694,
+      "learning_rate": 3.967363623511238e-06,
+      "loss": 0.2165,
+      "step": 2652
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8225766790454904,
+      "learning_rate": 3.96390232535033e-06,
+      "loss": 0.2243,
+      "step": 2653
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7850855144806796,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 0.2083,
+      "step": 2654
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8355763987646152,
+      "learning_rate": 3.956981286926499e-06,
+      "loss": 0.2249,
+      "step": 2655
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7487095180803912,
+      "learning_rate": 3.953521550128591e-06,
+      "loss": 0.1986,
+      "step": 2656
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7761684268773672,
+      "learning_rate": 3.950062337249596e-06,
+      "loss": 0.2054,
+      "step": 2657
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7901846845990047,
+      "learning_rate": 3.94660365002137e-06,
+      "loss": 0.2091,
+      "step": 2658
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8401474547019882,
+      "learning_rate": 3.943145490175497e-06,
+      "loss": 0.2293,
+      "step": 2659
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8927808934022511,
+      "learning_rate": 3.939687859443308e-06,
+      "loss": 0.2153,
+      "step": 2660
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8312230598729524,
+      "learning_rate": 3.93623075955586e-06,
+      "loss": 0.2354,
+      "step": 2661
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8734796104432941,
+      "learning_rate": 3.932774192243953e-06,
+      "loss": 0.2231,
+      "step": 2662
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9999896995710064,
+      "learning_rate": 3.929318159238113e-06,
+      "loss": 0.198,
+      "step": 2663
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7809118405031471,
+      "learning_rate": 3.925862662268602e-06,
+      "loss": 0.2038,
+      "step": 2664
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7622881275373817,
+      "learning_rate": 3.922407703065411e-06,
+      "loss": 0.2185,
+      "step": 2665
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7758620229097286,
+      "learning_rate": 3.918953283358266e-06,
+      "loss": 0.2232,
+      "step": 2666
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9183220481292341,
+      "learning_rate": 3.915499404876619e-06,
+      "loss": 0.208,
+      "step": 2667
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.817292929559574,
+      "learning_rate": 3.912046069349654e-06,
+      "loss": 0.2132,
+      "step": 2668
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8696779547981769,
+      "learning_rate": 3.908593278506279e-06,
+      "loss": 0.1863,
+      "step": 2669
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8499583948578705,
+      "learning_rate": 3.905141034075135e-06,
+      "loss": 0.2206,
+      "step": 2670
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8587429659584066,
+      "learning_rate": 3.901689337784585e-06,
+      "loss": 0.2115,
+      "step": 2671
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7743943454255556,
+      "learning_rate": 3.898238191362718e-06,
+      "loss": 0.2094,
+      "step": 2672
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7604576967749466,
+      "learning_rate": 3.894787596537352e-06,
+      "loss": 0.1997,
+      "step": 2673
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9398422336087029,
+      "learning_rate": 3.891337555036022e-06,
+      "loss": 0.2228,
+      "step": 2674
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8472698951248894,
+      "learning_rate": 3.887888068585992e-06,
+      "loss": 0.2033,
+      "step": 2675
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8666084641390832,
+      "learning_rate": 3.884439138914243e-06,
+      "loss": 0.218,
+      "step": 2676
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8219024175354562,
+      "learning_rate": 3.880990767747484e-06,
+      "loss": 0.2184,
+      "step": 2677
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9496205000949945,
+      "learning_rate": 3.877542956812137e-06,
+      "loss": 0.2059,
+      "step": 2678
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8210765261496925,
+      "learning_rate": 3.874095707834349e-06,
+      "loss": 0.2008,
+      "step": 2679
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7697768547946381,
+      "learning_rate": 3.870649022539979e-06,
+      "loss": 0.2184,
+      "step": 2680
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9205811402317242,
+      "learning_rate": 3.867202902654616e-06,
+      "loss": 0.2292,
+      "step": 2681
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7495326064505534,
+      "learning_rate": 3.863757349903551e-06,
+      "loss": 0.2335,
+      "step": 2682
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7702974408619584,
+      "learning_rate": 3.860312366011802e-06,
+      "loss": 0.2066,
+      "step": 2683
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8232460203640426,
+      "learning_rate": 3.856867952704095e-06,
+      "loss": 0.2041,
+      "step": 2684
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8250170596350289,
+      "learning_rate": 3.853424111704879e-06,
+      "loss": 0.214,
+      "step": 2685
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.789766692977329,
+      "learning_rate": 3.849980844738308e-06,
+      "loss": 0.1923,
+      "step": 2686
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7684060124203161,
+      "learning_rate": 3.846538153528253e-06,
+      "loss": 0.2049,
+      "step": 2687
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8022171139456236,
+      "learning_rate": 3.843096039798293e-06,
+      "loss": 0.2203,
+      "step": 2688
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7531670641769167,
+      "learning_rate": 3.839654505271724e-06,
+      "loss": 0.2098,
+      "step": 2689
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8624389530355157,
+      "learning_rate": 3.836213551671547e-06,
+      "loss": 0.2089,
+      "step": 2690
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8363938738709132,
+      "learning_rate": 3.832773180720475e-06,
+      "loss": 0.2264,
+      "step": 2691
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8901364074672697,
+      "learning_rate": 3.829333394140924e-06,
+      "loss": 0.215,
+      "step": 2692
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9174872450270926,
+      "learning_rate": 3.825894193655026e-06,
+      "loss": 0.2108,
+      "step": 2693
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9179067908785036,
+      "learning_rate": 3.822455580984613e-06,
+      "loss": 0.2056,
+      "step": 2694
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8242085262404026,
+      "learning_rate": 3.819017557851226e-06,
+      "loss": 0.2307,
+      "step": 2695
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8930091853683368,
+      "learning_rate": 3.8155801259761065e-06,
+      "loss": 0.2191,
+      "step": 2696
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0607483866972187,
+      "learning_rate": 3.812143287080205e-06,
+      "loss": 0.2054,
+      "step": 2697
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8284875194496906,
+      "learning_rate": 3.808707042884176e-06,
+      "loss": 0.2207,
+      "step": 2698
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.751861844377927,
+      "learning_rate": 3.805271395108368e-06,
+      "loss": 0.197,
+      "step": 2699
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7968277298327556,
+      "learning_rate": 3.8018363454728415e-06,
+      "loss": 0.2112,
+      "step": 2700
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.782236480738653,
+      "learning_rate": 3.7984018956973467e-06,
+      "loss": 0.2138,
+      "step": 2701
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8055326380244255,
+      "learning_rate": 3.794968047501345e-06,
+      "loss": 0.2133,
+      "step": 2702
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8410589550324171,
+      "learning_rate": 3.7915348026039877e-06,
+      "loss": 0.2203,
+      "step": 2703
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8644007681922948,
+      "learning_rate": 3.7881021627241283e-06,
+      "loss": 0.2377,
+      "step": 2704
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8894125556812655,
+      "learning_rate": 3.7846701295803136e-06,
+      "loss": 0.2263,
+      "step": 2705
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7992052339053118,
+      "learning_rate": 3.781238704890793e-06,
+      "loss": 0.2216,
+      "step": 2706
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8003924821129506,
+      "learning_rate": 3.7778078903735044e-06,
+      "loss": 0.2112,
+      "step": 2707
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.777879250999321,
+      "learning_rate": 3.7743776877460864e-06,
+      "loss": 0.2088,
+      "step": 2708
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9089829696459732,
+      "learning_rate": 3.770948098725864e-06,
+      "loss": 0.2256,
+      "step": 2709
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7628088876042391,
+      "learning_rate": 3.767519125029865e-06,
+      "loss": 0.2218,
+      "step": 2710
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9218064436316533,
+      "learning_rate": 3.7640907683747977e-06,
+      "loss": 0.2143,
+      "step": 2711
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7842350004331204,
+      "learning_rate": 3.760663030477072e-06,
+      "loss": 0.2125,
+      "step": 2712
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8557266822093587,
+      "learning_rate": 3.757235913052778e-06,
+      "loss": 0.2318,
+      "step": 2713
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7872626439790684,
+      "learning_rate": 3.7538094178177054e-06,
+      "loss": 0.2048,
+      "step": 2714
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7625279242278993,
+      "learning_rate": 3.750383546487324e-06,
+      "loss": 0.2158,
+      "step": 2715
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7475563159580398,
+      "learning_rate": 3.7469583007767964e-06,
+      "loss": 0.2079,
+      "step": 2716
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7886699262075394,
+      "learning_rate": 3.743533682400968e-06,
+      "loss": 0.2064,
+      "step": 2717
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8687639664861609,
+      "learning_rate": 3.7401096930743753e-06,
+      "loss": 0.2198,
+      "step": 2718
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8026490972681886,
+      "learning_rate": 3.7366863345112334e-06,
+      "loss": 0.2355,
+      "step": 2719
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8235233294813372,
+      "learning_rate": 3.733263608425447e-06,
+      "loss": 0.2161,
+      "step": 2720
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7378118749018426,
+      "learning_rate": 3.729841516530604e-06,
+      "loss": 0.2011,
+      "step": 2721
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8062445683282076,
+      "learning_rate": 3.72642006053997e-06,
+      "loss": 0.2074,
+      "step": 2722
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7640462170561314,
+      "learning_rate": 3.722999242166497e-06,
+      "loss": 0.2065,
+      "step": 2723
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8583164965637274,
+      "learning_rate": 3.7195790631228133e-06,
+      "loss": 0.2283,
+      "step": 2724
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8192044952321693,
+      "learning_rate": 3.7161595251212334e-06,
+      "loss": 0.235,
+      "step": 2725
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0737986645810607,
+      "learning_rate": 3.7127406298737458e-06,
+      "loss": 0.2081,
+      "step": 2726
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0047447236045286,
+      "learning_rate": 3.709322379092019e-06,
+      "loss": 0.1965,
+      "step": 2727
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7498639835860041,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.227,
+      "step": 2728
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7863594699321826,
+      "learning_rate": 3.702487817770903e-06,
+      "loss": 0.2002,
+      "step": 2729
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8722895658325807,
+      "learning_rate": 3.6990715106532353e-06,
+      "loss": 0.2222,
+      "step": 2730
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9042550773153962,
+      "learning_rate": 3.6956558548447665e-06,
+      "loss": 0.2152,
+      "step": 2731
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8207573916621742,
+      "learning_rate": 3.69224085205554e-06,
+      "loss": 0.2227,
+      "step": 2732
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7894456883541627,
+      "learning_rate": 3.6888265039952796e-06,
+      "loss": 0.2019,
+      "step": 2733
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7945884666770742,
+      "learning_rate": 3.6854128123733736e-06,
+      "loss": 0.2035,
+      "step": 2734
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7873938854386998,
+      "learning_rate": 3.681999778898887e-06,
+      "loss": 0.21,
+      "step": 2735
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8075574838364137,
+      "learning_rate": 3.6785874052805515e-06,
+      "loss": 0.2126,
+      "step": 2736
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8422908656622291,
+      "learning_rate": 3.6751756932267725e-06,
+      "loss": 0.227,
+      "step": 2737
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0890629661388975,
+      "learning_rate": 3.6717646444456196e-06,
+      "loss": 0.2147,
+      "step": 2738
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.168179302557617,
+      "learning_rate": 3.6683542606448347e-06,
+      "loss": 0.225,
+      "step": 2739
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7560219225777031,
+      "learning_rate": 3.664944543531822e-06,
+      "loss": 0.2006,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7833714412026251,
+      "learning_rate": 3.6615354948136584e-06,
+      "loss": 0.2259,
+      "step": 2741
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.824126519518729,
+      "learning_rate": 3.658127116197079e-06,
+      "loss": 0.2035,
+      "step": 2742
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8380871281344319,
+      "learning_rate": 3.6547194093884907e-06,
+      "loss": 0.2292,
+      "step": 2743
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.808523131298041,
+      "learning_rate": 3.651312376093955e-06,
+      "loss": 0.222,
+      "step": 2744
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8053456040381153,
+      "learning_rate": 3.6479060180192038e-06,
+      "loss": 0.2189,
+      "step": 2745
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8257237328447368,
+      "learning_rate": 3.6445003368696317e-06,
+      "loss": 0.221,
+      "step": 2746
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8529833844505347,
+      "learning_rate": 3.6410953343502875e-06,
+      "loss": 0.1962,
+      "step": 2747
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7587566326588328,
+      "learning_rate": 3.6376910121658867e-06,
+      "loss": 0.1906,
+      "step": 2748
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7715162420569016,
+      "learning_rate": 3.6342873720207972e-06,
+      "loss": 0.1935,
+      "step": 2749
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.925315603073503,
+      "learning_rate": 3.6308844156190546e-06,
+      "loss": 0.2069,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8213358492967576,
+      "learning_rate": 3.6274821446643445e-06,
+      "loss": 0.1905,
+      "step": 2751
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7646888885341483,
+      "learning_rate": 3.6240805608600137e-06,
+      "loss": 0.205,
+      "step": 2752
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8249740842799052,
+      "learning_rate": 3.6206796659090605e-06,
+      "loss": 0.1913,
+      "step": 2753
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7856231985282333,
+      "learning_rate": 3.6172794615141448e-06,
+      "loss": 0.199,
+      "step": 2754
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8382855119044373,
+      "learning_rate": 3.6138799493775758e-06,
+      "loss": 0.2045,
+      "step": 2755
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8620882892953495,
+      "learning_rate": 3.6104811312013175e-06,
+      "loss": 0.2172,
+      "step": 2756
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9100666307384393,
+      "learning_rate": 3.607083008686985e-06,
+      "loss": 0.211,
+      "step": 2757
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8014815158820463,
+      "learning_rate": 3.60368558353585e-06,
+      "loss": 0.2137,
+      "step": 2758
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8525537874270745,
+      "learning_rate": 3.6002888574488294e-06,
+      "loss": 0.2106,
+      "step": 2759
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8295517799011605,
+      "learning_rate": 3.596892832126494e-06,
+      "loss": 0.2095,
+      "step": 2760
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8464753011418003,
+      "learning_rate": 3.5934975092690585e-06,
+      "loss": 0.2102,
+      "step": 2761
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7500633774953718,
+      "learning_rate": 3.590102890576394e-06,
+      "loss": 0.2193,
+      "step": 2762
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8168078383300919,
+      "learning_rate": 3.5867089777480124e-06,
+      "loss": 0.2315,
+      "step": 2763
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7674131568797931,
+      "learning_rate": 3.5833157724830757e-06,
+      "loss": 0.2251,
+      "step": 2764
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8285083986400734,
+      "learning_rate": 3.579923276480387e-06,
+      "loss": 0.195,
+      "step": 2765
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8264527190354325,
+      "learning_rate": 3.576531491438403e-06,
+      "loss": 0.2058,
+      "step": 2766
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9047883854013061,
+      "learning_rate": 3.5731404190552154e-06,
+      "loss": 0.2049,
+      "step": 2767
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.81354452984204,
+      "learning_rate": 3.569750061028565e-06,
+      "loss": 0.2112,
+      "step": 2768
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7357270736869519,
+      "learning_rate": 3.56636041905583e-06,
+      "loss": 0.1998,
+      "step": 2769
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7888351671283148,
+      "learning_rate": 3.5629714948340354e-06,
+      "loss": 0.1969,
+      "step": 2770
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8092879626060412,
+      "learning_rate": 3.5595832900598453e-06,
+      "loss": 0.2061,
+      "step": 2771
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8770583302925253,
+      "learning_rate": 3.5561958064295585e-06,
+      "loss": 0.2155,
+      "step": 2772
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0551199583741129,
+      "learning_rate": 3.552809045639123e-06,
+      "loss": 0.2053,
+      "step": 2773
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.4446290328432512,
+      "learning_rate": 3.5494230093841144e-06,
+      "loss": 0.2297,
+      "step": 2774
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8138983975365861,
+      "learning_rate": 3.546037699359751e-06,
+      "loss": 0.2071,
+      "step": 2775
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.2383927350741737,
+      "learning_rate": 3.5426531172608858e-06,
+      "loss": 0.2008,
+      "step": 2776
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9343738112156047,
+      "learning_rate": 3.5392692647820113e-06,
+      "loss": 0.2148,
+      "step": 2777
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8801096045524948,
+      "learning_rate": 3.5358861436172487e-06,
+      "loss": 0.2069,
+      "step": 2778
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.257837225174746,
+      "learning_rate": 3.5325037554603576e-06,
+      "loss": 0.2076,
+      "step": 2779
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8881055040863564,
+      "learning_rate": 3.5291221020047257e-06,
+      "loss": 0.2198,
+      "step": 2780
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8318155944355023,
+      "learning_rate": 3.525741184943381e-06,
+      "loss": 0.2128,
+      "step": 2781
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8044154370054986,
+      "learning_rate": 3.5223610059689733e-06,
+      "loss": 0.2156,
+      "step": 2782
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.795474996817346,
+      "learning_rate": 3.5189815667737916e-06,
+      "loss": 0.2198,
+      "step": 2783
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9063106789418123,
+      "learning_rate": 3.515602869049746e-06,
+      "loss": 0.1813,
+      "step": 2784
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.859082493513603,
+      "learning_rate": 3.5122249144883848e-06,
+      "loss": 0.2191,
+      "step": 2785
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9383517012001301,
+      "learning_rate": 3.5088477047808738e-06,
+      "loss": 0.2268,
+      "step": 2786
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8090537403893789,
+      "learning_rate": 3.505471241618018e-06,
+      "loss": 0.2119,
+      "step": 2787
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7994057305100699,
+      "learning_rate": 3.5020955266902344e-06,
+      "loss": 0.1921,
+      "step": 2788
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.820362548934755,
+      "learning_rate": 3.4987205616875803e-06,
+      "loss": 0.1962,
+      "step": 2789
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9487500419937607,
+      "learning_rate": 3.495346348299724e-06,
+      "loss": 0.2002,
+      "step": 2790
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.769170201592449,
+      "learning_rate": 3.4919728882159685e-06,
+      "loss": 0.2112,
+      "step": 2791
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8584623214952484,
+      "learning_rate": 3.4886001831252324e-06,
+      "loss": 0.2,
+      "step": 2792
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7704708766190941,
+      "learning_rate": 3.485228234716058e-06,
+      "loss": 0.2163,
+      "step": 2793
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8697074810965529,
+      "learning_rate": 3.481857044676613e-06,
+      "loss": 0.2288,
+      "step": 2794
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7892362518298853,
+      "learning_rate": 3.478486614694679e-06,
+      "loss": 0.2165,
+      "step": 2795
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.031211885076577,
+      "learning_rate": 3.475116946457662e-06,
+      "loss": 0.2192,
+      "step": 2796
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7983405624895716,
+      "learning_rate": 3.4717480416525815e-06,
+      "loss": 0.2119,
+      "step": 2797
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8331655199132011,
+      "learning_rate": 3.4683799019660834e-06,
+      "loss": 0.2199,
+      "step": 2798
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7943604321937568,
+      "learning_rate": 3.4650125290844207e-06,
+      "loss": 0.1895,
+      "step": 2799
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7717375922540782,
+      "learning_rate": 3.4616459246934686e-06,
+      "loss": 0.2082,
+      "step": 2800
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8017836509465229,
+      "learning_rate": 3.4582800904787138e-06,
+      "loss": 0.2189,
+      "step": 2801
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8038280350137886,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.2038,
+      "step": 2802
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8180199071041132,
+      "learning_rate": 3.4515507393178316e-06,
+      "loss": 0.2127,
+      "step": 2803
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7087132598069035,
+      "learning_rate": 3.448187225740749e-06,
+      "loss": 0.1989,
+      "step": 2804
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7874444836488699,
+      "learning_rate": 3.4448244890779544e-06,
+      "loss": 0.2095,
+      "step": 2805
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7672821553300931,
+      "learning_rate": 3.4414625310130038e-06,
+      "loss": 0.217,
+      "step": 2806
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8106756033215952,
+      "learning_rate": 3.4381013532290575e-06,
+      "loss": 0.2034,
+      "step": 2807
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0268925042223673,
+      "learning_rate": 3.4347409574088896e-06,
+      "loss": 0.2148,
+      "step": 2808
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8078968722736017,
+      "learning_rate": 3.4313813452348774e-06,
+      "loss": 0.2176,
+      "step": 2809
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9095869503983214,
+      "learning_rate": 3.4280225183890124e-06,
+      "loss": 0.2203,
+      "step": 2810
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7965474866009801,
+      "learning_rate": 3.424664478552887e-06,
+      "loss": 0.2093,
+      "step": 2811
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8385626901121965,
+      "learning_rate": 3.4213072274077047e-06,
+      "loss": 0.221,
+      "step": 2812
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0054854881237683,
+      "learning_rate": 3.417950766634268e-06,
+      "loss": 0.2131,
+      "step": 2813
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9708274104079531,
+      "learning_rate": 3.4145950979129916e-06,
+      "loss": 0.2138,
+      "step": 2814
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8530198656102499,
+      "learning_rate": 3.411240222923886e-06,
+      "loss": 0.2394,
+      "step": 2815
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7807931749460651,
+      "learning_rate": 3.4078861433465693e-06,
+      "loss": 0.1994,
+      "step": 2816
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8207139070201468,
+      "learning_rate": 3.404532860860258e-06,
+      "loss": 0.2268,
+      "step": 2817
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7696307446805436,
+      "learning_rate": 3.401180377143774e-06,
+      "loss": 0.2188,
+      "step": 2818
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8107930026307134,
+      "learning_rate": 3.3978286938755357e-06,
+      "loss": 0.2152,
+      "step": 2819
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8308033832542051,
+      "learning_rate": 3.39447781273356e-06,
+      "loss": 0.223,
+      "step": 2820
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9074587088026733,
+      "learning_rate": 3.3911277353954665e-06,
+      "loss": 0.218,
+      "step": 2821
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8240663397596868,
+      "learning_rate": 3.3877784635384668e-06,
+      "loss": 0.215,
+      "step": 2822
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7272519922375972,
+      "learning_rate": 3.3844299988393757e-06,
+      "loss": 0.2107,
+      "step": 2823
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7581516774408502,
+      "learning_rate": 3.3810823429745955e-06,
+      "loss": 0.2014,
+      "step": 2824
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7938744863258874,
+      "learning_rate": 3.3777354976201338e-06,
+      "loss": 0.2098,
+      "step": 2825
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1152120435671204,
+      "learning_rate": 3.374389464451583e-06,
+      "loss": 0.2038,
+      "step": 2826
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7490356402595377,
+      "learning_rate": 3.3710442451441374e-06,
+      "loss": 0.2016,
+      "step": 2827
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9343813686179344,
+      "learning_rate": 3.3676998413725726e-06,
+      "loss": 0.2236,
+      "step": 2828
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8493733132035036,
+      "learning_rate": 3.36435625481127e-06,
+      "loss": 0.2224,
+      "step": 2829
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8692870460485675,
+      "learning_rate": 3.361013487134188e-06,
+      "loss": 0.1949,
+      "step": 2830
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8357456641730054,
+      "learning_rate": 3.3576715400148863e-06,
+      "loss": 0.217,
+      "step": 2831
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8265222680892984,
+      "learning_rate": 3.354330415126508e-06,
+      "loss": 0.2184,
+      "step": 2832
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8004658420662647,
+      "learning_rate": 3.3509901141417845e-06,
+      "loss": 0.2246,
+      "step": 2833
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7627389703257154,
+      "learning_rate": 3.3476506387330355e-06,
+      "loss": 0.1985,
+      "step": 2834
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.826368784062843,
+      "learning_rate": 3.34431199057217e-06,
+      "loss": 0.2045,
+      "step": 2835
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7741983129226397,
+      "learning_rate": 3.3409741713306777e-06,
+      "loss": 0.2191,
+      "step": 2836
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9174956927942635,
+      "learning_rate": 3.3376371826796393e-06,
+      "loss": 0.228,
+      "step": 2837
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8442943044336269,
+      "learning_rate": 3.3343010262897125e-06,
+      "loss": 0.2224,
+      "step": 2838
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8374390463103089,
+      "learning_rate": 3.330965703831146e-06,
+      "loss": 0.2045,
+      "step": 2839
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8073994007170352,
+      "learning_rate": 3.327631216973766e-06,
+      "loss": 0.2165,
+      "step": 2840
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9459782259564232,
+      "learning_rate": 3.324297567386979e-06,
+      "loss": 0.2135,
+      "step": 2841
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.455618060619485,
+      "learning_rate": 3.320964756739781e-06,
+      "loss": 0.1922,
+      "step": 2842
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7881845705328274,
+      "learning_rate": 3.3176327867007376e-06,
+      "loss": 0.2057,
+      "step": 2843
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7644482263233994,
+      "learning_rate": 3.3143016589379984e-06,
+      "loss": 0.1959,
+      "step": 2844
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7926361977733939,
+      "learning_rate": 3.310971375119291e-06,
+      "loss": 0.2214,
+      "step": 2845
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.770737080489669,
+      "learning_rate": 3.3076419369119216e-06,
+      "loss": 0.2022,
+      "step": 2846
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9628586282917673,
+      "learning_rate": 3.3043133459827703e-06,
+      "loss": 0.2151,
+      "step": 2847
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.120525691368218,
+      "learning_rate": 3.300985603998296e-06,
+      "loss": 0.2028,
+      "step": 2848
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.522616310638525,
+      "learning_rate": 3.2976587126245275e-06,
+      "loss": 0.1982,
+      "step": 2849
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9215433474975548,
+      "learning_rate": 3.2943326735270766e-06,
+      "loss": 0.1824,
+      "step": 2850
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8824943364737484,
+      "learning_rate": 3.2910074883711186e-06,
+      "loss": 0.2238,
+      "step": 2851
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8591385892532555,
+      "learning_rate": 3.2876831588214093e-06,
+      "loss": 0.2115,
+      "step": 2852
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8157937071559049,
+      "learning_rate": 3.2843596865422687e-06,
+      "loss": 0.2058,
+      "step": 2853
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8433628029952621,
+      "learning_rate": 3.2810370731975956e-06,
+      "loss": 0.2042,
+      "step": 2854
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8222083079502595,
+      "learning_rate": 3.2777153204508516e-06,
+      "loss": 0.2334,
+      "step": 2855
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7694144421714925,
+      "learning_rate": 3.2743944299650733e-06,
+      "loss": 0.1973,
+      "step": 2856
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9327678773273042,
+      "learning_rate": 3.2710744034028598e-06,
+      "loss": 0.2068,
+      "step": 2857
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8206881066599493,
+      "learning_rate": 3.2677552424263836e-06,
+      "loss": 0.2324,
+      "step": 2858
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7965491863202331,
+      "learning_rate": 3.2644369486973786e-06,
+      "loss": 0.2131,
+      "step": 2859
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8724099438136644,
+      "learning_rate": 3.2611195238771494e-06,
+      "loss": 0.2234,
+      "step": 2860
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8431898531793434,
+      "learning_rate": 3.2578029696265596e-06,
+      "loss": 0.1936,
+      "step": 2861
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8673959852461565,
+      "learning_rate": 3.254487287606044e-06,
+      "loss": 0.2115,
+      "step": 2862
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8160537313563696,
+      "learning_rate": 3.251172479475595e-06,
+      "loss": 0.2187,
+      "step": 2863
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8348664651142109,
+      "learning_rate": 3.247858546894772e-06,
+      "loss": 0.2082,
+      "step": 2864
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7785917053320802,
+      "learning_rate": 3.244545491522689e-06,
+      "loss": 0.1925,
+      "step": 2865
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8001266074188069,
+      "learning_rate": 3.2412333150180298e-06,
+      "loss": 0.1909,
+      "step": 2866
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4054316139368854,
+      "learning_rate": 3.2379220190390337e-06,
+      "loss": 0.2124,
+      "step": 2867
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7678717240282915,
+      "learning_rate": 3.234611605243496e-06,
+      "loss": 0.1758,
+      "step": 2868
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8648807246427335,
+      "learning_rate": 3.231302075288778e-06,
+      "loss": 0.233,
+      "step": 2869
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7983806788813163,
+      "learning_rate": 3.2279934308317905e-06,
+      "loss": 0.2042,
+      "step": 2870
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9997638007643157,
+      "learning_rate": 3.2246856735290096e-06,
+      "loss": 0.2269,
+      "step": 2871
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8708347214245391,
+      "learning_rate": 3.2213788050364557e-06,
+      "loss": 0.2162,
+      "step": 2872
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8992070926422012,
+      "learning_rate": 3.2180728270097163e-06,
+      "loss": 0.2279,
+      "step": 2873
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8263007225870138,
+      "learning_rate": 3.2147677411039236e-06,
+      "loss": 0.2089,
+      "step": 2874
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7895979073180245,
+      "learning_rate": 3.2114635489737715e-06,
+      "loss": 0.2138,
+      "step": 2875
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.800447611226406,
+      "learning_rate": 3.2081602522734987e-06,
+      "loss": 0.1886,
+      "step": 2876
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8031344204426796,
+      "learning_rate": 3.204857852656902e-06,
+      "loss": 0.2173,
+      "step": 2877
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8469409960576493,
+      "learning_rate": 3.2015563517773214e-06,
+      "loss": 0.1834,
+      "step": 2878
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8452527541463382,
+      "learning_rate": 3.1982557512876567e-06,
+      "loss": 0.2001,
+      "step": 2879
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7885876336272336,
+      "learning_rate": 3.1949560528403487e-06,
+      "loss": 0.2057,
+      "step": 2880
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7390680381458439,
+      "learning_rate": 3.1916572580873927e-06,
+      "loss": 0.2044,
+      "step": 2881
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7883179827529321,
+      "learning_rate": 3.1883593686803244e-06,
+      "loss": 0.2029,
+      "step": 2882
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8175460474338475,
+      "learning_rate": 3.1850623862702344e-06,
+      "loss": 0.2201,
+      "step": 2883
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.876214402126467,
+      "learning_rate": 3.1817663125077534e-06,
+      "loss": 0.1968,
+      "step": 2884
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0901333707949346,
+      "learning_rate": 3.178471149043061e-06,
+      "loss": 0.2353,
+      "step": 2885
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8323768413230866,
+      "learning_rate": 3.1751768975258745e-06,
+      "loss": 0.1977,
+      "step": 2886
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.818184869630506,
+      "learning_rate": 3.171883559605465e-06,
+      "loss": 0.2152,
+      "step": 2887
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.314727881712965,
+      "learning_rate": 3.1685911369306364e-06,
+      "loss": 0.1912,
+      "step": 2888
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8840837992883462,
+      "learning_rate": 3.16529963114974e-06,
+      "loss": 0.1944,
+      "step": 2889
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7965007599408117,
+      "learning_rate": 3.162009043910668e-06,
+      "loss": 0.2017,
+      "step": 2890
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7986781673244749,
+      "learning_rate": 3.158719376860848e-06,
+      "loss": 0.1882,
+      "step": 2891
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.823466215310121,
+      "learning_rate": 3.155430631647254e-06,
+      "loss": 0.2099,
+      "step": 2892
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7589234678520347,
+      "learning_rate": 3.1521428099163897e-06,
+      "loss": 0.2052,
+      "step": 2893
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.866879450281189,
+      "learning_rate": 3.1488559133143065e-06,
+      "loss": 0.2274,
+      "step": 2894
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8268631457808928,
+      "learning_rate": 3.1455699434865844e-06,
+      "loss": 0.2062,
+      "step": 2895
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8874755041026242,
+      "learning_rate": 3.1422849020783442e-06,
+      "loss": 0.194,
+      "step": 2896
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8120941992261204,
+      "learning_rate": 3.139000790734238e-06,
+      "loss": 0.223,
+      "step": 2897
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8186848159095814,
+      "learning_rate": 3.1357176110984578e-06,
+      "loss": 0.2177,
+      "step": 2898
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9324707965792035,
+      "learning_rate": 3.132435364814723e-06,
+      "loss": 0.2223,
+      "step": 2899
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9040049393560066,
+      "learning_rate": 3.1291540535262905e-06,
+      "loss": 0.2166,
+      "step": 2900
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7531530044476393,
+      "learning_rate": 3.1258736788759443e-06,
+      "loss": 0.2013,
+      "step": 2901
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7545410400535517,
+      "learning_rate": 3.122594242506006e-06,
+      "loss": 0.2059,
+      "step": 2902
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9177530174030278,
+      "learning_rate": 3.1193157460583217e-06,
+      "loss": 0.2258,
+      "step": 2903
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8184140846778939,
+      "learning_rate": 3.11603819117427e-06,
+      "loss": 0.2039,
+      "step": 2904
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.019865945808397,
+      "learning_rate": 3.1127615794947546e-06,
+      "loss": 0.222,
+      "step": 2905
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7790758474309414,
+      "learning_rate": 3.1094859126602132e-06,
+      "loss": 0.1852,
+      "step": 2906
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8184845932731135,
+      "learning_rate": 3.1062111923106044e-06,
+      "loss": 0.1925,
+      "step": 2907
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8997262243246836,
+      "learning_rate": 3.1029374200854167e-06,
+      "loss": 0.2012,
+      "step": 2908
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8170131292887797,
+      "learning_rate": 3.0996645976236593e-06,
+      "loss": 0.2316,
+      "step": 2909
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8513443895186591,
+      "learning_rate": 3.096392726563874e-06,
+      "loss": 0.2111,
+      "step": 2910
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.800414666000089,
+      "learning_rate": 3.0931218085441174e-06,
+      "loss": 0.2214,
+      "step": 2911
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9514433083368672,
+      "learning_rate": 3.089851845201976e-06,
+      "loss": 0.2179,
+      "step": 2912
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7876499770546943,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.1967,
+      "step": 2913
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8427186362894981,
+      "learning_rate": 3.0833147890984735e-06,
+      "loss": 0.2309,
+      "step": 2914
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.098042435688673,
+      "learning_rate": 3.08004769960989e-06,
+      "loss": 0.2056,
+      "step": 2915
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.834672146088912,
+      "learning_rate": 3.0767815713444655e-06,
+      "loss": 0.2089,
+      "step": 2916
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7642789181355669,
+      "learning_rate": 3.0735164059373872e-06,
+      "loss": 0.2078,
+      "step": 2917
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8941663943733157,
+      "learning_rate": 3.070252205023356e-06,
+      "loss": 0.2263,
+      "step": 2918
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7818872621497082,
+      "learning_rate": 3.066988970236596e-06,
+      "loss": 0.195,
+      "step": 2919
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8106151063298087,
+      "learning_rate": 3.063726703210841e-06,
+      "loss": 0.2126,
+      "step": 2920
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8270373845797028,
+      "learning_rate": 3.0604654055793446e-06,
+      "loss": 0.2073,
+      "step": 2921
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7908878893270318,
+      "learning_rate": 3.0572050789748732e-06,
+      "loss": 0.2084,
+      "step": 2922
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7813334381726155,
+      "learning_rate": 3.0539457250297095e-06,
+      "loss": 0.2051,
+      "step": 2923
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8691637265220079,
+      "learning_rate": 3.050687345375645e-06,
+      "loss": 0.2443,
+      "step": 2924
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.767239237045816,
+      "learning_rate": 3.0474299416439885e-06,
+      "loss": 0.2106,
+      "step": 2925
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.889950754540831,
+      "learning_rate": 3.044173515465554e-06,
+      "loss": 0.2164,
+      "step": 2926
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9144201304487327,
+      "learning_rate": 3.0409180684706728e-06,
+      "loss": 0.2612,
+      "step": 2927
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9082517662681858,
+      "learning_rate": 3.0376636022891813e-06,
+      "loss": 0.2384,
+      "step": 2928
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.0682680731306011,
+      "learning_rate": 3.034410118550428e-06,
+      "loss": 0.2164,
+      "step": 2929
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7834550587850849,
+      "learning_rate": 3.031157618883265e-06,
+      "loss": 0.2191,
+      "step": 2930
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7747947399327755,
+      "learning_rate": 3.027906104916057e-06,
+      "loss": 0.192,
+      "step": 2931
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7466369789275669,
+      "learning_rate": 3.0246555782766723e-06,
+      "loss": 0.1948,
+      "step": 2932
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8742980317938105,
+      "learning_rate": 3.0214060405924863e-06,
+      "loss": 0.2283,
+      "step": 2933
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7512587356624814,
+      "learning_rate": 3.0181574934903745e-06,
+      "loss": 0.1967,
+      "step": 2934
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8033649153549249,
+      "learning_rate": 3.0149099385967262e-06,
+      "loss": 0.2138,
+      "step": 2935
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8542524534739685,
+      "learning_rate": 3.0116633775374237e-06,
+      "loss": 0.2143,
+      "step": 2936
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8489850573023486,
+      "learning_rate": 3.0084178119378584e-06,
+      "loss": 0.1985,
+      "step": 2937
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7874167132796998,
+      "learning_rate": 3.0051732434229185e-06,
+      "loss": 0.2051,
+      "step": 2938
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.762124797401922,
+      "learning_rate": 3.001929673616999e-06,
+      "loss": 0.2208,
+      "step": 2939
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7638137001085864,
+      "learning_rate": 2.998687104143991e-06,
+      "loss": 0.1998,
+      "step": 2940
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7582412271812757,
+      "learning_rate": 2.995445536627282e-06,
+      "loss": 0.1723,
+      "step": 2941
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7677871772705294,
+      "learning_rate": 2.992204972689765e-06,
+      "loss": 0.2163,
+      "step": 2942
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.800589097164039,
+      "learning_rate": 2.988965413953825e-06,
+      "loss": 0.2123,
+      "step": 2943
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8810546387713858,
+      "learning_rate": 2.985726862041346e-06,
+      "loss": 0.2104,
+      "step": 2944
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.83653960834497,
+      "learning_rate": 2.9824893185737046e-06,
+      "loss": 0.1988,
+      "step": 2945
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8335092083816997,
+      "learning_rate": 2.9792527851717805e-06,
+      "loss": 0.2274,
+      "step": 2946
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8694844172672435,
+      "learning_rate": 2.976017263455939e-06,
+      "loss": 0.1951,
+      "step": 2947
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7910645113232483,
+      "learning_rate": 2.972782755046043e-06,
+      "loss": 0.2024,
+      "step": 2948
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9111514158457229,
+      "learning_rate": 2.969549261561447e-06,
+      "loss": 0.2087,
+      "step": 2949
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7776743223330237,
+      "learning_rate": 2.966316784621e-06,
+      "loss": 0.2104,
+      "step": 2950
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9741260405116585,
+      "learning_rate": 2.9630853258430376e-06,
+      "loss": 0.1985,
+      "step": 2951
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8056070285973302,
+      "learning_rate": 2.9598548868453904e-06,
+      "loss": 0.1999,
+      "step": 2952
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9874318528656963,
+      "learning_rate": 2.956625469245372e-06,
+      "loss": 0.2087,
+      "step": 2953
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.422678933886513,
+      "learning_rate": 2.9533970746597935e-06,
+      "loss": 0.2174,
+      "step": 2954
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8799884634878385,
+      "learning_rate": 2.950169704704945e-06,
+      "loss": 0.1979,
+      "step": 2955
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7074717093005782,
+      "learning_rate": 2.9469433609966125e-06,
+      "loss": 0.2136,
+      "step": 2956
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.869928348627164,
+      "learning_rate": 2.943718045150058e-06,
+      "loss": 0.2119,
+      "step": 2957
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9459169507278187,
+      "learning_rate": 2.9404937587800374e-06,
+      "loss": 0.2029,
+      "step": 2958
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8616734248965919,
+      "learning_rate": 2.9372705035007863e-06,
+      "loss": 0.2002,
+      "step": 2959
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0253349226411435,
+      "learning_rate": 2.9340482809260294e-06,
+      "loss": 0.197,
+      "step": 2960
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8210676210928378,
+      "learning_rate": 2.9308270926689663e-06,
+      "loss": 0.2185,
+      "step": 2961
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9652384864864089,
+      "learning_rate": 2.9276069403422846e-06,
+      "loss": 0.2038,
+      "step": 2962
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8077990767343116,
+      "learning_rate": 2.924387825558155e-06,
+      "loss": 0.2158,
+      "step": 2963
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8562211680674319,
+      "learning_rate": 2.9211697499282223e-06,
+      "loss": 0.2022,
+      "step": 2964
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.886167552788557,
+      "learning_rate": 2.9179527150636155e-06,
+      "loss": 0.2196,
+      "step": 2965
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7438096099851934,
+      "learning_rate": 2.9147367225749425e-06,
+      "loss": 0.1964,
+      "step": 2966
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.800249671881866,
+      "learning_rate": 2.9115217740722866e-06,
+      "loss": 0.2078,
+      "step": 2967
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8087610158092343,
+      "learning_rate": 2.90830787116521e-06,
+      "loss": 0.2086,
+      "step": 2968
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8005793800234628,
+      "learning_rate": 2.9050950154627546e-06,
+      "loss": 0.2082,
+      "step": 2969
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7691700099881706,
+      "learning_rate": 2.90188320857343e-06,
+      "loss": 0.2054,
+      "step": 2970
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8261783254806907,
+      "learning_rate": 2.898672452105231e-06,
+      "loss": 0.2203,
+      "step": 2971
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7733720594350387,
+      "learning_rate": 2.895462747665617e-06,
+      "loss": 0.1981,
+      "step": 2972
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8632246548942037,
+      "learning_rate": 2.892254096861529e-06,
+      "loss": 0.2134,
+      "step": 2973
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7999492179998566,
+      "learning_rate": 2.889046501299373e-06,
+      "loss": 0.21,
+      "step": 2974
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8364586691859411,
+      "learning_rate": 2.885839962585032e-06,
+      "loss": 0.2212,
+      "step": 2975
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8462249184833399,
+      "learning_rate": 2.882634482323856e-06,
+      "loss": 0.2138,
+      "step": 2976
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7879892354405302,
+      "learning_rate": 2.8794300621206695e-06,
+      "loss": 0.2016,
+      "step": 2977
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.9123861407506638,
+      "learning_rate": 2.8762267035797607e-06,
+      "loss": 0.2152,
+      "step": 2978
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8766055075154338,
+      "learning_rate": 2.873024408304894e-06,
+      "loss": 0.2138,
+      "step": 2979
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7937032321072666,
+      "learning_rate": 2.8698231778992925e-06,
+      "loss": 0.2069,
+      "step": 2980
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.0213536848525122,
+      "learning_rate": 2.866623013965655e-06,
+      "loss": 0.2035,
+      "step": 2981
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.810234490173411,
+      "learning_rate": 2.8634239181061384e-06,
+      "loss": 0.2047,
+      "step": 2982
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8145518180751014,
+      "learning_rate": 2.8602258919223703e-06,
+      "loss": 0.2126,
+      "step": 2983
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7684423473907966,
+      "learning_rate": 2.857028937015438e-06,
+      "loss": 0.2027,
+      "step": 2984
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.788439625778795,
+      "learning_rate": 2.8538330549858993e-06,
+      "loss": 0.201,
+      "step": 2985
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8990496384347014,
+      "learning_rate": 2.850638247433768e-06,
+      "loss": 0.2127,
+      "step": 2986
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8788851436781256,
+      "learning_rate": 2.8474445159585235e-06,
+      "loss": 0.2232,
+      "step": 2987
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.9162817614948497,
+      "learning_rate": 2.8442518621591085e-06,
+      "loss": 0.2172,
+      "step": 2988
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7761929976262789,
+      "learning_rate": 2.8410602876339214e-06,
+      "loss": 0.2114,
+      "step": 2989
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.903454171256981,
+      "learning_rate": 2.837869793980823e-06,
+      "loss": 0.2031,
+      "step": 2990
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8185863797446901,
+      "learning_rate": 2.834680382797129e-06,
+      "loss": 0.2043,
+      "step": 2991
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7827453924215438,
+      "learning_rate": 2.831492055679622e-06,
+      "loss": 0.2145,
+      "step": 2992
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7723305036792252,
+      "learning_rate": 2.828304814224532e-06,
+      "loss": 0.2033,
+      "step": 2993
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8644665907308277,
+      "learning_rate": 2.8251186600275533e-06,
+      "loss": 0.2044,
+      "step": 2994
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8355616489925088,
+      "learning_rate": 2.821933594683829e-06,
+      "loss": 0.2064,
+      "step": 2995
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.9481009205109034,
+      "learning_rate": 2.8187496197879637e-06,
+      "loss": 0.1948,
+      "step": 2996
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7602271610007271,
+      "learning_rate": 2.815566736934009e-06,
+      "loss": 0.1793,
+      "step": 2997
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8475545550518697,
+      "learning_rate": 2.8123849477154808e-06,
+      "loss": 0.2145,
+      "step": 2998
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7856372667978032,
+      "learning_rate": 2.8092042537253306e-06,
+      "loss": 0.1999,
+      "step": 2999
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7552246240937519,
+      "learning_rate": 2.806024656555978e-06,
+      "loss": 0.1939,
+      "step": 3000
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8005046504671984,
+      "learning_rate": 2.8028461577992837e-06,
+      "loss": 0.222,
+      "step": 3001
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8991915090027928,
+      "learning_rate": 2.799668759046564e-06,
+      "loss": 0.2034,
+      "step": 3002
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.831784619277771,
+      "learning_rate": 2.796492461888578e-06,
+      "loss": 0.2118,
+      "step": 3003
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8730386848936544,
+      "learning_rate": 2.7933172679155427e-06,
+      "loss": 0.2148,
+      "step": 3004
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7663633613604429,
+      "learning_rate": 2.790143178717113e-06,
+      "loss": 0.1905,
+      "step": 3005
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7886307240407918,
+      "learning_rate": 2.786970195882398e-06,
+      "loss": 0.2035,
+      "step": 3006
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8419191611615269,
+      "learning_rate": 2.7837983209999484e-06,
+      "loss": 0.2175,
+      "step": 3007
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.088777681256128,
+      "learning_rate": 2.7806275556577624e-06,
+      "loss": 0.2148,
+      "step": 3008
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.9303445179955662,
+      "learning_rate": 2.777457901443279e-06,
+      "loss": 0.226,
+      "step": 3009
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8850410159146159,
+      "learning_rate": 2.7742893599433875e-06,
+      "loss": 0.2085,
+      "step": 3010
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8913724905116569,
+      "learning_rate": 2.7711219327444174e-06,
+      "loss": 0.198,
+      "step": 3011
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8961499566976321,
+      "learning_rate": 2.7679556214321357e-06,
+      "loss": 0.1972,
+      "step": 3012
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7719291064639178,
+      "learning_rate": 2.764790427591759e-06,
+      "loss": 0.216,
+      "step": 3013
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7774583557138853,
+      "learning_rate": 2.761626352807938e-06,
+      "loss": 0.1964,
+      "step": 3014
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8055736990018637,
+      "learning_rate": 2.7584633986647645e-06,
+      "loss": 0.1958,
+      "step": 3015
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8281913491697434,
+      "learning_rate": 2.7553015667457694e-06,
+      "loss": 0.2114,
+      "step": 3016
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7580558260117922,
+      "learning_rate": 2.752140858633924e-06,
+      "loss": 0.186,
+      "step": 3017
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0084404497520019,
+      "learning_rate": 2.748981275911633e-06,
+      "loss": 0.2047,
+      "step": 3018
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.954198393612559,
+      "learning_rate": 2.745822820160744e-06,
+      "loss": 0.2077,
+      "step": 3019
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7416692846123539,
+      "learning_rate": 2.7426654929625307e-06,
+      "loss": 0.1918,
+      "step": 3020
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0164207507181655,
+      "learning_rate": 2.739509295897713e-06,
+      "loss": 0.214,
+      "step": 3021
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9210450379873891,
+      "learning_rate": 2.736354230546437e-06,
+      "loss": 0.2363,
+      "step": 3022
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8072038977680153,
+      "learning_rate": 2.733200298488284e-06,
+      "loss": 0.1867,
+      "step": 3023
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8636499564002508,
+      "learning_rate": 2.7300475013022666e-06,
+      "loss": 0.2055,
+      "step": 3024
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8135219966804593,
+      "learning_rate": 2.7268958405668356e-06,
+      "loss": 0.2072,
+      "step": 3025
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8316352281402635,
+      "learning_rate": 2.7237453178598645e-06,
+      "loss": 0.2093,
+      "step": 3026
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0388570263136245,
+      "learning_rate": 2.7205959347586642e-06,
+      "loss": 0.1958,
+      "step": 3027
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0104616923083285,
+      "learning_rate": 2.7174476928399685e-06,
+      "loss": 0.2173,
+      "step": 3028
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8147652200871783,
+      "learning_rate": 2.714300593679946e-06,
+      "loss": 0.2006,
+      "step": 3029
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0258507926938247,
+      "learning_rate": 2.71115463885419e-06,
+      "loss": 0.2242,
+      "step": 3030
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9538258478090904,
+      "learning_rate": 2.70800982993772e-06,
+      "loss": 0.215,
+      "step": 3031
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9154394633142257,
+      "learning_rate": 2.7048661685049814e-06,
+      "loss": 0.2091,
+      "step": 3032
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7857969020830272,
+      "learning_rate": 2.701723656129851e-06,
+      "loss": 0.2124,
+      "step": 3033
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7727338085947075,
+      "learning_rate": 2.698582294385621e-06,
+      "loss": 0.2004,
+      "step": 3034
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8405838501697791,
+      "learning_rate": 2.695442084845016e-06,
+      "loss": 0.1916,
+      "step": 3035
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8552677889515188,
+      "learning_rate": 2.6923030290801817e-06,
+      "loss": 0.2251,
+      "step": 3036
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7768239379763998,
+      "learning_rate": 2.6891651286626823e-06,
+      "loss": 0.2012,
+      "step": 3037
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8857793395887262,
+      "learning_rate": 2.6860283851635067e-06,
+      "loss": 0.2025,
+      "step": 3038
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8213548218128602,
+      "learning_rate": 2.6828928001530617e-06,
+      "loss": 0.1956,
+      "step": 3039
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1026820937795836,
+      "learning_rate": 2.679758375201181e-06,
+      "loss": 0.2011,
+      "step": 3040
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8465780028678093,
+      "learning_rate": 2.676625111877108e-06,
+      "loss": 0.2057,
+      "step": 3041
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7621701029981595,
+      "learning_rate": 2.6734930117495133e-06,
+      "loss": 0.1904,
+      "step": 3042
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8041448701260598,
+      "learning_rate": 2.670362076386478e-06,
+      "loss": 0.2183,
+      "step": 3043
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8617395508732649,
+      "learning_rate": 2.6672323073555074e-06,
+      "loss": 0.2107,
+      "step": 3044
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9571828182964619,
+      "learning_rate": 2.6641037062235143e-06,
+      "loss": 0.231,
+      "step": 3045
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7691019149323092,
+      "learning_rate": 2.6609762745568358e-06,
+      "loss": 0.197,
+      "step": 3046
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.2925235027761417,
+      "learning_rate": 2.657850013921216e-06,
+      "loss": 0.2094,
+      "step": 3047
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9808728023581575,
+      "learning_rate": 2.6547249258818162e-06,
+      "loss": 0.2036,
+      "step": 3048
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7733734052695692,
+      "learning_rate": 2.6516010120032094e-06,
+      "loss": 0.2083,
+      "step": 3049
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9067912659771177,
+      "learning_rate": 2.6484782738493843e-06,
+      "loss": 0.2084,
+      "step": 3050
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7740876750745924,
+      "learning_rate": 2.6453567129837333e-06,
+      "loss": 0.1996,
+      "step": 3051
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.762453490644448,
+      "learning_rate": 2.6422363309690702e-06,
+      "loss": 0.1955,
+      "step": 3052
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8046666311407799,
+      "learning_rate": 2.6391171293676077e-06,
+      "loss": 0.1932,
+      "step": 3053
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8241038830199734,
+      "learning_rate": 2.6359991097409767e-06,
+      "loss": 0.194,
+      "step": 3054
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.9867667844518735,
+      "learning_rate": 2.6328822736502102e-06,
+      "loss": 0.2072,
+      "step": 3055
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8365274820020729,
+      "learning_rate": 2.6297666226557505e-06,
+      "loss": 0.2416,
+      "step": 3056
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8017378501499673,
+      "learning_rate": 2.6266521583174443e-06,
+      "loss": 0.2042,
+      "step": 3057
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.125890412447613,
+      "learning_rate": 2.6235388821945497e-06,
+      "loss": 0.2099,
+      "step": 3058
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7733684020918966,
+      "learning_rate": 2.6204267958457288e-06,
+      "loss": 0.1936,
+      "step": 3059
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8047133183702092,
+      "learning_rate": 2.6173159008290407e-06,
+      "loss": 0.2146,
+      "step": 3060
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8071226874679802,
+      "learning_rate": 2.614206198701958e-06,
+      "loss": 0.2347,
+      "step": 3061
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8738502758796656,
+      "learning_rate": 2.61109769102135e-06,
+      "loss": 0.1943,
+      "step": 3062
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7998032164891477,
+      "learning_rate": 2.607990379343489e-06,
+      "loss": 0.1986,
+      "step": 3063
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8185008437066287,
+      "learning_rate": 2.604884265224048e-06,
+      "loss": 0.2162,
+      "step": 3064
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7590568488271531,
+      "learning_rate": 2.6017793502181043e-06,
+      "loss": 0.212,
+      "step": 3065
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7991992839239818,
+      "learning_rate": 2.598675635880129e-06,
+      "loss": 0.1912,
+      "step": 3066
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8881403355044615,
+      "learning_rate": 2.595573123763998e-06,
+      "loss": 0.2001,
+      "step": 3067
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8043172094272145,
+      "learning_rate": 2.59247181542298e-06,
+      "loss": 0.1895,
+      "step": 3068
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.821667480854544,
+      "learning_rate": 2.5893717124097463e-06,
+      "loss": 0.2213,
+      "step": 3069
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.823812610821691,
+      "learning_rate": 2.5862728162763595e-06,
+      "loss": 0.2066,
+      "step": 3070
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7516461430661372,
+      "learning_rate": 2.5831751285742825e-06,
+      "loss": 0.1972,
+      "step": 3071
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.172756279618626,
+      "learning_rate": 2.580078650854366e-06,
+      "loss": 0.2326,
+      "step": 3072
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8929307845299073,
+      "learning_rate": 2.576983384666867e-06,
+      "loss": 0.2085,
+      "step": 3073
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7736543841758085,
+      "learning_rate": 2.573889331561423e-06,
+      "loss": 0.2164,
+      "step": 3074
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7943622792654158,
+      "learning_rate": 2.5707964930870753e-06,
+      "loss": 0.192,
+      "step": 3075
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8141509641941209,
+      "learning_rate": 2.5677048707922482e-06,
+      "loss": 0.2085,
+      "step": 3076
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7781340336076751,
+      "learning_rate": 2.564614466224763e-06,
+      "loss": 0.2162,
+      "step": 3077
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8418142417137037,
+      "learning_rate": 2.5615252809318287e-06,
+      "loss": 0.2125,
+      "step": 3078
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6919397481188864,
+      "learning_rate": 2.558437316460044e-06,
+      "loss": 0.1753,
+      "step": 3079
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8866344653899453,
+      "learning_rate": 2.555350574355395e-06,
+      "loss": 0.2,
+      "step": 3080
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8468650906209103,
+      "learning_rate": 2.5522650561632613e-06,
+      "loss": 0.2153,
+      "step": 3081
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8265544908604783,
+      "learning_rate": 2.5491807634284016e-06,
+      "loss": 0.1994,
+      "step": 3082
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.844655877987455,
+      "learning_rate": 2.5460976976949686e-06,
+      "loss": 0.218,
+      "step": 3083
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7647810639972368,
+      "learning_rate": 2.543015860506499e-06,
+      "loss": 0.1815,
+      "step": 3084
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7612889125880175,
+      "learning_rate": 2.5399352534059087e-06,
+      "loss": 0.195,
+      "step": 3085
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0107729368680227,
+      "learning_rate": 2.536855877935507e-06,
+      "loss": 0.2266,
+      "step": 3086
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9304147806558607,
+      "learning_rate": 2.5337777356369785e-06,
+      "loss": 0.2213,
+      "step": 3087
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9424567021225977,
+      "learning_rate": 2.5307008280513956e-06,
+      "loss": 0.2076,
+      "step": 3088
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.803054059151844,
+      "learning_rate": 2.5276251567192063e-06,
+      "loss": 0.2263,
+      "step": 3089
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7835261143504805,
+      "learning_rate": 2.524550723180249e-06,
+      "loss": 0.1896,
+      "step": 3090
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8707931589991942,
+      "learning_rate": 2.5214775289737335e-06,
+      "loss": 0.2176,
+      "step": 3091
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8325161438993461,
+      "learning_rate": 2.5184055756382566e-06,
+      "loss": 0.2134,
+      "step": 3092
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0116672487870508,
+      "learning_rate": 2.515334864711786e-06,
+      "loss": 0.1874,
+      "step": 3093
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8234790002004878,
+      "learning_rate": 2.5122653977316758e-06,
+      "loss": 0.1975,
+      "step": 3094
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8302823157904956,
+      "learning_rate": 2.5091971762346505e-06,
+      "loss": 0.2287,
+      "step": 3095
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9307487739704947,
+      "learning_rate": 2.5061302017568153e-06,
+      "loss": 0.2055,
+      "step": 3096
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8896993330109665,
+      "learning_rate": 2.503064475833646e-06,
+      "loss": 0.2151,
+      "step": 3097
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8537153330584717,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.21,
+      "step": 3098
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8276474260815257,
+      "learning_rate": 2.4969367757901054e-06,
+      "loss": 0.2166,
+      "step": 3099
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8453157404978698,
+      "learning_rate": 2.4938748047375637e-06,
+      "loss": 0.2113,
+      "step": 3100
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8157112428911031,
+      "learning_rate": 2.4908140883753468e-06,
+      "loss": 0.2053,
+      "step": 3101
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9355270535635507,
+      "learning_rate": 2.487754628235805e-06,
+      "loss": 0.2206,
+      "step": 3102
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.3070905066897267,
+      "learning_rate": 2.484696425850653e-06,
+      "loss": 0.1878,
+      "step": 3103
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8335250047472076,
+      "learning_rate": 2.481639482750979e-06,
+      "loss": 0.2099,
+      "step": 3104
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8345695263656018,
+      "learning_rate": 2.478583800467237e-06,
+      "loss": 0.2115,
+      "step": 3105
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.766005960266728,
+      "learning_rate": 2.4755293805292572e-06,
+      "loss": 0.2065,
+      "step": 3106
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8268356710766961,
+      "learning_rate": 2.4724762244662304e-06,
+      "loss": 0.221,
+      "step": 3107
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7767186203400164,
+      "learning_rate": 2.469424333806718e-06,
+      "loss": 0.2075,
+      "step": 3108
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8171669135876729,
+      "learning_rate": 2.466373710078651e-06,
+      "loss": 0.2092,
+      "step": 3109
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8116978818104398,
+      "learning_rate": 2.4633243548093195e-06,
+      "loss": 0.2134,
+      "step": 3110
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.341052783989684,
+      "learning_rate": 2.4602762695253825e-06,
+      "loss": 0.1998,
+      "step": 3111
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7629748367169008,
+      "learning_rate": 2.4572294557528604e-06,
+      "loss": 0.2054,
+      "step": 3112
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.9299464913728804,
+      "learning_rate": 2.454183915017142e-06,
+      "loss": 0.2286,
+      "step": 3113
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.853995240807254,
+      "learning_rate": 2.4511396488429727e-06,
+      "loss": 0.2318,
+      "step": 3114
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7675365360854232,
+      "learning_rate": 2.448096658754467e-06,
+      "loss": 0.2294,
+      "step": 3115
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8302556337572986,
+      "learning_rate": 2.4450549462750917e-06,
+      "loss": 0.2203,
+      "step": 3116
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8040162734121414,
+      "learning_rate": 2.442014512927682e-06,
+      "loss": 0.2157,
+      "step": 3117
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8612529516619589,
+      "learning_rate": 2.4389753602344298e-06,
+      "loss": 0.1996,
+      "step": 3118
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8443916044165317,
+      "learning_rate": 2.4359374897168823e-06,
+      "loss": 0.2194,
+      "step": 3119
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8027338326456102,
+      "learning_rate": 2.4329009028959484e-06,
+      "loss": 0.1964,
+      "step": 3120
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8452950352452875,
+      "learning_rate": 2.4298656012918963e-06,
+      "loss": 0.2294,
+      "step": 3121
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7975561157942317,
+      "learning_rate": 2.4268315864243447e-06,
+      "loss": 0.2165,
+      "step": 3122
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.473127800752957,
+      "learning_rate": 2.423798859812275e-06,
+      "loss": 0.225,
+      "step": 3123
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.755436599337083,
+      "learning_rate": 2.420767422974018e-06,
+      "loss": 0.199,
+      "step": 3124
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8599584570163421,
+      "learning_rate": 2.4177372774272627e-06,
+      "loss": 0.2286,
+      "step": 3125
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7820229220383547,
+      "learning_rate": 2.4147084246890478e-06,
+      "loss": 0.2243,
+      "step": 3126
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8515330047521719,
+      "learning_rate": 2.411680866275772e-06,
+      "loss": 0.2038,
+      "step": 3127
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.793089989130816,
+      "learning_rate": 2.4086546037031734e-06,
+      "loss": 0.2035,
+      "step": 3128
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7549325263039716,
+      "learning_rate": 2.405629638486354e-06,
+      "loss": 0.2098,
+      "step": 3129
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.3609509669673343,
+      "learning_rate": 2.402605972139759e-06,
+      "loss": 0.2022,
+      "step": 3130
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8363603353135879,
+      "learning_rate": 2.399583606177185e-06,
+      "loss": 0.1993,
+      "step": 3131
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8126452941602579,
+      "learning_rate": 2.396562542111781e-06,
+      "loss": 0.191,
+      "step": 3132
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9531097709594196,
+      "learning_rate": 2.393542781456038e-06,
+      "loss": 0.2028,
+      "step": 3133
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8622446180181914,
+      "learning_rate": 2.3905243257218007e-06,
+      "loss": 0.2154,
+      "step": 3134
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7772479938189563,
+      "learning_rate": 2.387507176420256e-06,
+      "loss": 0.2045,
+      "step": 3135
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7893923086408783,
+      "learning_rate": 2.3844913350619382e-06,
+      "loss": 0.204,
+      "step": 3136
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.1930448729042564,
+      "learning_rate": 2.381476803156724e-06,
+      "loss": 0.1957,
+      "step": 3137
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7670545327433946,
+      "learning_rate": 2.3784635822138424e-06,
+      "loss": 0.2199,
+      "step": 3138
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.859640794661714,
+      "learning_rate": 2.3754516737418563e-06,
+      "loss": 0.2038,
+      "step": 3139
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.718809183646963,
+      "learning_rate": 2.37244107924868e-06,
+      "loss": 0.1925,
+      "step": 3140
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8547119015733966,
+      "learning_rate": 2.3694318002415627e-06,
+      "loss": 0.2255,
+      "step": 3141
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9709120699646466,
+      "learning_rate": 2.366423838227102e-06,
+      "loss": 0.2108,
+      "step": 3142
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8711537558411829,
+      "learning_rate": 2.3634171947112307e-06,
+      "loss": 0.2268,
+      "step": 3143
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7530653638077148,
+      "learning_rate": 2.3604118711992237e-06,
+      "loss": 0.2005,
+      "step": 3144
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8222497532675002,
+      "learning_rate": 2.357407869195692e-06,
+      "loss": 0.21,
+      "step": 3145
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.104867511007057,
+      "learning_rate": 2.354405190204592e-06,
+      "loss": 0.1965,
+      "step": 3146
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7699230271847658,
+      "learning_rate": 2.3514038357292103e-06,
+      "loss": 0.1989,
+      "step": 3147
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7807702556129498,
+      "learning_rate": 2.348403807272176e-06,
+      "loss": 0.2082,
+      "step": 3148
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8054041647108304,
+      "learning_rate": 2.345405106335449e-06,
+      "loss": 0.2067,
+      "step": 3149
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8930884292767063,
+      "learning_rate": 2.342407734420331e-06,
+      "loss": 0.2102,
+      "step": 3150
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7902547236493344,
+      "learning_rate": 2.3394116930274523e-06,
+      "loss": 0.2149,
+      "step": 3151
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8180759268240664,
+      "learning_rate": 2.3364169836567805e-06,
+      "loss": 0.2081,
+      "step": 3152
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.972223660128982,
+      "learning_rate": 2.3334236078076126e-06,
+      "loss": 0.2165,
+      "step": 3153
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9825062847363517,
+      "learning_rate": 2.3304315669785853e-06,
+      "loss": 0.1945,
+      "step": 3154
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7632916875896303,
+      "learning_rate": 2.327440862667658e-06,
+      "loss": 0.2125,
+      "step": 3155
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7744491155870843,
+      "learning_rate": 2.3244514963721276e-06,
+      "loss": 0.2014,
+      "step": 3156
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7738192037490644,
+      "learning_rate": 2.32146346958862e-06,
+      "loss": 0.2105,
+      "step": 3157
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7759405070639642,
+      "learning_rate": 2.318476783813088e-06,
+      "loss": 0.2156,
+      "step": 3158
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1275161860606406,
+      "learning_rate": 2.315491440540815e-06,
+      "loss": 0.2027,
+      "step": 3159
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7912412501464616,
+      "learning_rate": 2.3125074412664085e-06,
+      "loss": 0.2269,
+      "step": 3160
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7915995234936646,
+      "learning_rate": 2.30952478748381e-06,
+      "loss": 0.2114,
+      "step": 3161
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.77094446337307,
+      "learning_rate": 2.306543480686281e-06,
+      "loss": 0.2156,
+      "step": 3162
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8057062478019477,
+      "learning_rate": 2.3035635223664136e-06,
+      "loss": 0.2275,
+      "step": 3163
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9159963956339656,
+      "learning_rate": 2.30058491401612e-06,
+      "loss": 0.2083,
+      "step": 3164
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8030899154938572,
+      "learning_rate": 2.2976076571266424e-06,
+      "loss": 0.1904,
+      "step": 3165
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8837240482056102,
+      "learning_rate": 2.294631753188541e-06,
+      "loss": 0.2292,
+      "step": 3166
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.879157415727393,
+      "learning_rate": 2.2916572036917e-06,
+      "loss": 0.2041,
+      "step": 3167
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0123822421480149,
+      "learning_rate": 2.288684010125325e-06,
+      "loss": 0.2074,
+      "step": 3168
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.825002360865364,
+      "learning_rate": 2.2857121739779474e-06,
+      "loss": 0.1921,
+      "step": 3169
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.859904124279098,
+      "learning_rate": 2.2827416967374115e-06,
+      "loss": 0.2062,
+      "step": 3170
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.774260735722781,
+      "learning_rate": 2.2797725798908897e-06,
+      "loss": 0.2099,
+      "step": 3171
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7630195281985123,
+      "learning_rate": 2.2768048249248648e-06,
+      "loss": 0.1948,
+      "step": 3172
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8011586360391141,
+      "learning_rate": 2.2738384333251447e-06,
+      "loss": 0.2025,
+      "step": 3173
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8095184301123792,
+      "learning_rate": 2.270873406576849e-06,
+      "loss": 0.205,
+      "step": 3174
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8631819962991627,
+      "learning_rate": 2.2679097461644204e-06,
+      "loss": 0.2021,
+      "step": 3175
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8479907752641959,
+      "learning_rate": 2.2649474535716128e-06,
+      "loss": 0.2043,
+      "step": 3176
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8324186019183961,
+      "learning_rate": 2.2619865302814955e-06,
+      "loss": 0.2161,
+      "step": 3177
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7620298793814823,
+      "learning_rate": 2.2590269777764516e-06,
+      "loss": 0.1925,
+      "step": 3178
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.9517231645111064,
+      "learning_rate": 2.256068797538182e-06,
+      "loss": 0.2157,
+      "step": 3179
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7804147156057747,
+      "learning_rate": 2.2531119910476994e-06,
+      "loss": 0.2261,
+      "step": 3180
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7520850781811256,
+      "learning_rate": 2.250156559785324e-06,
+      "loss": 0.1957,
+      "step": 3181
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7754201549208866,
+      "learning_rate": 2.247202505230695e-06,
+      "loss": 0.1941,
+      "step": 3182
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.0023277333747453,
+      "learning_rate": 2.2442498288627555e-06,
+      "loss": 0.2105,
+      "step": 3183
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8440743701427125,
+      "learning_rate": 2.241298532159763e-06,
+      "loss": 0.2182,
+      "step": 3184
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8386840681995981,
+      "learning_rate": 2.2383486165992796e-06,
+      "loss": 0.1875,
+      "step": 3185
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8097244439628282,
+      "learning_rate": 2.2354000836581834e-06,
+      "loss": 0.1884,
+      "step": 3186
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.0560824166286376,
+      "learning_rate": 2.2324529348126523e-06,
+      "loss": 0.2041,
+      "step": 3187
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.802708095131495,
+      "learning_rate": 2.229507171538178e-06,
+      "loss": 0.2112,
+      "step": 3188
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7881178064454469,
+      "learning_rate": 2.226562795309552e-06,
+      "loss": 0.1899,
+      "step": 3189
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8742831488964093,
+      "learning_rate": 2.223619807600878e-06,
+      "loss": 0.2215,
+      "step": 3190
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8744353068321438,
+      "learning_rate": 2.2206782098855605e-06,
+      "loss": 0.2068,
+      "step": 3191
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.9829306125936487,
+      "learning_rate": 2.217738003636309e-06,
+      "loss": 0.2147,
+      "step": 3192
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8376179196091236,
+      "learning_rate": 2.214799190325133e-06,
+      "loss": 0.2011,
+      "step": 3193
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7265655786830119,
+      "learning_rate": 2.2118617714233524e-06,
+      "loss": 0.1853,
+      "step": 3194
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8641612284386482,
+      "learning_rate": 2.208925748401582e-06,
+      "loss": 0.2135,
+      "step": 3195
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7772275112714924,
+      "learning_rate": 2.205991122729742e-06,
+      "loss": 0.2216,
+      "step": 3196
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8505635048842062,
+      "learning_rate": 2.2030578958770484e-06,
+      "loss": 0.2223,
+      "step": 3197
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8836007056693609,
+      "learning_rate": 2.2001260693120236e-06,
+      "loss": 0.2177,
+      "step": 3198
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9012512892578552,
+      "learning_rate": 2.1971956445024826e-06,
+      "loss": 0.2166,
+      "step": 3199
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8882704921770721,
+      "learning_rate": 2.1942666229155424e-06,
+      "loss": 0.2077,
+      "step": 3200
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8621184634870268,
+      "learning_rate": 2.1913390060176135e-06,
+      "loss": 0.2181,
+      "step": 3201
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.844022792572928,
+      "learning_rate": 2.1884127952744095e-06,
+      "loss": 0.2164,
+      "step": 3202
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8011996331629333,
+      "learning_rate": 2.185487992150933e-06,
+      "loss": 0.2128,
+      "step": 3203
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0521005220717294,
+      "learning_rate": 2.1825645981114862e-06,
+      "loss": 0.2097,
+      "step": 3204
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7734502744448922,
+      "learning_rate": 2.179642614619668e-06,
+      "loss": 0.1927,
+      "step": 3205
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7576598462680747,
+      "learning_rate": 2.1767220431383652e-06,
+      "loss": 0.2007,
+      "step": 3206
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.953407410890585,
+      "learning_rate": 2.173802885129761e-06,
+      "loss": 0.197,
+      "step": 3207
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8468257406906902,
+      "learning_rate": 2.1708851420553277e-06,
+      "loss": 0.2209,
+      "step": 3208
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8121120122660982,
+      "learning_rate": 2.1679688153758373e-06,
+      "loss": 0.2048,
+      "step": 3209
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8275387173301326,
+      "learning_rate": 2.1650539065513415e-06,
+      "loss": 0.2196,
+      "step": 3210
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8222028664597474,
+      "learning_rate": 2.162140417041193e-06,
+      "loss": 0.1913,
+      "step": 3211
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8344272684755182,
+      "learning_rate": 2.159228348304025e-06,
+      "loss": 0.2106,
+      "step": 3212
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8060344952566492,
+      "learning_rate": 2.156317701797766e-06,
+      "loss": 0.1975,
+      "step": 3213
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8276789528294942,
+      "learning_rate": 2.153408478979628e-06,
+      "loss": 0.2017,
+      "step": 3214
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9029012509764688,
+      "learning_rate": 2.150500681306113e-06,
+      "loss": 0.1971,
+      "step": 3215
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8711250450428643,
+      "learning_rate": 2.147594310233009e-06,
+      "loss": 0.2067,
+      "step": 3216
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8501994795964988,
+      "learning_rate": 2.144689367215388e-06,
+      "loss": 0.2055,
+      "step": 3217
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7772234664759253,
+      "learning_rate": 2.141785853707607e-06,
+      "loss": 0.2075,
+      "step": 3218
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8199309714142324,
+      "learning_rate": 2.1388837711633114e-06,
+      "loss": 0.1952,
+      "step": 3219
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8380341496025484,
+      "learning_rate": 2.1359831210354247e-06,
+      "loss": 0.2059,
+      "step": 3220
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8149823609835521,
+      "learning_rate": 2.133083904776159e-06,
+      "loss": 0.2241,
+      "step": 3221
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8077749533894638,
+      "learning_rate": 2.130186123837002e-06,
+      "loss": 0.2017,
+      "step": 3222
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7974798783301,
+      "learning_rate": 2.12728977966873e-06,
+      "loss": 0.2106,
+      "step": 3223
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9195202283547274,
+      "learning_rate": 2.124394873721394e-06,
+      "loss": 0.2195,
+      "step": 3224
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9221546344931357,
+      "learning_rate": 2.121501407444328e-06,
+      "loss": 0.1885,
+      "step": 3225
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8069130904876847,
+      "learning_rate": 2.118609382286142e-06,
+      "loss": 0.2047,
+      "step": 3226
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8159607610700541,
+      "learning_rate": 2.1157187996947293e-06,
+      "loss": 0.2088,
+      "step": 3227
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7934163682912723,
+      "learning_rate": 2.1128296611172593e-06,
+      "loss": 0.2151,
+      "step": 3228
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8615410582796292,
+      "learning_rate": 2.1099419680001754e-06,
+      "loss": 0.2148,
+      "step": 3229
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9947380084639743,
+      "learning_rate": 2.107055721789203e-06,
+      "loss": 0.2182,
+      "step": 3230
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7682896570091206,
+      "learning_rate": 2.104170923929338e-06,
+      "loss": 0.197,
+      "step": 3231
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.867638403583962,
+      "learning_rate": 2.101287575864853e-06,
+      "loss": 0.222,
+      "step": 3232
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7541241109485453,
+      "learning_rate": 2.0984056790392926e-06,
+      "loss": 0.2076,
+      "step": 3233
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.836860607996358,
+      "learning_rate": 2.0955252348954806e-06,
+      "loss": 0.2009,
+      "step": 3234
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8369419550450745,
+      "learning_rate": 2.092646244875507e-06,
+      "loss": 0.2316,
+      "step": 3235
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7975029954078579,
+      "learning_rate": 2.08976871042074e-06,
+      "loss": 0.2135,
+      "step": 3236
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8850077460085356,
+      "learning_rate": 2.086892632971812e-06,
+      "loss": 0.2066,
+      "step": 3237
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9250288937137455,
+      "learning_rate": 2.0840180139686333e-06,
+      "loss": 0.19,
+      "step": 3238
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8106983510023307,
+      "learning_rate": 2.0811448548503793e-06,
+      "loss": 0.2155,
+      "step": 3239
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7968210837336025,
+      "learning_rate": 2.0782731570554946e-06,
+      "loss": 0.2118,
+      "step": 3240
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.814239327913518,
+      "learning_rate": 2.0754029220216924e-06,
+      "loss": 0.2178,
+      "step": 3241
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8056965556317693,
+      "learning_rate": 2.0725341511859576e-06,
+      "loss": 0.2115,
+      "step": 3242
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8444592527639645,
+      "learning_rate": 2.0696668459845354e-06,
+      "loss": 0.2112,
+      "step": 3243
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7679445699607277,
+      "learning_rate": 2.066801007852944e-06,
+      "loss": 0.2244,
+      "step": 3244
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9589122284312126,
+      "learning_rate": 2.0639366382259613e-06,
+      "loss": 0.2069,
+      "step": 3245
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9854480183659478,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.2166,
+      "step": 3246
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7781911956731242,
+      "learning_rate": 2.0582123102212726e-06,
+      "loss": 0.1944,
+      "step": 3247
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7388083304807147,
+      "learning_rate": 2.0553523547094473e-06,
+      "loss": 0.2033,
+      "step": 3248
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7903279905242337,
+      "learning_rate": 2.052493873433993e-06,
+      "loss": 0.2316,
+      "step": 3249
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8349326117720296,
+      "learning_rate": 2.0496368678260094e-06,
+      "loss": 0.1926,
+      "step": 3250
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8399620873714326,
+      "learning_rate": 2.0467813393158516e-06,
+      "loss": 0.2169,
+      "step": 3251
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8026526597112721,
+      "learning_rate": 2.043927289333141e-06,
+      "loss": 0.2186,
+      "step": 3252
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7853269901213448,
+      "learning_rate": 2.041074719306757e-06,
+      "loss": 0.2136,
+      "step": 3253
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8653838243790698,
+      "learning_rate": 2.038223630664835e-06,
+      "loss": 0.2095,
+      "step": 3254
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9304915267859243,
+      "learning_rate": 2.0353740248347758e-06,
+      "loss": 0.2075,
+      "step": 3255
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8282259678275214,
+      "learning_rate": 2.0325259032432264e-06,
+      "loss": 0.2082,
+      "step": 3256
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7989331933480582,
+      "learning_rate": 2.0296792673161026e-06,
+      "loss": 0.197,
+      "step": 3257
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8196633559139862,
+      "learning_rate": 2.0268341184785674e-06,
+      "loss": 0.2104,
+      "step": 3258
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8020231646043837,
+      "learning_rate": 2.0239904581550463e-06,
+      "loss": 0.1995,
+      "step": 3259
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8614708879025961,
+      "learning_rate": 2.021148287769214e-06,
+      "loss": 0.2158,
+      "step": 3260
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.825506865008838,
+      "learning_rate": 2.0183076087440044e-06,
+      "loss": 0.2057,
+      "step": 3261
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9512589594614995,
+      "learning_rate": 2.015468422501599e-06,
+      "loss": 0.2072,
+      "step": 3262
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8510289317296952,
+      "learning_rate": 2.0126307304634383e-06,
+      "loss": 0.2039,
+      "step": 3263
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8567531171332677,
+      "learning_rate": 2.00979453405021e-06,
+      "loss": 0.1914,
+      "step": 3264
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8517184920310976,
+      "learning_rate": 2.0069598346818537e-06,
+      "loss": 0.2087,
+      "step": 3265
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8518801317835938,
+      "learning_rate": 2.004126633777559e-06,
+      "loss": 0.237,
+      "step": 3266
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7785193828843381,
+      "learning_rate": 2.0012949327557694e-06,
+      "loss": 0.1865,
+      "step": 3267
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8954189673714956,
+      "learning_rate": 1.998464733034172e-06,
+      "loss": 0.1877,
+      "step": 3268
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9432242086327545,
+      "learning_rate": 1.995636036029708e-06,
+      "loss": 0.2197,
+      "step": 3269
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8814047066092197,
+      "learning_rate": 1.992808843158559e-06,
+      "loss": 0.2151,
+      "step": 3270
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7968111208742598,
+      "learning_rate": 1.989983155836162e-06,
+      "loss": 0.2029,
+      "step": 3271
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7861394610246052,
+      "learning_rate": 1.987158975477194e-06,
+      "loss": 0.2156,
+      "step": 3272
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7250623716509548,
+      "learning_rate": 1.98433630349558e-06,
+      "loss": 0.1838,
+      "step": 3273
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8117945655269141,
+      "learning_rate": 1.9815151413044865e-06,
+      "loss": 0.2206,
+      "step": 3274
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8456378605754116,
+      "learning_rate": 1.9786954903163293e-06,
+      "loss": 0.2007,
+      "step": 3275
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8327891596828579,
+      "learning_rate": 1.9758773519427664e-06,
+      "loss": 0.2301,
+      "step": 3276
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8418499386432918,
+      "learning_rate": 1.9730607275946952e-06,
+      "loss": 0.2131,
+      "step": 3277
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.0054847034571774,
+      "learning_rate": 1.9702456186822595e-06,
+      "loss": 0.2122,
+      "step": 3278
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8478331371009393,
+      "learning_rate": 1.9674320266148404e-06,
+      "loss": 0.1957,
+      "step": 3279
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8566512300202195,
+      "learning_rate": 1.9646199528010625e-06,
+      "loss": 0.1838,
+      "step": 3280
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8565426326594301,
+      "learning_rate": 1.961809398648787e-06,
+      "loss": 0.2134,
+      "step": 3281
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8490900363469699,
+      "learning_rate": 1.959000365565119e-06,
+      "loss": 0.2236,
+      "step": 3282
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.800349517552001,
+      "learning_rate": 1.956192854956397e-06,
+      "loss": 0.2132,
+      "step": 3283
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7442932101010588,
+      "learning_rate": 1.9533868682282026e-06,
+      "loss": 0.1934,
+      "step": 3284
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8743136583462924,
+      "learning_rate": 1.9505824067853485e-06,
+      "loss": 0.2027,
+      "step": 3285
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7940996870536857,
+      "learning_rate": 1.947779472031891e-06,
+      "loss": 0.2101,
+      "step": 3286
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8034710017061045,
+      "learning_rate": 1.944978065371115e-06,
+      "loss": 0.2092,
+      "step": 3287
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9079000286925298,
+      "learning_rate": 1.9421781882055447e-06,
+      "loss": 0.2265,
+      "step": 3288
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8274484367761239,
+      "learning_rate": 1.9393798419369345e-06,
+      "loss": 0.2087,
+      "step": 3289
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9286772067451234,
+      "learning_rate": 1.936583027966279e-06,
+      "loss": 0.2095,
+      "step": 3290
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.868383103082128,
+      "learning_rate": 1.9337877476937976e-06,
+      "loss": 0.1812,
+      "step": 3291
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7731429013784945,
+      "learning_rate": 1.93099400251895e-06,
+      "loss": 0.2144,
+      "step": 3292
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7632374316295278,
+      "learning_rate": 1.9282017938404202e-06,
+      "loss": 0.1909,
+      "step": 3293
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7963238674116132,
+      "learning_rate": 1.9254111230561282e-06,
+      "loss": 0.199,
+      "step": 3294
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.82013274946942,
+      "learning_rate": 1.922621991563222e-06,
+      "loss": 0.2022,
+      "step": 3295
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.760476038426535,
+      "learning_rate": 1.9198344007580776e-06,
+      "loss": 0.202,
+      "step": 3296
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8638881345573433,
+      "learning_rate": 1.9170483520362997e-06,
+      "loss": 0.2124,
+      "step": 3297
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7685851447482087,
+      "learning_rate": 1.9142638467927254e-06,
+      "loss": 0.1985,
+      "step": 3298
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7391543251675341,
+      "learning_rate": 1.911480886421413e-06,
+      "loss": 0.198,
+      "step": 3299
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7888019814873322,
+      "learning_rate": 1.908699472315651e-06,
+      "loss": 0.2169,
+      "step": 3300
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8156410925312664,
+      "learning_rate": 1.9059196058679558e-06,
+      "loss": 0.2187,
+      "step": 3301
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7904501431121618,
+      "learning_rate": 1.9031412884700612e-06,
+      "loss": 0.2011,
+      "step": 3302
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7488330766055353,
+      "learning_rate": 1.9003645215129356e-06,
+      "loss": 0.2141,
+      "step": 3303
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7885355192235719,
+      "learning_rate": 1.8975893063867635e-06,
+      "loss": 0.1825,
+      "step": 3304
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3690519403155148,
+      "learning_rate": 1.8948156444809545e-06,
+      "loss": 0.1928,
+      "step": 3305
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7618246127299082,
+      "learning_rate": 1.8920435371841394e-06,
+      "loss": 0.2091,
+      "step": 3306
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7954547227984143,
+      "learning_rate": 1.8892729858841768e-06,
+      "loss": 0.2012,
+      "step": 3307
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7686160189430921,
+      "learning_rate": 1.8865039919681377e-06,
+      "loss": 0.2104,
+      "step": 3308
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7589853799932874,
+      "learning_rate": 1.8837365568223204e-06,
+      "loss": 0.1993,
+      "step": 3309
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8412486914235118,
+      "learning_rate": 1.8809706818322375e-06,
+      "loss": 0.1843,
+      "step": 3310
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7801189119686156,
+      "learning_rate": 1.8782063683826256e-06,
+      "loss": 0.2159,
+      "step": 3311
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7758518570190216,
+      "learning_rate": 1.8754436178574358e-06,
+      "loss": 0.2019,
+      "step": 3312
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.740043516945997,
+      "learning_rate": 1.8726824316398372e-06,
+      "loss": 0.1884,
+      "step": 3313
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.801561530820336,
+      "learning_rate": 1.8699228111122146e-06,
+      "loss": 0.2082,
+      "step": 3314
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8004339012964724,
+      "learning_rate": 1.8671647576561747e-06,
+      "loss": 0.1905,
+      "step": 3315
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8897923895862614,
+      "learning_rate": 1.8644082726525313e-06,
+      "loss": 0.1872,
+      "step": 3316
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.71434147433435,
+      "learning_rate": 1.8616533574813216e-06,
+      "loss": 0.1879,
+      "step": 3317
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8238140495829853,
+      "learning_rate": 1.8589000135217882e-06,
+      "loss": 0.1978,
+      "step": 3318
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7759961617039973,
+      "learning_rate": 1.8561482421523952e-06,
+      "loss": 0.2001,
+      "step": 3319
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8831098883299466,
+      "learning_rate": 1.8533980447508138e-06,
+      "loss": 0.2212,
+      "step": 3320
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8188259302532211,
+      "learning_rate": 1.8506494226939292e-06,
+      "loss": 0.2066,
+      "step": 3321
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.841167169312486,
+      "learning_rate": 1.8479023773578354e-06,
+      "loss": 0.2189,
+      "step": 3322
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8050964310368286,
+      "learning_rate": 1.845156910117843e-06,
+      "loss": 0.209,
+      "step": 3323
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7603322253340515,
+      "learning_rate": 1.8424130223484655e-06,
+      "loss": 0.1989,
+      "step": 3324
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8740542710488863,
+      "learning_rate": 1.8396707154234306e-06,
+      "loss": 0.2209,
+      "step": 3325
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8253660078379524,
+      "learning_rate": 1.8369299907156745e-06,
+      "loss": 0.2047,
+      "step": 3326
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8165013650626817,
+      "learning_rate": 1.8341908495973381e-06,
+      "loss": 0.2097,
+      "step": 3327
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8196176079684382,
+      "learning_rate": 1.831453293439771e-06,
+      "loss": 0.2109,
+      "step": 3328
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.812778640581865,
+      "learning_rate": 1.8287173236135276e-06,
+      "loss": 0.1933,
+      "step": 3329
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7779559182206821,
+      "learning_rate": 1.8259829414883728e-06,
+      "loss": 0.2091,
+      "step": 3330
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7752225084191672,
+      "learning_rate": 1.8232501484332709e-06,
+      "loss": 0.1922,
+      "step": 3331
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8112264861966072,
+      "learning_rate": 1.8205189458163964e-06,
+      "loss": 0.2128,
+      "step": 3332
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8046695056006833,
+      "learning_rate": 1.8177893350051213e-06,
+      "loss": 0.2104,
+      "step": 3333
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7666341239383426,
+      "learning_rate": 1.815061317366027e-06,
+      "loss": 0.1882,
+      "step": 3334
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.174902929142859,
+      "learning_rate": 1.8123348942648928e-06,
+      "loss": 0.1948,
+      "step": 3335
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7852615666700088,
+      "learning_rate": 1.809610067066701e-06,
+      "loss": 0.1973,
+      "step": 3336
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8411662878127395,
+      "learning_rate": 1.8068868371356335e-06,
+      "loss": 0.214,
+      "step": 3337
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.9211516860800817,
+      "learning_rate": 1.8041652058350768e-06,
+      "loss": 0.2154,
+      "step": 3338
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8506791860544992,
+      "learning_rate": 1.8014451745276107e-06,
+      "loss": 0.2122,
+      "step": 3339
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8011736871216636,
+      "learning_rate": 1.7987267445750224e-06,
+      "loss": 0.2133,
+      "step": 3340
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.803351772330926,
+      "learning_rate": 1.7960099173382873e-06,
+      "loss": 0.2027,
+      "step": 3341
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.779652060813132,
+      "learning_rate": 1.7932946941775882e-06,
+      "loss": 0.2201,
+      "step": 3342
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7926453327542258,
+      "learning_rate": 1.7905810764522963e-06,
+      "loss": 0.2202,
+      "step": 3343
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8261028360811151,
+      "learning_rate": 1.7878690655209862e-06,
+      "loss": 0.2274,
+      "step": 3344
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0141678427395608,
+      "learning_rate": 1.7851586627414237e-06,
+      "loss": 0.2088,
+      "step": 3345
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8089185736439358,
+      "learning_rate": 1.78244986947057e-06,
+      "loss": 0.2207,
+      "step": 3346
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.362747420277286,
+      "learning_rate": 1.7797426870645802e-06,
+      "loss": 0.2038,
+      "step": 3347
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8134397029299315,
+      "learning_rate": 1.7770371168788042e-06,
+      "loss": 0.2275,
+      "step": 3348
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8333071168288685,
+      "learning_rate": 1.7743331602677877e-06,
+      "loss": 0.2004,
+      "step": 3349
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8704726996693642,
+      "learning_rate": 1.7716308185852605e-06,
+      "loss": 0.227,
+      "step": 3350
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8024968155235331,
+      "learning_rate": 1.7689300931841525e-06,
+      "loss": 0.2061,
+      "step": 3351
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8124007740714414,
+      "learning_rate": 1.7662309854165788e-06,
+      "loss": 0.1926,
+      "step": 3352
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7791513631448548,
+      "learning_rate": 1.7635334966338463e-06,
+      "loss": 0.2092,
+      "step": 3353
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7972694475869957,
+      "learning_rate": 1.7608376281864502e-06,
+      "loss": 0.1986,
+      "step": 3354
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8364852305218088,
+      "learning_rate": 1.7581433814240783e-06,
+      "loss": 0.214,
+      "step": 3355
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8026403472656206,
+      "learning_rate": 1.755450757695601e-06,
+      "loss": 0.195,
+      "step": 3356
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7693819784773925,
+      "learning_rate": 1.7527597583490825e-06,
+      "loss": 0.19,
+      "step": 3357
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8736758908578457,
+      "learning_rate": 1.7500703847317663e-06,
+      "loss": 0.2057,
+      "step": 3358
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7533961922047684,
+      "learning_rate": 1.74738263819009e-06,
+      "loss": 0.1847,
+      "step": 3359
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.1543978344404755,
+      "learning_rate": 1.7446965200696702e-06,
+      "loss": 0.209,
+      "step": 3360
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.805265020204263,
+      "learning_rate": 1.742012031715311e-06,
+      "loss": 0.218,
+      "step": 3361
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7544891081451538,
+      "learning_rate": 1.739329174470999e-06,
+      "loss": 0.2007,
+      "step": 3362
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9689417981718482,
+      "learning_rate": 1.7366479496799076e-06,
+      "loss": 0.1896,
+      "step": 3363
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.785512385590833,
+      "learning_rate": 1.733968358684388e-06,
+      "loss": 0.2088,
+      "step": 3364
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7767264044023039,
+      "learning_rate": 1.7312904028259797e-06,
+      "loss": 0.199,
+      "step": 3365
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8329400830718305,
+      "learning_rate": 1.7286140834453958e-06,
+      "loss": 0.1912,
+      "step": 3366
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7180479998208326,
+      "learning_rate": 1.7259394018825382e-06,
+      "loss": 0.1825,
+      "step": 3367
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.053902937598688,
+      "learning_rate": 1.723266359476483e-06,
+      "loss": 0.2124,
+      "step": 3368
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8656524070478164,
+      "learning_rate": 1.7205949575654874e-06,
+      "loss": 0.2017,
+      "step": 3369
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.873908889851164,
+      "learning_rate": 1.7179251974869858e-06,
+      "loss": 0.1968,
+      "step": 3370
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9796584928277199,
+      "learning_rate": 1.7152570805775965e-06,
+      "loss": 0.1972,
+      "step": 3371
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8276948851524922,
+      "learning_rate": 1.7125906081731064e-06,
+      "loss": 0.1955,
+      "step": 3372
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8864942787130538,
+      "learning_rate": 1.7099257816084851e-06,
+      "loss": 0.2171,
+      "step": 3373
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8243581280974943,
+      "learning_rate": 1.7072626022178796e-06,
+      "loss": 0.2057,
+      "step": 3374
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9664438044882248,
+      "learning_rate": 1.7046010713346067e-06,
+      "loss": 0.2124,
+      "step": 3375
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.1661656916847312,
+      "learning_rate": 1.7019411902911609e-06,
+      "loss": 0.2179,
+      "step": 3376
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8193602069493761,
+      "learning_rate": 1.6992829604192078e-06,
+      "loss": 0.2037,
+      "step": 3377
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8362432705715591,
+      "learning_rate": 1.6966263830495939e-06,
+      "loss": 0.2064,
+      "step": 3378
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8178472203355794,
+      "learning_rate": 1.6939714595123286e-06,
+      "loss": 0.2172,
+      "step": 3379
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0525938965120512,
+      "learning_rate": 1.6913181911366017e-06,
+      "loss": 0.2321,
+      "step": 3380
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8431063032259721,
+      "learning_rate": 1.6886665792507673e-06,
+      "loss": 0.1893,
+      "step": 3381
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8628036901156144,
+      "learning_rate": 1.6860166251823574e-06,
+      "loss": 0.2289,
+      "step": 3382
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8010464052525165,
+      "learning_rate": 1.6833683302580661e-06,
+      "loss": 0.2064,
+      "step": 3383
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0186888300275314,
+      "learning_rate": 1.6807216958037664e-06,
+      "loss": 0.2072,
+      "step": 3384
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.946623120418213,
+      "learning_rate": 1.6780767231444878e-06,
+      "loss": 0.22,
+      "step": 3385
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.802255166272075,
+      "learning_rate": 1.6754334136044398e-06,
+      "loss": 0.2047,
+      "step": 3386
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7937388098489746,
+      "learning_rate": 1.67279176850699e-06,
+      "loss": 0.207,
+      "step": 3387
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8075185435907749,
+      "learning_rate": 1.6701517891746805e-06,
+      "loss": 0.2048,
+      "step": 3388
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8286460883101916,
+      "learning_rate": 1.6675134769292118e-06,
+      "loss": 0.2043,
+      "step": 3389
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8212664479637964,
+      "learning_rate": 1.6648768330914578e-06,
+      "loss": 0.2037,
+      "step": 3390
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0137714605439248,
+      "learning_rate": 1.6622418589814488e-06,
+      "loss": 0.1938,
+      "step": 3391
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8001847198882011,
+      "learning_rate": 1.6596085559183872e-06,
+      "loss": 0.2022,
+      "step": 3392
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8531521133802972,
+      "learning_rate": 1.656976925220633e-06,
+      "loss": 0.2274,
+      "step": 3393
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8965691966248163,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.2172,
+      "step": 3394
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.5420719678828736,
+      "learning_rate": 1.651718686190305e-06,
+      "loss": 0.1928,
+      "step": 3395
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8481175122072203,
+      "learning_rate": 1.649092080490266e-06,
+      "loss": 0.1925,
+      "step": 3396
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8818475509793297,
+      "learning_rate": 1.6464671524206045e-06,
+      "loss": 0.2147,
+      "step": 3397
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8579994612105364,
+      "learning_rate": 1.6438439032954857e-06,
+      "loss": 0.2132,
+      "step": 3398
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.817764262567123,
+      "learning_rate": 1.6412223344282408e-06,
+      "loss": 0.2136,
+      "step": 3399
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7724067931548769,
+      "learning_rate": 1.6386024471313561e-06,
+      "loss": 0.1882,
+      "step": 3400
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7951043137394653,
+      "learning_rate": 1.635984242716477e-06,
+      "loss": 0.1957,
+      "step": 3401
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9096016735797896,
+      "learning_rate": 1.633367722494404e-06,
+      "loss": 0.222,
+      "step": 3402
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7851445671442756,
+      "learning_rate": 1.6307528877751e-06,
+      "loss": 0.1935,
+      "step": 3403
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7981824484464611,
+      "learning_rate": 1.6281397398676774e-06,
+      "loss": 0.2087,
+      "step": 3404
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8257145772280945,
+      "learning_rate": 1.6255282800804106e-06,
+      "loss": 0.2009,
+      "step": 3405
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8358313127586263,
+      "learning_rate": 1.6229185097207228e-06,
+      "loss": 0.1953,
+      "step": 3406
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8372511588315872,
+      "learning_rate": 1.6203104300951977e-06,
+      "loss": 0.22,
+      "step": 3407
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7297601020779889,
+      "learning_rate": 1.6177040425095664e-06,
+      "loss": 0.1795,
+      "step": 3408
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8187690930906967,
+      "learning_rate": 1.615099348268716e-06,
+      "loss": 0.1997,
+      "step": 3409
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8349366819741094,
+      "learning_rate": 1.6124963486766844e-06,
+      "loss": 0.2172,
+      "step": 3410
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9526039541824979,
+      "learning_rate": 1.6098950450366646e-06,
+      "loss": 0.2088,
+      "step": 3411
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7643186143682315,
+      "learning_rate": 1.6072954386509948e-06,
+      "loss": 0.191,
+      "step": 3412
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7763849277525393,
+      "learning_rate": 1.6046975308211699e-06,
+      "loss": 0.2018,
+      "step": 3413
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8069045942159537,
+      "learning_rate": 1.6021013228478277e-06,
+      "loss": 0.2027,
+      "step": 3414
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7670864174871269,
+      "learning_rate": 1.5995068160307626e-06,
+      "loss": 0.218,
+      "step": 3415
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.120353584364422,
+      "learning_rate": 1.5969140116689102e-06,
+      "loss": 0.2199,
+      "step": 3416
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9782191501580447,
+      "learning_rate": 1.594322911060358e-06,
+      "loss": 0.212,
+      "step": 3417
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.3002434663086055,
+      "learning_rate": 1.5917335155023368e-06,
+      "loss": 0.2103,
+      "step": 3418
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8665752578629694,
+      "learning_rate": 1.589145826291229e-06,
+      "loss": 0.2043,
+      "step": 3419
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7619838322979566,
+      "learning_rate": 1.5865598447225576e-06,
+      "loss": 0.1896,
+      "step": 3420
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8165689636900055,
+      "learning_rate": 1.583975572090995e-06,
+      "loss": 0.1867,
+      "step": 3421
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8027240532598897,
+      "learning_rate": 1.5813930096903562e-06,
+      "loss": 0.2051,
+      "step": 3422
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8307299189926738,
+      "learning_rate": 1.5788121588135975e-06,
+      "loss": 0.2087,
+      "step": 3423
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.319113463636292,
+      "learning_rate": 1.5762330207528249e-06,
+      "loss": 0.1983,
+      "step": 3424
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8269848516559437,
+      "learning_rate": 1.5736555967992767e-06,
+      "loss": 0.2,
+      "step": 3425
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8372428671282304,
+      "learning_rate": 1.5710798882433432e-06,
+      "loss": 0.2345,
+      "step": 3426
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.780368819211805,
+      "learning_rate": 1.5685058963745485e-06,
+      "loss": 0.1866,
+      "step": 3427
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8276023998014325,
+      "learning_rate": 1.5659336224815642e-06,
+      "loss": 0.2352,
+      "step": 3428
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8401235796527401,
+      "learning_rate": 1.5633630678521944e-06,
+      "loss": 0.215,
+      "step": 3429
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7577518044717185,
+      "learning_rate": 1.5607942337733884e-06,
+      "loss": 0.2118,
+      "step": 3430
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.867300192640808,
+      "learning_rate": 1.5582271215312294e-06,
+      "loss": 0.2053,
+      "step": 3431
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.924391865569222,
+      "learning_rate": 1.555661732410944e-06,
+      "loss": 0.21,
+      "step": 3432
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8232797542390393,
+      "learning_rate": 1.553098067696891e-06,
+      "loss": 0.22,
+      "step": 3433
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.820105135930157,
+      "learning_rate": 1.5505361286725679e-06,
+      "loss": 0.1906,
+      "step": 3434
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7724534243795683,
+      "learning_rate": 1.5479759166206066e-06,
+      "loss": 0.1967,
+      "step": 3435
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7899343682083337,
+      "learning_rate": 1.5454174328227794e-06,
+      "loss": 0.2161,
+      "step": 3436
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.750236768498842,
+      "learning_rate": 1.542860678559987e-06,
+      "loss": 0.2094,
+      "step": 3437
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7885457603957196,
+      "learning_rate": 1.5403056551122697e-06,
+      "loss": 0.191,
+      "step": 3438
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8372802919883708,
+      "learning_rate": 1.537752363758796e-06,
+      "loss": 0.2077,
+      "step": 3439
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.9534512890056626,
+      "learning_rate": 1.5352008057778723e-06,
+      "loss": 0.2087,
+      "step": 3440
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8124154375622823,
+      "learning_rate": 1.5326509824469344e-06,
+      "loss": 0.2016,
+      "step": 3441
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8520239843138975,
+      "learning_rate": 1.5301028950425494e-06,
+      "loss": 0.2102,
+      "step": 3442
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7597685491712003,
+      "learning_rate": 1.5275565448404146e-06,
+      "loss": 0.1979,
+      "step": 3443
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8435121280984034,
+      "learning_rate": 1.5250119331153606e-06,
+      "loss": 0.2223,
+      "step": 3444
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8447686909157063,
+      "learning_rate": 1.5224690611413473e-06,
+      "loss": 0.2205,
+      "step": 3445
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8370952533705471,
+      "learning_rate": 1.5199279301914592e-06,
+      "loss": 0.2071,
+      "step": 3446
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7729393004010132,
+      "learning_rate": 1.5173885415379158e-06,
+      "loss": 0.2115,
+      "step": 3447
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8233846334752233,
+      "learning_rate": 1.5148508964520586e-06,
+      "loss": 0.212,
+      "step": 3448
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9322331717838734,
+      "learning_rate": 1.5123149962043586e-06,
+      "loss": 0.2422,
+      "step": 3449
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1456779617542083,
+      "learning_rate": 1.5097808420644117e-06,
+      "loss": 0.2081,
+      "step": 3450
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.0217732844256677,
+      "learning_rate": 1.5072484353009432e-06,
+      "loss": 0.2054,
+      "step": 3451
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7900926142164549,
+      "learning_rate": 1.5047177771817989e-06,
+      "loss": 0.1986,
+      "step": 3452
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.0201822972789896,
+      "learning_rate": 1.502188868973955e-06,
+      "loss": 0.2016,
+      "step": 3453
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8210230800808721,
+      "learning_rate": 1.4996617119435036e-06,
+      "loss": 0.2117,
+      "step": 3454
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7857713342546472,
+      "learning_rate": 1.4971363073556695e-06,
+      "loss": 0.1999,
+      "step": 3455
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7822815615128043,
+      "learning_rate": 1.494612656474792e-06,
+      "loss": 0.1989,
+      "step": 3456
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7614857534425913,
+      "learning_rate": 1.492090760564337e-06,
+      "loss": 0.2057,
+      "step": 3457
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8272760234423757,
+      "learning_rate": 1.4895706208868876e-06,
+      "loss": 0.2083,
+      "step": 3458
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7913364960540578,
+      "learning_rate": 1.4870522387041547e-06,
+      "loss": 0.2055,
+      "step": 3459
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8893914656269921,
+      "learning_rate": 1.4845356152769618e-06,
+      "loss": 0.2222,
+      "step": 3460
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7485807539081458,
+      "learning_rate": 1.4820207518652579e-06,
+      "loss": 0.1911,
+      "step": 3461
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8326643128891993,
+      "learning_rate": 1.4795076497281052e-06,
+      "loss": 0.2091,
+      "step": 3462
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8422258448531256,
+      "learning_rate": 1.4769963101236894e-06,
+      "loss": 0.1944,
+      "step": 3463
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8105778141313266,
+      "learning_rate": 1.4744867343093112e-06,
+      "loss": 0.2059,
+      "step": 3464
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8749747878894896,
+      "learning_rate": 1.4719789235413878e-06,
+      "loss": 0.2055,
+      "step": 3465
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9705698152581488,
+      "learning_rate": 1.4694728790754514e-06,
+      "loss": 0.2171,
+      "step": 3466
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8476178302170252,
+      "learning_rate": 1.466968602166156e-06,
+      "loss": 0.206,
+      "step": 3467
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8224026257897947,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.2019,
+      "step": 3468
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8412430378210753,
+      "learning_rate": 1.4619653560316526e-06,
+      "loss": 0.2049,
+      "step": 3469
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.774068204102223,
+      "learning_rate": 1.459466389311321e-06,
+      "loss": 0.1992,
+      "step": 3470
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7796751460408429,
+      "learning_rate": 1.4569691951573706e-06,
+      "loss": 0.1929,
+      "step": 3471
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8219871575304707,
+      "learning_rate": 1.4544737748200238e-06,
+      "loss": 0.1998,
+      "step": 3472
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7929977641684712,
+      "learning_rate": 1.4519801295486102e-06,
+      "loss": 0.1819,
+      "step": 3473
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.77939373841552,
+      "learning_rate": 1.4494882605915717e-06,
+      "loss": 0.2048,
+      "step": 3474
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8027365047716699,
+      "learning_rate": 1.4469981691964597e-06,
+      "loss": 0.1988,
+      "step": 3475
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8187522788588909,
+      "learning_rate": 1.4445098566099409e-06,
+      "loss": 0.1976,
+      "step": 3476
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8123397983341452,
+      "learning_rate": 1.4420233240777837e-06,
+      "loss": 0.2167,
+      "step": 3477
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.0042076698559055,
+      "learning_rate": 1.439538572844873e-06,
+      "loss": 0.2009,
+      "step": 3478
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7736945835605942,
+      "learning_rate": 1.4370556041551954e-06,
+      "loss": 0.2017,
+      "step": 3479
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9992813336286926,
+      "learning_rate": 1.4345744192518507e-06,
+      "loss": 0.2229,
+      "step": 3480
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7926589877235521,
+      "learning_rate": 1.432095019377041e-06,
+      "loss": 0.1995,
+      "step": 3481
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8815110213413043,
+      "learning_rate": 1.4296174057720775e-06,
+      "loss": 0.2046,
+      "step": 3482
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8605052112184932,
+      "learning_rate": 1.4271415796773742e-06,
+      "loss": 0.2046,
+      "step": 3483
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8566376865758489,
+      "learning_rate": 1.4246675423324546e-06,
+      "loss": 0.204,
+      "step": 3484
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.825729509274473,
+      "learning_rate": 1.422195294975942e-06,
+      "loss": 0.2148,
+      "step": 3485
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8854436748369858,
+      "learning_rate": 1.4197248388455693e-06,
+      "loss": 0.2096,
+      "step": 3486
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8814268928124271,
+      "learning_rate": 1.417256175178165e-06,
+      "loss": 0.2095,
+      "step": 3487
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7757642378841095,
+      "learning_rate": 1.4147893052096684e-06,
+      "loss": 0.194,
+      "step": 3488
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8352620872254193,
+      "learning_rate": 1.412324230175114e-06,
+      "loss": 0.2074,
+      "step": 3489
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7878847876836989,
+      "learning_rate": 1.4098609513086414e-06,
+      "loss": 0.1815,
+      "step": 3490
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8256932912279322,
+      "learning_rate": 1.4073994698434878e-06,
+      "loss": 0.2178,
+      "step": 3491
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9682088262729036,
+      "learning_rate": 1.4049397870119962e-06,
+      "loss": 0.2189,
+      "step": 3492
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.795974439007975,
+      "learning_rate": 1.4024819040456023e-06,
+      "loss": 0.208,
+      "step": 3493
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9614707626719363,
+      "learning_rate": 1.4000258221748452e-06,
+      "loss": 0.2038,
+      "step": 3494
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8186898432804607,
+      "learning_rate": 1.3975715426293636e-06,
+      "loss": 0.2048,
+      "step": 3495
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8081595548542846,
+      "learning_rate": 1.3951190666378884e-06,
+      "loss": 0.1981,
+      "step": 3496
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9010710984067137,
+      "learning_rate": 1.392668395428251e-06,
+      "loss": 0.2167,
+      "step": 3497
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7995729180175998,
+      "learning_rate": 1.390219530227378e-06,
+      "loss": 0.2064,
+      "step": 3498
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9840464278083798,
+      "learning_rate": 1.3877724722612946e-06,
+      "loss": 0.2078,
+      "step": 3499
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8357745782925916,
+      "learning_rate": 1.3853272227551167e-06,
+      "loss": 0.218,
+      "step": 3500
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8047796537819546,
+      "learning_rate": 1.38288378293306e-06,
+      "loss": 0.2077,
+      "step": 3501
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.755031115231481,
+      "learning_rate": 1.3804421540184282e-06,
+      "loss": 0.2028,
+      "step": 3502
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.0971158277072703,
+      "learning_rate": 1.378002337233625e-06,
+      "loss": 0.2138,
+      "step": 3503
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7807061504298903,
+      "learning_rate": 1.3755643338001411e-06,
+      "loss": 0.2023,
+      "step": 3504
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8656902601933734,
+      "learning_rate": 1.373128144938563e-06,
+      "loss": 0.1888,
+      "step": 3505
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.031521010344397,
+      "learning_rate": 1.3706937718685654e-06,
+      "loss": 0.1941,
+      "step": 3506
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.899935939344936,
+      "learning_rate": 1.3682612158089193e-06,
+      "loss": 0.2111,
+      "step": 3507
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9221398938132912,
+      "learning_rate": 1.3658304779774784e-06,
+      "loss": 0.211,
+      "step": 3508
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9644054856485513,
+      "learning_rate": 1.3634015595911949e-06,
+      "loss": 0.2215,
+      "step": 3509
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.096033081917837,
+      "learning_rate": 1.3609744618661013e-06,
+      "loss": 0.2225,
+      "step": 3510
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8633323123407886,
+      "learning_rate": 1.3585491860173266e-06,
+      "loss": 0.2292,
+      "step": 3511
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8291248678482006,
+      "learning_rate": 1.3561257332590816e-06,
+      "loss": 0.2051,
+      "step": 3512
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9033381741044483,
+      "learning_rate": 1.3537041048046696e-06,
+      "loss": 0.2054,
+      "step": 3513
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9268819127963769,
+      "learning_rate": 1.3512843018664728e-06,
+      "loss": 0.2211,
+      "step": 3514
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8139214964308515,
+      "learning_rate": 1.3488663256559686e-06,
+      "loss": 0.2053,
+      "step": 3515
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.808047396968743,
+      "learning_rate": 1.3464501773837124e-06,
+      "loss": 0.2009,
+      "step": 3516
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8611374390206479,
+      "learning_rate": 1.3440358582593494e-06,
+      "loss": 0.2143,
+      "step": 3517
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8089010666437245,
+      "learning_rate": 1.3416233694916086e-06,
+      "loss": 0.1966,
+      "step": 3518
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8165067335926468,
+      "learning_rate": 1.339212712288298e-06,
+      "loss": 0.213,
+      "step": 3519
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7987161348304072,
+      "learning_rate": 1.3368038878563154e-06,
+      "loss": 0.1931,
+      "step": 3520
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8040124384608907,
+      "learning_rate": 1.3343968974016363e-06,
+      "loss": 0.226,
+      "step": 3521
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1333750446158743,
+      "learning_rate": 1.3319917421293182e-06,
+      "loss": 0.2117,
+      "step": 3522
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8737546243548677,
+      "learning_rate": 1.3295884232435008e-06,
+      "loss": 0.2018,
+      "step": 3523
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8061908344028794,
+      "learning_rate": 1.3271869419474059e-06,
+      "loss": 0.1903,
+      "step": 3524
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8528369438779891,
+      "learning_rate": 1.3247872994433326e-06,
+      "loss": 0.2043,
+      "step": 3525
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8305601562702487,
+      "learning_rate": 1.3223894969326623e-06,
+      "loss": 0.2074,
+      "step": 3526
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9684403819154285,
+      "learning_rate": 1.319993535615851e-06,
+      "loss": 0.217,
+      "step": 3527
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7968138022749343,
+      "learning_rate": 1.3175994166924394e-06,
+      "loss": 0.1979,
+      "step": 3528
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7690472819918537,
+      "learning_rate": 1.315207141361039e-06,
+      "loss": 0.1862,
+      "step": 3529
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9758582448587286,
+      "learning_rate": 1.312816710819343e-06,
+      "loss": 0.2193,
+      "step": 3530
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7689536706858033,
+      "learning_rate": 1.310428126264116e-06,
+      "loss": 0.1978,
+      "step": 3531
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8327994298199546,
+      "learning_rate": 1.3080413888912057e-06,
+      "loss": 0.1921,
+      "step": 3532
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.979708986631606,
+      "learning_rate": 1.3056564998955274e-06,
+      "loss": 0.217,
+      "step": 3533
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7803521370093142,
+      "learning_rate": 1.3032734604710784e-06,
+      "loss": 0.2067,
+      "step": 3534
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8607140755029349,
+      "learning_rate": 1.3008922718109224e-06,
+      "loss": 0.2019,
+      "step": 3535
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8596086779684551,
+      "learning_rate": 1.298512935107204e-06,
+      "loss": 0.2008,
+      "step": 3536
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1042397284503669,
+      "learning_rate": 1.2961354515511342e-06,
+      "loss": 0.2284,
+      "step": 3537
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7919105409015194,
+      "learning_rate": 1.2937598223330006e-06,
+      "loss": 0.2154,
+      "step": 3538
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7861956651995161,
+      "learning_rate": 1.2913860486421582e-06,
+      "loss": 0.1935,
+      "step": 3539
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7799427573035573,
+      "learning_rate": 1.289014131667039e-06,
+      "loss": 0.2066,
+      "step": 3540
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9681879590218875,
+      "learning_rate": 1.286644072595139e-06,
+      "loss": 0.2096,
+      "step": 3541
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.820376834221769,
+      "learning_rate": 1.2842758726130283e-06,
+      "loss": 0.205,
+      "step": 3542
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0332328833332276,
+      "learning_rate": 1.2819095329063469e-06,
+      "loss": 0.2032,
+      "step": 3543
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8269043782175944,
+      "learning_rate": 1.2795450546597999e-06,
+      "loss": 0.2222,
+      "step": 3544
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.267771836097729,
+      "learning_rate": 1.277182439057162e-06,
+      "loss": 0.1951,
+      "step": 3545
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.861476784953918,
+      "learning_rate": 1.2748216872812747e-06,
+      "loss": 0.2165,
+      "step": 3546
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8293675720220853,
+      "learning_rate": 1.2724628005140493e-06,
+      "loss": 0.2005,
+      "step": 3547
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8285243141940684,
+      "learning_rate": 1.2701057799364591e-06,
+      "loss": 0.2042,
+      "step": 3548
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7651915327180964,
+      "learning_rate": 1.2677506267285479e-06,
+      "loss": 0.1749,
+      "step": 3549
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9526645787512711,
+      "learning_rate": 1.2653973420694182e-06,
+      "loss": 0.1724,
+      "step": 3550
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0558723901488347,
+      "learning_rate": 1.2630459271372452e-06,
+      "loss": 0.195,
+      "step": 3551
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8024244788379525,
+      "learning_rate": 1.26069638310926e-06,
+      "loss": 0.1911,
+      "step": 3552
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7976452647214893,
+      "learning_rate": 1.2583487111617647e-06,
+      "loss": 0.2147,
+      "step": 3553
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9156139921380538,
+      "learning_rate": 1.256002912470115e-06,
+      "loss": 0.2031,
+      "step": 3554
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8078829343801378,
+      "learning_rate": 1.2536589882087373e-06,
+      "loss": 0.2008,
+      "step": 3555
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8339401246494644,
+      "learning_rate": 1.2513169395511143e-06,
+      "loss": 0.2,
+      "step": 3556
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7915477379397834,
+      "learning_rate": 1.2489767676697933e-06,
+      "loss": 0.2142,
+      "step": 3557
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8725135022111107,
+      "learning_rate": 1.246638473736378e-06,
+      "loss": 0.2265,
+      "step": 3558
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8137624088065074,
+      "learning_rate": 1.244302058921537e-06,
+      "loss": 0.1977,
+      "step": 3559
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1074754258104627,
+      "learning_rate": 1.2419675243949914e-06,
+      "loss": 0.2121,
+      "step": 3560
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7780668813160131,
+      "learning_rate": 1.239634871325529e-06,
+      "loss": 0.2118,
+      "step": 3561
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8007451123360613,
+      "learning_rate": 1.2373041008809888e-06,
+      "loss": 0.2146,
+      "step": 3562
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0409957842636757,
+      "learning_rate": 1.2349752142282706e-06,
+      "loss": 0.2116,
+      "step": 3563
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8171938168922067,
+      "learning_rate": 1.2326482125333284e-06,
+      "loss": 0.2144,
+      "step": 3564
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8646783439733442,
+      "learning_rate": 1.2303230969611762e-06,
+      "loss": 0.2135,
+      "step": 3565
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7973310555649603,
+      "learning_rate": 1.2279998686758833e-06,
+      "loss": 0.1903,
+      "step": 3566
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9753177957301877,
+      "learning_rate": 1.2256785288405699e-06,
+      "loss": 0.2106,
+      "step": 3567
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8066336038488723,
+      "learning_rate": 1.223359078617416e-06,
+      "loss": 0.1976,
+      "step": 3568
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8684321476007432,
+      "learning_rate": 1.2210415191676528e-06,
+      "loss": 0.2006,
+      "step": 3569
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8092425533463993,
+      "learning_rate": 1.2187258516515644e-06,
+      "loss": 0.1874,
+      "step": 3570
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9399340383182222,
+      "learning_rate": 1.2164120772284875e-06,
+      "loss": 0.1993,
+      "step": 3571
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8144659789422408,
+      "learning_rate": 1.214100197056815e-06,
+      "loss": 0.2031,
+      "step": 3572
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8385791682493909,
+      "learning_rate": 1.2117902122939861e-06,
+      "loss": 0.2132,
+      "step": 3573
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7936766903035203,
+      "learning_rate": 1.2094821240964955e-06,
+      "loss": 0.1941,
+      "step": 3574
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8095237760363129,
+      "learning_rate": 1.207175933619884e-06,
+      "loss": 0.1955,
+      "step": 3575
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7576197633819444,
+      "learning_rate": 1.2048716420187473e-06,
+      "loss": 0.192,
+      "step": 3576
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8027512718064951,
+      "learning_rate": 1.2025692504467269e-06,
+      "loss": 0.1962,
+      "step": 3577
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8391023717681856,
+      "learning_rate": 1.2002687600565138e-06,
+      "loss": 0.2106,
+      "step": 3578
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8435942512662753,
+      "learning_rate": 1.1979701719998454e-06,
+      "loss": 0.2066,
+      "step": 3579
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9428467327997415,
+      "learning_rate": 1.1956734874275127e-06,
+      "loss": 0.2278,
+      "step": 3580
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8575603331831663,
+      "learning_rate": 1.1933787074893456e-06,
+      "loss": 0.218,
+      "step": 3581
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9078280989759726,
+      "learning_rate": 1.1910858333342279e-06,
+      "loss": 0.205,
+      "step": 3582
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7891066954358616,
+      "learning_rate": 1.1887948661100833e-06,
+      "loss": 0.2042,
+      "step": 3583
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8156392151731457,
+      "learning_rate": 1.1865058069638857e-06,
+      "loss": 0.2134,
+      "step": 3584
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.124435952371972,
+      "learning_rate": 1.1842186570416508e-06,
+      "loss": 0.2048,
+      "step": 3585
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8776043979512734,
+      "learning_rate": 1.1819334174884378e-06,
+      "loss": 0.2238,
+      "step": 3586
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8384733616249512,
+      "learning_rate": 1.1796500894483509e-06,
+      "loss": 0.2021,
+      "step": 3587
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8723432205393217,
+      "learning_rate": 1.1773686740645384e-06,
+      "loss": 0.2047,
+      "step": 3588
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8092010548413108,
+      "learning_rate": 1.175089172479188e-06,
+      "loss": 0.205,
+      "step": 3589
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0911560188158096,
+      "learning_rate": 1.172811585833532e-06,
+      "loss": 0.2103,
+      "step": 3590
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.9641210030710429,
+      "learning_rate": 1.1705359152678447e-06,
+      "loss": 0.2057,
+      "step": 3591
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0067864100139765,
+      "learning_rate": 1.168262161921438e-06,
+      "loss": 0.2145,
+      "step": 3592
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8707594591788022,
+      "learning_rate": 1.165990326932665e-06,
+      "loss": 0.208,
+      "step": 3593
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7905319165167928,
+      "learning_rate": 1.1637204114389179e-06,
+      "loss": 0.208,
+      "step": 3594
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8815228821103789,
+      "learning_rate": 1.161452416576631e-06,
+      "loss": 0.204,
+      "step": 3595
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8228684281331501,
+      "learning_rate": 1.1591863434812722e-06,
+      "loss": 0.2213,
+      "step": 3596
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.9296572309299613,
+      "learning_rate": 1.1569221932873532e-06,
+      "loss": 0.2036,
+      "step": 3597
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8192671976092413,
+      "learning_rate": 1.1546599671284158e-06,
+      "loss": 0.2035,
+      "step": 3598
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7952126300788089,
+      "learning_rate": 1.1523996661370461e-06,
+      "loss": 0.2034,
+      "step": 3599
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8059362887195344,
+      "learning_rate": 1.1501412914448596e-06,
+      "loss": 0.1653,
+      "step": 3600
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6968565935301788,
+      "learning_rate": 1.147884844182513e-06,
+      "loss": 0.1809,
+      "step": 3601
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8115473082145499,
+      "learning_rate": 1.1456303254796941e-06,
+      "loss": 0.1983,
+      "step": 3602
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8498106650073951,
+      "learning_rate": 1.1433777364651272e-06,
+      "loss": 0.2231,
+      "step": 3603
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.9055494362860586,
+      "learning_rate": 1.141127078266568e-06,
+      "loss": 0.2056,
+      "step": 3604
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7767402520626958,
+      "learning_rate": 1.1388783520108099e-06,
+      "loss": 0.1932,
+      "step": 3605
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8346745815385201,
+      "learning_rate": 1.1366315588236743e-06,
+      "loss": 0.1865,
+      "step": 3606
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8118276972446513,
+      "learning_rate": 1.1343866998300197e-06,
+      "loss": 0.198,
+      "step": 3607
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.796520180623774,
+      "learning_rate": 1.1321437761537307e-06,
+      "loss": 0.2,
+      "step": 3608
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8486086885753186,
+      "learning_rate": 1.1299027889177294e-06,
+      "loss": 0.2324,
+      "step": 3609
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8658922921943732,
+      "learning_rate": 1.127663739243962e-06,
+      "loss": 0.2057,
+      "step": 3610
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7508713642397388,
+      "learning_rate": 1.1254266282534093e-06,
+      "loss": 0.1983,
+      "step": 3611
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8097290895424888,
+      "learning_rate": 1.1231914570660774e-06,
+      "loss": 0.1914,
+      "step": 3612
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8104974394212128,
+      "learning_rate": 1.1209582268010056e-06,
+      "loss": 0.1947,
+      "step": 3613
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7679544665339595,
+      "learning_rate": 1.1187269385762605e-06,
+      "loss": 0.1993,
+      "step": 3614
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.859962307542535,
+      "learning_rate": 1.1164975935089328e-06,
+      "loss": 0.2163,
+      "step": 3615
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9903223737163112,
+      "learning_rate": 1.1142701927151456e-06,
+      "loss": 0.2067,
+      "step": 3616
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9844393456116584,
+      "learning_rate": 1.1120447373100446e-06,
+      "loss": 0.1916,
+      "step": 3617
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8027324597899479,
+      "learning_rate": 1.1098212284078037e-06,
+      "loss": 0.2067,
+      "step": 3618
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7820781512583674,
+      "learning_rate": 1.1075996671216188e-06,
+      "loss": 0.208,
+      "step": 3619
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8443056418451189,
+      "learning_rate": 1.1053800545637173e-06,
+      "loss": 0.2029,
+      "step": 3620
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9472883916953864,
+      "learning_rate": 1.1031623918453443e-06,
+      "loss": 0.2094,
+      "step": 3621
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8289555634803486,
+      "learning_rate": 1.1009466800767738e-06,
+      "loss": 0.2096,
+      "step": 3622
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8072473943438971,
+      "learning_rate": 1.098732920367298e-06,
+      "loss": 0.214,
+      "step": 3623
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9195456978114837,
+      "learning_rate": 1.0965211138252373e-06,
+      "loss": 0.2086,
+      "step": 3624
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8808086479544656,
+      "learning_rate": 1.094311261557931e-06,
+      "loss": 0.2119,
+      "step": 3625
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8194269970572332,
+      "learning_rate": 1.09210336467174e-06,
+      "loss": 0.2007,
+      "step": 3626
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7978026152273885,
+      "learning_rate": 1.089897424272045e-06,
+      "loss": 0.2132,
+      "step": 3627
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8488874415793412,
+      "learning_rate": 1.0876934414632523e-06,
+      "loss": 0.2029,
+      "step": 3628
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7999440837229275,
+      "learning_rate": 1.0854914173487818e-06,
+      "loss": 0.2153,
+      "step": 3629
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8286368507981562,
+      "learning_rate": 1.0832913530310784e-06,
+      "loss": 0.2035,
+      "step": 3630
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8847132309617382,
+      "learning_rate": 1.081093249611601e-06,
+      "loss": 0.2121,
+      "step": 3631
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8020459809026471,
+      "learning_rate": 1.0788971081908306e-06,
+      "loss": 0.1943,
+      "step": 3632
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8487115384443211,
+      "learning_rate": 1.0767029298682642e-06,
+      "loss": 0.2102,
+      "step": 3633
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8118672554593933,
+      "learning_rate": 1.0745107157424155e-06,
+      "loss": 0.223,
+      "step": 3634
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7769522094613674,
+      "learning_rate": 1.072320466910814e-06,
+      "loss": 0.1865,
+      "step": 3635
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9589782812870415,
+      "learning_rate": 1.0701321844700097e-06,
+      "loss": 0.1949,
+      "step": 3636
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7951027148316914,
+      "learning_rate": 1.0679458695155615e-06,
+      "loss": 0.1923,
+      "step": 3637
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8987626687742695,
+      "learning_rate": 1.0657615231420492e-06,
+      "loss": 0.2086,
+      "step": 3638
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8159729677007637,
+      "learning_rate": 1.0635791464430662e-06,
+      "loss": 0.2002,
+      "step": 3639
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.774333080650485,
+      "learning_rate": 1.0613987405112153e-06,
+      "loss": 0.1963,
+      "step": 3640
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8051028458497546,
+      "learning_rate": 1.0592203064381173e-06,
+      "loss": 0.2058,
+      "step": 3641
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8170841617742868,
+      "learning_rate": 1.0570438453144044e-06,
+      "loss": 0.2159,
+      "step": 3642
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.791844593124062,
+      "learning_rate": 1.0548693582297203e-06,
+      "loss": 0.2128,
+      "step": 3643
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.1118591544528837,
+      "learning_rate": 1.0526968462727189e-06,
+      "loss": 0.2084,
+      "step": 3644
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7832870169592823,
+      "learning_rate": 1.0505263105310703e-06,
+      "loss": 0.1927,
+      "step": 3645
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9831239512914707,
+      "learning_rate": 1.0483577520914495e-06,
+      "loss": 0.1942,
+      "step": 3646
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.090541707063517,
+      "learning_rate": 1.0461911720395473e-06,
+      "loss": 0.2081,
+      "step": 3647
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9446796158532035,
+      "learning_rate": 1.0440265714600573e-06,
+      "loss": 0.213,
+      "step": 3648
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7907447466661203,
+      "learning_rate": 1.041863951436689e-06,
+      "loss": 0.1913,
+      "step": 3649
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8661213692870352,
+      "learning_rate": 1.0397033130521554e-06,
+      "loss": 0.2009,
+      "step": 3650
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8078273264527825,
+      "learning_rate": 1.0375446573881799e-06,
+      "loss": 0.1918,
+      "step": 3651
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7716586982727934,
+      "learning_rate": 1.0353879855254895e-06,
+      "loss": 0.1928,
+      "step": 3652
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.798997258975493,
+      "learning_rate": 1.0332332985438248e-06,
+      "loss": 0.2182,
+      "step": 3653
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7731597351245862,
+      "learning_rate": 1.0310805975219256e-06,
+      "loss": 0.1945,
+      "step": 3654
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8000652231064437,
+      "learning_rate": 1.028929883537544e-06,
+      "loss": 0.1919,
+      "step": 3655
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7684390898514347,
+      "learning_rate": 1.0267811576674298e-06,
+      "loss": 0.1764,
+      "step": 3656
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8093237570186982,
+      "learning_rate": 1.0246344209873448e-06,
+      "loss": 0.1972,
+      "step": 3657
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.2153880093051543,
+      "learning_rate": 1.0224896745720513e-06,
+      "loss": 0.2273,
+      "step": 3658
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9037509384169398,
+      "learning_rate": 1.0203469194953142e-06,
+      "loss": 0.2145,
+      "step": 3659
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8030371191153859,
+      "learning_rate": 1.0182061568299019e-06,
+      "loss": 0.1894,
+      "step": 3660
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8135204711994154,
+      "learning_rate": 1.0160673876475892e-06,
+      "loss": 0.2035,
+      "step": 3661
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8677374019379261,
+      "learning_rate": 1.0139306130191463e-06,
+      "loss": 0.2021,
+      "step": 3662
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8576472856211185,
+      "learning_rate": 1.0117958340143508e-06,
+      "loss": 0.2013,
+      "step": 3663
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8784681491481449,
+      "learning_rate": 1.009663051701979e-06,
+      "loss": 0.2005,
+      "step": 3664
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.842071964716862,
+      "learning_rate": 1.0075322671498066e-06,
+      "loss": 0.2029,
+      "step": 3665
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.967074925887901,
+      "learning_rate": 1.0054034814246093e-06,
+      "loss": 0.2083,
+      "step": 3666
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7889482664317842,
+      "learning_rate": 1.0032766955921618e-06,
+      "loss": 0.1958,
+      "step": 3667
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8080021005537612,
+      "learning_rate": 1.0011519107172413e-06,
+      "loss": 0.2012,
+      "step": 3668
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9046681041574066,
+      "learning_rate": 9.990291278636172e-07,
+      "loss": 0.2163,
+      "step": 3669
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7943574367451625,
+      "learning_rate": 9.969083480940617e-07,
+      "loss": 0.2113,
+      "step": 3670
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8024997829179281,
+      "learning_rate": 9.947895724703405e-07,
+      "loss": 0.2182,
+      "step": 3671
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8500174633083788,
+      "learning_rate": 9.926728020532196e-07,
+      "loss": 0.2141,
+      "step": 3672
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8460750685691629,
+      "learning_rate": 9.905580379024581e-07,
+      "loss": 0.2128,
+      "step": 3673
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7709603458527973,
+      "learning_rate": 9.884452810768115e-07,
+      "loss": 0.1964,
+      "step": 3674
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8479935201406378,
+      "learning_rate": 9.86334532634029e-07,
+      "loss": 0.1914,
+      "step": 3675
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8013268090881425,
+      "learning_rate": 9.842257936308586e-07,
+      "loss": 0.2068,
+      "step": 3676
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8387738337434317,
+      "learning_rate": 9.82119065123036e-07,
+      "loss": 0.1901,
+      "step": 3677
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8290057138358154,
+      "learning_rate": 9.80014348165298e-07,
+      "loss": 0.2095,
+      "step": 3678
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.5112561746451716,
+      "learning_rate": 9.77911643811366e-07,
+      "loss": 0.2097,
+      "step": 3679
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8318329708085995,
+      "learning_rate": 9.758109531139614e-07,
+      "loss": 0.2081,
+      "step": 3680
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7902647869297437,
+      "learning_rate": 9.737122771247902e-07,
+      "loss": 0.218,
+      "step": 3681
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8043231827512183,
+      "learning_rate": 9.71615616894559e-07,
+      "loss": 0.1905,
+      "step": 3682
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8366333282917933,
+      "learning_rate": 9.695209734729533e-07,
+      "loss": 0.1998,
+      "step": 3683
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.1100784358132785,
+      "learning_rate": 9.6742834790866e-07,
+      "loss": 0.2087,
+      "step": 3684
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9680089947720858,
+      "learning_rate": 9.653377412493475e-07,
+      "loss": 0.2106,
+      "step": 3685
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9374502060880311,
+      "learning_rate": 9.632491545416788e-07,
+      "loss": 0.2212,
+      "step": 3686
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8672045307314835,
+      "learning_rate": 9.611625888313053e-07,
+      "loss": 0.2279,
+      "step": 3687
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9927088701551656,
+      "learning_rate": 9.590780451628617e-07,
+      "loss": 0.2044,
+      "step": 3688
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7665049501292929,
+      "learning_rate": 9.569955245799772e-07,
+      "loss": 0.1952,
+      "step": 3689
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8570924116932979,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.2198,
+      "step": 3690
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8213389965384976,
+      "learning_rate": 9.528365568403192e-07,
+      "loss": 0.204,
+      "step": 3691
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9541633539827719,
+      "learning_rate": 9.507601117657289e-07,
+      "loss": 0.2274,
+      "step": 3692
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8076097259671791,
+      "learning_rate": 9.486856939410672e-07,
+      "loss": 0.1872,
+      "step": 3693
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8206085192121794,
+      "learning_rate": 9.466133044048865e-07,
+      "loss": 0.202,
+      "step": 3694
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.793885108075915,
+      "learning_rate": 9.445429441947312e-07,
+      "loss": 0.1873,
+      "step": 3695
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8237562097322644,
+      "learning_rate": 9.424746143471231e-07,
+      "loss": 0.1955,
+      "step": 3696
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8566500331921099,
+      "learning_rate": 9.40408315897573e-07,
+      "loss": 0.2103,
+      "step": 3697
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8118488042599434,
+      "learning_rate": 9.383440498805712e-07,
+      "loss": 0.1901,
+      "step": 3698
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7825206273720617,
+      "learning_rate": 9.362818173295913e-07,
+      "loss": 0.1974,
+      "step": 3699
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8343968055992778,
+      "learning_rate": 9.34221619277087e-07,
+      "loss": 0.1887,
+      "step": 3700
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9309266745555499,
+      "learning_rate": 9.321634567544991e-07,
+      "loss": 0.2084,
+      "step": 3701
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.2162724916476546,
+      "learning_rate": 9.301073307922432e-07,
+      "loss": 0.2075,
+      "step": 3702
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8442287417639899,
+      "learning_rate": 9.280532424197192e-07,
+      "loss": 0.2075,
+      "step": 3703
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.762330935854017,
+      "learning_rate": 9.260011926653034e-07,
+      "loss": 0.202,
+      "step": 3704
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8839810669284089,
+      "learning_rate": 9.239511825563557e-07,
+      "loss": 0.2186,
+      "step": 3705
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8492312638956099,
+      "learning_rate": 9.219032131192124e-07,
+      "loss": 0.2227,
+      "step": 3706
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8995211074720286,
+      "learning_rate": 9.19857285379187e-07,
+      "loss": 0.2173,
+      "step": 3707
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0600843029334939,
+      "learning_rate": 9.178134003605721e-07,
+      "loss": 0.19,
+      "step": 3708
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8139780439335652,
+      "learning_rate": 9.157715590866396e-07,
+      "loss": 0.2047,
+      "step": 3709
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7615918643187999,
+      "learning_rate": 9.137317625796338e-07,
+      "loss": 0.1938,
+      "step": 3710
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.833307436003548,
+      "learning_rate": 9.116940118607792e-07,
+      "loss": 0.2055,
+      "step": 3711
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.047999031048492,
+      "learning_rate": 9.096583079502763e-07,
+      "loss": 0.2158,
+      "step": 3712
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0051136529935405,
+      "learning_rate": 9.076246518672971e-07,
+      "loss": 0.1972,
+      "step": 3713
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8562286955718115,
+      "learning_rate": 9.055930446299916e-07,
+      "loss": 0.2018,
+      "step": 3714
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8708516555491815,
+      "learning_rate": 9.035634872554805e-07,
+      "loss": 0.2095,
+      "step": 3715
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.782798154922052,
+      "learning_rate": 9.015359807598645e-07,
+      "loss": 0.2105,
+      "step": 3716
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.869057387737003,
+      "learning_rate": 8.995105261582099e-07,
+      "loss": 0.1976,
+      "step": 3717
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3717048311422353,
+      "learning_rate": 8.974871244645628e-07,
+      "loss": 0.2058,
+      "step": 3718
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7384399264571577,
+      "learning_rate": 8.954657766919361e-07,
+      "loss": 0.1909,
+      "step": 3719
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7954486886194144,
+      "learning_rate": 8.934464838523183e-07,
+      "loss": 0.2119,
+      "step": 3720
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7783189881384129,
+      "learning_rate": 8.91429246956666e-07,
+      "loss": 0.1843,
+      "step": 3721
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8485549939291552,
+      "learning_rate": 8.894140670149082e-07,
+      "loss": 0.2098,
+      "step": 3722
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8414693478287831,
+      "learning_rate": 8.874009450359428e-07,
+      "loss": 0.221,
+      "step": 3723
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7721776135119247,
+      "learning_rate": 8.853898820276402e-07,
+      "loss": 0.1884,
+      "step": 3724
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7854851820256735,
+      "learning_rate": 8.833808789968362e-07,
+      "loss": 0.2104,
+      "step": 3725
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8433500020177149,
+      "learning_rate": 8.813739369493395e-07,
+      "loss": 0.1901,
+      "step": 3726
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.889269386543488,
+      "learning_rate": 8.793690568899216e-07,
+      "loss": 0.2171,
+      "step": 3727
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7730538446106687,
+      "learning_rate": 8.773662398223276e-07,
+      "loss": 0.1964,
+      "step": 3728
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.2630239324641603,
+      "learning_rate": 8.753654867492645e-07,
+      "loss": 0.2053,
+      "step": 3729
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8063119148196102,
+      "learning_rate": 8.733667986724109e-07,
+      "loss": 0.209,
+      "step": 3730
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8742976655301959,
+      "learning_rate": 8.713701765924071e-07,
+      "loss": 0.2182,
+      "step": 3731
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7582762309401178,
+      "learning_rate": 8.693756215088617e-07,
+      "loss": 0.1876,
+      "step": 3732
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7877900679442391,
+      "learning_rate": 8.673831344203454e-07,
+      "loss": 0.1995,
+      "step": 3733
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.895868246935937,
+      "learning_rate": 8.653927163243986e-07,
+      "loss": 0.2024,
+      "step": 3734
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8554626814224279,
+      "learning_rate": 8.634043682175231e-07,
+      "loss": 0.2013,
+      "step": 3735
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8424386713144641,
+      "learning_rate": 8.61418091095183e-07,
+      "loss": 0.2028,
+      "step": 3736
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8825143301547952,
+      "learning_rate": 8.594338859518081e-07,
+      "loss": 0.2264,
+      "step": 3737
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7363240472006344,
+      "learning_rate": 8.574517537807897e-07,
+      "loss": 0.1909,
+      "step": 3738
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8058587454985411,
+      "learning_rate": 8.554716955744808e-07,
+      "loss": 0.2091,
+      "step": 3739
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.746650032737436,
+      "learning_rate": 8.534937123241949e-07,
+      "loss": 0.1977,
+      "step": 3740
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9322444847241745,
+      "learning_rate": 8.515178050202116e-07,
+      "loss": 0.1962,
+      "step": 3741
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7662245116014795,
+      "learning_rate": 8.495439746517648e-07,
+      "loss": 0.1936,
+      "step": 3742
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7992709191770518,
+      "learning_rate": 8.475722222070542e-07,
+      "loss": 0.2138,
+      "step": 3743
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7855362131323891,
+      "learning_rate": 8.45602548673235e-07,
+      "loss": 0.1863,
+      "step": 3744
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8198704326827374,
+      "learning_rate": 8.436349550364248e-07,
+      "loss": 0.2188,
+      "step": 3745
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8574845025581416,
+      "learning_rate": 8.416694422816974e-07,
+      "loss": 0.2152,
+      "step": 3746
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8734761601902286,
+      "learning_rate": 8.39706011393086e-07,
+      "loss": 0.2058,
+      "step": 3747
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8699411893182128,
+      "learning_rate": 8.377446633535797e-07,
+      "loss": 0.2305,
+      "step": 3748
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.872049380766428,
+      "learning_rate": 8.357853991451292e-07,
+      "loss": 0.1891,
+      "step": 3749
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9703365752449679,
+      "learning_rate": 8.338282197486364e-07,
+      "loss": 0.2031,
+      "step": 3750
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9597747728992018,
+      "learning_rate": 8.31873126143965e-07,
+      "loss": 0.2001,
+      "step": 3751
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7781429248836178,
+      "learning_rate": 8.299201193099287e-07,
+      "loss": 0.1892,
+      "step": 3752
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8910428465624551,
+      "learning_rate": 8.279692002243028e-07,
+      "loss": 0.2087,
+      "step": 3753
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7876573419615257,
+      "learning_rate": 8.260203698638114e-07,
+      "loss": 0.1902,
+      "step": 3754
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8652165438749898,
+      "learning_rate": 8.240736292041368e-07,
+      "loss": 0.2051,
+      "step": 3755
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7982767976167535,
+      "learning_rate": 8.221289792199117e-07,
+      "loss": 0.2035,
+      "step": 3756
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8064849658822881,
+      "learning_rate": 8.20186420884726e-07,
+      "loss": 0.1951,
+      "step": 3757
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8516380143723365,
+      "learning_rate": 8.182459551711197e-07,
+      "loss": 0.1961,
+      "step": 3758
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8170274694214515,
+      "learning_rate": 8.163075830505862e-07,
+      "loss": 0.2084,
+      "step": 3759
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7848119868482271,
+      "learning_rate": 8.143713054935715e-07,
+      "loss": 0.2089,
+      "step": 3760
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8681651173376379,
+      "learning_rate": 8.124371234694716e-07,
+      "loss": 0.2174,
+      "step": 3761
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.2276031862576908,
+      "learning_rate": 8.105050379466334e-07,
+      "loss": 0.2112,
+      "step": 3762
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.0397829049670677,
+      "learning_rate": 8.085750498923528e-07,
+      "loss": 0.207,
+      "step": 3763
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8869452684719509,
+      "learning_rate": 8.066471602728804e-07,
+      "loss": 0.2139,
+      "step": 3764
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7939137189076256,
+      "learning_rate": 8.047213700534113e-07,
+      "loss": 0.2083,
+      "step": 3765
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8146138760011529,
+      "learning_rate": 8.027976801980924e-07,
+      "loss": 0.2062,
+      "step": 3766
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9361726965691364,
+      "learning_rate": 8.008760916700164e-07,
+      "loss": 0.2098,
+      "step": 3767
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7954309858807205,
+      "learning_rate": 7.989566054312286e-07,
+      "loss": 0.1952,
+      "step": 3768
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.0748408476484792,
+      "learning_rate": 7.970392224427154e-07,
+      "loss": 0.2044,
+      "step": 3769
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8762758740417925,
+      "learning_rate": 7.95123943664417e-07,
+      "loss": 0.1922,
+      "step": 3770
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8529059164525692,
+      "learning_rate": 7.932107700552145e-07,
+      "loss": 0.2148,
+      "step": 3771
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7544800978025945,
+      "learning_rate": 7.91299702572938e-07,
+      "loss": 0.1855,
+      "step": 3772
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8082615312602527,
+      "learning_rate": 7.893907421743613e-07,
+      "loss": 0.1945,
+      "step": 3773
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.201293386337643,
+      "learning_rate": 7.87483889815207e-07,
+      "loss": 0.2062,
+      "step": 3774
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.871228167568139,
+      "learning_rate": 7.855791464501367e-07,
+      "loss": 0.2309,
+      "step": 3775
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.75176775993887,
+      "learning_rate": 7.836765130327622e-07,
+      "loss": 0.1962,
+      "step": 3776
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7713547123928347,
+      "learning_rate": 7.817759905156335e-07,
+      "loss": 0.1943,
+      "step": 3777
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8374360385486832,
+      "learning_rate": 7.798775798502484e-07,
+      "loss": 0.2014,
+      "step": 3778
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8515984466196056,
+      "learning_rate": 7.779812819870442e-07,
+      "loss": 0.2047,
+      "step": 3779
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.800703699472245,
+      "learning_rate": 7.760870978754015e-07,
+      "loss": 0.2138,
+      "step": 3780
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8483940049187502,
+      "learning_rate": 7.741950284636418e-07,
+      "loss": 0.2086,
+      "step": 3781
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.799667251002074,
+      "learning_rate": 7.723050746990291e-07,
+      "loss": 0.2249,
+      "step": 3782
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9946186357912877,
+      "learning_rate": 7.704172375277691e-07,
+      "loss": 0.2154,
+      "step": 3783
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1765673579911435,
+      "learning_rate": 7.685315178950048e-07,
+      "loss": 0.1985,
+      "step": 3784
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8123966902768448,
+      "learning_rate": 7.666479167448232e-07,
+      "loss": 0.1935,
+      "step": 3785
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7537669288259652,
+      "learning_rate": 7.647664350202461e-07,
+      "loss": 0.1867,
+      "step": 3786
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7647671372533134,
+      "learning_rate": 7.628870736632371e-07,
+      "loss": 0.1879,
+      "step": 3787
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9845487069266511,
+      "learning_rate": 7.610098336146965e-07,
+      "loss": 0.212,
+      "step": 3788
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8280720882960786,
+      "learning_rate": 7.591347158144657e-07,
+      "loss": 0.2035,
+      "step": 3789
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.025387279050998,
+      "learning_rate": 7.572617212013189e-07,
+      "loss": 0.2099,
+      "step": 3790
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7890041517552925,
+      "learning_rate": 7.553908507129726e-07,
+      "loss": 0.1974,
+      "step": 3791
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7933855564351378,
+      "learning_rate": 7.535221052860747e-07,
+      "loss": 0.2086,
+      "step": 3792
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9193059241132462,
+      "learning_rate": 7.516554858562142e-07,
+      "loss": 0.1936,
+      "step": 3793
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8348593049220868,
+      "learning_rate": 7.497909933579117e-07,
+      "loss": 0.2104,
+      "step": 3794
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8509112522957173,
+      "learning_rate": 7.479286287246251e-07,
+      "loss": 0.202,
+      "step": 3795
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8343529624343005,
+      "learning_rate": 7.460683928887441e-07,
+      "loss": 0.2115,
+      "step": 3796
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0403250421941344,
+      "learning_rate": 7.442102867815976e-07,
+      "loss": 0.2028,
+      "step": 3797
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9601938468505777,
+      "learning_rate": 7.423543113334436e-07,
+      "loss": 0.2112,
+      "step": 3798
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8081755499783799,
+      "learning_rate": 7.405004674734773e-07,
+      "loss": 0.1998,
+      "step": 3799
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7792484396325824,
+      "learning_rate": 7.386487561298223e-07,
+      "loss": 0.1965,
+      "step": 3800
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8558122552855584,
+      "learning_rate": 7.367991782295392e-07,
+      "loss": 0.2079,
+      "step": 3801
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8322477656948252,
+      "learning_rate": 7.349517346986173e-07,
+      "loss": 0.1982,
+      "step": 3802
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8339041509179176,
+      "learning_rate": 7.331064264619786e-07,
+      "loss": 0.1997,
+      "step": 3803
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7312271091474731,
+      "learning_rate": 7.312632544434739e-07,
+      "loss": 0.2077,
+      "step": 3804
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8224690466288591,
+      "learning_rate": 7.29422219565889e-07,
+      "loss": 0.1943,
+      "step": 3805
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8890720072653991,
+      "learning_rate": 7.275833227509344e-07,
+      "loss": 0.2153,
+      "step": 3806
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.831813817782938,
+      "learning_rate": 7.257465649192547e-07,
+      "loss": 0.1993,
+      "step": 3807
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9393004761756156,
+      "learning_rate": 7.239119469904227e-07,
+      "loss": 0.2173,
+      "step": 3808
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8883094743491976,
+      "learning_rate": 7.220794698829364e-07,
+      "loss": 0.2069,
+      "step": 3809
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8315462828219292,
+      "learning_rate": 7.202491345142288e-07,
+      "loss": 0.195,
+      "step": 3810
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.093515590443107,
+      "learning_rate": 7.184209418006505e-07,
+      "loss": 0.205,
+      "step": 3811
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8858876866917612,
+      "learning_rate": 7.165948926574895e-07,
+      "loss": 0.2019,
+      "step": 3812
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.821316519876693,
+      "learning_rate": 7.147709879989539e-07,
+      "loss": 0.1975,
+      "step": 3813
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8479866459243267,
+      "learning_rate": 7.129492287381828e-07,
+      "loss": 0.1937,
+      "step": 3814
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8326126919630332,
+      "learning_rate": 7.111296157872377e-07,
+      "loss": 0.2149,
+      "step": 3815
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8239638032038116,
+      "learning_rate": 7.093121500571082e-07,
+      "loss": 0.2007,
+      "step": 3816
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9343617154628604,
+      "learning_rate": 7.074968324577053e-07,
+      "loss": 0.2085,
+      "step": 3817
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8347499458324309,
+      "learning_rate": 7.056836638978698e-07,
+      "loss": 0.1894,
+      "step": 3818
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8427570883397526,
+      "learning_rate": 7.038726452853617e-07,
+      "loss": 0.217,
+      "step": 3819
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.758231515877324,
+      "learning_rate": 7.020637775268674e-07,
+      "loss": 0.2029,
+      "step": 3820
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8106161401967246,
+      "learning_rate": 7.002570615279942e-07,
+      "loss": 0.213,
+      "step": 3821
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8288152298146064,
+      "learning_rate": 6.984524981932756e-07,
+      "loss": 0.2212,
+      "step": 3822
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.072435086837615,
+      "learning_rate": 6.966500884261635e-07,
+      "loss": 0.1971,
+      "step": 3823
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8436488239457399,
+      "learning_rate": 6.948498331290349e-07,
+      "loss": 0.2056,
+      "step": 3824
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8424681071602617,
+      "learning_rate": 6.93051733203185e-07,
+      "loss": 0.2168,
+      "step": 3825
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.9425824029259076,
+      "learning_rate": 6.912557895488342e-07,
+      "loss": 0.2137,
+      "step": 3826
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8657745010361109,
+      "learning_rate": 6.894620030651194e-07,
+      "loss": 0.1935,
+      "step": 3827
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8085830086450677,
+      "learning_rate": 6.876703746500984e-07,
+      "loss": 0.1951,
+      "step": 3828
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7651929658290535,
+      "learning_rate": 6.858809052007487e-07,
+      "loss": 0.1909,
+      "step": 3829
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7938713005695339,
+      "learning_rate": 6.84093595612968e-07,
+      "loss": 0.1845,
+      "step": 3830
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7909073871605091,
+      "learning_rate": 6.823084467815732e-07,
+      "loss": 0.1976,
+      "step": 3831
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.0176713298418427,
+      "learning_rate": 6.805254596002959e-07,
+      "loss": 0.1934,
+      "step": 3832
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9091069117832286,
+      "learning_rate": 6.787446349617899e-07,
+      "loss": 0.2138,
+      "step": 3833
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7667174435069745,
+      "learning_rate": 6.769659737576229e-07,
+      "loss": 0.1698,
+      "step": 3834
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.869850620844387,
+      "learning_rate": 6.751894768782813e-07,
+      "loss": 0.1933,
+      "step": 3835
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8203528930431416,
+      "learning_rate": 6.73415145213166e-07,
+      "loss": 0.2009,
+      "step": 3836
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8883219838228861,
+      "learning_rate": 6.716429796505974e-07,
+      "loss": 0.1893,
+      "step": 3837
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8589403643561135,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.2133,
+      "step": 3838
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.834322096250237,
+      "learning_rate": 6.681051503809455e-07,
+      "loss": 0.1989,
+      "step": 3839
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8487891941733847,
+      "learning_rate": 6.663394884450752e-07,
+      "loss": 0.2011,
+      "step": 3840
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.782415106109103,
+      "learning_rate": 6.645759961541759e-07,
+      "loss": 0.1776,
+      "step": 3841
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8025368703584524,
+      "learning_rate": 6.628146743911374e-07,
+      "loss": 0.2083,
+      "step": 3842
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8746725348466036,
+      "learning_rate": 6.610555240377653e-07,
+      "loss": 0.1927,
+      "step": 3843
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8745749497724613,
+      "learning_rate": 6.592985459747753e-07,
+      "loss": 0.1977,
+      "step": 3844
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.824480544115484,
+      "learning_rate": 6.575437410818008e-07,
+      "loss": 0.2001,
+      "step": 3845
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8036888002584697,
+      "learning_rate": 6.55791110237381e-07,
+      "loss": 0.2015,
+      "step": 3846
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8601142637747644,
+      "learning_rate": 6.540406543189726e-07,
+      "loss": 0.1982,
+      "step": 3847
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7967538109239728,
+      "learning_rate": 6.522923742029374e-07,
+      "loss": 0.215,
+      "step": 3848
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.827540600902556,
+      "learning_rate": 6.505462707645538e-07,
+      "loss": 0.1836,
+      "step": 3849
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8056432015113425,
+      "learning_rate": 6.488023448780068e-07,
+      "loss": 0.2063,
+      "step": 3850
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9663753260447989,
+      "learning_rate": 6.470605974163913e-07,
+      "loss": 0.202,
+      "step": 3851
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.819820278821711,
+      "learning_rate": 6.453210292517114e-07,
+      "loss": 0.2045,
+      "step": 3852
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9415573780778222,
+      "learning_rate": 6.435836412548835e-07,
+      "loss": 0.2013,
+      "step": 3853
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9137740476074608,
+      "learning_rate": 6.418484342957277e-07,
+      "loss": 0.1961,
+      "step": 3854
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.858805326385652,
+      "learning_rate": 6.401154092429751e-07,
+      "loss": 0.2111,
+      "step": 3855
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8838214997386564,
+      "learning_rate": 6.383845669642657e-07,
+      "loss": 0.2142,
+      "step": 3856
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8503049688875259,
+      "learning_rate": 6.36655908326142e-07,
+      "loss": 0.2121,
+      "step": 3857
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6177377011446528,
+      "learning_rate": 6.349294341940593e-07,
+      "loss": 0.1943,
+      "step": 3858
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7346950479194562,
+      "learning_rate": 6.332051454323734e-07,
+      "loss": 0.1984,
+      "step": 3859
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8409538844298663,
+      "learning_rate": 6.3148304290435e-07,
+      "loss": 0.2179,
+      "step": 3860
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.871318753888756,
+      "learning_rate": 6.297631274721567e-07,
+      "loss": 0.2017,
+      "step": 3861
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7909115465670457,
+      "learning_rate": 6.280453999968711e-07,
+      "loss": 0.203,
+      "step": 3862
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9159764845961813,
+      "learning_rate": 6.263298613384705e-07,
+      "loss": 0.2137,
+      "step": 3863
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8229568254503754,
+      "learning_rate": 6.246165123558401e-07,
+      "loss": 0.2021,
+      "step": 3864
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8579591876010748,
+      "learning_rate": 6.229053539067659e-07,
+      "loss": 0.2078,
+      "step": 3865
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.1037543313647,
+      "learning_rate": 6.211963868479393e-07,
+      "loss": 0.23,
+      "step": 3866
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8310258921658458,
+      "learning_rate": 6.194896120349542e-07,
+      "loss": 0.2135,
+      "step": 3867
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.2623280199862326,
+      "learning_rate": 6.177850303223059e-07,
+      "loss": 0.2124,
+      "step": 3868
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9964418882413476,
+      "learning_rate": 6.160826425633914e-07,
+      "loss": 0.2282,
+      "step": 3869
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8549676593653617,
+      "learning_rate": 6.143824496105122e-07,
+      "loss": 0.192,
+      "step": 3870
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9229001317910464,
+      "learning_rate": 6.126844523148672e-07,
+      "loss": 0.2003,
+      "step": 3871
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8392703310226469,
+      "learning_rate": 6.109886515265595e-07,
+      "loss": 0.1819,
+      "step": 3872
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8340437624149339,
+      "learning_rate": 6.092950480945897e-07,
+      "loss": 0.2217,
+      "step": 3873
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8000268985636866,
+      "learning_rate": 6.076036428668613e-07,
+      "loss": 0.2069,
+      "step": 3874
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9022017399787988,
+      "learning_rate": 6.059144366901737e-07,
+      "loss": 0.2221,
+      "step": 3875
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4626915541223438,
+      "learning_rate": 6.042274304102281e-07,
+      "loss": 0.2282,
+      "step": 3876
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7693320221845886,
+      "learning_rate": 6.025426248716216e-07,
+      "loss": 0.2065,
+      "step": 3877
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.75357497018012,
+      "learning_rate": 6.008600209178539e-07,
+      "loss": 0.1917,
+      "step": 3878
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9755680478743025,
+      "learning_rate": 5.991796193913175e-07,
+      "loss": 0.2106,
+      "step": 3879
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8617131880701999,
+      "learning_rate": 5.975014211333053e-07,
+      "loss": 0.2101,
+      "step": 3880
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7990965923351903,
+      "learning_rate": 5.95825426984008e-07,
+      "loss": 0.2125,
+      "step": 3881
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8276272883213156,
+      "learning_rate": 5.941516377825102e-07,
+      "loss": 0.2108,
+      "step": 3882
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7815220450546825,
+      "learning_rate": 5.92480054366793e-07,
+      "loss": 0.204,
+      "step": 3883
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8107710212124583,
+      "learning_rate": 5.908106775737332e-07,
+      "loss": 0.2007,
+      "step": 3884
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8600664828735772,
+      "learning_rate": 5.891435082391056e-07,
+      "loss": 0.2199,
+      "step": 3885
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8234917423402174,
+      "learning_rate": 5.874785471975753e-07,
+      "loss": 0.2108,
+      "step": 3886
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8017790896838503,
+      "learning_rate": 5.858157952827064e-07,
+      "loss": 0.2026,
+      "step": 3887
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8127573471823886,
+      "learning_rate": 5.841552533269534e-07,
+      "loss": 0.1979,
+      "step": 3888
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9537588192959175,
+      "learning_rate": 5.824969221616666e-07,
+      "loss": 0.1841,
+      "step": 3889
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8480704861539714,
+      "learning_rate": 5.808408026170892e-07,
+      "loss": 0.1991,
+      "step": 3890
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8829074216083773,
+      "learning_rate": 5.791868955223551e-07,
+      "loss": 0.1956,
+      "step": 3891
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.903129924705151,
+      "learning_rate": 5.775352017054919e-07,
+      "loss": 0.2255,
+      "step": 3892
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0570016589658677,
+      "learning_rate": 5.75885721993421e-07,
+      "loss": 0.2031,
+      "step": 3893
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7897074337985064,
+      "learning_rate": 5.742384572119519e-07,
+      "loss": 0.2108,
+      "step": 3894
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7671673128159777,
+      "learning_rate": 5.725934081857887e-07,
+      "loss": 0.1819,
+      "step": 3895
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1560761708819736,
+      "learning_rate": 5.70950575738522e-07,
+      "loss": 0.193,
+      "step": 3896
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8266104100427756,
+      "learning_rate": 5.693099606926378e-07,
+      "loss": 0.2138,
+      "step": 3897
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9400626160647357,
+      "learning_rate": 5.676715638695063e-07,
+      "loss": 0.2072,
+      "step": 3898
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1423973566585084,
+      "learning_rate": 5.66035386089393e-07,
+      "loss": 0.1927,
+      "step": 3899
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.82746304361677,
+      "learning_rate": 5.644014281714483e-07,
+      "loss": 0.2253,
+      "step": 3900
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8169668299216283,
+      "learning_rate": 5.627696909337116e-07,
+      "loss": 0.199,
+      "step": 3901
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9023605214661183,
+      "learning_rate": 5.611401751931112e-07,
+      "loss": 0.191,
+      "step": 3902
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7986897526512989,
+      "learning_rate": 5.595128817654638e-07,
+      "loss": 0.2078,
+      "step": 3903
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8040005815015872,
+      "learning_rate": 5.578878114654746e-07,
+      "loss": 0.2039,
+      "step": 3904
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.858360557943728,
+      "learning_rate": 5.56264965106732e-07,
+      "loss": 0.2068,
+      "step": 3905
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8706865083336387,
+      "learning_rate": 5.546443435017146e-07,
+      "loss": 0.2168,
+      "step": 3906
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7717822179710967,
+      "learning_rate": 5.530259474617855e-07,
+      "loss": 0.1919,
+      "step": 3907
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8210650331210764,
+      "learning_rate": 5.514097777971939e-07,
+      "loss": 0.1829,
+      "step": 3908
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9768157127123505,
+      "learning_rate": 5.49795835317073e-07,
+      "loss": 0.202,
+      "step": 3909
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8048703836246344,
+      "learning_rate": 5.481841208294447e-07,
+      "loss": 0.2035,
+      "step": 3910
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1486695170808887,
+      "learning_rate": 5.465746351412104e-07,
+      "loss": 0.2245,
+      "step": 3911
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7571312605671712,
+      "learning_rate": 5.449673790581611e-07,
+      "loss": 0.1935,
+      "step": 3912
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8397170397228072,
+      "learning_rate": 5.433623533849658e-07,
+      "loss": 0.2008,
+      "step": 3913
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.7444825236367847,
+      "learning_rate": 5.417595589251828e-07,
+      "loss": 0.207,
+      "step": 3914
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8071567524539929,
+      "learning_rate": 5.401589964812487e-07,
+      "loss": 0.1943,
+      "step": 3915
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8228529711926907,
+      "learning_rate": 5.385606668544846e-07,
+      "loss": 0.1958,
+      "step": 3916
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9129528375869991,
+      "learning_rate": 5.369645708450921e-07,
+      "loss": 0.1896,
+      "step": 3917
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8845229574488706,
+      "learning_rate": 5.353707092521581e-07,
+      "loss": 0.1828,
+      "step": 3918
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7541373949720337,
+      "learning_rate": 5.337790828736466e-07,
+      "loss": 0.1925,
+      "step": 3919
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.0397183913808665,
+      "learning_rate": 5.321896925064069e-07,
+      "loss": 0.199,
+      "step": 3920
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8247989468634176,
+      "learning_rate": 5.306025389461639e-07,
+      "loss": 0.2109,
+      "step": 3921
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.0443913373255327,
+      "learning_rate": 5.290176229875271e-07,
+      "loss": 0.2118,
+      "step": 3922
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8572407372338757,
+      "learning_rate": 5.274349454239836e-07,
+      "loss": 0.2055,
+      "step": 3923
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8050551684893036,
+      "learning_rate": 5.258545070478999e-07,
+      "loss": 0.223,
+      "step": 3924
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1428706365557089,
+      "learning_rate": 5.242763086505204e-07,
+      "loss": 0.2037,
+      "step": 3925
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7944787625645583,
+      "learning_rate": 5.227003510219714e-07,
+      "loss": 0.1817,
+      "step": 3926
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8320405880723029,
+      "learning_rate": 5.211266349512534e-07,
+      "loss": 0.1941,
+      "step": 3927
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.0669326789781646,
+      "learning_rate": 5.195551612262478e-07,
+      "loss": 0.1994,
+      "step": 3928
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.851629054160342,
+      "learning_rate": 5.179859306337126e-07,
+      "loss": 0.2226,
+      "step": 3929
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.796345223447863,
+      "learning_rate": 5.16418943959282e-07,
+      "loss": 0.1923,
+      "step": 3930
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8083653391836518,
+      "learning_rate": 5.148542019874664e-07,
+      "loss": 0.2122,
+      "step": 3931
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8176011231151137,
+      "learning_rate": 5.132917055016523e-07,
+      "loss": 0.1904,
+      "step": 3932
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7828436099089595,
+      "learning_rate": 5.117314552841052e-07,
+      "loss": 0.2093,
+      "step": 3933
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8899260283472261,
+      "learning_rate": 5.101734521159613e-07,
+      "loss": 0.1952,
+      "step": 3934
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7853313953915956,
+      "learning_rate": 5.086176967772355e-07,
+      "loss": 0.1864,
+      "step": 3935
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8315092634447973,
+      "learning_rate": 5.070641900468149e-07,
+      "loss": 0.203,
+      "step": 3936
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8658607175539407,
+      "learning_rate": 5.055129327024638e-07,
+      "loss": 0.2041,
+      "step": 3937
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.3549164217991263,
+      "learning_rate": 5.039639255208156e-07,
+      "loss": 0.1841,
+      "step": 3938
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8458062028970617,
+      "learning_rate": 5.024171692773838e-07,
+      "loss": 0.2127,
+      "step": 3939
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7754061731629885,
+      "learning_rate": 5.008726647465468e-07,
+      "loss": 0.2051,
+      "step": 3940
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8509210561752191,
+      "learning_rate": 4.993304127015625e-07,
+      "loss": 0.1974,
+      "step": 3941
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8669291395009998,
+      "learning_rate": 4.977904139145579e-07,
+      "loss": 0.2053,
+      "step": 3942
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8087545604252768,
+      "learning_rate": 4.962526691565333e-07,
+      "loss": 0.2083,
+      "step": 3943
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7458371561544055,
+      "learning_rate": 4.947171791973587e-07,
+      "loss": 0.1989,
+      "step": 3944
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.0337221906533698,
+      "learning_rate": 4.931839448057779e-07,
+      "loss": 0.1885,
+      "step": 3945
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8080534575407161,
+      "learning_rate": 4.916529667494024e-07,
+      "loss": 0.2125,
+      "step": 3946
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8016966329393496,
+      "learning_rate": 4.90124245794717e-07,
+      "loss": 0.2018,
+      "step": 3947
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8294599072143369,
+      "learning_rate": 4.885977827070748e-07,
+      "loss": 0.2051,
+      "step": 3948
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8152538381593493,
+      "learning_rate": 4.87073578250698e-07,
+      "loss": 0.2007,
+      "step": 3949
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7766415747197416,
+      "learning_rate": 4.855516331886783e-07,
+      "loss": 0.203,
+      "step": 3950
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8262522620856726,
+      "learning_rate": 4.840319482829775e-07,
+      "loss": 0.1984,
+      "step": 3951
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8990240593180416,
+      "learning_rate": 4.825145242944262e-07,
+      "loss": 0.2121,
+      "step": 3952
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7776435423537928,
+      "learning_rate": 4.809993619827203e-07,
+      "loss": 0.1843,
+      "step": 3953
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8603876078938943,
+      "learning_rate": 4.794864621064266e-07,
+      "loss": 0.2318,
+      "step": 3954
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8184030749612083,
+      "learning_rate": 4.779758254229766e-07,
+      "loss": 0.2095,
+      "step": 3955
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8320142627438151,
+      "learning_rate": 4.7646745268867034e-07,
+      "loss": 0.2048,
+      "step": 3956
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9616725326968707,
+      "learning_rate": 4.749613446586726e-07,
+      "loss": 0.2081,
+      "step": 3957
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.859828561633418,
+      "learning_rate": 4.734575020870169e-07,
+      "loss": 0.189,
+      "step": 3958
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8926801427787361,
+      "learning_rate": 4.719559257266004e-07,
+      "loss": 0.2155,
+      "step": 3959
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8037717329541498,
+      "learning_rate": 4.704566163291879e-07,
+      "loss": 0.1903,
+      "step": 3960
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8440098832226688,
+      "learning_rate": 4.689595746454062e-07,
+      "loss": 0.174,
+      "step": 3961
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8489461212740874,
+      "learning_rate": 4.6746480142475015e-07,
+      "loss": 0.2035,
+      "step": 3962
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7798905778654694,
+      "learning_rate": 4.659722974155767e-07,
+      "loss": 0.2027,
+      "step": 3963
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8441987648332953,
+      "learning_rate": 4.6448206336510646e-07,
+      "loss": 0.1813,
+      "step": 3964
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7714088503164557,
+      "learning_rate": 4.6299410001942436e-07,
+      "loss": 0.1959,
+      "step": 3965
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8392635975332138,
+      "learning_rate": 4.6150840812348e-07,
+      "loss": 0.2083,
+      "step": 3966
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7659298230306423,
+      "learning_rate": 4.600249884210822e-07,
+      "loss": 0.2035,
+      "step": 3967
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8572929434374059,
+      "learning_rate": 4.5854384165490596e-07,
+      "loss": 0.2203,
+      "step": 3968
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8262853451648589,
+      "learning_rate": 4.5706496856648517e-07,
+      "loss": 0.2178,
+      "step": 3969
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7444520909475757,
+      "learning_rate": 4.5558836989621933e-07,
+      "loss": 0.1781,
+      "step": 3970
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8740174422967789,
+      "learning_rate": 4.541140463833649e-07,
+      "loss": 0.2105,
+      "step": 3971
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8909391407252222,
+      "learning_rate": 4.526419987660419e-07,
+      "loss": 0.2086,
+      "step": 3972
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8388246911174866,
+      "learning_rate": 4.511722277812286e-07,
+      "loss": 0.2168,
+      "step": 3973
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0226606882700247,
+      "learning_rate": 4.497047341647676e-07,
+      "loss": 0.2109,
+      "step": 3974
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8567427082776955,
+      "learning_rate": 4.4823951865135663e-07,
+      "loss": 0.2162,
+      "step": 3975
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8046771993491212,
+      "learning_rate": 4.467765819745562e-07,
+      "loss": 0.1977,
+      "step": 3976
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7991978634565645,
+      "learning_rate": 4.4531592486678597e-07,
+      "loss": 0.18,
+      "step": 3977
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8267469442414247,
+      "learning_rate": 4.43857548059321e-07,
+      "loss": 0.1935,
+      "step": 3978
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9124000824902273,
+      "learning_rate": 4.424014522822995e-07,
+      "loss": 0.2036,
+      "step": 3979
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.5105292693961567,
+      "learning_rate": 4.409476382647121e-07,
+      "loss": 0.2026,
+      "step": 3980
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8326972449505394,
+      "learning_rate": 4.3949610673441223e-07,
+      "loss": 0.1946,
+      "step": 3981
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8607337573273262,
+      "learning_rate": 4.380468584181069e-07,
+      "loss": 0.2053,
+      "step": 3982
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8406036526741832,
+      "learning_rate": 4.365998940413629e-07,
+      "loss": 0.2063,
+      "step": 3983
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.539713478160863,
+      "learning_rate": 4.351552143286014e-07,
+      "loss": 0.199,
+      "step": 3984
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8446264490184778,
+      "learning_rate": 4.33712820003101e-07,
+      "loss": 0.2036,
+      "step": 3985
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9532379828850427,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 0.2201,
+      "step": 3986
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8738826771027229,
+      "learning_rate": 4.3083489040127435e-07,
+      "loss": 0.2221,
+      "step": 3987
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.1842454818663806,
+      "learning_rate": 4.293993565657828e-07,
+      "loss": 0.2059,
+      "step": 3988
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7931529022003093,
+      "learning_rate": 4.2796611099921947e-07,
+      "loss": 0.2017,
+      "step": 3989
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8237265397220098,
+      "learning_rate": 4.265351544191365e-07,
+      "loss": 0.2019,
+      "step": 3990
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8525069013215144,
+      "learning_rate": 4.2510648754194474e-07,
+      "loss": 0.1984,
+      "step": 3991
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8653758417136634,
+      "learning_rate": 4.2368011108290306e-07,
+      "loss": 0.215,
+      "step": 3992
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9429577896230705,
+      "learning_rate": 4.222560257561276e-07,
+      "loss": 0.2003,
+      "step": 3993
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7856258158459324,
+      "learning_rate": 4.208342322745845e-07,
+      "loss": 0.2054,
+      "step": 3994
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7645592056500898,
+      "learning_rate": 4.1941473135009627e-07,
+      "loss": 0.1864,
+      "step": 3995
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8193316095825404,
+      "learning_rate": 4.1799752369333313e-07,
+      "loss": 0.2,
+      "step": 3996
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7685453523223607,
+      "learning_rate": 4.1658261001382096e-07,
+      "loss": 0.1803,
+      "step": 3997
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.802215142881008,
+      "learning_rate": 4.151699910199336e-07,
+      "loss": 0.183,
+      "step": 3998
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8363022565009625,
+      "learning_rate": 4.1375966741889874e-07,
+      "loss": 0.2109,
+      "step": 3999
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8181665212924627,
+      "learning_rate": 4.123516399167965e-07,
+      "loss": 0.2069,
+      "step": 4000
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.816008497542104,
+      "learning_rate": 4.109459092185519e-07,
+      "loss": 0.1978,
+      "step": 4001
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9871997972617493,
+      "learning_rate": 4.095424760279454e-07,
+      "loss": 0.2084,
+      "step": 4002
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9167169935208108,
+      "learning_rate": 4.0814134104760483e-07,
+      "loss": 0.2053,
+      "step": 4003
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8185705230639599,
+      "learning_rate": 4.067425049790075e-07,
+      "loss": 0.2071,
+      "step": 4004
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9049281152028358,
+      "learning_rate": 4.0534596852247867e-07,
+      "loss": 0.2231,
+      "step": 4005
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8092498948286584,
+      "learning_rate": 4.039517323771958e-07,
+      "loss": 0.2034,
+      "step": 4006
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8919151416554929,
+      "learning_rate": 4.0255979724118043e-07,
+      "loss": 0.2024,
+      "step": 4007
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7810190579772004,
+      "learning_rate": 4.0117016381130636e-07,
+      "loss": 0.1863,
+      "step": 4008
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8200126366165429,
+      "learning_rate": 3.997828327832909e-07,
+      "loss": 0.1896,
+      "step": 4009
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9769866482028404,
+      "learning_rate": 3.983978048517029e-07,
+      "loss": 0.2023,
+      "step": 4010
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.84281026321624,
+      "learning_rate": 3.970150807099538e-07,
+      "loss": 0.2028,
+      "step": 4011
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8402668106701563,
+      "learning_rate": 3.956346610503048e-07,
+      "loss": 0.2065,
+      "step": 4012
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8774358533915942,
+      "learning_rate": 3.9425654656386094e-07,
+      "loss": 0.2248,
+      "step": 4013
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.952572379347578,
+      "learning_rate": 3.9288073794057634e-07,
+      "loss": 0.1934,
+      "step": 4014
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.0120960895159956,
+      "learning_rate": 3.915072358692479e-07,
+      "loss": 0.2022,
+      "step": 4015
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9435792361938131,
+      "learning_rate": 3.9013604103751933e-07,
+      "loss": 0.1835,
+      "step": 4016
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7797716821713614,
+      "learning_rate": 3.887671541318777e-07,
+      "loss": 0.1875,
+      "step": 4017
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7898855338408389,
+      "learning_rate": 3.87400575837657e-07,
+      "loss": 0.1993,
+      "step": 4018
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1018463547736062,
+      "learning_rate": 3.860363068390327e-07,
+      "loss": 0.209,
+      "step": 4019
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8560954183496267,
+      "learning_rate": 3.8467434781902624e-07,
+      "loss": 0.2272,
+      "step": 4020
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8950292858362376,
+      "learning_rate": 3.833146994595005e-07,
+      "loss": 0.2089,
+      "step": 4021
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7758595738618209,
+      "learning_rate": 3.819573624411643e-07,
+      "loss": 0.1893,
+      "step": 4022
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.791461453623267,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 0.204,
+      "step": 4023
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7794960327594284,
+      "learning_rate": 3.792496251451e-07,
+      "loss": 0.1972,
+      "step": 4024
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9795948632455799,
+      "learning_rate": 3.7789922622300023e-07,
+      "loss": 0.2336,
+      "step": 4025
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.811796500342143,
+      "learning_rate": 3.765511413533429e-07,
+      "loss": 0.2057,
+      "step": 4026
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8547985496713922,
+      "learning_rate": 3.7520537121104707e-07,
+      "loss": 0.1992,
+      "step": 4027
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.866756839839874,
+      "learning_rate": 3.7386191646987094e-07,
+      "loss": 0.2043,
+      "step": 4028
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8193657620692705,
+      "learning_rate": 3.725207778024148e-07,
+      "loss": 0.2029,
+      "step": 4029
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1669543881884354,
+      "learning_rate": 3.711819558801183e-07,
+      "loss": 0.2024,
+      "step": 4030
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8754252641413152,
+      "learning_rate": 3.69845451373263e-07,
+      "loss": 0.1704,
+      "step": 4031
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7819328713805946,
+      "learning_rate": 3.685112649509681e-07,
+      "loss": 0.1916,
+      "step": 4032
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9273506355629815,
+      "learning_rate": 3.671793972811954e-07,
+      "loss": 0.2072,
+      "step": 4033
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9283250303146794,
+      "learning_rate": 3.6584984903074104e-07,
+      "loss": 0.2262,
+      "step": 4034
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8022456854805685,
+      "learning_rate": 3.645226208652458e-07,
+      "loss": 0.2061,
+      "step": 4035
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9099961151052507,
+      "learning_rate": 3.631977134491832e-07,
+      "loss": 0.1968,
+      "step": 4036
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7972179526475431,
+      "learning_rate": 3.6187512744586873e-07,
+      "loss": 0.2168,
+      "step": 4037
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7513454449950976,
+      "learning_rate": 3.6055486351745327e-07,
+      "loss": 0.1969,
+      "step": 4038
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8311959210392026,
+      "learning_rate": 3.592369223249281e-07,
+      "loss": 0.2258,
+      "step": 4039
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8694847031987086,
+      "learning_rate": 3.5792130452811825e-07,
+      "loss": 0.2037,
+      "step": 4040
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1496497172937292,
+      "learning_rate": 3.5660801078568785e-07,
+      "loss": 0.1985,
+      "step": 4041
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8437910551504004,
+      "learning_rate": 3.5529704175513604e-07,
+      "loss": 0.2067,
+      "step": 4042
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8545810190934139,
+      "learning_rate": 3.539883980928005e-07,
+      "loss": 0.1976,
+      "step": 4043
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8074444675133978,
+      "learning_rate": 3.5268208045385167e-07,
+      "loss": 0.201,
+      "step": 4044
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7972387981243508,
+      "learning_rate": 3.513780894922969e-07,
+      "loss": 0.1901,
+      "step": 4045
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8276084421551333,
+      "learning_rate": 3.500764258609779e-07,
+      "loss": 0.2033,
+      "step": 4046
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9068690006391373,
+      "learning_rate": 3.487770902115739e-07,
+      "loss": 0.1975,
+      "step": 4047
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8600377865524359,
+      "learning_rate": 3.4748008319459457e-07,
+      "loss": 0.2004,
+      "step": 4048
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.847491288567035,
+      "learning_rate": 3.4618540545938664e-07,
+      "loss": 0.2278,
+      "step": 4049
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.075709827999873,
+      "learning_rate": 3.4489305765413097e-07,
+      "loss": 0.2115,
+      "step": 4050
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8140198465286225,
+      "learning_rate": 3.4360304042583946e-07,
+      "loss": 0.2084,
+      "step": 4051
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8136309888002193,
+      "learning_rate": 3.4231535442035947e-07,
+      "loss": 0.2069,
+      "step": 4052
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.822864457395013,
+      "learning_rate": 3.410300002823691e-07,
+      "loss": 0.202,
+      "step": 4053
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7831291893860677,
+      "learning_rate": 3.3974697865538195e-07,
+      "loss": 0.1954,
+      "step": 4054
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.0449709010959651,
+      "learning_rate": 3.3846629018174037e-07,
+      "loss": 0.2135,
+      "step": 4055
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8150230963439088,
+      "learning_rate": 3.37187935502622e-07,
+      "loss": 0.2106,
+      "step": 4056
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7782227323266929,
+      "learning_rate": 3.3591191525803255e-07,
+      "loss": 0.189,
+      "step": 4057
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8531261783708025,
+      "learning_rate": 3.346382300868134e-07,
+      "loss": 0.1996,
+      "step": 4058
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8701497909251914,
+      "learning_rate": 3.3336688062663327e-07,
+      "loss": 0.2206,
+      "step": 4059
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8290264359551217,
+      "learning_rate": 3.320978675139919e-07,
+      "loss": 0.2197,
+      "step": 4060
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9091683119177891,
+      "learning_rate": 3.308311913842205e-07,
+      "loss": 0.2089,
+      "step": 4061
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8630885156994113,
+      "learning_rate": 3.2956685287148014e-07,
+      "loss": 0.2125,
+      "step": 4062
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7983941828493861,
+      "learning_rate": 3.2830485260876064e-07,
+      "loss": 0.2004,
+      "step": 4063
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.878752610285,
+      "learning_rate": 3.2704519122788324e-07,
+      "loss": 0.1919,
+      "step": 4064
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8530859190097071,
+      "learning_rate": 3.2578786935949514e-07,
+      "loss": 0.2021,
+      "step": 4065
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.0380567285473317,
+      "learning_rate": 3.24532887633075e-07,
+      "loss": 0.1903,
+      "step": 4066
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8217711823433189,
+      "learning_rate": 3.2328024667692804e-07,
+      "loss": 0.1971,
+      "step": 4067
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8146412476366669,
+      "learning_rate": 3.220299471181898e-07,
+      "loss": 0.2041,
+      "step": 4068
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8105164105560853,
+      "learning_rate": 3.207819895828196e-07,
+      "loss": 0.2054,
+      "step": 4069
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8513924045503527,
+      "learning_rate": 3.195363746956087e-07,
+      "loss": 0.2226,
+      "step": 4070
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8090537985382755,
+      "learning_rate": 3.182931030801728e-07,
+      "loss": 0.2029,
+      "step": 4071
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8111647254341061,
+      "learning_rate": 3.1705217535895504e-07,
+      "loss": 0.197,
+      "step": 4072
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7878230264025552,
+      "learning_rate": 3.158135921532268e-07,
+      "loss": 0.2164,
+      "step": 4073
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7783445310530277,
+      "learning_rate": 3.1457735408308153e-07,
+      "loss": 0.1893,
+      "step": 4074
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8488899061914819,
+      "learning_rate": 3.133434617674441e-07,
+      "loss": 0.2152,
+      "step": 4075
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7874291292828716,
+      "learning_rate": 3.121119158240604e-07,
+      "loss": 0.2076,
+      "step": 4076
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.0715493766371045,
+      "learning_rate": 3.108827168695039e-07,
+      "loss": 0.2034,
+      "step": 4077
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8072796592656187,
+      "learning_rate": 3.096558655191706e-07,
+      "loss": 0.1887,
+      "step": 4078
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7926121723809008,
+      "learning_rate": 3.084313623872859e-07,
+      "loss": 0.2013,
+      "step": 4079
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8014619768255167,
+      "learning_rate": 3.0720920808689435e-07,
+      "loss": 0.1953,
+      "step": 4080
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8224520758177808,
+      "learning_rate": 3.059894032298688e-07,
+      "loss": 0.2051,
+      "step": 4081
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5463662010291972,
+      "learning_rate": 3.047719484269018e-07,
+      "loss": 0.2027,
+      "step": 4082
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.587294322651313,
+      "learning_rate": 3.035568442875136e-07,
+      "loss": 0.2052,
+      "step": 4083
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7637705146631981,
+      "learning_rate": 3.023440914200448e-07,
+      "loss": 0.1978,
+      "step": 4084
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7916118317023755,
+      "learning_rate": 3.011336904316592e-07,
+      "loss": 0.1956,
+      "step": 4085
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.79590833449748,
+      "learning_rate": 2.9992564192834253e-07,
+      "loss": 0.1882,
+      "step": 4086
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.855246961537655,
+      "learning_rate": 2.9871994651490543e-07,
+      "loss": 0.201,
+      "step": 4087
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8621021345395968,
+      "learning_rate": 2.9751660479497737e-07,
+      "loss": 0.1837,
+      "step": 4088
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.04692489190766,
+      "learning_rate": 2.963156173710119e-07,
+      "loss": 0.1887,
+      "step": 4089
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9015523421188091,
+      "learning_rate": 2.9511698484428144e-07,
+      "loss": 0.2,
+      "step": 4090
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4039299235649725,
+      "learning_rate": 2.93920707814882e-07,
+      "loss": 0.1915,
+      "step": 4091
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7377025794286661,
+      "learning_rate": 2.9272678688172905e-07,
+      "loss": 0.1896,
+      "step": 4092
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7969972361187933,
+      "learning_rate": 2.915352226425583e-07,
+      "loss": 0.2242,
+      "step": 4093
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8820131183472176,
+      "learning_rate": 2.9034601569392417e-07,
+      "loss": 0.2114,
+      "step": 4094
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8204166008343741,
+      "learning_rate": 2.891591666312049e-07,
+      "loss": 0.2051,
+      "step": 4095
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7115129845206252,
+      "learning_rate": 2.879746760485946e-07,
+      "loss": 0.1887,
+      "step": 4096
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9867132005357729,
+      "learning_rate": 2.867925445391079e-07,
+      "loss": 0.2219,
+      "step": 4097
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8384580050489613,
+      "learning_rate": 2.85612772694579e-07,
+      "loss": 0.1822,
+      "step": 4098
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9031718961635928,
+      "learning_rate": 2.8443536110565993e-07,
+      "loss": 0.1956,
+      "step": 4099
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7989110113658948,
+      "learning_rate": 2.832603103618214e-07,
+      "loss": 0.1915,
+      "step": 4100
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9930514568363022,
+      "learning_rate": 2.820876210513507e-07,
+      "loss": 0.1976,
+      "step": 4101
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.785104857144247,
+      "learning_rate": 2.809172937613552e-07,
+      "loss": 0.1979,
+      "step": 4102
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8880068002715205,
+      "learning_rate": 2.7974932907775863e-07,
+      "loss": 0.2193,
+      "step": 4103
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8018004185217695,
+      "learning_rate": 2.7858372758530206e-07,
+      "loss": 0.2088,
+      "step": 4104
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.823567412170167,
+      "learning_rate": 2.77420489867542e-07,
+      "loss": 0.1993,
+      "step": 4105
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9245627622270036,
+      "learning_rate": 2.762596165068548e-07,
+      "loss": 0.2181,
+      "step": 4106
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.117502495015584,
+      "learning_rate": 2.7510110808443014e-07,
+      "loss": 0.2338,
+      "step": 4107
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8928002563456033,
+      "learning_rate": 2.739449651802756e-07,
+      "loss": 0.2235,
+      "step": 4108
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8014459355448333,
+      "learning_rate": 2.7279118837321094e-07,
+      "loss": 0.2113,
+      "step": 4109
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9130398337053962,
+      "learning_rate": 2.7163977824087696e-07,
+      "loss": 0.1995,
+      "step": 4110
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7890608418363301,
+      "learning_rate": 2.7049073535972415e-07,
+      "loss": 0.1992,
+      "step": 4111
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7793817472289398,
+      "learning_rate": 2.6934406030502235e-07,
+      "loss": 0.2072,
+      "step": 4112
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8498506527020736,
+      "learning_rate": 2.6819975365085237e-07,
+      "loss": 0.2174,
+      "step": 4113
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8777420835845929,
+      "learning_rate": 2.6705781597011147e-07,
+      "loss": 0.2185,
+      "step": 4114
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8233917333616306,
+      "learning_rate": 2.659182478345096e-07,
+      "loss": 0.197,
+      "step": 4115
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7292596818735203,
+      "learning_rate": 2.6478104981457254e-07,
+      "loss": 0.1832,
+      "step": 4116
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.789017066718651,
+      "learning_rate": 2.63646222479636e-07,
+      "loss": 0.1947,
+      "step": 4117
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8174777293285836,
+      "learning_rate": 2.6251376639785163e-07,
+      "loss": 0.1992,
+      "step": 4118
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8423757577512875,
+      "learning_rate": 2.613836821361826e-07,
+      "loss": 0.2017,
+      "step": 4119
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4259339690009898,
+      "learning_rate": 2.6025597026040463e-07,
+      "loss": 0.2113,
+      "step": 4120
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9071230165050423,
+      "learning_rate": 2.5913063133510776e-07,
+      "loss": 0.2102,
+      "step": 4121
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8043109371412812,
+      "learning_rate": 2.5800766592369077e-07,
+      "loss": 0.1862,
+      "step": 4122
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7968977616106517,
+      "learning_rate": 2.5688707458836724e-07,
+      "loss": 0.2042,
+      "step": 4123
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.819725181814112,
+      "learning_rate": 2.5576885789015905e-07,
+      "loss": 0.1733,
+      "step": 4124
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8581505494066862,
+      "learning_rate": 2.5465301638890216e-07,
+      "loss": 0.2117,
+      "step": 4125
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8074135776786505,
+      "learning_rate": 2.535395506432403e-07,
+      "loss": 0.2062,
+      "step": 4126
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8684734538589884,
+      "learning_rate": 2.524284612106315e-07,
+      "loss": 0.2168,
+      "step": 4127
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8288583168367408,
+      "learning_rate": 2.5131974864734063e-07,
+      "loss": 0.2273,
+      "step": 4128
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8393400849155237,
+      "learning_rate": 2.50213413508445e-07,
+      "loss": 0.2154,
+      "step": 4129
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7873485432676535,
+      "learning_rate": 2.4910945634783024e-07,
+      "loss": 0.1956,
+      "step": 4130
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8443282662115535,
+      "learning_rate": 2.4800787771819266e-07,
+      "loss": 0.2096,
+      "step": 4131
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8952730054098736,
+      "learning_rate": 2.469086781710367e-07,
+      "loss": 0.21,
+      "step": 4132
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8100945367303942,
+      "learning_rate": 2.45811858256676e-07,
+      "loss": 0.1999,
+      "step": 4133
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8285141329397564,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.2052,
+      "step": 4134
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8262424770430412,
+      "learning_rate": 2.436253595216376e-07,
+      "loss": 0.1921,
+      "step": 4135
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8175661286515162,
+      "learning_rate": 2.425356817956287e-07,
+      "loss": 0.2232,
+      "step": 4136
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.0994874313383547,
+      "learning_rate": 2.4144838589175433e-07,
+      "loss": 0.205,
+      "step": 4137
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8827937018748824,
+      "learning_rate": 2.403634723543674e-07,
+      "loss": 0.1951,
+      "step": 4138
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8147082112766576,
+      "learning_rate": 2.3928094172663005e-07,
+      "loss": 0.204,
+      "step": 4139
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7871187218372314,
+      "learning_rate": 2.3820079455051048e-07,
+      "loss": 0.1992,
+      "step": 4140
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7914214043614588,
+      "learning_rate": 2.3712303136678383e-07,
+      "loss": 0.2074,
+      "step": 4141
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8240587231891875,
+      "learning_rate": 2.3604765271503073e-07,
+      "loss": 0.2096,
+      "step": 4142
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.828801515989596,
+      "learning_rate": 2.3497465913364047e-07,
+      "loss": 0.1964,
+      "step": 4143
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8235848857000313,
+      "learning_rate": 2.3390405115980563e-07,
+      "loss": 0.1952,
+      "step": 4144
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8911828206454864,
+      "learning_rate": 2.3283582932952575e-07,
+      "loss": 0.2057,
+      "step": 4145
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9097715793489488,
+      "learning_rate": 2.3176999417760637e-07,
+      "loss": 0.2055,
+      "step": 4146
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7860497390735859,
+      "learning_rate": 2.3070654623765677e-07,
+      "loss": 0.1892,
+      "step": 4147
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9805113324852046,
+      "learning_rate": 2.2964548604209214e-07,
+      "loss": 0.196,
+      "step": 4148
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7629509866238189,
+      "learning_rate": 2.2858681412213034e-07,
+      "loss": 0.1974,
+      "step": 4149
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9386644456494928,
+      "learning_rate": 2.2753053100779632e-07,
+      "loss": 0.2049,
+      "step": 4150
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8139201038364291,
+      "learning_rate": 2.2647663722791645e-07,
+      "loss": 0.1995,
+      "step": 4151
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9859784686765883,
+      "learning_rate": 2.2542513331012372e-07,
+      "loss": 0.2014,
+      "step": 4152
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7621576170565827,
+      "learning_rate": 2.2437601978085144e-07,
+      "loss": 0.1989,
+      "step": 4153
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.827129174284319,
+      "learning_rate": 2.2332929716533947e-07,
+      "loss": 0.1995,
+      "step": 4154
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7855018164807245,
+      "learning_rate": 2.2228496598762696e-07,
+      "loss": 0.1925,
+      "step": 4155
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7875362522427479,
+      "learning_rate": 2.2124302677055954e-07,
+      "loss": 0.2048,
+      "step": 4156
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9885306836245875,
+      "learning_rate": 2.202034800357833e-07,
+      "loss": 0.2346,
+      "step": 4157
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2016353764145098,
+      "learning_rate": 2.1916632630374579e-07,
+      "loss": 0.1882,
+      "step": 4158
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8505742820576561,
+      "learning_rate": 2.1813156609369778e-07,
+      "loss": 0.2211,
+      "step": 4159
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8040158792490142,
+      "learning_rate": 2.1709919992369211e-07,
+      "loss": 0.2115,
+      "step": 4160
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8390183171765612,
+      "learning_rate": 2.1606922831058198e-07,
+      "loss": 0.2326,
+      "step": 4161
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8606981804539549,
+      "learning_rate": 2.1504165177002212e-07,
+      "loss": 0.2082,
+      "step": 4162
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8198131133972822,
+      "learning_rate": 2.1401647081646825e-07,
+      "loss": 0.2108,
+      "step": 4163
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.896650313702402,
+      "learning_rate": 2.1299368596317705e-07,
+      "loss": 0.215,
+      "step": 4164
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8859019957051937,
+      "learning_rate": 2.1197329772220555e-07,
+      "loss": 0.198,
+      "step": 4165
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7906853686039236,
+      "learning_rate": 2.1095530660440954e-07,
+      "loss": 0.177,
+      "step": 4166
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7790926103943034,
+      "learning_rate": 2.0993971311944528e-07,
+      "loss": 0.2147,
+      "step": 4167
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8201154419516158,
+      "learning_rate": 2.0892651777577045e-07,
+      "loss": 0.2031,
+      "step": 4168
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8206856328468557,
+      "learning_rate": 2.0791572108064152e-07,
+      "loss": 0.1947,
+      "step": 4169
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9875941492994058,
+      "learning_rate": 2.069073235401109e-07,
+      "loss": 0.2114,
+      "step": 4170
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8243896534642525,
+      "learning_rate": 2.0590132565903475e-07,
+      "loss": 0.2049,
+      "step": 4171
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.832983669290131,
+      "learning_rate": 2.0489772794106411e-07,
+      "loss": 0.2148,
+      "step": 4172
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8591176780630098,
+      "learning_rate": 2.0389653088865035e-07,
+      "loss": 0.2031,
+      "step": 4173
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9588169297025346,
+      "learning_rate": 2.028977350030409e-07,
+      "loss": 0.226,
+      "step": 4174
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.850681523446547,
+      "learning_rate": 2.0190134078428402e-07,
+      "loss": 0.1899,
+      "step": 4175
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8081795111571465,
+      "learning_rate": 2.0090734873122243e-07,
+      "loss": 0.211,
+      "step": 4176
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7557542395631186,
+      "learning_rate": 1.9991575934149965e-07,
+      "loss": 0.1762,
+      "step": 4177
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2340571176015243,
+      "learning_rate": 1.989265731115525e-07,
+      "loss": 0.2109,
+      "step": 4178
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7820928480865981,
+      "learning_rate": 1.9793979053661815e-07,
+      "loss": 0.2053,
+      "step": 4179
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8606703554402582,
+      "learning_rate": 1.969554121107281e-07,
+      "loss": 0.193,
+      "step": 4180
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8418066047077226,
+      "learning_rate": 1.9597343832671144e-07,
+      "loss": 0.2177,
+      "step": 4181
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.086391659854931,
+      "learning_rate": 1.9499386967619104e-07,
+      "loss": 0.2086,
+      "step": 4182
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9450698516642865,
+      "learning_rate": 1.940167066495896e-07,
+      "loss": 0.1983,
+      "step": 4183
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7056253670820704,
+      "learning_rate": 1.9304194973612244e-07,
+      "loss": 0.1832,
+      "step": 4184
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.782991966873025,
+      "learning_rate": 1.920695994238009e-07,
+      "loss": 0.1933,
+      "step": 4185
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8573609660712112,
+      "learning_rate": 1.910996561994316e-07,
+      "loss": 0.1934,
+      "step": 4186
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.829769055439044,
+      "learning_rate": 1.901321205486173e-07,
+      "loss": 0.1963,
+      "step": 4187
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8426681669347346,
+      "learning_rate": 1.8916699295575324e-07,
+      "loss": 0.2182,
+      "step": 4188
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9302262334168854,
+      "learning_rate": 1.8820427390403074e-07,
+      "loss": 0.2071,
+      "step": 4189
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8473500834780492,
+      "learning_rate": 1.872439638754331e-07,
+      "loss": 0.1901,
+      "step": 4190
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7996164512913472,
+      "learning_rate": 1.8628606335074017e-07,
+      "loss": 0.2024,
+      "step": 4191
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7847514083776127,
+      "learning_rate": 1.853305728095245e-07,
+      "loss": 0.2069,
+      "step": 4192
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.031610681576725,
+      "learning_rate": 1.8437749273015116e-07,
+      "loss": 0.215,
+      "step": 4193
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.805539579521808,
+      "learning_rate": 1.834268235897807e-07,
+      "loss": 0.1804,
+      "step": 4194
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.887077146390733,
+      "learning_rate": 1.8247856586436352e-07,
+      "loss": 0.1893,
+      "step": 4195
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7653124511822842,
+      "learning_rate": 1.8153272002864542e-07,
+      "loss": 0.1898,
+      "step": 4196
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.791335375464853,
+      "learning_rate": 1.805892865561637e-07,
+      "loss": 0.1922,
+      "step": 4197
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.786372606213497,
+      "learning_rate": 1.7964826591924722e-07,
+      "loss": 0.197,
+      "step": 4198
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8234488306562243,
+      "learning_rate": 1.787096585890169e-07,
+      "loss": 0.2085,
+      "step": 4199
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7993468969042029,
+      "learning_rate": 1.7777346503538796e-07,
+      "loss": 0.2154,
+      "step": 4200
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8316064278173886,
+      "learning_rate": 1.7683968572706323e-07,
+      "loss": 0.2074,
+      "step": 4201
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9187582348839382,
+      "learning_rate": 1.7590832113154045e-07,
+      "loss": 0.1879,
+      "step": 4202
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8315349316873673,
+      "learning_rate": 1.749793717151055e-07,
+      "loss": 0.2046,
+      "step": 4203
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8848335415154217,
+      "learning_rate": 1.7405283794283745e-07,
+      "loss": 0.1933,
+      "step": 4204
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.78807524838793,
+      "learning_rate": 1.731287202786053e-07,
+      "loss": 0.2093,
+      "step": 4205
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8826635913531817,
+      "learning_rate": 1.7220701918506666e-07,
+      "loss": 0.2176,
+      "step": 4206
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7926420529353946,
+      "learning_rate": 1.7128773512367137e-07,
+      "loss": 0.1789,
+      "step": 4207
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0525137389602433,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.2138,
+      "step": 4208
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8663704803526238,
+      "learning_rate": 1.6945641993705743e-07,
+      "loss": 0.224,
+      "step": 4209
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8391387054779983,
+      "learning_rate": 1.685443897286859e-07,
+      "loss": 0.1978,
+      "step": 4210
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.885819579178114,
+      "learning_rate": 1.6763477838615093e-07,
+      "loss": 0.1971,
+      "step": 4211
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7858147987711983,
+      "learning_rate": 1.6672758636485043e-07,
+      "loss": 0.2031,
+      "step": 4212
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.000006944020778,
+      "learning_rate": 1.6582281411896827e-07,
+      "loss": 0.221,
+      "step": 4213
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7912122257276801,
+      "learning_rate": 1.6492046210147884e-07,
+      "loss": 0.2088,
+      "step": 4214
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8490281381625883,
+      "learning_rate": 1.6402053076414294e-07,
+      "loss": 0.1959,
+      "step": 4215
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8328294942344804,
+      "learning_rate": 1.6312302055751295e-07,
+      "loss": 0.2032,
+      "step": 4216
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0421647356294217,
+      "learning_rate": 1.62227931930925e-07,
+      "loss": 0.1888,
+      "step": 4217
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7922207550870968,
+      "learning_rate": 1.6133526533250566e-07,
+      "loss": 0.1837,
+      "step": 4218
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8049941924765668,
+      "learning_rate": 1.604450212091685e-07,
+      "loss": 0.212,
+      "step": 4219
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7567240493690913,
+      "learning_rate": 1.5955720000661312e-07,
+      "loss": 0.1949,
+      "step": 4220
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7922885187351076,
+      "learning_rate": 1.5867180216932676e-07,
+      "loss": 0.1968,
+      "step": 4221
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.917386285547315,
+      "learning_rate": 1.577888281405826e-07,
+      "loss": 0.1841,
+      "step": 4222
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8161445404705024,
+      "learning_rate": 1.5690827836244317e-07,
+      "loss": 0.2013,
+      "step": 4223
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.809485819663372,
+      "learning_rate": 1.5603015327575354e-07,
+      "loss": 0.2033,
+      "step": 4224
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.810360399346151,
+      "learning_rate": 1.551544533201471e-07,
+      "loss": 0.1968,
+      "step": 4225
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7766166914746732,
+      "learning_rate": 1.5428117893404315e-07,
+      "loss": 0.2041,
+      "step": 4226
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8024287828428096,
+      "learning_rate": 1.5341033055464527e-07,
+      "loss": 0.1926,
+      "step": 4227
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8055022175664555,
+      "learning_rate": 1.5254190861794415e-07,
+      "loss": 0.1985,
+      "step": 4228
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8627157343120081,
+      "learning_rate": 1.5167591355871425e-07,
+      "loss": 0.2111,
+      "step": 4229
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.453449685006157,
+      "learning_rate": 1.5081234581051484e-07,
+      "loss": 0.2148,
+      "step": 4230
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8294759909548192,
+      "learning_rate": 1.4995120580569178e-07,
+      "loss": 0.2072,
+      "step": 4231
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8562373753813514,
+      "learning_rate": 1.490924939753735e-07,
+      "loss": 0.2276,
+      "step": 4232
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9963290612536272,
+      "learning_rate": 1.4823621074947503e-07,
+      "loss": 0.2147,
+      "step": 4233
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8387714928039369,
+      "learning_rate": 1.473823565566923e-07,
+      "loss": 0.1988,
+      "step": 4234
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8409647201782021,
+      "learning_rate": 1.4653093182450896e-07,
+      "loss": 0.1993,
+      "step": 4235
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8273291475236493,
+      "learning_rate": 1.4568193697918843e-07,
+      "loss": 0.2122,
+      "step": 4236
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7623212555055621,
+      "learning_rate": 1.4483537244578183e-07,
+      "loss": 0.1922,
+      "step": 4237
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.759953619147308,
+      "learning_rate": 1.4399123864811904e-07,
+      "loss": 0.1976,
+      "step": 4238
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7988001854552175,
+      "learning_rate": 1.4314953600881641e-07,
+      "loss": 0.1983,
+      "step": 4239
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9922995974955094,
+      "learning_rate": 1.4231026494927082e-07,
+      "loss": 0.21,
+      "step": 4240
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9619602080251539,
+      "learning_rate": 1.414734258896644e-07,
+      "loss": 0.2415,
+      "step": 4241
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8512927261573442,
+      "learning_rate": 1.4063901924895985e-07,
+      "loss": 0.2148,
+      "step": 4242
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9022481267706122,
+      "learning_rate": 1.398070454449013e-07,
+      "loss": 0.2057,
+      "step": 4243
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0241248735506996,
+      "learning_rate": 1.3897750489401785e-07,
+      "loss": 0.2066,
+      "step": 4244
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8621603722211275,
+      "learning_rate": 1.3815039801161723e-07,
+      "loss": 0.2031,
+      "step": 4245
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8709267103435034,
+      "learning_rate": 1.3732572521179043e-07,
+      "loss": 0.2227,
+      "step": 4246
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9494319134778009,
+      "learning_rate": 1.3650348690740888e-07,
+      "loss": 0.2071,
+      "step": 4247
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8895451494310613,
+      "learning_rate": 1.3568368351012718e-07,
+      "loss": 0.2081,
+      "step": 4248
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7906949405967503,
+      "learning_rate": 1.348663154303781e-07,
+      "loss": 0.2173,
+      "step": 4249
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.862394142097186,
+      "learning_rate": 1.3405138307737765e-07,
+      "loss": 0.2138,
+      "step": 4250
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3889252057743213,
+      "learning_rate": 1.3323888685912e-07,
+      "loss": 0.2052,
+      "step": 4251
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8356933110865523,
+      "learning_rate": 1.324288271823826e-07,
+      "loss": 0.2252,
+      "step": 4252
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8194552959005947,
+      "learning_rate": 1.3162120445272096e-07,
+      "loss": 0.1893,
+      "step": 4253
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8029613071077275,
+      "learning_rate": 1.3081601907447007e-07,
+      "loss": 0.1962,
+      "step": 4254
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.91058977026081,
+      "learning_rate": 1.3001327145074572e-07,
+      "loss": 0.199,
+      "step": 4255
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.789578733122997,
+      "learning_rate": 1.2921296198344424e-07,
+      "loss": 0.2146,
+      "step": 4256
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8272423391959876,
+      "learning_rate": 1.2841509107323845e-07,
+      "loss": 0.1907,
+      "step": 4257
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.842025492128621,
+      "learning_rate": 1.2761965911958385e-07,
+      "loss": 0.1896,
+      "step": 4258
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.291652478360475,
+      "learning_rate": 1.2682666652071184e-07,
+      "loss": 0.1952,
+      "step": 4259
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9239927421269486,
+      "learning_rate": 1.260361136736349e-07,
+      "loss": 0.2128,
+      "step": 4260
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7838177539747502,
+      "learning_rate": 1.2524800097414193e-07,
+      "loss": 0.1975,
+      "step": 4261
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7873491351431597,
+      "learning_rate": 1.244623288168012e-07,
+      "loss": 0.2058,
+      "step": 4262
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.855192411079509,
+      "learning_rate": 1.236790975949592e-07,
+      "loss": 0.2165,
+      "step": 4263
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7541051438106032,
+      "learning_rate": 1.2289830770074064e-07,
+      "loss": 0.1814,
+      "step": 4264
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.833254182586492,
+      "learning_rate": 1.2211995952504663e-07,
+      "loss": 0.2154,
+      "step": 4265
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7854138555330467,
+      "learning_rate": 1.2134405345755773e-07,
+      "loss": 0.2057,
+      "step": 4266
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8364236436410111,
+      "learning_rate": 1.20570589886731e-07,
+      "loss": 0.2076,
+      "step": 4267
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8519029750269763,
+      "learning_rate": 1.1979956919979996e-07,
+      "loss": 0.1964,
+      "step": 4268
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8361296851862722,
+      "learning_rate": 1.1903099178277644e-07,
+      "loss": 0.2108,
+      "step": 4269
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8183855821269452,
+      "learning_rate": 1.1826485802044707e-07,
+      "loss": 0.1873,
+      "step": 4270
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7675057024440398,
+      "learning_rate": 1.1750116829637837e-07,
+      "loss": 0.1805,
+      "step": 4271
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0498629669119803,
+      "learning_rate": 1.1673992299290893e-07,
+      "loss": 0.2008,
+      "step": 4272
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8148694573564313,
+      "learning_rate": 1.1598112249115723e-07,
+      "loss": 0.2021,
+      "step": 4273
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9546448567812671,
+      "learning_rate": 1.1522476717101605e-07,
+      "loss": 0.2063,
+      "step": 4274
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8282863176877058,
+      "learning_rate": 1.144708574111547e-07,
+      "loss": 0.2064,
+      "step": 4275
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9507271597389455,
+      "learning_rate": 1.137193935890174e-07,
+      "loss": 0.2049,
+      "step": 4276
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8513643801224212,
+      "learning_rate": 1.1297037608082372e-07,
+      "loss": 0.201,
+      "step": 4277
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.899111170799662,
+      "learning_rate": 1.1222380526156929e-07,
+      "loss": 0.2058,
+      "step": 4278
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8300072874619441,
+      "learning_rate": 1.1147968150502453e-07,
+      "loss": 0.2034,
+      "step": 4279
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8445980988705365,
+      "learning_rate": 1.1073800518373423e-07,
+      "loss": 0.1965,
+      "step": 4280
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7701317522212532,
+      "learning_rate": 1.099987766690186e-07,
+      "loss": 0.1988,
+      "step": 4281
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8554606624469464,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 0.1978,
+      "step": 4282
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8099979998750768,
+      "learning_rate": 1.0852766453846308e-07,
+      "loss": 0.1837,
+      "step": 4283
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.78862395433,
+      "learning_rate": 1.0779578165913463e-07,
+      "loss": 0.1893,
+      "step": 4284
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7814933201254386,
+      "learning_rate": 1.0706634805940363e-07,
+      "loss": 0.1892,
+      "step": 4285
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8920751389521329,
+      "learning_rate": 1.0633936410446077e-07,
+      "loss": 0.2141,
+      "step": 4286
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8379601172923967,
+      "learning_rate": 1.056148301582699e-07,
+      "loss": 0.2157,
+      "step": 4287
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8722097032903416,
+      "learning_rate": 1.0489274658356808e-07,
+      "loss": 0.2007,
+      "step": 4288
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9451538560308121,
+      "learning_rate": 1.0417311374186668e-07,
+      "loss": 0.225,
+      "step": 4289
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.220498363044524,
+      "learning_rate": 1.0345593199344972e-07,
+      "loss": 0.1976,
+      "step": 4290
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8977496964741644,
+      "learning_rate": 1.0274120169737334e-07,
+      "loss": 0.2046,
+      "step": 4291
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.9272643244826841,
+      "learning_rate": 1.0202892321146796e-07,
+      "loss": 0.1998,
+      "step": 4292
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8031092785598125,
+      "learning_rate": 1.0131909689233444e-07,
+      "loss": 0.2051,
+      "step": 4293
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.78627273922342,
+      "learning_rate": 1.0061172309534739e-07,
+      "loss": 0.1943,
+      "step": 4294
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7873020109056682,
+      "learning_rate": 9.990680217465243e-08,
+      "loss": 0.181,
+      "step": 4295
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7997785074167669,
+      "learning_rate": 9.92043344831689e-08,
+      "loss": 0.2048,
+      "step": 4296
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8260040443544108,
+      "learning_rate": 9.850432037258607e-08,
+      "loss": 0.1998,
+      "step": 4297
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8558640535994294,
+      "learning_rate": 9.780676019336632e-08,
+      "loss": 0.217,
+      "step": 4298
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7825287467776454,
+      "learning_rate": 9.711165429474256e-08,
+      "loss": 0.2056,
+      "step": 4299
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.2272565270640492,
+      "learning_rate": 9.641900302471918e-08,
+      "loss": 0.1783,
+      "step": 4300
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8220703775395123,
+      "learning_rate": 9.572880673007156e-08,
+      "loss": 0.1987,
+      "step": 4301
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9772114594719895,
+      "learning_rate": 9.504106575634664e-08,
+      "loss": 0.2153,
+      "step": 4302
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8517040372283946,
+      "learning_rate": 9.435578044786009e-08,
+      "loss": 0.193,
+      "step": 4303
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8160717421172,
+      "learning_rate": 9.367295114770191e-08,
+      "loss": 0.1984,
+      "step": 4304
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8456698407669494,
+      "learning_rate": 9.299257819772811e-08,
+      "loss": 0.2153,
+      "step": 4305
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8534577623169042,
+      "learning_rate": 9.231466193856897e-08,
+      "loss": 0.2177,
+      "step": 4306
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8202323168155831,
+      "learning_rate": 9.163920270962134e-08,
+      "loss": 0.2075,
+      "step": 4307
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8251220105895755,
+      "learning_rate": 9.096620084905472e-08,
+      "loss": 0.2109,
+      "step": 4308
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.1210097300397406,
+      "learning_rate": 9.029565669380625e-08,
+      "loss": 0.2176,
+      "step": 4309
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8023759644061593,
+      "learning_rate": 8.96275705795846e-08,
+      "loss": 0.1946,
+      "step": 4310
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8287191528633663,
+      "learning_rate": 8.896194284086501e-08,
+      "loss": 0.1925,
+      "step": 4311
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8922149767246479,
+      "learning_rate": 8.829877381089536e-08,
+      "loss": 0.2077,
+      "step": 4312
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9489478979822509,
+      "learning_rate": 8.763806382169005e-08,
+      "loss": 0.1806,
+      "step": 4313
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8414835233178939,
+      "learning_rate": 8.697981320403337e-08,
+      "loss": 0.2173,
+      "step": 4314
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8267171834461201,
+      "learning_rate": 8.632402228747893e-08,
+      "loss": 0.2079,
+      "step": 4315
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8539201696362041,
+      "learning_rate": 8.56706914003469e-08,
+      "loss": 0.1866,
+      "step": 4316
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8001398422686491,
+      "learning_rate": 8.501982086972782e-08,
+      "loss": 0.192,
+      "step": 4317
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.4607263803038326,
+      "learning_rate": 8.437141102147883e-08,
+      "loss": 0.1828,
+      "step": 4318
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8324831629154482,
+      "learning_rate": 8.372546218022747e-08,
+      "loss": 0.2011,
+      "step": 4319
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.854682609731579,
+      "learning_rate": 8.308197466936619e-08,
+      "loss": 0.2141,
+      "step": 4320
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8486773988922032,
+      "learning_rate": 8.244094881105779e-08,
+      "loss": 0.2109,
+      "step": 4321
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8736114856156321,
+      "learning_rate": 8.18023849262306e-08,
+      "loss": 0.1909,
+      "step": 4322
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7865589218051244,
+      "learning_rate": 8.11662833345822e-08,
+      "loss": 0.1839,
+      "step": 4323
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.81810792101549,
+      "learning_rate": 8.05326443545762e-08,
+      "loss": 0.1981,
+      "step": 4324
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.844069657300875,
+      "learning_rate": 7.990146830344325e-08,
+      "loss": 0.202,
+      "step": 4325
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8144883807217171,
+      "learning_rate": 7.927275549718228e-08,
+      "loss": 0.1862,
+      "step": 4326
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7728147958767654,
+      "learning_rate": 7.864650625055703e-08,
+      "loss": 0.1836,
+      "step": 4327
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.9612295571904864,
+      "learning_rate": 7.802272087709951e-08,
+      "loss": 0.2044,
+      "step": 4328
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8356836385569563,
+      "learning_rate": 7.740139968910709e-08,
+      "loss": 0.1998,
+      "step": 4329
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8777331277334701,
+      "learning_rate": 7.678254299764431e-08,
+      "loss": 0.2045,
+      "step": 4330
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8525310420886059,
+      "learning_rate": 7.616615111254111e-08,
+      "loss": 0.2044,
+      "step": 4331
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7727176107623636,
+      "learning_rate": 7.555222434239395e-08,
+      "loss": 0.1842,
+      "step": 4332
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8145000154280653,
+      "learning_rate": 7.494076299456531e-08,
+      "loss": 0.2047,
+      "step": 4333
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8128185405178724,
+      "learning_rate": 7.433176737518255e-08,
+      "loss": 0.2108,
+      "step": 4334
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8127863879831895,
+      "learning_rate": 7.372523778913898e-08,
+      "loss": 0.2039,
+      "step": 4335
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8219240522996707,
+      "learning_rate": 7.312117454009337e-08,
+      "loss": 0.2009,
+      "step": 4336
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7650207218139149,
+      "learning_rate": 7.251957793046993e-08,
+      "loss": 0.1981,
+      "step": 4337
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.795454272679756,
+      "learning_rate": 7.192044826145772e-08,
+      "loss": 0.1906,
+      "step": 4338
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8211907835009088,
+      "learning_rate": 7.132378583301014e-08,
+      "loss": 0.2111,
+      "step": 4339
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.103928838727255,
+      "learning_rate": 7.072959094384713e-08,
+      "loss": 0.1905,
+      "step": 4340
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9291895245191503,
+      "learning_rate": 7.013786389145127e-08,
+      "loss": 0.1961,
+      "step": 4341
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.518681073829251,
+      "learning_rate": 6.954860497207006e-08,
+      "loss": 0.2206,
+      "step": 4342
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9257636765711496,
+      "learning_rate": 6.896181448071582e-08,
+      "loss": 0.2028,
+      "step": 4343
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7650209065429256,
+      "learning_rate": 6.837749271116578e-08,
+      "loss": 0.1984,
+      "step": 4344
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9701360696018885,
+      "learning_rate": 6.779563995595927e-08,
+      "loss": 0.2061,
+      "step": 4345
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8033079789573468,
+      "learning_rate": 6.721625650640163e-08,
+      "loss": 0.2155,
+      "step": 4346
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.164709890010626,
+      "learning_rate": 6.663934265255966e-08,
+      "loss": 0.1944,
+      "step": 4347
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8479445790735926,
+      "learning_rate": 6.606489868326571e-08,
+      "loss": 0.1926,
+      "step": 4348
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7825957798379245,
+      "learning_rate": 6.549292488611525e-08,
+      "loss": 0.1852,
+      "step": 4349
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8070708732230681,
+      "learning_rate": 6.49234215474659e-08,
+      "loss": 0.1916,
+      "step": 4350
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8346431486344151,
+      "learning_rate": 6.435638895243957e-08,
+      "loss": 0.2112,
+      "step": 4351
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7966583696967651,
+      "learning_rate": 6.379182738492085e-08,
+      "loss": 0.1903,
+      "step": 4352
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8800618399486471,
+      "learning_rate": 6.322973712755698e-08,
+      "loss": 0.2023,
+      "step": 4353
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8422961987954524,
+      "learning_rate": 6.267011846175896e-08,
+      "loss": 0.198,
+      "step": 4354
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9092056315517442,
+      "learning_rate": 6.211297166769825e-08,
+      "loss": 0.2115,
+      "step": 4355
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7424438241911018,
+      "learning_rate": 6.15582970243117e-08,
+      "loss": 0.2047,
+      "step": 4356
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8291793064368945,
+      "learning_rate": 6.100609480929553e-08,
+      "loss": 0.1848,
+      "step": 4357
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9030894330812148,
+      "learning_rate": 6.045636529911025e-08,
+      "loss": 0.2097,
+      "step": 4358
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8054910058779746,
+      "learning_rate": 5.990910876897682e-08,
+      "loss": 0.2039,
+      "step": 4359
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9197107665996948,
+      "learning_rate": 5.9364325492880514e-08,
+      "loss": 0.219,
+      "step": 4360
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8400750808456374,
+      "learning_rate": 5.882201574356483e-08,
+      "loss": 0.1995,
+      "step": 4361
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8504335989561778,
+      "learning_rate": 5.82821797925387e-08,
+      "loss": 0.2266,
+      "step": 4362
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.795391677499905,
+      "learning_rate": 5.7744817910069804e-08,
+      "loss": 0.2003,
+      "step": 4363
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7434215941655571,
+      "learning_rate": 5.720993036518796e-08,
+      "loss": 0.1888,
+      "step": 4364
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8406245762246916,
+      "learning_rate": 5.667751742568506e-08,
+      "loss": 0.218,
+      "step": 4365
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8446230725935892,
+      "learning_rate": 5.6147579358112324e-08,
+      "loss": 0.214,
+      "step": 4366
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.871832269659328,
+      "learning_rate": 5.56201164277842e-08,
+      "loss": 0.2083,
+      "step": 4367
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8696313818676698,
+      "learning_rate": 5.509512889877333e-08,
+      "loss": 0.2232,
+      "step": 4368
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8584047883741758,
+      "learning_rate": 5.457261703391614e-08,
+      "loss": 0.2076,
+      "step": 4369
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8620927015898214,
+      "learning_rate": 5.4052581094806713e-08,
+      "loss": 0.2181,
+      "step": 4370
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7818547185915923,
+      "learning_rate": 5.353502134180122e-08,
+      "loss": 0.1753,
+      "step": 4371
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.237938713410867,
+      "learning_rate": 5.301993803401517e-08,
+      "loss": 0.2101,
+      "step": 4372
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.966139195886675,
+      "learning_rate": 5.250733142932562e-08,
+      "loss": 0.2025,
+      "step": 4373
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8162919576069905,
+      "learning_rate": 5.19972017843684e-08,
+      "loss": 0.184,
+      "step": 4374
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.82634177197088,
+      "learning_rate": 5.148954935453976e-08,
+      "loss": 0.1924,
+      "step": 4375
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7857417149785898,
+      "learning_rate": 5.098437439399473e-08,
+      "loss": 0.1861,
+      "step": 4376
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8536544665012806,
+      "learning_rate": 5.0481677155650445e-08,
+      "loss": 0.2073,
+      "step": 4377
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8227354749479117,
+      "learning_rate": 4.998145789118114e-08,
+      "loss": 0.1926,
+      "step": 4378
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7919060333382815,
+      "learning_rate": 4.948371685102149e-08,
+      "loss": 0.2032,
+      "step": 4379
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8071215688192208,
+      "learning_rate": 4.898845428436494e-08,
+      "loss": 0.199,
+      "step": 4380
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8610340301688427,
+      "learning_rate": 4.849567043916481e-08,
+      "loss": 0.2188,
+      "step": 4381
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7477997196592773,
+      "learning_rate": 4.800536556213375e-08,
+      "loss": 0.1986,
+      "step": 4382
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8457055677785161,
+      "learning_rate": 4.751753989874153e-08,
+      "loss": 0.2225,
+      "step": 4383
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8831504117177361,
+      "learning_rate": 4.7032193693217765e-08,
+      "loss": 0.1833,
+      "step": 4384
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7624044296106454,
+      "learning_rate": 4.654932718855143e-08,
+      "loss": 0.1892,
+      "step": 4385
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9113363347009648,
+      "learning_rate": 4.60689406264897e-08,
+      "loss": 0.2078,
+      "step": 4386
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.2420284259359295,
+      "learning_rate": 4.559103424753742e-08,
+      "loss": 0.2121,
+      "step": 4387
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9484476598715104,
+      "learning_rate": 4.511560829095818e-08,
+      "loss": 0.2053,
+      "step": 4388
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.784575706234333,
+      "learning_rate": 4.4642662994773264e-08,
+      "loss": 0.1978,
+      "step": 4389
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7783120941504319,
+      "learning_rate": 4.417219859576383e-08,
+      "loss": 0.1967,
+      "step": 4390
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.0725438309536843,
+      "learning_rate": 4.370421532946589e-08,
+      "loss": 0.2054,
+      "step": 4391
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8554226078051486,
+      "learning_rate": 4.3238713430176494e-08,
+      "loss": 0.2237,
+      "step": 4392
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.811224358429652,
+      "learning_rate": 4.2775693130948094e-08,
+      "loss": 0.1971,
+      "step": 4393
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.174971809189537,
+      "learning_rate": 4.231515466359193e-08,
+      "loss": 0.1924,
+      "step": 4394
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7975115905815806,
+      "learning_rate": 4.1857098258676345e-08,
+      "loss": 0.2008,
+      "step": 4395
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8286631290413113,
+      "learning_rate": 4.140152414552678e-08,
+      "loss": 0.1909,
+      "step": 4396
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8293242376895925,
+      "learning_rate": 4.094843255222691e-08,
+      "loss": 0.2147,
+      "step": 4397
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.108098465227292,
+      "learning_rate": 4.0497823705615836e-08,
+      "loss": 0.2179,
+      "step": 4398
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8074684025359189,
+      "learning_rate": 4.004969783129087e-08,
+      "loss": 0.2037,
+      "step": 4399
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8721092216019799,
+      "learning_rate": 3.960405515360699e-08,
+      "loss": 0.1831,
+      "step": 4400
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.818918008586044,
+      "learning_rate": 3.9160895895673515e-08,
+      "loss": 0.1978,
+      "step": 4401
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.791372822265506,
+      "learning_rate": 3.8720220279359065e-08,
+      "loss": 0.1845,
+      "step": 4402
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8169743302469691,
+      "learning_rate": 3.828202852528717e-08,
+      "loss": 0.1937,
+      "step": 4403
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8533238194605183,
+      "learning_rate": 3.784632085283846e-08,
+      "loss": 0.1921,
+      "step": 4404
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.924561383047835,
+      "learning_rate": 3.741309748014954e-08,
+      "loss": 0.1984,
+      "step": 4405
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7911616168051684,
+      "learning_rate": 3.698235862411359e-08,
+      "loss": 0.202,
+      "step": 4406
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7968483227370201,
+      "learning_rate": 3.6554104500379775e-08,
+      "loss": 0.2163,
+      "step": 4407
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9209933878141395,
+      "learning_rate": 3.6128335323353804e-08,
+      "loss": 0.1921,
+      "step": 4408
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.186494785400701,
+      "learning_rate": 3.570505130619573e-08,
+      "loss": 0.1786,
+      "step": 4409
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8782966049609051,
+      "learning_rate": 3.528425266082325e-08,
+      "loss": 0.2055,
+      "step": 4410
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9099504330414767,
+      "learning_rate": 3.486593959790896e-08,
+      "loss": 0.2035,
+      "step": 4411
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8073580655163208,
+      "learning_rate": 3.445011232688089e-08,
+      "loss": 0.1939,
+      "step": 4412
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0384052913183162,
+      "learning_rate": 3.4036771055923066e-08,
+      "loss": 0.2178,
+      "step": 4413
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.77307431670331,
+      "learning_rate": 3.36259159919744e-08,
+      "loss": 0.1801,
+      "step": 4414
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0889092205442377,
+      "learning_rate": 3.321754734072924e-08,
+      "loss": 0.2229,
+      "step": 4415
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8524057486797181,
+      "learning_rate": 3.281166530663682e-08,
+      "loss": 0.1927,
+      "step": 4416
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.79746635308267,
+      "learning_rate": 3.240827009290237e-08,
+      "loss": 0.2111,
+      "step": 4417
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7554088797776188,
+      "learning_rate": 3.2007361901485455e-08,
+      "loss": 0.1934,
+      "step": 4418
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8388182424733743,
+      "learning_rate": 3.160894093309996e-08,
+      "loss": 0.1999,
+      "step": 4419
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7936603820544055,
+      "learning_rate": 3.121300738721578e-08,
+      "loss": 0.1919,
+      "step": 4420
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8246485210446494,
+      "learning_rate": 3.0819561462056555e-08,
+      "loss": 0.2017,
+      "step": 4421
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.87650979792615,
+      "learning_rate": 3.042860335460085e-08,
+      "loss": 0.236,
+      "step": 4422
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8100632623335381,
+      "learning_rate": 3.004013326058153e-08,
+      "loss": 0.2123,
+      "step": 4423
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8747087668836719,
+      "learning_rate": 2.9654151374485797e-08,
+      "loss": 0.2106,
+      "step": 4424
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8026426180070924,
+      "learning_rate": 2.92706578895563e-08,
+      "loss": 0.1904,
+      "step": 4425
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8580657804760349,
+      "learning_rate": 2.8889652997787233e-08,
+      "loss": 0.2001,
+      "step": 4426
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8405599426948956,
+      "learning_rate": 2.8511136889929902e-08,
+      "loss": 0.1998,
+      "step": 4427
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7979792716345043,
+      "learning_rate": 2.8135109755487723e-08,
+      "loss": 0.2004,
+      "step": 4428
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8795933514344532,
+      "learning_rate": 2.7761571782718433e-08,
+      "loss": 0.2235,
+      "step": 4429
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.77255150504307,
+      "learning_rate": 2.7390523158633552e-08,
+      "loss": 0.1816,
+      "step": 4430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8057765481605579,
+      "learning_rate": 2.7021964068998375e-08,
+      "loss": 0.1872,
+      "step": 4431
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8027176617930745,
+      "learning_rate": 2.6655894698331962e-08,
+      "loss": 0.1996,
+      "step": 4432
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7842527169086394,
+      "learning_rate": 2.629231522990716e-08,
+      "loss": 0.194,
+      "step": 4433
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7561336185435995,
+      "learning_rate": 2.593122584574892e-08,
+      "loss": 0.1821,
+      "step": 4434
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8375285824278015,
+      "learning_rate": 2.5572626726637074e-08,
+      "loss": 0.2128,
+      "step": 4435
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8026728526836475,
+      "learning_rate": 2.5216518052104122e-08,
+      "loss": 0.2059,
+      "step": 4436
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8965794404551357,
+      "learning_rate": 2.486290000043523e-08,
+      "loss": 0.2261,
+      "step": 4437
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8446646009618801,
+      "learning_rate": 2.4511772748669894e-08,
+      "loss": 0.1898,
+      "step": 4438
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8237818058534104,
+      "learning_rate": 2.4163136472599157e-08,
+      "loss": 0.2063,
+      "step": 4439
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8245545767605245,
+      "learning_rate": 2.381699134676785e-08,
+      "loss": 0.1881,
+      "step": 4440
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7777214312866056,
+      "learning_rate": 2.3473337544472897e-08,
+      "loss": 0.2068,
+      "step": 4441
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7755079907713476,
+      "learning_rate": 2.3132175237765564e-08,
+      "loss": 0.198,
+      "step": 4442
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.9512591301572402,
+      "learning_rate": 2.2793504597447003e-08,
+      "loss": 0.2041,
+      "step": 4443
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.256811696411824,
+      "learning_rate": 2.2457325793073803e-08,
+      "loss": 0.1982,
+      "step": 4444
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.9760880670290429,
+      "learning_rate": 2.2123638992953e-08,
+      "loss": 0.2111,
+      "step": 4445
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8423124210156124,
+      "learning_rate": 2.179244436414485e-08,
+      "loss": 0.2129,
+      "step": 4446
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7812774531780088,
+      "learning_rate": 2.146374207246227e-08,
+      "loss": 0.1952,
+      "step": 4447
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7819471189617627,
+      "learning_rate": 2.1137532282469176e-08,
+      "loss": 0.1902,
+      "step": 4448
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7601180417746587,
+      "learning_rate": 2.081381515748271e-08,
+      "loss": 0.1746,
+      "step": 4449
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7841768154969317,
+      "learning_rate": 2.0492590859571558e-08,
+      "loss": 0.1795,
+      "step": 4450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7802282829928464,
+      "learning_rate": 2.017385954955653e-08,
+      "loss": 0.2116,
+      "step": 4451
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.9401853021509832,
+      "learning_rate": 1.9857621387010528e-08,
+      "loss": 0.2213,
+      "step": 4452
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8598876358664671,
+      "learning_rate": 1.954387653025802e-08,
+      "loss": 0.219,
+      "step": 4453
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8517431608273662,
+      "learning_rate": 1.9232625136374472e-08,
+      "loss": 0.1841,
+      "step": 4454
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.1946922665283066,
+      "learning_rate": 1.8923867361189118e-08,
+      "loss": 0.2002,
+      "step": 4455
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.781312049707781,
+      "learning_rate": 1.8617603359279977e-08,
+      "loss": 0.1871,
+      "step": 4456
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8399116729510478,
+      "learning_rate": 1.8313833283978844e-08,
+      "loss": 0.2152,
+      "step": 4457
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7791344113195685,
+      "learning_rate": 1.8012557287367394e-08,
+      "loss": 0.2051,
+      "step": 4458
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.861026012581039,
+      "learning_rate": 1.7713775520279976e-08,
+      "loss": 0.197,
+      "step": 4459
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8708701468900847,
+      "learning_rate": 1.7417488132301375e-08,
+      "loss": 0.1902,
+      "step": 4460
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.1010290188785832,
+      "learning_rate": 1.712369527176738e-08,
+      "loss": 0.1905,
+      "step": 4461
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.886974370725903,
+      "learning_rate": 1.6832397085765893e-08,
+      "loss": 0.2035,
+      "step": 4462
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8006407574013145,
+      "learning_rate": 1.6543593720134142e-08,
+      "loss": 0.2102,
+      "step": 4463
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.85202047911503,
+      "learning_rate": 1.6257285319462578e-08,
+      "loss": 0.1963,
+      "step": 4464
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8463813208162998,
+      "learning_rate": 1.597347202709043e-08,
+      "loss": 0.2184,
+      "step": 4465
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8536022952660971,
+      "learning_rate": 1.5692153985109037e-08,
+      "loss": 0.2099,
+      "step": 4466
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7708282065822465,
+      "learning_rate": 1.541333133436018e-08,
+      "loss": 0.1919,
+      "step": 4467
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8282528205129762,
+      "learning_rate": 1.513700421443609e-08,
+      "loss": 0.2188,
+      "step": 4468
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9277205723553366,
+      "learning_rate": 1.4863172763680545e-08,
+      "loss": 0.1901,
+      "step": 4469
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8012648394286224,
+      "learning_rate": 1.4591837119186104e-08,
+      "loss": 0.1996,
+      "step": 4470
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8474780778478656,
+      "learning_rate": 1.4322997416797435e-08,
+      "loss": 0.1948,
+      "step": 4471
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9378269006777065,
+      "learning_rate": 1.4056653791108543e-08,
+      "loss": 0.2013,
+      "step": 4472
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9808071497211823,
+      "learning_rate": 1.379280637546443e-08,
+      "loss": 0.2073,
+      "step": 4473
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.823508755501204,
+      "learning_rate": 1.3531455301960539e-08,
+      "loss": 0.2022,
+      "step": 4474
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7652214059229656,
+      "learning_rate": 1.3272600701441651e-08,
+      "loss": 0.2049,
+      "step": 4475
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8618012457347418,
+      "learning_rate": 1.3016242703503545e-08,
+      "loss": 0.1937,
+      "step": 4476
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8257825224906502,
+      "learning_rate": 1.2762381436491888e-08,
+      "loss": 0.2163,
+      "step": 4477
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.0842927567347542,
+      "learning_rate": 1.2511017027501682e-08,
+      "loss": 0.2072,
+      "step": 4478
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7631124898644732,
+      "learning_rate": 1.226214960237837e-08,
+      "loss": 0.2023,
+      "step": 4479
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8683902528256007,
+      "learning_rate": 1.2015779285717844e-08,
+      "loss": 0.2136,
+      "step": 4480
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.4582238775506597,
+      "learning_rate": 1.1771906200864769e-08,
+      "loss": 0.2045,
+      "step": 4481
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8445527714970521,
+      "learning_rate": 1.1530530469914258e-08,
+      "loss": 0.2051,
+      "step": 4482
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8413663218206109,
+      "learning_rate": 1.1291652213710758e-08,
+      "loss": 0.2178,
+      "step": 4483
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8032125873819468,
+      "learning_rate": 1.1055271551849156e-08,
+      "loss": 0.1768,
+      "step": 4484
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9110276444935175,
+      "learning_rate": 1.0821388602673121e-08,
+      "loss": 0.2097,
+      "step": 4485
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9016738766641257,
+      "learning_rate": 1.0590003483275657e-08,
+      "loss": 0.2139,
+      "step": 4486
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8244080154284859,
+      "learning_rate": 1.0361116309499653e-08,
+      "loss": 0.2084,
+      "step": 4487
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9444782541291822,
+      "learning_rate": 1.0134727195937332e-08,
+      "loss": 0.1943,
+      "step": 4488
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.854699025822739,
+      "learning_rate": 9.91083625593081e-09,
+      "loss": 0.1983,
+      "step": 4489
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8720945843328476,
+      "learning_rate": 9.689443601570425e-09,
+      "loss": 0.1975,
+      "step": 4490
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9306189030840065,
+      "learning_rate": 9.470549343696401e-09,
+      "loss": 0.1948,
+      "step": 4491
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8099432203666295,
+      "learning_rate": 9.254153591898852e-09,
+      "loss": 0.2045,
+      "step": 4492
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7853863220651058,
+      "learning_rate": 9.04025645451445e-09,
+      "loss": 0.2084,
+      "step": 4493
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9226245721462727,
+      "learning_rate": 8.828858038632538e-09,
+      "loss": 0.201,
+      "step": 4494
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2341460918247351,
+      "learning_rate": 8.61995845008845e-09,
+      "loss": 0.1996,
+      "step": 4495
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7973149650580209,
+      "learning_rate": 8.413557793467974e-09,
+      "loss": 0.2101,
+      "step": 4496
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8738397288902738,
+      "learning_rate": 8.209656172105673e-09,
+      "loss": 0.1899,
+      "step": 4497
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.7412177579708981,
+      "learning_rate": 8.008253688084888e-09,
+      "loss": 0.2059,
+      "step": 4498
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8259607955695522,
+      "learning_rate": 7.80935044223663e-09,
+      "loss": 0.2057,
+      "step": 4499
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8262383303147334,
+      "learning_rate": 7.612946534143461e-09,
+      "loss": 0.1972,
+      "step": 4500
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7899917147229134,
+      "learning_rate": 7.4190420621333965e-09,
+      "loss": 0.191,
+      "step": 4501
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.980296064179857,
+      "learning_rate": 7.227637123285447e-09,
+      "loss": 0.2128,
+      "step": 4502
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7964214665396392,
+      "learning_rate": 7.038731813426292e-09,
+      "loss": 0.1856,
+      "step": 4503
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8220537303970679,
+      "learning_rate": 6.852326227130835e-09,
+      "loss": 0.2152,
+      "step": 4504
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8938133428626523,
+      "learning_rate": 6.668420457723312e-09,
+      "loss": 0.2247,
+      "step": 4505
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7369254948271431,
+      "learning_rate": 6.487014597275631e-09,
+      "loss": 0.1774,
+      "step": 4506
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8972113303856036,
+      "learning_rate": 6.308108736609031e-09,
+      "loss": 0.2236,
+      "step": 4507
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7942584797923552,
+      "learning_rate": 6.1317029652929734e-09,
+      "loss": 0.2059,
+      "step": 4508
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.91256877115628,
+      "learning_rate": 5.957797371644036e-09,
+      "loss": 0.2113,
+      "step": 4509
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.0610743408502665,
+      "learning_rate": 5.7863920427286834e-09,
+      "loss": 0.1952,
+      "step": 4510
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.897646962572014,
+      "learning_rate": 5.617487064360494e-09,
+      "loss": 0.2027,
+      "step": 4511
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8079938729488058,
+      "learning_rate": 5.45108252110127e-09,
+      "loss": 0.1993,
+      "step": 4512
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.902689249324902,
+      "learning_rate": 5.2871784962627015e-09,
+      "loss": 0.2263,
+      "step": 4513
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.800999156908289,
+      "learning_rate": 5.125775071901373e-09,
+      "loss": 0.2019,
+      "step": 4514
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7732529652596019,
+      "learning_rate": 4.966872328825978e-09,
+      "loss": 0.1993,
+      "step": 4515
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.1807788923271527,
+      "learning_rate": 4.810470346588992e-09,
+      "loss": 0.2065,
+      "step": 4516
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9355350742193872,
+      "learning_rate": 4.6565692034944435e-09,
+      "loss": 0.2049,
+      "step": 4517
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8137638282310965,
+      "learning_rate": 4.505168976592922e-09,
+      "loss": 0.1781,
+      "step": 4518
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8583244037169239,
+      "learning_rate": 4.356269741681573e-09,
+      "loss": 0.2098,
+      "step": 4519
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7710465895237573,
+      "learning_rate": 4.209871573307989e-09,
+      "loss": 0.1989,
+      "step": 4520
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8619539870150191,
+      "learning_rate": 4.065974544765761e-09,
+      "loss": 0.211,
+      "step": 4521
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.924724665565591,
+      "learning_rate": 3.924578728096707e-09,
+      "loss": 0.21,
+      "step": 4522
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8046236520577087,
+      "learning_rate": 3.785684194090866e-09,
+      "loss": 0.1961,
+      "step": 4523
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.3180574603356483,
+      "learning_rate": 3.6492910122859493e-09,
+      "loss": 0.2053,
+      "step": 4524
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7900451380554656,
+      "learning_rate": 3.515399250966223e-09,
+      "loss": 0.202,
+      "step": 4525
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8316349143722477,
+      "learning_rate": 3.3840089771658424e-09,
+      "loss": 0.2073,
+      "step": 4526
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8198708418054829,
+      "learning_rate": 3.2551202566644125e-09,
+      "loss": 0.2082,
+      "step": 4527
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8151551093442948,
+      "learning_rate": 3.1287331539903155e-09,
+      "loss": 0.2116,
+      "step": 4528
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9106345792999407,
+      "learning_rate": 3.004847732419047e-09,
+      "loss": 0.2004,
+      "step": 4529
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8668046445438777,
+      "learning_rate": 2.8834640539737723e-09,
+      "loss": 0.1951,
+      "step": 4530
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9860754574399874,
+      "learning_rate": 2.7645821794247684e-09,
+      "loss": 0.2075,
+      "step": 4531
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8185978742781929,
+      "learning_rate": 2.648202168291092e-09,
+      "loss": 0.1956,
+      "step": 4532
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8464050529434357,
+      "learning_rate": 2.534324078837802e-09,
+      "loss": 0.2095,
+      "step": 4533
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8111589679192016,
+      "learning_rate": 2.422947968078182e-09,
+      "loss": 0.2099,
+      "step": 4534
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8052195453161209,
+      "learning_rate": 2.314073891772073e-09,
+      "loss": 0.1953,
+      "step": 4535
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.1226016200431905,
+      "learning_rate": 2.2077019044280946e-09,
+      "loss": 0.2234,
+      "step": 4536
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.801892742283442,
+      "learning_rate": 2.1038320593003146e-09,
+      "loss": 0.2153,
+      "step": 4537
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8733342628870312,
+      "learning_rate": 2.002464408392135e-09,
+      "loss": 0.1944,
+      "step": 4538
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8499280584410499,
+      "learning_rate": 1.90359900245185e-09,
+      "loss": 0.1887,
+      "step": 4539
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8358625521048361,
+      "learning_rate": 1.8072358909781983e-09,
+      "loss": 0.2137,
+      "step": 4540
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8273873816215992,
+      "learning_rate": 1.7133751222137007e-09,
+      "loss": 0.2025,
+      "step": 4541
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.2817219616262665,
+      "learning_rate": 1.622016743150212e-09,
+      "loss": 0.2042,
+      "step": 4542
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.812444414803269,
+      "learning_rate": 1.5331607995267006e-09,
+      "loss": 0.2158,
+      "step": 4543
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7564673611632567,
+      "learning_rate": 1.4468073358286926e-09,
+      "loss": 0.1839,
+      "step": 4544
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.8813056574740219,
+      "learning_rate": 1.362956395288828e-09,
+      "loss": 0.2012,
+      "step": 4545
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8752466778759325,
+      "learning_rate": 1.2816080198868596e-09,
+      "loss": 0.1916,
+      "step": 4546
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.867187800362047,
+      "learning_rate": 1.202762250350209e-09,
+      "loss": 0.1973,
+      "step": 4547
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8957982052367516,
+      "learning_rate": 1.1264191261528557e-09,
+      "loss": 0.2111,
+      "step": 4548
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8953466818167085,
+      "learning_rate": 1.0525786855158926e-09,
+      "loss": 0.2098,
+      "step": 4549
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8249529837044098,
+      "learning_rate": 9.812409654075261e-10,
+      "loss": 0.217,
+      "step": 4550
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.1108775703626435,
+      "learning_rate": 9.124060015425206e-10,
+      "loss": 0.2095,
+      "step": 4551
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.770870388249719,
+      "learning_rate": 8.460738283838643e-10,
+      "loss": 0.1937,
+      "step": 4552
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8338945368353098,
+      "learning_rate": 7.82244479139993e-10,
+      "loss": 0.1862,
+      "step": 4553
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8934523568229983,
+      "learning_rate": 7.209179857675664e-10,
+      "loss": 0.2054,
+      "step": 4554
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.02351471367892,
+      "learning_rate": 6.620943789692469e-10,
+      "loss": 0.2044,
+      "step": 4555
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.9562333092916492,
+      "learning_rate": 6.057736881959209e-10,
+      "loss": 0.2266,
+      "step": 4556
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8784272316742409,
+      "learning_rate": 5.519559416433673e-10,
+      "loss": 0.2205,
+      "step": 4557
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8980639943442136,
+      "learning_rate": 5.006411662555888e-10,
+      "loss": 0.2074,
+      "step": 4558
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.0230131315188493,
+      "learning_rate": 4.518293877242563e-10,
+      "loss": 0.2043,
+      "step": 4559
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8513051502405525,
+      "learning_rate": 4.0552063048593384e-10,
+      "loss": 0.2014,
+      "step": 4560
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8316641711654165,
+      "learning_rate": 3.6171491772540865e-10,
+      "loss": 0.2263,
+      "step": 4561
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7621859370937432,
+      "learning_rate": 3.204122713740265e-10,
+      "loss": 0.1931,
+      "step": 4562
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7843969228367804,
+      "learning_rate": 2.816127121102463e-10,
+      "loss": 0.1842,
+      "step": 4563
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8377848791749765,
+      "learning_rate": 2.453162593585301e-10,
+      "loss": 0.2071,
+      "step": 4564
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7866713580071245,
+      "learning_rate": 2.1152293129100832e-10,
+      "loss": 0.1891,
+      "step": 4565
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8370003067041355,
+      "learning_rate": 1.8023274482636965e-10,
+      "loss": 0.2106,
+      "step": 4566
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.9111748984033715,
+      "learning_rate": 1.5144571562930587e-10,
+      "loss": 0.1996,
+      "step": 4567
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8067891836799083,
+      "learning_rate": 1.251618581127323e-10,
+      "loss": 0.2038,
+      "step": 4568
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7987528347245983,
+      "learning_rate": 1.013811854361224e-10,
+      "loss": 0.1995,
+      "step": 4569
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9338257141197195,
+      "learning_rate": 8.01037095038426e-11,
+      "loss": 0.2131,
+      "step": 4570
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.795738349622226,
+      "learning_rate": 6.132944096959304e-11,
+      "loss": 0.2048,
+      "step": 4571
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8054348002133643,
+      "learning_rate": 4.505838923252181e-11,
+      "loss": 0.1937,
+      "step": 4572
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8094854616597447,
+      "learning_rate": 3.129056243833528e-11,
+      "loss": 0.199,
+      "step": 4573
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7653238421851912,
+      "learning_rate": 2.0025967479853082e-11,
+      "loss": 0.1877,
+      "step": 4574
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9556375060294325,
+      "learning_rate": 1.126460999700818e-11,
+      "loss": 0.1875,
+      "step": 4575
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8035255480125802,
+      "learning_rate": 5.006494376291748e-12,
+      "loss": 0.2028,
+      "step": 4576
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7939659174841657,
+      "learning_rate": 1.2516237507531615e-12,
+      "loss": 0.2069,
+      "step": 4577
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9720981463508033,
+      "learning_rate": 0.0,
+      "loss": 0.2017,
+      "step": 4578
+    },
+    {
+      "epoch": 2.0,
+      "step": 4578,
+      "total_flos": 2116616889237504.0,
+      "train_loss": 0.253446443596356,
+      "train_runtime": 85432.5485,
+      "train_samples_per_second": 6.862,
+      "train_steps_per_second": 0.054
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4578,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "total_flos": 2116616889237504.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}