diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22798 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3252,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009225092250922509,
+      "grad_norm": 12.955489203853885,
+      "learning_rate": 0.0,
+      "loss": 1.4185,
+      "step": 1
+    },
+    {
+      "epoch": 0.0018450184501845018,
+      "grad_norm": 15.45613889769327,
+      "learning_rate": 1.5337423312883438e-07,
+      "loss": 1.5687,
+      "step": 2
+    },
+    {
+      "epoch": 0.0027675276752767526,
+      "grad_norm": 12.92656381552016,
+      "learning_rate": 3.0674846625766876e-07,
+      "loss": 1.4134,
+      "step": 3
+    },
+    {
+      "epoch": 0.0036900369003690036,
+      "grad_norm": 12.98532338060555,
+      "learning_rate": 4.601226993865031e-07,
+      "loss": 1.4595,
+      "step": 4
+    },
+    {
+      "epoch": 0.004612546125461255,
+      "grad_norm": 12.338984619051718,
+      "learning_rate": 6.134969325153375e-07,
+      "loss": 1.4031,
+      "step": 5
+    },
+    {
+      "epoch": 0.005535055350553505,
+      "grad_norm": 14.85148132005717,
+      "learning_rate": 7.668711656441718e-07,
+      "loss": 1.4832,
+      "step": 6
+    },
+    {
+      "epoch": 0.006457564575645757,
+      "grad_norm": 17.072025860539494,
+      "learning_rate": 9.202453987730062e-07,
+      "loss": 1.6873,
+      "step": 7
+    },
+    {
+      "epoch": 0.007380073800738007,
+      "grad_norm": 16.431366793672808,
+      "learning_rate": 1.0736196319018406e-06,
+      "loss": 1.6399,
+      "step": 8
+    },
+    {
+      "epoch": 0.008302583025830259,
+      "grad_norm": 12.014901856256495,
+      "learning_rate": 1.226993865030675e-06,
+      "loss": 1.3723,
+      "step": 9
+    },
+    {
+      "epoch": 0.00922509225092251,
+      "grad_norm": 11.848609763944259,
+      "learning_rate": 1.3803680981595093e-06,
+      "loss": 1.4094,
+      "step": 10
+    },
+    {
+      "epoch": 0.01014760147601476,
+      "grad_norm": 11.765507368785373,
+      "learning_rate": 1.5337423312883435e-06,
+      "loss": 1.3755,
+      "step": 11
+    },
+    {
+      "epoch": 0.01107011070110701,
+      "grad_norm": 10.590795739479944,
+      "learning_rate": 1.687116564417178e-06,
+      "loss": 1.279,
+      "step": 12
+    },
+    {
+      "epoch": 0.011992619926199263,
+      "grad_norm": 12.149753919483338,
+      "learning_rate": 1.8404907975460124e-06,
+      "loss": 1.4244,
+      "step": 13
+    },
+    {
+      "epoch": 0.012915129151291513,
+      "grad_norm": 10.6231739703075,
+      "learning_rate": 1.9938650306748465e-06,
+      "loss": 1.2602,
+      "step": 14
+    },
+    {
+      "epoch": 0.013837638376383764,
+      "grad_norm": 11.087376374846873,
+      "learning_rate": 2.147239263803681e-06,
+      "loss": 1.2945,
+      "step": 15
+    },
+    {
+      "epoch": 0.014760147601476014,
+      "grad_norm": 8.797053743582785,
+      "learning_rate": 2.3006134969325154e-06,
+      "loss": 0.9855,
+      "step": 16
+    },
+    {
+      "epoch": 0.015682656826568265,
+      "grad_norm": 9.222973996497377,
+      "learning_rate": 2.45398773006135e-06,
+      "loss": 0.9322,
+      "step": 17
+    },
+    {
+      "epoch": 0.016605166051660517,
+      "grad_norm": 8.741210685946207,
+      "learning_rate": 2.607361963190184e-06,
+      "loss": 0.8991,
+      "step": 18
+    },
+    {
+      "epoch": 0.017527675276752766,
+      "grad_norm": 7.784987898808541,
+      "learning_rate": 2.7607361963190186e-06,
+      "loss": 0.8354,
+      "step": 19
+    },
+    {
+      "epoch": 0.01845018450184502,
+      "grad_norm": 8.114903017788382,
+      "learning_rate": 2.914110429447853e-06,
+      "loss": 0.7483,
+      "step": 20
+    },
+    {
+      "epoch": 0.01937269372693727,
+      "grad_norm": 6.985942947652761,
+      "learning_rate": 3.067484662576687e-06,
+      "loss": 0.7405,
+      "step": 21
+    },
+    {
+      "epoch": 0.02029520295202952,
+      "grad_norm": 5.057560376950641,
+      "learning_rate": 3.2208588957055217e-06,
+      "loss": 0.523,
+      "step": 22
+    },
+    {
+      "epoch": 0.021217712177121772,
+      "grad_norm": 4.30867356704448,
+      "learning_rate": 3.374233128834356e-06,
+      "loss": 0.5054,
+      "step": 23
+    },
+    {
+      "epoch": 0.02214022140221402,
+      "grad_norm": 3.186678341527648,
+      "learning_rate": 3.52760736196319e-06,
+      "loss": 0.4204,
+      "step": 24
+    },
+    {
+      "epoch": 0.023062730627306273,
+      "grad_norm": 2.547181779371563,
+      "learning_rate": 3.680981595092025e-06,
+      "loss": 0.471,
+      "step": 25
+    },
+    {
+      "epoch": 0.023985239852398525,
+      "grad_norm": 1.4892578932038187,
+      "learning_rate": 3.834355828220859e-06,
+      "loss": 0.3852,
+      "step": 26
+    },
+    {
+      "epoch": 0.024907749077490774,
+      "grad_norm": 1.6034561918455752,
+      "learning_rate": 3.987730061349693e-06,
+      "loss": 0.447,
+      "step": 27
+    },
+    {
+      "epoch": 0.025830258302583026,
+      "grad_norm": 1.3040786609892756,
+      "learning_rate": 4.141104294478528e-06,
+      "loss": 0.4171,
+      "step": 28
+    },
+    {
+      "epoch": 0.026752767527675275,
+      "grad_norm": 1.5718209161660188,
+      "learning_rate": 4.294478527607362e-06,
+      "loss": 0.4146,
+      "step": 29
+    },
+    {
+      "epoch": 0.027675276752767528,
+      "grad_norm": 1.5226011146024725,
+      "learning_rate": 4.447852760736196e-06,
+      "loss": 0.3771,
+      "step": 30
+    },
+    {
+      "epoch": 0.02859778597785978,
+      "grad_norm": 1.2181738630849195,
+      "learning_rate": 4.601226993865031e-06,
+      "loss": 0.4243,
+      "step": 31
+    },
+    {
+      "epoch": 0.02952029520295203,
+      "grad_norm": 1.3702765242872075,
+      "learning_rate": 4.7546012269938654e-06,
+      "loss": 0.393,
+      "step": 32
+    },
+    {
+      "epoch": 0.03044280442804428,
+      "grad_norm": 1.066008129856778,
+      "learning_rate": 4.9079754601227e-06,
+      "loss": 0.3754,
+      "step": 33
+    },
+    {
+      "epoch": 0.03136531365313653,
+      "grad_norm": 0.9407582448390286,
+      "learning_rate": 5.061349693251534e-06,
+      "loss": 0.3492,
+      "step": 34
+    },
+    {
+      "epoch": 0.03228782287822878,
+      "grad_norm": 0.9599309117793499,
+      "learning_rate": 5.214723926380368e-06,
+      "loss": 0.3554,
+      "step": 35
+    },
+    {
+      "epoch": 0.033210332103321034,
+      "grad_norm": 1.0092310826139268,
+      "learning_rate": 5.368098159509203e-06,
+      "loss": 0.3804,
+      "step": 36
+    },
+    {
+      "epoch": 0.03413284132841329,
+      "grad_norm": 0.8244638212070418,
+      "learning_rate": 5.521472392638037e-06,
+      "loss": 0.34,
+      "step": 37
+    },
+    {
+      "epoch": 0.03505535055350553,
+      "grad_norm": 0.8247247080222685,
+      "learning_rate": 5.674846625766871e-06,
+      "loss": 0.3395,
+      "step": 38
+    },
+    {
+      "epoch": 0.035977859778597784,
+      "grad_norm": 0.8400106617176237,
+      "learning_rate": 5.828220858895706e-06,
+      "loss": 0.3418,
+      "step": 39
+    },
+    {
+      "epoch": 0.03690036900369004,
+      "grad_norm": 0.7727224893246134,
+      "learning_rate": 5.98159509202454e-06,
+      "loss": 0.3096,
+      "step": 40
+    },
+    {
+      "epoch": 0.03782287822878229,
+      "grad_norm": 0.9087099282488907,
+      "learning_rate": 6.134969325153374e-06,
+      "loss": 0.319,
+      "step": 41
+    },
+    {
+      "epoch": 0.03874538745387454,
+      "grad_norm": 0.9932295126911083,
+      "learning_rate": 6.288343558282209e-06,
+      "loss": 0.3221,
+      "step": 42
+    },
+    {
+      "epoch": 0.03966789667896679,
+      "grad_norm": 0.6995800325598962,
+      "learning_rate": 6.4417177914110434e-06,
+      "loss": 0.3066,
+      "step": 43
+    },
+    {
+      "epoch": 0.04059040590405904,
+      "grad_norm": 0.769126559231979,
+      "learning_rate": 6.595092024539877e-06,
+      "loss": 0.2997,
+      "step": 44
+    },
+    {
+      "epoch": 0.04151291512915129,
+      "grad_norm": 0.8854396311491329,
+      "learning_rate": 6.748466257668712e-06,
+      "loss": 0.308,
+      "step": 45
+    },
+    {
+      "epoch": 0.042435424354243544,
+      "grad_norm": 0.9106864581523603,
+      "learning_rate": 6.901840490797547e-06,
+      "loss": 0.3079,
+      "step": 46
+    },
+    {
+      "epoch": 0.043357933579335796,
+      "grad_norm": 0.9132279319467866,
+      "learning_rate": 7.05521472392638e-06,
+      "loss": 0.2851,
+      "step": 47
+    },
+    {
+      "epoch": 0.04428044280442804,
+      "grad_norm": 0.7097422801782368,
+      "learning_rate": 7.208588957055215e-06,
+      "loss": 0.3063,
+      "step": 48
+    },
+    {
+      "epoch": 0.045202952029520294,
+      "grad_norm": 0.8888301377889726,
+      "learning_rate": 7.36196319018405e-06,
+      "loss": 0.3064,
+      "step": 49
+    },
+    {
+      "epoch": 0.046125461254612546,
+      "grad_norm": 0.8326651435551397,
+      "learning_rate": 7.5153374233128836e-06,
+      "loss": 0.3381,
+      "step": 50
+    },
+    {
+      "epoch": 0.0470479704797048,
+      "grad_norm": 0.7823625505280065,
+      "learning_rate": 7.668711656441718e-06,
+      "loss": 0.2976,
+      "step": 51
+    },
+    {
+      "epoch": 0.04797047970479705,
+      "grad_norm": 0.7726039404102877,
+      "learning_rate": 7.822085889570554e-06,
+      "loss": 0.2933,
+      "step": 52
+    },
+    {
+      "epoch": 0.048892988929889296,
+      "grad_norm": 0.7610651536301617,
+      "learning_rate": 7.975460122699386e-06,
+      "loss": 0.2882,
+      "step": 53
+    },
+    {
+      "epoch": 0.04981549815498155,
+      "grad_norm": 1.5278867981943542,
+      "learning_rate": 8.128834355828221e-06,
+      "loss": 0.2891,
+      "step": 54
+    },
+    {
+      "epoch": 0.0507380073800738,
+      "grad_norm": 0.6933285259760709,
+      "learning_rate": 8.282208588957055e-06,
+      "loss": 0.2631,
+      "step": 55
+    },
+    {
+      "epoch": 0.05166051660516605,
+      "grad_norm": 0.8315242149931074,
+      "learning_rate": 8.435582822085889e-06,
+      "loss": 0.3194,
+      "step": 56
+    },
+    {
+      "epoch": 0.052583025830258305,
+      "grad_norm": 0.8014840042460718,
+      "learning_rate": 8.588957055214725e-06,
+      "loss": 0.327,
+      "step": 57
+    },
+    {
+      "epoch": 0.05350553505535055,
+      "grad_norm": 0.7542043085568613,
+      "learning_rate": 8.742331288343558e-06,
+      "loss": 0.2902,
+      "step": 58
+    },
+    {
+      "epoch": 0.0544280442804428,
+      "grad_norm": 0.8021745207785408,
+      "learning_rate": 8.895705521472392e-06,
+      "loss": 0.2884,
+      "step": 59
+    },
+    {
+      "epoch": 0.055350553505535055,
+      "grad_norm": 0.7935265214617545,
+      "learning_rate": 9.049079754601228e-06,
+      "loss": 0.2938,
+      "step": 60
+    },
+    {
+      "epoch": 0.05627306273062731,
+      "grad_norm": 0.7945059096047613,
+      "learning_rate": 9.202453987730062e-06,
+      "loss": 0.3169,
+      "step": 61
+    },
+    {
+      "epoch": 0.05719557195571956,
+      "grad_norm": 0.6844890656867241,
+      "learning_rate": 9.355828220858897e-06,
+      "loss": 0.2787,
+      "step": 62
+    },
+    {
+      "epoch": 0.058118081180811805,
+      "grad_norm": 0.6759065932541566,
+      "learning_rate": 9.509202453987731e-06,
+      "loss": 0.3023,
+      "step": 63
+    },
+    {
+      "epoch": 0.05904059040590406,
+      "grad_norm": 0.7068063979581338,
+      "learning_rate": 9.662576687116565e-06,
+      "loss": 0.2876,
+      "step": 64
+    },
+    {
+      "epoch": 0.05996309963099631,
+      "grad_norm": 0.8019897278737798,
+      "learning_rate": 9.8159509202454e-06,
+      "loss": 0.2697,
+      "step": 65
+    },
+    {
+      "epoch": 0.06088560885608856,
+      "grad_norm": 0.7851152567049443,
+      "learning_rate": 9.969325153374232e-06,
+      "loss": 0.3117,
+      "step": 66
+    },
+    {
+      "epoch": 0.061808118081180814,
+      "grad_norm": 0.8443069282860475,
+      "learning_rate": 1.0122699386503068e-05,
+      "loss": 0.3379,
+      "step": 67
+    },
+    {
+      "epoch": 0.06273062730627306,
+      "grad_norm": 0.6844260224687486,
+      "learning_rate": 1.0276073619631903e-05,
+      "loss": 0.2864,
+      "step": 68
+    },
+    {
+      "epoch": 0.06365313653136531,
+      "grad_norm": 0.6764987660001486,
+      "learning_rate": 1.0429447852760736e-05,
+      "loss": 0.2632,
+      "step": 69
+    },
+    {
+      "epoch": 0.06457564575645756,
+      "grad_norm": 0.6735304741507071,
+      "learning_rate": 1.0582822085889571e-05,
+      "loss": 0.2689,
+      "step": 70
+    },
+    {
+      "epoch": 0.06549815498154982,
+      "grad_norm": 0.9313719592818234,
+      "learning_rate": 1.0736196319018407e-05,
+      "loss": 0.2969,
+      "step": 71
+    },
+    {
+      "epoch": 0.06642066420664207,
+      "grad_norm": 0.7047477227444052,
+      "learning_rate": 1.0889570552147239e-05,
+      "loss": 0.2802,
+      "step": 72
+    },
+    {
+      "epoch": 0.06734317343173432,
+      "grad_norm": 0.7511066786918772,
+      "learning_rate": 1.1042944785276074e-05,
+      "loss": 0.2942,
+      "step": 73
+    },
+    {
+      "epoch": 0.06826568265682657,
+      "grad_norm": 0.7753312122163863,
+      "learning_rate": 1.119631901840491e-05,
+      "loss": 0.3039,
+      "step": 74
+    },
+    {
+      "epoch": 0.06918819188191883,
+      "grad_norm": 0.7898944701574543,
+      "learning_rate": 1.1349693251533742e-05,
+      "loss": 0.3168,
+      "step": 75
+    },
+    {
+      "epoch": 0.07011070110701106,
+      "grad_norm": 0.7608450558743302,
+      "learning_rate": 1.1503067484662577e-05,
+      "loss": 0.3159,
+      "step": 76
+    },
+    {
+      "epoch": 0.07103321033210332,
+      "grad_norm": 0.6851187018726027,
+      "learning_rate": 1.1656441717791411e-05,
+      "loss": 0.302,
+      "step": 77
+    },
+    {
+      "epoch": 0.07195571955719557,
+      "grad_norm": 0.7009343522823712,
+      "learning_rate": 1.1809815950920245e-05,
+      "loss": 0.2858,
+      "step": 78
+    },
+    {
+      "epoch": 0.07287822878228782,
+      "grad_norm": 0.7106703686063037,
+      "learning_rate": 1.196319018404908e-05,
+      "loss": 0.2833,
+      "step": 79
+    },
+    {
+      "epoch": 0.07380073800738007,
+      "grad_norm": 0.6860438191626305,
+      "learning_rate": 1.2116564417177914e-05,
+      "loss": 0.2836,
+      "step": 80
+    },
+    {
+      "epoch": 0.07472324723247233,
+      "grad_norm": 0.6265893388163131,
+      "learning_rate": 1.2269938650306748e-05,
+      "loss": 0.274,
+      "step": 81
+    },
+    {
+      "epoch": 0.07564575645756458,
+      "grad_norm": 0.7216235699208722,
+      "learning_rate": 1.2423312883435584e-05,
+      "loss": 0.2814,
+      "step": 82
+    },
+    {
+      "epoch": 0.07656826568265683,
+      "grad_norm": 0.6406089500922234,
+      "learning_rate": 1.2576687116564418e-05,
+      "loss": 0.2927,
+      "step": 83
+    },
+    {
+      "epoch": 0.07749077490774908,
+      "grad_norm": 0.7439376335847888,
+      "learning_rate": 1.2730061349693251e-05,
+      "loss": 0.2664,
+      "step": 84
+    },
+    {
+      "epoch": 0.07841328413284133,
+      "grad_norm": 0.738376581767653,
+      "learning_rate": 1.2883435582822087e-05,
+      "loss": 0.2694,
+      "step": 85
+    },
+    {
+      "epoch": 0.07933579335793357,
+      "grad_norm": 0.8230580812584041,
+      "learning_rate": 1.303680981595092e-05,
+      "loss": 0.3308,
+      "step": 86
+    },
+    {
+      "epoch": 0.08025830258302583,
+      "grad_norm": 0.7665711642660273,
+      "learning_rate": 1.3190184049079754e-05,
+      "loss": 0.279,
+      "step": 87
+    },
+    {
+      "epoch": 0.08118081180811808,
+      "grad_norm": 1.0937041621239993,
+      "learning_rate": 1.334355828220859e-05,
+      "loss": 0.316,
+      "step": 88
+    },
+    {
+      "epoch": 0.08210332103321033,
+      "grad_norm": 0.8111056435505324,
+      "learning_rate": 1.3496932515337424e-05,
+      "loss": 0.298,
+      "step": 89
+    },
+    {
+      "epoch": 0.08302583025830258,
+      "grad_norm": 0.7327882215407213,
+      "learning_rate": 1.3650306748466258e-05,
+      "loss": 0.3052,
+      "step": 90
+    },
+    {
+      "epoch": 0.08394833948339483,
+      "grad_norm": 0.7183449677066613,
+      "learning_rate": 1.3803680981595093e-05,
+      "loss": 0.2996,
+      "step": 91
+    },
+    {
+      "epoch": 0.08487084870848709,
+      "grad_norm": 0.7227952480990556,
+      "learning_rate": 1.3957055214723927e-05,
+      "loss": 0.31,
+      "step": 92
+    },
+    {
+      "epoch": 0.08579335793357934,
+      "grad_norm": 0.6773404497705654,
+      "learning_rate": 1.411042944785276e-05,
+      "loss": 0.2641,
+      "step": 93
+    },
+    {
+      "epoch": 0.08671586715867159,
+      "grad_norm": 0.7237883572445348,
+      "learning_rate": 1.4263803680981596e-05,
+      "loss": 0.28,
+      "step": 94
+    },
+    {
+      "epoch": 0.08763837638376384,
+      "grad_norm": 0.6642562423013476,
+      "learning_rate": 1.441717791411043e-05,
+      "loss": 0.2477,
+      "step": 95
+    },
+    {
+      "epoch": 0.08856088560885608,
+      "grad_norm": 0.6417741591931398,
+      "learning_rate": 1.4570552147239264e-05,
+      "loss": 0.2697,
+      "step": 96
+    },
+    {
+      "epoch": 0.08948339483394833,
+      "grad_norm": 0.8172595640272041,
+      "learning_rate": 1.47239263803681e-05,
+      "loss": 0.2883,
+      "step": 97
+    },
+    {
+      "epoch": 0.09040590405904059,
+      "grad_norm": 0.8387869401006764,
+      "learning_rate": 1.4877300613496933e-05,
+      "loss": 0.2861,
+      "step": 98
+    },
+    {
+      "epoch": 0.09132841328413284,
+      "grad_norm": 0.7425167902118797,
+      "learning_rate": 1.5030674846625767e-05,
+      "loss": 0.2932,
+      "step": 99
+    },
+    {
+      "epoch": 0.09225092250922509,
+      "grad_norm": 0.7529530576926313,
+      "learning_rate": 1.5184049079754603e-05,
+      "loss": 0.2732,
+      "step": 100
+    },
+    {
+      "epoch": 0.09317343173431734,
+      "grad_norm": 0.7394624338802493,
+      "learning_rate": 1.5337423312883436e-05,
+      "loss": 0.2962,
+      "step": 101
+    },
+    {
+      "epoch": 0.0940959409594096,
+      "grad_norm": 0.9937576615108098,
+      "learning_rate": 1.549079754601227e-05,
+      "loss": 0.3119,
+      "step": 102
+    },
+    {
+      "epoch": 0.09501845018450185,
+      "grad_norm": 0.7799933355610369,
+      "learning_rate": 1.5644171779141108e-05,
+      "loss": 0.2782,
+      "step": 103
+    },
+    {
+      "epoch": 0.0959409594095941,
+      "grad_norm": 0.6363177894137105,
+      "learning_rate": 1.579754601226994e-05,
+      "loss": 0.2607,
+      "step": 104
+    },
+    {
+      "epoch": 0.09686346863468635,
+      "grad_norm": 0.7861983695816259,
+      "learning_rate": 1.5950920245398772e-05,
+      "loss": 0.2857,
+      "step": 105
+    },
+    {
+      "epoch": 0.09778597785977859,
+      "grad_norm": 0.7498718582416898,
+      "learning_rate": 1.6104294478527606e-05,
+      "loss": 0.2839,
+      "step": 106
+    },
+    {
+      "epoch": 0.09870848708487084,
+      "grad_norm": 0.7855761683296874,
+      "learning_rate": 1.6257668711656443e-05,
+      "loss": 0.3017,
+      "step": 107
+    },
+    {
+      "epoch": 0.0996309963099631,
+      "grad_norm": 0.7242477830142006,
+      "learning_rate": 1.6411042944785277e-05,
+      "loss": 0.263,
+      "step": 108
+    },
+    {
+      "epoch": 0.10055350553505535,
+      "grad_norm": 0.734020068542186,
+      "learning_rate": 1.656441717791411e-05,
+      "loss": 0.3022,
+      "step": 109
+    },
+    {
+      "epoch": 0.1014760147601476,
+      "grad_norm": 0.6281436660404887,
+      "learning_rate": 1.6717791411042948e-05,
+      "loss": 0.3031,
+      "step": 110
+    },
+    {
+      "epoch": 0.10239852398523985,
+      "grad_norm": 0.6664367279847719,
+      "learning_rate": 1.6871165644171778e-05,
+      "loss": 0.2709,
+      "step": 111
+    },
+    {
+      "epoch": 0.1033210332103321,
+      "grad_norm": 0.8705405825083281,
+      "learning_rate": 1.7024539877300612e-05,
+      "loss": 0.3019,
+      "step": 112
+    },
+    {
+      "epoch": 0.10424354243542436,
+      "grad_norm": 0.7455378186922866,
+      "learning_rate": 1.717791411042945e-05,
+      "loss": 0.3137,
+      "step": 113
+    },
+    {
+      "epoch": 0.10516605166051661,
+      "grad_norm": 0.7445813146349921,
+      "learning_rate": 1.7331288343558283e-05,
+      "loss": 0.2933,
+      "step": 114
+    },
+    {
+      "epoch": 0.10608856088560886,
+      "grad_norm": 0.7649673992573114,
+      "learning_rate": 1.7484662576687117e-05,
+      "loss": 0.3064,
+      "step": 115
+    },
+    {
+      "epoch": 0.1070110701107011,
+      "grad_norm": 0.6673700237310717,
+      "learning_rate": 1.7638036809815954e-05,
+      "loss": 0.2859,
+      "step": 116
+    },
+    {
+      "epoch": 0.10793357933579335,
+      "grad_norm": 0.7228195303864201,
+      "learning_rate": 1.7791411042944784e-05,
+      "loss": 0.2604,
+      "step": 117
+    },
+    {
+      "epoch": 0.1088560885608856,
+      "grad_norm": 0.6731996318063017,
+      "learning_rate": 1.7944785276073618e-05,
+      "loss": 0.2868,
+      "step": 118
+    },
+    {
+      "epoch": 0.10977859778597786,
+      "grad_norm": 0.7037771197992196,
+      "learning_rate": 1.8098159509202455e-05,
+      "loss": 0.254,
+      "step": 119
+    },
+    {
+      "epoch": 0.11070110701107011,
+      "grad_norm": 0.893157319529324,
+      "learning_rate": 1.825153374233129e-05,
+      "loss": 0.3371,
+      "step": 120
+    },
+    {
+      "epoch": 0.11162361623616236,
+      "grad_norm": 0.7109039573324791,
+      "learning_rate": 1.8404907975460123e-05,
+      "loss": 0.2896,
+      "step": 121
+    },
+    {
+      "epoch": 0.11254612546125461,
+      "grad_norm": 0.694703757975936,
+      "learning_rate": 1.855828220858896e-05,
+      "loss": 0.3052,
+      "step": 122
+    },
+    {
+      "epoch": 0.11346863468634687,
+      "grad_norm": 0.7069453047379004,
+      "learning_rate": 1.8711656441717794e-05,
+      "loss": 0.2659,
+      "step": 123
+    },
+    {
+      "epoch": 0.11439114391143912,
+      "grad_norm": 0.7668929032786836,
+      "learning_rate": 1.8865030674846625e-05,
+      "loss": 0.3011,
+      "step": 124
+    },
+    {
+      "epoch": 0.11531365313653137,
+      "grad_norm": 0.7800112789686109,
+      "learning_rate": 1.9018404907975462e-05,
+      "loss": 0.3154,
+      "step": 125
+    },
+    {
+      "epoch": 0.11623616236162361,
+      "grad_norm": 0.6656177080819063,
+      "learning_rate": 1.9171779141104296e-05,
+      "loss": 0.2746,
+      "step": 126
+    },
+    {
+      "epoch": 0.11715867158671586,
+      "grad_norm": 0.6439698740687717,
+      "learning_rate": 1.932515337423313e-05,
+      "loss": 0.2348,
+      "step": 127
+    },
+    {
+      "epoch": 0.11808118081180811,
+      "grad_norm": 0.72561609461152,
+      "learning_rate": 1.9478527607361967e-05,
+      "loss": 0.2767,
+      "step": 128
+    },
+    {
+      "epoch": 0.11900369003690037,
+      "grad_norm": 0.6902197277603564,
+      "learning_rate": 1.96319018404908e-05,
+      "loss": 0.2771,
+      "step": 129
+    },
+    {
+      "epoch": 0.11992619926199262,
+      "grad_norm": 0.693463471274204,
+      "learning_rate": 1.978527607361963e-05,
+      "loss": 0.2956,
+      "step": 130
+    },
+    {
+      "epoch": 0.12084870848708487,
+      "grad_norm": 0.6930227234795218,
+      "learning_rate": 1.9938650306748465e-05,
+      "loss": 0.2901,
+      "step": 131
+    },
+    {
+      "epoch": 0.12177121771217712,
+      "grad_norm": 0.6499770864096507,
+      "learning_rate": 2.0092024539877302e-05,
+      "loss": 0.2628,
+      "step": 132
+    },
+    {
+      "epoch": 0.12269372693726938,
+      "grad_norm": 0.6294074861857037,
+      "learning_rate": 2.0245398773006136e-05,
+      "loss": 0.2844,
+      "step": 133
+    },
+    {
+      "epoch": 0.12361623616236163,
+      "grad_norm": 0.7055663181910957,
+      "learning_rate": 2.039877300613497e-05,
+      "loss": 0.2824,
+      "step": 134
+    },
+    {
+      "epoch": 0.12453874538745388,
+      "grad_norm": 0.6464646932599035,
+      "learning_rate": 2.0552147239263807e-05,
+      "loss": 0.287,
+      "step": 135
+    },
+    {
+      "epoch": 0.12546125461254612,
+      "grad_norm": 0.6453636704346579,
+      "learning_rate": 2.0705521472392637e-05,
+      "loss": 0.2768,
+      "step": 136
+    },
+    {
+      "epoch": 0.12638376383763839,
+      "grad_norm": 0.6707939485094121,
+      "learning_rate": 2.085889570552147e-05,
+      "loss": 0.3016,
+      "step": 137
+    },
+    {
+      "epoch": 0.12730627306273062,
+      "grad_norm": 0.6481600913545116,
+      "learning_rate": 2.1012269938650308e-05,
+      "loss": 0.303,
+      "step": 138
+    },
+    {
+      "epoch": 0.1282287822878229,
+      "grad_norm": 0.6358781587661787,
+      "learning_rate": 2.1165644171779142e-05,
+      "loss": 0.2668,
+      "step": 139
+    },
+    {
+      "epoch": 0.12915129151291513,
+      "grad_norm": 0.6091204160554227,
+      "learning_rate": 2.1319018404907976e-05,
+      "loss": 0.2717,
+      "step": 140
+    },
+    {
+      "epoch": 0.13007380073800737,
+      "grad_norm": 0.6483601411774305,
+      "learning_rate": 2.1472392638036813e-05,
+      "loss": 0.2676,
+      "step": 141
+    },
+    {
+      "epoch": 0.13099630996309963,
+      "grad_norm": 0.6569203594746024,
+      "learning_rate": 2.1625766871165647e-05,
+      "loss": 0.2896,
+      "step": 142
+    },
+    {
+      "epoch": 0.13191881918819187,
+      "grad_norm": 0.6524579068823675,
+      "learning_rate": 2.1779141104294477e-05,
+      "loss": 0.287,
+      "step": 143
+    },
+    {
+      "epoch": 0.13284132841328414,
+      "grad_norm": 0.6686273011266631,
+      "learning_rate": 2.1932515337423315e-05,
+      "loss": 0.3011,
+      "step": 144
+    },
+    {
+      "epoch": 0.13376383763837638,
+      "grad_norm": 0.6425281862252585,
+      "learning_rate": 2.208588957055215e-05,
+      "loss": 0.2974,
+      "step": 145
+    },
+    {
+      "epoch": 0.13468634686346864,
+      "grad_norm": 0.6601519038414725,
+      "learning_rate": 2.2239263803680982e-05,
+      "loss": 0.287,
+      "step": 146
+    },
+    {
+      "epoch": 0.13560885608856088,
+      "grad_norm": 0.6003217899565634,
+      "learning_rate": 2.239263803680982e-05,
+      "loss": 0.2787,
+      "step": 147
+    },
+    {
+      "epoch": 0.13653136531365315,
+      "grad_norm": 0.7129866882084958,
+      "learning_rate": 2.2546012269938653e-05,
+      "loss": 0.3,
+      "step": 148
+    },
+    {
+      "epoch": 0.13745387453874539,
+      "grad_norm": 0.7644906844783096,
+      "learning_rate": 2.2699386503067484e-05,
+      "loss": 0.284,
+      "step": 149
+    },
+    {
+      "epoch": 0.13837638376383765,
+      "grad_norm": 0.7232617216110071,
+      "learning_rate": 2.285276073619632e-05,
+      "loss": 0.2606,
+      "step": 150
+    },
+    {
+      "epoch": 0.1392988929889299,
+      "grad_norm": 0.6829835249098904,
+      "learning_rate": 2.3006134969325155e-05,
+      "loss": 0.2909,
+      "step": 151
+    },
+    {
+      "epoch": 0.14022140221402213,
+      "grad_norm": 0.7576335079961564,
+      "learning_rate": 2.315950920245399e-05,
+      "loss": 0.2974,
+      "step": 152
+    },
+    {
+      "epoch": 0.1411439114391144,
+      "grad_norm": 0.6914655420763508,
+      "learning_rate": 2.3312883435582822e-05,
+      "loss": 0.2603,
+      "step": 153
+    },
+    {
+      "epoch": 0.14206642066420663,
+      "grad_norm": 0.6160441569790551,
+      "learning_rate": 2.346625766871166e-05,
+      "loss": 0.255,
+      "step": 154
+    },
+    {
+      "epoch": 0.1429889298892989,
+      "grad_norm": 0.6639332944719932,
+      "learning_rate": 2.361963190184049e-05,
+      "loss": 0.2553,
+      "step": 155
+    },
+    {
+      "epoch": 0.14391143911439114,
+      "grad_norm": 0.6751721350055936,
+      "learning_rate": 2.3773006134969324e-05,
+      "loss": 0.2785,
+      "step": 156
+    },
+    {
+      "epoch": 0.1448339483394834,
+      "grad_norm": 0.6773724192208138,
+      "learning_rate": 2.392638036809816e-05,
+      "loss": 0.2874,
+      "step": 157
+    },
+    {
+      "epoch": 0.14575645756457564,
+      "grad_norm": 0.7490573812895311,
+      "learning_rate": 2.4079754601226995e-05,
+      "loss": 0.3087,
+      "step": 158
+    },
+    {
+      "epoch": 0.1466789667896679,
+      "grad_norm": 0.6953581548517865,
+      "learning_rate": 2.423312883435583e-05,
+      "loss": 0.3007,
+      "step": 159
+    },
+    {
+      "epoch": 0.14760147601476015,
+      "grad_norm": 0.6499092512394959,
+      "learning_rate": 2.4386503067484666e-05,
+      "loss": 0.2694,
+      "step": 160
+    },
+    {
+      "epoch": 0.14852398523985239,
+      "grad_norm": 0.6977503740590972,
+      "learning_rate": 2.4539877300613496e-05,
+      "loss": 0.2922,
+      "step": 161
+    },
+    {
+      "epoch": 0.14944649446494465,
+      "grad_norm": 0.6671277451323265,
+      "learning_rate": 2.469325153374233e-05,
+      "loss": 0.2679,
+      "step": 162
+    },
+    {
+      "epoch": 0.1503690036900369,
+      "grad_norm": 0.6289487835271185,
+      "learning_rate": 2.4846625766871167e-05,
+      "loss": 0.2914,
+      "step": 163
+    },
+    {
+      "epoch": 0.15129151291512916,
+      "grad_norm": 0.6130378431780416,
+      "learning_rate": 2.5e-05,
+      "loss": 0.277,
+      "step": 164
+    },
+    {
+      "epoch": 0.1522140221402214,
+      "grad_norm": 0.6893998165404203,
+      "learning_rate": 2.5153374233128835e-05,
+      "loss": 0.2689,
+      "step": 165
+    },
+    {
+      "epoch": 0.15313653136531366,
+      "grad_norm": 0.6431972368587816,
+      "learning_rate": 2.530674846625767e-05,
+      "loss": 0.2588,
+      "step": 166
+    },
+    {
+      "epoch": 0.1540590405904059,
+      "grad_norm": 0.7442869589662978,
+      "learning_rate": 2.5460122699386503e-05,
+      "loss": 0.2891,
+      "step": 167
+    },
+    {
+      "epoch": 0.15498154981549817,
+      "grad_norm": 0.6513268252137925,
+      "learning_rate": 2.561349693251534e-05,
+      "loss": 0.3177,
+      "step": 168
+    },
+    {
+      "epoch": 0.1559040590405904,
+      "grad_norm": 0.7489389936139911,
+      "learning_rate": 2.5766871165644174e-05,
+      "loss": 0.2742,
+      "step": 169
+    },
+    {
+      "epoch": 0.15682656826568267,
+      "grad_norm": 0.6932053862394971,
+      "learning_rate": 2.5920245398773008e-05,
+      "loss": 0.2731,
+      "step": 170
+    },
+    {
+      "epoch": 0.1577490774907749,
+      "grad_norm": 0.708245071323059,
+      "learning_rate": 2.607361963190184e-05,
+      "loss": 0.2862,
+      "step": 171
+    },
+    {
+      "epoch": 0.15867158671586715,
+      "grad_norm": 0.6060615887381059,
+      "learning_rate": 2.6226993865030675e-05,
+      "loss": 0.2653,
+      "step": 172
+    },
+    {
+      "epoch": 0.1595940959409594,
+      "grad_norm": 0.726730037504758,
+      "learning_rate": 2.638036809815951e-05,
+      "loss": 0.2977,
+      "step": 173
+    },
+    {
+      "epoch": 0.16051660516605165,
+      "grad_norm": 0.6593317139268164,
+      "learning_rate": 2.6533742331288346e-05,
+      "loss": 0.2813,
+      "step": 174
+    },
+    {
+      "epoch": 0.16143911439114392,
+      "grad_norm": 0.6632254325238431,
+      "learning_rate": 2.668711656441718e-05,
+      "loss": 0.2797,
+      "step": 175
+    },
+    {
+      "epoch": 0.16236162361623616,
+      "grad_norm": 0.6922954978124917,
+      "learning_rate": 2.6840490797546014e-05,
+      "loss": 0.2825,
+      "step": 176
+    },
+    {
+      "epoch": 0.16328413284132842,
+      "grad_norm": 0.6667884515269213,
+      "learning_rate": 2.6993865030674848e-05,
+      "loss": 0.28,
+      "step": 177
+    },
+    {
+      "epoch": 0.16420664206642066,
+      "grad_norm": 0.6277808142578968,
+      "learning_rate": 2.714723926380368e-05,
+      "loss": 0.2862,
+      "step": 178
+    },
+    {
+      "epoch": 0.16512915129151293,
+      "grad_norm": 0.616336601612985,
+      "learning_rate": 2.7300613496932515e-05,
+      "loss": 0.2885,
+      "step": 179
+    },
+    {
+      "epoch": 0.16605166051660517,
+      "grad_norm": 0.6723978733223379,
+      "learning_rate": 2.7453987730061353e-05,
+      "loss": 0.2868,
+      "step": 180
+    },
+    {
+      "epoch": 0.1669741697416974,
+      "grad_norm": 0.6133837567397908,
+      "learning_rate": 2.7607361963190186e-05,
+      "loss": 0.272,
+      "step": 181
+    },
+    {
+      "epoch": 0.16789667896678967,
+      "grad_norm": 0.5228829672328914,
+      "learning_rate": 2.776073619631902e-05,
+      "loss": 0.2436,
+      "step": 182
+    },
+    {
+      "epoch": 0.1688191881918819,
+      "grad_norm": 0.6538390530329945,
+      "learning_rate": 2.7914110429447854e-05,
+      "loss": 0.2729,
+      "step": 183
+    },
+    {
+      "epoch": 0.16974169741697417,
+      "grad_norm": 0.7257778344514123,
+      "learning_rate": 2.8067484662576688e-05,
+      "loss": 0.2804,
+      "step": 184
+    },
+    {
+      "epoch": 0.1706642066420664,
+      "grad_norm": 0.602334945645816,
+      "learning_rate": 2.822085889570552e-05,
+      "loss": 0.2771,
+      "step": 185
+    },
+    {
+      "epoch": 0.17158671586715868,
+      "grad_norm": 0.6731703893298568,
+      "learning_rate": 2.837423312883436e-05,
+      "loss": 0.2643,
+      "step": 186
+    },
+    {
+      "epoch": 0.17250922509225092,
+      "grad_norm": 0.7561284769218707,
+      "learning_rate": 2.8527607361963193e-05,
+      "loss": 0.2931,
+      "step": 187
+    },
+    {
+      "epoch": 0.17343173431734318,
+      "grad_norm": 0.5856064222603531,
+      "learning_rate": 2.8680981595092026e-05,
+      "loss": 0.2506,
+      "step": 188
+    },
+    {
+      "epoch": 0.17435424354243542,
+      "grad_norm": 0.7093532995617934,
+      "learning_rate": 2.883435582822086e-05,
+      "loss": 0.2721,
+      "step": 189
+    },
+    {
+      "epoch": 0.1752767527675277,
+      "grad_norm": 0.6182706972943631,
+      "learning_rate": 2.8987730061349694e-05,
+      "loss": 0.2295,
+      "step": 190
+    },
+    {
+      "epoch": 0.17619926199261993,
+      "grad_norm": 0.6029101223776,
+      "learning_rate": 2.9141104294478528e-05,
+      "loss": 0.2719,
+      "step": 191
+    },
+    {
+      "epoch": 0.17712177121771217,
+      "grad_norm": 0.5449248192061773,
+      "learning_rate": 2.9294478527607362e-05,
+      "loss": 0.2528,
+      "step": 192
+    },
+    {
+      "epoch": 0.17804428044280443,
+      "grad_norm": 0.71239166097286,
+      "learning_rate": 2.94478527607362e-05,
+      "loss": 0.3122,
+      "step": 193
+    },
+    {
+      "epoch": 0.17896678966789667,
+      "grad_norm": 0.7133992819416917,
+      "learning_rate": 2.9601226993865033e-05,
+      "loss": 0.2744,
+      "step": 194
+    },
+    {
+      "epoch": 0.17988929889298894,
+      "grad_norm": 0.5886007480549781,
+      "learning_rate": 2.9754601226993867e-05,
+      "loss": 0.2749,
+      "step": 195
+    },
+    {
+      "epoch": 0.18081180811808117,
+      "grad_norm": 0.5833471661434647,
+      "learning_rate": 2.99079754601227e-05,
+      "loss": 0.2564,
+      "step": 196
+    },
+    {
+      "epoch": 0.18173431734317344,
+      "grad_norm": 0.576357377285588,
+      "learning_rate": 3.0061349693251534e-05,
+      "loss": 0.2868,
+      "step": 197
+    },
+    {
+      "epoch": 0.18265682656826568,
+      "grad_norm": 0.68431739062317,
+      "learning_rate": 3.0214723926380368e-05,
+      "loss": 0.3071,
+      "step": 198
+    },
+    {
+      "epoch": 0.18357933579335795,
+      "grad_norm": 0.6726685646776212,
+      "learning_rate": 3.0368098159509205e-05,
+      "loss": 0.3044,
+      "step": 199
+    },
+    {
+      "epoch": 0.18450184501845018,
+      "grad_norm": 0.6380774098448085,
+      "learning_rate": 3.052147239263804e-05,
+      "loss": 0.2919,
+      "step": 200
+    },
+    {
+      "epoch": 0.18542435424354242,
+      "grad_norm": 0.6378258091381934,
+      "learning_rate": 3.067484662576687e-05,
+      "loss": 0.2668,
+      "step": 201
+    },
+    {
+      "epoch": 0.1863468634686347,
+      "grad_norm": 0.602025433101351,
+      "learning_rate": 3.0828220858895703e-05,
+      "loss": 0.2844,
+      "step": 202
+    },
+    {
+      "epoch": 0.18726937269372693,
+      "grad_norm": 0.6017459414288094,
+      "learning_rate": 3.098159509202454e-05,
+      "loss": 0.3055,
+      "step": 203
+    },
+    {
+      "epoch": 0.1881918819188192,
+      "grad_norm": 0.6633309714534229,
+      "learning_rate": 3.113496932515337e-05,
+      "loss": 0.2985,
+      "step": 204
+    },
+    {
+      "epoch": 0.18911439114391143,
+      "grad_norm": 0.8175145037559346,
+      "learning_rate": 3.1288343558282215e-05,
+      "loss": 0.2697,
+      "step": 205
+    },
+    {
+      "epoch": 0.1900369003690037,
+      "grad_norm": 0.7648529536365435,
+      "learning_rate": 3.1441717791411045e-05,
+      "loss": 0.254,
+      "step": 206
+    },
+    {
+      "epoch": 0.19095940959409594,
+      "grad_norm": 0.6193845502232096,
+      "learning_rate": 3.159509202453988e-05,
+      "loss": 0.2812,
+      "step": 207
+    },
+    {
+      "epoch": 0.1918819188191882,
+      "grad_norm": 0.6029512076439135,
+      "learning_rate": 3.174846625766871e-05,
+      "loss": 0.2945,
+      "step": 208
+    },
+    {
+      "epoch": 0.19280442804428044,
+      "grad_norm": 0.5592254753004365,
+      "learning_rate": 3.1901840490797544e-05,
+      "loss": 0.2651,
+      "step": 209
+    },
+    {
+      "epoch": 0.1937269372693727,
+      "grad_norm": 0.5710111832264806,
+      "learning_rate": 3.205521472392638e-05,
+      "loss": 0.2609,
+      "step": 210
+    },
+    {
+      "epoch": 0.19464944649446494,
+      "grad_norm": 0.6118237476493752,
+      "learning_rate": 3.220858895705521e-05,
+      "loss": 0.2735,
+      "step": 211
+    },
+    {
+      "epoch": 0.19557195571955718,
+      "grad_norm": 0.6586039796536416,
+      "learning_rate": 3.2361963190184055e-05,
+      "loss": 0.2912,
+      "step": 212
+    },
+    {
+      "epoch": 0.19649446494464945,
+      "grad_norm": 0.7370528603672779,
+      "learning_rate": 3.2515337423312886e-05,
+      "loss": 0.3192,
+      "step": 213
+    },
+    {
+      "epoch": 0.1974169741697417,
+      "grad_norm": 0.6274437717194886,
+      "learning_rate": 3.266871165644172e-05,
+      "loss": 0.2956,
+      "step": 214
+    },
+    {
+      "epoch": 0.19833948339483395,
+      "grad_norm": 0.6532944820897233,
+      "learning_rate": 3.282208588957055e-05,
+      "loss": 0.2816,
+      "step": 215
+    },
+    {
+      "epoch": 0.1992619926199262,
+      "grad_norm": 0.7051377169549334,
+      "learning_rate": 3.2975460122699384e-05,
+      "loss": 0.2919,
+      "step": 216
+    },
+    {
+      "epoch": 0.20018450184501846,
+      "grad_norm": 0.6131283518507381,
+      "learning_rate": 3.312883435582822e-05,
+      "loss": 0.3063,
+      "step": 217
+    },
+    {
+      "epoch": 0.2011070110701107,
+      "grad_norm": 0.5643760380208747,
+      "learning_rate": 3.328220858895706e-05,
+      "loss": 0.2767,
+      "step": 218
+    },
+    {
+      "epoch": 0.20202952029520296,
+      "grad_norm": 0.5360618443762574,
+      "learning_rate": 3.3435582822085895e-05,
+      "loss": 0.2605,
+      "step": 219
+    },
+    {
+      "epoch": 0.2029520295202952,
+      "grad_norm": 0.6345677981765382,
+      "learning_rate": 3.3588957055214726e-05,
+      "loss": 0.3078,
+      "step": 220
+    },
+    {
+      "epoch": 0.20387453874538744,
+      "grad_norm": 0.699130888028388,
+      "learning_rate": 3.3742331288343556e-05,
+      "loss": 0.2986,
+      "step": 221
+    },
+    {
+      "epoch": 0.2047970479704797,
+      "grad_norm": 0.6583162616034626,
+      "learning_rate": 3.3895705521472393e-05,
+      "loss": 0.3105,
+      "step": 222
+    },
+    {
+      "epoch": 0.20571955719557194,
+      "grad_norm": 0.6133101318835855,
+      "learning_rate": 3.4049079754601224e-05,
+      "loss": 0.2785,
+      "step": 223
+    },
+    {
+      "epoch": 0.2066420664206642,
+      "grad_norm": 0.5312778105673021,
+      "learning_rate": 3.420245398773007e-05,
+      "loss": 0.2745,
+      "step": 224
+    },
+    {
+      "epoch": 0.20756457564575645,
+      "grad_norm": 0.6083100303550006,
+      "learning_rate": 3.43558282208589e-05,
+      "loss": 0.2718,
+      "step": 225
+    },
+    {
+      "epoch": 0.20848708487084872,
+      "grad_norm": 0.6655989531281572,
+      "learning_rate": 3.4509202453987735e-05,
+      "loss": 0.3118,
+      "step": 226
+    },
+    {
+      "epoch": 0.20940959409594095,
+      "grad_norm": 0.6528009026939667,
+      "learning_rate": 3.4662576687116566e-05,
+      "loss": 0.2962,
+      "step": 227
+    },
+    {
+      "epoch": 0.21033210332103322,
+      "grad_norm": 0.6638162654461088,
+      "learning_rate": 3.4815950920245396e-05,
+      "loss": 0.3234,
+      "step": 228
+    },
+    {
+      "epoch": 0.21125461254612546,
+      "grad_norm": 0.6010271418513191,
+      "learning_rate": 3.4969325153374234e-05,
+      "loss": 0.252,
+      "step": 229
+    },
+    {
+      "epoch": 0.21217712177121772,
+      "grad_norm": 0.5670469197127859,
+      "learning_rate": 3.512269938650307e-05,
+      "loss": 0.2859,
+      "step": 230
+    },
+    {
+      "epoch": 0.21309963099630996,
+      "grad_norm": 0.5359328549287473,
+      "learning_rate": 3.527607361963191e-05,
+      "loss": 0.2844,
+      "step": 231
+    },
+    {
+      "epoch": 0.2140221402214022,
+      "grad_norm": 0.5901864103525112,
+      "learning_rate": 3.542944785276074e-05,
+      "loss": 0.2813,
+      "step": 232
+    },
+    {
+      "epoch": 0.21494464944649447,
+      "grad_norm": 0.5238540986476823,
+      "learning_rate": 3.558282208588957e-05,
+      "loss": 0.2548,
+      "step": 233
+    },
+    {
+      "epoch": 0.2158671586715867,
+      "grad_norm": 0.7105076419864915,
+      "learning_rate": 3.5736196319018406e-05,
+      "loss": 0.3286,
+      "step": 234
+    },
+    {
+      "epoch": 0.21678966789667897,
+      "grad_norm": 0.6690175862737519,
+      "learning_rate": 3.5889570552147236e-05,
+      "loss": 0.2773,
+      "step": 235
+    },
+    {
+      "epoch": 0.2177121771217712,
+      "grad_norm": 0.6574628031652334,
+      "learning_rate": 3.6042944785276074e-05,
+      "loss": 0.3139,
+      "step": 236
+    },
+    {
+      "epoch": 0.21863468634686348,
+      "grad_norm": 0.5693984692687943,
+      "learning_rate": 3.619631901840491e-05,
+      "loss": 0.2981,
+      "step": 237
+    },
+    {
+      "epoch": 0.21955719557195572,
+      "grad_norm": 0.6371295925172662,
+      "learning_rate": 3.634969325153375e-05,
+      "loss": 0.2946,
+      "step": 238
+    },
+    {
+      "epoch": 0.22047970479704798,
+      "grad_norm": 0.6273102976578564,
+      "learning_rate": 3.650306748466258e-05,
+      "loss": 0.2981,
+      "step": 239
+    },
+    {
+      "epoch": 0.22140221402214022,
+      "grad_norm": 0.6016146989383563,
+      "learning_rate": 3.665644171779141e-05,
+      "loss": 0.2823,
+      "step": 240
+    },
+    {
+      "epoch": 0.22232472324723246,
+      "grad_norm": 0.5506463175407609,
+      "learning_rate": 3.6809815950920246e-05,
+      "loss": 0.2766,
+      "step": 241
+    },
+    {
+      "epoch": 0.22324723247232472,
+      "grad_norm": 0.6144790265354202,
+      "learning_rate": 3.696319018404908e-05,
+      "loss": 0.2818,
+      "step": 242
+    },
+    {
+      "epoch": 0.22416974169741696,
+      "grad_norm": 0.5410857080611227,
+      "learning_rate": 3.711656441717792e-05,
+      "loss": 0.3034,
+      "step": 243
+    },
+    {
+      "epoch": 0.22509225092250923,
+      "grad_norm": 0.7090071961863476,
+      "learning_rate": 3.726993865030675e-05,
+      "loss": 0.3308,
+      "step": 244
+    },
+    {
+      "epoch": 0.22601476014760147,
+      "grad_norm": 0.5817986077508676,
+      "learning_rate": 3.742331288343559e-05,
+      "loss": 0.2839,
+      "step": 245
+    },
+    {
+      "epoch": 0.22693726937269373,
+      "grad_norm": 0.5378634317218598,
+      "learning_rate": 3.757668711656442e-05,
+      "loss": 0.2614,
+      "step": 246
+    },
+    {
+      "epoch": 0.22785977859778597,
+      "grad_norm": 0.7089287955511329,
+      "learning_rate": 3.773006134969325e-05,
+      "loss": 0.3018,
+      "step": 247
+    },
+    {
+      "epoch": 0.22878228782287824,
+      "grad_norm": 0.5790229588961053,
+      "learning_rate": 3.7883435582822086e-05,
+      "loss": 0.2742,
+      "step": 248
+    },
+    {
+      "epoch": 0.22970479704797048,
+      "grad_norm": 0.5226751178463883,
+      "learning_rate": 3.8036809815950924e-05,
+      "loss": 0.3015,
+      "step": 249
+    },
+    {
+      "epoch": 0.23062730627306274,
+      "grad_norm": 0.5585042684687114,
+      "learning_rate": 3.819018404907976e-05,
+      "loss": 0.2803,
+      "step": 250
+    },
+    {
+      "epoch": 0.23154981549815498,
+      "grad_norm": 0.5266474842246636,
+      "learning_rate": 3.834355828220859e-05,
+      "loss": 0.2964,
+      "step": 251
+    },
+    {
+      "epoch": 0.23247232472324722,
+      "grad_norm": 0.5773690989302118,
+      "learning_rate": 3.849693251533742e-05,
+      "loss": 0.2986,
+      "step": 252
+    },
+    {
+      "epoch": 0.2333948339483395,
+      "grad_norm": 0.48670222947233316,
+      "learning_rate": 3.865030674846626e-05,
+      "loss": 0.2531,
+      "step": 253
+    },
+    {
+      "epoch": 0.23431734317343172,
+      "grad_norm": 0.6421764020323035,
+      "learning_rate": 3.880368098159509e-05,
+      "loss": 0.2806,
+      "step": 254
+    },
+    {
+      "epoch": 0.235239852398524,
+      "grad_norm": 0.5922553489710686,
+      "learning_rate": 3.895705521472393e-05,
+      "loss": 0.2713,
+      "step": 255
+    },
+    {
+      "epoch": 0.23616236162361623,
+      "grad_norm": 0.6324457275724104,
+      "learning_rate": 3.9110429447852764e-05,
+      "loss": 0.3018,
+      "step": 256
+    },
+    {
+      "epoch": 0.2370848708487085,
+      "grad_norm": 0.571837105622298,
+      "learning_rate": 3.92638036809816e-05,
+      "loss": 0.3122,
+      "step": 257
+    },
+    {
+      "epoch": 0.23800738007380073,
+      "grad_norm": 0.677008364241822,
+      "learning_rate": 3.941717791411043e-05,
+      "loss": 0.2748,
+      "step": 258
+    },
+    {
+      "epoch": 0.238929889298893,
+      "grad_norm": 0.5516744368754908,
+      "learning_rate": 3.957055214723926e-05,
+      "loss": 0.3112,
+      "step": 259
+    },
+    {
+      "epoch": 0.23985239852398524,
+      "grad_norm": 0.5635734178386571,
+      "learning_rate": 3.97239263803681e-05,
+      "loss": 0.2418,
+      "step": 260
+    },
+    {
+      "epoch": 0.24077490774907748,
+      "grad_norm": 0.579042615236464,
+      "learning_rate": 3.987730061349693e-05,
+      "loss": 0.2678,
+      "step": 261
+    },
+    {
+      "epoch": 0.24169741697416974,
+      "grad_norm": 0.6679946243134188,
+      "learning_rate": 4.0030674846625773e-05,
+      "loss": 0.2858,
+      "step": 262
+    },
+    {
+      "epoch": 0.24261992619926198,
+      "grad_norm": 0.5964573605703893,
+      "learning_rate": 4.0184049079754604e-05,
+      "loss": 0.2849,
+      "step": 263
+    },
+    {
+      "epoch": 0.24354243542435425,
+      "grad_norm": 0.5530986279028629,
+      "learning_rate": 4.033742331288344e-05,
+      "loss": 0.2878,
+      "step": 264
+    },
+    {
+      "epoch": 0.2444649446494465,
+      "grad_norm": 0.5728949927481534,
+      "learning_rate": 4.049079754601227e-05,
+      "loss": 0.2683,
+      "step": 265
+    },
+    {
+      "epoch": 0.24538745387453875,
+      "grad_norm": 0.5613742323890516,
+      "learning_rate": 4.06441717791411e-05,
+      "loss": 0.2786,
+      "step": 266
+    },
+    {
+      "epoch": 0.246309963099631,
+      "grad_norm": 0.5054081322613757,
+      "learning_rate": 4.079754601226994e-05,
+      "loss": 0.2916,
+      "step": 267
+    },
+    {
+      "epoch": 0.24723247232472326,
+      "grad_norm": 0.5177290916182169,
+      "learning_rate": 4.0950920245398776e-05,
+      "loss": 0.2751,
+      "step": 268
+    },
+    {
+      "epoch": 0.2481549815498155,
+      "grad_norm": 0.49608484582705986,
+      "learning_rate": 4.1104294478527614e-05,
+      "loss": 0.2657,
+      "step": 269
+    },
+    {
+      "epoch": 0.24907749077490776,
+      "grad_norm": 0.5951411165866146,
+      "learning_rate": 4.1257668711656444e-05,
+      "loss": 0.312,
+      "step": 270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5778971143375377,
+      "learning_rate": 4.1411042944785274e-05,
+      "loss": 0.2479,
+      "step": 271
+    },
+    {
+      "epoch": 0.25092250922509224,
+      "grad_norm": 0.5363308179443078,
+      "learning_rate": 4.156441717791411e-05,
+      "loss": 0.282,
+      "step": 272
+    },
+    {
+      "epoch": 0.2518450184501845,
+      "grad_norm": 0.5690998644749469,
+      "learning_rate": 4.171779141104294e-05,
+      "loss": 0.3032,
+      "step": 273
+    },
+    {
+      "epoch": 0.25276752767527677,
+      "grad_norm": 0.591776250914278,
+      "learning_rate": 4.1871165644171786e-05,
+      "loss": 0.3127,
+      "step": 274
+    },
+    {
+      "epoch": 0.253690036900369,
+      "grad_norm": 0.6059632015040894,
+      "learning_rate": 4.2024539877300617e-05,
+      "loss": 0.3166,
+      "step": 275
+    },
+    {
+      "epoch": 0.25461254612546125,
+      "grad_norm": 0.6083637647170296,
+      "learning_rate": 4.2177914110429454e-05,
+      "loss": 0.2654,
+      "step": 276
+    },
+    {
+      "epoch": 0.2555350553505535,
+      "grad_norm": 0.5890457273173466,
+      "learning_rate": 4.2331288343558284e-05,
+      "loss": 0.2591,
+      "step": 277
+    },
+    {
+      "epoch": 0.2564575645756458,
+      "grad_norm": 0.4944776942353346,
+      "learning_rate": 4.2484662576687115e-05,
+      "loss": 0.2517,
+      "step": 278
+    },
+    {
+      "epoch": 0.257380073800738,
+      "grad_norm": 0.6643790613200776,
+      "learning_rate": 4.263803680981595e-05,
+      "loss": 0.3287,
+      "step": 279
+    },
+    {
+      "epoch": 0.25830258302583026,
+      "grad_norm": 0.5781262612195905,
+      "learning_rate": 4.279141104294479e-05,
+      "loss": 0.2908,
+      "step": 280
+    },
+    {
+      "epoch": 0.2592250922509225,
+      "grad_norm": 0.5079966726432265,
+      "learning_rate": 4.2944785276073626e-05,
+      "loss": 0.2437,
+      "step": 281
+    },
+    {
+      "epoch": 0.26014760147601473,
+      "grad_norm": 0.6008854116248558,
+      "learning_rate": 4.309815950920246e-05,
+      "loss": 0.2783,
+      "step": 282
+    },
+    {
+      "epoch": 0.261070110701107,
+      "grad_norm": 0.5490644205236153,
+      "learning_rate": 4.3251533742331294e-05,
+      "loss": 0.2773,
+      "step": 283
+    },
+    {
+      "epoch": 0.26199261992619927,
+      "grad_norm": 0.5990005074423055,
+      "learning_rate": 4.3404907975460124e-05,
+      "loss": 0.342,
+      "step": 284
+    },
+    {
+      "epoch": 0.2629151291512915,
+      "grad_norm": 0.6264105081239723,
+      "learning_rate": 4.3558282208588955e-05,
+      "loss": 0.306,
+      "step": 285
+    },
+    {
+      "epoch": 0.26383763837638374,
+      "grad_norm": 0.5033413509882185,
+      "learning_rate": 4.371165644171779e-05,
+      "loss": 0.26,
+      "step": 286
+    },
+    {
+      "epoch": 0.26476014760147604,
+      "grad_norm": 0.554351447814042,
+      "learning_rate": 4.386503067484663e-05,
+      "loss": 0.2854,
+      "step": 287
+    },
+    {
+      "epoch": 0.2656826568265683,
+      "grad_norm": 0.5302773165225984,
+      "learning_rate": 4.4018404907975466e-05,
+      "loss": 0.2562,
+      "step": 288
+    },
+    {
+      "epoch": 0.2666051660516605,
+      "grad_norm": 0.5183141955611247,
+      "learning_rate": 4.41717791411043e-05,
+      "loss": 0.2559,
+      "step": 289
+    },
+    {
+      "epoch": 0.26752767527675275,
+      "grad_norm": 0.5409284590210018,
+      "learning_rate": 4.432515337423313e-05,
+      "loss": 0.2563,
+      "step": 290
+    },
+    {
+      "epoch": 0.26845018450184505,
+      "grad_norm": 0.6393051011337837,
+      "learning_rate": 4.4478527607361964e-05,
+      "loss": 0.3165,
+      "step": 291
+    },
+    {
+      "epoch": 0.2693726937269373,
+      "grad_norm": 0.5559185337854181,
+      "learning_rate": 4.4631901840490795e-05,
+      "loss": 0.3145,
+      "step": 292
+    },
+    {
+      "epoch": 0.2702952029520295,
+      "grad_norm": 0.48707012990996873,
+      "learning_rate": 4.478527607361964e-05,
+      "loss": 0.286,
+      "step": 293
+    },
+    {
+      "epoch": 0.27121771217712176,
+      "grad_norm": 1.1765119695816575,
+      "learning_rate": 4.493865030674847e-05,
+      "loss": 0.2925,
+      "step": 294
+    },
+    {
+      "epoch": 0.272140221402214,
+      "grad_norm": 0.5092749730753052,
+      "learning_rate": 4.5092024539877307e-05,
+      "loss": 0.2678,
+      "step": 295
+    },
+    {
+      "epoch": 0.2730627306273063,
+      "grad_norm": 0.49642035968887865,
+      "learning_rate": 4.524539877300614e-05,
+      "loss": 0.2772,
+      "step": 296
+    },
+    {
+      "epoch": 0.27398523985239853,
+      "grad_norm": 0.5680854452902957,
+      "learning_rate": 4.539877300613497e-05,
+      "loss": 0.2835,
+      "step": 297
+    },
+    {
+      "epoch": 0.27490774907749077,
+      "grad_norm": 0.6033918944567603,
+      "learning_rate": 4.5552147239263805e-05,
+      "loss": 0.3041,
+      "step": 298
+    },
+    {
+      "epoch": 0.275830258302583,
+      "grad_norm": 0.5920736549312533,
+      "learning_rate": 4.570552147239264e-05,
+      "loss": 0.3011,
+      "step": 299
+    },
+    {
+      "epoch": 0.2767527675276753,
+      "grad_norm": 0.5570819697177297,
+      "learning_rate": 4.585889570552148e-05,
+      "loss": 0.2795,
+      "step": 300
+    },
+    {
+      "epoch": 0.27767527675276754,
+      "grad_norm": 0.5521443698108845,
+      "learning_rate": 4.601226993865031e-05,
+      "loss": 0.2877,
+      "step": 301
+    },
+    {
+      "epoch": 0.2785977859778598,
+      "grad_norm": 0.6494127895380299,
+      "learning_rate": 4.616564417177914e-05,
+      "loss": 0.3129,
+      "step": 302
+    },
+    {
+      "epoch": 0.279520295202952,
+      "grad_norm": 0.5642988518110355,
+      "learning_rate": 4.631901840490798e-05,
+      "loss": 0.2724,
+      "step": 303
+    },
+    {
+      "epoch": 0.28044280442804426,
+      "grad_norm": 0.5641545725751617,
+      "learning_rate": 4.647239263803681e-05,
+      "loss": 0.2747,
+      "step": 304
+    },
+    {
+      "epoch": 0.28136531365313655,
+      "grad_norm": 0.651919658466901,
+      "learning_rate": 4.6625766871165645e-05,
+      "loss": 0.297,
+      "step": 305
+    },
+    {
+      "epoch": 0.2822878228782288,
+      "grad_norm": 0.5718174535396574,
+      "learning_rate": 4.677914110429448e-05,
+      "loss": 0.3027,
+      "step": 306
+    },
+    {
+      "epoch": 0.283210332103321,
+      "grad_norm": 0.48701493995012424,
+      "learning_rate": 4.693251533742332e-05,
+      "loss": 0.278,
+      "step": 307
+    },
+    {
+      "epoch": 0.28413284132841327,
+      "grad_norm": 0.5284403071949798,
+      "learning_rate": 4.708588957055215e-05,
+      "loss": 0.3056,
+      "step": 308
+    },
+    {
+      "epoch": 0.28505535055350556,
+      "grad_norm": 0.5913958628137351,
+      "learning_rate": 4.723926380368098e-05,
+      "loss": 0.303,
+      "step": 309
+    },
+    {
+      "epoch": 0.2859778597785978,
+      "grad_norm": 0.4965955407068457,
+      "learning_rate": 4.739263803680982e-05,
+      "loss": 0.2499,
+      "step": 310
+    },
+    {
+      "epoch": 0.28690036900369004,
+      "grad_norm": 0.5326185286236886,
+      "learning_rate": 4.754601226993865e-05,
+      "loss": 0.2947,
+      "step": 311
+    },
+    {
+      "epoch": 0.2878228782287823,
+      "grad_norm": 0.5123442651860254,
+      "learning_rate": 4.769938650306749e-05,
+      "loss": 0.2888,
+      "step": 312
+    },
+    {
+      "epoch": 0.2887453874538745,
+      "grad_norm": 0.5538925573268508,
+      "learning_rate": 4.785276073619632e-05,
+      "loss": 0.2939,
+      "step": 313
+    },
+    {
+      "epoch": 0.2896678966789668,
+      "grad_norm": 0.5218442693746654,
+      "learning_rate": 4.800613496932516e-05,
+      "loss": 0.3041,
+      "step": 314
+    },
+    {
+      "epoch": 0.29059040590405905,
+      "grad_norm": 0.562035249405384,
+      "learning_rate": 4.815950920245399e-05,
+      "loss": 0.2851,
+      "step": 315
+    },
+    {
+      "epoch": 0.2915129151291513,
+      "grad_norm": 0.536590531309307,
+      "learning_rate": 4.831288343558282e-05,
+      "loss": 0.3011,
+      "step": 316
+    },
+    {
+      "epoch": 0.2924354243542435,
+      "grad_norm": 0.5728269165630729,
+      "learning_rate": 4.846625766871166e-05,
+      "loss": 0.2891,
+      "step": 317
+    },
+    {
+      "epoch": 0.2933579335793358,
+      "grad_norm": 0.5526588133095315,
+      "learning_rate": 4.8619631901840495e-05,
+      "loss": 0.2873,
+      "step": 318
+    },
+    {
+      "epoch": 0.29428044280442806,
+      "grad_norm": 0.6078040240981145,
+      "learning_rate": 4.877300613496933e-05,
+      "loss": 0.2559,
+      "step": 319
+    },
+    {
+      "epoch": 0.2952029520295203,
+      "grad_norm": 0.5684796023103894,
+      "learning_rate": 4.892638036809816e-05,
+      "loss": 0.3157,
+      "step": 320
+    },
+    {
+      "epoch": 0.29612546125461253,
+      "grad_norm": 0.5847394173098899,
+      "learning_rate": 4.907975460122699e-05,
+      "loss": 0.3068,
+      "step": 321
+    },
+    {
+      "epoch": 0.29704797047970477,
+      "grad_norm": 0.5444187455969381,
+      "learning_rate": 4.923312883435583e-05,
+      "loss": 0.2886,
+      "step": 322
+    },
+    {
+      "epoch": 0.29797047970479706,
+      "grad_norm": 0.5453378770392785,
+      "learning_rate": 4.938650306748466e-05,
+      "loss": 0.2962,
+      "step": 323
+    },
+    {
+      "epoch": 0.2988929889298893,
+      "grad_norm": 0.5145521435893671,
+      "learning_rate": 4.9539877300613504e-05,
+      "loss": 0.2884,
+      "step": 324
+    },
+    {
+      "epoch": 0.29981549815498154,
+      "grad_norm": 0.54156802954864,
+      "learning_rate": 4.9693251533742335e-05,
+      "loss": 0.2766,
+      "step": 325
+    },
+    {
+      "epoch": 0.3007380073800738,
+      "grad_norm": 0.548183894892192,
+      "learning_rate": 4.984662576687117e-05,
+      "loss": 0.2851,
+      "step": 326
+    },
+    {
+      "epoch": 0.3016605166051661,
+      "grad_norm": 0.5241708687339598,
+      "learning_rate": 5e-05,
+      "loss": 0.3003,
+      "step": 327
+    },
+    {
+      "epoch": 0.3025830258302583,
+      "grad_norm": 0.5247623186991769,
+      "learning_rate": 4.999998559009648e-05,
+      "loss": 0.2907,
+      "step": 328
+    },
+    {
+      "epoch": 0.30350553505535055,
+      "grad_norm": 0.46550756716924957,
+      "learning_rate": 4.999994236040253e-05,
+      "loss": 0.2773,
+      "step": 329
+    },
+    {
+      "epoch": 0.3044280442804428,
+      "grad_norm": 0.5464343854379993,
+      "learning_rate": 4.999987031096798e-05,
+      "loss": 0.2838,
+      "step": 330
+    },
+    {
+      "epoch": 0.3053505535055351,
+      "grad_norm": 0.5408842372543923,
+      "learning_rate": 4.999976944187589e-05,
+      "loss": 0.3108,
+      "step": 331
+    },
+    {
+      "epoch": 0.3062730627306273,
+      "grad_norm": 0.47661724256796645,
+      "learning_rate": 4.999963975324254e-05,
+      "loss": 0.2778,
+      "step": 332
+    },
+    {
+      "epoch": 0.30719557195571956,
+      "grad_norm": 0.5485844438480383,
+      "learning_rate": 4.9999481245217444e-05,
+      "loss": 0.2658,
+      "step": 333
+    },
+    {
+      "epoch": 0.3081180811808118,
+      "grad_norm": 0.7530365626718195,
+      "learning_rate": 4.999929391798332e-05,
+      "loss": 0.3133,
+      "step": 334
+    },
+    {
+      "epoch": 0.30904059040590404,
+      "grad_norm": 0.6008376648478589,
+      "learning_rate": 4.999907777175612e-05,
+      "loss": 0.2975,
+      "step": 335
+    },
+    {
+      "epoch": 0.30996309963099633,
+      "grad_norm": 0.5622146369546329,
+      "learning_rate": 4.999883280678501e-05,
+      "loss": 0.2682,
+      "step": 336
+    },
+    {
+      "epoch": 0.31088560885608857,
+      "grad_norm": 0.5895201399404543,
+      "learning_rate": 4.999855902335239e-05,
+      "loss": 0.2942,
+      "step": 337
+    },
+    {
+      "epoch": 0.3118081180811808,
+      "grad_norm": 0.5241668276774941,
+      "learning_rate": 4.999825642177387e-05,
+      "loss": 0.3131,
+      "step": 338
+    },
+    {
+      "epoch": 0.31273062730627305,
+      "grad_norm": 0.499627002160098,
+      "learning_rate": 4.9997925002398295e-05,
+      "loss": 0.2928,
+      "step": 339
+    },
+    {
+      "epoch": 0.31365313653136534,
+      "grad_norm": 0.550188057582933,
+      "learning_rate": 4.9997564765607716e-05,
+      "loss": 0.2991,
+      "step": 340
+    },
+    {
+      "epoch": 0.3145756457564576,
+      "grad_norm": 0.5371175592847117,
+      "learning_rate": 4.999717571181742e-05,
+      "loss": 0.295,
+      "step": 341
+    },
+    {
+      "epoch": 0.3154981549815498,
+      "grad_norm": 0.5493905636808714,
+      "learning_rate": 4.9996757841475894e-05,
+      "loss": 0.2905,
+      "step": 342
+    },
+    {
+      "epoch": 0.31642066420664205,
+      "grad_norm": 0.5253806832374459,
+      "learning_rate": 4.9996311155064856e-05,
+      "loss": 0.2948,
+      "step": 343
+    },
+    {
+      "epoch": 0.3173431734317343,
+      "grad_norm": 0.5539799821025654,
+      "learning_rate": 4.9995835653099254e-05,
+      "loss": 0.291,
+      "step": 344
+    },
+    {
+      "epoch": 0.3182656826568266,
+      "grad_norm": 0.5550654258039699,
+      "learning_rate": 4.999533133612723e-05,
+      "loss": 0.2821,
+      "step": 345
+    },
+    {
+      "epoch": 0.3191881918819188,
+      "grad_norm": 0.7137968310290836,
+      "learning_rate": 4.9994798204730166e-05,
+      "loss": 0.3089,
+      "step": 346
+    },
+    {
+      "epoch": 0.32011070110701106,
+      "grad_norm": 0.5139241368756094,
+      "learning_rate": 4.999423625952264e-05,
+      "loss": 0.2847,
+      "step": 347
+    },
+    {
+      "epoch": 0.3210332103321033,
+      "grad_norm": 0.5262978328667104,
+      "learning_rate": 4.999364550115248e-05,
+      "loss": 0.2452,
+      "step": 348
+    },
+    {
+      "epoch": 0.3219557195571956,
+      "grad_norm": 0.5347337657622876,
+      "learning_rate": 4.9993025930300686e-05,
+      "loss": 0.2763,
+      "step": 349
+    },
+    {
+      "epoch": 0.32287822878228783,
+      "grad_norm": 0.6030031508392452,
+      "learning_rate": 4.9992377547681505e-05,
+      "loss": 0.3049,
+      "step": 350
+    },
+    {
+      "epoch": 0.3238007380073801,
+      "grad_norm": 0.5946252763444188,
+      "learning_rate": 4.999170035404239e-05,
+      "loss": 0.2984,
+      "step": 351
+    },
+    {
+      "epoch": 0.3247232472324723,
+      "grad_norm": 0.5719874620235956,
+      "learning_rate": 4.999099435016399e-05,
+      "loss": 0.275,
+      "step": 352
+    },
+    {
+      "epoch": 0.32564575645756455,
+      "grad_norm": 0.7364304283462457,
+      "learning_rate": 4.99902595368602e-05,
+      "loss": 0.2895,
+      "step": 353
+    },
+    {
+      "epoch": 0.32656826568265684,
+      "grad_norm": 0.5543236697442724,
+      "learning_rate": 4.998949591497809e-05,
+      "loss": 0.339,
+      "step": 354
+    },
+    {
+      "epoch": 0.3274907749077491,
+      "grad_norm": 0.4761992054593626,
+      "learning_rate": 4.998870348539797e-05,
+      "loss": 0.2633,
+      "step": 355
+    },
+    {
+      "epoch": 0.3284132841328413,
+      "grad_norm": 0.5726867972200776,
+      "learning_rate": 4.998788224903334e-05,
+      "loss": 0.2848,
+      "step": 356
+    },
+    {
+      "epoch": 0.32933579335793356,
+      "grad_norm": 0.4876367361430421,
+      "learning_rate": 4.9987032206830906e-05,
+      "loss": 0.2923,
+      "step": 357
+    },
+    {
+      "epoch": 0.33025830258302585,
+      "grad_norm": 0.5039945497900252,
+      "learning_rate": 4.9986153359770614e-05,
+      "loss": 0.2762,
+      "step": 358
+    },
+    {
+      "epoch": 0.3311808118081181,
+      "grad_norm": 0.5138969084818104,
+      "learning_rate": 4.998524570886558e-05,
+      "loss": 0.2611,
+      "step": 359
+    },
+    {
+      "epoch": 0.33210332103321033,
+      "grad_norm": 0.5593828298420278,
+      "learning_rate": 4.998430925516213e-05,
+      "loss": 0.2837,
+      "step": 360
+    },
+    {
+      "epoch": 0.33302583025830257,
+      "grad_norm": 0.5676564634771994,
+      "learning_rate": 4.9983343999739805e-05,
+      "loss": 0.3008,
+      "step": 361
+    },
+    {
+      "epoch": 0.3339483394833948,
+      "grad_norm": 0.5540762769745929,
+      "learning_rate": 4.998234994371135e-05,
+      "loss": 0.2901,
+      "step": 362
+    },
+    {
+      "epoch": 0.3348708487084871,
+      "grad_norm": 0.5275548687528755,
+      "learning_rate": 4.99813270882227e-05,
+      "loss": 0.294,
+      "step": 363
+    },
+    {
+      "epoch": 0.33579335793357934,
+      "grad_norm": 0.4698081370394266,
+      "learning_rate": 4.9980275434452995e-05,
+      "loss": 0.2585,
+      "step": 364
+    },
+    {
+      "epoch": 0.3367158671586716,
+      "grad_norm": 0.5518300809849733,
+      "learning_rate": 4.997919498361457e-05,
+      "loss": 0.294,
+      "step": 365
+    },
+    {
+      "epoch": 0.3376383763837638,
+      "grad_norm": 0.4924798980807726,
+      "learning_rate": 4.997808573695297e-05,
+      "loss": 0.2931,
+      "step": 366
+    },
+    {
+      "epoch": 0.3385608856088561,
+      "grad_norm": 0.4971688337772927,
+      "learning_rate": 4.997694769574692e-05,
+      "loss": 0.2935,
+      "step": 367
+    },
+    {
+      "epoch": 0.33948339483394835,
+      "grad_norm": 0.4955383153372864,
+      "learning_rate": 4.997578086130834e-05,
+      "loss": 0.2842,
+      "step": 368
+    },
+    {
+      "epoch": 0.3404059040590406,
+      "grad_norm": 0.5577743523602456,
+      "learning_rate": 4.997458523498236e-05,
+      "loss": 0.2781,
+      "step": 369
+    },
+    {
+      "epoch": 0.3413284132841328,
+      "grad_norm": 0.5039551863232465,
+      "learning_rate": 4.9973360818147276e-05,
+      "loss": 0.2839,
+      "step": 370
+    },
+    {
+      "epoch": 0.3422509225092251,
+      "grad_norm": 0.6590350760855003,
+      "learning_rate": 4.99721076122146e-05,
+      "loss": 0.303,
+      "step": 371
+    },
+    {
+      "epoch": 0.34317343173431736,
+      "grad_norm": 0.5389630694407243,
+      "learning_rate": 4.9970825618629e-05,
+      "loss": 0.2997,
+      "step": 372
+    },
+    {
+      "epoch": 0.3440959409594096,
+      "grad_norm": 0.5616168426598683,
+      "learning_rate": 4.9969514838868364e-05,
+      "loss": 0.2928,
+      "step": 373
+    },
+    {
+      "epoch": 0.34501845018450183,
+      "grad_norm": 0.4749345164297593,
+      "learning_rate": 4.996817527444374e-05,
+      "loss": 0.2891,
+      "step": 374
+    },
+    {
+      "epoch": 0.3459409594095941,
+      "grad_norm": 0.5570269133618422,
+      "learning_rate": 4.996680692689938e-05,
+      "loss": 0.3089,
+      "step": 375
+    },
+    {
+      "epoch": 0.34686346863468637,
+      "grad_norm": 0.5620682199124284,
+      "learning_rate": 4.996540979781269e-05,
+      "loss": 0.2734,
+      "step": 376
+    },
+    {
+      "epoch": 0.3477859778597786,
+      "grad_norm": 0.5092748080272552,
+      "learning_rate": 4.996398388879427e-05,
+      "loss": 0.3017,
+      "step": 377
+    },
+    {
+      "epoch": 0.34870848708487084,
+      "grad_norm": 0.47199034787323235,
+      "learning_rate": 4.996252920148791e-05,
+      "loss": 0.2699,
+      "step": 378
+    },
+    {
+      "epoch": 0.3496309963099631,
+      "grad_norm": 0.521363279262548,
+      "learning_rate": 4.996104573757054e-05,
+      "loss": 0.3283,
+      "step": 379
+    },
+    {
+      "epoch": 0.3505535055350554,
+      "grad_norm": 0.513514483372052,
+      "learning_rate": 4.995953349875232e-05,
+      "loss": 0.2764,
+      "step": 380
+    },
+    {
+      "epoch": 0.3514760147601476,
+      "grad_norm": 0.4611726498410872,
+      "learning_rate": 4.9957992486776516e-05,
+      "loss": 0.2941,
+      "step": 381
+    },
+    {
+      "epoch": 0.35239852398523985,
+      "grad_norm": 0.439926621612698,
+      "learning_rate": 4.995642270341961e-05,
+      "loss": 0.2696,
+      "step": 382
+    },
+    {
+      "epoch": 0.3533210332103321,
+      "grad_norm": 0.47508444380843656,
+      "learning_rate": 4.995482415049123e-05,
+      "loss": 0.2904,
+      "step": 383
+    },
+    {
+      "epoch": 0.35424354243542433,
+      "grad_norm": 0.47065356253686386,
+      "learning_rate": 4.995319682983418e-05,
+      "loss": 0.2967,
+      "step": 384
+    },
+    {
+      "epoch": 0.3551660516605166,
+      "grad_norm": 0.5298362343070417,
+      "learning_rate": 4.995154074332441e-05,
+      "loss": 0.2865,
+      "step": 385
+    },
+    {
+      "epoch": 0.35608856088560886,
+      "grad_norm": 0.4686677587974533,
+      "learning_rate": 4.994985589287107e-05,
+      "loss": 0.2953,
+      "step": 386
+    },
+    {
+      "epoch": 0.3570110701107011,
+      "grad_norm": 0.4622899566823007,
+      "learning_rate": 4.994814228041641e-05,
+      "loss": 0.2846,
+      "step": 387
+    },
+    {
+      "epoch": 0.35793357933579334,
+      "grad_norm": 0.5023602796309572,
+      "learning_rate": 4.9946399907935894e-05,
+      "loss": 0.2982,
+      "step": 388
+    },
+    {
+      "epoch": 0.35885608856088563,
+      "grad_norm": 0.47765997988304104,
+      "learning_rate": 4.9944628777438104e-05,
+      "loss": 0.2631,
+      "step": 389
+    },
+    {
+      "epoch": 0.35977859778597787,
+      "grad_norm": 0.5233065519746218,
+      "learning_rate": 4.99428288909648e-05,
+      "loss": 0.3156,
+      "step": 390
+    },
+    {
+      "epoch": 0.3607011070110701,
+      "grad_norm": 0.5077270448652319,
+      "learning_rate": 4.994100025059085e-05,
+      "loss": 0.3018,
+      "step": 391
+    },
+    {
+      "epoch": 0.36162361623616235,
+      "grad_norm": 0.501254905607358,
+      "learning_rate": 4.993914285842433e-05,
+      "loss": 0.2854,
+      "step": 392
+    },
+    {
+      "epoch": 0.3625461254612546,
+      "grad_norm": 0.462697133871905,
+      "learning_rate": 4.9937256716606394e-05,
+      "loss": 0.2777,
+      "step": 393
+    },
+    {
+      "epoch": 0.3634686346863469,
+      "grad_norm": 0.5345478233734732,
+      "learning_rate": 4.99353418273114e-05,
+      "loss": 0.2797,
+      "step": 394
+    },
+    {
+      "epoch": 0.3643911439114391,
+      "grad_norm": 0.5614041609883599,
+      "learning_rate": 4.993339819274679e-05,
+      "loss": 0.3221,
+      "step": 395
+    },
+    {
+      "epoch": 0.36531365313653136,
+      "grad_norm": 0.5208539310813163,
+      "learning_rate": 4.9931425815153205e-05,
+      "loss": 0.2901,
+      "step": 396
+    },
+    {
+      "epoch": 0.3662361623616236,
+      "grad_norm": 0.4757548139725171,
+      "learning_rate": 4.992942469680436e-05,
+      "loss": 0.274,
+      "step": 397
+    },
+    {
+      "epoch": 0.3671586715867159,
+      "grad_norm": 0.5076300925523486,
+      "learning_rate": 4.992739484000714e-05,
+      "loss": 0.2836,
+      "step": 398
+    },
+    {
+      "epoch": 0.36808118081180813,
+      "grad_norm": 0.49140369112610577,
+      "learning_rate": 4.992533624710154e-05,
+      "loss": 0.2572,
+      "step": 399
+    },
+    {
+      "epoch": 0.36900369003690037,
+      "grad_norm": 0.4384787991074871,
+      "learning_rate": 4.992324892046069e-05,
+      "loss": 0.2559,
+      "step": 400
+    },
+    {
+      "epoch": 0.3699261992619926,
+      "grad_norm": 0.47598607693350403,
+      "learning_rate": 4.992113286249086e-05,
+      "loss": 0.2611,
+      "step": 401
+    },
+    {
+      "epoch": 0.37084870848708484,
+      "grad_norm": 0.49850175098728877,
+      "learning_rate": 4.9918988075631404e-05,
+      "loss": 0.2901,
+      "step": 402
+    },
+    {
+      "epoch": 0.37177121771217714,
+      "grad_norm": 0.5659124528427997,
+      "learning_rate": 4.991681456235483e-05,
+      "loss": 0.2994,
+      "step": 403
+    },
+    {
+      "epoch": 0.3726937269372694,
+      "grad_norm": 0.4704789200086404,
+      "learning_rate": 4.991461232516675e-05,
+      "loss": 0.275,
+      "step": 404
+    },
+    {
+      "epoch": 0.3736162361623616,
+      "grad_norm": 0.5042793616248943,
+      "learning_rate": 4.9912381366605876e-05,
+      "loss": 0.2801,
+      "step": 405
+    },
+    {
+      "epoch": 0.37453874538745385,
+      "grad_norm": 0.5183087948851817,
+      "learning_rate": 4.991012168924404e-05,
+      "loss": 0.299,
+      "step": 406
+    },
+    {
+      "epoch": 0.37546125461254615,
+      "grad_norm": 0.49887343246856986,
+      "learning_rate": 4.9907833295686185e-05,
+      "loss": 0.2726,
+      "step": 407
+    },
+    {
+      "epoch": 0.3763837638376384,
+      "grad_norm": 0.5472149090701182,
+      "learning_rate": 4.990551618857035e-05,
+      "loss": 0.2977,
+      "step": 408
+    },
+    {
+      "epoch": 0.3773062730627306,
+      "grad_norm": 0.4205940791760018,
+      "learning_rate": 4.990317037056769e-05,
+      "loss": 0.2619,
+      "step": 409
+    },
+    {
+      "epoch": 0.37822878228782286,
+      "grad_norm": 0.4628735099185641,
+      "learning_rate": 4.990079584438243e-05,
+      "loss": 0.2777,
+      "step": 410
+    },
+    {
+      "epoch": 0.37915129151291516,
+      "grad_norm": 0.44009518829262617,
+      "learning_rate": 4.989839261275191e-05,
+      "loss": 0.2605,
+      "step": 411
+    },
+    {
+      "epoch": 0.3800738007380074,
+      "grad_norm": 0.5271666577742473,
+      "learning_rate": 4.989596067844656e-05,
+      "loss": 0.3104,
+      "step": 412
+    },
+    {
+      "epoch": 0.38099630996309963,
+      "grad_norm": 0.49637247500333725,
+      "learning_rate": 4.989350004426989e-05,
+      "loss": 0.2911,
+      "step": 413
+    },
+    {
+      "epoch": 0.38191881918819187,
+      "grad_norm": 0.8195228166776699,
+      "learning_rate": 4.9891010713058506e-05,
+      "loss": 0.2697,
+      "step": 414
+    },
+    {
+      "epoch": 0.3828413284132841,
+      "grad_norm": 0.5735252085658141,
+      "learning_rate": 4.9888492687682096e-05,
+      "loss": 0.2996,
+      "step": 415
+    },
+    {
+      "epoch": 0.3837638376383764,
+      "grad_norm": 0.4625217340159638,
+      "learning_rate": 4.98859459710434e-05,
+      "loss": 0.2433,
+      "step": 416
+    },
+    {
+      "epoch": 0.38468634686346864,
+      "grad_norm": 0.5189849572409303,
+      "learning_rate": 4.988337056607827e-05,
+      "loss": 0.2974,
+      "step": 417
+    },
+    {
+      "epoch": 0.3856088560885609,
+      "grad_norm": 0.5636750135554274,
+      "learning_rate": 4.988076647575562e-05,
+      "loss": 0.3159,
+      "step": 418
+    },
+    {
+      "epoch": 0.3865313653136531,
+      "grad_norm": 0.4692607149916364,
+      "learning_rate": 4.987813370307739e-05,
+      "loss": 0.2933,
+      "step": 419
+    },
+    {
+      "epoch": 0.3874538745387454,
+      "grad_norm": 0.4625490071520199,
+      "learning_rate": 4.987547225107866e-05,
+      "loss": 0.2913,
+      "step": 420
+    },
+    {
+      "epoch": 0.38837638376383765,
+      "grad_norm": 0.5310824156987997,
+      "learning_rate": 4.987278212282751e-05,
+      "loss": 0.2797,
+      "step": 421
+    },
+    {
+      "epoch": 0.3892988929889299,
+      "grad_norm": 0.45783594777827885,
+      "learning_rate": 4.9870063321425105e-05,
+      "loss": 0.279,
+      "step": 422
+    },
+    {
+      "epoch": 0.39022140221402213,
+      "grad_norm": 0.4546793722346438,
+      "learning_rate": 4.986731585000566e-05,
+      "loss": 0.2727,
+      "step": 423
+    },
+    {
+      "epoch": 0.39114391143911437,
+      "grad_norm": 0.5003814360244799,
+      "learning_rate": 4.9864539711736425e-05,
+      "loss": 0.2842,
+      "step": 424
+    },
+    {
+      "epoch": 0.39206642066420666,
+      "grad_norm": 0.48676719250433703,
+      "learning_rate": 4.986173490981773e-05,
+      "loss": 0.3123,
+      "step": 425
+    },
+    {
+      "epoch": 0.3929889298892989,
+      "grad_norm": 0.44752651152763834,
+      "learning_rate": 4.985890144748292e-05,
+      "loss": 0.2714,
+      "step": 426
+    },
+    {
+      "epoch": 0.39391143911439114,
+      "grad_norm": 0.4524183699229073,
+      "learning_rate": 4.985603932799839e-05,
+      "loss": 0.2521,
+      "step": 427
+    },
+    {
+      "epoch": 0.3948339483394834,
+      "grad_norm": 0.6702816647145312,
+      "learning_rate": 4.9853148554663564e-05,
+      "loss": 0.2766,
+      "step": 428
+    },
+    {
+      "epoch": 0.39575645756457567,
+      "grad_norm": 0.45992023501143997,
+      "learning_rate": 4.985022913081091e-05,
+      "loss": 0.2684,
+      "step": 429
+    },
+    {
+      "epoch": 0.3966789667896679,
+      "grad_norm": 0.49294214318656143,
+      "learning_rate": 4.9847281059805914e-05,
+      "loss": 0.2782,
+      "step": 430
+    },
+    {
+      "epoch": 0.39760147601476015,
+      "grad_norm": 0.42437535528454967,
+      "learning_rate": 4.98443043450471e-05,
+      "loss": 0.2648,
+      "step": 431
+    },
+    {
+      "epoch": 0.3985239852398524,
+      "grad_norm": 0.45541868520627043,
+      "learning_rate": 4.9841298989965984e-05,
+      "loss": 0.2667,
+      "step": 432
+    },
+    {
+      "epoch": 0.3994464944649446,
+      "grad_norm": 0.5286063432706287,
+      "learning_rate": 4.983826499802712e-05,
+      "loss": 0.2709,
+      "step": 433
+    },
+    {
+      "epoch": 0.4003690036900369,
+      "grad_norm": 0.4543353369578046,
+      "learning_rate": 4.9835202372728086e-05,
+      "loss": 0.2474,
+      "step": 434
+    },
+    {
+      "epoch": 0.40129151291512916,
+      "grad_norm": 0.52459663995117,
+      "learning_rate": 4.9832111117599436e-05,
+      "loss": 0.2831,
+      "step": 435
+    },
+    {
+      "epoch": 0.4022140221402214,
+      "grad_norm": 0.5242387005681102,
+      "learning_rate": 4.982899123620475e-05,
+      "loss": 0.2708,
+      "step": 436
+    },
+    {
+      "epoch": 0.40313653136531363,
+      "grad_norm": 0.507220401515919,
+      "learning_rate": 4.982584273214061e-05,
+      "loss": 0.3001,
+      "step": 437
+    },
+    {
+      "epoch": 0.4040590405904059,
+      "grad_norm": 0.44289090289551086,
+      "learning_rate": 4.982266560903657e-05,
+      "loss": 0.2749,
+      "step": 438
+    },
+    {
+      "epoch": 0.40498154981549817,
+      "grad_norm": 0.4860542261240817,
+      "learning_rate": 4.981945987055521e-05,
+      "loss": 0.2865,
+      "step": 439
+    },
+    {
+      "epoch": 0.4059040590405904,
+      "grad_norm": 0.48611623956271083,
+      "learning_rate": 4.981622552039207e-05,
+      "loss": 0.2704,
+      "step": 440
+    },
+    {
+      "epoch": 0.40682656826568264,
+      "grad_norm": 0.5231243198031841,
+      "learning_rate": 4.981296256227569e-05,
+      "loss": 0.2918,
+      "step": 441
+    },
+    {
+      "epoch": 0.4077490774907749,
+      "grad_norm": 0.4645772233906956,
+      "learning_rate": 4.980967099996759e-05,
+      "loss": 0.2333,
+      "step": 442
+    },
+    {
+      "epoch": 0.4086715867158672,
+      "grad_norm": 0.5057641477236923,
+      "learning_rate": 4.980635083726225e-05,
+      "loss": 0.2866,
+      "step": 443
+    },
+    {
+      "epoch": 0.4095940959409594,
+      "grad_norm": 0.5746343137271073,
+      "learning_rate": 4.980300207798711e-05,
+      "loss": 0.2751,
+      "step": 444
+    },
+    {
+      "epoch": 0.41051660516605165,
+      "grad_norm": 0.4494986104587851,
+      "learning_rate": 4.979962472600263e-05,
+      "loss": 0.2815,
+      "step": 445
+    },
+    {
+      "epoch": 0.4114391143911439,
+      "grad_norm": 0.6074090221600998,
+      "learning_rate": 4.979621878520216e-05,
+      "loss": 0.3311,
+      "step": 446
+    },
+    {
+      "epoch": 0.4123616236162362,
+      "grad_norm": 0.43702703064995774,
+      "learning_rate": 4.979278425951207e-05,
+      "loss": 0.2654,
+      "step": 447
+    },
+    {
+      "epoch": 0.4132841328413284,
+      "grad_norm": 0.4835542229937427,
+      "learning_rate": 4.978932115289164e-05,
+      "loss": 0.2823,
+      "step": 448
+    },
+    {
+      "epoch": 0.41420664206642066,
+      "grad_norm": 0.48416396231619063,
+      "learning_rate": 4.9785829469333116e-05,
+      "loss": 0.2637,
+      "step": 449
+    },
+    {
+      "epoch": 0.4151291512915129,
+      "grad_norm": 0.4661535425065205,
+      "learning_rate": 4.978230921286168e-05,
+      "loss": 0.3024,
+      "step": 450
+    },
+    {
+      "epoch": 0.4160516605166052,
+      "grad_norm": 0.5035766846654282,
+      "learning_rate": 4.9778760387535465e-05,
+      "loss": 0.3318,
+      "step": 451
+    },
+    {
+      "epoch": 0.41697416974169743,
+      "grad_norm": 0.45587434855728826,
+      "learning_rate": 4.977518299744552e-05,
+      "loss": 0.2778,
+      "step": 452
+    },
+    {
+      "epoch": 0.41789667896678967,
+      "grad_norm": 0.5226509979413794,
+      "learning_rate": 4.9771577046715846e-05,
+      "loss": 0.3112,
+      "step": 453
+    },
+    {
+      "epoch": 0.4188191881918819,
+      "grad_norm": 0.45316940391861665,
+      "learning_rate": 4.976794253950334e-05,
+      "loss": 0.2761,
+      "step": 454
+    },
+    {
+      "epoch": 0.41974169741697415,
+      "grad_norm": 0.5061175030402905,
+      "learning_rate": 4.976427947999784e-05,
+      "loss": 0.2869,
+      "step": 455
+    },
+    {
+      "epoch": 0.42066420664206644,
+      "grad_norm": 0.5434356789335424,
+      "learning_rate": 4.976058787242209e-05,
+      "loss": 0.2655,
+      "step": 456
+    },
+    {
+      "epoch": 0.4215867158671587,
+      "grad_norm": 0.452576602253526,
+      "learning_rate": 4.9756867721031756e-05,
+      "loss": 0.2708,
+      "step": 457
+    },
+    {
+      "epoch": 0.4225092250922509,
+      "grad_norm": 0.4961825947084803,
+      "learning_rate": 4.975311903011539e-05,
+      "loss": 0.2806,
+      "step": 458
+    },
+    {
+      "epoch": 0.42343173431734316,
+      "grad_norm": 0.4630990428690511,
+      "learning_rate": 4.9749341803994465e-05,
+      "loss": 0.2655,
+      "step": 459
+    },
+    {
+      "epoch": 0.42435424354243545,
+      "grad_norm": 0.4956010831171726,
+      "learning_rate": 4.9745536047023324e-05,
+      "loss": 0.2816,
+      "step": 460
+    },
+    {
+      "epoch": 0.4252767527675277,
+      "grad_norm": 0.43560042606186716,
+      "learning_rate": 4.974170176358922e-05,
+      "loss": 0.2691,
+      "step": 461
+    },
+    {
+      "epoch": 0.4261992619926199,
+      "grad_norm": 0.4878621995774692,
+      "learning_rate": 4.973783895811228e-05,
+      "loss": 0.2893,
+      "step": 462
+    },
+    {
+      "epoch": 0.42712177121771217,
+      "grad_norm": 0.4718466074702949,
+      "learning_rate": 4.9733947635045534e-05,
+      "loss": 0.2697,
+      "step": 463
+    },
+    {
+      "epoch": 0.4280442804428044,
+      "grad_norm": 0.502923872783246,
+      "learning_rate": 4.9730027798874856e-05,
+      "loss": 0.2944,
+      "step": 464
+    },
+    {
+      "epoch": 0.4289667896678967,
+      "grad_norm": 0.5007071278997294,
+      "learning_rate": 4.9726079454119e-05,
+      "loss": 0.2807,
+      "step": 465
+    },
+    {
+      "epoch": 0.42988929889298894,
+      "grad_norm": 0.4815546251566584,
+      "learning_rate": 4.97221026053296e-05,
+      "loss": 0.2637,
+      "step": 466
+    },
+    {
+      "epoch": 0.4308118081180812,
+      "grad_norm": 0.7239974358847797,
+      "learning_rate": 4.971809725709112e-05,
+      "loss": 0.2806,
+      "step": 467
+    },
+    {
+      "epoch": 0.4317343173431734,
+      "grad_norm": 0.4779514843339991,
+      "learning_rate": 4.971406341402091e-05,
+      "loss": 0.3156,
+      "step": 468
+    },
+    {
+      "epoch": 0.4326568265682657,
+      "grad_norm": 0.5319628872166638,
+      "learning_rate": 4.9710001080769145e-05,
+      "loss": 0.2581,
+      "step": 469
+    },
+    {
+      "epoch": 0.43357933579335795,
+      "grad_norm": 0.4165504237124002,
+      "learning_rate": 4.970591026201884e-05,
+      "loss": 0.2662,
+      "step": 470
+    },
+    {
+      "epoch": 0.4345018450184502,
+      "grad_norm": 0.42652362506523284,
+      "learning_rate": 4.970179096248588e-05,
+      "loss": 0.2427,
+      "step": 471
+    },
+    {
+      "epoch": 0.4354243542435424,
+      "grad_norm": 0.426956519303912,
+      "learning_rate": 4.969764318691896e-05,
+      "loss": 0.2592,
+      "step": 472
+    },
+    {
+      "epoch": 0.43634686346863466,
+      "grad_norm": 0.4964034230546089,
+      "learning_rate": 4.9693466940099596e-05,
+      "loss": 0.2825,
+      "step": 473
+    },
+    {
+      "epoch": 0.43726937269372695,
+      "grad_norm": 0.45671563154158606,
+      "learning_rate": 4.968926222684213e-05,
+      "loss": 0.2711,
+      "step": 474
+    },
+    {
+      "epoch": 0.4381918819188192,
+      "grad_norm": 0.46933639358843837,
+      "learning_rate": 4.968502905199373e-05,
+      "loss": 0.2746,
+      "step": 475
+    },
+    {
+      "epoch": 0.43911439114391143,
+      "grad_norm": 0.47054551084355595,
+      "learning_rate": 4.968076742043437e-05,
+      "loss": 0.2962,
+      "step": 476
+    },
+    {
+      "epoch": 0.44003690036900367,
+      "grad_norm": 0.5576416962434739,
+      "learning_rate": 4.967647733707681e-05,
+      "loss": 0.2647,
+      "step": 477
+    },
+    {
+      "epoch": 0.44095940959409596,
+      "grad_norm": 0.5280087869537966,
+      "learning_rate": 4.9672158806866645e-05,
+      "loss": 0.3067,
+      "step": 478
+    },
+    {
+      "epoch": 0.4418819188191882,
+      "grad_norm": 0.44384985941092053,
+      "learning_rate": 4.9667811834782224e-05,
+      "loss": 0.2678,
+      "step": 479
+    },
+    {
+      "epoch": 0.44280442804428044,
+      "grad_norm": 0.5030864516669159,
+      "learning_rate": 4.966343642583472e-05,
+      "loss": 0.3055,
+      "step": 480
+    },
+    {
+      "epoch": 0.4437269372693727,
+      "grad_norm": 0.48741228397240355,
+      "learning_rate": 4.965903258506806e-05,
+      "loss": 0.2601,
+      "step": 481
+    },
+    {
+      "epoch": 0.4446494464944649,
+      "grad_norm": 0.4955807220446558,
+      "learning_rate": 4.9654600317558965e-05,
+      "loss": 0.2773,
+      "step": 482
+    },
+    {
+      "epoch": 0.4455719557195572,
+      "grad_norm": 0.47270716571562477,
+      "learning_rate": 4.9650139628416916e-05,
+      "loss": 0.3089,
+      "step": 483
+    },
+    {
+      "epoch": 0.44649446494464945,
+      "grad_norm": 0.4865711142061868,
+      "learning_rate": 4.9645650522784156e-05,
+      "loss": 0.2761,
+      "step": 484
+    },
+    {
+      "epoch": 0.4474169741697417,
+      "grad_norm": 0.48214633442910015,
+      "learning_rate": 4.9641133005835696e-05,
+      "loss": 0.2772,
+      "step": 485
+    },
+    {
+      "epoch": 0.4483394833948339,
+      "grad_norm": 0.44599604929644415,
+      "learning_rate": 4.963658708277929e-05,
+      "loss": 0.2337,
+      "step": 486
+    },
+    {
+      "epoch": 0.4492619926199262,
+      "grad_norm": 0.5107631952533427,
+      "learning_rate": 4.963201275885545e-05,
+      "loss": 0.2595,
+      "step": 487
+    },
+    {
+      "epoch": 0.45018450184501846,
+      "grad_norm": 0.4856998967219853,
+      "learning_rate": 4.962741003933742e-05,
+      "loss": 0.2616,
+      "step": 488
+    },
+    {
+      "epoch": 0.4511070110701107,
+      "grad_norm": 0.4716394691734046,
+      "learning_rate": 4.962277892953118e-05,
+      "loss": 0.2817,
+      "step": 489
+    },
+    {
+      "epoch": 0.45202952029520294,
+      "grad_norm": 0.4449737230991853,
+      "learning_rate": 4.9618119434775436e-05,
+      "loss": 0.2523,
+      "step": 490
+    },
+    {
+      "epoch": 0.45295202952029523,
+      "grad_norm": 0.5038411750840269,
+      "learning_rate": 4.961343156044161e-05,
+      "loss": 0.2948,
+      "step": 491
+    },
+    {
+      "epoch": 0.45387453874538747,
+      "grad_norm": 0.47723990277212436,
+      "learning_rate": 4.960871531193386e-05,
+      "loss": 0.275,
+      "step": 492
+    },
+    {
+      "epoch": 0.4547970479704797,
+      "grad_norm": 0.4874275599114698,
+      "learning_rate": 4.9603970694689036e-05,
+      "loss": 0.2658,
+      "step": 493
+    },
+    {
+      "epoch": 0.45571955719557194,
+      "grad_norm": 0.4604630895511357,
+      "learning_rate": 4.959919771417669e-05,
+      "loss": 0.2473,
+      "step": 494
+    },
+    {
+      "epoch": 0.4566420664206642,
+      "grad_norm": 0.4433615779641383,
+      "learning_rate": 4.959439637589909e-05,
+      "loss": 0.2648,
+      "step": 495
+    },
+    {
+      "epoch": 0.4575645756457565,
+      "grad_norm": 0.47951428273104185,
+      "learning_rate": 4.958956668539117e-05,
+      "loss": 0.2852,
+      "step": 496
+    },
+    {
+      "epoch": 0.4584870848708487,
+      "grad_norm": 0.436709393894546,
+      "learning_rate": 4.9584708648220554e-05,
+      "loss": 0.2782,
+      "step": 497
+    },
+    {
+      "epoch": 0.45940959409594095,
+      "grad_norm": 0.7256225184114476,
+      "learning_rate": 4.9579822269987574e-05,
+      "loss": 0.2632,
+      "step": 498
+    },
+    {
+      "epoch": 0.4603321033210332,
+      "grad_norm": 0.3961328509767162,
+      "learning_rate": 4.9574907556325186e-05,
+      "loss": 0.2534,
+      "step": 499
+    },
+    {
+      "epoch": 0.4612546125461255,
+      "grad_norm": 0.41311737848784036,
+      "learning_rate": 4.956996451289906e-05,
+      "loss": 0.2759,
+      "step": 500
+    },
+    {
+      "epoch": 0.4621771217712177,
+      "grad_norm": 0.5557796830484363,
+      "learning_rate": 4.956499314540747e-05,
+      "loss": 0.2736,
+      "step": 501
+    },
+    {
+      "epoch": 0.46309963099630996,
+      "grad_norm": 0.5150229800472483,
+      "learning_rate": 4.9559993459581375e-05,
+      "loss": 0.3181,
+      "step": 502
+    },
+    {
+      "epoch": 0.4640221402214022,
+      "grad_norm": 0.4501951026255143,
+      "learning_rate": 4.955496546118439e-05,
+      "loss": 0.2556,
+      "step": 503
+    },
+    {
+      "epoch": 0.46494464944649444,
+      "grad_norm": 0.8463552432381145,
+      "learning_rate": 4.954990915601274e-05,
+      "loss": 0.2482,
+      "step": 504
+    },
+    {
+      "epoch": 0.46586715867158673,
+      "grad_norm": 0.43951620309458156,
+      "learning_rate": 4.95448245498953e-05,
+      "loss": 0.2865,
+      "step": 505
+    },
+    {
+      "epoch": 0.466789667896679,
+      "grad_norm": 0.46321874246749795,
+      "learning_rate": 4.9539711648693555e-05,
+      "loss": 0.2847,
+      "step": 506
+    },
+    {
+      "epoch": 0.4677121771217712,
+      "grad_norm": 0.42236752170429886,
+      "learning_rate": 4.953457045830163e-05,
+      "loss": 0.2766,
+      "step": 507
+    },
+    {
+      "epoch": 0.46863468634686345,
+      "grad_norm": 0.5338015589876643,
+      "learning_rate": 4.9529400984646244e-05,
+      "loss": 0.3102,
+      "step": 508
+    },
+    {
+      "epoch": 0.46955719557195574,
+      "grad_norm": 0.45534183065412587,
+      "learning_rate": 4.952420323368673e-05,
+      "loss": 0.2976,
+      "step": 509
+    },
+    {
+      "epoch": 0.470479704797048,
+      "grad_norm": 0.46562018951031153,
+      "learning_rate": 4.951897721141502e-05,
+      "loss": 0.2961,
+      "step": 510
+    },
+    {
+      "epoch": 0.4714022140221402,
+      "grad_norm": 0.4038765927587688,
+      "learning_rate": 4.951372292385561e-05,
+      "loss": 0.2531,
+      "step": 511
+    },
+    {
+      "epoch": 0.47232472324723246,
+      "grad_norm": 0.4934140912362845,
+      "learning_rate": 4.950844037706563e-05,
+      "loss": 0.2767,
+      "step": 512
+    },
+    {
+      "epoch": 0.4732472324723247,
+      "grad_norm": 0.3985419688294112,
+      "learning_rate": 4.950312957713474e-05,
+      "loss": 0.2293,
+      "step": 513
+    },
+    {
+      "epoch": 0.474169741697417,
+      "grad_norm": 0.44235538916203665,
+      "learning_rate": 4.9497790530185194e-05,
+      "loss": 0.2712,
+      "step": 514
+    },
+    {
+      "epoch": 0.47509225092250923,
+      "grad_norm": 0.4345617378206702,
+      "learning_rate": 4.9492423242371814e-05,
+      "loss": 0.293,
+      "step": 515
+    },
+    {
+      "epoch": 0.47601476014760147,
+      "grad_norm": 0.47623291965082337,
+      "learning_rate": 4.948702771988195e-05,
+      "loss": 0.2989,
+      "step": 516
+    },
+    {
+      "epoch": 0.4769372693726937,
+      "grad_norm": 0.5004270904084257,
+      "learning_rate": 4.948160396893553e-05,
+      "loss": 0.2819,
+      "step": 517
+    },
+    {
+      "epoch": 0.477859778597786,
+      "grad_norm": 0.43365473204761534,
+      "learning_rate": 4.9476151995785016e-05,
+      "loss": 0.2779,
+      "step": 518
+    },
+    {
+      "epoch": 0.47878228782287824,
+      "grad_norm": 0.432061270625897,
+      "learning_rate": 4.9470671806715386e-05,
+      "loss": 0.2705,
+      "step": 519
+    },
+    {
+      "epoch": 0.4797047970479705,
+      "grad_norm": 0.511804435991744,
+      "learning_rate": 4.946516340804417e-05,
+      "loss": 0.2963,
+      "step": 520
+    },
+    {
+      "epoch": 0.4806273062730627,
+      "grad_norm": 0.4571692558782327,
+      "learning_rate": 4.945962680612142e-05,
+      "loss": 0.2918,
+      "step": 521
+    },
+    {
+      "epoch": 0.48154981549815495,
+      "grad_norm": 0.422817092719061,
+      "learning_rate": 4.945406200732966e-05,
+      "loss": 0.2861,
+      "step": 522
+    },
+    {
+      "epoch": 0.48247232472324725,
+      "grad_norm": 0.3882985681149562,
+      "learning_rate": 4.9448469018083965e-05,
+      "loss": 0.284,
+      "step": 523
+    },
+    {
+      "epoch": 0.4833948339483395,
+      "grad_norm": 0.4059792503571466,
+      "learning_rate": 4.9442847844831884e-05,
+      "loss": 0.2587,
+      "step": 524
+    },
+    {
+      "epoch": 0.4843173431734317,
+      "grad_norm": 0.5029319147515259,
+      "learning_rate": 4.9437198494053464e-05,
+      "loss": 0.3026,
+      "step": 525
+    },
+    {
+      "epoch": 0.48523985239852396,
+      "grad_norm": 0.501433585973909,
+      "learning_rate": 4.9431520972261236e-05,
+      "loss": 0.2803,
+      "step": 526
+    },
+    {
+      "epoch": 0.48616236162361626,
+      "grad_norm": 0.49813327319291245,
+      "learning_rate": 4.94258152860002e-05,
+      "loss": 0.2757,
+      "step": 527
+    },
+    {
+      "epoch": 0.4870848708487085,
+      "grad_norm": 0.43818296091000847,
+      "learning_rate": 4.942008144184783e-05,
+      "loss": 0.2547,
+      "step": 528
+    },
+    {
+      "epoch": 0.48800738007380073,
+      "grad_norm": 0.4817891496312095,
+      "learning_rate": 4.941431944641405e-05,
+      "loss": 0.2709,
+      "step": 529
+    },
+    {
+      "epoch": 0.488929889298893,
+      "grad_norm": 0.45278995667382466,
+      "learning_rate": 4.9408529306341255e-05,
+      "loss": 0.2676,
+      "step": 530
+    },
+    {
+      "epoch": 0.48985239852398527,
+      "grad_norm": 0.4683646977976737,
+      "learning_rate": 4.940271102830426e-05,
+      "loss": 0.2736,
+      "step": 531
+    },
+    {
+      "epoch": 0.4907749077490775,
+      "grad_norm": 0.5129772082178625,
+      "learning_rate": 4.939686461901034e-05,
+      "loss": 0.3174,
+      "step": 532
+    },
+    {
+      "epoch": 0.49169741697416974,
+      "grad_norm": 0.41287356302324596,
+      "learning_rate": 4.9390990085199197e-05,
+      "loss": 0.2495,
+      "step": 533
+    },
+    {
+      "epoch": 0.492619926199262,
+      "grad_norm": 0.4634074703994402,
+      "learning_rate": 4.938508743364293e-05,
+      "loss": 0.2689,
+      "step": 534
+    },
+    {
+      "epoch": 0.4935424354243542,
+      "grad_norm": 0.44341222256163587,
+      "learning_rate": 4.9379156671146084e-05,
+      "loss": 0.27,
+      "step": 535
+    },
+    {
+      "epoch": 0.4944649446494465,
+      "grad_norm": 0.405366223217404,
+      "learning_rate": 4.937319780454559e-05,
+      "loss": 0.2827,
+      "step": 536
+    },
+    {
+      "epoch": 0.49538745387453875,
+      "grad_norm": 0.6408259161108691,
+      "learning_rate": 4.936721084071079e-05,
+      "loss": 0.2761,
+      "step": 537
+    },
+    {
+      "epoch": 0.496309963099631,
+      "grad_norm": 0.4285821096818319,
+      "learning_rate": 4.936119578654341e-05,
+      "loss": 0.2903,
+      "step": 538
+    },
+    {
+      "epoch": 0.49723247232472323,
+      "grad_norm": 0.4718131706229978,
+      "learning_rate": 4.935515264897754e-05,
+      "loss": 0.2527,
+      "step": 539
+    },
+    {
+      "epoch": 0.4981549815498155,
+      "grad_norm": 0.4238454146716917,
+      "learning_rate": 4.934908143497969e-05,
+      "loss": 0.2697,
+      "step": 540
+    },
+    {
+      "epoch": 0.49907749077490776,
+      "grad_norm": 0.4619380626601982,
+      "learning_rate": 4.934298215154869e-05,
+      "loss": 0.2633,
+      "step": 541
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4654607007586769,
+      "learning_rate": 4.933685480571575e-05,
+      "loss": 0.2553,
+      "step": 542
+    },
+    {
+      "epoch": 0.5009225092250923,
+      "grad_norm": 0.4758385588029141,
+      "learning_rate": 4.933069940454443e-05,
+      "loss": 0.2542,
+      "step": 543
+    },
+    {
+      "epoch": 0.5018450184501845,
+      "grad_norm": 0.3888905402060683,
+      "learning_rate": 4.932451595513062e-05,
+      "loss": 0.237,
+      "step": 544
+    },
+    {
+      "epoch": 0.5027675276752768,
+      "grad_norm": 0.4446242533662745,
+      "learning_rate": 4.931830446460257e-05,
+      "loss": 0.2769,
+      "step": 545
+    },
+    {
+      "epoch": 0.503690036900369,
+      "grad_norm": 0.4179986241757249,
+      "learning_rate": 4.9312064940120825e-05,
+      "loss": 0.2665,
+      "step": 546
+    },
+    {
+      "epoch": 0.5046125461254612,
+      "grad_norm": 0.4704726176124594,
+      "learning_rate": 4.9305797388878264e-05,
+      "loss": 0.2698,
+      "step": 547
+    },
+    {
+      "epoch": 0.5055350553505535,
+      "grad_norm": 0.42816781624738687,
+      "learning_rate": 4.929950181810008e-05,
+      "loss": 0.2841,
+      "step": 548
+    },
+    {
+      "epoch": 0.5064575645756457,
+      "grad_norm": 0.4786245676444808,
+      "learning_rate": 4.929317823504373e-05,
+      "loss": 0.28,
+      "step": 549
+    },
+    {
+      "epoch": 0.507380073800738,
+      "grad_norm": 0.48781934756147605,
+      "learning_rate": 4.928682664699904e-05,
+      "loss": 0.2825,
+      "step": 550
+    },
+    {
+      "epoch": 0.5083025830258303,
+      "grad_norm": 0.5559869669654536,
+      "learning_rate": 4.928044706128803e-05,
+      "loss": 0.311,
+      "step": 551
+    },
+    {
+      "epoch": 0.5092250922509225,
+      "grad_norm": 0.4748900536986569,
+      "learning_rate": 4.927403948526504e-05,
+      "loss": 0.2624,
+      "step": 552
+    },
+    {
+      "epoch": 0.5101476014760148,
+      "grad_norm": 0.4957615812794995,
+      "learning_rate": 4.92676039263167e-05,
+      "loss": 0.2369,
+      "step": 553
+    },
+    {
+      "epoch": 0.511070110701107,
+      "grad_norm": 0.4130723638312345,
+      "learning_rate": 4.926114039186185e-05,
+      "loss": 0.2626,
+      "step": 554
+    },
+    {
+      "epoch": 0.5119926199261993,
+      "grad_norm": 0.40832203191352767,
+      "learning_rate": 4.925464888935162e-05,
+      "loss": 0.2652,
+      "step": 555
+    },
+    {
+      "epoch": 0.5129151291512916,
+      "grad_norm": 0.42179080972296734,
+      "learning_rate": 4.924812942626934e-05,
+      "loss": 0.2836,
+      "step": 556
+    },
+    {
+      "epoch": 0.5138376383763837,
+      "grad_norm": 0.47362251067172373,
+      "learning_rate": 4.924158201013062e-05,
+      "loss": 0.2822,
+      "step": 557
+    },
+    {
+      "epoch": 0.514760147601476,
+      "grad_norm": 0.4585248532810261,
+      "learning_rate": 4.923500664848326e-05,
+      "loss": 0.263,
+      "step": 558
+    },
+    {
+      "epoch": 0.5156826568265682,
+      "grad_norm": 0.4612972283763447,
+      "learning_rate": 4.922840334890729e-05,
+      "loss": 0.2869,
+      "step": 559
+    },
+    {
+      "epoch": 0.5166051660516605,
+      "grad_norm": 0.4912111054419741,
+      "learning_rate": 4.922177211901494e-05,
+      "loss": 0.2808,
+      "step": 560
+    },
+    {
+      "epoch": 0.5175276752767528,
+      "grad_norm": 0.4096288303101279,
+      "learning_rate": 4.921511296645064e-05,
+      "loss": 0.2728,
+      "step": 561
+    },
+    {
+      "epoch": 0.518450184501845,
+      "grad_norm": 0.4231251335442949,
+      "learning_rate": 4.920842589889102e-05,
+      "loss": 0.2457,
+      "step": 562
+    },
+    {
+      "epoch": 0.5193726937269373,
+      "grad_norm": 0.4714855709651646,
+      "learning_rate": 4.9201710924044865e-05,
+      "loss": 0.254,
+      "step": 563
+    },
+    {
+      "epoch": 0.5202952029520295,
+      "grad_norm": 0.4582728610686477,
+      "learning_rate": 4.9194968049653144e-05,
+      "loss": 0.2885,
+      "step": 564
+    },
+    {
+      "epoch": 0.5212177121771218,
+      "grad_norm": 0.4365802042178772,
+      "learning_rate": 4.9188197283489015e-05,
+      "loss": 0.2736,
+      "step": 565
+    },
+    {
+      "epoch": 0.522140221402214,
+      "grad_norm": 0.39054096278502237,
+      "learning_rate": 4.918139863335774e-05,
+      "loss": 0.2512,
+      "step": 566
+    },
+    {
+      "epoch": 0.5230627306273062,
+      "grad_norm": 0.3881999562207738,
+      "learning_rate": 4.917457210709675e-05,
+      "loss": 0.2457,
+      "step": 567
+    },
+    {
+      "epoch": 0.5239852398523985,
+      "grad_norm": 0.4230922310859938,
+      "learning_rate": 4.9167717712575635e-05,
+      "loss": 0.2852,
+      "step": 568
+    },
+    {
+      "epoch": 0.5249077490774908,
+      "grad_norm": 0.4539687703260704,
+      "learning_rate": 4.916083545769607e-05,
+      "loss": 0.276,
+      "step": 569
+    },
+    {
+      "epoch": 0.525830258302583,
+      "grad_norm": 0.4634154156646078,
+      "learning_rate": 4.915392535039187e-05,
+      "loss": 0.2664,
+      "step": 570
+    },
+    {
+      "epoch": 0.5267527675276753,
+      "grad_norm": 0.4106928537097177,
+      "learning_rate": 4.914698739862895e-05,
+      "loss": 0.253,
+      "step": 571
+    },
+    {
+      "epoch": 0.5276752767527675,
+      "grad_norm": 0.4801224680728872,
+      "learning_rate": 4.9140021610405326e-05,
+      "loss": 0.2962,
+      "step": 572
+    },
+    {
+      "epoch": 0.5285977859778598,
+      "grad_norm": 0.4660747634206696,
+      "learning_rate": 4.913302799375112e-05,
+      "loss": 0.2839,
+      "step": 573
+    },
+    {
+      "epoch": 0.5295202952029521,
+      "grad_norm": 0.4643565869221289,
+      "learning_rate": 4.91260065567285e-05,
+      "loss": 0.2967,
+      "step": 574
+    },
+    {
+      "epoch": 0.5304428044280443,
+      "grad_norm": 0.4565991699458606,
+      "learning_rate": 4.911895730743174e-05,
+      "loss": 0.2828,
+      "step": 575
+    },
+    {
+      "epoch": 0.5313653136531366,
+      "grad_norm": 0.4906222604967463,
+      "learning_rate": 4.9111880253987144e-05,
+      "loss": 0.2629,
+      "step": 576
+    },
+    {
+      "epoch": 0.5322878228782287,
+      "grad_norm": 0.5290433906880226,
+      "learning_rate": 4.9104775404553096e-05,
+      "loss": 0.3066,
+      "step": 577
+    },
+    {
+      "epoch": 0.533210332103321,
+      "grad_norm": 0.38292404341294,
+      "learning_rate": 4.909764276732001e-05,
+      "loss": 0.2509,
+      "step": 578
+    },
+    {
+      "epoch": 0.5341328413284133,
+      "grad_norm": 0.470498772870862,
+      "learning_rate": 4.9090482350510336e-05,
+      "loss": 0.2686,
+      "step": 579
+    },
+    {
+      "epoch": 0.5350553505535055,
+      "grad_norm": 0.49653229082754774,
+      "learning_rate": 4.908329416237855e-05,
+      "loss": 0.3074,
+      "step": 580
+    },
+    {
+      "epoch": 0.5359778597785978,
+      "grad_norm": 0.4672373400263147,
+      "learning_rate": 4.907607821121112e-05,
+      "loss": 0.3088,
+      "step": 581
+    },
+    {
+      "epoch": 0.5369003690036901,
+      "grad_norm": 0.4532750252264721,
+      "learning_rate": 4.906883450532657e-05,
+      "loss": 0.3046,
+      "step": 582
+    },
+    {
+      "epoch": 0.5378228782287823,
+      "grad_norm": 0.39484226178695664,
+      "learning_rate": 4.906156305307536e-05,
+      "loss": 0.2675,
+      "step": 583
+    },
+    {
+      "epoch": 0.5387453874538746,
+      "grad_norm": 0.46409690143882776,
+      "learning_rate": 4.905426386283998e-05,
+      "loss": 0.2799,
+      "step": 584
+    },
+    {
+      "epoch": 0.5396678966789668,
+      "grad_norm": 0.42793988061634014,
+      "learning_rate": 4.904693694303488e-05,
+      "loss": 0.2804,
+      "step": 585
+    },
+    {
+      "epoch": 0.540590405904059,
+      "grad_norm": 0.4563227359982457,
+      "learning_rate": 4.9039582302106465e-05,
+      "loss": 0.2755,
+      "step": 586
+    },
+    {
+      "epoch": 0.5415129151291513,
+      "grad_norm": 0.5135009107400244,
+      "learning_rate": 4.903219994853313e-05,
+      "loss": 0.286,
+      "step": 587
+    },
+    {
+      "epoch": 0.5424354243542435,
+      "grad_norm": 0.46841319933602094,
+      "learning_rate": 4.902478989082517e-05,
+      "loss": 0.2903,
+      "step": 588
+    },
+    {
+      "epoch": 0.5433579335793358,
+      "grad_norm": 0.43650400749097834,
+      "learning_rate": 4.901735213752486e-05,
+      "loss": 0.2392,
+      "step": 589
+    },
+    {
+      "epoch": 0.544280442804428,
+      "grad_norm": 0.43694449569568666,
+      "learning_rate": 4.900988669720637e-05,
+      "loss": 0.2756,
+      "step": 590
+    },
+    {
+      "epoch": 0.5452029520295203,
+      "grad_norm": 0.4318039876501152,
+      "learning_rate": 4.9002393578475816e-05,
+      "loss": 0.2898,
+      "step": 591
+    },
+    {
+      "epoch": 0.5461254612546126,
+      "grad_norm": 0.5479363128059827,
+      "learning_rate": 4.89948727899712e-05,
+      "loss": 0.3107,
+      "step": 592
+    },
+    {
+      "epoch": 0.5470479704797048,
+      "grad_norm": 0.4227935698497923,
+      "learning_rate": 4.898732434036244e-05,
+      "loss": 0.2884,
+      "step": 593
+    },
+    {
+      "epoch": 0.5479704797047971,
+      "grad_norm": 0.4238342532266894,
+      "learning_rate": 4.897974823835131e-05,
+      "loss": 0.2528,
+      "step": 594
+    },
+    {
+      "epoch": 0.5488929889298892,
+      "grad_norm": 0.47695834978409135,
+      "learning_rate": 4.89721444926715e-05,
+      "loss": 0.2873,
+      "step": 595
+    },
+    {
+      "epoch": 0.5498154981549815,
+      "grad_norm": 0.4838457590674877,
+      "learning_rate": 4.896451311208854e-05,
+      "loss": 0.2601,
+      "step": 596
+    },
+    {
+      "epoch": 0.5507380073800738,
+      "grad_norm": 0.4332730618581192,
+      "learning_rate": 4.895685410539983e-05,
+      "loss": 0.2478,
+      "step": 597
+    },
+    {
+      "epoch": 0.551660516605166,
+      "grad_norm": 0.4375563350114113,
+      "learning_rate": 4.894916748143461e-05,
+      "loss": 0.2638,
+      "step": 598
+    },
+    {
+      "epoch": 0.5525830258302583,
+      "grad_norm": 0.45076712957616827,
+      "learning_rate": 4.894145324905396e-05,
+      "loss": 0.2684,
+      "step": 599
+    },
+    {
+      "epoch": 0.5535055350553506,
+      "grad_norm": 0.4636336178782477,
+      "learning_rate": 4.89337114171508e-05,
+      "loss": 0.2721,
+      "step": 600
+    },
+    {
+      "epoch": 0.5544280442804428,
+      "grad_norm": 0.4391886154375099,
+      "learning_rate": 4.892594199464984e-05,
+      "loss": 0.2637,
+      "step": 601
+    },
+    {
+      "epoch": 0.5553505535055351,
+      "grad_norm": 0.43384904186127465,
+      "learning_rate": 4.891814499050762e-05,
+      "loss": 0.2743,
+      "step": 602
+    },
+    {
+      "epoch": 0.5562730627306273,
+      "grad_norm": 0.4524629128775071,
+      "learning_rate": 4.891032041371246e-05,
+      "loss": 0.2697,
+      "step": 603
+    },
+    {
+      "epoch": 0.5571955719557196,
+      "grad_norm": 0.4392319773444855,
+      "learning_rate": 4.8902468273284475e-05,
+      "loss": 0.2702,
+      "step": 604
+    },
+    {
+      "epoch": 0.5581180811808119,
+      "grad_norm": 0.4382414715952929,
+      "learning_rate": 4.8894588578275544e-05,
+      "loss": 0.2442,
+      "step": 605
+    },
+    {
+      "epoch": 0.559040590405904,
+      "grad_norm": 0.4274378197084093,
+      "learning_rate": 4.888668133776934e-05,
+      "loss": 0.2704,
+      "step": 606
+    },
+    {
+      "epoch": 0.5599630996309963,
+      "grad_norm": 0.4486738805320487,
+      "learning_rate": 4.887874656088124e-05,
+      "loss": 0.257,
+      "step": 607
+    },
+    {
+      "epoch": 0.5608856088560885,
+      "grad_norm": 0.46043272436522703,
+      "learning_rate": 4.88707842567584e-05,
+      "loss": 0.2906,
+      "step": 608
+    },
+    {
+      "epoch": 0.5618081180811808,
+      "grad_norm": 0.4479495902950971,
+      "learning_rate": 4.8862794434579726e-05,
+      "loss": 0.2642,
+      "step": 609
+    },
+    {
+      "epoch": 0.5627306273062731,
+      "grad_norm": 0.4497957687832158,
+      "learning_rate": 4.8854777103555804e-05,
+      "loss": 0.2613,
+      "step": 610
+    },
+    {
+      "epoch": 0.5636531365313653,
+      "grad_norm": 0.4203226668395155,
+      "learning_rate": 4.884673227292895e-05,
+      "loss": 0.2425,
+      "step": 611
+    },
+    {
+      "epoch": 0.5645756457564576,
+      "grad_norm": 0.4445645750758397,
+      "learning_rate": 4.883865995197319e-05,
+      "loss": 0.2921,
+      "step": 612
+    },
+    {
+      "epoch": 0.5654981549815498,
+      "grad_norm": 0.5884497201558954,
+      "learning_rate": 4.883056014999423e-05,
+      "loss": 0.29,
+      "step": 613
+    },
+    {
+      "epoch": 0.566420664206642,
+      "grad_norm": 0.3975901321158924,
+      "learning_rate": 4.882243287632947e-05,
+      "loss": 0.2516,
+      "step": 614
+    },
+    {
+      "epoch": 0.5673431734317343,
+      "grad_norm": 0.3938164318797877,
+      "learning_rate": 4.881427814034795e-05,
+      "loss": 0.2498,
+      "step": 615
+    },
+    {
+      "epoch": 0.5682656826568265,
+      "grad_norm": 0.4512900807352941,
+      "learning_rate": 4.880609595145039e-05,
+      "loss": 0.2555,
+      "step": 616
+    },
+    {
+      "epoch": 0.5691881918819188,
+      "grad_norm": 0.39020092611130675,
+      "learning_rate": 4.8797886319069164e-05,
+      "loss": 0.2531,
+      "step": 617
+    },
+    {
+      "epoch": 0.5701107011070111,
+      "grad_norm": 0.460118191888693,
+      "learning_rate": 4.8789649252668267e-05,
+      "loss": 0.2666,
+      "step": 618
+    },
+    {
+      "epoch": 0.5710332103321033,
+      "grad_norm": 0.3934694887996066,
+      "learning_rate": 4.878138476174333e-05,
+      "loss": 0.2503,
+      "step": 619
+    },
+    {
+      "epoch": 0.5719557195571956,
+      "grad_norm": 0.4259182945379848,
+      "learning_rate": 4.877309285582159e-05,
+      "loss": 0.29,
+      "step": 620
+    },
+    {
+      "epoch": 0.5728782287822878,
+      "grad_norm": 0.518219509749948,
+      "learning_rate": 4.8764773544461886e-05,
+      "loss": 0.2675,
+      "step": 621
+    },
+    {
+      "epoch": 0.5738007380073801,
+      "grad_norm": 0.4193442527427061,
+      "learning_rate": 4.875642683725467e-05,
+      "loss": 0.2424,
+      "step": 622
+    },
+    {
+      "epoch": 0.5747232472324724,
+      "grad_norm": 0.4162366808810705,
+      "learning_rate": 4.874805274382196e-05,
+      "loss": 0.2711,
+      "step": 623
+    },
+    {
+      "epoch": 0.5756457564575646,
+      "grad_norm": 0.40037988979271777,
+      "learning_rate": 4.8739651273817335e-05,
+      "loss": 0.266,
+      "step": 624
+    },
+    {
+      "epoch": 0.5765682656826568,
+      "grad_norm": 0.5577640987449097,
+      "learning_rate": 4.8731222436925946e-05,
+      "loss": 0.2472,
+      "step": 625
+    },
+    {
+      "epoch": 0.577490774907749,
+      "grad_norm": 0.4356354148828068,
+      "learning_rate": 4.87227662428645e-05,
+      "loss": 0.2789,
+      "step": 626
+    },
+    {
+      "epoch": 0.5784132841328413,
+      "grad_norm": 0.42006007196665196,
+      "learning_rate": 4.871428270138123e-05,
+      "loss": 0.2682,
+      "step": 627
+    },
+    {
+      "epoch": 0.5793357933579336,
+      "grad_norm": 0.43660375602062723,
+      "learning_rate": 4.870577182225589e-05,
+      "loss": 0.2736,
+      "step": 628
+    },
+    {
+      "epoch": 0.5802583025830258,
+      "grad_norm": 0.39778255799712997,
+      "learning_rate": 4.8697233615299765e-05,
+      "loss": 0.2489,
+      "step": 629
+    },
+    {
+      "epoch": 0.5811808118081181,
+      "grad_norm": 0.4616450616405046,
+      "learning_rate": 4.8688668090355626e-05,
+      "loss": 0.2623,
+      "step": 630
+    },
+    {
+      "epoch": 0.5821033210332104,
+      "grad_norm": 0.4865535225730099,
+      "learning_rate": 4.868007525729775e-05,
+      "loss": 0.2675,
+      "step": 631
+    },
+    {
+      "epoch": 0.5830258302583026,
+      "grad_norm": 0.4568154435508115,
+      "learning_rate": 4.8671455126031896e-05,
+      "loss": 0.2712,
+      "step": 632
+    },
+    {
+      "epoch": 0.5839483394833949,
+      "grad_norm": 0.4223709303814713,
+      "learning_rate": 4.8662807706495264e-05,
+      "loss": 0.2748,
+      "step": 633
+    },
+    {
+      "epoch": 0.584870848708487,
+      "grad_norm": 0.4072049582485937,
+      "learning_rate": 4.865413300865655e-05,
+      "loss": 0.2626,
+      "step": 634
+    },
+    {
+      "epoch": 0.5857933579335793,
+      "grad_norm": 0.43673565956983407,
+      "learning_rate": 4.864543104251587e-05,
+      "loss": 0.2922,
+      "step": 635
+    },
+    {
+      "epoch": 0.5867158671586716,
+      "grad_norm": 0.47814222534404105,
+      "learning_rate": 4.863670181810479e-05,
+      "loss": 0.2847,
+      "step": 636
+    },
+    {
+      "epoch": 0.5876383763837638,
+      "grad_norm": 0.47852847689852657,
+      "learning_rate": 4.862794534548628e-05,
+      "loss": 0.2558,
+      "step": 637
+    },
+    {
+      "epoch": 0.5885608856088561,
+      "grad_norm": 0.4306830707239445,
+      "learning_rate": 4.861916163475475e-05,
+      "loss": 0.2771,
+      "step": 638
+    },
+    {
+      "epoch": 0.5894833948339483,
+      "grad_norm": 0.4313832268646541,
+      "learning_rate": 4.861035069603599e-05,
+      "loss": 0.2464,
+      "step": 639
+    },
+    {
+      "epoch": 0.5904059040590406,
+      "grad_norm": 0.4061200350798131,
+      "learning_rate": 4.860151253948717e-05,
+      "loss": 0.2571,
+      "step": 640
+    },
+    {
+      "epoch": 0.5913284132841329,
+      "grad_norm": 0.4160150362331395,
+      "learning_rate": 4.859264717529686e-05,
+      "loss": 0.2345,
+      "step": 641
+    },
+    {
+      "epoch": 0.5922509225092251,
+      "grad_norm": 0.4886974969126436,
+      "learning_rate": 4.858375461368499e-05,
+      "loss": 0.2596,
+      "step": 642
+    },
+    {
+      "epoch": 0.5931734317343174,
+      "grad_norm": 0.4106943768640075,
+      "learning_rate": 4.8574834864902816e-05,
+      "loss": 0.2263,
+      "step": 643
+    },
+    {
+      "epoch": 0.5940959409594095,
+      "grad_norm": 0.4557579946573745,
+      "learning_rate": 4.856588793923297e-05,
+      "loss": 0.2565,
+      "step": 644
+    },
+    {
+      "epoch": 0.5950184501845018,
+      "grad_norm": 0.4413330841053196,
+      "learning_rate": 4.8556913846989394e-05,
+      "loss": 0.2804,
+      "step": 645
+    },
+    {
+      "epoch": 0.5959409594095941,
+      "grad_norm": 0.38143228991076894,
+      "learning_rate": 4.854791259851735e-05,
+      "loss": 0.2454,
+      "step": 646
+    },
+    {
+      "epoch": 0.5968634686346863,
+      "grad_norm": 0.46491021788622433,
+      "learning_rate": 4.8538884204193426e-05,
+      "loss": 0.2946,
+      "step": 647
+    },
+    {
+      "epoch": 0.5977859778597786,
+      "grad_norm": 0.4521473135967776,
+      "learning_rate": 4.852982867442546e-05,
+      "loss": 0.2673,
+      "step": 648
+    },
+    {
+      "epoch": 0.5987084870848709,
+      "grad_norm": 0.4154034194245415,
+      "learning_rate": 4.8520746019652605e-05,
+      "loss": 0.2607,
+      "step": 649
+    },
+    {
+      "epoch": 0.5996309963099631,
+      "grad_norm": 0.4491837463124351,
+      "learning_rate": 4.8511636250345294e-05,
+      "loss": 0.2858,
+      "step": 650
+    },
+    {
+      "epoch": 0.6005535055350554,
+      "grad_norm": 0.4044243243068525,
+      "learning_rate": 4.850249937700517e-05,
+      "loss": 0.2695,
+      "step": 651
+    },
+    {
+      "epoch": 0.6014760147601476,
+      "grad_norm": 0.430403757037781,
+      "learning_rate": 4.849333541016516e-05,
+      "loss": 0.2715,
+      "step": 652
+    },
+    {
+      "epoch": 0.6023985239852399,
+      "grad_norm": 0.45438956743986286,
+      "learning_rate": 4.8484144360389425e-05,
+      "loss": 0.2622,
+      "step": 653
+    },
+    {
+      "epoch": 0.6033210332103321,
+      "grad_norm": 0.4539691622407592,
+      "learning_rate": 4.847492623827333e-05,
+      "loss": 0.2569,
+      "step": 654
+    },
+    {
+      "epoch": 0.6042435424354243,
+      "grad_norm": 0.46659023853004006,
+      "learning_rate": 4.846568105444345e-05,
+      "loss": 0.2603,
+      "step": 655
+    },
+    {
+      "epoch": 0.6051660516605166,
+      "grad_norm": 0.47799503385597875,
+      "learning_rate": 4.8456408819557564e-05,
+      "loss": 0.2813,
+      "step": 656
+    },
+    {
+      "epoch": 0.6060885608856088,
+      "grad_norm": 0.49542981968905914,
+      "learning_rate": 4.8447109544304636e-05,
+      "loss": 0.3191,
+      "step": 657
+    },
+    {
+      "epoch": 0.6070110701107011,
+      "grad_norm": 0.40468346171857267,
+      "learning_rate": 4.84377832394048e-05,
+      "loss": 0.2524,
+      "step": 658
+    },
+    {
+      "epoch": 0.6079335793357934,
+      "grad_norm": 0.4069524816036995,
+      "learning_rate": 4.8428429915609336e-05,
+      "loss": 0.2643,
+      "step": 659
+    },
+    {
+      "epoch": 0.6088560885608856,
+      "grad_norm": 0.40713370506278707,
+      "learning_rate": 4.8419049583700696e-05,
+      "loss": 0.2681,
+      "step": 660
+    },
+    {
+      "epoch": 0.6097785977859779,
+      "grad_norm": 0.4519195307988597,
+      "learning_rate": 4.840964225449245e-05,
+      "loss": 0.2779,
+      "step": 661
+    },
+    {
+      "epoch": 0.6107011070110702,
+      "grad_norm": 0.4558992574499373,
+      "learning_rate": 4.84002079388293e-05,
+      "loss": 0.2811,
+      "step": 662
+    },
+    {
+      "epoch": 0.6116236162361623,
+      "grad_norm": 0.3903183870666297,
+      "learning_rate": 4.839074664758704e-05,
+      "loss": 0.2579,
+      "step": 663
+    },
+    {
+      "epoch": 0.6125461254612546,
+      "grad_norm": 0.3808535356695177,
+      "learning_rate": 4.838125839167259e-05,
+      "loss": 0.2498,
+      "step": 664
+    },
+    {
+      "epoch": 0.6134686346863468,
+      "grad_norm": 0.41419592574548914,
+      "learning_rate": 4.837174318202392e-05,
+      "loss": 0.2699,
+      "step": 665
+    },
+    {
+      "epoch": 0.6143911439114391,
+      "grad_norm": 0.43374303904431827,
+      "learning_rate": 4.836220102961011e-05,
+      "loss": 0.2745,
+      "step": 666
+    },
+    {
+      "epoch": 0.6153136531365314,
+      "grad_norm": 0.4469220968657081,
+      "learning_rate": 4.835263194543126e-05,
+      "loss": 0.2857,
+      "step": 667
+    },
+    {
+      "epoch": 0.6162361623616236,
+      "grad_norm": 0.4620147290242077,
+      "learning_rate": 4.834303594051854e-05,
+      "loss": 0.2621,
+      "step": 668
+    },
+    {
+      "epoch": 0.6171586715867159,
+      "grad_norm": 0.4216662617349631,
+      "learning_rate": 4.833341302593417e-05,
+      "loss": 0.2775,
+      "step": 669
+    },
+    {
+      "epoch": 0.6180811808118081,
+      "grad_norm": 0.47013460047624434,
+      "learning_rate": 4.8323763212771354e-05,
+      "loss": 0.3009,
+      "step": 670
+    },
+    {
+      "epoch": 0.6190036900369004,
+      "grad_norm": 0.401712384643802,
+      "learning_rate": 4.8314086512154325e-05,
+      "loss": 0.266,
+      "step": 671
+    },
+    {
+      "epoch": 0.6199261992619927,
+      "grad_norm": 0.39452738403372956,
+      "learning_rate": 4.83043829352383e-05,
+      "loss": 0.2757,
+      "step": 672
+    },
+    {
+      "epoch": 0.6208487084870848,
+      "grad_norm": 0.4541490118997505,
+      "learning_rate": 4.829465249320951e-05,
+      "loss": 0.2768,
+      "step": 673
+    },
+    {
+      "epoch": 0.6217712177121771,
+      "grad_norm": 0.3987990087632013,
+      "learning_rate": 4.8284895197285116e-05,
+      "loss": 0.2574,
+      "step": 674
+    },
+    {
+      "epoch": 0.6226937269372693,
+      "grad_norm": 0.4191622596571706,
+      "learning_rate": 4.827511105871325e-05,
+      "loss": 0.2453,
+      "step": 675
+    },
+    {
+      "epoch": 0.6236162361623616,
+      "grad_norm": 0.40689152241903564,
+      "learning_rate": 4.826530008877301e-05,
+      "loss": 0.2709,
+      "step": 676
+    },
+    {
+      "epoch": 0.6245387453874539,
+      "grad_norm": 0.424621539220838,
+      "learning_rate": 4.825546229877439e-05,
+      "loss": 0.253,
+      "step": 677
+    },
+    {
+      "epoch": 0.6254612546125461,
+      "grad_norm": 0.4292291438742583,
+      "learning_rate": 4.824559770005833e-05,
+      "loss": 0.2763,
+      "step": 678
+    },
+    {
+      "epoch": 0.6263837638376384,
+      "grad_norm": 0.3954778440075958,
+      "learning_rate": 4.823570630399665e-05,
+      "loss": 0.2345,
+      "step": 679
+    },
+    {
+      "epoch": 0.6273062730627307,
+      "grad_norm": 0.6001086880801821,
+      "learning_rate": 4.822578812199208e-05,
+      "loss": 0.2674,
+      "step": 680
+    },
+    {
+      "epoch": 0.6282287822878229,
+      "grad_norm": 0.40696009845252346,
+      "learning_rate": 4.821584316547824e-05,
+      "loss": 0.2796,
+      "step": 681
+    },
+    {
+      "epoch": 0.6291512915129152,
+      "grad_norm": 0.41493478120098126,
+      "learning_rate": 4.820587144591957e-05,
+      "loss": 0.2739,
+      "step": 682
+    },
+    {
+      "epoch": 0.6300738007380073,
+      "grad_norm": 0.4095714288542147,
+      "learning_rate": 4.819587297481141e-05,
+      "loss": 0.2797,
+      "step": 683
+    },
+    {
+      "epoch": 0.6309963099630996,
+      "grad_norm": 0.46390279747131585,
+      "learning_rate": 4.818584776367993e-05,
+      "loss": 0.2716,
+      "step": 684
+    },
+    {
+      "epoch": 0.6319188191881919,
+      "grad_norm": 0.40183433756081804,
+      "learning_rate": 4.817579582408208e-05,
+      "loss": 0.2501,
+      "step": 685
+    },
+    {
+      "epoch": 0.6328413284132841,
+      "grad_norm": 0.37094621112893805,
+      "learning_rate": 4.8165717167605694e-05,
+      "loss": 0.2356,
+      "step": 686
+    },
+    {
+      "epoch": 0.6337638376383764,
+      "grad_norm": 0.4427415896084445,
+      "learning_rate": 4.815561180586936e-05,
+      "loss": 0.2584,
+      "step": 687
+    },
+    {
+      "epoch": 0.6346863468634686,
+      "grad_norm": 0.3941215954984464,
+      "learning_rate": 4.814547975052245e-05,
+      "loss": 0.2525,
+      "step": 688
+    },
+    {
+      "epoch": 0.6356088560885609,
+      "grad_norm": 0.381367093297642,
+      "learning_rate": 4.813532101324514e-05,
+      "loss": 0.2563,
+      "step": 689
+    },
+    {
+      "epoch": 0.6365313653136532,
+      "grad_norm": 0.38140384507782954,
+      "learning_rate": 4.8125135605748314e-05,
+      "loss": 0.243,
+      "step": 690
+    },
+    {
+      "epoch": 0.6374538745387454,
+      "grad_norm": 0.4505634418832179,
+      "learning_rate": 4.811492353977366e-05,
+      "loss": 0.2599,
+      "step": 691
+    },
+    {
+      "epoch": 0.6383763837638377,
+      "grad_norm": 0.4528289733421791,
+      "learning_rate": 4.810468482709355e-05,
+      "loss": 0.252,
+      "step": 692
+    },
+    {
+      "epoch": 0.6392988929889298,
+      "grad_norm": 0.46057740923035445,
+      "learning_rate": 4.80944194795111e-05,
+      "loss": 0.252,
+      "step": 693
+    },
+    {
+      "epoch": 0.6402214022140221,
+      "grad_norm": 0.4481110788137588,
+      "learning_rate": 4.808412750886013e-05,
+      "loss": 0.2569,
+      "step": 694
+    },
+    {
+      "epoch": 0.6411439114391144,
+      "grad_norm": 0.44907624769789195,
+      "learning_rate": 4.8073808927005125e-05,
+      "loss": 0.2708,
+      "step": 695
+    },
+    {
+      "epoch": 0.6420664206642066,
+      "grad_norm": 0.4255246259068333,
+      "learning_rate": 4.806346374584129e-05,
+      "loss": 0.2808,
+      "step": 696
+    },
+    {
+      "epoch": 0.6429889298892989,
+      "grad_norm": 0.44614343193719974,
+      "learning_rate": 4.8053091977294456e-05,
+      "loss": 0.2471,
+      "step": 697
+    },
+    {
+      "epoch": 0.6439114391143912,
+      "grad_norm": 0.3894364062272488,
+      "learning_rate": 4.804269363332112e-05,
+      "loss": 0.2385,
+      "step": 698
+    },
+    {
+      "epoch": 0.6448339483394834,
+      "grad_norm": 0.4214046028628608,
+      "learning_rate": 4.803226872590841e-05,
+      "loss": 0.2743,
+      "step": 699
+    },
+    {
+      "epoch": 0.6457564575645757,
+      "grad_norm": 0.4135698277984648,
+      "learning_rate": 4.8021817267074084e-05,
+      "loss": 0.254,
+      "step": 700
+    },
+    {
+      "epoch": 0.6466789667896679,
+      "grad_norm": 0.44485124011855504,
+      "learning_rate": 4.8011339268866505e-05,
+      "loss": 0.2809,
+      "step": 701
+    },
+    {
+      "epoch": 0.6476014760147601,
+      "grad_norm": 0.47409527158946313,
+      "learning_rate": 4.800083474336463e-05,
+      "loss": 0.2565,
+      "step": 702
+    },
+    {
+      "epoch": 0.6485239852398524,
+      "grad_norm": 0.39316505907870625,
+      "learning_rate": 4.7990303702677976e-05,
+      "loss": 0.274,
+      "step": 703
+    },
+    {
+      "epoch": 0.6494464944649446,
+      "grad_norm": 0.4578784316296411,
+      "learning_rate": 4.797974615894667e-05,
+      "loss": 0.2806,
+      "step": 704
+    },
+    {
+      "epoch": 0.6503690036900369,
+      "grad_norm": 0.38103820283177153,
+      "learning_rate": 4.796916212434135e-05,
+      "loss": 0.2297,
+      "step": 705
+    },
+    {
+      "epoch": 0.6512915129151291,
+      "grad_norm": 0.394140134989682,
+      "learning_rate": 4.795855161106322e-05,
+      "loss": 0.2561,
+      "step": 706
+    },
+    {
+      "epoch": 0.6522140221402214,
+      "grad_norm": 0.3976966065031595,
+      "learning_rate": 4.794791463134399e-05,
+      "loss": 0.2821,
+      "step": 707
+    },
+    {
+      "epoch": 0.6531365313653137,
+      "grad_norm": 0.4163144717479845,
+      "learning_rate": 4.7937251197445886e-05,
+      "loss": 0.2421,
+      "step": 708
+    },
+    {
+      "epoch": 0.6540590405904059,
+      "grad_norm": 0.4158977933033687,
+      "learning_rate": 4.7926561321661646e-05,
+      "loss": 0.26,
+      "step": 709
+    },
+    {
+      "epoch": 0.6549815498154982,
+      "grad_norm": 0.4427865277567835,
+      "learning_rate": 4.791584501631447e-05,
+      "loss": 0.2707,
+      "step": 710
+    },
+    {
+      "epoch": 0.6559040590405905,
+      "grad_norm": 0.46200166734221765,
+      "learning_rate": 4.790510229375802e-05,
+      "loss": 0.2909,
+      "step": 711
+    },
+    {
+      "epoch": 0.6568265682656826,
+      "grad_norm": 0.4285513877878524,
+      "learning_rate": 4.789433316637644e-05,
+      "loss": 0.2833,
+      "step": 712
+    },
+    {
+      "epoch": 0.6577490774907749,
+      "grad_norm": 0.4644538463412512,
+      "learning_rate": 4.7883537646584285e-05,
+      "loss": 0.2932,
+      "step": 713
+    },
+    {
+      "epoch": 0.6586715867158671,
+      "grad_norm": 0.42937930052684714,
+      "learning_rate": 4.787271574682656e-05,
+      "loss": 0.2826,
+      "step": 714
+    },
+    {
+      "epoch": 0.6595940959409594,
+      "grad_norm": 0.4186608552294606,
+      "learning_rate": 4.786186747957866e-05,
+      "loss": 0.2436,
+      "step": 715
+    },
+    {
+      "epoch": 0.6605166051660517,
+      "grad_norm": 0.4067942162950451,
+      "learning_rate": 4.785099285734638e-05,
+      "loss": 0.2712,
+      "step": 716
+    },
+    {
+      "epoch": 0.6614391143911439,
+      "grad_norm": 0.40470907170911935,
+      "learning_rate": 4.7840091892665904e-05,
+      "loss": 0.2416,
+      "step": 717
+    },
+    {
+      "epoch": 0.6623616236162362,
+      "grad_norm": 0.37668928961572784,
+      "learning_rate": 4.782916459810378e-05,
+      "loss": 0.2568,
+      "step": 718
+    },
+    {
+      "epoch": 0.6632841328413284,
+      "grad_norm": 0.4176007426824119,
+      "learning_rate": 4.78182109862569e-05,
+      "loss": 0.2656,
+      "step": 719
+    },
+    {
+      "epoch": 0.6642066420664207,
+      "grad_norm": 0.4602953981120611,
+      "learning_rate": 4.7807231069752536e-05,
+      "loss": 0.2703,
+      "step": 720
+    },
+    {
+      "epoch": 0.665129151291513,
+      "grad_norm": 0.45123474203908875,
+      "learning_rate": 4.7796224861248214e-05,
+      "loss": 0.2771,
+      "step": 721
+    },
+    {
+      "epoch": 0.6660516605166051,
+      "grad_norm": 0.3715107067242283,
+      "learning_rate": 4.778519237343182e-05,
+      "loss": 0.2285,
+      "step": 722
+    },
+    {
+      "epoch": 0.6669741697416974,
+      "grad_norm": 0.4295792736840619,
+      "learning_rate": 4.7774133619021514e-05,
+      "loss": 0.2679,
+      "step": 723
+    },
+    {
+      "epoch": 0.6678966789667896,
+      "grad_norm": 0.5222961706066085,
+      "learning_rate": 4.776304861076576e-05,
+      "loss": 0.2422,
+      "step": 724
+    },
+    {
+      "epoch": 0.6688191881918819,
+      "grad_norm": 0.4273618356221251,
+      "learning_rate": 4.775193736144326e-05,
+      "loss": 0.2858,
+      "step": 725
+    },
+    {
+      "epoch": 0.6697416974169742,
+      "grad_norm": 0.4580406802259534,
+      "learning_rate": 4.774079988386296e-05,
+      "loss": 0.2676,
+      "step": 726
+    },
+    {
+      "epoch": 0.6706642066420664,
+      "grad_norm": 0.43261963264608283,
+      "learning_rate": 4.7729636190864085e-05,
+      "loss": 0.2546,
+      "step": 727
+    },
+    {
+      "epoch": 0.6715867158671587,
+      "grad_norm": 0.4023257232553763,
+      "learning_rate": 4.7718446295316044e-05,
+      "loss": 0.2725,
+      "step": 728
+    },
+    {
+      "epoch": 0.672509225092251,
+      "grad_norm": 0.41518162426686667,
+      "learning_rate": 4.770723021011846e-05,
+      "loss": 0.2755,
+      "step": 729
+    },
+    {
+      "epoch": 0.6734317343173432,
+      "grad_norm": 0.40050411915943995,
+      "learning_rate": 4.769598794820114e-05,
+      "loss": 0.2579,
+      "step": 730
+    },
+    {
+      "epoch": 0.6743542435424354,
+      "grad_norm": 0.4016016628876197,
+      "learning_rate": 4.76847195225241e-05,
+      "loss": 0.2746,
+      "step": 731
+    },
+    {
+      "epoch": 0.6752767527675276,
+      "grad_norm": 0.4181553089033055,
+      "learning_rate": 4.7673424946077474e-05,
+      "loss": 0.2638,
+      "step": 732
+    },
+    {
+      "epoch": 0.6761992619926199,
+      "grad_norm": 0.37434014798856674,
+      "learning_rate": 4.7662104231881574e-05,
+      "loss": 0.2499,
+      "step": 733
+    },
+    {
+      "epoch": 0.6771217712177122,
+      "grad_norm": 0.42639642254570276,
+      "learning_rate": 4.765075739298683e-05,
+      "loss": 0.2718,
+      "step": 734
+    },
+    {
+      "epoch": 0.6780442804428044,
+      "grad_norm": 0.4440804113794752,
+      "learning_rate": 4.763938444247378e-05,
+      "loss": 0.2453,
+      "step": 735
+    },
+    {
+      "epoch": 0.6789667896678967,
+      "grad_norm": 0.4192906925708955,
+      "learning_rate": 4.762798539345309e-05,
+      "loss": 0.2885,
+      "step": 736
+    },
+    {
+      "epoch": 0.6798892988929889,
+      "grad_norm": 0.482625341846353,
+      "learning_rate": 4.7616560259065486e-05,
+      "loss": 0.2716,
+      "step": 737
+    },
+    {
+      "epoch": 0.6808118081180812,
+      "grad_norm": 0.4165196350508539,
+      "learning_rate": 4.760510905248177e-05,
+      "loss": 0.2698,
+      "step": 738
+    },
+    {
+      "epoch": 0.6817343173431735,
+      "grad_norm": 0.38642332048977557,
+      "learning_rate": 4.759363178690282e-05,
+      "loss": 0.2637,
+      "step": 739
+    },
+    {
+      "epoch": 0.6826568265682657,
+      "grad_norm": 0.4076451590146458,
+      "learning_rate": 4.758212847555953e-05,
+      "loss": 0.252,
+      "step": 740
+    },
+    {
+      "epoch": 0.683579335793358,
+      "grad_norm": 0.45785352706377896,
+      "learning_rate": 4.757059913171282e-05,
+      "loss": 0.2608,
+      "step": 741
+    },
+    {
+      "epoch": 0.6845018450184502,
+      "grad_norm": 0.4543916372183251,
+      "learning_rate": 4.755904376865364e-05,
+      "loss": 0.2412,
+      "step": 742
+    },
+    {
+      "epoch": 0.6854243542435424,
+      "grad_norm": 0.4356719346714614,
+      "learning_rate": 4.754746239970292e-05,
+      "loss": 0.2443,
+      "step": 743
+    },
+    {
+      "epoch": 0.6863468634686347,
+      "grad_norm": 0.390093208524888,
+      "learning_rate": 4.753585503821157e-05,
+      "loss": 0.2588,
+      "step": 744
+    },
+    {
+      "epoch": 0.6872693726937269,
+      "grad_norm": 0.42573916131343337,
+      "learning_rate": 4.752422169756048e-05,
+      "loss": 0.2507,
+      "step": 745
+    },
+    {
+      "epoch": 0.6881918819188192,
+      "grad_norm": 0.36877784679859316,
+      "learning_rate": 4.751256239116046e-05,
+      "loss": 0.2669,
+      "step": 746
+    },
+    {
+      "epoch": 0.6891143911439115,
+      "grad_norm": 0.3889726868456164,
+      "learning_rate": 4.750087713245227e-05,
+      "loss": 0.2385,
+      "step": 747
+    },
+    {
+      "epoch": 0.6900369003690037,
+      "grad_norm": 0.42525791717206574,
+      "learning_rate": 4.74891659349066e-05,
+      "loss": 0.2601,
+      "step": 748
+    },
+    {
+      "epoch": 0.690959409594096,
+      "grad_norm": 0.42840628871678277,
+      "learning_rate": 4.7477428812024e-05,
+      "loss": 0.281,
+      "step": 749
+    },
+    {
+      "epoch": 0.6918819188191881,
+      "grad_norm": 0.42504599248665903,
+      "learning_rate": 4.746566577733497e-05,
+      "loss": 0.3042,
+      "step": 750
+    },
+    {
+      "epoch": 0.6928044280442804,
+      "grad_norm": 0.42452760937026107,
+      "learning_rate": 4.7453876844399824e-05,
+      "loss": 0.2447,
+      "step": 751
+    },
+    {
+      "epoch": 0.6937269372693727,
+      "grad_norm": 0.4594757150735403,
+      "learning_rate": 4.7442062026808756e-05,
+      "loss": 0.2488,
+      "step": 752
+    },
+    {
+      "epoch": 0.6946494464944649,
+      "grad_norm": 0.410268565822323,
+      "learning_rate": 4.743022133818179e-05,
+      "loss": 0.261,
+      "step": 753
+    },
+    {
+      "epoch": 0.6955719557195572,
+      "grad_norm": 0.4545436622681227,
+      "learning_rate": 4.7418354792168794e-05,
+      "loss": 0.2171,
+      "step": 754
+    },
+    {
+      "epoch": 0.6964944649446494,
+      "grad_norm": 0.40396416205888286,
+      "learning_rate": 4.7406462402449426e-05,
+      "loss": 0.2523,
+      "step": 755
+    },
+    {
+      "epoch": 0.6974169741697417,
+      "grad_norm": 0.3943645189708737,
+      "learning_rate": 4.7394544182733144e-05,
+      "loss": 0.2813,
+      "step": 756
+    },
+    {
+      "epoch": 0.698339483394834,
+      "grad_norm": 0.40320692754111276,
+      "learning_rate": 4.7382600146759174e-05,
+      "loss": 0.2638,
+      "step": 757
+    },
+    {
+      "epoch": 0.6992619926199262,
+      "grad_norm": 0.40025770221592594,
+      "learning_rate": 4.7370630308296505e-05,
+      "loss": 0.2623,
+      "step": 758
+    },
+    {
+      "epoch": 0.7001845018450185,
+      "grad_norm": 0.41160142244941034,
+      "learning_rate": 4.735863468114388e-05,
+      "loss": 0.2878,
+      "step": 759
+    },
+    {
+      "epoch": 0.7011070110701108,
+      "grad_norm": 0.4099065053363144,
+      "learning_rate": 4.734661327912976e-05,
+      "loss": 0.2576,
+      "step": 760
+    },
+    {
+      "epoch": 0.7020295202952029,
+      "grad_norm": 0.3916032675432416,
+      "learning_rate": 4.733456611611233e-05,
+      "loss": 0.2644,
+      "step": 761
+    },
+    {
+      "epoch": 0.7029520295202952,
+      "grad_norm": 0.4127119444291852,
+      "learning_rate": 4.732249320597948e-05,
+      "loss": 0.2597,
+      "step": 762
+    },
+    {
+      "epoch": 0.7038745387453874,
+      "grad_norm": 0.4158472026504385,
+      "learning_rate": 4.731039456264874e-05,
+      "loss": 0.2596,
+      "step": 763
+    },
+    {
+      "epoch": 0.7047970479704797,
+      "grad_norm": 0.44282321593212026,
+      "learning_rate": 4.729827020006735e-05,
+      "loss": 0.2547,
+      "step": 764
+    },
+    {
+      "epoch": 0.705719557195572,
+      "grad_norm": 0.4210528058327828,
+      "learning_rate": 4.7286120132212176e-05,
+      "loss": 0.2665,
+      "step": 765
+    },
+    {
+      "epoch": 0.7066420664206642,
+      "grad_norm": 0.4103670912984461,
+      "learning_rate": 4.7273944373089724e-05,
+      "loss": 0.2479,
+      "step": 766
+    },
+    {
+      "epoch": 0.7075645756457565,
+      "grad_norm": 0.459416682150275,
+      "learning_rate": 4.726174293673612e-05,
+      "loss": 0.2716,
+      "step": 767
+    },
+    {
+      "epoch": 0.7084870848708487,
+      "grad_norm": 0.3766584559940595,
+      "learning_rate": 4.724951583721707e-05,
+      "loss": 0.25,
+      "step": 768
+    },
+    {
+      "epoch": 0.709409594095941,
+      "grad_norm": 0.4111985223859097,
+      "learning_rate": 4.7237263088627905e-05,
+      "loss": 0.2599,
+      "step": 769
+    },
+    {
+      "epoch": 0.7103321033210332,
+      "grad_norm": 0.39176492040102723,
+      "learning_rate": 4.722498470509348e-05,
+      "loss": 0.2336,
+      "step": 770
+    },
+    {
+      "epoch": 0.7112546125461254,
+      "grad_norm": 0.4432494114714155,
+      "learning_rate": 4.721268070076822e-05,
+      "loss": 0.2886,
+      "step": 771
+    },
+    {
+      "epoch": 0.7121771217712177,
+      "grad_norm": 0.42411795005176356,
+      "learning_rate": 4.720035108983609e-05,
+      "loss": 0.2656,
+      "step": 772
+    },
+    {
+      "epoch": 0.7130996309963099,
+      "grad_norm": 0.4078323707771864,
+      "learning_rate": 4.718799588651058e-05,
+      "loss": 0.2735,
+      "step": 773
+    },
+    {
+      "epoch": 0.7140221402214022,
+      "grad_norm": 0.4627564835371624,
+      "learning_rate": 4.717561510503466e-05,
+      "loss": 0.2563,
+      "step": 774
+    },
+    {
+      "epoch": 0.7149446494464945,
+      "grad_norm": 0.41029118918141744,
+      "learning_rate": 4.716320875968081e-05,
+      "loss": 0.2567,
+      "step": 775
+    },
+    {
+      "epoch": 0.7158671586715867,
+      "grad_norm": 0.4205427816104139,
+      "learning_rate": 4.7150776864750956e-05,
+      "loss": 0.263,
+      "step": 776
+    },
+    {
+      "epoch": 0.716789667896679,
+      "grad_norm": 0.5414972668050808,
+      "learning_rate": 4.71383194345765e-05,
+      "loss": 0.2393,
+      "step": 777
+    },
+    {
+      "epoch": 0.7177121771217713,
+      "grad_norm": 0.3885768505484422,
+      "learning_rate": 4.7125836483518276e-05,
+      "loss": 0.2572,
+      "step": 778
+    },
+    {
+      "epoch": 0.7186346863468634,
+      "grad_norm": 0.4161037898367446,
+      "learning_rate": 4.711332802596652e-05,
+      "loss": 0.2506,
+      "step": 779
+    },
+    {
+      "epoch": 0.7195571955719557,
+      "grad_norm": 0.38852555170461517,
+      "learning_rate": 4.7100794076340896e-05,
+      "loss": 0.2526,
+      "step": 780
+    },
+    {
+      "epoch": 0.7204797047970479,
+      "grad_norm": 0.398585165108261,
+      "learning_rate": 4.708823464909045e-05,
+      "loss": 0.229,
+      "step": 781
+    },
+    {
+      "epoch": 0.7214022140221402,
+      "grad_norm": 0.4195912312646104,
+      "learning_rate": 4.7075649758693565e-05,
+      "loss": 0.2544,
+      "step": 782
+    },
+    {
+      "epoch": 0.7223247232472325,
+      "grad_norm": 0.38972929452659466,
+      "learning_rate": 4.7063039419658035e-05,
+      "loss": 0.2686,
+      "step": 783
+    },
+    {
+      "epoch": 0.7232472324723247,
+      "grad_norm": 0.4057648954307941,
+      "learning_rate": 4.7050403646520944e-05,
+      "loss": 0.2521,
+      "step": 784
+    },
+    {
+      "epoch": 0.724169741697417,
+      "grad_norm": 0.3881229279065408,
+      "learning_rate": 4.703774245384873e-05,
+      "loss": 0.2668,
+      "step": 785
+    },
+    {
+      "epoch": 0.7250922509225092,
+      "grad_norm": 0.46964517708980524,
+      "learning_rate": 4.70250558562371e-05,
+      "loss": 0.2724,
+      "step": 786
+    },
+    {
+      "epoch": 0.7260147601476015,
+      "grad_norm": 0.387501710974925,
+      "learning_rate": 4.701234386831108e-05,
+      "loss": 0.2369,
+      "step": 787
+    },
+    {
+      "epoch": 0.7269372693726938,
+      "grad_norm": 0.42745253437867164,
+      "learning_rate": 4.6999606504724944e-05,
+      "loss": 0.2631,
+      "step": 788
+    },
+    {
+      "epoch": 0.727859778597786,
+      "grad_norm": 0.35781009260671326,
+      "learning_rate": 4.698684378016222e-05,
+      "loss": 0.2435,
+      "step": 789
+    },
+    {
+      "epoch": 0.7287822878228782,
+      "grad_norm": 0.41820542281472733,
+      "learning_rate": 4.6974055709335705e-05,
+      "loss": 0.2591,
+      "step": 790
+    },
+    {
+      "epoch": 0.7297047970479705,
+      "grad_norm": 0.4024910437468129,
+      "learning_rate": 4.696124230698736e-05,
+      "loss": 0.249,
+      "step": 791
+    },
+    {
+      "epoch": 0.7306273062730627,
+      "grad_norm": 0.395320824058086,
+      "learning_rate": 4.694840358788839e-05,
+      "loss": 0.2558,
+      "step": 792
+    },
+    {
+      "epoch": 0.731549815498155,
+      "grad_norm": 0.4373305480576554,
+      "learning_rate": 4.693553956683916e-05,
+      "loss": 0.2599,
+      "step": 793
+    },
+    {
+      "epoch": 0.7324723247232472,
+      "grad_norm": 0.41630398217727865,
+      "learning_rate": 4.692265025866923e-05,
+      "loss": 0.2776,
+      "step": 794
+    },
+    {
+      "epoch": 0.7333948339483395,
+      "grad_norm": 0.38719793187775475,
+      "learning_rate": 4.6909735678237284e-05,
+      "loss": 0.2635,
+      "step": 795
+    },
+    {
+      "epoch": 0.7343173431734318,
+      "grad_norm": 0.3749824756461014,
+      "learning_rate": 4.689679584043115e-05,
+      "loss": 0.2453,
+      "step": 796
+    },
+    {
+      "epoch": 0.735239852398524,
+      "grad_norm": 0.6396474148444344,
+      "learning_rate": 4.688383076016778e-05,
+      "loss": 0.2525,
+      "step": 797
+    },
+    {
+      "epoch": 0.7361623616236163,
+      "grad_norm": 0.41477012509152505,
+      "learning_rate": 4.687084045239322e-05,
+      "loss": 0.245,
+      "step": 798
+    },
+    {
+      "epoch": 0.7370848708487084,
+      "grad_norm": 0.3666873244609313,
+      "learning_rate": 4.6857824932082586e-05,
+      "loss": 0.2361,
+      "step": 799
+    },
+    {
+      "epoch": 0.7380073800738007,
+      "grad_norm": 0.3776802233167813,
+      "learning_rate": 4.6844784214240076e-05,
+      "loss": 0.2556,
+      "step": 800
+    },
+    {
+      "epoch": 0.738929889298893,
+      "grad_norm": 0.39241143181846977,
+      "learning_rate": 4.683171831389892e-05,
+      "loss": 0.2504,
+      "step": 801
+    },
+    {
+      "epoch": 0.7398523985239852,
+      "grad_norm": 0.33691745713939714,
+      "learning_rate": 4.681862724612141e-05,
+      "loss": 0.2177,
+      "step": 802
+    },
+    {
+      "epoch": 0.7407749077490775,
+      "grad_norm": 0.36087914762741574,
+      "learning_rate": 4.68055110259988e-05,
+      "loss": 0.2475,
+      "step": 803
+    },
+    {
+      "epoch": 0.7416974169741697,
+      "grad_norm": 0.3803088774224962,
+      "learning_rate": 4.6792369668651384e-05,
+      "loss": 0.2574,
+      "step": 804
+    },
+    {
+      "epoch": 0.742619926199262,
+      "grad_norm": 0.45272098427208934,
+      "learning_rate": 4.6779203189228417e-05,
+      "loss": 0.2966,
+      "step": 805
+    },
+    {
+      "epoch": 0.7435424354243543,
+      "grad_norm": 0.3677471946148316,
+      "learning_rate": 4.6766011602908114e-05,
+      "loss": 0.234,
+      "step": 806
+    },
+    {
+      "epoch": 0.7444649446494465,
+      "grad_norm": 0.42506463356607843,
+      "learning_rate": 4.6752794924897624e-05,
+      "loss": 0.2668,
+      "step": 807
+    },
+    {
+      "epoch": 0.7453874538745388,
+      "grad_norm": 0.3614067518204027,
+      "learning_rate": 4.6739553170433045e-05,
+      "loss": 0.2316,
+      "step": 808
+    },
+    {
+      "epoch": 0.746309963099631,
+      "grad_norm": 0.4618757772460389,
+      "learning_rate": 4.672628635477936e-05,
+      "loss": 0.2652,
+      "step": 809
+    },
+    {
+      "epoch": 0.7472324723247232,
+      "grad_norm": 0.3924011015972566,
+      "learning_rate": 4.671299449323045e-05,
+      "loss": 0.2415,
+      "step": 810
+    },
+    {
+      "epoch": 0.7481549815498155,
+      "grad_norm": 0.406502684771447,
+      "learning_rate": 4.669967760110908e-05,
+      "loss": 0.2582,
+      "step": 811
+    },
+    {
+      "epoch": 0.7490774907749077,
+      "grad_norm": 0.44920857107907514,
+      "learning_rate": 4.668633569376685e-05,
+      "loss": 0.2794,
+      "step": 812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41631190647871946,
+      "learning_rate": 4.667296878658423e-05,
+      "loss": 0.2592,
+      "step": 813
+    },
+    {
+      "epoch": 0.7509225092250923,
+      "grad_norm": 0.3633834642775942,
+      "learning_rate": 4.665957689497045e-05,
+      "loss": 0.2424,
+      "step": 814
+    },
+    {
+      "epoch": 0.7518450184501845,
+      "grad_norm": 0.424235479072311,
+      "learning_rate": 4.664616003436361e-05,
+      "loss": 0.2391,
+      "step": 815
+    },
+    {
+      "epoch": 0.7527675276752768,
+      "grad_norm": 0.40557799197958655,
+      "learning_rate": 4.663271822023055e-05,
+      "loss": 0.2551,
+      "step": 816
+    },
+    {
+      "epoch": 0.753690036900369,
+      "grad_norm": 0.4167542469925017,
+      "learning_rate": 4.66192514680669e-05,
+      "loss": 0.2633,
+      "step": 817
+    },
+    {
+      "epoch": 0.7546125461254612,
+      "grad_norm": 0.4207560114138412,
+      "learning_rate": 4.660575979339701e-05,
+      "loss": 0.2691,
+      "step": 818
+    },
+    {
+      "epoch": 0.7555350553505535,
+      "grad_norm": 0.4183650561450747,
+      "learning_rate": 4.6592243211774e-05,
+      "loss": 0.2702,
+      "step": 819
+    },
+    {
+      "epoch": 0.7564575645756457,
+      "grad_norm": 0.4320429078035428,
+      "learning_rate": 4.657870173877967e-05,
+      "loss": 0.2856,
+      "step": 820
+    },
+    {
+      "epoch": 0.757380073800738,
+      "grad_norm": 0.4243496691378848,
+      "learning_rate": 4.6565135390024515e-05,
+      "loss": 0.2729,
+      "step": 821
+    },
+    {
+      "epoch": 0.7583025830258303,
+      "grad_norm": 0.40806346530521626,
+      "learning_rate": 4.6551544181147744e-05,
+      "loss": 0.2381,
+      "step": 822
+    },
+    {
+      "epoch": 0.7592250922509225,
+      "grad_norm": 0.37050493027188813,
+      "learning_rate": 4.653792812781717e-05,
+      "loss": 0.2472,
+      "step": 823
+    },
+    {
+      "epoch": 0.7601476014760148,
+      "grad_norm": 0.42478917712919395,
+      "learning_rate": 4.6524287245729295e-05,
+      "loss": 0.2696,
+      "step": 824
+    },
+    {
+      "epoch": 0.761070110701107,
+      "grad_norm": 0.39453849669994284,
+      "learning_rate": 4.65106215506092e-05,
+      "loss": 0.2507,
+      "step": 825
+    },
+    {
+      "epoch": 0.7619926199261993,
+      "grad_norm": 0.3985150630506928,
+      "learning_rate": 4.6496931058210615e-05,
+      "loss": 0.2441,
+      "step": 826
+    },
+    {
+      "epoch": 0.7629151291512916,
+      "grad_norm": 0.43537042016071226,
+      "learning_rate": 4.6483215784315826e-05,
+      "loss": 0.26,
+      "step": 827
+    },
+    {
+      "epoch": 0.7638376383763837,
+      "grad_norm": 0.39034045876332363,
+      "learning_rate": 4.646947574473569e-05,
+      "loss": 0.2328,
+      "step": 828
+    },
+    {
+      "epoch": 0.764760147601476,
+      "grad_norm": 0.4179709095045547,
+      "learning_rate": 4.645571095530963e-05,
+      "loss": 0.268,
+      "step": 829
+    },
+    {
+      "epoch": 0.7656826568265682,
+      "grad_norm": 0.39056066444947624,
+      "learning_rate": 4.644192143190558e-05,
+      "loss": 0.2689,
+      "step": 830
+    },
+    {
+      "epoch": 0.7666051660516605,
+      "grad_norm": 0.41534205342903163,
+      "learning_rate": 4.642810719041999e-05,
+      "loss": 0.2419,
+      "step": 831
+    },
+    {
+      "epoch": 0.7675276752767528,
+      "grad_norm": 0.38688656660125625,
+      "learning_rate": 4.6414268246777824e-05,
+      "loss": 0.2533,
+      "step": 832
+    },
+    {
+      "epoch": 0.768450184501845,
+      "grad_norm": 0.38205343347520876,
+      "learning_rate": 4.6400404616932505e-05,
+      "loss": 0.2462,
+      "step": 833
+    },
+    {
+      "epoch": 0.7693726937269373,
+      "grad_norm": 0.38784424587097893,
+      "learning_rate": 4.6386516316865916e-05,
+      "loss": 0.2285,
+      "step": 834
+    },
+    {
+      "epoch": 0.7702952029520295,
+      "grad_norm": 0.45292745730319345,
+      "learning_rate": 4.637260336258838e-05,
+      "loss": 0.2774,
+      "step": 835
+    },
+    {
+      "epoch": 0.7712177121771218,
+      "grad_norm": 0.39512053052710583,
+      "learning_rate": 4.6358665770138664e-05,
+      "loss": 0.2568,
+      "step": 836
+    },
+    {
+      "epoch": 0.772140221402214,
+      "grad_norm": 0.43157081923032486,
+      "learning_rate": 4.6344703555583884e-05,
+      "loss": 0.2372,
+      "step": 837
+    },
+    {
+      "epoch": 0.7730627306273062,
+      "grad_norm": 0.3467182809262665,
+      "learning_rate": 4.63307167350196e-05,
+      "loss": 0.2308,
+      "step": 838
+    },
+    {
+      "epoch": 0.7739852398523985,
+      "grad_norm": 0.41682462805052434,
+      "learning_rate": 4.6316705324569687e-05,
+      "loss": 0.2481,
+      "step": 839
+    },
+    {
+      "epoch": 0.7749077490774908,
+      "grad_norm": 0.4292084985479234,
+      "learning_rate": 4.630266934038642e-05,
+      "loss": 0.2765,
+      "step": 840
+    },
+    {
+      "epoch": 0.775830258302583,
+      "grad_norm": 0.40443659411602667,
+      "learning_rate": 4.628860879865035e-05,
+      "loss": 0.2575,
+      "step": 841
+    },
+    {
+      "epoch": 0.7767527675276753,
+      "grad_norm": 0.44602392736308993,
+      "learning_rate": 4.627452371557036e-05,
+      "loss": 0.2345,
+      "step": 842
+    },
+    {
+      "epoch": 0.7776752767527675,
+      "grad_norm": 0.39494373687964907,
+      "learning_rate": 4.6260414107383646e-05,
+      "loss": 0.2538,
+      "step": 843
+    },
+    {
+      "epoch": 0.7785977859778598,
+      "grad_norm": 0.39973400621846794,
+      "learning_rate": 4.624627999035563e-05,
+      "loss": 0.2567,
+      "step": 844
+    },
+    {
+      "epoch": 0.7795202952029521,
+      "grad_norm": 0.3926013086674455,
+      "learning_rate": 4.6232121380780034e-05,
+      "loss": 0.2263,
+      "step": 845
+    },
+    {
+      "epoch": 0.7804428044280443,
+      "grad_norm": 0.3712337226687765,
+      "learning_rate": 4.621793829497879e-05,
+      "loss": 0.2214,
+      "step": 846
+    },
+    {
+      "epoch": 0.7813653136531366,
+      "grad_norm": 0.3842637168616991,
+      "learning_rate": 4.6203730749302043e-05,
+      "loss": 0.2395,
+      "step": 847
+    },
+    {
+      "epoch": 0.7822878228782287,
+      "grad_norm": 0.3549206951977871,
+      "learning_rate": 4.6189498760128136e-05,
+      "loss": 0.2417,
+      "step": 848
+    },
+    {
+      "epoch": 0.783210332103321,
+      "grad_norm": 0.37894061281979396,
+      "learning_rate": 4.617524234386361e-05,
+      "loss": 0.2378,
+      "step": 849
+    },
+    {
+      "epoch": 0.7841328413284133,
+      "grad_norm": 0.4168344672554966,
+      "learning_rate": 4.6160961516943145e-05,
+      "loss": 0.2354,
+      "step": 850
+    },
+    {
+      "epoch": 0.7850553505535055,
+      "grad_norm": 0.4080512446894384,
+      "learning_rate": 4.614665629582958e-05,
+      "loss": 0.2468,
+      "step": 851
+    },
+    {
+      "epoch": 0.7859778597785978,
+      "grad_norm": 0.40911776930832033,
+      "learning_rate": 4.613232669701384e-05,
+      "loss": 0.2691,
+      "step": 852
+    },
+    {
+      "epoch": 0.7869003690036901,
+      "grad_norm": 0.39657146630350776,
+      "learning_rate": 4.6117972737014993e-05,
+      "loss": 0.2588,
+      "step": 853
+    },
+    {
+      "epoch": 0.7878228782287823,
+      "grad_norm": 0.3676678279956454,
+      "learning_rate": 4.610359443238017e-05,
+      "loss": 0.2633,
+      "step": 854
+    },
+    {
+      "epoch": 0.7887453874538746,
+      "grad_norm": 0.4220328616701844,
+      "learning_rate": 4.608919179968457e-05,
+      "loss": 0.2179,
+      "step": 855
+    },
+    {
+      "epoch": 0.7896678966789668,
+      "grad_norm": 0.3873623022895507,
+      "learning_rate": 4.6074764855531435e-05,
+      "loss": 0.255,
+      "step": 856
+    },
+    {
+      "epoch": 0.790590405904059,
+      "grad_norm": 0.39345781924934875,
+      "learning_rate": 4.606031361655203e-05,
+      "loss": 0.2321,
+      "step": 857
+    },
+    {
+      "epoch": 0.7915129151291513,
+      "grad_norm": 0.3641248075979448,
+      "learning_rate": 4.604583809940565e-05,
+      "loss": 0.2359,
+      "step": 858
+    },
+    {
+      "epoch": 0.7924354243542435,
+      "grad_norm": 0.3795834365154066,
+      "learning_rate": 4.6031338320779534e-05,
+      "loss": 0.2187,
+      "step": 859
+    },
+    {
+      "epoch": 0.7933579335793358,
+      "grad_norm": 0.3748626112020789,
+      "learning_rate": 4.601681429738893e-05,
+      "loss": 0.2537,
+      "step": 860
+    },
+    {
+      "epoch": 0.794280442804428,
+      "grad_norm": 0.38057309992880556,
+      "learning_rate": 4.6002266045977015e-05,
+      "loss": 0.2633,
+      "step": 861
+    },
+    {
+      "epoch": 0.7952029520295203,
+      "grad_norm": 0.39693638718508983,
+      "learning_rate": 4.598769358331491e-05,
+      "loss": 0.2528,
+      "step": 862
+    },
+    {
+      "epoch": 0.7961254612546126,
+      "grad_norm": 0.41350889386382905,
+      "learning_rate": 4.597309692620163e-05,
+      "loss": 0.2627,
+      "step": 863
+    },
+    {
+      "epoch": 0.7970479704797048,
+      "grad_norm": 0.38247449539113726,
+      "learning_rate": 4.5958476091464086e-05,
+      "loss": 0.2568,
+      "step": 864
+    },
+    {
+      "epoch": 0.7979704797047971,
+      "grad_norm": 0.4104295900334278,
+      "learning_rate": 4.5943831095957066e-05,
+      "loss": 0.2742,
+      "step": 865
+    },
+    {
+      "epoch": 0.7988929889298892,
+      "grad_norm": 0.3959010980286953,
+      "learning_rate": 4.592916195656322e-05,
+      "loss": 0.2423,
+      "step": 866
+    },
+    {
+      "epoch": 0.7998154981549815,
+      "grad_norm": 0.397778263487737,
+      "learning_rate": 4.5914468690192994e-05,
+      "loss": 0.2829,
+      "step": 867
+    },
+    {
+      "epoch": 0.8007380073800738,
+      "grad_norm": 0.3589113580764412,
+      "learning_rate": 4.5899751313784693e-05,
+      "loss": 0.2367,
+      "step": 868
+    },
+    {
+      "epoch": 0.801660516605166,
+      "grad_norm": 0.35817034843856266,
+      "learning_rate": 4.5885009844304386e-05,
+      "loss": 0.2241,
+      "step": 869
+    },
+    {
+      "epoch": 0.8025830258302583,
+      "grad_norm": 0.4181099949219438,
+      "learning_rate": 4.5870244298745926e-05,
+      "loss": 0.2649,
+      "step": 870
+    },
+    {
+      "epoch": 0.8035055350553506,
+      "grad_norm": 0.41552754910798373,
+      "learning_rate": 4.585545469413092e-05,
+      "loss": 0.263,
+      "step": 871
+    },
+    {
+      "epoch": 0.8044280442804428,
+      "grad_norm": 0.4000016435818405,
+      "learning_rate": 4.584064104750872e-05,
+      "loss": 0.2596,
+      "step": 872
+    },
+    {
+      "epoch": 0.8053505535055351,
+      "grad_norm": 0.4301391513536784,
+      "learning_rate": 4.582580337595636e-05,
+      "loss": 0.2369,
+      "step": 873
+    },
+    {
+      "epoch": 0.8062730627306273,
+      "grad_norm": 0.3771770992053558,
+      "learning_rate": 4.5810941696578616e-05,
+      "loss": 0.2679,
+      "step": 874
+    },
+    {
+      "epoch": 0.8071955719557196,
+      "grad_norm": 0.3681314252976609,
+      "learning_rate": 4.57960560265079e-05,
+      "loss": 0.2453,
+      "step": 875
+    },
+    {
+      "epoch": 0.8081180811808119,
+      "grad_norm": 0.3960434602601055,
+      "learning_rate": 4.5781146382904314e-05,
+      "loss": 0.2459,
+      "step": 876
+    },
+    {
+      "epoch": 0.809040590405904,
+      "grad_norm": 0.3651264724900847,
+      "learning_rate": 4.576621278295558e-05,
+      "loss": 0.2293,
+      "step": 877
+    },
+    {
+      "epoch": 0.8099630996309963,
+      "grad_norm": 0.3645341896411552,
+      "learning_rate": 4.5751255243877015e-05,
+      "loss": 0.2426,
+      "step": 878
+    },
+    {
+      "epoch": 0.8108856088560885,
+      "grad_norm": 0.37810515308621573,
+      "learning_rate": 4.5736273782911575e-05,
+      "loss": 0.2508,
+      "step": 879
+    },
+    {
+      "epoch": 0.8118081180811808,
+      "grad_norm": 0.4134745374916864,
+      "learning_rate": 4.572126841732976e-05,
+      "loss": 0.2536,
+      "step": 880
+    },
+    {
+      "epoch": 0.8127306273062731,
+      "grad_norm": 0.3897201458683745,
+      "learning_rate": 4.570623916442966e-05,
+      "loss": 0.2801,
+      "step": 881
+    },
+    {
+      "epoch": 0.8136531365313653,
+      "grad_norm": 0.3499836209573072,
+      "learning_rate": 4.569118604153686e-05,
+      "loss": 0.2224,
+      "step": 882
+    },
+    {
+      "epoch": 0.8145756457564576,
+      "grad_norm": 0.3991348914664004,
+      "learning_rate": 4.567610906600449e-05,
+      "loss": 0.2454,
+      "step": 883
+    },
+    {
+      "epoch": 0.8154981549815498,
+      "grad_norm": 0.39310366746476055,
+      "learning_rate": 4.566100825521317e-05,
+      "loss": 0.2343,
+      "step": 884
+    },
+    {
+      "epoch": 0.816420664206642,
+      "grad_norm": 0.4209272421254392,
+      "learning_rate": 4.564588362657101e-05,
+      "loss": 0.2634,
+      "step": 885
+    },
+    {
+      "epoch": 0.8173431734317343,
+      "grad_norm": 0.36902568189665436,
+      "learning_rate": 4.5630735197513554e-05,
+      "loss": 0.2492,
+      "step": 886
+    },
+    {
+      "epoch": 0.8182656826568265,
+      "grad_norm": 0.35839307483214716,
+      "learning_rate": 4.561556298550379e-05,
+      "loss": 0.2408,
+      "step": 887
+    },
+    {
+      "epoch": 0.8191881918819188,
+      "grad_norm": 0.3632283103696501,
+      "learning_rate": 4.560036700803213e-05,
+      "loss": 0.2124,
+      "step": 888
+    },
+    {
+      "epoch": 0.8201107011070111,
+      "grad_norm": 0.3683588449535609,
+      "learning_rate": 4.558514728261639e-05,
+      "loss": 0.2399,
+      "step": 889
+    },
+    {
+      "epoch": 0.8210332103321033,
+      "grad_norm": 0.416135107412475,
+      "learning_rate": 4.556990382680174e-05,
+      "loss": 0.2488,
+      "step": 890
+    },
+    {
+      "epoch": 0.8219557195571956,
+      "grad_norm": 0.38767776714892904,
+      "learning_rate": 4.555463665816073e-05,
+      "loss": 0.2451,
+      "step": 891
+    },
+    {
+      "epoch": 0.8228782287822878,
+      "grad_norm": 0.38655898937170086,
+      "learning_rate": 4.553934579429322e-05,
+      "loss": 0.2551,
+      "step": 892
+    },
+    {
+      "epoch": 0.8238007380073801,
+      "grad_norm": 0.4237419311315628,
+      "learning_rate": 4.552403125282641e-05,
+      "loss": 0.2788,
+      "step": 893
+    },
+    {
+      "epoch": 0.8247232472324724,
+      "grad_norm": 0.41798421330220237,
+      "learning_rate": 4.550869305141478e-05,
+      "loss": 0.2413,
+      "step": 894
+    },
+    {
+      "epoch": 0.8256457564575646,
+      "grad_norm": 0.41305156789240494,
+      "learning_rate": 4.54933312077401e-05,
+      "loss": 0.2372,
+      "step": 895
+    },
+    {
+      "epoch": 0.8265682656826568,
+      "grad_norm": 0.4248949007952841,
+      "learning_rate": 4.547794573951136e-05,
+      "loss": 0.2637,
+      "step": 896
+    },
+    {
+      "epoch": 0.827490774907749,
+      "grad_norm": 0.36047701179072617,
+      "learning_rate": 4.546253666446484e-05,
+      "loss": 0.2558,
+      "step": 897
+    },
+    {
+      "epoch": 0.8284132841328413,
+      "grad_norm": 0.43732481006930657,
+      "learning_rate": 4.5447104000363985e-05,
+      "loss": 0.2856,
+      "step": 898
+    },
+    {
+      "epoch": 0.8293357933579336,
+      "grad_norm": 0.4187175037943844,
+      "learning_rate": 4.5431647764999455e-05,
+      "loss": 0.2464,
+      "step": 899
+    },
+    {
+      "epoch": 0.8302583025830258,
+      "grad_norm": 0.3940722356549932,
+      "learning_rate": 4.541616797618907e-05,
+      "loss": 0.2404,
+      "step": 900
+    },
+    {
+      "epoch": 0.8311808118081181,
+      "grad_norm": 0.39931432095847397,
+      "learning_rate": 4.5400664651777835e-05,
+      "loss": 0.2628,
+      "step": 901
+    },
+    {
+      "epoch": 0.8321033210332104,
+      "grad_norm": 0.410685067676723,
+      "learning_rate": 4.538513780963784e-05,
+      "loss": 0.2384,
+      "step": 902
+    },
+    {
+      "epoch": 0.8330258302583026,
+      "grad_norm": 0.42948688546306324,
+      "learning_rate": 4.5369587467668315e-05,
+      "loss": 0.2486,
+      "step": 903
+    },
+    {
+      "epoch": 0.8339483394833949,
+      "grad_norm": 0.3922072991864588,
+      "learning_rate": 4.535401364379558e-05,
+      "loss": 0.2296,
+      "step": 904
+    },
+    {
+      "epoch": 0.834870848708487,
+      "grad_norm": 0.43030836009249707,
+      "learning_rate": 4.5338416355973006e-05,
+      "loss": 0.269,
+      "step": 905
+    },
+    {
+      "epoch": 0.8357933579335793,
+      "grad_norm": 0.4216060941407115,
+      "learning_rate": 4.5322795622181044e-05,
+      "loss": 0.2564,
+      "step": 906
+    },
+    {
+      "epoch": 0.8367158671586716,
+      "grad_norm": 0.42802377697733435,
+      "learning_rate": 4.530715146042713e-05,
+      "loss": 0.2479,
+      "step": 907
+    },
+    {
+      "epoch": 0.8376383763837638,
+      "grad_norm": 0.4266861366633069,
+      "learning_rate": 4.529148388874577e-05,
+      "loss": 0.2652,
+      "step": 908
+    },
+    {
+      "epoch": 0.8385608856088561,
+      "grad_norm": 0.41261864191830294,
+      "learning_rate": 4.5275792925198383e-05,
+      "loss": 0.2716,
+      "step": 909
+    },
+    {
+      "epoch": 0.8394833948339483,
+      "grad_norm": 0.3646033560843698,
+      "learning_rate": 4.526007858787341e-05,
+      "loss": 0.2242,
+      "step": 910
+    },
+    {
+      "epoch": 0.8404059040590406,
+      "grad_norm": 0.4570957874662083,
+      "learning_rate": 4.5244340894886215e-05,
+      "loss": 0.2694,
+      "step": 911
+    },
+    {
+      "epoch": 0.8413284132841329,
+      "grad_norm": 0.3911930239160912,
+      "learning_rate": 4.522857986437909e-05,
+      "loss": 0.2566,
+      "step": 912
+    },
+    {
+      "epoch": 0.8422509225092251,
+      "grad_norm": 0.36326227597308736,
+      "learning_rate": 4.521279551452122e-05,
+      "loss": 0.2629,
+      "step": 913
+    },
+    {
+      "epoch": 0.8431734317343174,
+      "grad_norm": 0.37918166872103803,
+      "learning_rate": 4.51969878635087e-05,
+      "loss": 0.238,
+      "step": 914
+    },
+    {
+      "epoch": 0.8440959409594095,
+      "grad_norm": 0.43316472238041337,
+      "learning_rate": 4.518115692956445e-05,
+      "loss": 0.2676,
+      "step": 915
+    },
+    {
+      "epoch": 0.8450184501845018,
+      "grad_norm": 0.42341768992528533,
+      "learning_rate": 4.516530273093825e-05,
+      "loss": 0.2394,
+      "step": 916
+    },
+    {
+      "epoch": 0.8459409594095941,
+      "grad_norm": 0.44622071611241565,
+      "learning_rate": 4.514942528590671e-05,
+      "loss": 0.2486,
+      "step": 917
+    },
+    {
+      "epoch": 0.8468634686346863,
+      "grad_norm": 0.35345649117366434,
+      "learning_rate": 4.513352461277323e-05,
+      "loss": 0.246,
+      "step": 918
+    },
+    {
+      "epoch": 0.8477859778597786,
+      "grad_norm": 0.3880135352024817,
+      "learning_rate": 4.511760072986795e-05,
+      "loss": 0.2601,
+      "step": 919
+    },
+    {
+      "epoch": 0.8487084870848709,
+      "grad_norm": 0.3965918262583217,
+      "learning_rate": 4.5101653655547834e-05,
+      "loss": 0.2701,
+      "step": 920
+    },
+    {
+      "epoch": 0.8496309963099631,
+      "grad_norm": 0.4381694713773363,
+      "learning_rate": 4.5085683408196535e-05,
+      "loss": 0.2731,
+      "step": 921
+    },
+    {
+      "epoch": 0.8505535055350554,
+      "grad_norm": 0.3650583316068512,
+      "learning_rate": 4.5069690006224424e-05,
+      "loss": 0.2383,
+      "step": 922
+    },
+    {
+      "epoch": 0.8514760147601476,
+      "grad_norm": 0.46663226029659377,
+      "learning_rate": 4.505367346806858e-05,
+      "loss": 0.2495,
+      "step": 923
+    },
+    {
+      "epoch": 0.8523985239852399,
+      "grad_norm": 0.3775521089743212,
+      "learning_rate": 4.503763381219275e-05,
+      "loss": 0.2545,
+      "step": 924
+    },
+    {
+      "epoch": 0.8533210332103321,
+      "grad_norm": 0.3597708317746619,
+      "learning_rate": 4.502157105708731e-05,
+      "loss": 0.2513,
+      "step": 925
+    },
+    {
+      "epoch": 0.8542435424354243,
+      "grad_norm": 0.40647398411997965,
+      "learning_rate": 4.5005485221269285e-05,
+      "loss": 0.2476,
+      "step": 926
+    },
+    {
+      "epoch": 0.8551660516605166,
+      "grad_norm": 0.33447531379059603,
+      "learning_rate": 4.498937632328231e-05,
+      "loss": 0.2145,
+      "step": 927
+    },
+    {
+      "epoch": 0.8560885608856088,
+      "grad_norm": 0.3512162886026165,
+      "learning_rate": 4.4973244381696585e-05,
+      "loss": 0.254,
+      "step": 928
+    },
+    {
+      "epoch": 0.8570110701107011,
+      "grad_norm": 0.35575023354053653,
+      "learning_rate": 4.49570894151089e-05,
+      "loss": 0.2421,
+      "step": 929
+    },
+    {
+      "epoch": 0.8579335793357934,
+      "grad_norm": 0.32325446414884945,
+      "learning_rate": 4.494091144214258e-05,
+      "loss": 0.2154,
+      "step": 930
+    },
+    {
+      "epoch": 0.8588560885608856,
+      "grad_norm": 0.4100276190850683,
+      "learning_rate": 4.492471048144744e-05,
+      "loss": 0.2787,
+      "step": 931
+    },
+    {
+      "epoch": 0.8597785977859779,
+      "grad_norm": 0.38579462589053465,
+      "learning_rate": 4.490848655169986e-05,
+      "loss": 0.2248,
+      "step": 932
+    },
+    {
+      "epoch": 0.8607011070110702,
+      "grad_norm": 0.3981022513100881,
+      "learning_rate": 4.489223967160263e-05,
+      "loss": 0.2792,
+      "step": 933
+    },
+    {
+      "epoch": 0.8616236162361623,
+      "grad_norm": 0.3558274930747902,
+      "learning_rate": 4.487596985988505e-05,
+      "loss": 0.2283,
+      "step": 934
+    },
+    {
+      "epoch": 0.8625461254612546,
+      "grad_norm": 0.36787697605706837,
+      "learning_rate": 4.485967713530281e-05,
+      "loss": 0.2542,
+      "step": 935
+    },
+    {
+      "epoch": 0.8634686346863468,
+      "grad_norm": 0.3990125166173122,
+      "learning_rate": 4.4843361516638075e-05,
+      "loss": 0.262,
+      "step": 936
+    },
+    {
+      "epoch": 0.8643911439114391,
+      "grad_norm": 0.42844550257397196,
+      "learning_rate": 4.4827023022699323e-05,
+      "loss": 0.2717,
+      "step": 937
+    },
+    {
+      "epoch": 0.8653136531365314,
+      "grad_norm": 0.4345063745058787,
+      "learning_rate": 4.4810661672321466e-05,
+      "loss": 0.2602,
+      "step": 938
+    },
+    {
+      "epoch": 0.8662361623616236,
+      "grad_norm": 0.38743451727159217,
+      "learning_rate": 4.4794277484365724e-05,
+      "loss": 0.2461,
+      "step": 939
+    },
+    {
+      "epoch": 0.8671586715867159,
+      "grad_norm": 0.3797476635576325,
+      "learning_rate": 4.477787047771969e-05,
+      "loss": 0.2194,
+      "step": 940
+    },
+    {
+      "epoch": 0.8680811808118081,
+      "grad_norm": 0.4388668977771248,
+      "learning_rate": 4.476144067129722e-05,
+      "loss": 0.2822,
+      "step": 941
+    },
+    {
+      "epoch": 0.8690036900369004,
+      "grad_norm": 0.3420994737833039,
+      "learning_rate": 4.474498808403846e-05,
+      "loss": 0.2155,
+      "step": 942
+    },
+    {
+      "epoch": 0.8699261992619927,
+      "grad_norm": 0.3750141322967599,
+      "learning_rate": 4.4728512734909844e-05,
+      "loss": 0.2319,
+      "step": 943
+    },
+    {
+      "epoch": 0.8708487084870848,
+      "grad_norm": 0.34705027085406503,
+      "learning_rate": 4.471201464290401e-05,
+      "loss": 0.2106,
+      "step": 944
+    },
+    {
+      "epoch": 0.8717712177121771,
+      "grad_norm": 0.3979496693198322,
+      "learning_rate": 4.4695493827039846e-05,
+      "loss": 0.2379,
+      "step": 945
+    },
+    {
+      "epoch": 0.8726937269372693,
+      "grad_norm": 0.379676451402185,
+      "learning_rate": 4.4678950306362405e-05,
+      "loss": 0.2554,
+      "step": 946
+    },
+    {
+      "epoch": 0.8736162361623616,
+      "grad_norm": 0.41340574010421133,
+      "learning_rate": 4.4662384099942946e-05,
+      "loss": 0.2688,
+      "step": 947
+    },
+    {
+      "epoch": 0.8745387453874539,
+      "grad_norm": 0.3408035104441574,
+      "learning_rate": 4.464579522687885e-05,
+      "loss": 0.2254,
+      "step": 948
+    },
+    {
+      "epoch": 0.8754612546125461,
+      "grad_norm": 0.3967662981537898,
+      "learning_rate": 4.462918370629365e-05,
+      "loss": 0.2745,
+      "step": 949
+    },
+    {
+      "epoch": 0.8763837638376384,
+      "grad_norm": 0.40779663351573675,
+      "learning_rate": 4.4612549557336974e-05,
+      "loss": 0.266,
+      "step": 950
+    },
+    {
+      "epoch": 0.8773062730627307,
+      "grad_norm": 0.38684475616778263,
+      "learning_rate": 4.4595892799184546e-05,
+      "loss": 0.2559,
+      "step": 951
+    },
+    {
+      "epoch": 0.8782287822878229,
+      "grad_norm": 0.35900223578342305,
+      "learning_rate": 4.457921345103815e-05,
+      "loss": 0.2714,
+      "step": 952
+    },
+    {
+      "epoch": 0.8791512915129152,
+      "grad_norm": 0.42653085046234424,
+      "learning_rate": 4.456251153212561e-05,
+      "loss": 0.2694,
+      "step": 953
+    },
+    {
+      "epoch": 0.8800738007380073,
+      "grad_norm": 0.48911189024620294,
+      "learning_rate": 4.454578706170075e-05,
+      "loss": 0.2365,
+      "step": 954
+    },
+    {
+      "epoch": 0.8809963099630996,
+      "grad_norm": 0.3699032935486172,
+      "learning_rate": 4.4529040059043424e-05,
+      "loss": 0.212,
+      "step": 955
+    },
+    {
+      "epoch": 0.8819188191881919,
+      "grad_norm": 0.3603449065440313,
+      "learning_rate": 4.451227054345946e-05,
+      "loss": 0.2258,
+      "step": 956
+    },
+    {
+      "epoch": 0.8828413284132841,
+      "grad_norm": 0.3805358233443923,
+      "learning_rate": 4.449547853428061e-05,
+      "loss": 0.2197,
+      "step": 957
+    },
+    {
+      "epoch": 0.8837638376383764,
+      "grad_norm": 0.34817693270934846,
+      "learning_rate": 4.4478664050864586e-05,
+      "loss": 0.241,
+      "step": 958
+    },
+    {
+      "epoch": 0.8846863468634686,
+      "grad_norm": 0.4123436517213217,
+      "learning_rate": 4.4461827112594974e-05,
+      "loss": 0.258,
+      "step": 959
+    },
+    {
+      "epoch": 0.8856088560885609,
+      "grad_norm": 0.36338763462519325,
+      "learning_rate": 4.444496773888128e-05,
+      "loss": 0.2463,
+      "step": 960
+    },
+    {
+      "epoch": 0.8865313653136532,
+      "grad_norm": 0.3870641913052104,
+      "learning_rate": 4.442808594915886e-05,
+      "loss": 0.2417,
+      "step": 961
+    },
+    {
+      "epoch": 0.8874538745387454,
+      "grad_norm": 0.3569445908716236,
+      "learning_rate": 4.441118176288891e-05,
+      "loss": 0.2504,
+      "step": 962
+    },
+    {
+      "epoch": 0.8883763837638377,
+      "grad_norm": 0.3680184801953975,
+      "learning_rate": 4.439425519955844e-05,
+      "loss": 0.2466,
+      "step": 963
+    },
+    {
+      "epoch": 0.8892988929889298,
+      "grad_norm": 0.4728595394165681,
+      "learning_rate": 4.437730627868027e-05,
+      "loss": 0.3084,
+      "step": 964
+    },
+    {
+      "epoch": 0.8902214022140221,
+      "grad_norm": 0.4015573191764765,
+      "learning_rate": 4.436033501979299e-05,
+      "loss": 0.2364,
+      "step": 965
+    },
+    {
+      "epoch": 0.8911439114391144,
+      "grad_norm": 0.4016704048673385,
+      "learning_rate": 4.434334144246092e-05,
+      "loss": 0.2433,
+      "step": 966
+    },
+    {
+      "epoch": 0.8920664206642066,
+      "grad_norm": 0.40381267757102457,
+      "learning_rate": 4.432632556627413e-05,
+      "loss": 0.2663,
+      "step": 967
+    },
+    {
+      "epoch": 0.8929889298892989,
+      "grad_norm": 0.41218953298312716,
+      "learning_rate": 4.430928741084839e-05,
+      "loss": 0.2321,
+      "step": 968
+    },
+    {
+      "epoch": 0.8939114391143912,
+      "grad_norm": 0.3932985646232469,
+      "learning_rate": 4.429222699582517e-05,
+      "loss": 0.2552,
+      "step": 969
+    },
+    {
+      "epoch": 0.8948339483394834,
+      "grad_norm": 0.3592400920563191,
+      "learning_rate": 4.4275144340871556e-05,
+      "loss": 0.2403,
+      "step": 970
+    },
+    {
+      "epoch": 0.8957564575645757,
+      "grad_norm": 0.3806295572930225,
+      "learning_rate": 4.4258039465680326e-05,
+      "loss": 0.2373,
+      "step": 971
+    },
+    {
+      "epoch": 0.8966789667896679,
+      "grad_norm": 0.35209635408855017,
+      "learning_rate": 4.4240912389969833e-05,
+      "loss": 0.2287,
+      "step": 972
+    },
+    {
+      "epoch": 0.8976014760147601,
+      "grad_norm": 0.381360988218551,
+      "learning_rate": 4.422376313348405e-05,
+      "loss": 0.2476,
+      "step": 973
+    },
+    {
+      "epoch": 0.8985239852398524,
+      "grad_norm": 0.3774398810196574,
+      "learning_rate": 4.42065917159925e-05,
+      "loss": 0.2525,
+      "step": 974
+    },
+    {
+      "epoch": 0.8994464944649446,
+      "grad_norm": 0.39437701105031125,
+      "learning_rate": 4.418939815729026e-05,
+      "loss": 0.2372,
+      "step": 975
+    },
+    {
+      "epoch": 0.9003690036900369,
+      "grad_norm": 0.42489208182485905,
+      "learning_rate": 4.417218247719794e-05,
+      "loss": 0.2454,
+      "step": 976
+    },
+    {
+      "epoch": 0.9012915129151291,
+      "grad_norm": 0.3951313616214242,
+      "learning_rate": 4.415494469556163e-05,
+      "loss": 0.2388,
+      "step": 977
+    },
+    {
+      "epoch": 0.9022140221402214,
+      "grad_norm": 0.4328782740433957,
+      "learning_rate": 4.413768483225292e-05,
+      "loss": 0.2811,
+      "step": 978
+    },
+    {
+      "epoch": 0.9031365313653137,
+      "grad_norm": 0.41015549722125544,
+      "learning_rate": 4.412040290716884e-05,
+      "loss": 0.2578,
+      "step": 979
+    },
+    {
+      "epoch": 0.9040590405904059,
+      "grad_norm": 0.3733772113740058,
+      "learning_rate": 4.410309894023187e-05,
+      "loss": 0.2295,
+      "step": 980
+    },
+    {
+      "epoch": 0.9049815498154982,
+      "grad_norm": 0.33251844742181735,
+      "learning_rate": 4.408577295138988e-05,
+      "loss": 0.2392,
+      "step": 981
+    },
+    {
+      "epoch": 0.9059040590405905,
+      "grad_norm": 0.33664811795093086,
+      "learning_rate": 4.406842496061615e-05,
+      "loss": 0.249,
+      "step": 982
+    },
+    {
+      "epoch": 0.9068265682656826,
+      "grad_norm": 0.4100127469946453,
+      "learning_rate": 4.4051054987909295e-05,
+      "loss": 0.2538,
+      "step": 983
+    },
+    {
+      "epoch": 0.9077490774907749,
+      "grad_norm": 0.4038276025334575,
+      "learning_rate": 4.40336630532933e-05,
+      "loss": 0.2671,
+      "step": 984
+    },
+    {
+      "epoch": 0.9086715867158671,
+      "grad_norm": 0.4130873622798777,
+      "learning_rate": 4.4016249176817424e-05,
+      "loss": 0.2482,
+      "step": 985
+    },
+    {
+      "epoch": 0.9095940959409594,
+      "grad_norm": 0.3438323781857651,
+      "learning_rate": 4.399881337855629e-05,
+      "loss": 0.2379,
+      "step": 986
+    },
+    {
+      "epoch": 0.9105166051660517,
+      "grad_norm": 0.3771480958562822,
+      "learning_rate": 4.398135567860972e-05,
+      "loss": 0.2384,
+      "step": 987
+    },
+    {
+      "epoch": 0.9114391143911439,
+      "grad_norm": 0.385766779169622,
+      "learning_rate": 4.396387609710283e-05,
+      "loss": 0.23,
+      "step": 988
+    },
+    {
+      "epoch": 0.9123616236162362,
+      "grad_norm": 0.42205705134674754,
+      "learning_rate": 4.394637465418594e-05,
+      "loss": 0.2624,
+      "step": 989
+    },
+    {
+      "epoch": 0.9132841328413284,
+      "grad_norm": 0.3862877149306351,
+      "learning_rate": 4.392885137003459e-05,
+      "loss": 0.2461,
+      "step": 990
+    },
+    {
+      "epoch": 0.9142066420664207,
+      "grad_norm": 0.3459414889329022,
+      "learning_rate": 4.391130626484947e-05,
+      "loss": 0.2346,
+      "step": 991
+    },
+    {
+      "epoch": 0.915129151291513,
+      "grad_norm": 0.38016164148453147,
+      "learning_rate": 4.389373935885646e-05,
+      "loss": 0.236,
+      "step": 992
+    },
+    {
+      "epoch": 0.9160516605166051,
+      "grad_norm": 0.3841933827116207,
+      "learning_rate": 4.387615067230654e-05,
+      "loss": 0.2696,
+      "step": 993
+    },
+    {
+      "epoch": 0.9169741697416974,
+      "grad_norm": 0.41688074113579665,
+      "learning_rate": 4.3858540225475817e-05,
+      "loss": 0.2563,
+      "step": 994
+    },
+    {
+      "epoch": 0.9178966789667896,
+      "grad_norm": 0.3849169335820403,
+      "learning_rate": 4.384090803866547e-05,
+      "loss": 0.2593,
+      "step": 995
+    },
+    {
+      "epoch": 0.9188191881918819,
+      "grad_norm": 0.38188251977300347,
+      "learning_rate": 4.3823254132201763e-05,
+      "loss": 0.248,
+      "step": 996
+    },
+    {
+      "epoch": 0.9197416974169742,
+      "grad_norm": 0.4147197292335253,
+      "learning_rate": 4.380557852643598e-05,
+      "loss": 0.2401,
+      "step": 997
+    },
+    {
+      "epoch": 0.9206642066420664,
+      "grad_norm": 0.4440822567873978,
+      "learning_rate": 4.378788124174441e-05,
+      "loss": 0.253,
+      "step": 998
+    },
+    {
+      "epoch": 0.9215867158671587,
+      "grad_norm": 0.4169692472216945,
+      "learning_rate": 4.377016229852836e-05,
+      "loss": 0.2335,
+      "step": 999
+    },
+    {
+      "epoch": 0.922509225092251,
+      "grad_norm": 0.41047303343206154,
+      "learning_rate": 4.3752421717214085e-05,
+      "loss": 0.2563,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9234317343173432,
+      "grad_norm": 0.41813382044338954,
+      "learning_rate": 4.3734659518252787e-05,
+      "loss": 0.2834,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9243542435424354,
+      "grad_norm": 0.41427781364888505,
+      "learning_rate": 4.371687572212059e-05,
+      "loss": 0.261,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9252767527675276,
+      "grad_norm": 0.4623243834462149,
+      "learning_rate": 4.3699070349318537e-05,
+      "loss": 0.2837,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9261992619926199,
+      "grad_norm": 0.399358869629028,
+      "learning_rate": 4.36812434203725e-05,
+      "loss": 0.2581,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9271217712177122,
+      "grad_norm": 0.42411862628878405,
+      "learning_rate": 4.3663394955833235e-05,
+      "loss": 0.2694,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9280442804428044,
+      "grad_norm": 0.3571473870625449,
+      "learning_rate": 4.364552497627632e-05,
+      "loss": 0.2242,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9289667896678967,
+      "grad_norm": 0.3723780793429299,
+      "learning_rate": 4.3627633502302124e-05,
+      "loss": 0.238,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9298892988929889,
+      "grad_norm": 0.3957445137280655,
+      "learning_rate": 4.360972055453579e-05,
+      "loss": 0.2537,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9308118081180812,
+      "grad_norm": 0.38717717292059817,
+      "learning_rate": 4.3591786153627247e-05,
+      "loss": 0.2225,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9317343173431735,
+      "grad_norm": 0.3863531822945893,
+      "learning_rate": 4.357383032025112e-05,
+      "loss": 0.2461,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9326568265682657,
+      "grad_norm": 0.3517513341161581,
+      "learning_rate": 4.355585307510675e-05,
+      "loss": 0.2115,
+      "step": 1011
+    },
+    {
+      "epoch": 0.933579335793358,
+      "grad_norm": 0.394077914968105,
+      "learning_rate": 4.353785443891818e-05,
+      "loss": 0.2469,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9345018450184502,
+      "grad_norm": 0.4000681433541021,
+      "learning_rate": 4.351983443243409e-05,
+      "loss": 0.238,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9354243542435424,
+      "grad_norm": 0.38273149497237724,
+      "learning_rate": 4.350179307642781e-05,
+      "loss": 0.256,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9363468634686347,
+      "grad_norm": 0.39191957054373927,
+      "learning_rate": 4.3483730391697275e-05,
+      "loss": 0.2546,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9372693726937269,
+      "grad_norm": 0.3654054474137455,
+      "learning_rate": 4.346564639906501e-05,
+      "loss": 0.2292,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9381918819188192,
+      "grad_norm": 0.38086987242860304,
+      "learning_rate": 4.344754111937809e-05,
+      "loss": 0.2337,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9391143911439115,
+      "grad_norm": 0.3891942472846791,
+      "learning_rate": 4.342941457350816e-05,
+      "loss": 0.2579,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9400369003690037,
+      "grad_norm": 0.43555445451666,
+      "learning_rate": 4.3411266782351346e-05,
+      "loss": 0.2411,
+      "step": 1019
+    },
+    {
+      "epoch": 0.940959409594096,
+      "grad_norm": 0.3664039746636333,
+      "learning_rate": 4.3393097766828293e-05,
+      "loss": 0.2468,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9418819188191881,
+      "grad_norm": 0.3723820455291896,
+      "learning_rate": 4.3374907547884095e-05,
+      "loss": 0.2426,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9428044280442804,
+      "grad_norm": 0.4054503641813245,
+      "learning_rate": 4.3356696146488304e-05,
+      "loss": 0.2491,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9437269372693727,
+      "grad_norm": 0.38516524752919157,
+      "learning_rate": 4.333846358363487e-05,
+      "loss": 0.222,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9446494464944649,
+      "grad_norm": 0.40103836099698303,
+      "learning_rate": 4.3320209880342156e-05,
+      "loss": 0.2545,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9455719557195572,
+      "grad_norm": 0.38058729358893373,
+      "learning_rate": 4.33019350576529e-05,
+      "loss": 0.2791,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9464944649446494,
+      "grad_norm": 0.37402738285722437,
+      "learning_rate": 4.3283639136634167e-05,
+      "loss": 0.237,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9474169741697417,
+      "grad_norm": 0.3697120672939745,
+      "learning_rate": 4.3265322138377354e-05,
+      "loss": 0.2524,
+      "step": 1027
+    },
+    {
+      "epoch": 0.948339483394834,
+      "grad_norm": 0.3820254627029573,
+      "learning_rate": 4.3246984083998154e-05,
+      "loss": 0.258,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9492619926199262,
+      "grad_norm": 0.3523371586640625,
+      "learning_rate": 4.322862499463654e-05,
+      "loss": 0.2347,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9501845018450185,
+      "grad_norm": 0.3781888289848851,
+      "learning_rate": 4.321024489145673e-05,
+      "loss": 0.2446,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9511070110701108,
+      "grad_norm": 0.4317197990012946,
+      "learning_rate": 4.319184379564716e-05,
+      "loss": 0.2389,
+      "step": 1031
+    },
+    {
+      "epoch": 0.9520295202952029,
+      "grad_norm": 0.4410472946030106,
+      "learning_rate": 4.3173421728420464e-05,
+      "loss": 0.2875,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9529520295202952,
+      "grad_norm": 0.3812828110364232,
+      "learning_rate": 4.315497871101347e-05,
+      "loss": 0.2505,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9538745387453874,
+      "grad_norm": 0.3998281184321331,
+      "learning_rate": 4.313651476468715e-05,
+      "loss": 0.2456,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9547970479704797,
+      "grad_norm": 0.3726950564281805,
+      "learning_rate": 4.311802991072659e-05,
+      "loss": 0.2439,
+      "step": 1035
+    },
+    {
+      "epoch": 0.955719557195572,
+      "grad_norm": 0.37085693904059847,
+      "learning_rate": 4.309952417044099e-05,
+      "loss": 0.2481,
+      "step": 1036
+    },
+    {
+      "epoch": 0.9566420664206642,
+      "grad_norm": 0.41930432385724087,
+      "learning_rate": 4.308099756516361e-05,
+      "loss": 0.2627,
+      "step": 1037
+    },
+    {
+      "epoch": 0.9575645756457565,
+      "grad_norm": 0.4012474686060809,
+      "learning_rate": 4.306245011625181e-05,
+      "loss": 0.2836,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9584870848708487,
+      "grad_norm": 0.39256009864556235,
+      "learning_rate": 4.304388184508691e-05,
+      "loss": 0.2618,
+      "step": 1039
+    },
+    {
+      "epoch": 0.959409594095941,
+      "grad_norm": 0.4018260203344265,
+      "learning_rate": 4.3025292773074294e-05,
+      "loss": 0.2672,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9603321033210332,
+      "grad_norm": 0.3796571177233923,
+      "learning_rate": 4.3006682921643296e-05,
+      "loss": 0.2587,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9612546125461254,
+      "grad_norm": 0.3296707035409425,
+      "learning_rate": 4.298805231224721e-05,
+      "loss": 0.2033,
+      "step": 1042
+    },
+    {
+      "epoch": 0.9621771217712177,
+      "grad_norm": 0.36564967878510296,
+      "learning_rate": 4.296940096636324e-05,
+      "loss": 0.2393,
+      "step": 1043
+    },
+    {
+      "epoch": 0.9630996309963099,
+      "grad_norm": 0.4024183037228938,
+      "learning_rate": 4.2950728905492544e-05,
+      "loss": 0.2545,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9640221402214022,
+      "grad_norm": 0.359357908219692,
+      "learning_rate": 4.2932036151160104e-05,
+      "loss": 0.2306,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9649446494464945,
+      "grad_norm": 0.41006569313717495,
+      "learning_rate": 4.291332272491479e-05,
+      "loss": 0.2496,
+      "step": 1046
+    },
+    {
+      "epoch": 0.9658671586715867,
+      "grad_norm": 0.40258702606237634,
+      "learning_rate": 4.289458864832931e-05,
+      "loss": 0.2562,
+      "step": 1047
+    },
+    {
+      "epoch": 0.966789667896679,
+      "grad_norm": 0.39459325003198475,
+      "learning_rate": 4.287583394300015e-05,
+      "loss": 0.2526,
+      "step": 1048
+    },
+    {
+      "epoch": 0.9677121771217713,
+      "grad_norm": 0.4638042314426276,
+      "learning_rate": 4.2857058630547594e-05,
+      "loss": 0.2695,
+      "step": 1049
+    },
+    {
+      "epoch": 0.9686346863468634,
+      "grad_norm": 0.4025697845338391,
+      "learning_rate": 4.283826273261567e-05,
+      "loss": 0.2713,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9695571955719557,
+      "grad_norm": 0.3669837154426985,
+      "learning_rate": 4.281944627087214e-05,
+      "loss": 0.2529,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9704797047970479,
+      "grad_norm": 0.3972939537424461,
+      "learning_rate": 4.28006092670085e-05,
+      "loss": 0.2467,
+      "step": 1052
+    },
+    {
+      "epoch": 0.9714022140221402,
+      "grad_norm": 0.4099812196589974,
+      "learning_rate": 4.2781751742739885e-05,
+      "loss": 0.2642,
+      "step": 1053
+    },
+    {
+      "epoch": 0.9723247232472325,
+      "grad_norm": 0.40327491814518013,
+      "learning_rate": 4.27628737198051e-05,
+      "loss": 0.2476,
+      "step": 1054
+    },
+    {
+      "epoch": 0.9732472324723247,
+      "grad_norm": 0.3542847438268756,
+      "learning_rate": 4.274397521996658e-05,
+      "loss": 0.2318,
+      "step": 1055
+    },
+    {
+      "epoch": 0.974169741697417,
+      "grad_norm": 0.357301905791408,
+      "learning_rate": 4.272505626501039e-05,
+      "loss": 0.2267,
+      "step": 1056
+    },
+    {
+      "epoch": 0.9750922509225092,
+      "grad_norm": 0.4057193033011623,
+      "learning_rate": 4.270611687674615e-05,
+      "loss": 0.2655,
+      "step": 1057
+    },
+    {
+      "epoch": 0.9760147601476015,
+      "grad_norm": 0.4210452062719313,
+      "learning_rate": 4.268715707700703e-05,
+      "loss": 0.271,
+      "step": 1058
+    },
+    {
+      "epoch": 0.9769372693726938,
+      "grad_norm": 0.37462001216192053,
+      "learning_rate": 4.266817688764974e-05,
+      "loss": 0.2188,
+      "step": 1059
+    },
+    {
+      "epoch": 0.977859778597786,
+      "grad_norm": 0.384305636910666,
+      "learning_rate": 4.2649176330554505e-05,
+      "loss": 0.2358,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9787822878228782,
+      "grad_norm": 0.4023223260591023,
+      "learning_rate": 4.263015542762502e-05,
+      "loss": 0.2253,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9797047970479705,
+      "grad_norm": 0.3582360155741272,
+      "learning_rate": 4.261111420078843e-05,
+      "loss": 0.2323,
+      "step": 1062
+    },
+    {
+      "epoch": 0.9806273062730627,
+      "grad_norm": 0.34309753558813766,
+      "learning_rate": 4.259205267199532e-05,
+      "loss": 0.2312,
+      "step": 1063
+    },
+    {
+      "epoch": 0.981549815498155,
+      "grad_norm": 0.342868960758366,
+      "learning_rate": 4.257297086321967e-05,
+      "loss": 0.2117,
+      "step": 1064
+    },
+    {
+      "epoch": 0.9824723247232472,
+      "grad_norm": 0.3756603692103656,
+      "learning_rate": 4.255386879645884e-05,
+      "loss": 0.2511,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9833948339483395,
+      "grad_norm": 0.36504578675556676,
+      "learning_rate": 4.2534746493733544e-05,
+      "loss": 0.2623,
+      "step": 1066
+    },
+    {
+      "epoch": 0.9843173431734318,
+      "grad_norm": 0.3877827233780789,
+      "learning_rate": 4.2515603977087834e-05,
+      "loss": 0.2484,
+      "step": 1067
+    },
+    {
+      "epoch": 0.985239852398524,
+      "grad_norm": 0.4000115235566352,
+      "learning_rate": 4.2496441268589046e-05,
+      "loss": 0.2322,
+      "step": 1068
+    },
+    {
+      "epoch": 0.9861623616236163,
+      "grad_norm": 0.3916507964424266,
+      "learning_rate": 4.247725839032781e-05,
+      "loss": 0.2486,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9870848708487084,
+      "grad_norm": 0.3672624822476246,
+      "learning_rate": 4.245805536441799e-05,
+      "loss": 0.246,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9880073800738007,
+      "grad_norm": 0.424336758506717,
+      "learning_rate": 4.243883221299669e-05,
+      "loss": 0.256,
+      "step": 1071
+    },
+    {
+      "epoch": 0.988929889298893,
+      "grad_norm": 0.3925976329361435,
+      "learning_rate": 4.241958895822422e-05,
+      "loss": 0.2327,
+      "step": 1072
+    },
+    {
+      "epoch": 0.9898523985239852,
+      "grad_norm": 0.36882062594137904,
+      "learning_rate": 4.240032562228405e-05,
+      "loss": 0.24,
+      "step": 1073
+    },
+    {
+      "epoch": 0.9907749077490775,
+      "grad_norm": 0.34872009808112114,
+      "learning_rate": 4.23810422273828e-05,
+      "loss": 0.237,
+      "step": 1074
+    },
+    {
+      "epoch": 0.9916974169741697,
+      "grad_norm": 0.3600321823367888,
+      "learning_rate": 4.2361738795750214e-05,
+      "loss": 0.2533,
+      "step": 1075
+    },
+    {
+      "epoch": 0.992619926199262,
+      "grad_norm": 0.3744568379901036,
+      "learning_rate": 4.234241534963916e-05,
+      "loss": 0.244,
+      "step": 1076
+    },
+    {
+      "epoch": 0.9935424354243543,
+      "grad_norm": 0.35579109010529775,
+      "learning_rate": 4.2323071911325535e-05,
+      "loss": 0.2362,
+      "step": 1077
+    },
+    {
+      "epoch": 0.9944649446494465,
+      "grad_norm": 0.34114175573589445,
+      "learning_rate": 4.230370850310832e-05,
+      "loss": 0.2144,
+      "step": 1078
+    },
+    {
+      "epoch": 0.9953874538745388,
+      "grad_norm": 0.4020001689070888,
+      "learning_rate": 4.228432514730949e-05,
+      "loss": 0.2648,
+      "step": 1079
+    },
+    {
+      "epoch": 0.996309963099631,
+      "grad_norm": 0.33804693122463964,
+      "learning_rate": 4.2264921866274046e-05,
+      "loss": 0.2684,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9972324723247232,
+      "grad_norm": 0.3775262235809972,
+      "learning_rate": 4.224549868236993e-05,
+      "loss": 0.2539,
+      "step": 1081
+    },
+    {
+      "epoch": 0.9981549815498155,
+      "grad_norm": 0.36410628062196304,
+      "learning_rate": 4.2226055617988024e-05,
+      "loss": 0.2561,
+      "step": 1082
+    },
+    {
+      "epoch": 0.9990774907749077,
+      "grad_norm": 0.37350818965624166,
+      "learning_rate": 4.220659269554217e-05,
+      "loss": 0.2529,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3419738536638325,
+      "learning_rate": 4.218710993746906e-05,
+      "loss": 0.2375,
+      "step": 1084
+    },
+    {
+      "epoch": 1.0009225092250922,
+      "grad_norm": 0.3769368882169451,
+      "learning_rate": 4.2167607366228266e-05,
+      "loss": 0.1954,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0018450184501846,
+      "grad_norm": 0.37813463650633805,
+      "learning_rate": 4.2148085004302205e-05,
+      "loss": 0.1987,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0027675276752768,
+      "grad_norm": 0.33502446350380966,
+      "learning_rate": 4.212854287419611e-05,
+      "loss": 0.1669,
+      "step": 1087
+    },
+    {
+      "epoch": 1.003690036900369,
+      "grad_norm": 0.38269711360360914,
+      "learning_rate": 4.2108980998437984e-05,
+      "loss": 0.1816,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0046125461254614,
+      "grad_norm": 0.4259402307903112,
+      "learning_rate": 4.208939939957862e-05,
+      "loss": 0.1729,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0055350553505535,
+      "grad_norm": 0.5129546312352509,
+      "learning_rate": 4.2069798100191525e-05,
+      "loss": 0.1661,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0064575645756457,
+      "grad_norm": 0.4389264706914863,
+      "learning_rate": 4.2050177122872934e-05,
+      "loss": 0.1952,
+      "step": 1091
+    },
+    {
+      "epoch": 1.007380073800738,
+      "grad_norm": 0.41996401878411477,
+      "learning_rate": 4.2030536490241754e-05,
+      "loss": 0.1639,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0083025830258303,
+      "grad_norm": 0.40587370232707043,
+      "learning_rate": 4.2010876224939556e-05,
+      "loss": 0.183,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0092250922509225,
+      "grad_norm": 0.4216925521611639,
+      "learning_rate": 4.1991196349630536e-05,
+      "loss": 0.1777,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0101476014760147,
+      "grad_norm": 0.354670234146558,
+      "learning_rate": 4.19714968870015e-05,
+      "loss": 0.1556,
+      "step": 1095
+    },
+    {
+      "epoch": 1.011070110701107,
+      "grad_norm": 0.37794728109822767,
+      "learning_rate": 4.195177785976185e-05,
+      "loss": 0.1615,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0119926199261993,
+      "grad_norm": 0.4704906880862343,
+      "learning_rate": 4.193203929064353e-05,
+      "loss": 0.1987,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0129151291512914,
+      "grad_norm": 0.3772186034127118,
+      "learning_rate": 4.191228120240099e-05,
+      "loss": 0.1643,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0138376383763839,
+      "grad_norm": 0.42250435630940364,
+      "learning_rate": 4.1892503617811216e-05,
+      "loss": 0.1789,
+      "step": 1099
+    },
+    {
+      "epoch": 1.014760147601476,
+      "grad_norm": 0.40835588337077483,
+      "learning_rate": 4.1872706559673665e-05,
+      "loss": 0.1671,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0156826568265682,
+      "grad_norm": 0.3971438386755592,
+      "learning_rate": 4.185289005081021e-05,
+      "loss": 0.1758,
+      "step": 1101
+    },
+    {
+      "epoch": 1.0166051660516606,
+      "grad_norm": 0.42954003229122456,
+      "learning_rate": 4.1833054114065175e-05,
+      "loss": 0.191,
+      "step": 1102
+    },
+    {
+      "epoch": 1.0175276752767528,
+      "grad_norm": 0.5067713954677341,
+      "learning_rate": 4.1813198772305284e-05,
+      "loss": 0.1785,
+      "step": 1103
+    },
+    {
+      "epoch": 1.018450184501845,
+      "grad_norm": 0.36309220773044926,
+      "learning_rate": 4.1793324048419626e-05,
+      "loss": 0.1654,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0193726937269372,
+      "grad_norm": 0.3651840162756363,
+      "learning_rate": 4.177342996531961e-05,
+      "loss": 0.1507,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0202952029520296,
+      "grad_norm": 0.4051567508664911,
+      "learning_rate": 4.175351654593899e-05,
+      "loss": 0.1719,
+      "step": 1106
+    },
+    {
+      "epoch": 1.0212177121771218,
+      "grad_norm": 0.3881943625337821,
+      "learning_rate": 4.1733583813233815e-05,
+      "loss": 0.1742,
+      "step": 1107
+    },
+    {
+      "epoch": 1.022140221402214,
+      "grad_norm": 0.37905275668416616,
+      "learning_rate": 4.1713631790182364e-05,
+      "loss": 0.1635,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0230627306273063,
+      "grad_norm": 0.3427845280225927,
+      "learning_rate": 4.169366049978519e-05,
+      "loss": 0.1579,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0239852398523985,
+      "grad_norm": 0.3538821348972919,
+      "learning_rate": 4.167366996506503e-05,
+      "loss": 0.1776,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0249077490774907,
+      "grad_norm": 0.36124493881262104,
+      "learning_rate": 4.1653660209066835e-05,
+      "loss": 0.1815,
+      "step": 1111
+    },
+    {
+      "epoch": 1.0258302583025831,
+      "grad_norm": 0.3934661273597267,
+      "learning_rate": 4.16336312548577e-05,
+      "loss": 0.1672,
+      "step": 1112
+    },
+    {
+      "epoch": 1.0267527675276753,
+      "grad_norm": 0.35071002280057345,
+      "learning_rate": 4.161358312552682e-05,
+      "loss": 0.1409,
+      "step": 1113
+    },
+    {
+      "epoch": 1.0276752767527675,
+      "grad_norm": 0.3631582571734631,
+      "learning_rate": 4.1593515844185536e-05,
+      "loss": 0.1526,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0285977859778597,
+      "grad_norm": 0.4493670419396526,
+      "learning_rate": 4.157342943396728e-05,
+      "loss": 0.1829,
+      "step": 1115
+    },
+    {
+      "epoch": 1.029520295202952,
+      "grad_norm": 0.39113684237210034,
+      "learning_rate": 4.155332391802748e-05,
+      "loss": 0.153,
+      "step": 1116
+    },
+    {
+      "epoch": 1.0304428044280443,
+      "grad_norm": 0.3407921649948736,
+      "learning_rate": 4.153319931954363e-05,
+      "loss": 0.1448,
+      "step": 1117
+    },
+    {
+      "epoch": 1.0313653136531364,
+      "grad_norm": 0.35735453759259145,
+      "learning_rate": 4.1513055661715214e-05,
+      "loss": 0.1642,
+      "step": 1118
+    },
+    {
+      "epoch": 1.0322878228782288,
+      "grad_norm": 0.378033416589244,
+      "learning_rate": 4.1492892967763686e-05,
+      "loss": 0.17,
+      "step": 1119
+    },
+    {
+      "epoch": 1.033210332103321,
+      "grad_norm": 0.32755535064867,
+      "learning_rate": 4.147271126093243e-05,
+      "loss": 0.1763,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0341328413284132,
+      "grad_norm": 0.33184993219665576,
+      "learning_rate": 4.145251056448678e-05,
+      "loss": 0.1641,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0350553505535056,
+      "grad_norm": 0.34342127146183604,
+      "learning_rate": 4.1432290901713944e-05,
+      "loss": 0.1599,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0359778597785978,
+      "grad_norm": 0.31650953682305155,
+      "learning_rate": 4.141205229592298e-05,
+      "loss": 0.1522,
+      "step": 1123
+    },
+    {
+      "epoch": 1.03690036900369,
+      "grad_norm": 0.3884987790776325,
+      "learning_rate": 4.13917947704448e-05,
+      "loss": 0.1784,
+      "step": 1124
+    },
+    {
+      "epoch": 1.0378228782287824,
+      "grad_norm": 0.39352618968889314,
+      "learning_rate": 4.137151834863213e-05,
+      "loss": 0.1832,
+      "step": 1125
+    },
+    {
+      "epoch": 1.0387453874538746,
+      "grad_norm": 0.40731228868134994,
+      "learning_rate": 4.1351223053859465e-05,
+      "loss": 0.164,
+      "step": 1126
+    },
+    {
+      "epoch": 1.0396678966789668,
+      "grad_norm": 0.35681228331776604,
+      "learning_rate": 4.133090890952306e-05,
+      "loss": 0.1737,
+      "step": 1127
+    },
+    {
+      "epoch": 1.040590405904059,
+      "grad_norm": 0.38427644358885815,
+      "learning_rate": 4.131057593904092e-05,
+      "loss": 0.1525,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0415129151291513,
+      "grad_norm": 0.3735181284156466,
+      "learning_rate": 4.129022416585272e-05,
+      "loss": 0.1689,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0424354243542435,
+      "grad_norm": 0.3841245872081137,
+      "learning_rate": 4.126985361341984e-05,
+      "loss": 0.1679,
+      "step": 1130
+    },
+    {
+      "epoch": 1.0433579335793357,
+      "grad_norm": 0.4394062794454207,
+      "learning_rate": 4.1249464305225294e-05,
+      "loss": 0.1701,
+      "step": 1131
+    },
+    {
+      "epoch": 1.044280442804428,
+      "grad_norm": 0.34635693765690534,
+      "learning_rate": 4.1229056264773705e-05,
+      "loss": 0.1483,
+      "step": 1132
+    },
+    {
+      "epoch": 1.0452029520295203,
+      "grad_norm": 0.3689927275176747,
+      "learning_rate": 4.1208629515591316e-05,
+      "loss": 0.173,
+      "step": 1133
+    },
+    {
+      "epoch": 1.0461254612546125,
+      "grad_norm": 0.42258687449431287,
+      "learning_rate": 4.118818408122592e-05,
+      "loss": 0.189,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0470479704797049,
+      "grad_norm": 0.3591825951560322,
+      "learning_rate": 4.116771998524688e-05,
+      "loss": 0.1768,
+      "step": 1135
+    },
+    {
+      "epoch": 1.047970479704797,
+      "grad_norm": 0.3531927142004278,
+      "learning_rate": 4.114723725124501e-05,
+      "loss": 0.1765,
+      "step": 1136
+    },
+    {
+      "epoch": 1.0488929889298892,
+      "grad_norm": 0.3829638749984938,
+      "learning_rate": 4.112673590283267e-05,
+      "loss": 0.1816,
+      "step": 1137
+    },
+    {
+      "epoch": 1.0498154981549817,
+      "grad_norm": 0.37254171936921016,
+      "learning_rate": 4.1106215963643645e-05,
+      "loss": 0.1709,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0507380073800738,
+      "grad_norm": 0.36194452587323034,
+      "learning_rate": 4.108567745733318e-05,
+      "loss": 0.1576,
+      "step": 1139
+    },
+    {
+      "epoch": 1.051660516605166,
+      "grad_norm": 0.41177269350486917,
+      "learning_rate": 4.106512040757789e-05,
+      "loss": 0.1664,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0525830258302582,
+      "grad_norm": 0.42734114893298225,
+      "learning_rate": 4.1044544838075794e-05,
+      "loss": 0.1702,
+      "step": 1141
+    },
+    {
+      "epoch": 1.0535055350553506,
+      "grad_norm": 0.37337649729707373,
+      "learning_rate": 4.102395077254624e-05,
+      "loss": 0.1489,
+      "step": 1142
+    },
+    {
+      "epoch": 1.0544280442804428,
+      "grad_norm": 0.39261391909709,
+      "learning_rate": 4.100333823472992e-05,
+      "loss": 0.1749,
+      "step": 1143
+    },
+    {
+      "epoch": 1.055350553505535,
+      "grad_norm": 0.42549411406276944,
+      "learning_rate": 4.098270724838879e-05,
+      "loss": 0.1549,
+      "step": 1144
+    },
+    {
+      "epoch": 1.0562730627306274,
+      "grad_norm": 0.5053855322170004,
+      "learning_rate": 4.096205783730611e-05,
+      "loss": 0.1779,
+      "step": 1145
+    },
+    {
+      "epoch": 1.0571955719557196,
+      "grad_norm": 0.45748588362554227,
+      "learning_rate": 4.094139002528635e-05,
+      "loss": 0.163,
+      "step": 1146
+    },
+    {
+      "epoch": 1.0581180811808117,
+      "grad_norm": 0.4068618648456191,
+      "learning_rate": 4.092070383615522e-05,
+      "loss": 0.1813,
+      "step": 1147
+    },
+    {
+      "epoch": 1.0590405904059041,
+      "grad_norm": 0.40950624908793243,
+      "learning_rate": 4.089999929375957e-05,
+      "loss": 0.1853,
+      "step": 1148
+    },
+    {
+      "epoch": 1.0599630996309963,
+      "grad_norm": 0.38836686198772674,
+      "learning_rate": 4.0879276421967475e-05,
+      "loss": 0.1708,
+      "step": 1149
+    },
+    {
+      "epoch": 1.0608856088560885,
+      "grad_norm": 0.3955714676343847,
+      "learning_rate": 4.0858535244668066e-05,
+      "loss": 0.1574,
+      "step": 1150
+    },
+    {
+      "epoch": 1.061808118081181,
+      "grad_norm": 0.3783893233613329,
+      "learning_rate": 4.083777578577164e-05,
+      "loss": 0.1761,
+      "step": 1151
+    },
+    {
+      "epoch": 1.062730627306273,
+      "grad_norm": 0.40335215045068984,
+      "learning_rate": 4.081699806920951e-05,
+      "loss": 0.1697,
+      "step": 1152
+    },
+    {
+      "epoch": 1.0636531365313653,
+      "grad_norm": 0.3969568516698353,
+      "learning_rate": 4.0796202118934105e-05,
+      "loss": 0.1621,
+      "step": 1153
+    },
+    {
+      "epoch": 1.0645756457564575,
+      "grad_norm": 0.4575477135581521,
+      "learning_rate": 4.077538795891881e-05,
+      "loss": 0.1622,
+      "step": 1154
+    },
+    {
+      "epoch": 1.0654981549815499,
+      "grad_norm": 0.3922931124328152,
+      "learning_rate": 4.075455561315803e-05,
+      "loss": 0.1777,
+      "step": 1155
+    },
+    {
+      "epoch": 1.066420664206642,
+      "grad_norm": 0.39111563963326046,
+      "learning_rate": 4.073370510566714e-05,
+      "loss": 0.1577,
+      "step": 1156
+    },
+    {
+      "epoch": 1.0673431734317342,
+      "grad_norm": 0.3783621355518027,
+      "learning_rate": 4.071283646048244e-05,
+      "loss": 0.1733,
+      "step": 1157
+    },
+    {
+      "epoch": 1.0682656826568266,
+      "grad_norm": 0.3977090300343326,
+      "learning_rate": 4.0691949701661145e-05,
+      "loss": 0.1654,
+      "step": 1158
+    },
+    {
+      "epoch": 1.0691881918819188,
+      "grad_norm": 0.38215847988575297,
+      "learning_rate": 4.067104485328135e-05,
+      "loss": 0.1733,
+      "step": 1159
+    },
+    {
+      "epoch": 1.070110701107011,
+      "grad_norm": 0.3926697225841087,
+      "learning_rate": 4.065012193944201e-05,
+      "loss": 0.1704,
+      "step": 1160
+    },
+    {
+      "epoch": 1.0710332103321034,
+      "grad_norm": 0.46480611520311177,
+      "learning_rate": 4.062918098426288e-05,
+      "loss": 0.1901,
+      "step": 1161
+    },
+    {
+      "epoch": 1.0719557195571956,
+      "grad_norm": 0.3888251222505632,
+      "learning_rate": 4.0608222011884545e-05,
+      "loss": 0.1768,
+      "step": 1162
+    },
+    {
+      "epoch": 1.0728782287822878,
+      "grad_norm": 0.39571832407283675,
+      "learning_rate": 4.058724504646834e-05,
+      "loss": 0.1859,
+      "step": 1163
+    },
+    {
+      "epoch": 1.07380073800738,
+      "grad_norm": 0.40515368230454263,
+      "learning_rate": 4.056625011219636e-05,
+      "loss": 0.184,
+      "step": 1164
+    },
+    {
+      "epoch": 1.0747232472324724,
+      "grad_norm": 0.4075951012369609,
+      "learning_rate": 4.0545237233271383e-05,
+      "loss": 0.1898,
+      "step": 1165
+    },
+    {
+      "epoch": 1.0756457564575646,
+      "grad_norm": 0.41305306544440507,
+      "learning_rate": 4.052420643391692e-05,
+      "loss": 0.1888,
+      "step": 1166
+    },
+    {
+      "epoch": 1.0765682656826567,
+      "grad_norm": 0.45050394410648004,
+      "learning_rate": 4.050315773837708e-05,
+      "loss": 0.1882,
+      "step": 1167
+    },
+    {
+      "epoch": 1.0774907749077491,
+      "grad_norm": 0.4168100597899049,
+      "learning_rate": 4.048209117091668e-05,
+      "loss": 0.1991,
+      "step": 1168
+    },
+    {
+      "epoch": 1.0784132841328413,
+      "grad_norm": 0.38736925330372834,
+      "learning_rate": 4.0461006755821066e-05,
+      "loss": 0.1768,
+      "step": 1169
+    },
+    {
+      "epoch": 1.0793357933579335,
+      "grad_norm": 0.39062853541467635,
+      "learning_rate": 4.043990451739619e-05,
+      "loss": 0.1596,
+      "step": 1170
+    },
+    {
+      "epoch": 1.080258302583026,
+      "grad_norm": 0.3859713176984828,
+      "learning_rate": 4.041878447996855e-05,
+      "loss": 0.1723,
+      "step": 1171
+    },
+    {
+      "epoch": 1.081180811808118,
+      "grad_norm": 0.4365972706893191,
+      "learning_rate": 4.039764666788518e-05,
+      "loss": 0.1648,
+      "step": 1172
+    },
+    {
+      "epoch": 1.0821033210332103,
+      "grad_norm": 0.3972484486373133,
+      "learning_rate": 4.037649110551357e-05,
+      "loss": 0.1728,
+      "step": 1173
+    },
+    {
+      "epoch": 1.0830258302583027,
+      "grad_norm": 0.4318672652196214,
+      "learning_rate": 4.03553178172417e-05,
+      "loss": 0.198,
+      "step": 1174
+    },
+    {
+      "epoch": 1.0839483394833949,
+      "grad_norm": 0.3699873653140849,
+      "learning_rate": 4.033412682747796e-05,
+      "loss": 0.1626,
+      "step": 1175
+    },
+    {
+      "epoch": 1.084870848708487,
+      "grad_norm": 0.36511846754595956,
+      "learning_rate": 4.031291816065117e-05,
+      "loss": 0.1674,
+      "step": 1176
+    },
+    {
+      "epoch": 1.0857933579335795,
+      "grad_norm": 0.4059047610587183,
+      "learning_rate": 4.029169184121051e-05,
+      "loss": 0.1856,
+      "step": 1177
+    },
+    {
+      "epoch": 1.0867158671586716,
+      "grad_norm": 0.37772960175348175,
+      "learning_rate": 4.027044789362552e-05,
+      "loss": 0.1762,
+      "step": 1178
+    },
+    {
+      "epoch": 1.0876383763837638,
+      "grad_norm": 0.36173658775918344,
+      "learning_rate": 4.024918634238606e-05,
+      "loss": 0.1657,
+      "step": 1179
+    },
+    {
+      "epoch": 1.088560885608856,
+      "grad_norm": 0.39888987811999554,
+      "learning_rate": 4.022790721200229e-05,
+      "loss": 0.1879,
+      "step": 1180
+    },
+    {
+      "epoch": 1.0894833948339484,
+      "grad_norm": 0.3878771144901679,
+      "learning_rate": 4.020661052700461e-05,
+      "loss": 0.1663,
+      "step": 1181
+    },
+    {
+      "epoch": 1.0904059040590406,
+      "grad_norm": 0.4059867833781443,
+      "learning_rate": 4.018529631194369e-05,
+      "loss": 0.1758,
+      "step": 1182
+    },
+    {
+      "epoch": 1.0913284132841328,
+      "grad_norm": 0.3822778038862252,
+      "learning_rate": 4.016396459139038e-05,
+      "loss": 0.1987,
+      "step": 1183
+    },
+    {
+      "epoch": 1.0922509225092252,
+      "grad_norm": 0.37823723096845563,
+      "learning_rate": 4.0142615389935736e-05,
+      "loss": 0.1609,
+      "step": 1184
+    },
+    {
+      "epoch": 1.0931734317343174,
+      "grad_norm": 0.4041649814551321,
+      "learning_rate": 4.012124873219094e-05,
+      "loss": 0.1831,
+      "step": 1185
+    },
+    {
+      "epoch": 1.0940959409594095,
+      "grad_norm": 0.4081409306127812,
+      "learning_rate": 4.0099864642787324e-05,
+      "loss": 0.1739,
+      "step": 1186
+    },
+    {
+      "epoch": 1.095018450184502,
+      "grad_norm": 0.41445508633820266,
+      "learning_rate": 4.0078463146376277e-05,
+      "loss": 0.1607,
+      "step": 1187
+    },
+    {
+      "epoch": 1.0959409594095941,
+      "grad_norm": 0.40050046405665934,
+      "learning_rate": 4.005704426762931e-05,
+      "loss": 0.1666,
+      "step": 1188
+    },
+    {
+      "epoch": 1.0968634686346863,
+      "grad_norm": 0.421627251814422,
+      "learning_rate": 4.003560803123791e-05,
+      "loss": 0.18,
+      "step": 1189
+    },
+    {
+      "epoch": 1.0977859778597785,
+      "grad_norm": 0.3775866601395446,
+      "learning_rate": 4.001415446191363e-05,
+      "loss": 0.1585,
+      "step": 1190
+    },
+    {
+      "epoch": 1.098708487084871,
+      "grad_norm": 0.3949171725100949,
+      "learning_rate": 3.999268358438797e-05,
+      "loss": 0.1801,
+      "step": 1191
+    },
+    {
+      "epoch": 1.099630996309963,
+      "grad_norm": 0.445238838334624,
+      "learning_rate": 3.997119542341239e-05,
+      "loss": 0.1852,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1005535055350553,
+      "grad_norm": 0.4079689080027289,
+      "learning_rate": 3.994969000375828e-05,
+      "loss": 0.1667,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1014760147601477,
+      "grad_norm": 0.3955839693352635,
+      "learning_rate": 3.992816735021692e-05,
+      "loss": 0.1898,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1023985239852399,
+      "grad_norm": 0.40343859359856044,
+      "learning_rate": 3.990662748759946e-05,
+      "loss": 0.1733,
+      "step": 1195
+    },
+    {
+      "epoch": 1.103321033210332,
+      "grad_norm": 0.36049582413990144,
+      "learning_rate": 3.988507044073687e-05,
+      "loss": 0.1634,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1042435424354244,
+      "grad_norm": 0.3744445972145401,
+      "learning_rate": 3.986349623447998e-05,
+      "loss": 0.1803,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1051660516605166,
+      "grad_norm": 0.3506582654711279,
+      "learning_rate": 3.9841904893699346e-05,
+      "loss": 0.1526,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1060885608856088,
+      "grad_norm": 0.44016397693457415,
+      "learning_rate": 3.9820296443285306e-05,
+      "loss": 0.1767,
+      "step": 1199
+    },
+    {
+      "epoch": 1.1070110701107012,
+      "grad_norm": 0.3599239979681691,
+      "learning_rate": 3.979867090814791e-05,
+      "loss": 0.1682,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1079335793357934,
+      "grad_norm": 0.4117378061959942,
+      "learning_rate": 3.977702831321692e-05,
+      "loss": 0.1764,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1088560885608856,
+      "grad_norm": 0.3655808844946291,
+      "learning_rate": 3.9755368683441735e-05,
+      "loss": 0.164,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1097785977859778,
+      "grad_norm": 0.40877835704731236,
+      "learning_rate": 3.9733692043791414e-05,
+      "loss": 0.1819,
+      "step": 1203
+    },
+    {
+      "epoch": 1.1107011070110702,
+      "grad_norm": 0.37098664695584865,
+      "learning_rate": 3.9711998419254634e-05,
+      "loss": 0.1716,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1116236162361623,
+      "grad_norm": 0.4192369944446082,
+      "learning_rate": 3.969028783483962e-05,
+      "loss": 0.1853,
+      "step": 1205
+    },
+    {
+      "epoch": 1.1125461254612545,
+      "grad_norm": 0.4238854228544376,
+      "learning_rate": 3.966856031557418e-05,
+      "loss": 0.1713,
+      "step": 1206
+    },
+    {
+      "epoch": 1.113468634686347,
+      "grad_norm": 0.43343780566972534,
+      "learning_rate": 3.964681588650562e-05,
+      "loss": 0.186,
+      "step": 1207
+    },
+    {
+      "epoch": 1.1143911439114391,
+      "grad_norm": 0.3741644034657715,
+      "learning_rate": 3.9625054572700757e-05,
+      "loss": 0.1675,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1153136531365313,
+      "grad_norm": 0.4416771819445561,
+      "learning_rate": 3.960327639924586e-05,
+      "loss": 0.1761,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1162361623616237,
+      "grad_norm": 0.39769798121436667,
+      "learning_rate": 3.958148139124664e-05,
+      "loss": 0.1724,
+      "step": 1210
+    },
+    {
+      "epoch": 1.117158671586716,
+      "grad_norm": 0.3647395932311634,
+      "learning_rate": 3.9559669573828225e-05,
+      "loss": 0.1602,
+      "step": 1211
+    },
+    {
+      "epoch": 1.118081180811808,
+      "grad_norm": 0.4120209918268371,
+      "learning_rate": 3.9537840972135094e-05,
+      "loss": 0.1638,
+      "step": 1212
+    },
+    {
+      "epoch": 1.1190036900369003,
+      "grad_norm": 0.37394155614506874,
+      "learning_rate": 3.95159956113311e-05,
+      "loss": 0.1841,
+      "step": 1213
+    },
+    {
+      "epoch": 1.1199261992619927,
+      "grad_norm": 0.38930236349935654,
+      "learning_rate": 3.94941335165994e-05,
+      "loss": 0.1609,
+      "step": 1214
+    },
+    {
+      "epoch": 1.1208487084870848,
+      "grad_norm": 0.40402129092996436,
+      "learning_rate": 3.9472254713142455e-05,
+      "loss": 0.1758,
+      "step": 1215
+    },
+    {
+      "epoch": 1.121771217712177,
+      "grad_norm": 0.40121844035137655,
+      "learning_rate": 3.945035922618197e-05,
+      "loss": 0.169,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1226937269372694,
+      "grad_norm": 0.3843359826294688,
+      "learning_rate": 3.942844708095892e-05,
+      "loss": 0.1868,
+      "step": 1217
+    },
+    {
+      "epoch": 1.1236162361623616,
+      "grad_norm": 0.37991679506107584,
+      "learning_rate": 3.9406518302733416e-05,
+      "loss": 0.1855,
+      "step": 1218
+    },
+    {
+      "epoch": 1.1245387453874538,
+      "grad_norm": 0.3431532616018786,
+      "learning_rate": 3.938457291678482e-05,
+      "loss": 0.1463,
+      "step": 1219
+    },
+    {
+      "epoch": 1.1254612546125462,
+      "grad_norm": 0.39247797781305566,
+      "learning_rate": 3.9362610948411585e-05,
+      "loss": 0.1563,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1263837638376384,
+      "grad_norm": 0.4455152042959939,
+      "learning_rate": 3.93406324229313e-05,
+      "loss": 0.194,
+      "step": 1221
+    },
+    {
+      "epoch": 1.1273062730627306,
+      "grad_norm": 0.3646911500010225,
+      "learning_rate": 3.931863736568065e-05,
+      "loss": 0.1487,
+      "step": 1222
+    },
+    {
+      "epoch": 1.128228782287823,
+      "grad_norm": 0.3751401018519618,
+      "learning_rate": 3.9296625802015356e-05,
+      "loss": 0.1659,
+      "step": 1223
+    },
+    {
+      "epoch": 1.1291512915129152,
+      "grad_norm": 0.37936272235043617,
+      "learning_rate": 3.9274597757310186e-05,
+      "loss": 0.1777,
+      "step": 1224
+    },
+    {
+      "epoch": 1.1300738007380073,
+      "grad_norm": 0.39440103333960363,
+      "learning_rate": 3.925255325695889e-05,
+      "loss": 0.1643,
+      "step": 1225
+    },
+    {
+      "epoch": 1.1309963099630997,
+      "grad_norm": 0.4200580796108357,
+      "learning_rate": 3.923049232637421e-05,
+      "loss": 0.1814,
+      "step": 1226
+    },
+    {
+      "epoch": 1.131918819188192,
+      "grad_norm": 0.3973096309997986,
+      "learning_rate": 3.920841499098781e-05,
+      "loss": 0.1814,
+      "step": 1227
+    },
+    {
+      "epoch": 1.132841328413284,
+      "grad_norm": 0.38341587007083405,
+      "learning_rate": 3.9186321276250274e-05,
+      "loss": 0.1707,
+      "step": 1228
+    },
+    {
+      "epoch": 1.1337638376383763,
+      "grad_norm": 0.360186558259523,
+      "learning_rate": 3.916421120763106e-05,
+      "loss": 0.1535,
+      "step": 1229
+    },
+    {
+      "epoch": 1.1346863468634687,
+      "grad_norm": 0.450775459207837,
+      "learning_rate": 3.9142084810618495e-05,
+      "loss": 0.1827,
+      "step": 1230
+    },
+    {
+      "epoch": 1.1356088560885609,
+      "grad_norm": 0.4002646614014494,
+      "learning_rate": 3.911994211071971e-05,
+      "loss": 0.1614,
+      "step": 1231
+    },
+    {
+      "epoch": 1.136531365313653,
+      "grad_norm": 0.3459076477142216,
+      "learning_rate": 3.909778313346064e-05,
+      "loss": 0.1468,
+      "step": 1232
+    },
+    {
+      "epoch": 1.1374538745387455,
+      "grad_norm": 0.36252214847000114,
+      "learning_rate": 3.907560790438598e-05,
+      "loss": 0.167,
+      "step": 1233
+    },
+    {
+      "epoch": 1.1383763837638377,
+      "grad_norm": 0.35362107841991497,
+      "learning_rate": 3.905341644905918e-05,
+      "loss": 0.1565,
+      "step": 1234
+    },
+    {
+      "epoch": 1.1392988929889298,
+      "grad_norm": 0.38456791825871967,
+      "learning_rate": 3.9031208793062354e-05,
+      "loss": 0.1778,
+      "step": 1235
+    },
+    {
+      "epoch": 1.140221402214022,
+      "grad_norm": 0.43754192987728807,
+      "learning_rate": 3.900898496199634e-05,
+      "loss": 0.155,
+      "step": 1236
+    },
+    {
+      "epoch": 1.1411439114391144,
+      "grad_norm": 0.37634931688449635,
+      "learning_rate": 3.898674498148058e-05,
+      "loss": 0.1738,
+      "step": 1237
+    },
+    {
+      "epoch": 1.1420664206642066,
+      "grad_norm": 0.37374035268129,
+      "learning_rate": 3.896448887715316e-05,
+      "loss": 0.1687,
+      "step": 1238
+    },
+    {
+      "epoch": 1.1429889298892988,
+      "grad_norm": 0.37830297443523053,
+      "learning_rate": 3.894221667467074e-05,
+      "loss": 0.1546,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1439114391143912,
+      "grad_norm": 0.33714399256913596,
+      "learning_rate": 3.891992839970855e-05,
+      "loss": 0.1668,
+      "step": 1240
+    },
+    {
+      "epoch": 1.1448339483394834,
+      "grad_norm": 0.3966714697489937,
+      "learning_rate": 3.889762407796034e-05,
+      "loss": 0.1734,
+      "step": 1241
+    },
+    {
+      "epoch": 1.1457564575645756,
+      "grad_norm": 0.33789584779157467,
+      "learning_rate": 3.8875303735138355e-05,
+      "loss": 0.1621,
+      "step": 1242
+    },
+    {
+      "epoch": 1.146678966789668,
+      "grad_norm": 0.38525392066538,
+      "learning_rate": 3.885296739697332e-05,
+      "loss": 0.2028,
+      "step": 1243
+    },
+    {
+      "epoch": 1.1476014760147601,
+      "grad_norm": 0.4074987067807514,
+      "learning_rate": 3.883061508921439e-05,
+      "loss": 0.1746,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1485239852398523,
+      "grad_norm": 0.3499613895655867,
+      "learning_rate": 3.880824683762914e-05,
+      "loss": 0.1565,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1494464944649447,
+      "grad_norm": 0.35799322013488066,
+      "learning_rate": 3.87858626680035e-05,
+      "loss": 0.1771,
+      "step": 1246
+    },
+    {
+      "epoch": 1.150369003690037,
+      "grad_norm": 0.41086514997229345,
+      "learning_rate": 3.876346260614179e-05,
+      "loss": 0.1508,
+      "step": 1247
+    },
+    {
+      "epoch": 1.151291512915129,
+      "grad_norm": 0.39013582251282924,
+      "learning_rate": 3.874104667786661e-05,
+      "loss": 0.175,
+      "step": 1248
+    },
+    {
+      "epoch": 1.1522140221402215,
+      "grad_norm": 0.3543398516359713,
+      "learning_rate": 3.871861490901888e-05,
+      "loss": 0.1592,
+      "step": 1249
+    },
+    {
+      "epoch": 1.1531365313653137,
+      "grad_norm": 0.3951549154139269,
+      "learning_rate": 3.869616732545777e-05,
+      "loss": 0.1762,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1540590405904059,
+      "grad_norm": 0.40136372535159015,
+      "learning_rate": 3.867370395306068e-05,
+      "loss": 0.1703,
+      "step": 1251
+    },
+    {
+      "epoch": 1.1549815498154983,
+      "grad_norm": 0.353467602286537,
+      "learning_rate": 3.8651224817723194e-05,
+      "loss": 0.1662,
+      "step": 1252
+    },
+    {
+      "epoch": 1.1559040590405905,
+      "grad_norm": 0.384811288693674,
+      "learning_rate": 3.862872994535912e-05,
+      "loss": 0.1784,
+      "step": 1253
+    },
+    {
+      "epoch": 1.1568265682656826,
+      "grad_norm": 0.37004350457797447,
+      "learning_rate": 3.860621936190035e-05,
+      "loss": 0.1595,
+      "step": 1254
+    },
+    {
+      "epoch": 1.1577490774907748,
+      "grad_norm": 0.38635850669837396,
+      "learning_rate": 3.8583693093296914e-05,
+      "loss": 0.1712,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1586715867158672,
+      "grad_norm": 0.436738373842028,
+      "learning_rate": 3.8561151165516925e-05,
+      "loss": 0.187,
+      "step": 1256
+    },
+    {
+      "epoch": 1.1595940959409594,
+      "grad_norm": 0.37161247978898304,
+      "learning_rate": 3.853859360454654e-05,
+      "loss": 0.1781,
+      "step": 1257
+    },
+    {
+      "epoch": 1.1605166051660516,
+      "grad_norm": 0.3596937247805052,
+      "learning_rate": 3.851602043638994e-05,
+      "loss": 0.187,
+      "step": 1258
+    },
+    {
+      "epoch": 1.161439114391144,
+      "grad_norm": 0.3690606390753706,
+      "learning_rate": 3.84934316870693e-05,
+      "loss": 0.1481,
+      "step": 1259
+    },
+    {
+      "epoch": 1.1623616236162362,
+      "grad_norm": 0.3546601796031542,
+      "learning_rate": 3.847082738262477e-05,
+      "loss": 0.1629,
+      "step": 1260
+    },
+    {
+      "epoch": 1.1632841328413284,
+      "grad_norm": 0.32931616066822833,
+      "learning_rate": 3.84482075491144e-05,
+      "loss": 0.1668,
+      "step": 1261
+    },
+    {
+      "epoch": 1.1642066420664205,
+      "grad_norm": 0.33058419513964465,
+      "learning_rate": 3.842557221261415e-05,
+      "loss": 0.1495,
+      "step": 1262
+    },
+    {
+      "epoch": 1.165129151291513,
+      "grad_norm": 0.3739966388174159,
+      "learning_rate": 3.840292139921789e-05,
+      "loss": 0.1668,
+      "step": 1263
+    },
+    {
+      "epoch": 1.1660516605166051,
+      "grad_norm": 0.3945416380278711,
+      "learning_rate": 3.8380255135037285e-05,
+      "loss": 0.159,
+      "step": 1264
+    },
+    {
+      "epoch": 1.1669741697416973,
+      "grad_norm": 0.39559579609047163,
+      "learning_rate": 3.8357573446201825e-05,
+      "loss": 0.1769,
+      "step": 1265
+    },
+    {
+      "epoch": 1.1678966789667897,
+      "grad_norm": 0.3668179056990549,
+      "learning_rate": 3.833487635885881e-05,
+      "loss": 0.1634,
+      "step": 1266
+    },
+    {
+      "epoch": 1.168819188191882,
+      "grad_norm": 0.38166202060178234,
+      "learning_rate": 3.8312163899173234e-05,
+      "loss": 0.1572,
+      "step": 1267
+    },
+    {
+      "epoch": 1.169741697416974,
+      "grad_norm": 0.3770071600155405,
+      "learning_rate": 3.828943609332787e-05,
+      "loss": 0.1656,
+      "step": 1268
+    },
+    {
+      "epoch": 1.1706642066420665,
+      "grad_norm": 0.34029454152305877,
+      "learning_rate": 3.8266692967523156e-05,
+      "loss": 0.1631,
+      "step": 1269
+    },
+    {
+      "epoch": 1.1715867158671587,
+      "grad_norm": 0.3441263852387389,
+      "learning_rate": 3.824393454797718e-05,
+      "loss": 0.1722,
+      "step": 1270
+    },
+    {
+      "epoch": 1.1725092250922509,
+      "grad_norm": 0.45946161597400226,
+      "learning_rate": 3.8221160860925666e-05,
+      "loss": 0.1966,
+      "step": 1271
+    },
+    {
+      "epoch": 1.1734317343173433,
+      "grad_norm": 0.3743764303189791,
+      "learning_rate": 3.8198371932621965e-05,
+      "loss": 0.1725,
+      "step": 1272
+    },
+    {
+      "epoch": 1.1743542435424354,
+      "grad_norm": 0.3789734722999809,
+      "learning_rate": 3.817556778933698e-05,
+      "loss": 0.166,
+      "step": 1273
+    },
+    {
+      "epoch": 1.1752767527675276,
+      "grad_norm": 0.3540465811335595,
+      "learning_rate": 3.815274845735912e-05,
+      "loss": 0.1662,
+      "step": 1274
+    },
+    {
+      "epoch": 1.17619926199262,
+      "grad_norm": 0.3508442031273123,
+      "learning_rate": 3.812991396299437e-05,
+      "loss": 0.1666,
+      "step": 1275
+    },
+    {
+      "epoch": 1.1771217712177122,
+      "grad_norm": 0.39654898942218,
+      "learning_rate": 3.8107064332566136e-05,
+      "loss": 0.1814,
+      "step": 1276
+    },
+    {
+      "epoch": 1.1780442804428044,
+      "grad_norm": 0.3878983535558658,
+      "learning_rate": 3.8084199592415305e-05,
+      "loss": 0.1845,
+      "step": 1277
+    },
+    {
+      "epoch": 1.1789667896678966,
+      "grad_norm": 0.39812963500678517,
+      "learning_rate": 3.8061319768900175e-05,
+      "loss": 0.1774,
+      "step": 1278
+    },
+    {
+      "epoch": 1.179889298892989,
+      "grad_norm": 0.356504948337044,
+      "learning_rate": 3.8038424888396416e-05,
+      "loss": 0.1466,
+      "step": 1279
+    },
+    {
+      "epoch": 1.1808118081180812,
+      "grad_norm": 0.37203533726446586,
+      "learning_rate": 3.801551497729709e-05,
+      "loss": 0.1826,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1817343173431734,
+      "grad_norm": 0.39282813959728746,
+      "learning_rate": 3.799259006201255e-05,
+      "loss": 0.1681,
+      "step": 1281
+    },
+    {
+      "epoch": 1.1826568265682658,
+      "grad_norm": 0.42248503596639037,
+      "learning_rate": 3.796965016897047e-05,
+      "loss": 0.1794,
+      "step": 1282
+    },
+    {
+      "epoch": 1.183579335793358,
+      "grad_norm": 0.3418308160696917,
+      "learning_rate": 3.7946695324615775e-05,
+      "loss": 0.1525,
+      "step": 1283
+    },
+    {
+      "epoch": 1.1845018450184501,
+      "grad_norm": 0.3725828576686521,
+      "learning_rate": 3.7923725555410636e-05,
+      "loss": 0.1716,
+      "step": 1284
+    },
+    {
+      "epoch": 1.1854243542435423,
+      "grad_norm": 0.32818516577429657,
+      "learning_rate": 3.790074088783443e-05,
+      "loss": 0.1512,
+      "step": 1285
+    },
+    {
+      "epoch": 1.1863468634686347,
+      "grad_norm": 0.3564761683059801,
+      "learning_rate": 3.78777413483837e-05,
+      "loss": 0.153,
+      "step": 1286
+    },
+    {
+      "epoch": 1.187269372693727,
+      "grad_norm": 0.42092652414606263,
+      "learning_rate": 3.785472696357214e-05,
+      "loss": 0.1932,
+      "step": 1287
+    },
+    {
+      "epoch": 1.188191881918819,
+      "grad_norm": 0.4086237832286056,
+      "learning_rate": 3.783169775993055e-05,
+      "loss": 0.1823,
+      "step": 1288
+    },
+    {
+      "epoch": 1.1891143911439115,
+      "grad_norm": 0.37861293020683706,
+      "learning_rate": 3.780865376400682e-05,
+      "loss": 0.1705,
+      "step": 1289
+    },
+    {
+      "epoch": 1.1900369003690037,
+      "grad_norm": 0.4672151127767379,
+      "learning_rate": 3.7785595002365884e-05,
+      "loss": 0.2024,
+      "step": 1290
+    },
+    {
+      "epoch": 1.1909594095940959,
+      "grad_norm": 0.3678600637225433,
+      "learning_rate": 3.7762521501589723e-05,
+      "loss": 0.1683,
+      "step": 1291
+    },
+    {
+      "epoch": 1.1918819188191883,
+      "grad_norm": 0.38876778522083116,
+      "learning_rate": 3.773943328827728e-05,
+      "loss": 0.1681,
+      "step": 1292
+    },
+    {
+      "epoch": 1.1928044280442804,
+      "grad_norm": 0.4205936397419189,
+      "learning_rate": 3.771633038904446e-05,
+      "loss": 0.1882,
+      "step": 1293
+    },
+    {
+      "epoch": 1.1937269372693726,
+      "grad_norm": 0.41562563425985843,
+      "learning_rate": 3.769321283052412e-05,
+      "loss": 0.1657,
+      "step": 1294
+    },
+    {
+      "epoch": 1.194649446494465,
+      "grad_norm": 0.34433052253776564,
+      "learning_rate": 3.7670080639366004e-05,
+      "loss": 0.1528,
+      "step": 1295
+    },
+    {
+      "epoch": 1.1955719557195572,
+      "grad_norm": 0.38136188611696903,
+      "learning_rate": 3.764693384223671e-05,
+      "loss": 0.177,
+      "step": 1296
+    },
+    {
+      "epoch": 1.1964944649446494,
+      "grad_norm": 0.4362631823446143,
+      "learning_rate": 3.76237724658197e-05,
+      "loss": 0.1834,
+      "step": 1297
+    },
+    {
+      "epoch": 1.1974169741697418,
+      "grad_norm": 0.4490202044552004,
+      "learning_rate": 3.7600596536815224e-05,
+      "loss": 0.1883,
+      "step": 1298
+    },
+    {
+      "epoch": 1.198339483394834,
+      "grad_norm": 0.4159580498109915,
+      "learning_rate": 3.7577406081940314e-05,
+      "loss": 0.1821,
+      "step": 1299
+    },
+    {
+      "epoch": 1.1992619926199262,
+      "grad_norm": 0.3754259049777767,
+      "learning_rate": 3.7554201127928744e-05,
+      "loss": 0.1637,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2001845018450186,
+      "grad_norm": 0.39487303704772997,
+      "learning_rate": 3.753098170153102e-05,
+      "loss": 0.1751,
+      "step": 1301
+    },
+    {
+      "epoch": 1.2011070110701108,
+      "grad_norm": 0.38006135161592924,
+      "learning_rate": 3.750774782951431e-05,
+      "loss": 0.1832,
+      "step": 1302
+    },
+    {
+      "epoch": 1.202029520295203,
+      "grad_norm": 0.373794173672362,
+      "learning_rate": 3.7484499538662424e-05,
+      "loss": 0.1638,
+      "step": 1303
+    },
+    {
+      "epoch": 1.2029520295202951,
+      "grad_norm": 0.39680441858428456,
+      "learning_rate": 3.746123685577585e-05,
+      "loss": 0.1784,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2038745387453875,
+      "grad_norm": 0.36870033261938107,
+      "learning_rate": 3.743795980767159e-05,
+      "loss": 0.1682,
+      "step": 1305
+    },
+    {
+      "epoch": 1.2047970479704797,
+      "grad_norm": 0.4399212266489506,
+      "learning_rate": 3.741466842118327e-05,
+      "loss": 0.1771,
+      "step": 1306
+    },
+    {
+      "epoch": 1.205719557195572,
+      "grad_norm": 0.34685405203127034,
+      "learning_rate": 3.739136272316102e-05,
+      "loss": 0.1706,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2066420664206643,
+      "grad_norm": 0.36429649342017056,
+      "learning_rate": 3.736804274047145e-05,
+      "loss": 0.1596,
+      "step": 1308
+    },
+    {
+      "epoch": 1.2075645756457565,
+      "grad_norm": 0.3983863474001515,
+      "learning_rate": 3.734470849999767e-05,
+      "loss": 0.1708,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2084870848708487,
+      "grad_norm": 0.3655548087385067,
+      "learning_rate": 3.732136002863922e-05,
+      "loss": 0.1698,
+      "step": 1310
+    },
+    {
+      "epoch": 1.2094095940959408,
+      "grad_norm": 0.4039655923900139,
+      "learning_rate": 3.729799735331203e-05,
+      "loss": 0.1778,
+      "step": 1311
+    },
+    {
+      "epoch": 1.2103321033210332,
+      "grad_norm": 0.35516821761702266,
+      "learning_rate": 3.727462050094841e-05,
+      "loss": 0.1459,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2112546125461254,
+      "grad_norm": 0.44642489434454374,
+      "learning_rate": 3.7251229498497e-05,
+      "loss": 0.1767,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2121771217712176,
+      "grad_norm": 0.40901821529429044,
+      "learning_rate": 3.72278243729228e-05,
+      "loss": 0.1797,
+      "step": 1314
+    },
+    {
+      "epoch": 1.21309963099631,
+      "grad_norm": 0.3801414969463073,
+      "learning_rate": 3.7204405151207036e-05,
+      "loss": 0.1921,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2140221402214022,
+      "grad_norm": 0.35808170261872035,
+      "learning_rate": 3.718097186034721e-05,
+      "loss": 0.1663,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2149446494464944,
+      "grad_norm": 0.42564468017637885,
+      "learning_rate": 3.715752452735704e-05,
+      "loss": 0.1888,
+      "step": 1317
+    },
+    {
+      "epoch": 1.2158671586715868,
+      "grad_norm": 0.37950666546171086,
+      "learning_rate": 3.7134063179266425e-05,
+      "loss": 0.1562,
+      "step": 1318
+    },
+    {
+      "epoch": 1.216789667896679,
+      "grad_norm": 0.3843938105095001,
+      "learning_rate": 3.711058784312144e-05,
+      "loss": 0.1777,
+      "step": 1319
+    },
+    {
+      "epoch": 1.2177121771217712,
+      "grad_norm": 0.36524969537809376,
+      "learning_rate": 3.708709854598425e-05,
+      "loss": 0.1649,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2186346863468636,
+      "grad_norm": 0.4003937659265883,
+      "learning_rate": 3.706359531493316e-05,
+      "loss": 0.1688,
+      "step": 1321
+    },
+    {
+      "epoch": 1.2195571955719557,
+      "grad_norm": 0.39405861933597286,
+      "learning_rate": 3.7040078177062484e-05,
+      "loss": 0.1879,
+      "step": 1322
+    },
+    {
+      "epoch": 1.220479704797048,
+      "grad_norm": 0.37765746210969386,
+      "learning_rate": 3.701654715948264e-05,
+      "loss": 0.1865,
+      "step": 1323
+    },
+    {
+      "epoch": 1.2214022140221403,
+      "grad_norm": 0.4818548000488366,
+      "learning_rate": 3.6993002289319955e-05,
+      "loss": 0.1646,
+      "step": 1324
+    },
+    {
+      "epoch": 1.2223247232472325,
+      "grad_norm": 0.39356600986755663,
+      "learning_rate": 3.6969443593716804e-05,
+      "loss": 0.1728,
+      "step": 1325
+    },
+    {
+      "epoch": 1.2232472324723247,
+      "grad_norm": 0.5377447965710058,
+      "learning_rate": 3.694587109983147e-05,
+      "loss": 0.1827,
+      "step": 1326
+    },
+    {
+      "epoch": 1.2241697416974169,
+      "grad_norm": 0.4146277637023804,
+      "learning_rate": 3.692228483483812e-05,
+      "loss": 0.1821,
+      "step": 1327
+    },
+    {
+      "epoch": 1.2250922509225093,
+      "grad_norm": 0.4362453081571424,
+      "learning_rate": 3.689868482592684e-05,
+      "loss": 0.1963,
+      "step": 1328
+    },
+    {
+      "epoch": 1.2260147601476015,
+      "grad_norm": 0.36800404753226107,
+      "learning_rate": 3.6875071100303523e-05,
+      "loss": 0.18,
+      "step": 1329
+    },
+    {
+      "epoch": 1.2269372693726937,
+      "grad_norm": 0.36808130166836256,
+      "learning_rate": 3.685144368518991e-05,
+      "loss": 0.1481,
+      "step": 1330
+    },
+    {
+      "epoch": 1.227859778597786,
+      "grad_norm": 0.3631608839553618,
+      "learning_rate": 3.682780260782348e-05,
+      "loss": 0.177,
+      "step": 1331
+    },
+    {
+      "epoch": 1.2287822878228782,
+      "grad_norm": 0.32755840956082244,
+      "learning_rate": 3.680414789545749e-05,
+      "loss": 0.1573,
+      "step": 1332
+    },
+    {
+      "epoch": 1.2297047970479704,
+      "grad_norm": 0.41375648373583096,
+      "learning_rate": 3.678047957536092e-05,
+      "loss": 0.1833,
+      "step": 1333
+    },
+    {
+      "epoch": 1.2306273062730628,
+      "grad_norm": 0.399469699965947,
+      "learning_rate": 3.675679767481842e-05,
+      "loss": 0.187,
+      "step": 1334
+    },
+    {
+      "epoch": 1.231549815498155,
+      "grad_norm": 0.3659805491727511,
+      "learning_rate": 3.67331022211303e-05,
+      "loss": 0.1573,
+      "step": 1335
+    },
+    {
+      "epoch": 1.2324723247232472,
+      "grad_norm": 0.3799332849809668,
+      "learning_rate": 3.670939324161251e-05,
+      "loss": 0.1798,
+      "step": 1336
+    },
+    {
+      "epoch": 1.2333948339483394,
+      "grad_norm": 0.3972461594400006,
+      "learning_rate": 3.668567076359656e-05,
+      "loss": 0.1681,
+      "step": 1337
+    },
+    {
+      "epoch": 1.2343173431734318,
+      "grad_norm": 0.3914441459651774,
+      "learning_rate": 3.666193481442954e-05,
+      "loss": 0.17,
+      "step": 1338
+    },
+    {
+      "epoch": 1.235239852398524,
+      "grad_norm": 0.36242286305851845,
+      "learning_rate": 3.6638185421474084e-05,
+      "loss": 0.1712,
+      "step": 1339
+    },
+    {
+      "epoch": 1.2361623616236161,
+      "grad_norm": 0.36469642878502373,
+      "learning_rate": 3.66144226121083e-05,
+      "loss": 0.1654,
+      "step": 1340
+    },
+    {
+      "epoch": 1.2370848708487086,
+      "grad_norm": 0.3917313439062194,
+      "learning_rate": 3.659064641372576e-05,
+      "loss": 0.168,
+      "step": 1341
+    },
+    {
+      "epoch": 1.2380073800738007,
+      "grad_norm": 0.3617694791967727,
+      "learning_rate": 3.6566856853735516e-05,
+      "loss": 0.1726,
+      "step": 1342
+    },
+    {
+      "epoch": 1.238929889298893,
+      "grad_norm": 0.4141158252212677,
+      "learning_rate": 3.654305395956195e-05,
+      "loss": 0.1849,
+      "step": 1343
+    },
+    {
+      "epoch": 1.2398523985239853,
+      "grad_norm": 0.4059832013250797,
+      "learning_rate": 3.651923775864488e-05,
+      "loss": 0.1519,
+      "step": 1344
+    },
+    {
+      "epoch": 1.2407749077490775,
+      "grad_norm": 0.37433954496738203,
+      "learning_rate": 3.6495408278439426e-05,
+      "loss": 0.1625,
+      "step": 1345
+    },
+    {
+      "epoch": 1.2416974169741697,
+      "grad_norm": 0.3918972235044372,
+      "learning_rate": 3.647156554641603e-05,
+      "loss": 0.164,
+      "step": 1346
+    },
+    {
+      "epoch": 1.242619926199262,
+      "grad_norm": 0.41833676506254874,
+      "learning_rate": 3.644770959006042e-05,
+      "loss": 0.1872,
+      "step": 1347
+    },
+    {
+      "epoch": 1.2435424354243543,
+      "grad_norm": 0.4218150024250036,
+      "learning_rate": 3.642384043687356e-05,
+      "loss": 0.163,
+      "step": 1348
+    },
+    {
+      "epoch": 1.2444649446494465,
+      "grad_norm": 0.39570660118087275,
+      "learning_rate": 3.6399958114371595e-05,
+      "loss": 0.1816,
+      "step": 1349
+    },
+    {
+      "epoch": 1.2453874538745389,
+      "grad_norm": 0.3707416696700744,
+      "learning_rate": 3.637606265008592e-05,
+      "loss": 0.1765,
+      "step": 1350
+    },
+    {
+      "epoch": 1.246309963099631,
+      "grad_norm": 0.3740850457534175,
+      "learning_rate": 3.635215407156302e-05,
+      "loss": 0.1795,
+      "step": 1351
+    },
+    {
+      "epoch": 1.2472324723247232,
+      "grad_norm": 0.40097805285714844,
+      "learning_rate": 3.632823240636452e-05,
+      "loss": 0.1835,
+      "step": 1352
+    },
+    {
+      "epoch": 1.2481549815498154,
+      "grad_norm": 0.3811162867312523,
+      "learning_rate": 3.6304297682067144e-05,
+      "loss": 0.1752,
+      "step": 1353
+    },
+    {
+      "epoch": 1.2490774907749078,
+      "grad_norm": 0.33638864950228525,
+      "learning_rate": 3.628034992626265e-05,
+      "loss": 0.1514,
+      "step": 1354
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4225820180427333,
+      "learning_rate": 3.6256389166557825e-05,
+      "loss": 0.1761,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2509225092250922,
+      "grad_norm": 0.41138341726215444,
+      "learning_rate": 3.623241543057444e-05,
+      "loss": 0.1765,
+      "step": 1356
+    },
+    {
+      "epoch": 1.2518450184501844,
+      "grad_norm": 0.4001534683532543,
+      "learning_rate": 3.6208428745949255e-05,
+      "loss": 0.1642,
+      "step": 1357
+    },
+    {
+      "epoch": 1.2527675276752768,
+      "grad_norm": 0.43869036112274273,
+      "learning_rate": 3.618442914033392e-05,
+      "loss": 0.1895,
+      "step": 1358
+    },
+    {
+      "epoch": 1.253690036900369,
+      "grad_norm": 0.3739304339036808,
+      "learning_rate": 3.616041664139499e-05,
+      "loss": 0.1586,
+      "step": 1359
+    },
+    {
+      "epoch": 1.2546125461254611,
+      "grad_norm": 0.35109359825230413,
+      "learning_rate": 3.613639127681389e-05,
+      "loss": 0.1753,
+      "step": 1360
+    },
+    {
+      "epoch": 1.2555350553505535,
+      "grad_norm": 0.3577963740043067,
+      "learning_rate": 3.61123530742869e-05,
+      "loss": 0.1769,
+      "step": 1361
+    },
+    {
+      "epoch": 1.2564575645756457,
+      "grad_norm": 0.3843171510463367,
+      "learning_rate": 3.608830206152503e-05,
+      "loss": 0.1823,
+      "step": 1362
+    },
+    {
+      "epoch": 1.257380073800738,
+      "grad_norm": 0.3598594630906758,
+      "learning_rate": 3.6064238266254145e-05,
+      "loss": 0.1795,
+      "step": 1363
+    },
+    {
+      "epoch": 1.2583025830258303,
+      "grad_norm": 0.37345021729424344,
+      "learning_rate": 3.6040161716214774e-05,
+      "loss": 0.151,
+      "step": 1364
+    },
+    {
+      "epoch": 1.2592250922509225,
+      "grad_norm": 0.364594440421076,
+      "learning_rate": 3.601607243916219e-05,
+      "loss": 0.1564,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2601476014760147,
+      "grad_norm": 0.3760450872663813,
+      "learning_rate": 3.599197046286632e-05,
+      "loss": 0.169,
+      "step": 1366
+    },
+    {
+      "epoch": 1.261070110701107,
+      "grad_norm": 0.3349224229844502,
+      "learning_rate": 3.596785581511174e-05,
+      "loss": 0.1379,
+      "step": 1367
+    },
+    {
+      "epoch": 1.2619926199261993,
+      "grad_norm": 0.4197601137515256,
+      "learning_rate": 3.594372852369763e-05,
+      "loss": 0.1779,
+      "step": 1368
+    },
+    {
+      "epoch": 1.2629151291512914,
+      "grad_norm": 0.4252010238904596,
+      "learning_rate": 3.591958861643775e-05,
+      "loss": 0.1819,
+      "step": 1369
+    },
+    {
+      "epoch": 1.2638376383763839,
+      "grad_norm": 0.3726501790103951,
+      "learning_rate": 3.5895436121160386e-05,
+      "loss": 0.1709,
+      "step": 1370
+    },
+    {
+      "epoch": 1.264760147601476,
+      "grad_norm": 0.3649082841374998,
+      "learning_rate": 3.5871271065708354e-05,
+      "loss": 0.164,
+      "step": 1371
+    },
+    {
+      "epoch": 1.2656826568265682,
+      "grad_norm": 0.4709102508547808,
+      "learning_rate": 3.5847093477938956e-05,
+      "loss": 0.1951,
+      "step": 1372
+    },
+    {
+      "epoch": 1.2666051660516606,
+      "grad_norm": 0.36340144372771255,
+      "learning_rate": 3.5822903385723904e-05,
+      "loss": 0.1733,
+      "step": 1373
+    },
+    {
+      "epoch": 1.2675276752767528,
+      "grad_norm": 0.36838691907955257,
+      "learning_rate": 3.579870081694938e-05,
+      "loss": 0.1707,
+      "step": 1374
+    },
+    {
+      "epoch": 1.268450184501845,
+      "grad_norm": 0.379732273767842,
+      "learning_rate": 3.577448579951589e-05,
+      "loss": 0.1584,
+      "step": 1375
+    },
+    {
+      "epoch": 1.2693726937269374,
+      "grad_norm": 0.4069962634671663,
+      "learning_rate": 3.575025836133833e-05,
+      "loss": 0.1775,
+      "step": 1376
+    },
+    {
+      "epoch": 1.2702952029520296,
+      "grad_norm": 0.36298862782828534,
+      "learning_rate": 3.5726018530345915e-05,
+      "loss": 0.1565,
+      "step": 1377
+    },
+    {
+      "epoch": 1.2712177121771218,
+      "grad_norm": 0.35890423156426643,
+      "learning_rate": 3.5701766334482114e-05,
+      "loss": 0.1797,
+      "step": 1378
+    },
+    {
+      "epoch": 1.272140221402214,
+      "grad_norm": 0.404840860825177,
+      "learning_rate": 3.5677501801704685e-05,
+      "loss": 0.1996,
+      "step": 1379
+    },
+    {
+      "epoch": 1.2730627306273063,
+      "grad_norm": 0.34154877692677005,
+      "learning_rate": 3.565322495998559e-05,
+      "loss": 0.1577,
+      "step": 1380
+    },
+    {
+      "epoch": 1.2739852398523985,
+      "grad_norm": 0.37781924095596964,
+      "learning_rate": 3.5628935837310984e-05,
+      "loss": 0.1688,
+      "step": 1381
+    },
+    {
+      "epoch": 1.2749077490774907,
+      "grad_norm": 0.3574258267368524,
+      "learning_rate": 3.5604634461681184e-05,
+      "loss": 0.1646,
+      "step": 1382
+    },
+    {
+      "epoch": 1.275830258302583,
+      "grad_norm": 0.3885364748872123,
+      "learning_rate": 3.5580320861110625e-05,
+      "loss": 0.1756,
+      "step": 1383
+    },
+    {
+      "epoch": 1.2767527675276753,
+      "grad_norm": 0.3809434520908714,
+      "learning_rate": 3.555599506362784e-05,
+      "loss": 0.1548,
+      "step": 1384
+    },
+    {
+      "epoch": 1.2776752767527675,
+      "grad_norm": 0.3881374004257548,
+      "learning_rate": 3.5531657097275425e-05,
+      "loss": 0.1685,
+      "step": 1385
+    },
+    {
+      "epoch": 1.2785977859778597,
+      "grad_norm": 0.3878220749254127,
+      "learning_rate": 3.550730699010999e-05,
+      "loss": 0.1685,
+      "step": 1386
+    },
+    {
+      "epoch": 1.279520295202952,
+      "grad_norm": 0.566907399079411,
+      "learning_rate": 3.5482944770202145e-05,
+      "loss": 0.1975,
+      "step": 1387
+    },
+    {
+      "epoch": 1.2804428044280443,
+      "grad_norm": 0.36180413368577086,
+      "learning_rate": 3.545857046563649e-05,
+      "loss": 0.149,
+      "step": 1388
+    },
+    {
+      "epoch": 1.2813653136531364,
+      "grad_norm": 0.3552529788203019,
+      "learning_rate": 3.543418410451152e-05,
+      "loss": 0.1789,
+      "step": 1389
+    },
+    {
+      "epoch": 1.2822878228782288,
+      "grad_norm": 0.3985813856389354,
+      "learning_rate": 3.540978571493966e-05,
+      "loss": 0.1791,
+      "step": 1390
+    },
+    {
+      "epoch": 1.283210332103321,
+      "grad_norm": 0.3654350181181372,
+      "learning_rate": 3.5385375325047166e-05,
+      "loss": 0.1572,
+      "step": 1391
+    },
+    {
+      "epoch": 1.2841328413284132,
+      "grad_norm": 0.3651885453567304,
+      "learning_rate": 3.536095296297415e-05,
+      "loss": 0.15,
+      "step": 1392
+    },
+    {
+      "epoch": 1.2850553505535056,
+      "grad_norm": 0.41262729710097773,
+      "learning_rate": 3.533651865687454e-05,
+      "loss": 0.1771,
+      "step": 1393
+    },
+    {
+      "epoch": 1.2859778597785978,
+      "grad_norm": 0.3895770830415364,
+      "learning_rate": 3.5312072434915986e-05,
+      "loss": 0.1636,
+      "step": 1394
+    },
+    {
+      "epoch": 1.28690036900369,
+      "grad_norm": 0.4216840380459189,
+      "learning_rate": 3.528761432527992e-05,
+      "loss": 0.1901,
+      "step": 1395
+    },
+    {
+      "epoch": 1.2878228782287824,
+      "grad_norm": 0.370406659545189,
+      "learning_rate": 3.5263144356161476e-05,
+      "loss": 0.1783,
+      "step": 1396
+    },
+    {
+      "epoch": 1.2887453874538746,
+      "grad_norm": 0.3570954429973423,
+      "learning_rate": 3.523866255576943e-05,
+      "loss": 0.1706,
+      "step": 1397
+    },
+    {
+      "epoch": 1.2896678966789668,
+      "grad_norm": 0.343995922631194,
+      "learning_rate": 3.52141689523262e-05,
+      "loss": 0.1491,
+      "step": 1398
+    },
+    {
+      "epoch": 1.2905904059040592,
+      "grad_norm": 0.3900651531256659,
+      "learning_rate": 3.518966357406786e-05,
+      "loss": 0.1892,
+      "step": 1399
+    },
+    {
+      "epoch": 1.2915129151291513,
+      "grad_norm": 0.3536298519135728,
+      "learning_rate": 3.516514644924398e-05,
+      "loss": 0.1513,
+      "step": 1400
+    },
+    {
+      "epoch": 1.2924354243542435,
+      "grad_norm": 0.42665857944563257,
+      "learning_rate": 3.5140617606117736e-05,
+      "loss": 0.1705,
+      "step": 1401
+    },
+    {
+      "epoch": 1.293357933579336,
+      "grad_norm": 0.38719001646314355,
+      "learning_rate": 3.511607707296579e-05,
+      "loss": 0.1779,
+      "step": 1402
+    },
+    {
+      "epoch": 1.294280442804428,
+      "grad_norm": 0.3927927302690867,
+      "learning_rate": 3.509152487807826e-05,
+      "loss": 0.1593,
+      "step": 1403
+    },
+    {
+      "epoch": 1.2952029520295203,
+      "grad_norm": 0.3848428495517545,
+      "learning_rate": 3.506696104975875e-05,
+      "loss": 0.1832,
+      "step": 1404
+    },
+    {
+      "epoch": 1.2961254612546125,
+      "grad_norm": 0.3610052410007685,
+      "learning_rate": 3.504238561632424e-05,
+      "loss": 0.1592,
+      "step": 1405
+    },
+    {
+      "epoch": 1.2970479704797047,
+      "grad_norm": 0.3907677657069571,
+      "learning_rate": 3.5017798606105095e-05,
+      "loss": 0.1797,
+      "step": 1406
+    },
+    {
+      "epoch": 1.297970479704797,
+      "grad_norm": 0.35151063137831756,
+      "learning_rate": 3.499320004744505e-05,
+      "loss": 0.1687,
+      "step": 1407
+    },
+    {
+      "epoch": 1.2988929889298892,
+      "grad_norm": 0.40429797101240167,
+      "learning_rate": 3.496858996870111e-05,
+      "loss": 0.1915,
+      "step": 1408
+    },
+    {
+      "epoch": 1.2998154981549814,
+      "grad_norm": 0.38685864677248,
+      "learning_rate": 3.49439683982436e-05,
+      "loss": 0.1639,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3007380073800738,
+      "grad_norm": 0.41763155701235083,
+      "learning_rate": 3.491933536445606e-05,
+      "loss": 0.2078,
+      "step": 1410
+    },
+    {
+      "epoch": 1.301660516605166,
+      "grad_norm": 0.3897175145433721,
+      "learning_rate": 3.489469089573529e-05,
+      "loss": 0.1899,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3025830258302582,
+      "grad_norm": 0.3950918324246369,
+      "learning_rate": 3.487003502049122e-05,
+      "loss": 0.1789,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3035055350553506,
+      "grad_norm": 0.3741381176221496,
+      "learning_rate": 3.484536776714694e-05,
+      "loss": 0.1622,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3044280442804428,
+      "grad_norm": 0.3572545260636694,
+      "learning_rate": 3.482068916413871e-05,
+      "loss": 0.1677,
+      "step": 1414
+    },
+    {
+      "epoch": 1.305350553505535,
+      "grad_norm": 0.33245502360717216,
+      "learning_rate": 3.47959992399158e-05,
+      "loss": 0.1495,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3062730627306274,
+      "grad_norm": 0.37138142998504686,
+      "learning_rate": 3.477129802294057e-05,
+      "loss": 0.1591,
+      "step": 1416
+    },
+    {
+      "epoch": 1.3071955719557196,
+      "grad_norm": 0.3668631190741448,
+      "learning_rate": 3.47465855416884e-05,
+      "loss": 0.1586,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3081180811808117,
+      "grad_norm": 0.3441380348515503,
+      "learning_rate": 3.472186182464765e-05,
+      "loss": 0.16,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3090405904059041,
+      "grad_norm": 0.39242981438000046,
+      "learning_rate": 3.469712690031962e-05,
+      "loss": 0.186,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3099630996309963,
+      "grad_norm": 0.3818463039421942,
+      "learning_rate": 3.467238079721855e-05,
+      "loss": 0.1602,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3108856088560885,
+      "grad_norm": 0.40650210769142753,
+      "learning_rate": 3.464762354387155e-05,
+      "loss": 0.1713,
+      "step": 1421
+    },
+    {
+      "epoch": 1.311808118081181,
+      "grad_norm": 0.41220414363195734,
+      "learning_rate": 3.4622855168818586e-05,
+      "loss": 0.1785,
+      "step": 1422
+    },
+    {
+      "epoch": 1.312730627306273,
+      "grad_norm": 0.37792728496286754,
+      "learning_rate": 3.459807570061246e-05,
+      "loss": 0.1716,
+      "step": 1423
+    },
+    {
+      "epoch": 1.3136531365313653,
+      "grad_norm": 0.41173148841989027,
+      "learning_rate": 3.4573285167818744e-05,
+      "loss": 0.1645,
+      "step": 1424
+    },
+    {
+      "epoch": 1.3145756457564577,
+      "grad_norm": 0.3680120838951169,
+      "learning_rate": 3.454848359901578e-05,
+      "loss": 0.1679,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3154981549815499,
+      "grad_norm": 0.42016288755986253,
+      "learning_rate": 3.4523671022794616e-05,
+      "loss": 0.1776,
+      "step": 1426
+    },
+    {
+      "epoch": 1.316420664206642,
+      "grad_norm": 0.3497717481979851,
+      "learning_rate": 3.4498847467759e-05,
+      "loss": 0.162,
+      "step": 1427
+    },
+    {
+      "epoch": 1.3173431734317342,
+      "grad_norm": 0.36083858060108276,
+      "learning_rate": 3.447401296252535e-05,
+      "loss": 0.1675,
+      "step": 1428
+    },
+    {
+      "epoch": 1.3182656826568266,
+      "grad_norm": 0.3574519149758475,
+      "learning_rate": 3.444916753572266e-05,
+      "loss": 0.1585,
+      "step": 1429
+    },
+    {
+      "epoch": 1.3191881918819188,
+      "grad_norm": 0.4173982461504816,
+      "learning_rate": 3.442431121599259e-05,
+      "loss": 0.1781,
+      "step": 1430
+    },
+    {
+      "epoch": 1.320110701107011,
+      "grad_norm": 0.38466774843280815,
+      "learning_rate": 3.439944403198928e-05,
+      "loss": 0.1704,
+      "step": 1431
+    },
+    {
+      "epoch": 1.3210332103321032,
+      "grad_norm": 0.44274122553698614,
+      "learning_rate": 3.437456601237943e-05,
+      "loss": 0.1593,
+      "step": 1432
+    },
+    {
+      "epoch": 1.3219557195571956,
+      "grad_norm": 0.36275017892653194,
+      "learning_rate": 3.4349677185842245e-05,
+      "loss": 0.1534,
+      "step": 1433
+    },
+    {
+      "epoch": 1.3228782287822878,
+      "grad_norm": 0.407549047074853,
+      "learning_rate": 3.4324777581069356e-05,
+      "loss": 0.1761,
+      "step": 1434
+    },
+    {
+      "epoch": 1.32380073800738,
+      "grad_norm": 0.4293837842471811,
+      "learning_rate": 3.4299867226764845e-05,
+      "loss": 0.1693,
+      "step": 1435
+    },
+    {
+      "epoch": 1.3247232472324724,
+      "grad_norm": 0.39258990949042727,
+      "learning_rate": 3.427494615164518e-05,
+      "loss": 0.1784,
+      "step": 1436
+    },
+    {
+      "epoch": 1.3256457564575646,
+      "grad_norm": 0.3650336270383585,
+      "learning_rate": 3.4250014384439175e-05,
+      "loss": 0.1678,
+      "step": 1437
+    },
+    {
+      "epoch": 1.3265682656826567,
+      "grad_norm": 0.3894315314074764,
+      "learning_rate": 3.4225071953887976e-05,
+      "loss": 0.1934,
+      "step": 1438
+    },
+    {
+      "epoch": 1.3274907749077491,
+      "grad_norm": 0.36263806255056463,
+      "learning_rate": 3.4200118888745045e-05,
+      "loss": 0.1529,
+      "step": 1439
+    },
+    {
+      "epoch": 1.3284132841328413,
+      "grad_norm": 0.4263699101442625,
+      "learning_rate": 3.4175155217776055e-05,
+      "loss": 0.1602,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3293357933579335,
+      "grad_norm": 0.3239873105222901,
+      "learning_rate": 3.415018096975895e-05,
+      "loss": 0.1504,
+      "step": 1441
+    },
+    {
+      "epoch": 1.330258302583026,
+      "grad_norm": 0.41428571031323863,
+      "learning_rate": 3.412519617348384e-05,
+      "loss": 0.1554,
+      "step": 1442
+    },
+    {
+      "epoch": 1.331180811808118,
+      "grad_norm": 0.35636351141470096,
+      "learning_rate": 3.4100200857753026e-05,
+      "loss": 0.1511,
+      "step": 1443
+    },
+    {
+      "epoch": 1.3321033210332103,
+      "grad_norm": 0.4468881576942684,
+      "learning_rate": 3.407519505138089e-05,
+      "loss": 0.1688,
+      "step": 1444
+    },
+    {
+      "epoch": 1.3330258302583027,
+      "grad_norm": 0.3632320728678713,
+      "learning_rate": 3.4050178783193945e-05,
+      "loss": 0.187,
+      "step": 1445
+    },
+    {
+      "epoch": 1.3339483394833949,
+      "grad_norm": 0.40674020966682184,
+      "learning_rate": 3.402515208203076e-05,
+      "loss": 0.166,
+      "step": 1446
+    },
+    {
+      "epoch": 1.334870848708487,
+      "grad_norm": 0.3535751510672735,
+      "learning_rate": 3.4000114976741906e-05,
+      "loss": 0.1536,
+      "step": 1447
+    },
+    {
+      "epoch": 1.3357933579335795,
+      "grad_norm": 0.39981031351083296,
+      "learning_rate": 3.3975067496189965e-05,
+      "loss": 0.2023,
+      "step": 1448
+    },
+    {
+      "epoch": 1.3367158671586716,
+      "grad_norm": 0.3868102100726796,
+      "learning_rate": 3.3950009669249497e-05,
+      "loss": 0.1578,
+      "step": 1449
+    },
+    {
+      "epoch": 1.3376383763837638,
+      "grad_norm": 0.3707430060161426,
+      "learning_rate": 3.392494152480696e-05,
+      "loss": 0.1654,
+      "step": 1450
+    },
+    {
+      "epoch": 1.3385608856088562,
+      "grad_norm": 0.3418089472071561,
+      "learning_rate": 3.3899863091760715e-05,
+      "loss": 0.1735,
+      "step": 1451
+    },
+    {
+      "epoch": 1.3394833948339484,
+      "grad_norm": 0.4074415047811184,
+      "learning_rate": 3.387477439902099e-05,
+      "loss": 0.1602,
+      "step": 1452
+    },
+    {
+      "epoch": 1.3404059040590406,
+      "grad_norm": 0.39756053217292886,
+      "learning_rate": 3.384967547550984e-05,
+      "loss": 0.1507,
+      "step": 1453
+    },
+    {
+      "epoch": 1.3413284132841328,
+      "grad_norm": 0.4054167651529612,
+      "learning_rate": 3.38245663501611e-05,
+      "loss": 0.1834,
+      "step": 1454
+    },
+    {
+      "epoch": 1.3422509225092252,
+      "grad_norm": 0.3741082213467395,
+      "learning_rate": 3.379944705192039e-05,
+      "loss": 0.1724,
+      "step": 1455
+    },
+    {
+      "epoch": 1.3431734317343174,
+      "grad_norm": 0.3731975734916343,
+      "learning_rate": 3.377431760974503e-05,
+      "loss": 0.1818,
+      "step": 1456
+    },
+    {
+      "epoch": 1.3440959409594095,
+      "grad_norm": 0.4148220352374969,
+      "learning_rate": 3.3749178052604045e-05,
+      "loss": 0.1827,
+      "step": 1457
+    },
+    {
+      "epoch": 1.3450184501845017,
+      "grad_norm": 0.378581946781744,
+      "learning_rate": 3.372402840947814e-05,
+      "loss": 0.1557,
+      "step": 1458
+    },
+    {
+      "epoch": 1.3459409594095941,
+      "grad_norm": 0.3937199976821091,
+      "learning_rate": 3.3698868709359616e-05,
+      "loss": 0.175,
+      "step": 1459
+    },
+    {
+      "epoch": 1.3468634686346863,
+      "grad_norm": 0.38797450109320764,
+      "learning_rate": 3.367369898125238e-05,
+      "loss": 0.166,
+      "step": 1460
+    },
+    {
+      "epoch": 1.3477859778597785,
+      "grad_norm": 0.4216007726359451,
+      "learning_rate": 3.364851925417191e-05,
+      "loss": 0.192,
+      "step": 1461
+    },
+    {
+      "epoch": 1.348708487084871,
+      "grad_norm": 0.34470152781883956,
+      "learning_rate": 3.362332955714519e-05,
+      "loss": 0.1545,
+      "step": 1462
+    },
+    {
+      "epoch": 1.349630996309963,
+      "grad_norm": 0.36583757500717157,
+      "learning_rate": 3.359812991921072e-05,
+      "loss": 0.1602,
+      "step": 1463
+    },
+    {
+      "epoch": 1.3505535055350553,
+      "grad_norm": 0.5255217283137901,
+      "learning_rate": 3.357292036941844e-05,
+      "loss": 0.1775,
+      "step": 1464
+    },
+    {
+      "epoch": 1.3514760147601477,
+      "grad_norm": 0.38245817490974543,
+      "learning_rate": 3.3547700936829726e-05,
+      "loss": 0.1739,
+      "step": 1465
+    },
+    {
+      "epoch": 1.3523985239852399,
+      "grad_norm": 0.4061122342060611,
+      "learning_rate": 3.352247165051734e-05,
+      "loss": 0.1725,
+      "step": 1466
+    },
+    {
+      "epoch": 1.353321033210332,
+      "grad_norm": 0.3498602144068321,
+      "learning_rate": 3.349723253956542e-05,
+      "loss": 0.188,
+      "step": 1467
+    },
+    {
+      "epoch": 1.3542435424354244,
+      "grad_norm": 0.42855975241480126,
+      "learning_rate": 3.347198363306942e-05,
+      "loss": 0.1916,
+      "step": 1468
+    },
+    {
+      "epoch": 1.3551660516605166,
+      "grad_norm": 0.38571087950309607,
+      "learning_rate": 3.344672496013606e-05,
+      "loss": 0.1716,
+      "step": 1469
+    },
+    {
+      "epoch": 1.3560885608856088,
+      "grad_norm": 0.3909638801847444,
+      "learning_rate": 3.3421456549883366e-05,
+      "loss": 0.1714,
+      "step": 1470
+    },
+    {
+      "epoch": 1.3570110701107012,
+      "grad_norm": 0.3983408278731454,
+      "learning_rate": 3.339617843144057e-05,
+      "loss": 0.1637,
+      "step": 1471
+    },
+    {
+      "epoch": 1.3579335793357934,
+      "grad_norm": 0.40774414174127693,
+      "learning_rate": 3.337089063394807e-05,
+      "loss": 0.1554,
+      "step": 1472
+    },
+    {
+      "epoch": 1.3588560885608856,
+      "grad_norm": 0.38286568434794654,
+      "learning_rate": 3.334559318655746e-05,
+      "loss": 0.1706,
+      "step": 1473
+    },
+    {
+      "epoch": 1.359778597785978,
+      "grad_norm": 0.41029740965251404,
+      "learning_rate": 3.3320286118431444e-05,
+      "loss": 0.1837,
+      "step": 1474
+    },
+    {
+      "epoch": 1.3607011070110702,
+      "grad_norm": 0.3522151177885343,
+      "learning_rate": 3.32949694587438e-05,
+      "loss": 0.1641,
+      "step": 1475
+    },
+    {
+      "epoch": 1.3616236162361623,
+      "grad_norm": 0.41040542329855806,
+      "learning_rate": 3.3269643236679384e-05,
+      "loss": 0.1901,
+      "step": 1476
+    },
+    {
+      "epoch": 1.3625461254612545,
+      "grad_norm": 0.3986149445543998,
+      "learning_rate": 3.324430748143409e-05,
+      "loss": 0.1799,
+      "step": 1477
+    },
+    {
+      "epoch": 1.363468634686347,
+      "grad_norm": 0.3231908952222806,
+      "learning_rate": 3.321896222221475e-05,
+      "loss": 0.1502,
+      "step": 1478
+    },
+    {
+      "epoch": 1.3643911439114391,
+      "grad_norm": 0.3585100260807637,
+      "learning_rate": 3.3193607488239196e-05,
+      "loss": 0.1691,
+      "step": 1479
+    },
+    {
+      "epoch": 1.3653136531365313,
+      "grad_norm": 0.3724697410027662,
+      "learning_rate": 3.3168243308736174e-05,
+      "loss": 0.1649,
+      "step": 1480
+    },
+    {
+      "epoch": 1.3662361623616235,
+      "grad_norm": 0.43836765114750026,
+      "learning_rate": 3.3142869712945314e-05,
+      "loss": 0.1949,
+      "step": 1481
+    },
+    {
+      "epoch": 1.367158671586716,
+      "grad_norm": 0.38698908313357655,
+      "learning_rate": 3.311748673011709e-05,
+      "loss": 0.1785,
+      "step": 1482
+    },
+    {
+      "epoch": 1.368081180811808,
+      "grad_norm": 0.40085667062326424,
+      "learning_rate": 3.3092094389512815e-05,
+      "loss": 0.1706,
+      "step": 1483
+    },
+    {
+      "epoch": 1.3690036900369003,
+      "grad_norm": 0.4215350852247018,
+      "learning_rate": 3.306669272040459e-05,
+      "loss": 0.1605,
+      "step": 1484
+    },
+    {
+      "epoch": 1.3699261992619927,
+      "grad_norm": 0.38755558872265455,
+      "learning_rate": 3.304128175207526e-05,
+      "loss": 0.177,
+      "step": 1485
+    },
+    {
+      "epoch": 1.3708487084870848,
+      "grad_norm": 0.33245606825587154,
+      "learning_rate": 3.301586151381839e-05,
+      "loss": 0.1584,
+      "step": 1486
+    },
+    {
+      "epoch": 1.371771217712177,
+      "grad_norm": 0.3968505143633735,
+      "learning_rate": 3.2990432034938235e-05,
+      "loss": 0.1731,
+      "step": 1487
+    },
+    {
+      "epoch": 1.3726937269372694,
+      "grad_norm": 0.3545579887998813,
+      "learning_rate": 3.29649933447497e-05,
+      "loss": 0.1558,
+      "step": 1488
+    },
+    {
+      "epoch": 1.3736162361623616,
+      "grad_norm": 0.3538986320266718,
+      "learning_rate": 3.293954547257832e-05,
+      "loss": 0.1609,
+      "step": 1489
+    },
+    {
+      "epoch": 1.3745387453874538,
+      "grad_norm": 0.4036471913431026,
+      "learning_rate": 3.2914088447760194e-05,
+      "loss": 0.184,
+      "step": 1490
+    },
+    {
+      "epoch": 1.3754612546125462,
+      "grad_norm": 0.39631473739938977,
+      "learning_rate": 3.288862229964198e-05,
+      "loss": 0.1834,
+      "step": 1491
+    },
+    {
+      "epoch": 1.3763837638376384,
+      "grad_norm": 0.37679883371092154,
+      "learning_rate": 3.2863147057580875e-05,
+      "loss": 0.1723,
+      "step": 1492
+    },
+    {
+      "epoch": 1.3773062730627306,
+      "grad_norm": 0.33910583495721097,
+      "learning_rate": 3.2837662750944535e-05,
+      "loss": 0.1611,
+      "step": 1493
+    },
+    {
+      "epoch": 1.378228782287823,
+      "grad_norm": 0.3449252182028951,
+      "learning_rate": 3.281216940911106e-05,
+      "loss": 0.1555,
+      "step": 1494
+    },
+    {
+      "epoch": 1.3791512915129152,
+      "grad_norm": 0.40143260634074024,
+      "learning_rate": 3.2786667061469e-05,
+      "loss": 0.198,
+      "step": 1495
+    },
+    {
+      "epoch": 1.3800738007380073,
+      "grad_norm": 0.33285579323659764,
+      "learning_rate": 3.276115573741724e-05,
+      "loss": 0.1453,
+      "step": 1496
+    },
+    {
+      "epoch": 1.3809963099630997,
+      "grad_norm": 0.36713778474185194,
+      "learning_rate": 3.2735635466365046e-05,
+      "loss": 0.1747,
+      "step": 1497
+    },
+    {
+      "epoch": 1.381918819188192,
+      "grad_norm": 0.3583536746177072,
+      "learning_rate": 3.2710106277732e-05,
+      "loss": 0.1672,
+      "step": 1498
+    },
+    {
+      "epoch": 1.382841328413284,
+      "grad_norm": 0.3237529683832653,
+      "learning_rate": 3.268456820094794e-05,
+      "loss": 0.1521,
+      "step": 1499
+    },
+    {
+      "epoch": 1.3837638376383765,
+      "grad_norm": 0.3432367392597323,
+      "learning_rate": 3.2659021265452974e-05,
+      "loss": 0.1691,
+      "step": 1500
+    },
+    {
+      "epoch": 1.3846863468634687,
+      "grad_norm": 0.34058885505884795,
+      "learning_rate": 3.263346550069741e-05,
+      "loss": 0.1648,
+      "step": 1501
+    },
+    {
+      "epoch": 1.3856088560885609,
+      "grad_norm": 0.372009222581819,
+      "learning_rate": 3.2607900936141725e-05,
+      "loss": 0.1831,
+      "step": 1502
+    },
+    {
+      "epoch": 1.386531365313653,
+      "grad_norm": 0.34954399449943774,
+      "learning_rate": 3.258232760125657e-05,
+      "loss": 0.1474,
+      "step": 1503
+    },
+    {
+      "epoch": 1.3874538745387455,
+      "grad_norm": 0.41213354250385276,
+      "learning_rate": 3.255674552552267e-05,
+      "loss": 0.1805,
+      "step": 1504
+    },
+    {
+      "epoch": 1.3883763837638377,
+      "grad_norm": 0.3590923607406116,
+      "learning_rate": 3.253115473843086e-05,
+      "loss": 0.1636,
+      "step": 1505
+    },
+    {
+      "epoch": 1.3892988929889298,
+      "grad_norm": 0.5195077376564883,
+      "learning_rate": 3.2505555269481993e-05,
+      "loss": 0.1746,
+      "step": 1506
+    },
+    {
+      "epoch": 1.390221402214022,
+      "grad_norm": 0.35726683093834866,
+      "learning_rate": 3.247994714818694e-05,
+      "loss": 0.1728,
+      "step": 1507
+    },
+    {
+      "epoch": 1.3911439114391144,
+      "grad_norm": 0.3431429130379043,
+      "learning_rate": 3.2454330404066545e-05,
+      "loss": 0.1489,
+      "step": 1508
+    },
+    {
+      "epoch": 1.3920664206642066,
+      "grad_norm": 0.3404992287293656,
+      "learning_rate": 3.2428705066651603e-05,
+      "loss": 0.1455,
+      "step": 1509
+    },
+    {
+      "epoch": 1.3929889298892988,
+      "grad_norm": 0.4157171836412933,
+      "learning_rate": 3.240307116548279e-05,
+      "loss": 0.1697,
+      "step": 1510
+    },
+    {
+      "epoch": 1.3939114391143912,
+      "grad_norm": 0.3851045240063217,
+      "learning_rate": 3.2377428730110684e-05,
+      "loss": 0.1602,
+      "step": 1511
+    },
+    {
+      "epoch": 1.3948339483394834,
+      "grad_norm": 0.36563293785621676,
+      "learning_rate": 3.235177779009567e-05,
+      "loss": 0.1466,
+      "step": 1512
+    },
+    {
+      "epoch": 1.3957564575645756,
+      "grad_norm": 0.37769315758092786,
+      "learning_rate": 3.2326118375007965e-05,
+      "loss": 0.1444,
+      "step": 1513
+    },
+    {
+      "epoch": 1.396678966789668,
+      "grad_norm": 0.404148742775024,
+      "learning_rate": 3.230045051442754e-05,
+      "loss": 0.1512,
+      "step": 1514
+    },
+    {
+      "epoch": 1.3976014760147601,
+      "grad_norm": 0.40149150310085274,
+      "learning_rate": 3.227477423794412e-05,
+      "loss": 0.16,
+      "step": 1515
+    },
+    {
+      "epoch": 1.3985239852398523,
+      "grad_norm": 0.3652165868891584,
+      "learning_rate": 3.2249089575157095e-05,
+      "loss": 0.1621,
+      "step": 1516
+    },
+    {
+      "epoch": 1.3994464944649447,
+      "grad_norm": 0.38068047916450476,
+      "learning_rate": 3.222339655567556e-05,
+      "loss": 0.1552,
+      "step": 1517
+    },
+    {
+      "epoch": 1.400369003690037,
+      "grad_norm": 0.42498426136576395,
+      "learning_rate": 3.2197695209118236e-05,
+      "loss": 0.1744,
+      "step": 1518
+    },
+    {
+      "epoch": 1.401291512915129,
+      "grad_norm": 0.4101821448145483,
+      "learning_rate": 3.2171985565113415e-05,
+      "loss": 0.1681,
+      "step": 1519
+    },
+    {
+      "epoch": 1.4022140221402215,
+      "grad_norm": 0.4070341768264358,
+      "learning_rate": 3.2146267653299e-05,
+      "loss": 0.18,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4031365313653137,
+      "grad_norm": 0.3400012868189246,
+      "learning_rate": 3.212054150332239e-05,
+      "loss": 0.1678,
+      "step": 1521
+    },
+    {
+      "epoch": 1.4040590405904059,
+      "grad_norm": 0.433022903329026,
+      "learning_rate": 3.209480714484049e-05,
+      "loss": 0.1774,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4049815498154983,
+      "grad_norm": 0.31030208242511237,
+      "learning_rate": 3.206906460751968e-05,
+      "loss": 0.1567,
+      "step": 1523
+    },
+    {
+      "epoch": 1.4059040590405905,
+      "grad_norm": 0.34001679021223896,
+      "learning_rate": 3.2043313921035743e-05,
+      "loss": 0.159,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4068265682656826,
+      "grad_norm": 0.38322630552346637,
+      "learning_rate": 3.201755511507389e-05,
+      "loss": 0.1669,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4077490774907748,
+      "grad_norm": 0.3693576434894572,
+      "learning_rate": 3.199178821932865e-05,
+      "loss": 0.1604,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4086715867158672,
+      "grad_norm": 0.3995499629249807,
+      "learning_rate": 3.196601326350393e-05,
+      "loss": 0.1676,
+      "step": 1527
+    },
+    {
+      "epoch": 1.4095940959409594,
+      "grad_norm": 0.34824438998827817,
+      "learning_rate": 3.194023027731288e-05,
+      "loss": 0.1432,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4105166051660516,
+      "grad_norm": 0.36116120388148987,
+      "learning_rate": 3.191443929047793e-05,
+      "loss": 0.169,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4114391143911438,
+      "grad_norm": 0.3560442086436447,
+      "learning_rate": 3.188864033273074e-05,
+      "loss": 0.1731,
+      "step": 1530
+    },
+    {
+      "epoch": 1.4123616236162362,
+      "grad_norm": 0.3706163750507851,
+      "learning_rate": 3.186283343381213e-05,
+      "loss": 0.1388,
+      "step": 1531
+    },
+    {
+      "epoch": 1.4132841328413284,
+      "grad_norm": 0.38187197667421363,
+      "learning_rate": 3.1837018623472116e-05,
+      "loss": 0.1537,
+      "step": 1532
+    },
+    {
+      "epoch": 1.4142066420664205,
+      "grad_norm": 0.42580664179707683,
+      "learning_rate": 3.1811195931469804e-05,
+      "loss": 0.1768,
+      "step": 1533
+    },
+    {
+      "epoch": 1.415129151291513,
+      "grad_norm": 0.4084666432032316,
+      "learning_rate": 3.178536538757339e-05,
+      "loss": 0.1887,
+      "step": 1534
+    },
+    {
+      "epoch": 1.4160516605166051,
+      "grad_norm": 0.35942118774969967,
+      "learning_rate": 3.1759527021560126e-05,
+      "loss": 0.1804,
+      "step": 1535
+    },
+    {
+      "epoch": 1.4169741697416973,
+      "grad_norm": 0.3692777759704592,
+      "learning_rate": 3.173368086321629e-05,
+      "loss": 0.1849,
+      "step": 1536
+    },
+    {
+      "epoch": 1.4178966789667897,
+      "grad_norm": 0.3636443925361855,
+      "learning_rate": 3.170782694233712e-05,
+      "loss": 0.1743,
+      "step": 1537
+    },
+    {
+      "epoch": 1.418819188191882,
+      "grad_norm": 0.40942980389322786,
+      "learning_rate": 3.168196528872682e-05,
+      "loss": 0.1725,
+      "step": 1538
+    },
+    {
+      "epoch": 1.419741697416974,
+      "grad_norm": 0.33833653641281153,
+      "learning_rate": 3.165609593219852e-05,
+      "loss": 0.1743,
+      "step": 1539
+    },
+    {
+      "epoch": 1.4206642066420665,
+      "grad_norm": 0.34662617508546745,
+      "learning_rate": 3.16302189025742e-05,
+      "loss": 0.1665,
+      "step": 1540
+    },
+    {
+      "epoch": 1.4215867158671587,
+      "grad_norm": 0.3837517105262154,
+      "learning_rate": 3.1604334229684705e-05,
+      "loss": 0.1811,
+      "step": 1541
+    },
+    {
+      "epoch": 1.4225092250922509,
+      "grad_norm": 0.3660541015601058,
+      "learning_rate": 3.157844194336968e-05,
+      "loss": 0.1662,
+      "step": 1542
+    },
+    {
+      "epoch": 1.4234317343173433,
+      "grad_norm": 0.4041655451243474,
+      "learning_rate": 3.1552542073477555e-05,
+      "loss": 0.1661,
+      "step": 1543
+    },
+    {
+      "epoch": 1.4243542435424354,
+      "grad_norm": 0.38054326648318904,
+      "learning_rate": 3.1526634649865514e-05,
+      "loss": 0.1596,
+      "step": 1544
+    },
+    {
+      "epoch": 1.4252767527675276,
+      "grad_norm": 0.3849710778859087,
+      "learning_rate": 3.150071970239941e-05,
+      "loss": 0.1609,
+      "step": 1545
+    },
+    {
+      "epoch": 1.42619926199262,
+      "grad_norm": 0.41310874387916263,
+      "learning_rate": 3.1474797260953806e-05,
+      "loss": 0.1604,
+      "step": 1546
+    },
+    {
+      "epoch": 1.4271217712177122,
+      "grad_norm": 0.3824298243642993,
+      "learning_rate": 3.144886735541191e-05,
+      "loss": 0.1414,
+      "step": 1547
+    },
+    {
+      "epoch": 1.4280442804428044,
+      "grad_norm": 0.438248106970847,
+      "learning_rate": 3.1422930015665484e-05,
+      "loss": 0.1707,
+      "step": 1548
+    },
+    {
+      "epoch": 1.4289667896678968,
+      "grad_norm": 0.47825328997554417,
+      "learning_rate": 3.1396985271614914e-05,
+      "loss": 0.1933,
+      "step": 1549
+    },
+    {
+      "epoch": 1.429889298892989,
+      "grad_norm": 0.395617655243006,
+      "learning_rate": 3.13710331531691e-05,
+      "loss": 0.182,
+      "step": 1550
+    },
+    {
+      "epoch": 1.4308118081180812,
+      "grad_norm": 0.36187891165428526,
+      "learning_rate": 3.134507369024543e-05,
+      "loss": 0.1464,
+      "step": 1551
+    },
+    {
+      "epoch": 1.4317343173431734,
+      "grad_norm": 0.42839944714948797,
+      "learning_rate": 3.13191069127698e-05,
+      "loss": 0.1694,
+      "step": 1552
+    },
+    {
+      "epoch": 1.4326568265682658,
+      "grad_norm": 0.38965828762765414,
+      "learning_rate": 3.1293132850676484e-05,
+      "loss": 0.1714,
+      "step": 1553
+    },
+    {
+      "epoch": 1.433579335793358,
+      "grad_norm": 0.39447447361198096,
+      "learning_rate": 3.126715153390819e-05,
+      "loss": 0.187,
+      "step": 1554
+    },
+    {
+      "epoch": 1.4345018450184501,
+      "grad_norm": 0.357097085573246,
+      "learning_rate": 3.124116299241598e-05,
+      "loss": 0.1798,
+      "step": 1555
+    },
+    {
+      "epoch": 1.4354243542435423,
+      "grad_norm": 0.3825572959600671,
+      "learning_rate": 3.1215167256159245e-05,
+      "loss": 0.1611,
+      "step": 1556
+    },
+    {
+      "epoch": 1.4363468634686347,
+      "grad_norm": 0.3546728581900947,
+      "learning_rate": 3.118916435510567e-05,
+      "loss": 0.1813,
+      "step": 1557
+    },
+    {
+      "epoch": 1.437269372693727,
+      "grad_norm": 0.36315138657808316,
+      "learning_rate": 3.1163154319231194e-05,
+      "loss": 0.1532,
+      "step": 1558
+    },
+    {
+      "epoch": 1.438191881918819,
+      "grad_norm": 0.39610627077159327,
+      "learning_rate": 3.1137137178519985e-05,
+      "loss": 0.1717,
+      "step": 1559
+    },
+    {
+      "epoch": 1.4391143911439115,
+      "grad_norm": 0.4748758696309284,
+      "learning_rate": 3.111111296296441e-05,
+      "loss": 0.1557,
+      "step": 1560
+    },
+    {
+      "epoch": 1.4400369003690037,
+      "grad_norm": 0.3137228098987481,
+      "learning_rate": 3.1085081702564966e-05,
+      "loss": 0.1489,
+      "step": 1561
+    },
+    {
+      "epoch": 1.4409594095940959,
+      "grad_norm": 0.40963764410421305,
+      "learning_rate": 3.105904342733032e-05,
+      "loss": 0.1652,
+      "step": 1562
+    },
+    {
+      "epoch": 1.4418819188191883,
+      "grad_norm": 0.3559814030662488,
+      "learning_rate": 3.103299816727716e-05,
+      "loss": 0.17,
+      "step": 1563
+    },
+    {
+      "epoch": 1.4428044280442804,
+      "grad_norm": 0.4173167640148135,
+      "learning_rate": 3.100694595243028e-05,
+      "loss": 0.1809,
+      "step": 1564
+    },
+    {
+      "epoch": 1.4437269372693726,
+      "grad_norm": 0.3214124947242933,
+      "learning_rate": 3.0980886812822474e-05,
+      "loss": 0.1542,
+      "step": 1565
+    },
+    {
+      "epoch": 1.444649446494465,
+      "grad_norm": 0.399532478767107,
+      "learning_rate": 3.0954820778494516e-05,
+      "loss": 0.182,
+      "step": 1566
+    },
+    {
+      "epoch": 1.4455719557195572,
+      "grad_norm": 0.42170975126734245,
+      "learning_rate": 3.0928747879495115e-05,
+      "loss": 0.182,
+      "step": 1567
+    },
+    {
+      "epoch": 1.4464944649446494,
+      "grad_norm": 0.3615743704404705,
+      "learning_rate": 3.0902668145880924e-05,
+      "loss": 0.1593,
+      "step": 1568
+    },
+    {
+      "epoch": 1.4474169741697418,
+      "grad_norm": 0.35225646041206127,
+      "learning_rate": 3.0876581607716456e-05,
+      "loss": 0.1579,
+      "step": 1569
+    },
+    {
+      "epoch": 1.448339483394834,
+      "grad_norm": 0.39706475767264293,
+      "learning_rate": 3.085048829507406e-05,
+      "loss": 0.1691,
+      "step": 1570
+    },
+    {
+      "epoch": 1.4492619926199262,
+      "grad_norm": 0.4822514269767401,
+      "learning_rate": 3.082438823803392e-05,
+      "loss": 0.1897,
+      "step": 1571
+    },
+    {
+      "epoch": 1.4501845018450186,
+      "grad_norm": 0.3723450625164562,
+      "learning_rate": 3.079828146668397e-05,
+      "loss": 0.1832,
+      "step": 1572
+    },
+    {
+      "epoch": 1.4511070110701108,
+      "grad_norm": 0.4106009436359365,
+      "learning_rate": 3.07721680111199e-05,
+      "loss": 0.1728,
+      "step": 1573
+    },
+    {
+      "epoch": 1.452029520295203,
+      "grad_norm": 0.34347681763677174,
+      "learning_rate": 3.07460479014451e-05,
+      "loss": 0.1614,
+      "step": 1574
+    },
+    {
+      "epoch": 1.4529520295202953,
+      "grad_norm": 0.381661507373486,
+      "learning_rate": 3.0719921167770624e-05,
+      "loss": 0.1815,
+      "step": 1575
+    },
+    {
+      "epoch": 1.4538745387453875,
+      "grad_norm": 0.44435761234773025,
+      "learning_rate": 3.069378784021518e-05,
+      "loss": 0.1875,
+      "step": 1576
+    },
+    {
+      "epoch": 1.4547970479704797,
+      "grad_norm": 0.3624200072470951,
+      "learning_rate": 3.066764794890505e-05,
+      "loss": 0.1644,
+      "step": 1577
+    },
+    {
+      "epoch": 1.455719557195572,
+      "grad_norm": 0.36808113934931824,
+      "learning_rate": 3.064150152397412e-05,
+      "loss": 0.1501,
+      "step": 1578
+    },
+    {
+      "epoch": 1.456642066420664,
+      "grad_norm": 0.3666267965852813,
+      "learning_rate": 3.061534859556377e-05,
+      "loss": 0.1796,
+      "step": 1579
+    },
+    {
+      "epoch": 1.4575645756457565,
+      "grad_norm": 0.392902157755771,
+      "learning_rate": 3.0589189193822895e-05,
+      "loss": 0.19,
+      "step": 1580
+    },
+    {
+      "epoch": 1.4584870848708487,
+      "grad_norm": 0.3913033345885448,
+      "learning_rate": 3.056302334890786e-05,
+      "loss": 0.1794,
+      "step": 1581
+    },
+    {
+      "epoch": 1.4594095940959408,
+      "grad_norm": 0.3496689337140897,
+      "learning_rate": 3.053685109098245e-05,
+      "loss": 0.16,
+      "step": 1582
+    },
+    {
+      "epoch": 1.4603321033210332,
+      "grad_norm": 0.39419972511865325,
+      "learning_rate": 3.051067245021783e-05,
+      "loss": 0.1546,
+      "step": 1583
+    },
+    {
+      "epoch": 1.4612546125461254,
+      "grad_norm": 0.41736336497176446,
+      "learning_rate": 3.048448745679255e-05,
+      "loss": 0.1972,
+      "step": 1584
+    },
+    {
+      "epoch": 1.4621771217712176,
+      "grad_norm": 0.35073426253376633,
+      "learning_rate": 3.045829614089246e-05,
+      "loss": 0.1669,
+      "step": 1585
+    },
+    {
+      "epoch": 1.46309963099631,
+      "grad_norm": 0.4021785887416188,
+      "learning_rate": 3.04320985327107e-05,
+      "loss": 0.1731,
+      "step": 1586
+    },
+    {
+      "epoch": 1.4640221402214022,
+      "grad_norm": 0.40666184044352294,
+      "learning_rate": 3.040589466244768e-05,
+      "loss": 0.1901,
+      "step": 1587
+    },
+    {
+      "epoch": 1.4649446494464944,
+      "grad_norm": 0.3696675854986723,
+      "learning_rate": 3.0379684560311027e-05,
+      "loss": 0.1521,
+      "step": 1588
+    },
+    {
+      "epoch": 1.4658671586715868,
+      "grad_norm": 0.38100821412378033,
+      "learning_rate": 3.035346825651552e-05,
+      "loss": 0.1482,
+      "step": 1589
+    },
+    {
+      "epoch": 1.466789667896679,
+      "grad_norm": 0.3983505131281807,
+      "learning_rate": 3.0327245781283136e-05,
+      "loss": 0.1712,
+      "step": 1590
+    },
+    {
+      "epoch": 1.4677121771217712,
+      "grad_norm": 0.41094545194443866,
+      "learning_rate": 3.0301017164842932e-05,
+      "loss": 0.1786,
+      "step": 1591
+    },
+    {
+      "epoch": 1.4686346863468636,
+      "grad_norm": 0.3687323192999056,
+      "learning_rate": 3.027478243743106e-05,
+      "loss": 0.1708,
+      "step": 1592
+    },
+    {
+      "epoch": 1.4695571955719557,
+      "grad_norm": 0.343827240699272,
+      "learning_rate": 3.0248541629290693e-05,
+      "loss": 0.1628,
+      "step": 1593
+    },
+    {
+      "epoch": 1.470479704797048,
+      "grad_norm": 0.3260547541598996,
+      "learning_rate": 3.0222294770672053e-05,
+      "loss": 0.1657,
+      "step": 1594
+    },
+    {
+      "epoch": 1.4714022140221403,
+      "grad_norm": 0.35540004378286827,
+      "learning_rate": 3.0196041891832312e-05,
+      "loss": 0.1344,
+      "step": 1595
+    },
+    {
+      "epoch": 1.4723247232472325,
+      "grad_norm": 0.3527802448027834,
+      "learning_rate": 3.0169783023035577e-05,
+      "loss": 0.1693,
+      "step": 1596
+    },
+    {
+      "epoch": 1.4732472324723247,
+      "grad_norm": 0.38113471097611723,
+      "learning_rate": 3.0143518194552873e-05,
+      "loss": 0.1703,
+      "step": 1597
+    },
+    {
+      "epoch": 1.474169741697417,
+      "grad_norm": 0.35839469193166723,
+      "learning_rate": 3.01172474366621e-05,
+      "loss": 0.1674,
+      "step": 1598
+    },
+    {
+      "epoch": 1.4750922509225093,
+      "grad_norm": 0.37083937938017913,
+      "learning_rate": 3.009097077964797e-05,
+      "loss": 0.1568,
+      "step": 1599
+    },
+    {
+      "epoch": 1.4760147601476015,
+      "grad_norm": 0.39238812133174916,
+      "learning_rate": 3.0064688253802026e-05,
+      "loss": 0.1711,
+      "step": 1600
+    },
+    {
+      "epoch": 1.4769372693726937,
+      "grad_norm": 0.3355525417090908,
+      "learning_rate": 3.0038399889422553e-05,
+      "loss": 0.1507,
+      "step": 1601
+    },
+    {
+      "epoch": 1.477859778597786,
+      "grad_norm": 0.36063956430414745,
+      "learning_rate": 3.001210571681457e-05,
+      "loss": 0.1492,
+      "step": 1602
+    },
+    {
+      "epoch": 1.4787822878228782,
+      "grad_norm": 0.3710232857863549,
+      "learning_rate": 2.9985805766289817e-05,
+      "loss": 0.1685,
+      "step": 1603
+    },
+    {
+      "epoch": 1.4797047970479704,
+      "grad_norm": 0.42301545914956,
+      "learning_rate": 2.995950006816664e-05,
+      "loss": 0.1862,
+      "step": 1604
+    },
+    {
+      "epoch": 1.4806273062730626,
+      "grad_norm": 0.43511270029205984,
+      "learning_rate": 2.9933188652770068e-05,
+      "loss": 0.1947,
+      "step": 1605
+    },
+    {
+      "epoch": 1.481549815498155,
+      "grad_norm": 0.38886533606243734,
+      "learning_rate": 2.9906871550431697e-05,
+      "loss": 0.1481,
+      "step": 1606
+    },
+    {
+      "epoch": 1.4824723247232472,
+      "grad_norm": 0.329252996048036,
+      "learning_rate": 2.988054879148967e-05,
+      "loss": 0.1656,
+      "step": 1607
+    },
+    {
+      "epoch": 1.4833948339483394,
+      "grad_norm": 0.3395919414815041,
+      "learning_rate": 2.9854220406288668e-05,
+      "loss": 0.1695,
+      "step": 1608
+    },
+    {
+      "epoch": 1.4843173431734318,
+      "grad_norm": 0.3409522992599561,
+      "learning_rate": 2.9827886425179848e-05,
+      "loss": 0.1649,
+      "step": 1609
+    },
+    {
+      "epoch": 1.485239852398524,
+      "grad_norm": 0.33811409950873744,
+      "learning_rate": 2.980154687852082e-05,
+      "loss": 0.1569,
+      "step": 1610
+    },
+    {
+      "epoch": 1.4861623616236161,
+      "grad_norm": 0.3673414750977099,
+      "learning_rate": 2.977520179667561e-05,
+      "loss": 0.1749,
+      "step": 1611
+    },
+    {
+      "epoch": 1.4870848708487086,
+      "grad_norm": 0.40741286834007223,
+      "learning_rate": 2.9748851210014628e-05,
+      "loss": 0.1768,
+      "step": 1612
+    },
+    {
+      "epoch": 1.4880073800738007,
+      "grad_norm": 0.33719593043968854,
+      "learning_rate": 2.972249514891462e-05,
+      "loss": 0.1698,
+      "step": 1613
+    },
+    {
+      "epoch": 1.488929889298893,
+      "grad_norm": 0.4098015224299873,
+      "learning_rate": 2.9696133643758662e-05,
+      "loss": 0.1781,
+      "step": 1614
+    },
+    {
+      "epoch": 1.4898523985239853,
+      "grad_norm": 0.3378128069918369,
+      "learning_rate": 2.966976672493607e-05,
+      "loss": 0.1642,
+      "step": 1615
+    },
+    {
+      "epoch": 1.4907749077490775,
+      "grad_norm": 0.3317008949395516,
+      "learning_rate": 2.964339442284245e-05,
+      "loss": 0.1531,
+      "step": 1616
+    },
+    {
+      "epoch": 1.4916974169741697,
+      "grad_norm": 0.31973761938103357,
+      "learning_rate": 2.961701676787958e-05,
+      "loss": 0.1463,
+      "step": 1617
+    },
+    {
+      "epoch": 1.492619926199262,
+      "grad_norm": 0.40153379072024,
+      "learning_rate": 2.9590633790455413e-05,
+      "loss": 0.1799,
+      "step": 1618
+    },
+    {
+      "epoch": 1.4935424354243543,
+      "grad_norm": 0.37354165237734216,
+      "learning_rate": 2.9564245520984047e-05,
+      "loss": 0.1633,
+      "step": 1619
+    },
+    {
+      "epoch": 1.4944649446494465,
+      "grad_norm": 0.3082489787747408,
+      "learning_rate": 2.9537851989885667e-05,
+      "loss": 0.1449,
+      "step": 1620
+    },
+    {
+      "epoch": 1.4953874538745389,
+      "grad_norm": 0.3370920265749821,
+      "learning_rate": 2.951145322758654e-05,
+      "loss": 0.1447,
+      "step": 1621
+    },
+    {
+      "epoch": 1.496309963099631,
+      "grad_norm": 0.38639008934505387,
+      "learning_rate": 2.948504926451896e-05,
+      "loss": 0.1556,
+      "step": 1622
+    },
+    {
+      "epoch": 1.4972324723247232,
+      "grad_norm": 0.39041551437746863,
+      "learning_rate": 2.945864013112119e-05,
+      "loss": 0.1731,
+      "step": 1623
+    },
+    {
+      "epoch": 1.4981549815498156,
+      "grad_norm": 0.38232994861502023,
+      "learning_rate": 2.943222585783749e-05,
+      "loss": 0.1621,
+      "step": 1624
+    },
+    {
+      "epoch": 1.4990774907749078,
+      "grad_norm": 0.38654601080386863,
+      "learning_rate": 2.9405806475118048e-05,
+      "loss": 0.142,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.33722181119539885,
+      "learning_rate": 2.9379382013418892e-05,
+      "loss": 0.1591,
+      "step": 1626
+    },
+    {
+      "epoch": 1.5009225092250924,
+      "grad_norm": 0.4095673988908001,
+      "learning_rate": 2.935295250320196e-05,
+      "loss": 0.1754,
+      "step": 1627
+    },
+    {
+      "epoch": 1.5018450184501844,
+      "grad_norm": 0.4197188700101124,
+      "learning_rate": 2.932651797493498e-05,
+      "loss": 0.1686,
+      "step": 1628
+    },
+    {
+      "epoch": 1.5027675276752768,
+      "grad_norm": 0.3595654552689153,
+      "learning_rate": 2.9300078459091462e-05,
+      "loss": 0.1708,
+      "step": 1629
+    },
+    {
+      "epoch": 1.503690036900369,
+      "grad_norm": 0.36307084098081793,
+      "learning_rate": 2.9273633986150696e-05,
+      "loss": 0.1656,
+      "step": 1630
+    },
+    {
+      "epoch": 1.5046125461254611,
+      "grad_norm": 0.383936016864393,
+      "learning_rate": 2.9247184586597648e-05,
+      "loss": 0.1669,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5055350553505535,
+      "grad_norm": 0.4118452867071001,
+      "learning_rate": 2.922073029092299e-05,
+      "loss": 0.1715,
+      "step": 1632
+    },
+    {
+      "epoch": 1.5064575645756457,
+      "grad_norm": 0.3875727639529233,
+      "learning_rate": 2.9194271129623034e-05,
+      "loss": 0.1759,
+      "step": 1633
+    },
+    {
+      "epoch": 1.507380073800738,
+      "grad_norm": 0.31670060362800706,
+      "learning_rate": 2.9167807133199686e-05,
+      "loss": 0.1561,
+      "step": 1634
+    },
+    {
+      "epoch": 1.5083025830258303,
+      "grad_norm": 0.4349756602309646,
+      "learning_rate": 2.914133833216045e-05,
+      "loss": 0.1745,
+      "step": 1635
+    },
+    {
+      "epoch": 1.5092250922509225,
+      "grad_norm": 0.36721395758123293,
+      "learning_rate": 2.9114864757018352e-05,
+      "loss": 0.1707,
+      "step": 1636
+    },
+    {
+      "epoch": 1.5101476014760147,
+      "grad_norm": 0.37569955946034134,
+      "learning_rate": 2.908838643829191e-05,
+      "loss": 0.1665,
+      "step": 1637
+    },
+    {
+      "epoch": 1.511070110701107,
+      "grad_norm": 0.3515764064184136,
+      "learning_rate": 2.9061903406505154e-05,
+      "loss": 0.1652,
+      "step": 1638
+    },
+    {
+      "epoch": 1.5119926199261993,
+      "grad_norm": 0.4000005503389113,
+      "learning_rate": 2.9035415692187485e-05,
+      "loss": 0.1642,
+      "step": 1639
+    },
+    {
+      "epoch": 1.5129151291512914,
+      "grad_norm": 0.3495958649285323,
+      "learning_rate": 2.9008923325873753e-05,
+      "loss": 0.1629,
+      "step": 1640
+    },
+    {
+      "epoch": 1.5138376383763839,
+      "grad_norm": 0.4242767363712676,
+      "learning_rate": 2.8982426338104168e-05,
+      "loss": 0.1856,
+      "step": 1641
+    },
+    {
+      "epoch": 1.514760147601476,
+      "grad_norm": 0.39603614592949704,
+      "learning_rate": 2.8955924759424225e-05,
+      "loss": 0.1546,
+      "step": 1642
+    },
+    {
+      "epoch": 1.5156826568265682,
+      "grad_norm": 0.3321707220729133,
+      "learning_rate": 2.8929418620384753e-05,
+      "loss": 0.1562,
+      "step": 1643
+    },
+    {
+      "epoch": 1.5166051660516606,
+      "grad_norm": 0.39501612838057204,
+      "learning_rate": 2.8902907951541834e-05,
+      "loss": 0.1787,
+      "step": 1644
+    },
+    {
+      "epoch": 1.5175276752767528,
+      "grad_norm": 0.4115223591573108,
+      "learning_rate": 2.887639278345674e-05,
+      "loss": 0.159,
+      "step": 1645
+    },
+    {
+      "epoch": 1.518450184501845,
+      "grad_norm": 0.3952771894733965,
+      "learning_rate": 2.8849873146695972e-05,
+      "loss": 0.1739,
+      "step": 1646
+    },
+    {
+      "epoch": 1.5193726937269374,
+      "grad_norm": 0.35382488689126607,
+      "learning_rate": 2.882334907183115e-05,
+      "loss": 0.1742,
+      "step": 1647
+    },
+    {
+      "epoch": 1.5202952029520294,
+      "grad_norm": 0.35980398974927635,
+      "learning_rate": 2.8796820589439027e-05,
+      "loss": 0.1663,
+      "step": 1648
+    },
+    {
+      "epoch": 1.5212177121771218,
+      "grad_norm": 0.351634834990234,
+      "learning_rate": 2.877028773010144e-05,
+      "loss": 0.163,
+      "step": 1649
+    },
+    {
+      "epoch": 1.5221402214022142,
+      "grad_norm": 0.36216094534186977,
+      "learning_rate": 2.8743750524405254e-05,
+      "loss": 0.1537,
+      "step": 1650
+    },
+    {
+      "epoch": 1.5230627306273061,
+      "grad_norm": 0.3820723737397306,
+      "learning_rate": 2.8717209002942357e-05,
+      "loss": 0.1797,
+      "step": 1651
+    },
+    {
+      "epoch": 1.5239852398523985,
+      "grad_norm": 0.36131648309341774,
+      "learning_rate": 2.8690663196309615e-05,
+      "loss": 0.1567,
+      "step": 1652
+    },
+    {
+      "epoch": 1.524907749077491,
+      "grad_norm": 0.4146918598402644,
+      "learning_rate": 2.866411313510882e-05,
+      "loss": 0.1655,
+      "step": 1653
+    },
+    {
+      "epoch": 1.525830258302583,
+      "grad_norm": 0.3872220874047178,
+      "learning_rate": 2.863755884994669e-05,
+      "loss": 0.1684,
+      "step": 1654
+    },
+    {
+      "epoch": 1.5267527675276753,
+      "grad_norm": 0.3488782618904229,
+      "learning_rate": 2.8611000371434794e-05,
+      "loss": 0.1621,
+      "step": 1655
+    },
+    {
+      "epoch": 1.5276752767527675,
+      "grad_norm": 0.3513312290077386,
+      "learning_rate": 2.8584437730189534e-05,
+      "loss": 0.148,
+      "step": 1656
+    },
+    {
+      "epoch": 1.5285977859778597,
+      "grad_norm": 0.4150137776028713,
+      "learning_rate": 2.8557870956832132e-05,
+      "loss": 0.1728,
+      "step": 1657
+    },
+    {
+      "epoch": 1.529520295202952,
+      "grad_norm": 0.4118880306886742,
+      "learning_rate": 2.853130008198855e-05,
+      "loss": 0.1669,
+      "step": 1658
+    },
+    {
+      "epoch": 1.5304428044280443,
+      "grad_norm": 0.39900580456854273,
+      "learning_rate": 2.850472513628948e-05,
+      "loss": 0.1645,
+      "step": 1659
+    },
+    {
+      "epoch": 1.5313653136531364,
+      "grad_norm": 0.34815868384829984,
+      "learning_rate": 2.8478146150370337e-05,
+      "loss": 0.1551,
+      "step": 1660
+    },
+    {
+      "epoch": 1.5322878228782288,
+      "grad_norm": 0.387257729505376,
+      "learning_rate": 2.8451563154871148e-05,
+      "loss": 0.1814,
+      "step": 1661
+    },
+    {
+      "epoch": 1.533210332103321,
+      "grad_norm": 0.3766488859307147,
+      "learning_rate": 2.8424976180436596e-05,
+      "loss": 0.1739,
+      "step": 1662
+    },
+    {
+      "epoch": 1.5341328413284132,
+      "grad_norm": 0.3858891394420731,
+      "learning_rate": 2.8398385257715942e-05,
+      "loss": 0.178,
+      "step": 1663
+    },
+    {
+      "epoch": 1.5350553505535056,
+      "grad_norm": 0.3623152374526288,
+      "learning_rate": 2.8371790417362987e-05,
+      "loss": 0.1736,
+      "step": 1664
+    },
+    {
+      "epoch": 1.5359778597785978,
+      "grad_norm": 0.3314968152447182,
+      "learning_rate": 2.8345191690036064e-05,
+      "loss": 0.1512,
+      "step": 1665
+    },
+    {
+      "epoch": 1.53690036900369,
+      "grad_norm": 0.42404689265015627,
+      "learning_rate": 2.8318589106397987e-05,
+      "loss": 0.1886,
+      "step": 1666
+    },
+    {
+      "epoch": 1.5378228782287824,
+      "grad_norm": 0.36779101131292224,
+      "learning_rate": 2.8291982697115986e-05,
+      "loss": 0.171,
+      "step": 1667
+    },
+    {
+      "epoch": 1.5387453874538746,
+      "grad_norm": 0.4139048273758788,
+      "learning_rate": 2.826537249286176e-05,
+      "loss": 0.1883,
+      "step": 1668
+    },
+    {
+      "epoch": 1.5396678966789668,
+      "grad_norm": 0.388458597327605,
+      "learning_rate": 2.8238758524311314e-05,
+      "loss": 0.1693,
+      "step": 1669
+    },
+    {
+      "epoch": 1.5405904059040592,
+      "grad_norm": 0.37003305542652387,
+      "learning_rate": 2.821214082214504e-05,
+      "loss": 0.1662,
+      "step": 1670
+    },
+    {
+      "epoch": 1.5415129151291513,
+      "grad_norm": 0.41708058167998435,
+      "learning_rate": 2.8185519417047624e-05,
+      "loss": 0.181,
+      "step": 1671
+    },
+    {
+      "epoch": 1.5424354243542435,
+      "grad_norm": 0.3875713861998777,
+      "learning_rate": 2.8158894339708004e-05,
+      "loss": 0.1732,
+      "step": 1672
+    },
+    {
+      "epoch": 1.543357933579336,
+      "grad_norm": 0.3953949690930639,
+      "learning_rate": 2.813226562081938e-05,
+      "loss": 0.176,
+      "step": 1673
+    },
+    {
+      "epoch": 1.5442804428044279,
+      "grad_norm": 0.33502194674439734,
+      "learning_rate": 2.8105633291079116e-05,
+      "loss": 0.1546,
+      "step": 1674
+    },
+    {
+      "epoch": 1.5452029520295203,
+      "grad_norm": 0.3262775436675679,
+      "learning_rate": 2.807899738118876e-05,
+      "loss": 0.1438,
+      "step": 1675
+    },
+    {
+      "epoch": 1.5461254612546127,
+      "grad_norm": 0.3862535028697546,
+      "learning_rate": 2.8052357921854e-05,
+      "loss": 0.1779,
+      "step": 1676
+    },
+    {
+      "epoch": 1.5470479704797047,
+      "grad_norm": 0.36875700215567225,
+      "learning_rate": 2.802571494378458e-05,
+      "loss": 0.159,
+      "step": 1677
+    },
+    {
+      "epoch": 1.547970479704797,
+      "grad_norm": 0.36609032887191695,
+      "learning_rate": 2.799906847769433e-05,
+      "loss": 0.1624,
+      "step": 1678
+    },
+    {
+      "epoch": 1.5488929889298892,
+      "grad_norm": 0.3719412064017298,
+      "learning_rate": 2.7972418554301084e-05,
+      "loss": 0.156,
+      "step": 1679
+    },
+    {
+      "epoch": 1.5498154981549814,
+      "grad_norm": 0.3968781036045197,
+      "learning_rate": 2.794576520432666e-05,
+      "loss": 0.167,
+      "step": 1680
+    },
+    {
+      "epoch": 1.5507380073800738,
+      "grad_norm": 0.4222511906972072,
+      "learning_rate": 2.791910845849686e-05,
+      "loss": 0.1975,
+      "step": 1681
+    },
+    {
+      "epoch": 1.551660516605166,
+      "grad_norm": 0.38799129053194403,
+      "learning_rate": 2.7892448347541354e-05,
+      "loss": 0.1696,
+      "step": 1682
+    },
+    {
+      "epoch": 1.5525830258302582,
+      "grad_norm": 0.3317047677083327,
+      "learning_rate": 2.7865784902193714e-05,
+      "loss": 0.1732,
+      "step": 1683
+    },
+    {
+      "epoch": 1.5535055350553506,
+      "grad_norm": 0.3343018202348186,
+      "learning_rate": 2.7839118153191362e-05,
+      "loss": 0.1633,
+      "step": 1684
+    },
+    {
+      "epoch": 1.5544280442804428,
+      "grad_norm": 0.3947024735698901,
+      "learning_rate": 2.781244813127552e-05,
+      "loss": 0.1719,
+      "step": 1685
+    },
+    {
+      "epoch": 1.555350553505535,
+      "grad_norm": 0.35244868140225677,
+      "learning_rate": 2.7785774867191172e-05,
+      "loss": 0.1691,
+      "step": 1686
+    },
+    {
+      "epoch": 1.5562730627306274,
+      "grad_norm": 0.3439655799420121,
+      "learning_rate": 2.775909839168706e-05,
+      "loss": 0.1827,
+      "step": 1687
+    },
+    {
+      "epoch": 1.5571955719557196,
+      "grad_norm": 0.3660175861653934,
+      "learning_rate": 2.7732418735515627e-05,
+      "loss": 0.1504,
+      "step": 1688
+    },
+    {
+      "epoch": 1.5581180811808117,
+      "grad_norm": 0.35395738515545716,
+      "learning_rate": 2.7705735929432953e-05,
+      "loss": 0.1573,
+      "step": 1689
+    },
+    {
+      "epoch": 1.5590405904059041,
+      "grad_norm": 0.34896194637545036,
+      "learning_rate": 2.7679050004198787e-05,
+      "loss": 0.1609,
+      "step": 1690
+    },
+    {
+      "epoch": 1.5599630996309963,
+      "grad_norm": 0.362312918114885,
+      "learning_rate": 2.7652360990576453e-05,
+      "loss": 0.1675,
+      "step": 1691
+    },
+    {
+      "epoch": 1.5608856088560885,
+      "grad_norm": 0.3919764217015699,
+      "learning_rate": 2.762566891933285e-05,
+      "loss": 0.1785,
+      "step": 1692
+    },
+    {
+      "epoch": 1.561808118081181,
+      "grad_norm": 0.39902352613928777,
+      "learning_rate": 2.7598973821238365e-05,
+      "loss": 0.1535,
+      "step": 1693
+    },
+    {
+      "epoch": 1.562730627306273,
+      "grad_norm": 0.33741091519673283,
+      "learning_rate": 2.7572275727066927e-05,
+      "loss": 0.1507,
+      "step": 1694
+    },
+    {
+      "epoch": 1.5636531365313653,
+      "grad_norm": 0.3704980091260743,
+      "learning_rate": 2.754557466759589e-05,
+      "loss": 0.1626,
+      "step": 1695
+    },
+    {
+      "epoch": 1.5645756457564577,
+      "grad_norm": 0.3923587627762875,
+      "learning_rate": 2.751887067360601e-05,
+      "loss": 0.1694,
+      "step": 1696
+    },
+    {
+      "epoch": 1.5654981549815496,
+      "grad_norm": 0.3725058521166386,
+      "learning_rate": 2.7492163775881475e-05,
+      "loss": 0.1638,
+      "step": 1697
+    },
+    {
+      "epoch": 1.566420664206642,
+      "grad_norm": 0.3538882141932905,
+      "learning_rate": 2.746545400520977e-05,
+      "loss": 0.1562,
+      "step": 1698
+    },
+    {
+      "epoch": 1.5673431734317345,
+      "grad_norm": 0.3582084206694265,
+      "learning_rate": 2.7438741392381705e-05,
+      "loss": 0.1666,
+      "step": 1699
+    },
+    {
+      "epoch": 1.5682656826568264,
+      "grad_norm": 0.3325709438632519,
+      "learning_rate": 2.74120259681914e-05,
+      "loss": 0.1598,
+      "step": 1700
+    },
+    {
+      "epoch": 1.5691881918819188,
+      "grad_norm": 0.3600940625464994,
+      "learning_rate": 2.7385307763436168e-05,
+      "loss": 0.1576,
+      "step": 1701
+    },
+    {
+      "epoch": 1.5701107011070112,
+      "grad_norm": 0.3303856865581989,
+      "learning_rate": 2.7358586808916557e-05,
+      "loss": 0.15,
+      "step": 1702
+    },
+    {
+      "epoch": 1.5710332103321032,
+      "grad_norm": 0.36761502472329155,
+      "learning_rate": 2.733186313543628e-05,
+      "loss": 0.1616,
+      "step": 1703
+    },
+    {
+      "epoch": 1.5719557195571956,
+      "grad_norm": 0.3488967311542213,
+      "learning_rate": 2.730513677380218e-05,
+      "loss": 0.1602,
+      "step": 1704
+    },
+    {
+      "epoch": 1.5728782287822878,
+      "grad_norm": 0.3744373709817195,
+      "learning_rate": 2.7278407754824194e-05,
+      "loss": 0.1653,
+      "step": 1705
+    },
+    {
+      "epoch": 1.57380073800738,
+      "grad_norm": 0.3426693420939731,
+      "learning_rate": 2.7251676109315338e-05,
+      "loss": 0.1576,
+      "step": 1706
+    },
+    {
+      "epoch": 1.5747232472324724,
+      "grad_norm": 0.3504742590283337,
+      "learning_rate": 2.7224941868091643e-05,
+      "loss": 0.1523,
+      "step": 1707
+    },
+    {
+      "epoch": 1.5756457564575646,
+      "grad_norm": 0.30178230668808714,
+      "learning_rate": 2.7198205061972132e-05,
+      "loss": 0.1482,
+      "step": 1708
+    },
+    {
+      "epoch": 1.5765682656826567,
+      "grad_norm": 0.42439136751990975,
+      "learning_rate": 2.7171465721778787e-05,
+      "loss": 0.174,
+      "step": 1709
+    },
+    {
+      "epoch": 1.5774907749077491,
+      "grad_norm": 0.37755044976708263,
+      "learning_rate": 2.7144723878336524e-05,
+      "loss": 0.1531,
+      "step": 1710
+    },
+    {
+      "epoch": 1.5784132841328413,
+      "grad_norm": 0.3686743204129699,
+      "learning_rate": 2.711797956247313e-05,
+      "loss": 0.1628,
+      "step": 1711
+    },
+    {
+      "epoch": 1.5793357933579335,
+      "grad_norm": 0.4007196085048192,
+      "learning_rate": 2.7091232805019235e-05,
+      "loss": 0.172,
+      "step": 1712
+    },
+    {
+      "epoch": 1.580258302583026,
+      "grad_norm": 0.36940706989551075,
+      "learning_rate": 2.7064483636808313e-05,
+      "loss": 0.1503,
+      "step": 1713
+    },
+    {
+      "epoch": 1.581180811808118,
+      "grad_norm": 0.3717310421409479,
+      "learning_rate": 2.7037732088676582e-05,
+      "loss": 0.1609,
+      "step": 1714
+    },
+    {
+      "epoch": 1.5821033210332103,
+      "grad_norm": 0.3728801357325129,
+      "learning_rate": 2.7010978191463025e-05,
+      "loss": 0.1649,
+      "step": 1715
+    },
+    {
+      "epoch": 1.5830258302583027,
+      "grad_norm": 0.3752575101090936,
+      "learning_rate": 2.698422197600934e-05,
+      "loss": 0.1821,
+      "step": 1716
+    },
+    {
+      "epoch": 1.5839483394833949,
+      "grad_norm": 0.4099370798173003,
+      "learning_rate": 2.695746347315987e-05,
+      "loss": 0.1767,
+      "step": 1717
+    },
+    {
+      "epoch": 1.584870848708487,
+      "grad_norm": 0.3803739879623954,
+      "learning_rate": 2.6930702713761612e-05,
+      "loss": 0.1739,
+      "step": 1718
+    },
+    {
+      "epoch": 1.5857933579335795,
+      "grad_norm": 0.4213106506726271,
+      "learning_rate": 2.6903939728664174e-05,
+      "loss": 0.1786,
+      "step": 1719
+    },
+    {
+      "epoch": 1.5867158671586716,
+      "grad_norm": 0.3525768768348475,
+      "learning_rate": 2.6877174548719706e-05,
+      "loss": 0.1598,
+      "step": 1720
+    },
+    {
+      "epoch": 1.5876383763837638,
+      "grad_norm": 0.3965576248495974,
+      "learning_rate": 2.6850407204782912e-05,
+      "loss": 0.1742,
+      "step": 1721
+    },
+    {
+      "epoch": 1.5885608856088562,
+      "grad_norm": 0.32956801808293157,
+      "learning_rate": 2.6823637727710972e-05,
+      "loss": 0.1397,
+      "step": 1722
+    },
+    {
+      "epoch": 1.5894833948339482,
+      "grad_norm": 0.4421425101053271,
+      "learning_rate": 2.6796866148363538e-05,
+      "loss": 0.1559,
+      "step": 1723
+    },
+    {
+      "epoch": 1.5904059040590406,
+      "grad_norm": 0.39009858519793944,
+      "learning_rate": 2.677009249760268e-05,
+      "loss": 0.1708,
+      "step": 1724
+    },
+    {
+      "epoch": 1.591328413284133,
+      "grad_norm": 0.40742855451442184,
+      "learning_rate": 2.674331680629284e-05,
+      "loss": 0.1715,
+      "step": 1725
+    },
+    {
+      "epoch": 1.592250922509225,
+      "grad_norm": 0.3891052404705072,
+      "learning_rate": 2.6716539105300853e-05,
+      "loss": 0.1639,
+      "step": 1726
+    },
+    {
+      "epoch": 1.5931734317343174,
+      "grad_norm": 0.31948820814791407,
+      "learning_rate": 2.668975942549583e-05,
+      "loss": 0.151,
+      "step": 1727
+    },
+    {
+      "epoch": 1.5940959409594095,
+      "grad_norm": 0.3722932165709747,
+      "learning_rate": 2.6662977797749178e-05,
+      "loss": 0.1779,
+      "step": 1728
+    },
+    {
+      "epoch": 1.5950184501845017,
+      "grad_norm": 0.36509476764263094,
+      "learning_rate": 2.663619425293456e-05,
+      "loss": 0.1639,
+      "step": 1729
+    },
+    {
+      "epoch": 1.5959409594095941,
+      "grad_norm": 0.3858310656003099,
+      "learning_rate": 2.6609408821927838e-05,
+      "loss": 0.1693,
+      "step": 1730
+    },
+    {
+      "epoch": 1.5968634686346863,
+      "grad_norm": 0.3331195088745354,
+      "learning_rate": 2.6582621535607043e-05,
+      "loss": 0.1581,
+      "step": 1731
+    },
+    {
+      "epoch": 1.5977859778597785,
+      "grad_norm": 0.3405081382004149,
+      "learning_rate": 2.655583242485236e-05,
+      "loss": 0.1595,
+      "step": 1732
+    },
+    {
+      "epoch": 1.598708487084871,
+      "grad_norm": 0.3574851821043332,
+      "learning_rate": 2.652904152054607e-05,
+      "loss": 0.1621,
+      "step": 1733
+    },
+    {
+      "epoch": 1.599630996309963,
+      "grad_norm": 0.381571319987199,
+      "learning_rate": 2.650224885357251e-05,
+      "loss": 0.1676,
+      "step": 1734
+    },
+    {
+      "epoch": 1.6005535055350553,
+      "grad_norm": 0.35821102821722783,
+      "learning_rate": 2.6475454454818073e-05,
+      "loss": 0.1557,
+      "step": 1735
+    },
+    {
+      "epoch": 1.6014760147601477,
+      "grad_norm": 0.3764251524627297,
+      "learning_rate": 2.6448658355171125e-05,
+      "loss": 0.1638,
+      "step": 1736
+    },
+    {
+      "epoch": 1.6023985239852399,
+      "grad_norm": 0.38347245125149365,
+      "learning_rate": 2.6421860585522e-05,
+      "loss": 0.1776,
+      "step": 1737
+    },
+    {
+      "epoch": 1.603321033210332,
+      "grad_norm": 0.3768996262490828,
+      "learning_rate": 2.6395061176762976e-05,
+      "loss": 0.1597,
+      "step": 1738
+    },
+    {
+      "epoch": 1.6042435424354244,
+      "grad_norm": 0.34816471747517486,
+      "learning_rate": 2.6368260159788195e-05,
+      "loss": 0.1551,
+      "step": 1739
+    },
+    {
+      "epoch": 1.6051660516605166,
+      "grad_norm": 0.35597562064113825,
+      "learning_rate": 2.6341457565493654e-05,
+      "loss": 0.1668,
+      "step": 1740
+    },
+    {
+      "epoch": 1.6060885608856088,
+      "grad_norm": 0.4190496695256457,
+      "learning_rate": 2.6314653424777193e-05,
+      "loss": 0.1706,
+      "step": 1741
+    },
+    {
+      "epoch": 1.6070110701107012,
+      "grad_norm": 0.3576242465558538,
+      "learning_rate": 2.628784776853841e-05,
+      "loss": 0.1654,
+      "step": 1742
+    },
+    {
+      "epoch": 1.6079335793357934,
+      "grad_norm": 0.34269936127558054,
+      "learning_rate": 2.6261040627678655e-05,
+      "loss": 0.1607,
+      "step": 1743
+    },
+    {
+      "epoch": 1.6088560885608856,
+      "grad_norm": 0.36752172633589664,
+      "learning_rate": 2.6234232033101e-05,
+      "loss": 0.1661,
+      "step": 1744
+    },
+    {
+      "epoch": 1.609778597785978,
+      "grad_norm": 0.3524255109254831,
+      "learning_rate": 2.620742201571018e-05,
+      "loss": 0.1513,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6107011070110702,
+      "grad_norm": 0.33457956119732485,
+      "learning_rate": 2.6180610606412587e-05,
+      "loss": 0.1676,
+      "step": 1746
+    },
+    {
+      "epoch": 1.6116236162361623,
+      "grad_norm": 0.36581897070107644,
+      "learning_rate": 2.615379783611619e-05,
+      "loss": 0.1722,
+      "step": 1747
+    },
+    {
+      "epoch": 1.6125461254612548,
+      "grad_norm": 0.3775378029415688,
+      "learning_rate": 2.612698373573056e-05,
+      "loss": 0.1726,
+      "step": 1748
+    },
+    {
+      "epoch": 1.6134686346863467,
+      "grad_norm": 0.36484098955173416,
+      "learning_rate": 2.610016833616678e-05,
+      "loss": 0.1632,
+      "step": 1749
+    },
+    {
+      "epoch": 1.6143911439114391,
+      "grad_norm": 0.37034868726475717,
+      "learning_rate": 2.6073351668337425e-05,
+      "loss": 0.1483,
+      "step": 1750
+    },
+    {
+      "epoch": 1.6153136531365315,
+      "grad_norm": 0.3509729304946948,
+      "learning_rate": 2.6046533763156556e-05,
+      "loss": 0.1478,
+      "step": 1751
+    },
+    {
+      "epoch": 1.6162361623616235,
+      "grad_norm": 0.3491459694747171,
+      "learning_rate": 2.6019714651539646e-05,
+      "loss": 0.1492,
+      "step": 1752
+    },
+    {
+      "epoch": 1.617158671586716,
+      "grad_norm": 0.43367934165324495,
+      "learning_rate": 2.599289436440355e-05,
+      "loss": 0.174,
+      "step": 1753
+    },
+    {
+      "epoch": 1.618081180811808,
+      "grad_norm": 0.3929404569119107,
+      "learning_rate": 2.5966072932666496e-05,
+      "loss": 0.1559,
+      "step": 1754
+    },
+    {
+      "epoch": 1.6190036900369003,
+      "grad_norm": 0.3504889879631341,
+      "learning_rate": 2.593925038724802e-05,
+      "loss": 0.1624,
+      "step": 1755
+    },
+    {
+      "epoch": 1.6199261992619927,
+      "grad_norm": 0.3530763221610121,
+      "learning_rate": 2.5912426759068942e-05,
+      "loss": 0.1628,
+      "step": 1756
+    },
+    {
+      "epoch": 1.6208487084870848,
+      "grad_norm": 0.35679727055157445,
+      "learning_rate": 2.5885602079051353e-05,
+      "loss": 0.1646,
+      "step": 1757
+    },
+    {
+      "epoch": 1.621771217712177,
+      "grad_norm": 0.39285731384307976,
+      "learning_rate": 2.585877637811851e-05,
+      "loss": 0.1542,
+      "step": 1758
+    },
+    {
+      "epoch": 1.6226937269372694,
+      "grad_norm": 0.379985099892189,
+      "learning_rate": 2.5831949687194896e-05,
+      "loss": 0.151,
+      "step": 1759
+    },
+    {
+      "epoch": 1.6236162361623616,
+      "grad_norm": 0.4129181801521749,
+      "learning_rate": 2.5805122037206093e-05,
+      "loss": 0.1825,
+      "step": 1760
+    },
+    {
+      "epoch": 1.6245387453874538,
+      "grad_norm": 0.403755626049169,
+      "learning_rate": 2.5778293459078828e-05,
+      "loss": 0.1549,
+      "step": 1761
+    },
+    {
+      "epoch": 1.6254612546125462,
+      "grad_norm": 0.3828008999461262,
+      "learning_rate": 2.575146398374087e-05,
+      "loss": 0.1591,
+      "step": 1762
+    },
+    {
+      "epoch": 1.6263837638376384,
+      "grad_norm": 0.38324101709104025,
+      "learning_rate": 2.5724633642121025e-05,
+      "loss": 0.1548,
+      "step": 1763
+    },
+    {
+      "epoch": 1.6273062730627306,
+      "grad_norm": 0.3695771356000055,
+      "learning_rate": 2.5697802465149117e-05,
+      "loss": 0.1537,
+      "step": 1764
+    },
+    {
+      "epoch": 1.628228782287823,
+      "grad_norm": 0.36402826938395866,
+      "learning_rate": 2.5670970483755912e-05,
+      "loss": 0.1593,
+      "step": 1765
+    },
+    {
+      "epoch": 1.6291512915129152,
+      "grad_norm": 0.367222780146665,
+      "learning_rate": 2.5644137728873107e-05,
+      "loss": 0.1615,
+      "step": 1766
+    },
+    {
+      "epoch": 1.6300738007380073,
+      "grad_norm": 0.3175180820245255,
+      "learning_rate": 2.5617304231433305e-05,
+      "loss": 0.1396,
+      "step": 1767
+    },
+    {
+      "epoch": 1.6309963099630997,
+      "grad_norm": 0.3579853168539979,
+      "learning_rate": 2.559047002236995e-05,
+      "loss": 0.1728,
+      "step": 1768
+    },
+    {
+      "epoch": 1.631918819188192,
+      "grad_norm": 0.34460467282407875,
+      "learning_rate": 2.5563635132617302e-05,
+      "loss": 0.1606,
+      "step": 1769
+    },
+    {
+      "epoch": 1.632841328413284,
+      "grad_norm": 0.3523090029542282,
+      "learning_rate": 2.553679959311044e-05,
+      "loss": 0.17,
+      "step": 1770
+    },
+    {
+      "epoch": 1.6337638376383765,
+      "grad_norm": 0.34425584939112036,
+      "learning_rate": 2.550996343478514e-05,
+      "loss": 0.169,
+      "step": 1771
+    },
+    {
+      "epoch": 1.6346863468634685,
+      "grad_norm": 0.36165025369113873,
+      "learning_rate": 2.5483126688577926e-05,
+      "loss": 0.1629,
+      "step": 1772
+    },
+    {
+      "epoch": 1.6356088560885609,
+      "grad_norm": 0.35481197568027084,
+      "learning_rate": 2.5456289385426e-05,
+      "loss": 0.1363,
+      "step": 1773
+    },
+    {
+      "epoch": 1.6365313653136533,
+      "grad_norm": 0.34246971926457204,
+      "learning_rate": 2.5429451556267187e-05,
+      "loss": 0.1535,
+      "step": 1774
+    },
+    {
+      "epoch": 1.6374538745387452,
+      "grad_norm": 0.3726840562658278,
+      "learning_rate": 2.5402613232039934e-05,
+      "loss": 0.1519,
+      "step": 1775
+    },
+    {
+      "epoch": 1.6383763837638377,
+      "grad_norm": 0.38379172397334155,
+      "learning_rate": 2.5375774443683265e-05,
+      "loss": 0.1708,
+      "step": 1776
+    },
+    {
+      "epoch": 1.6392988929889298,
+      "grad_norm": 0.34258217495610643,
+      "learning_rate": 2.5348935222136704e-05,
+      "loss": 0.1596,
+      "step": 1777
+    },
+    {
+      "epoch": 1.640221402214022,
+      "grad_norm": 0.38480369696157934,
+      "learning_rate": 2.5322095598340322e-05,
+      "loss": 0.1459,
+      "step": 1778
+    },
+    {
+      "epoch": 1.6411439114391144,
+      "grad_norm": 0.36182050713057723,
+      "learning_rate": 2.529525560323462e-05,
+      "loss": 0.1694,
+      "step": 1779
+    },
+    {
+      "epoch": 1.6420664206642066,
+      "grad_norm": 0.38787979484694496,
+      "learning_rate": 2.5268415267760526e-05,
+      "loss": 0.1599,
+      "step": 1780
+    },
+    {
+      "epoch": 1.6429889298892988,
+      "grad_norm": 0.3495537975863066,
+      "learning_rate": 2.5241574622859394e-05,
+      "loss": 0.1702,
+      "step": 1781
+    },
+    {
+      "epoch": 1.6439114391143912,
+      "grad_norm": 0.3569848539736059,
+      "learning_rate": 2.521473369947289e-05,
+      "loss": 0.1614,
+      "step": 1782
+    },
+    {
+      "epoch": 1.6448339483394834,
+      "grad_norm": 0.4155549926988143,
+      "learning_rate": 2.518789252854305e-05,
+      "loss": 0.1982,
+      "step": 1783
+    },
+    {
+      "epoch": 1.6457564575645756,
+      "grad_norm": 0.3795448375787088,
+      "learning_rate": 2.516105114101215e-05,
+      "loss": 0.1626,
+      "step": 1784
+    },
+    {
+      "epoch": 1.646678966789668,
+      "grad_norm": 0.371239016889001,
+      "learning_rate": 2.5134209567822724e-05,
+      "loss": 0.1607,
+      "step": 1785
+    },
+    {
+      "epoch": 1.6476014760147601,
+      "grad_norm": 0.36075383156663865,
+      "learning_rate": 2.510736783991756e-05,
+      "loss": 0.164,
+      "step": 1786
+    },
+    {
+      "epoch": 1.6485239852398523,
+      "grad_norm": 0.3735289734675146,
+      "learning_rate": 2.5080525988239574e-05,
+      "loss": 0.16,
+      "step": 1787
+    },
+    {
+      "epoch": 1.6494464944649447,
+      "grad_norm": 0.37814164032652253,
+      "learning_rate": 2.5053684043731847e-05,
+      "loss": 0.156,
+      "step": 1788
+    },
+    {
+      "epoch": 1.650369003690037,
+      "grad_norm": 0.40747785575746726,
+      "learning_rate": 2.502684203733758e-05,
+      "loss": 0.162,
+      "step": 1789
+    },
+    {
+      "epoch": 1.651291512915129,
+      "grad_norm": 0.3248782786503274,
+      "learning_rate": 2.5e-05,
+      "loss": 0.1421,
+      "step": 1790
+    },
+    {
+      "epoch": 1.6522140221402215,
+      "grad_norm": 0.37339369747781037,
+      "learning_rate": 2.4973157962662437e-05,
+      "loss": 0.1287,
+      "step": 1791
+    },
+    {
+      "epoch": 1.6531365313653137,
+      "grad_norm": 0.391019477959082,
+      "learning_rate": 2.4946315956268156e-05,
+      "loss": 0.1677,
+      "step": 1792
+    },
+    {
+      "epoch": 1.6540590405904059,
+      "grad_norm": 0.3579934657099326,
+      "learning_rate": 2.4919474011760432e-05,
+      "loss": 0.1776,
+      "step": 1793
+    },
+    {
+      "epoch": 1.6549815498154983,
+      "grad_norm": 0.40890887145853805,
+      "learning_rate": 2.4892632160082448e-05,
+      "loss": 0.1585,
+      "step": 1794
+    },
+    {
+      "epoch": 1.6559040590405905,
+      "grad_norm": 0.349869966597724,
+      "learning_rate": 2.486579043217727e-05,
+      "loss": 0.152,
+      "step": 1795
+    },
+    {
+      "epoch": 1.6568265682656826,
+      "grad_norm": 0.4295873001981432,
+      "learning_rate": 2.483894885898786e-05,
+      "loss": 0.181,
+      "step": 1796
+    },
+    {
+      "epoch": 1.657749077490775,
+      "grad_norm": 0.3512100711226925,
+      "learning_rate": 2.4812107471456954e-05,
+      "loss": 0.1618,
+      "step": 1797
+    },
+    {
+      "epoch": 1.658671586715867,
+      "grad_norm": 0.39208911596233437,
+      "learning_rate": 2.4785266300527105e-05,
+      "loss": 0.1915,
+      "step": 1798
+    },
+    {
+      "epoch": 1.6595940959409594,
+      "grad_norm": 0.3469507405655631,
+      "learning_rate": 2.4758425377140612e-05,
+      "loss": 0.1613,
+      "step": 1799
+    },
+    {
+      "epoch": 1.6605166051660518,
+      "grad_norm": 0.3661258766022745,
+      "learning_rate": 2.4731584732239486e-05,
+      "loss": 0.1479,
+      "step": 1800
+    },
+    {
+      "epoch": 1.6614391143911438,
+      "grad_norm": 0.3071602425954102,
+      "learning_rate": 2.470474439676539e-05,
+      "loss": 0.1536,
+      "step": 1801
+    },
+    {
+      "epoch": 1.6623616236162362,
+      "grad_norm": 0.34788042054454105,
+      "learning_rate": 2.4677904401659684e-05,
+      "loss": 0.1552,
+      "step": 1802
+    },
+    {
+      "epoch": 1.6632841328413284,
+      "grad_norm": 0.4010228125029169,
+      "learning_rate": 2.4651064777863305e-05,
+      "loss": 0.1753,
+      "step": 1803
+    },
+    {
+      "epoch": 1.6642066420664205,
+      "grad_norm": 0.4014536564406406,
+      "learning_rate": 2.4624225556316744e-05,
+      "loss": 0.175,
+      "step": 1804
+    },
+    {
+      "epoch": 1.665129151291513,
+      "grad_norm": 0.39391368995880854,
+      "learning_rate": 2.4597386767960075e-05,
+      "loss": 0.1469,
+      "step": 1805
+    },
+    {
+      "epoch": 1.6660516605166051,
+      "grad_norm": 0.3662948621635896,
+      "learning_rate": 2.4570548443732825e-05,
+      "loss": 0.1672,
+      "step": 1806
+    },
+    {
+      "epoch": 1.6669741697416973,
+      "grad_norm": 0.4180461633069466,
+      "learning_rate": 2.4543710614574005e-05,
+      "loss": 0.1689,
+      "step": 1807
+    },
+    {
+      "epoch": 1.6678966789667897,
+      "grad_norm": 0.3733317090949078,
+      "learning_rate": 2.4516873311422083e-05,
+      "loss": 0.1637,
+      "step": 1808
+    },
+    {
+      "epoch": 1.668819188191882,
+      "grad_norm": 0.37684961403111317,
+      "learning_rate": 2.4490036565214873e-05,
+      "loss": 0.1565,
+      "step": 1809
+    },
+    {
+      "epoch": 1.669741697416974,
+      "grad_norm": 0.37196143979647983,
+      "learning_rate": 2.4463200406889562e-05,
+      "loss": 0.1581,
+      "step": 1810
+    },
+    {
+      "epoch": 1.6706642066420665,
+      "grad_norm": 0.421690946853156,
+      "learning_rate": 2.44363648673827e-05,
+      "loss": 0.1847,
+      "step": 1811
+    },
+    {
+      "epoch": 1.6715867158671587,
+      "grad_norm": 0.37568701071123256,
+      "learning_rate": 2.440952997763005e-05,
+      "loss": 0.1752,
+      "step": 1812
+    },
+    {
+      "epoch": 1.6725092250922509,
+      "grad_norm": 0.3557026738053844,
+      "learning_rate": 2.4382695768566697e-05,
+      "loss": 0.1625,
+      "step": 1813
+    },
+    {
+      "epoch": 1.6734317343173433,
+      "grad_norm": 0.3917601421523301,
+      "learning_rate": 2.4355862271126896e-05,
+      "loss": 0.1679,
+      "step": 1814
+    },
+    {
+      "epoch": 1.6743542435424354,
+      "grad_norm": 0.406327508192895,
+      "learning_rate": 2.432902951624409e-05,
+      "loss": 0.1481,
+      "step": 1815
+    },
+    {
+      "epoch": 1.6752767527675276,
+      "grad_norm": 0.3788673145366441,
+      "learning_rate": 2.4302197534850892e-05,
+      "loss": 0.1507,
+      "step": 1816
+    },
+    {
+      "epoch": 1.67619926199262,
+      "grad_norm": 0.3531155351857008,
+      "learning_rate": 2.427536635787898e-05,
+      "loss": 0.1639,
+      "step": 1817
+    },
+    {
+      "epoch": 1.6771217712177122,
+      "grad_norm": 0.48329291467772933,
+      "learning_rate": 2.4248536016259135e-05,
+      "loss": 0.1714,
+      "step": 1818
+    },
+    {
+      "epoch": 1.6780442804428044,
+      "grad_norm": 0.3666417542770587,
+      "learning_rate": 2.4221706540921178e-05,
+      "loss": 0.1761,
+      "step": 1819
+    },
+    {
+      "epoch": 1.6789667896678968,
+      "grad_norm": 0.34653859689705124,
+      "learning_rate": 2.4194877962793913e-05,
+      "loss": 0.1801,
+      "step": 1820
+    },
+    {
+      "epoch": 1.6798892988929888,
+      "grad_norm": 0.4110613509102009,
+      "learning_rate": 2.416805031280511e-05,
+      "loss": 0.1616,
+      "step": 1821
+    },
+    {
+      "epoch": 1.6808118081180812,
+      "grad_norm": 0.36485843700318177,
+      "learning_rate": 2.4141223621881495e-05,
+      "loss": 0.1544,
+      "step": 1822
+    },
+    {
+      "epoch": 1.6817343173431736,
+      "grad_norm": 0.31499028962004866,
+      "learning_rate": 2.4114397920948657e-05,
+      "loss": 0.1595,
+      "step": 1823
+    },
+    {
+      "epoch": 1.6826568265682655,
+      "grad_norm": 0.40781315168296495,
+      "learning_rate": 2.4087573240931053e-05,
+      "loss": 0.173,
+      "step": 1824
+    },
+    {
+      "epoch": 1.683579335793358,
+      "grad_norm": 0.4175211608803691,
+      "learning_rate": 2.4060749612751988e-05,
+      "loss": 0.1805,
+      "step": 1825
+    },
+    {
+      "epoch": 1.6845018450184504,
+      "grad_norm": 0.3268684870697504,
+      "learning_rate": 2.4033927067333513e-05,
+      "loss": 0.1805,
+      "step": 1826
+    },
+    {
+      "epoch": 1.6854243542435423,
+      "grad_norm": 0.36505231798978677,
+      "learning_rate": 2.4007105635596454e-05,
+      "loss": 0.1626,
+      "step": 1827
+    },
+    {
+      "epoch": 1.6863468634686347,
+      "grad_norm": 0.33455639629313166,
+      "learning_rate": 2.3980285348460363e-05,
+      "loss": 0.1715,
+      "step": 1828
+    },
+    {
+      "epoch": 1.687269372693727,
+      "grad_norm": 0.29965711634892783,
+      "learning_rate": 2.395346623684345e-05,
+      "loss": 0.1497,
+      "step": 1829
+    },
+    {
+      "epoch": 1.688191881918819,
+      "grad_norm": 0.36809149141297764,
+      "learning_rate": 2.3926648331662578e-05,
+      "loss": 0.1728,
+      "step": 1830
+    },
+    {
+      "epoch": 1.6891143911439115,
+      "grad_norm": 0.3576239939689782,
+      "learning_rate": 2.389983166383323e-05,
+      "loss": 0.1642,
+      "step": 1831
+    },
+    {
+      "epoch": 1.6900369003690037,
+      "grad_norm": 0.3675704061928072,
+      "learning_rate": 2.387301626426944e-05,
+      "loss": 0.1563,
+      "step": 1832
+    },
+    {
+      "epoch": 1.6909594095940959,
+      "grad_norm": 0.3913242233199537,
+      "learning_rate": 2.3846202163883807e-05,
+      "loss": 0.1642,
+      "step": 1833
+    },
+    {
+      "epoch": 1.6918819188191883,
+      "grad_norm": 0.3540709417362491,
+      "learning_rate": 2.381938939358742e-05,
+      "loss": 0.1621,
+      "step": 1834
+    },
+    {
+      "epoch": 1.6928044280442804,
+      "grad_norm": 0.3618416232702187,
+      "learning_rate": 2.3792577984289825e-05,
+      "loss": 0.1787,
+      "step": 1835
+    },
+    {
+      "epoch": 1.6937269372693726,
+      "grad_norm": 0.35382274617948284,
+      "learning_rate": 2.3765767966899e-05,
+      "loss": 0.1676,
+      "step": 1836
+    },
+    {
+      "epoch": 1.694649446494465,
+      "grad_norm": 0.36513333624772454,
+      "learning_rate": 2.3738959372321347e-05,
+      "loss": 0.1533,
+      "step": 1837
+    },
+    {
+      "epoch": 1.6955719557195572,
+      "grad_norm": 0.408832356741295,
+      "learning_rate": 2.37121522314616e-05,
+      "loss": 0.1648,
+      "step": 1838
+    },
+    {
+      "epoch": 1.6964944649446494,
+      "grad_norm": 0.41739708583790197,
+      "learning_rate": 2.368534657522281e-05,
+      "loss": 0.1668,
+      "step": 1839
+    },
+    {
+      "epoch": 1.6974169741697418,
+      "grad_norm": 0.33318751572141103,
+      "learning_rate": 2.3658542434506352e-05,
+      "loss": 0.1446,
+      "step": 1840
+    },
+    {
+      "epoch": 1.698339483394834,
+      "grad_norm": 0.3521329697255817,
+      "learning_rate": 2.3631739840211817e-05,
+      "loss": 0.1613,
+      "step": 1841
+    },
+    {
+      "epoch": 1.6992619926199262,
+      "grad_norm": 0.40553912933042213,
+      "learning_rate": 2.3604938823237023e-05,
+      "loss": 0.1772,
+      "step": 1842
+    },
+    {
+      "epoch": 1.7001845018450186,
+      "grad_norm": 0.43117642985897287,
+      "learning_rate": 2.3578139414478002e-05,
+      "loss": 0.1779,
+      "step": 1843
+    },
+    {
+      "epoch": 1.7011070110701108,
+      "grad_norm": 0.3564227968003786,
+      "learning_rate": 2.3551341644828884e-05,
+      "loss": 0.1712,
+      "step": 1844
+    },
+    {
+      "epoch": 1.702029520295203,
+      "grad_norm": 0.3471065097683919,
+      "learning_rate": 2.3524545545181933e-05,
+      "loss": 0.1592,
+      "step": 1845
+    },
+    {
+      "epoch": 1.7029520295202953,
+      "grad_norm": 0.3743744653135974,
+      "learning_rate": 2.3497751146427493e-05,
+      "loss": 0.1706,
+      "step": 1846
+    },
+    {
+      "epoch": 1.7038745387453873,
+      "grad_norm": 0.35803695186074397,
+      "learning_rate": 2.3470958479453938e-05,
+      "loss": 0.1495,
+      "step": 1847
+    },
+    {
+      "epoch": 1.7047970479704797,
+      "grad_norm": 0.3949075012122784,
+      "learning_rate": 2.344416757514764e-05,
+      "loss": 0.1612,
+      "step": 1848
+    },
+    {
+      "epoch": 1.7057195571955721,
+      "grad_norm": 0.36782728105119133,
+      "learning_rate": 2.3417378464392963e-05,
+      "loss": 0.1478,
+      "step": 1849
+    },
+    {
+      "epoch": 1.706642066420664,
+      "grad_norm": 0.37391347662986596,
+      "learning_rate": 2.339059117807217e-05,
+      "loss": 0.1654,
+      "step": 1850
+    },
+    {
+      "epoch": 1.7075645756457565,
+      "grad_norm": 0.39306774002597317,
+      "learning_rate": 2.3363805747065443e-05,
+      "loss": 0.1808,
+      "step": 1851
+    },
+    {
+      "epoch": 1.7084870848708487,
+      "grad_norm": 0.38057434121566486,
+      "learning_rate": 2.3337022202250828e-05,
+      "loss": 0.1814,
+      "step": 1852
+    },
+    {
+      "epoch": 1.7094095940959408,
+      "grad_norm": 0.3339814040682833,
+      "learning_rate": 2.3310240574504185e-05,
+      "loss": 0.1373,
+      "step": 1853
+    },
+    {
+      "epoch": 1.7103321033210332,
+      "grad_norm": 0.3963449827644891,
+      "learning_rate": 2.3283460894699156e-05,
+      "loss": 0.167,
+      "step": 1854
+    },
+    {
+      "epoch": 1.7112546125461254,
+      "grad_norm": 0.40301591658685915,
+      "learning_rate": 2.3256683193707166e-05,
+      "loss": 0.1557,
+      "step": 1855
+    },
+    {
+      "epoch": 1.7121771217712176,
+      "grad_norm": 0.4226352086554742,
+      "learning_rate": 2.322990750239733e-05,
+      "loss": 0.1625,
+      "step": 1856
+    },
+    {
+      "epoch": 1.71309963099631,
+      "grad_norm": 0.34736340879699523,
+      "learning_rate": 2.3203133851636465e-05,
+      "loss": 0.1607,
+      "step": 1857
+    },
+    {
+      "epoch": 1.7140221402214022,
+      "grad_norm": 0.3681170700006526,
+      "learning_rate": 2.317636227228903e-05,
+      "loss": 0.1549,
+      "step": 1858
+    },
+    {
+      "epoch": 1.7149446494464944,
+      "grad_norm": 0.3443452415949501,
+      "learning_rate": 2.314959279521709e-05,
+      "loss": 0.1591,
+      "step": 1859
+    },
+    {
+      "epoch": 1.7158671586715868,
+      "grad_norm": 0.31743163801936963,
+      "learning_rate": 2.3122825451280296e-05,
+      "loss": 0.1608,
+      "step": 1860
+    },
+    {
+      "epoch": 1.716789667896679,
+      "grad_norm": 0.44848989979697457,
+      "learning_rate": 2.3096060271335832e-05,
+      "loss": 0.1847,
+      "step": 1861
+    },
+    {
+      "epoch": 1.7177121771217712,
+      "grad_norm": 0.36396283944358304,
+      "learning_rate": 2.306929728623839e-05,
+      "loss": 0.1762,
+      "step": 1862
+    },
+    {
+      "epoch": 1.7186346863468636,
+      "grad_norm": 0.38223590312959876,
+      "learning_rate": 2.3042536526840134e-05,
+      "loss": 0.1593,
+      "step": 1863
+    },
+    {
+      "epoch": 1.7195571955719557,
+      "grad_norm": 0.36568180560510427,
+      "learning_rate": 2.3015778023990667e-05,
+      "loss": 0.1707,
+      "step": 1864
+    },
+    {
+      "epoch": 1.720479704797048,
+      "grad_norm": 0.32810773073454375,
+      "learning_rate": 2.2989021808536974e-05,
+      "loss": 0.1597,
+      "step": 1865
+    },
+    {
+      "epoch": 1.7214022140221403,
+      "grad_norm": 0.4709336738948193,
+      "learning_rate": 2.296226791132342e-05,
+      "loss": 0.1704,
+      "step": 1866
+    },
+    {
+      "epoch": 1.7223247232472325,
+      "grad_norm": 0.34675317136160727,
+      "learning_rate": 2.2935516363191693e-05,
+      "loss": 0.1772,
+      "step": 1867
+    },
+    {
+      "epoch": 1.7232472324723247,
+      "grad_norm": 0.3926172439787084,
+      "learning_rate": 2.2908767194980764e-05,
+      "loss": 0.1717,
+      "step": 1868
+    },
+    {
+      "epoch": 1.724169741697417,
+      "grad_norm": 0.35368453034298086,
+      "learning_rate": 2.2882020437526873e-05,
+      "loss": 0.1585,
+      "step": 1869
+    },
+    {
+      "epoch": 1.725092250922509,
+      "grad_norm": 0.3639008053704692,
+      "learning_rate": 2.2855276121663485e-05,
+      "loss": 0.1644,
+      "step": 1870
+    },
+    {
+      "epoch": 1.7260147601476015,
+      "grad_norm": 0.367283149868555,
+      "learning_rate": 2.2828534278221212e-05,
+      "loss": 0.1617,
+      "step": 1871
+    },
+    {
+      "epoch": 1.7269372693726939,
+      "grad_norm": 0.35207757734068457,
+      "learning_rate": 2.2801794938027873e-05,
+      "loss": 0.1574,
+      "step": 1872
+    },
+    {
+      "epoch": 1.7278597785977858,
+      "grad_norm": 0.3313244822773851,
+      "learning_rate": 2.277505813190837e-05,
+      "loss": 0.1616,
+      "step": 1873
+    },
+    {
+      "epoch": 1.7287822878228782,
+      "grad_norm": 0.37018665631685116,
+      "learning_rate": 2.2748323890684665e-05,
+      "loss": 0.1495,
+      "step": 1874
+    },
+    {
+      "epoch": 1.7297047970479706,
+      "grad_norm": 0.35511624838409717,
+      "learning_rate": 2.2721592245175812e-05,
+      "loss": 0.1698,
+      "step": 1875
+    },
+    {
+      "epoch": 1.7306273062730626,
+      "grad_norm": 0.3674813704404082,
+      "learning_rate": 2.269486322619783e-05,
+      "loss": 0.1663,
+      "step": 1876
+    },
+    {
+      "epoch": 1.731549815498155,
+      "grad_norm": 0.32742579310033787,
+      "learning_rate": 2.266813686456372e-05,
+      "loss": 0.1685,
+      "step": 1877
+    },
+    {
+      "epoch": 1.7324723247232472,
+      "grad_norm": 0.3827716207681046,
+      "learning_rate": 2.2641413191083445e-05,
+      "loss": 0.1785,
+      "step": 1878
+    },
+    {
+      "epoch": 1.7333948339483394,
+      "grad_norm": 0.3530017875613452,
+      "learning_rate": 2.2614692236563838e-05,
+      "loss": 0.1582,
+      "step": 1879
+    },
+    {
+      "epoch": 1.7343173431734318,
+      "grad_norm": 0.39018328169038247,
+      "learning_rate": 2.2587974031808608e-05,
+      "loss": 0.151,
+      "step": 1880
+    },
+    {
+      "epoch": 1.735239852398524,
+      "grad_norm": 0.37853261835392593,
+      "learning_rate": 2.2561258607618297e-05,
+      "loss": 0.1657,
+      "step": 1881
+    },
+    {
+      "epoch": 1.7361623616236161,
+      "grad_norm": 0.35354906387489693,
+      "learning_rate": 2.2534545994790244e-05,
+      "loss": 0.1537,
+      "step": 1882
+    },
+    {
+      "epoch": 1.7370848708487086,
+      "grad_norm": 0.3990954409001275,
+      "learning_rate": 2.250783622411853e-05,
+      "loss": 0.1773,
+      "step": 1883
+    },
+    {
+      "epoch": 1.7380073800738007,
+      "grad_norm": 0.3944526491031225,
+      "learning_rate": 2.2481129326393992e-05,
+      "loss": 0.1646,
+      "step": 1884
+    },
+    {
+      "epoch": 1.738929889298893,
+      "grad_norm": 0.36893109396357665,
+      "learning_rate": 2.2454425332404122e-05,
+      "loss": 0.1664,
+      "step": 1885
+    },
+    {
+      "epoch": 1.7398523985239853,
+      "grad_norm": 0.339583638584494,
+      "learning_rate": 2.2427724272933075e-05,
+      "loss": 0.1602,
+      "step": 1886
+    },
+    {
+      "epoch": 1.7407749077490775,
+      "grad_norm": 0.34874589566586056,
+      "learning_rate": 2.240102617876164e-05,
+      "loss": 0.1472,
+      "step": 1887
+    },
+    {
+      "epoch": 1.7416974169741697,
+      "grad_norm": 0.3593168206981894,
+      "learning_rate": 2.2374331080667164e-05,
+      "loss": 0.1723,
+      "step": 1888
+    },
+    {
+      "epoch": 1.742619926199262,
+      "grad_norm": 0.43778282669439506,
+      "learning_rate": 2.2347639009423553e-05,
+      "loss": 0.1772,
+      "step": 1889
+    },
+    {
+      "epoch": 1.7435424354243543,
+      "grad_norm": 0.3387112834557388,
+      "learning_rate": 2.2320949995801222e-05,
+      "loss": 0.1544,
+      "step": 1890
+    },
+    {
+      "epoch": 1.7444649446494465,
+      "grad_norm": 0.3849249076716214,
+      "learning_rate": 2.2294264070567056e-05,
+      "loss": 0.1736,
+      "step": 1891
+    },
+    {
+      "epoch": 1.7453874538745389,
+      "grad_norm": 0.2938543747815116,
+      "learning_rate": 2.2267581264484382e-05,
+      "loss": 0.1334,
+      "step": 1892
+    },
+    {
+      "epoch": 1.746309963099631,
+      "grad_norm": 0.3683306778476559,
+      "learning_rate": 2.2240901608312942e-05,
+      "loss": 0.155,
+      "step": 1893
+    },
+    {
+      "epoch": 1.7472324723247232,
+      "grad_norm": 0.39697443182198616,
+      "learning_rate": 2.2214225132808837e-05,
+      "loss": 0.1666,
+      "step": 1894
+    },
+    {
+      "epoch": 1.7481549815498156,
+      "grad_norm": 0.3604022476778861,
+      "learning_rate": 2.2187551868724485e-05,
+      "loss": 0.1555,
+      "step": 1895
+    },
+    {
+      "epoch": 1.7490774907749076,
+      "grad_norm": 0.35881561561350755,
+      "learning_rate": 2.216088184680864e-05,
+      "loss": 0.1464,
+      "step": 1896
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.33670919990207204,
+      "learning_rate": 2.2134215097806295e-05,
+      "loss": 0.1494,
+      "step": 1897
+    },
+    {
+      "epoch": 1.7509225092250924,
+      "grad_norm": 0.33737836275487587,
+      "learning_rate": 2.2107551652458648e-05,
+      "loss": 0.1436,
+      "step": 1898
+    },
+    {
+      "epoch": 1.7518450184501844,
+      "grad_norm": 0.35033115584252883,
+      "learning_rate": 2.2080891541503145e-05,
+      "loss": 0.1533,
+      "step": 1899
+    },
+    {
+      "epoch": 1.7527675276752768,
+      "grad_norm": 0.39914141234672607,
+      "learning_rate": 2.2054234795673334e-05,
+      "loss": 0.1659,
+      "step": 1900
+    },
+    {
+      "epoch": 1.753690036900369,
+      "grad_norm": 0.39865041105966287,
+      "learning_rate": 2.2027581445698922e-05,
+      "loss": 0.1284,
+      "step": 1901
+    },
+    {
+      "epoch": 1.7546125461254611,
+      "grad_norm": 0.40139166331439674,
+      "learning_rate": 2.200093152230568e-05,
+      "loss": 0.186,
+      "step": 1902
+    },
+    {
+      "epoch": 1.7555350553505535,
+      "grad_norm": 0.3601541632619284,
+      "learning_rate": 2.197428505621542e-05,
+      "loss": 0.1481,
+      "step": 1903
+    },
+    {
+      "epoch": 1.7564575645756457,
+      "grad_norm": 0.37537784969448856,
+      "learning_rate": 2.1947642078146004e-05,
+      "loss": 0.1746,
+      "step": 1904
+    },
+    {
+      "epoch": 1.757380073800738,
+      "grad_norm": 0.3820863338038734,
+      "learning_rate": 2.1921002618811244e-05,
+      "loss": 0.1788,
+      "step": 1905
+    },
+    {
+      "epoch": 1.7583025830258303,
+      "grad_norm": 0.3454516502653024,
+      "learning_rate": 2.1894366708920886e-05,
+      "loss": 0.1612,
+      "step": 1906
+    },
+    {
+      "epoch": 1.7592250922509225,
+      "grad_norm": 0.35810941378697064,
+      "learning_rate": 2.1867734379180628e-05,
+      "loss": 0.1556,
+      "step": 1907
+    },
+    {
+      "epoch": 1.7601476014760147,
+      "grad_norm": 0.35804285212253945,
+      "learning_rate": 2.1841105660292e-05,
+      "loss": 0.1533,
+      "step": 1908
+    },
+    {
+      "epoch": 1.761070110701107,
+      "grad_norm": 0.332110775349014,
+      "learning_rate": 2.1814480582952375e-05,
+      "loss": 0.1279,
+      "step": 1909
+    },
+    {
+      "epoch": 1.7619926199261993,
+      "grad_norm": 0.3970112390451209,
+      "learning_rate": 2.1787859177854964e-05,
+      "loss": 0.1645,
+      "step": 1910
+    },
+    {
+      "epoch": 1.7629151291512914,
+      "grad_norm": 0.34157855663060777,
+      "learning_rate": 2.1761241475688695e-05,
+      "loss": 0.1612,
+      "step": 1911
+    },
+    {
+      "epoch": 1.7638376383763839,
+      "grad_norm": 0.3280523541309268,
+      "learning_rate": 2.1734627507138244e-05,
+      "loss": 0.1419,
+      "step": 1912
+    },
+    {
+      "epoch": 1.764760147601476,
+      "grad_norm": 0.3462767125893894,
+      "learning_rate": 2.1708017302884016e-05,
+      "loss": 0.1657,
+      "step": 1913
+    },
+    {
+      "epoch": 1.7656826568265682,
+      "grad_norm": 0.3248813728241005,
+      "learning_rate": 2.168141089360203e-05,
+      "loss": 0.1499,
+      "step": 1914
+    },
+    {
+      "epoch": 1.7666051660516606,
+      "grad_norm": 0.3851727529303274,
+      "learning_rate": 2.1654808309963938e-05,
+      "loss": 0.187,
+      "step": 1915
+    },
+    {
+      "epoch": 1.7675276752767528,
+      "grad_norm": 0.3771866949552074,
+      "learning_rate": 2.1628209582637022e-05,
+      "loss": 0.1625,
+      "step": 1916
+    },
+    {
+      "epoch": 1.768450184501845,
+      "grad_norm": 0.3475517371282469,
+      "learning_rate": 2.160161474228407e-05,
+      "loss": 0.1568,
+      "step": 1917
+    },
+    {
+      "epoch": 1.7693726937269374,
+      "grad_norm": 0.35822102164648034,
+      "learning_rate": 2.157502381956341e-05,
+      "loss": 0.1461,
+      "step": 1918
+    },
+    {
+      "epoch": 1.7702952029520294,
+      "grad_norm": 0.3570314929373,
+      "learning_rate": 2.1548436845128858e-05,
+      "loss": 0.1455,
+      "step": 1919
+    },
+    {
+      "epoch": 1.7712177121771218,
+      "grad_norm": 0.34562742110365285,
+      "learning_rate": 2.1521853849629675e-05,
+      "loss": 0.1593,
+      "step": 1920
+    },
+    {
+      "epoch": 1.7721402214022142,
+      "grad_norm": 0.3692231165374884,
+      "learning_rate": 2.1495274863710517e-05,
+      "loss": 0.1549,
+      "step": 1921
+    },
+    {
+      "epoch": 1.7730627306273061,
+      "grad_norm": 0.37465863169104,
+      "learning_rate": 2.146869991801146e-05,
+      "loss": 0.1708,
+      "step": 1922
+    },
+    {
+      "epoch": 1.7739852398523985,
+      "grad_norm": 0.35179057696040306,
+      "learning_rate": 2.1442129043167874e-05,
+      "loss": 0.1449,
+      "step": 1923
+    },
+    {
+      "epoch": 1.774907749077491,
+      "grad_norm": 0.3282554209480367,
+      "learning_rate": 2.1415562269810465e-05,
+      "loss": 0.1501,
+      "step": 1924
+    },
+    {
+      "epoch": 1.775830258302583,
+      "grad_norm": 0.3792811815541581,
+      "learning_rate": 2.1388999628565212e-05,
+      "loss": 0.1507,
+      "step": 1925
+    },
+    {
+      "epoch": 1.7767527675276753,
+      "grad_norm": 0.40143599228679433,
+      "learning_rate": 2.1362441150053312e-05,
+      "loss": 0.1652,
+      "step": 1926
+    },
+    {
+      "epoch": 1.7776752767527675,
+      "grad_norm": 0.37802858331533024,
+      "learning_rate": 2.1335886864891182e-05,
+      "loss": 0.1713,
+      "step": 1927
+    },
+    {
+      "epoch": 1.7785977859778597,
+      "grad_norm": 0.3525746877132614,
+      "learning_rate": 2.130933680369039e-05,
+      "loss": 0.1623,
+      "step": 1928
+    },
+    {
+      "epoch": 1.779520295202952,
+      "grad_norm": 0.39454226179717417,
+      "learning_rate": 2.128279099705765e-05,
+      "loss": 0.1544,
+      "step": 1929
+    },
+    {
+      "epoch": 1.7804428044280443,
+      "grad_norm": 0.34775886716768395,
+      "learning_rate": 2.125624947559475e-05,
+      "loss": 0.1474,
+      "step": 1930
+    },
+    {
+      "epoch": 1.7813653136531364,
+      "grad_norm": 0.36876573765007625,
+      "learning_rate": 2.1229712269898565e-05,
+      "loss": 0.1551,
+      "step": 1931
+    },
+    {
+      "epoch": 1.7822878228782288,
+      "grad_norm": 0.34869551059545806,
+      "learning_rate": 2.120317941056098e-05,
+      "loss": 0.1447,
+      "step": 1932
+    },
+    {
+      "epoch": 1.783210332103321,
+      "grad_norm": 0.36587891022317953,
+      "learning_rate": 2.117665092816885e-05,
+      "loss": 0.1671,
+      "step": 1933
+    },
+    {
+      "epoch": 1.7841328413284132,
+      "grad_norm": 0.3487945989291348,
+      "learning_rate": 2.1150126853304034e-05,
+      "loss": 0.1509,
+      "step": 1934
+    },
+    {
+      "epoch": 1.7850553505535056,
+      "grad_norm": 0.35613918691738616,
+      "learning_rate": 2.112360721654327e-05,
+      "loss": 0.1616,
+      "step": 1935
+    },
+    {
+      "epoch": 1.7859778597785978,
+      "grad_norm": 0.4013564724324399,
+      "learning_rate": 2.1097092048458172e-05,
+      "loss": 0.1742,
+      "step": 1936
+    },
+    {
+      "epoch": 1.78690036900369,
+      "grad_norm": 0.3151332094994997,
+      "learning_rate": 2.1070581379615253e-05,
+      "loss": 0.1426,
+      "step": 1937
+    },
+    {
+      "epoch": 1.7878228782287824,
+      "grad_norm": 0.3655321152134994,
+      "learning_rate": 2.1044075240575787e-05,
+      "loss": 0.1673,
+      "step": 1938
+    },
+    {
+      "epoch": 1.7887453874538746,
+      "grad_norm": 0.4302445109058553,
+      "learning_rate": 2.1017573661895838e-05,
+      "loss": 0.1696,
+      "step": 1939
+    },
+    {
+      "epoch": 1.7896678966789668,
+      "grad_norm": 0.3767103931313366,
+      "learning_rate": 2.099107667412625e-05,
+      "loss": 0.1689,
+      "step": 1940
+    },
+    {
+      "epoch": 1.7905904059040592,
+      "grad_norm": 0.35855837036481364,
+      "learning_rate": 2.0964584307812514e-05,
+      "loss": 0.16,
+      "step": 1941
+    },
+    {
+      "epoch": 1.7915129151291513,
+      "grad_norm": 0.3537270043252408,
+      "learning_rate": 2.0938096593494855e-05,
+      "loss": 0.1686,
+      "step": 1942
+    },
+    {
+      "epoch": 1.7924354243542435,
+      "grad_norm": 0.37247251035430284,
+      "learning_rate": 2.0911613561708093e-05,
+      "loss": 0.1624,
+      "step": 1943
+    },
+    {
+      "epoch": 1.793357933579336,
+      "grad_norm": 0.377132166654423,
+      "learning_rate": 2.088513524298165e-05,
+      "loss": 0.1726,
+      "step": 1944
+    },
+    {
+      "epoch": 1.7942804428044279,
+      "grad_norm": 0.3706806558600546,
+      "learning_rate": 2.0858661667839553e-05,
+      "loss": 0.1678,
+      "step": 1945
+    },
+    {
+      "epoch": 1.7952029520295203,
+      "grad_norm": 0.3819437189608818,
+      "learning_rate": 2.0832192866800316e-05,
+      "loss": 0.1696,
+      "step": 1946
+    },
+    {
+      "epoch": 1.7961254612546127,
+      "grad_norm": 0.37824618313239416,
+      "learning_rate": 2.0805728870376965e-05,
+      "loss": 0.157,
+      "step": 1947
+    },
+    {
+      "epoch": 1.7970479704797047,
+      "grad_norm": 0.40439075320316176,
+      "learning_rate": 2.077926970907701e-05,
+      "loss": 0.1767,
+      "step": 1948
+    },
+    {
+      "epoch": 1.797970479704797,
+      "grad_norm": 0.35369655441936054,
+      "learning_rate": 2.075281541340236e-05,
+      "loss": 0.1705,
+      "step": 1949
+    },
+    {
+      "epoch": 1.7988929889298892,
+      "grad_norm": 0.36162004862398006,
+      "learning_rate": 2.0726366013849313e-05,
+      "loss": 0.1741,
+      "step": 1950
+    },
+    {
+      "epoch": 1.7998154981549814,
+      "grad_norm": 0.33726035758550105,
+      "learning_rate": 2.0699921540908544e-05,
+      "loss": 0.1387,
+      "step": 1951
+    },
+    {
+      "epoch": 1.8007380073800738,
+      "grad_norm": 0.3707853938334104,
+      "learning_rate": 2.067348202506503e-05,
+      "loss": 0.1716,
+      "step": 1952
+    },
+    {
+      "epoch": 1.801660516605166,
+      "grad_norm": 0.3434117297353524,
+      "learning_rate": 2.0647047496798043e-05,
+      "loss": 0.164,
+      "step": 1953
+    },
+    {
+      "epoch": 1.8025830258302582,
+      "grad_norm": 0.3764964159000091,
+      "learning_rate": 2.062061798658111e-05,
+      "loss": 0.1606,
+      "step": 1954
+    },
+    {
+      "epoch": 1.8035055350553506,
+      "grad_norm": 0.3124660572761698,
+      "learning_rate": 2.059419352488196e-05,
+      "loss": 0.1451,
+      "step": 1955
+    },
+    {
+      "epoch": 1.8044280442804428,
+      "grad_norm": 0.33515911908453766,
+      "learning_rate": 2.0567774142162505e-05,
+      "loss": 0.1458,
+      "step": 1956
+    },
+    {
+      "epoch": 1.805350553505535,
+      "grad_norm": 0.368634745922285,
+      "learning_rate": 2.0541359868878815e-05,
+      "loss": 0.1473,
+      "step": 1957
+    },
+    {
+      "epoch": 1.8062730627306274,
+      "grad_norm": 0.3701635563982187,
+      "learning_rate": 2.0514950735481052e-05,
+      "loss": 0.1546,
+      "step": 1958
+    },
+    {
+      "epoch": 1.8071955719557196,
+      "grad_norm": 0.3651591871906541,
+      "learning_rate": 2.0488546772413462e-05,
+      "loss": 0.1565,
+      "step": 1959
+    },
+    {
+      "epoch": 1.8081180811808117,
+      "grad_norm": 0.38876955496214516,
+      "learning_rate": 2.046214801011434e-05,
+      "loss": 0.1718,
+      "step": 1960
+    },
+    {
+      "epoch": 1.8090405904059041,
+      "grad_norm": 0.36873405275038995,
+      "learning_rate": 2.0435754479015962e-05,
+      "loss": 0.1465,
+      "step": 1961
+    },
+    {
+      "epoch": 1.8099630996309963,
+      "grad_norm": 0.4358055520966555,
+      "learning_rate": 2.040936620954459e-05,
+      "loss": 0.1632,
+      "step": 1962
+    },
+    {
+      "epoch": 1.8108856088560885,
+      "grad_norm": 0.3819666578014411,
+      "learning_rate": 2.0382983232120422e-05,
+      "loss": 0.1642,
+      "step": 1963
+    },
+    {
+      "epoch": 1.811808118081181,
+      "grad_norm": 0.38078572817688405,
+      "learning_rate": 2.0356605577157552e-05,
+      "loss": 0.1498,
+      "step": 1964
+    },
+    {
+      "epoch": 1.812730627306273,
+      "grad_norm": 0.3554143708172271,
+      "learning_rate": 2.033023327506393e-05,
+      "loss": 0.1635,
+      "step": 1965
+    },
+    {
+      "epoch": 1.8136531365313653,
+      "grad_norm": 0.4513479691720528,
+      "learning_rate": 2.0303866356241347e-05,
+      "loss": 0.1577,
+      "step": 1966
+    },
+    {
+      "epoch": 1.8145756457564577,
+      "grad_norm": 0.36347844362261,
+      "learning_rate": 2.0277504851085388e-05,
+      "loss": 0.168,
+      "step": 1967
+    },
+    {
+      "epoch": 1.8154981549815496,
+      "grad_norm": 0.3971419529038244,
+      "learning_rate": 2.0251148789985374e-05,
+      "loss": 0.1634,
+      "step": 1968
+    },
+    {
+      "epoch": 1.816420664206642,
+      "grad_norm": 0.37177706325342263,
+      "learning_rate": 2.0224798203324392e-05,
+      "loss": 0.1709,
+      "step": 1969
+    },
+    {
+      "epoch": 1.8173431734317345,
+      "grad_norm": 0.4059762479232794,
+      "learning_rate": 2.019845312147919e-05,
+      "loss": 0.1576,
+      "step": 1970
+    },
+    {
+      "epoch": 1.8182656826568264,
+      "grad_norm": 0.38189613182315707,
+      "learning_rate": 2.017211357482015e-05,
+      "loss": 0.18,
+      "step": 1971
+    },
+    {
+      "epoch": 1.8191881918819188,
+      "grad_norm": 0.3413123858587887,
+      "learning_rate": 2.0145779593711338e-05,
+      "loss": 0.1733,
+      "step": 1972
+    },
+    {
+      "epoch": 1.8201107011070112,
+      "grad_norm": 0.34528499503103854,
+      "learning_rate": 2.011945120851034e-05,
+      "loss": 0.1523,
+      "step": 1973
+    },
+    {
+      "epoch": 1.8210332103321032,
+      "grad_norm": 0.3576523993441041,
+      "learning_rate": 2.0093128449568306e-05,
+      "loss": 0.1482,
+      "step": 1974
+    },
+    {
+      "epoch": 1.8219557195571956,
+      "grad_norm": 0.33455730748939116,
+      "learning_rate": 2.006681134722994e-05,
+      "loss": 0.157,
+      "step": 1975
+    },
+    {
+      "epoch": 1.8228782287822878,
+      "grad_norm": 0.3824153116538723,
+      "learning_rate": 2.0040499931833373e-05,
+      "loss": 0.1591,
+      "step": 1976
+    },
+    {
+      "epoch": 1.82380073800738,
+      "grad_norm": 0.44205789018039365,
+      "learning_rate": 2.0014194233710193e-05,
+      "loss": 0.1636,
+      "step": 1977
+    },
+    {
+      "epoch": 1.8247232472324724,
+      "grad_norm": 0.3486851989930659,
+      "learning_rate": 1.9987894283185434e-05,
+      "loss": 0.167,
+      "step": 1978
+    },
+    {
+      "epoch": 1.8256457564575646,
+      "grad_norm": 0.39737083923768923,
+      "learning_rate": 1.9961600110577456e-05,
+      "loss": 0.166,
+      "step": 1979
+    },
+    {
+      "epoch": 1.8265682656826567,
+      "grad_norm": 0.39316579916896227,
+      "learning_rate": 1.993531174619798e-05,
+      "loss": 0.157,
+      "step": 1980
+    },
+    {
+      "epoch": 1.8274907749077491,
+      "grad_norm": 0.3859558417881259,
+      "learning_rate": 1.9909029220352035e-05,
+      "loss": 0.1653,
+      "step": 1981
+    },
+    {
+      "epoch": 1.8284132841328413,
+      "grad_norm": 0.3662210447558923,
+      "learning_rate": 1.988275256333791e-05,
+      "loss": 0.1341,
+      "step": 1982
+    },
+    {
+      "epoch": 1.8293357933579335,
+      "grad_norm": 0.34513867329028414,
+      "learning_rate": 1.985648180544713e-05,
+      "loss": 0.1427,
+      "step": 1983
+    },
+    {
+      "epoch": 1.830258302583026,
+      "grad_norm": 0.38852188782944674,
+      "learning_rate": 1.9830216976964433e-05,
+      "loss": 0.167,
+      "step": 1984
+    },
+    {
+      "epoch": 1.831180811808118,
+      "grad_norm": 0.37931259597058387,
+      "learning_rate": 1.9803958108167694e-05,
+      "loss": 0.1723,
+      "step": 1985
+    },
+    {
+      "epoch": 1.8321033210332103,
+      "grad_norm": 0.3933102162504835,
+      "learning_rate": 1.9777705229327952e-05,
+      "loss": 0.1773,
+      "step": 1986
+    },
+    {
+      "epoch": 1.8330258302583027,
+      "grad_norm": 0.3214932647010383,
+      "learning_rate": 1.9751458370709313e-05,
+      "loss": 0.1565,
+      "step": 1987
+    },
+    {
+      "epoch": 1.8339483394833949,
+      "grad_norm": 0.41438090194219784,
+      "learning_rate": 1.9725217562568948e-05,
+      "loss": 0.1824,
+      "step": 1988
+    },
+    {
+      "epoch": 1.834870848708487,
+      "grad_norm": 0.3322969835242972,
+      "learning_rate": 1.969898283515707e-05,
+      "loss": 0.1488,
+      "step": 1989
+    },
+    {
+      "epoch": 1.8357933579335795,
+      "grad_norm": 0.34637329546270856,
+      "learning_rate": 1.967275421871687e-05,
+      "loss": 0.1623,
+      "step": 1990
+    },
+    {
+      "epoch": 1.8367158671586716,
+      "grad_norm": 0.35230432230838765,
+      "learning_rate": 1.9646531743484478e-05,
+      "loss": 0.1642,
+      "step": 1991
+    },
+    {
+      "epoch": 1.8376383763837638,
+      "grad_norm": 0.3796304358368079,
+      "learning_rate": 1.962031543968898e-05,
+      "loss": 0.1828,
+      "step": 1992
+    },
+    {
+      "epoch": 1.8385608856088562,
+      "grad_norm": 0.36214624107202387,
+      "learning_rate": 1.9594105337552323e-05,
+      "loss": 0.1664,
+      "step": 1993
+    },
+    {
+      "epoch": 1.8394833948339482,
+      "grad_norm": 0.37017048469642266,
+      "learning_rate": 1.9567901467289302e-05,
+      "loss": 0.1582,
+      "step": 1994
+    },
+    {
+      "epoch": 1.8404059040590406,
+      "grad_norm": 0.3789452226839255,
+      "learning_rate": 1.9541703859107545e-05,
+      "loss": 0.1632,
+      "step": 1995
+    },
+    {
+      "epoch": 1.841328413284133,
+      "grad_norm": 0.39124779668089543,
+      "learning_rate": 1.9515512543207453e-05,
+      "loss": 0.1547,
+      "step": 1996
+    },
+    {
+      "epoch": 1.842250922509225,
+      "grad_norm": 0.3694093100688777,
+      "learning_rate": 1.9489327549782168e-05,
+      "loss": 0.155,
+      "step": 1997
+    },
+    {
+      "epoch": 1.8431734317343174,
+      "grad_norm": 0.3865408499721915,
+      "learning_rate": 1.9463148909017553e-05,
+      "loss": 0.1774,
+      "step": 1998
+    },
+    {
+      "epoch": 1.8440959409594095,
+      "grad_norm": 0.3570437529844188,
+      "learning_rate": 1.9436976651092144e-05,
+      "loss": 0.1476,
+      "step": 1999
+    },
+    {
+      "epoch": 1.8450184501845017,
+      "grad_norm": 0.3423328317252141,
+      "learning_rate": 1.9410810806177104e-05,
+      "loss": 0.1484,
+      "step": 2000
+    },
+    {
+      "epoch": 1.8459409594095941,
+      "grad_norm": 0.4010672629924379,
+      "learning_rate": 1.9384651404436237e-05,
+      "loss": 0.1744,
+      "step": 2001
+    },
+    {
+      "epoch": 1.8468634686346863,
+      "grad_norm": 0.3544260848357162,
+      "learning_rate": 1.9358498476025895e-05,
+      "loss": 0.1475,
+      "step": 2002
+    },
+    {
+      "epoch": 1.8477859778597785,
+      "grad_norm": 0.38458565244615806,
+      "learning_rate": 1.9332352051094952e-05,
+      "loss": 0.175,
+      "step": 2003
+    },
+    {
+      "epoch": 1.848708487084871,
+      "grad_norm": 0.4008291354864782,
+      "learning_rate": 1.9306212159784828e-05,
+      "loss": 0.1605,
+      "step": 2004
+    },
+    {
+      "epoch": 1.849630996309963,
+      "grad_norm": 0.357938251421677,
+      "learning_rate": 1.9280078832229388e-05,
+      "loss": 0.1587,
+      "step": 2005
+    },
+    {
+      "epoch": 1.8505535055350553,
+      "grad_norm": 0.38562595776978303,
+      "learning_rate": 1.9253952098554903e-05,
+      "loss": 0.1465,
+      "step": 2006
+    },
+    {
+      "epoch": 1.8514760147601477,
+      "grad_norm": 0.3410195087635133,
+      "learning_rate": 1.9227831988880107e-05,
+      "loss": 0.1593,
+      "step": 2007
+    },
+    {
+      "epoch": 1.8523985239852399,
+      "grad_norm": 0.3682543630360419,
+      "learning_rate": 1.920171853331604e-05,
+      "loss": 0.1614,
+      "step": 2008
+    },
+    {
+      "epoch": 1.853321033210332,
+      "grad_norm": 0.33459865670495864,
+      "learning_rate": 1.9175611761966082e-05,
+      "loss": 0.1476,
+      "step": 2009
+    },
+    {
+      "epoch": 1.8542435424354244,
+      "grad_norm": 0.3327774582454587,
+      "learning_rate": 1.9149511704925942e-05,
+      "loss": 0.157,
+      "step": 2010
+    },
+    {
+      "epoch": 1.8551660516605166,
+      "grad_norm": 0.33415145348760156,
+      "learning_rate": 1.9123418392283553e-05,
+      "loss": 0.1707,
+      "step": 2011
+    },
+    {
+      "epoch": 1.8560885608856088,
+      "grad_norm": 0.34689944075535845,
+      "learning_rate": 1.9097331854119078e-05,
+      "loss": 0.1553,
+      "step": 2012
+    },
+    {
+      "epoch": 1.8570110701107012,
+      "grad_norm": 0.3515569170426396,
+      "learning_rate": 1.907125212050489e-05,
+      "loss": 0.1502,
+      "step": 2013
+    },
+    {
+      "epoch": 1.8579335793357934,
+      "grad_norm": 0.38749293290572323,
+      "learning_rate": 1.9045179221505497e-05,
+      "loss": 0.1766,
+      "step": 2014
+    },
+    {
+      "epoch": 1.8588560885608856,
+      "grad_norm": 0.3467288067329626,
+      "learning_rate": 1.901911318717753e-05,
+      "loss": 0.1575,
+      "step": 2015
+    },
+    {
+      "epoch": 1.859778597785978,
+      "grad_norm": 0.3705513922755851,
+      "learning_rate": 1.8993054047569726e-05,
+      "loss": 0.164,
+      "step": 2016
+    },
+    {
+      "epoch": 1.8607011070110702,
+      "grad_norm": 0.3948406618485568,
+      "learning_rate": 1.896700183272285e-05,
+      "loss": 0.155,
+      "step": 2017
+    },
+    {
+      "epoch": 1.8616236162361623,
+      "grad_norm": 0.3327445858175401,
+      "learning_rate": 1.8940956572669692e-05,
+      "loss": 0.1376,
+      "step": 2018
+    },
+    {
+      "epoch": 1.8625461254612548,
+      "grad_norm": 0.37760574403241454,
+      "learning_rate": 1.891491829743504e-05,
+      "loss": 0.1746,
+      "step": 2019
+    },
+    {
+      "epoch": 1.8634686346863467,
+      "grad_norm": 0.3887404163177847,
+      "learning_rate": 1.8888887037035607e-05,
+      "loss": 0.1813,
+      "step": 2020
+    },
+    {
+      "epoch": 1.8643911439114391,
+      "grad_norm": 0.39334239979851965,
+      "learning_rate": 1.8862862821480025e-05,
+      "loss": 0.1731,
+      "step": 2021
+    },
+    {
+      "epoch": 1.8653136531365315,
+      "grad_norm": 0.2978329393936861,
+      "learning_rate": 1.8836845680768815e-05,
+      "loss": 0.1383,
+      "step": 2022
+    },
+    {
+      "epoch": 1.8662361623616235,
+      "grad_norm": 0.4028454502450053,
+      "learning_rate": 1.8810835644894344e-05,
+      "loss": 0.1733,
+      "step": 2023
+    },
+    {
+      "epoch": 1.867158671586716,
+      "grad_norm": 0.331584454188422,
+      "learning_rate": 1.8784832743840757e-05,
+      "loss": 0.1498,
+      "step": 2024
+    },
+    {
+      "epoch": 1.868081180811808,
+      "grad_norm": 0.4106832662420921,
+      "learning_rate": 1.8758837007584023e-05,
+      "loss": 0.1796,
+      "step": 2025
+    },
+    {
+      "epoch": 1.8690036900369003,
+      "grad_norm": 0.3764091889563927,
+      "learning_rate": 1.8732848466091818e-05,
+      "loss": 0.1673,
+      "step": 2026
+    },
+    {
+      "epoch": 1.8699261992619927,
+      "grad_norm": 0.37190067044130726,
+      "learning_rate": 1.870686714932352e-05,
+      "loss": 0.1492,
+      "step": 2027
+    },
+    {
+      "epoch": 1.8708487084870848,
+      "grad_norm": 0.3441146253858929,
+      "learning_rate": 1.8680893087230204e-05,
+      "loss": 0.1529,
+      "step": 2028
+    },
+    {
+      "epoch": 1.871771217712177,
+      "grad_norm": 0.4040439209904815,
+      "learning_rate": 1.8654926309754566e-05,
+      "loss": 0.162,
+      "step": 2029
+    },
+    {
+      "epoch": 1.8726937269372694,
+      "grad_norm": 0.3931622956402532,
+      "learning_rate": 1.8628966846830907e-05,
+      "loss": 0.1676,
+      "step": 2030
+    },
+    {
+      "epoch": 1.8736162361623616,
+      "grad_norm": 0.3193108144849364,
+      "learning_rate": 1.8603014728385095e-05,
+      "loss": 0.169,
+      "step": 2031
+    },
+    {
+      "epoch": 1.8745387453874538,
+      "grad_norm": 0.3441639223339678,
+      "learning_rate": 1.8577069984334522e-05,
+      "loss": 0.1302,
+      "step": 2032
+    },
+    {
+      "epoch": 1.8754612546125462,
+      "grad_norm": 0.45250380780450933,
+      "learning_rate": 1.8551132644588102e-05,
+      "loss": 0.1702,
+      "step": 2033
+    },
+    {
+      "epoch": 1.8763837638376384,
+      "grad_norm": 0.3210029209059365,
+      "learning_rate": 1.8525202739046196e-05,
+      "loss": 0.1396,
+      "step": 2034
+    },
+    {
+      "epoch": 1.8773062730627306,
+      "grad_norm": 0.3711136600679912,
+      "learning_rate": 1.8499280297600594e-05,
+      "loss": 0.1622,
+      "step": 2035
+    },
+    {
+      "epoch": 1.878228782287823,
+      "grad_norm": 0.43656379806291445,
+      "learning_rate": 1.84733653501345e-05,
+      "loss": 0.1755,
+      "step": 2036
+    },
+    {
+      "epoch": 1.8791512915129152,
+      "grad_norm": 0.40124015959206116,
+      "learning_rate": 1.8447457926522454e-05,
+      "loss": 0.1593,
+      "step": 2037
+    },
+    {
+      "epoch": 1.8800738007380073,
+      "grad_norm": 0.3979470533271776,
+      "learning_rate": 1.8421558056630324e-05,
+      "loss": 0.1666,
+      "step": 2038
+    },
+    {
+      "epoch": 1.8809963099630997,
+      "grad_norm": 0.3631975944244946,
+      "learning_rate": 1.8395665770315298e-05,
+      "loss": 0.1678,
+      "step": 2039
+    },
+    {
+      "epoch": 1.881918819188192,
+      "grad_norm": 0.3493776189055461,
+      "learning_rate": 1.836978109742581e-05,
+      "loss": 0.1633,
+      "step": 2040
+    },
+    {
+      "epoch": 1.882841328413284,
+      "grad_norm": 0.36778378045879667,
+      "learning_rate": 1.8343904067801477e-05,
+      "loss": 0.1614,
+      "step": 2041
+    },
+    {
+      "epoch": 1.8837638376383765,
+      "grad_norm": 0.3748701187810373,
+      "learning_rate": 1.831803471127318e-05,
+      "loss": 0.1619,
+      "step": 2042
+    },
+    {
+      "epoch": 1.8846863468634685,
+      "grad_norm": 0.35954905741291965,
+      "learning_rate": 1.829217305766289e-05,
+      "loss": 0.171,
+      "step": 2043
+    },
+    {
+      "epoch": 1.8856088560885609,
+      "grad_norm": 0.3319221458297196,
+      "learning_rate": 1.8266319136783712e-05,
+      "loss": 0.1585,
+      "step": 2044
+    },
+    {
+      "epoch": 1.8865313653136533,
+      "grad_norm": 0.3794837799983876,
+      "learning_rate": 1.8240472978439883e-05,
+      "loss": 0.1452,
+      "step": 2045
+    },
+    {
+      "epoch": 1.8874538745387452,
+      "grad_norm": 0.3987585374780179,
+      "learning_rate": 1.8214634612426623e-05,
+      "loss": 0.1737,
+      "step": 2046
+    },
+    {
+      "epoch": 1.8883763837638377,
+      "grad_norm": 0.3683407966053583,
+      "learning_rate": 1.8188804068530206e-05,
+      "loss": 0.1576,
+      "step": 2047
+    },
+    {
+      "epoch": 1.8892988929889298,
+      "grad_norm": 0.34263755995174733,
+      "learning_rate": 1.8162981376527894e-05,
+      "loss": 0.1696,
+      "step": 2048
+    },
+    {
+      "epoch": 1.890221402214022,
+      "grad_norm": 0.314212820389449,
+      "learning_rate": 1.813716656618788e-05,
+      "loss": 0.1475,
+      "step": 2049
+    },
+    {
+      "epoch": 1.8911439114391144,
+      "grad_norm": 0.3829724497296785,
+      "learning_rate": 1.8111359667269275e-05,
+      "loss": 0.1592,
+      "step": 2050
+    },
+    {
+      "epoch": 1.8920664206642066,
+      "grad_norm": 0.3266759250780846,
+      "learning_rate": 1.8085560709522077e-05,
+      "loss": 0.1539,
+      "step": 2051
+    },
+    {
+      "epoch": 1.8929889298892988,
+      "grad_norm": 0.3628235399383464,
+      "learning_rate": 1.805976972268713e-05,
+      "loss": 0.1758,
+      "step": 2052
+    },
+    {
+      "epoch": 1.8939114391143912,
+      "grad_norm": 0.37032064120291297,
+      "learning_rate": 1.8033986736496078e-05,
+      "loss": 0.1758,
+      "step": 2053
+    },
+    {
+      "epoch": 1.8948339483394834,
+      "grad_norm": 0.3390476198553966,
+      "learning_rate": 1.8008211780671353e-05,
+      "loss": 0.1558,
+      "step": 2054
+    },
+    {
+      "epoch": 1.8957564575645756,
+      "grad_norm": 0.39703738351615603,
+      "learning_rate": 1.798244488492612e-05,
+      "loss": 0.1701,
+      "step": 2055
+    },
+    {
+      "epoch": 1.896678966789668,
+      "grad_norm": 0.35851108811570903,
+      "learning_rate": 1.795668607896426e-05,
+      "loss": 0.1706,
+      "step": 2056
+    },
+    {
+      "epoch": 1.8976014760147601,
+      "grad_norm": 0.3431270651037541,
+      "learning_rate": 1.7930935392480326e-05,
+      "loss": 0.1631,
+      "step": 2057
+    },
+    {
+      "epoch": 1.8985239852398523,
+      "grad_norm": 0.3416418505976903,
+      "learning_rate": 1.7905192855159514e-05,
+      "loss": 0.1461,
+      "step": 2058
+    },
+    {
+      "epoch": 1.8994464944649447,
+      "grad_norm": 0.32644775615725424,
+      "learning_rate": 1.7879458496677615e-05,
+      "loss": 0.134,
+      "step": 2059
+    },
+    {
+      "epoch": 1.900369003690037,
+      "grad_norm": 0.38332095835052227,
+      "learning_rate": 1.7853732346701003e-05,
+      "loss": 0.1551,
+      "step": 2060
+    },
+    {
+      "epoch": 1.901291512915129,
+      "grad_norm": 0.39235068169305326,
+      "learning_rate": 1.7828014434886588e-05,
+      "loss": 0.1534,
+      "step": 2061
+    },
+    {
+      "epoch": 1.9022140221402215,
+      "grad_norm": 0.34317444492086,
+      "learning_rate": 1.7802304790881773e-05,
+      "loss": 0.1457,
+      "step": 2062
+    },
+    {
+      "epoch": 1.9031365313653137,
+      "grad_norm": 0.4447190660701878,
+      "learning_rate": 1.7776603444324445e-05,
+      "loss": 0.1782,
+      "step": 2063
+    },
+    {
+      "epoch": 1.9040590405904059,
+      "grad_norm": 0.36598747335495846,
+      "learning_rate": 1.775091042484292e-05,
+      "loss": 0.1503,
+      "step": 2064
+    },
+    {
+      "epoch": 1.9049815498154983,
+      "grad_norm": 0.3369529655069667,
+      "learning_rate": 1.7725225762055887e-05,
+      "loss": 0.1452,
+      "step": 2065
+    },
+    {
+      "epoch": 1.9059040590405905,
+      "grad_norm": 0.3162265674227008,
+      "learning_rate": 1.7699549485572465e-05,
+      "loss": 0.146,
+      "step": 2066
+    },
+    {
+      "epoch": 1.9068265682656826,
+      "grad_norm": 0.41214926742766195,
+      "learning_rate": 1.7673881624992047e-05,
+      "loss": 0.1747,
+      "step": 2067
+    },
+    {
+      "epoch": 1.907749077490775,
+      "grad_norm": 0.38817916498204885,
+      "learning_rate": 1.7648222209904338e-05,
+      "loss": 0.1695,
+      "step": 2068
+    },
+    {
+      "epoch": 1.908671586715867,
+      "grad_norm": 0.3704083615912522,
+      "learning_rate": 1.7622571269889326e-05,
+      "loss": 0.17,
+      "step": 2069
+    },
+    {
+      "epoch": 1.9095940959409594,
+      "grad_norm": 0.4032782003700155,
+      "learning_rate": 1.759692883451721e-05,
+      "loss": 0.1721,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9105166051660518,
+      "grad_norm": 0.3698213808043485,
+      "learning_rate": 1.75712949333484e-05,
+      "loss": 0.1713,
+      "step": 2071
+    },
+    {
+      "epoch": 1.9114391143911438,
+      "grad_norm": 0.3370359745013529,
+      "learning_rate": 1.754566959593346e-05,
+      "loss": 0.1542,
+      "step": 2072
+    },
+    {
+      "epoch": 1.9123616236162362,
+      "grad_norm": 0.3643578033579248,
+      "learning_rate": 1.752005285181306e-05,
+      "loss": 0.1683,
+      "step": 2073
+    },
+    {
+      "epoch": 1.9132841328413284,
+      "grad_norm": 0.3662165203448614,
+      "learning_rate": 1.7494444730518012e-05,
+      "loss": 0.1721,
+      "step": 2074
+    },
+    {
+      "epoch": 1.9142066420664205,
+      "grad_norm": 0.33697578821573065,
+      "learning_rate": 1.746884526156915e-05,
+      "loss": 0.1353,
+      "step": 2075
+    },
+    {
+      "epoch": 1.915129151291513,
+      "grad_norm": 0.3702583026292005,
+      "learning_rate": 1.7443254474477327e-05,
+      "loss": 0.156,
+      "step": 2076
+    },
+    {
+      "epoch": 1.9160516605166051,
+      "grad_norm": 0.36325214626895036,
+      "learning_rate": 1.741767239874344e-05,
+      "loss": 0.1677,
+      "step": 2077
+    },
+    {
+      "epoch": 1.9169741697416973,
+      "grad_norm": 0.3693694666175546,
+      "learning_rate": 1.7392099063858284e-05,
+      "loss": 0.1852,
+      "step": 2078
+    },
+    {
+      "epoch": 1.9178966789667897,
+      "grad_norm": 0.3714584247062507,
+      "learning_rate": 1.7366534499302595e-05,
+      "loss": 0.1513,
+      "step": 2079
+    },
+    {
+      "epoch": 1.918819188191882,
+      "grad_norm": 0.31958997192401395,
+      "learning_rate": 1.7340978734547035e-05,
+      "loss": 0.1472,
+      "step": 2080
+    },
+    {
+      "epoch": 1.919741697416974,
+      "grad_norm": 0.3487306312338888,
+      "learning_rate": 1.7315431799052066e-05,
+      "loss": 0.1637,
+      "step": 2081
+    },
+    {
+      "epoch": 1.9206642066420665,
+      "grad_norm": 0.34585110546026254,
+      "learning_rate": 1.728989372226801e-05,
+      "loss": 0.1511,
+      "step": 2082
+    },
+    {
+      "epoch": 1.9215867158671587,
+      "grad_norm": 0.3865905336151693,
+      "learning_rate": 1.7264364533634956e-05,
+      "loss": 0.1665,
+      "step": 2083
+    },
+    {
+      "epoch": 1.9225092250922509,
+      "grad_norm": 0.3523667141912192,
+      "learning_rate": 1.723884426258277e-05,
+      "loss": 0.154,
+      "step": 2084
+    },
+    {
+      "epoch": 1.9234317343173433,
+      "grad_norm": 0.4303298108479506,
+      "learning_rate": 1.7213332938531012e-05,
+      "loss": 0.1641,
+      "step": 2085
+    },
+    {
+      "epoch": 1.9243542435424354,
+      "grad_norm": 0.3534015365860965,
+      "learning_rate": 1.718783059088894e-05,
+      "loss": 0.1403,
+      "step": 2086
+    },
+    {
+      "epoch": 1.9252767527675276,
+      "grad_norm": 0.3673903376438012,
+      "learning_rate": 1.7162337249055477e-05,
+      "loss": 0.1579,
+      "step": 2087
+    },
+    {
+      "epoch": 1.92619926199262,
+      "grad_norm": 0.342962694727826,
+      "learning_rate": 1.7136852942419127e-05,
+      "loss": 0.1631,
+      "step": 2088
+    },
+    {
+      "epoch": 1.9271217712177122,
+      "grad_norm": 0.3753024966381805,
+      "learning_rate": 1.7111377700358022e-05,
+      "loss": 0.1472,
+      "step": 2089
+    },
+    {
+      "epoch": 1.9280442804428044,
+      "grad_norm": 0.3364763848904482,
+      "learning_rate": 1.708591155223982e-05,
+      "loss": 0.1518,
+      "step": 2090
+    },
+    {
+      "epoch": 1.9289667896678968,
+      "grad_norm": 0.3866136533660062,
+      "learning_rate": 1.7060454527421688e-05,
+      "loss": 0.1713,
+      "step": 2091
+    },
+    {
+      "epoch": 1.9298892988929888,
+      "grad_norm": 0.4147150763053682,
+      "learning_rate": 1.7035006655250304e-05,
+      "loss": 0.1563,
+      "step": 2092
+    },
+    {
+      "epoch": 1.9308118081180812,
+      "grad_norm": 0.3756944304039137,
+      "learning_rate": 1.7009567965061774e-05,
+      "loss": 0.1565,
+      "step": 2093
+    },
+    {
+      "epoch": 1.9317343173431736,
+      "grad_norm": 0.3807925730918648,
+      "learning_rate": 1.698413848618161e-05,
+      "loss": 0.1593,
+      "step": 2094
+    },
+    {
+      "epoch": 1.9326568265682655,
+      "grad_norm": 0.38492062922264886,
+      "learning_rate": 1.6958718247924745e-05,
+      "loss": 0.1593,
+      "step": 2095
+    },
+    {
+      "epoch": 1.933579335793358,
+      "grad_norm": 0.3772907678044971,
+      "learning_rate": 1.6933307279595413e-05,
+      "loss": 0.1621,
+      "step": 2096
+    },
+    {
+      "epoch": 1.9345018450184504,
+      "grad_norm": 0.4084042020189238,
+      "learning_rate": 1.6907905610487184e-05,
+      "loss": 0.1708,
+      "step": 2097
+    },
+    {
+      "epoch": 1.9354243542435423,
+      "grad_norm": 0.38837729340567867,
+      "learning_rate": 1.6882513269882917e-05,
+      "loss": 0.1686,
+      "step": 2098
+    },
+    {
+      "epoch": 1.9363468634686347,
+      "grad_norm": 0.38721867620945977,
+      "learning_rate": 1.6857130287054702e-05,
+      "loss": 0.1726,
+      "step": 2099
+    },
+    {
+      "epoch": 1.937269372693727,
+      "grad_norm": 0.41436134979934947,
+      "learning_rate": 1.683175669126383e-05,
+      "loss": 0.161,
+      "step": 2100
+    },
+    {
+      "epoch": 1.938191881918819,
+      "grad_norm": 0.36101101262609697,
+      "learning_rate": 1.6806392511760803e-05,
+      "loss": 0.1426,
+      "step": 2101
+    },
+    {
+      "epoch": 1.9391143911439115,
+      "grad_norm": 0.36419907430338827,
+      "learning_rate": 1.678103777778526e-05,
+      "loss": 0.1574,
+      "step": 2102
+    },
+    {
+      "epoch": 1.9400369003690037,
+      "grad_norm": 0.2908242224470308,
+      "learning_rate": 1.6755692518565914e-05,
+      "loss": 0.1223,
+      "step": 2103
+    },
+    {
+      "epoch": 1.9409594095940959,
+      "grad_norm": 0.35436960934575523,
+      "learning_rate": 1.6730356763320615e-05,
+      "loss": 0.1472,
+      "step": 2104
+    },
+    {
+      "epoch": 1.9418819188191883,
+      "grad_norm": 0.3760954645026784,
+      "learning_rate": 1.670503054125621e-05,
+      "loss": 0.1505,
+      "step": 2105
+    },
+    {
+      "epoch": 1.9428044280442804,
+      "grad_norm": 0.3489617565462921,
+      "learning_rate": 1.667971388156856e-05,
+      "loss": 0.1569,
+      "step": 2106
+    },
+    {
+      "epoch": 1.9437269372693726,
+      "grad_norm": 0.27280654136664284,
+      "learning_rate": 1.6654406813442545e-05,
+      "loss": 0.1325,
+      "step": 2107
+    },
+    {
+      "epoch": 1.944649446494465,
+      "grad_norm": 0.3622553045951628,
+      "learning_rate": 1.662910936605194e-05,
+      "loss": 0.1629,
+      "step": 2108
+    },
+    {
+      "epoch": 1.9455719557195572,
+      "grad_norm": 0.336857190707021,
+      "learning_rate": 1.6603821568559437e-05,
+      "loss": 0.1441,
+      "step": 2109
+    },
+    {
+      "epoch": 1.9464944649446494,
+      "grad_norm": 0.40529479124812,
+      "learning_rate": 1.657854345011664e-05,
+      "loss": 0.164,
+      "step": 2110
+    },
+    {
+      "epoch": 1.9474169741697418,
+      "grad_norm": 0.3788435120348254,
+      "learning_rate": 1.655327503986395e-05,
+      "loss": 0.1569,
+      "step": 2111
+    },
+    {
+      "epoch": 1.948339483394834,
+      "grad_norm": 0.34957022347909955,
+      "learning_rate": 1.6528016366930592e-05,
+      "loss": 0.1688,
+      "step": 2112
+    },
+    {
+      "epoch": 1.9492619926199262,
+      "grad_norm": 0.347731366608516,
+      "learning_rate": 1.6502767460434588e-05,
+      "loss": 0.1534,
+      "step": 2113
+    },
+    {
+      "epoch": 1.9501845018450186,
+      "grad_norm": 0.3728756800631118,
+      "learning_rate": 1.6477528349482656e-05,
+      "loss": 0.1845,
+      "step": 2114
+    },
+    {
+      "epoch": 1.9511070110701108,
+      "grad_norm": 0.3563021187541003,
+      "learning_rate": 1.6452299063170283e-05,
+      "loss": 0.1581,
+      "step": 2115
+    },
+    {
+      "epoch": 1.952029520295203,
+      "grad_norm": 0.3505466294901255,
+      "learning_rate": 1.6427079630581572e-05,
+      "loss": 0.1655,
+      "step": 2116
+    },
+    {
+      "epoch": 1.9529520295202953,
+      "grad_norm": 0.3101500827340247,
+      "learning_rate": 1.6401870080789282e-05,
+      "loss": 0.1342,
+      "step": 2117
+    },
+    {
+      "epoch": 1.9538745387453873,
+      "grad_norm": 0.39658480605922347,
+      "learning_rate": 1.6376670442854815e-05,
+      "loss": 0.1574,
+      "step": 2118
+    },
+    {
+      "epoch": 1.9547970479704797,
+      "grad_norm": 0.33081154898048376,
+      "learning_rate": 1.63514807458281e-05,
+      "loss": 0.1487,
+      "step": 2119
+    },
+    {
+      "epoch": 1.9557195571955721,
+      "grad_norm": 0.35135230194523337,
+      "learning_rate": 1.6326301018747623e-05,
+      "loss": 0.1507,
+      "step": 2120
+    },
+    {
+      "epoch": 1.956642066420664,
+      "grad_norm": 0.34425650471796426,
+      "learning_rate": 1.6301131290640393e-05,
+      "loss": 0.1617,
+      "step": 2121
+    },
+    {
+      "epoch": 1.9575645756457565,
+      "grad_norm": 0.34768864572071606,
+      "learning_rate": 1.627597159052187e-05,
+      "loss": 0.168,
+      "step": 2122
+    },
+    {
+      "epoch": 1.9584870848708487,
+      "grad_norm": 0.3781965255314279,
+      "learning_rate": 1.6250821947395954e-05,
+      "loss": 0.1748,
+      "step": 2123
+    },
+    {
+      "epoch": 1.9594095940959408,
+      "grad_norm": 0.42719434703034126,
+      "learning_rate": 1.622568239025498e-05,
+      "loss": 0.1494,
+      "step": 2124
+    },
+    {
+      "epoch": 1.9603321033210332,
+      "grad_norm": 0.3241759352541061,
+      "learning_rate": 1.620055294807962e-05,
+      "loss": 0.1507,
+      "step": 2125
+    },
+    {
+      "epoch": 1.9612546125461254,
+      "grad_norm": 0.3711367189573145,
+      "learning_rate": 1.61754336498389e-05,
+      "loss": 0.1778,
+      "step": 2126
+    },
+    {
+      "epoch": 1.9621771217712176,
+      "grad_norm": 0.40511893012609534,
+      "learning_rate": 1.615032452449017e-05,
+      "loss": 0.1554,
+      "step": 2127
+    },
+    {
+      "epoch": 1.96309963099631,
+      "grad_norm": 0.3651020044924753,
+      "learning_rate": 1.6125225600979015e-05,
+      "loss": 0.1545,
+      "step": 2128
+    },
+    {
+      "epoch": 1.9640221402214022,
+      "grad_norm": 0.3389783590681625,
+      "learning_rate": 1.6100136908239284e-05,
+      "loss": 0.1523,
+      "step": 2129
+    },
+    {
+      "epoch": 1.9649446494464944,
+      "grad_norm": 0.32787767426737324,
+      "learning_rate": 1.6075058475193045e-05,
+      "loss": 0.1434,
+      "step": 2130
+    },
+    {
+      "epoch": 1.9658671586715868,
+      "grad_norm": 0.35453868294421087,
+      "learning_rate": 1.604999033075051e-05,
+      "loss": 0.1705,
+      "step": 2131
+    },
+    {
+      "epoch": 1.966789667896679,
+      "grad_norm": 0.3408667966426956,
+      "learning_rate": 1.602493250381003e-05,
+      "loss": 0.1408,
+      "step": 2132
+    },
+    {
+      "epoch": 1.9677121771217712,
+      "grad_norm": 0.37629081425467487,
+      "learning_rate": 1.59998850232581e-05,
+      "loss": 0.1574,
+      "step": 2133
+    },
+    {
+      "epoch": 1.9686346863468636,
+      "grad_norm": 0.31639473669869117,
+      "learning_rate": 1.5974847917969253e-05,
+      "loss": 0.1344,
+      "step": 2134
+    },
+    {
+      "epoch": 1.9695571955719557,
+      "grad_norm": 0.3894565590429641,
+      "learning_rate": 1.594982121680605e-05,
+      "loss": 0.1562,
+      "step": 2135
+    },
+    {
+      "epoch": 1.970479704797048,
+      "grad_norm": 0.33721420188956064,
+      "learning_rate": 1.592480494861911e-05,
+      "loss": 0.1454,
+      "step": 2136
+    },
+    {
+      "epoch": 1.9714022140221403,
+      "grad_norm": 0.3658903476299478,
+      "learning_rate": 1.5899799142246987e-05,
+      "loss": 0.161,
+      "step": 2137
+    },
+    {
+      "epoch": 1.9723247232472325,
+      "grad_norm": 0.36192956235609736,
+      "learning_rate": 1.5874803826516153e-05,
+      "loss": 0.172,
+      "step": 2138
+    },
+    {
+      "epoch": 1.9732472324723247,
+      "grad_norm": 0.35690008919131494,
+      "learning_rate": 1.584981903024106e-05,
+      "loss": 0.1569,
+      "step": 2139
+    },
+    {
+      "epoch": 1.974169741697417,
+      "grad_norm": 0.32918971311261697,
+      "learning_rate": 1.5824844782223954e-05,
+      "loss": 0.1541,
+      "step": 2140
+    },
+    {
+      "epoch": 1.975092250922509,
+      "grad_norm": 0.3585597257101039,
+      "learning_rate": 1.579988111125496e-05,
+      "loss": 0.1414,
+      "step": 2141
+    },
+    {
+      "epoch": 1.9760147601476015,
+      "grad_norm": 0.342128447426705,
+      "learning_rate": 1.5774928046112027e-05,
+      "loss": 0.1466,
+      "step": 2142
+    },
+    {
+      "epoch": 1.9769372693726939,
+      "grad_norm": 0.3672701644773694,
+      "learning_rate": 1.5749985615560837e-05,
+      "loss": 0.14,
+      "step": 2143
+    },
+    {
+      "epoch": 1.9778597785977858,
+      "grad_norm": 0.3570713683549541,
+      "learning_rate": 1.572505384835482e-05,
+      "loss": 0.1455,
+      "step": 2144
+    },
+    {
+      "epoch": 1.9787822878228782,
+      "grad_norm": 0.32429396370851155,
+      "learning_rate": 1.570013277323516e-05,
+      "loss": 0.1584,
+      "step": 2145
+    },
+    {
+      "epoch": 1.9797047970479706,
+      "grad_norm": 0.3506385107195562,
+      "learning_rate": 1.5675222418930653e-05,
+      "loss": 0.1341,
+      "step": 2146
+    },
+    {
+      "epoch": 1.9806273062730626,
+      "grad_norm": 0.37317718273227996,
+      "learning_rate": 1.5650322814157764e-05,
+      "loss": 0.1616,
+      "step": 2147
+    },
+    {
+      "epoch": 1.981549815498155,
+      "grad_norm": 0.353562182142919,
+      "learning_rate": 1.5625433987620577e-05,
+      "loss": 0.1486,
+      "step": 2148
+    },
+    {
+      "epoch": 1.9824723247232472,
+      "grad_norm": 0.42413792939965583,
+      "learning_rate": 1.5600555968010734e-05,
+      "loss": 0.1726,
+      "step": 2149
+    },
+    {
+      "epoch": 1.9833948339483394,
+      "grad_norm": 0.3622606610836388,
+      "learning_rate": 1.557568878400742e-05,
+      "loss": 0.1537,
+      "step": 2150
+    },
+    {
+      "epoch": 1.9843173431734318,
+      "grad_norm": 0.3572809406638548,
+      "learning_rate": 1.555083246427734e-05,
+      "loss": 0.1741,
+      "step": 2151
+    },
+    {
+      "epoch": 1.985239852398524,
+      "grad_norm": 0.33778560462753165,
+      "learning_rate": 1.5525987037474667e-05,
+      "loss": 0.1312,
+      "step": 2152
+    },
+    {
+      "epoch": 1.9861623616236161,
+      "grad_norm": 0.3740675742270338,
+      "learning_rate": 1.5501152532241005e-05,
+      "loss": 0.1542,
+      "step": 2153
+    },
+    {
+      "epoch": 1.9870848708487086,
+      "grad_norm": 0.3119493809349143,
+      "learning_rate": 1.5476328977205397e-05,
+      "loss": 0.1424,
+      "step": 2154
+    },
+    {
+      "epoch": 1.9880073800738007,
+      "grad_norm": 0.38111403421490336,
+      "learning_rate": 1.5451516400984235e-05,
+      "loss": 0.1573,
+      "step": 2155
+    },
+    {
+      "epoch": 1.988929889298893,
+      "grad_norm": 0.3224861320468959,
+      "learning_rate": 1.5426714832181262e-05,
+      "loss": 0.1412,
+      "step": 2156
+    },
+    {
+      "epoch": 1.9898523985239853,
+      "grad_norm": 0.3376414274010515,
+      "learning_rate": 1.540192429938755e-05,
+      "loss": 0.1604,
+      "step": 2157
+    },
+    {
+      "epoch": 1.9907749077490775,
+      "grad_norm": 0.36182822271175735,
+      "learning_rate": 1.5377144831181416e-05,
+      "loss": 0.1672,
+      "step": 2158
+    },
+    {
+      "epoch": 1.9916974169741697,
+      "grad_norm": 0.3688889212889081,
+      "learning_rate": 1.535237645612846e-05,
+      "loss": 0.1629,
+      "step": 2159
+    },
+    {
+      "epoch": 1.992619926199262,
+      "grad_norm": 0.35371931564594805,
+      "learning_rate": 1.5327619202781457e-05,
+      "loss": 0.1417,
+      "step": 2160
+    },
+    {
+      "epoch": 1.9935424354243543,
+      "grad_norm": 0.3872381312100595,
+      "learning_rate": 1.5302873099680377e-05,
+      "loss": 0.1533,
+      "step": 2161
+    },
+    {
+      "epoch": 1.9944649446494465,
+      "grad_norm": 0.34813821973923315,
+      "learning_rate": 1.5278138175352353e-05,
+      "loss": 0.1609,
+      "step": 2162
+    },
+    {
+      "epoch": 1.9953874538745389,
+      "grad_norm": 0.340179721949598,
+      "learning_rate": 1.52534144583116e-05,
+      "loss": 0.1327,
+      "step": 2163
+    },
+    {
+      "epoch": 1.996309963099631,
+      "grad_norm": 0.4044504632167899,
+      "learning_rate": 1.5228701977059428e-05,
+      "loss": 0.1735,
+      "step": 2164
+    },
+    {
+      "epoch": 1.9972324723247232,
+      "grad_norm": 0.3516746142082308,
+      "learning_rate": 1.5204000760084206e-05,
+      "loss": 0.1633,
+      "step": 2165
+    },
+    {
+      "epoch": 1.9981549815498156,
+      "grad_norm": 0.35824635814324857,
+      "learning_rate": 1.5179310835861299e-05,
+      "loss": 0.1534,
+      "step": 2166
+    },
+    {
+      "epoch": 1.9990774907749076,
+      "grad_norm": 0.3345091202124253,
+      "learning_rate": 1.5154632232853055e-05,
+      "loss": 0.148,
+      "step": 2167
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.30797574850596654,
+      "learning_rate": 1.5129964979508792e-05,
+      "loss": 0.1199,
+      "step": 2168
+    },
+    {
+      "epoch": 2.0009225092250924,
+      "grad_norm": 0.2976360537697981,
+      "learning_rate": 1.5105309104264725e-05,
+      "loss": 0.0703,
+      "step": 2169
+    },
+    {
+      "epoch": 2.0018450184501844,
+      "grad_norm": 0.3198873467806432,
+      "learning_rate": 1.5080664635543934e-05,
+      "loss": 0.0854,
+      "step": 2170
+    },
+    {
+      "epoch": 2.0027675276752768,
+      "grad_norm": 0.31263105502287075,
+      "learning_rate": 1.5056031601756405e-05,
+      "loss": 0.069,
+      "step": 2171
+    },
+    {
+      "epoch": 2.003690036900369,
+      "grad_norm": 0.3083167766286197,
+      "learning_rate": 1.5031410031298898e-05,
+      "loss": 0.0798,
+      "step": 2172
+    },
+    {
+      "epoch": 2.004612546125461,
+      "grad_norm": 0.3087797367319195,
+      "learning_rate": 1.5006799952554954e-05,
+      "loss": 0.0674,
+      "step": 2173
+    },
+    {
+      "epoch": 2.0055350553505535,
+      "grad_norm": 0.38139646201681016,
+      "learning_rate": 1.4982201393894906e-05,
+      "loss": 0.0761,
+      "step": 2174
+    },
+    {
+      "epoch": 2.006457564575646,
+      "grad_norm": 0.46131637088146976,
+      "learning_rate": 1.495761438367577e-05,
+      "loss": 0.0772,
+      "step": 2175
+    },
+    {
+      "epoch": 2.007380073800738,
+      "grad_norm": 0.46230756809420753,
+      "learning_rate": 1.4933038950241252e-05,
+      "loss": 0.0656,
+      "step": 2176
+    },
+    {
+      "epoch": 2.0083025830258303,
+      "grad_norm": 0.45122782926874755,
+      "learning_rate": 1.4908475121921744e-05,
+      "loss": 0.0676,
+      "step": 2177
+    },
+    {
+      "epoch": 2.0092250922509227,
+      "grad_norm": 0.41925272184214457,
+      "learning_rate": 1.4883922927034222e-05,
+      "loss": 0.0631,
+      "step": 2178
+    },
+    {
+      "epoch": 2.0101476014760147,
+      "grad_norm": 0.3936049787261394,
+      "learning_rate": 1.485938239388227e-05,
+      "loss": 0.062,
+      "step": 2179
+    },
+    {
+      "epoch": 2.011070110701107,
+      "grad_norm": 0.44263563462972005,
+      "learning_rate": 1.4834853550756029e-05,
+      "loss": 0.0793,
+      "step": 2180
+    },
+    {
+      "epoch": 2.011992619926199,
+      "grad_norm": 0.4146007752592815,
+      "learning_rate": 1.4810336425932155e-05,
+      "loss": 0.0707,
+      "step": 2181
+    },
+    {
+      "epoch": 2.0129151291512914,
+      "grad_norm": 0.3902394436660754,
+      "learning_rate": 1.4785831047673799e-05,
+      "loss": 0.0715,
+      "step": 2182
+    },
+    {
+      "epoch": 2.013837638376384,
+      "grad_norm": 0.4844314150823258,
+      "learning_rate": 1.4761337444230583e-05,
+      "loss": 0.0639,
+      "step": 2183
+    },
+    {
+      "epoch": 2.014760147601476,
+      "grad_norm": 0.3584547751069646,
+      "learning_rate": 1.4736855643838532e-05,
+      "loss": 0.0646,
+      "step": 2184
+    },
+    {
+      "epoch": 2.015682656826568,
+      "grad_norm": 0.38296790522159585,
+      "learning_rate": 1.471238567472008e-05,
+      "loss": 0.0754,
+      "step": 2185
+    },
+    {
+      "epoch": 2.0166051660516606,
+      "grad_norm": 0.3735311076171752,
+      "learning_rate": 1.4687927565084022e-05,
+      "loss": 0.074,
+      "step": 2186
+    },
+    {
+      "epoch": 2.0175276752767526,
+      "grad_norm": 0.3751760443336041,
+      "learning_rate": 1.4663481343125477e-05,
+      "loss": 0.0742,
+      "step": 2187
+    },
+    {
+      "epoch": 2.018450184501845,
+      "grad_norm": 0.31588170716651065,
+      "learning_rate": 1.4639047037025855e-05,
+      "loss": 0.0583,
+      "step": 2188
+    },
+    {
+      "epoch": 2.0193726937269374,
+      "grad_norm": 0.3515839610072372,
+      "learning_rate": 1.4614624674952842e-05,
+      "loss": 0.0618,
+      "step": 2189
+    },
+    {
+      "epoch": 2.0202952029520294,
+      "grad_norm": 0.3861408374210167,
+      "learning_rate": 1.4590214285060349e-05,
+      "loss": 0.0732,
+      "step": 2190
+    },
+    {
+      "epoch": 2.0212177121771218,
+      "grad_norm": 0.3927048337607836,
+      "learning_rate": 1.4565815895488476e-05,
+      "loss": 0.0608,
+      "step": 2191
+    },
+    {
+      "epoch": 2.022140221402214,
+      "grad_norm": 0.3531098164929209,
+      "learning_rate": 1.4541429534363515e-05,
+      "loss": 0.0579,
+      "step": 2192
+    },
+    {
+      "epoch": 2.023062730627306,
+      "grad_norm": 0.45167105898061277,
+      "learning_rate": 1.4517055229797857e-05,
+      "loss": 0.0751,
+      "step": 2193
+    },
+    {
+      "epoch": 2.0239852398523985,
+      "grad_norm": 0.4627100326669043,
+      "learning_rate": 1.4492693009890018e-05,
+      "loss": 0.0748,
+      "step": 2194
+    },
+    {
+      "epoch": 2.024907749077491,
+      "grad_norm": 0.46979209382263687,
+      "learning_rate": 1.4468342902724591e-05,
+      "loss": 0.0828,
+      "step": 2195
+    },
+    {
+      "epoch": 2.025830258302583,
+      "grad_norm": 0.44744051567035725,
+      "learning_rate": 1.4444004936372165e-05,
+      "loss": 0.0684,
+      "step": 2196
+    },
+    {
+      "epoch": 2.0267527675276753,
+      "grad_norm": 0.3403332049495621,
+      "learning_rate": 1.4419679138889378e-05,
+      "loss": 0.0589,
+      "step": 2197
+    },
+    {
+      "epoch": 2.0276752767527677,
+      "grad_norm": 0.41763785693758054,
+      "learning_rate": 1.4395365538318829e-05,
+      "loss": 0.0631,
+      "step": 2198
+    },
+    {
+      "epoch": 2.0285977859778597,
+      "grad_norm": 0.39019638549907776,
+      "learning_rate": 1.4371064162689024e-05,
+      "loss": 0.0671,
+      "step": 2199
+    },
+    {
+      "epoch": 2.029520295202952,
+      "grad_norm": 0.3860442514405348,
+      "learning_rate": 1.4346775040014415e-05,
+      "loss": 0.0669,
+      "step": 2200
+    },
+    {
+      "epoch": 2.0304428044280445,
+      "grad_norm": 0.3344238710082706,
+      "learning_rate": 1.4322498198295327e-05,
+      "loss": 0.05,
+      "step": 2201
+    },
+    {
+      "epoch": 2.0313653136531364,
+      "grad_norm": 0.40033249044017183,
+      "learning_rate": 1.4298233665517896e-05,
+      "loss": 0.0756,
+      "step": 2202
+    },
+    {
+      "epoch": 2.032287822878229,
+      "grad_norm": 0.39955797763735873,
+      "learning_rate": 1.4273981469654093e-05,
+      "loss": 0.0713,
+      "step": 2203
+    },
+    {
+      "epoch": 2.0332103321033212,
+      "grad_norm": 0.35323675199493065,
+      "learning_rate": 1.4249741638661679e-05,
+      "loss": 0.0625,
+      "step": 2204
+    },
+    {
+      "epoch": 2.034132841328413,
+      "grad_norm": 0.36589217113532685,
+      "learning_rate": 1.4225514200484116e-05,
+      "loss": 0.065,
+      "step": 2205
+    },
+    {
+      "epoch": 2.0350553505535056,
+      "grad_norm": 0.30664598609653493,
+      "learning_rate": 1.4201299183050626e-05,
+      "loss": 0.0575,
+      "step": 2206
+    },
+    {
+      "epoch": 2.0359778597785976,
+      "grad_norm": 0.386599914617671,
+      "learning_rate": 1.4177096614276097e-05,
+      "loss": 0.0777,
+      "step": 2207
+    },
+    {
+      "epoch": 2.03690036900369,
+      "grad_norm": 0.36339789042022713,
+      "learning_rate": 1.4152906522061048e-05,
+      "loss": 0.06,
+      "step": 2208
+    },
+    {
+      "epoch": 2.0378228782287824,
+      "grad_norm": 0.3961162554717477,
+      "learning_rate": 1.4128728934291641e-05,
+      "loss": 0.0673,
+      "step": 2209
+    },
+    {
+      "epoch": 2.0387453874538743,
+      "grad_norm": 0.4122745090428468,
+      "learning_rate": 1.4104563878839621e-05,
+      "loss": 0.0684,
+      "step": 2210
+    },
+    {
+      "epoch": 2.0396678966789668,
+      "grad_norm": 0.44028661517890205,
+      "learning_rate": 1.4080411383562258e-05,
+      "loss": 0.0746,
+      "step": 2211
+    },
+    {
+      "epoch": 2.040590405904059,
+      "grad_norm": 0.3942892239056397,
+      "learning_rate": 1.4056271476302368e-05,
+      "loss": 0.0628,
+      "step": 2212
+    },
+    {
+      "epoch": 2.041512915129151,
+      "grad_norm": 0.3712665887448281,
+      "learning_rate": 1.4032144184888269e-05,
+      "loss": 0.0581,
+      "step": 2213
+    },
+    {
+      "epoch": 2.0424354243542435,
+      "grad_norm": 0.43744992110441555,
+      "learning_rate": 1.4008029537133685e-05,
+      "loss": 0.0627,
+      "step": 2214
+    },
+    {
+      "epoch": 2.043357933579336,
+      "grad_norm": 0.4857776711592854,
+      "learning_rate": 1.3983927560837815e-05,
+      "loss": 0.0705,
+      "step": 2215
+    },
+    {
+      "epoch": 2.044280442804428,
+      "grad_norm": 0.37197542467309835,
+      "learning_rate": 1.3959838283785237e-05,
+      "loss": 0.0745,
+      "step": 2216
+    },
+    {
+      "epoch": 2.0452029520295203,
+      "grad_norm": 0.39177509656673304,
+      "learning_rate": 1.3935761733745865e-05,
+      "loss": 0.0637,
+      "step": 2217
+    },
+    {
+      "epoch": 2.0461254612546127,
+      "grad_norm": 0.4065437517107225,
+      "learning_rate": 1.3911697938474966e-05,
+      "loss": 0.0579,
+      "step": 2218
+    },
+    {
+      "epoch": 2.0470479704797047,
+      "grad_norm": 0.4002320100803212,
+      "learning_rate": 1.3887646925713116e-05,
+      "loss": 0.0754,
+      "step": 2219
+    },
+    {
+      "epoch": 2.047970479704797,
+      "grad_norm": 0.3549424979834139,
+      "learning_rate": 1.3863608723186108e-05,
+      "loss": 0.0707,
+      "step": 2220
+    },
+    {
+      "epoch": 2.0488929889298895,
+      "grad_norm": 0.3870244238874869,
+      "learning_rate": 1.3839583358605012e-05,
+      "loss": 0.0711,
+      "step": 2221
+    },
+    {
+      "epoch": 2.0498154981549814,
+      "grad_norm": 0.4444638854748495,
+      "learning_rate": 1.3815570859666091e-05,
+      "loss": 0.0688,
+      "step": 2222
+    },
+    {
+      "epoch": 2.050738007380074,
+      "grad_norm": 0.40841543971334465,
+      "learning_rate": 1.3791571254050747e-05,
+      "loss": 0.0847,
+      "step": 2223
+    },
+    {
+      "epoch": 2.0516605166051662,
+      "grad_norm": 0.34017711339234313,
+      "learning_rate": 1.3767584569425562e-05,
+      "loss": 0.0584,
+      "step": 2224
+    },
+    {
+      "epoch": 2.052583025830258,
+      "grad_norm": 0.40919307099781965,
+      "learning_rate": 1.3743610833442182e-05,
+      "loss": 0.0757,
+      "step": 2225
+    },
+    {
+      "epoch": 2.0535055350553506,
+      "grad_norm": 0.35418905688832636,
+      "learning_rate": 1.3719650073737352e-05,
+      "loss": 0.0686,
+      "step": 2226
+    },
+    {
+      "epoch": 2.054428044280443,
+      "grad_norm": 0.3812889972286595,
+      "learning_rate": 1.3695702317932862e-05,
+      "loss": 0.0751,
+      "step": 2227
+    },
+    {
+      "epoch": 2.055350553505535,
+      "grad_norm": 0.36828127241310443,
+      "learning_rate": 1.3671767593635482e-05,
+      "loss": 0.0585,
+      "step": 2228
+    },
+    {
+      "epoch": 2.0562730627306274,
+      "grad_norm": 0.4465144692445298,
+      "learning_rate": 1.3647845928436986e-05,
+      "loss": 0.0746,
+      "step": 2229
+    },
+    {
+      "epoch": 2.0571955719557193,
+      "grad_norm": 0.30379954883920934,
+      "learning_rate": 1.3623937349914093e-05,
+      "loss": 0.0556,
+      "step": 2230
+    },
+    {
+      "epoch": 2.0581180811808117,
+      "grad_norm": 0.41901500267106806,
+      "learning_rate": 1.3600041885628409e-05,
+      "loss": 0.0692,
+      "step": 2231
+    },
+    {
+      "epoch": 2.059040590405904,
+      "grad_norm": 0.41402687661233484,
+      "learning_rate": 1.357615956312645e-05,
+      "loss": 0.072,
+      "step": 2232
+    },
+    {
+      "epoch": 2.059963099630996,
+      "grad_norm": 0.40045954654873867,
+      "learning_rate": 1.355229040993959e-05,
+      "loss": 0.0672,
+      "step": 2233
+    },
+    {
+      "epoch": 2.0608856088560885,
+      "grad_norm": 0.4218320383494676,
+      "learning_rate": 1.3528434453583972e-05,
+      "loss": 0.0634,
+      "step": 2234
+    },
+    {
+      "epoch": 2.061808118081181,
+      "grad_norm": 0.3874555064614107,
+      "learning_rate": 1.3504591721560578e-05,
+      "loss": 0.0682,
+      "step": 2235
+    },
+    {
+      "epoch": 2.062730627306273,
+      "grad_norm": 0.4112000719178533,
+      "learning_rate": 1.3480762241355132e-05,
+      "loss": 0.0684,
+      "step": 2236
+    },
+    {
+      "epoch": 2.0636531365313653,
+      "grad_norm": 0.3738712320639851,
+      "learning_rate": 1.3456946040438057e-05,
+      "loss": 0.0625,
+      "step": 2237
+    },
+    {
+      "epoch": 2.0645756457564577,
+      "grad_norm": 0.43171012828868377,
+      "learning_rate": 1.3433143146264493e-05,
+      "loss": 0.0701,
+      "step": 2238
+    },
+    {
+      "epoch": 2.0654981549815496,
+      "grad_norm": 0.3524502423148047,
+      "learning_rate": 1.3409353586274243e-05,
+      "loss": 0.0634,
+      "step": 2239
+    },
+    {
+      "epoch": 2.066420664206642,
+      "grad_norm": 0.37612818023093847,
+      "learning_rate": 1.338557738789171e-05,
+      "loss": 0.0706,
+      "step": 2240
+    },
+    {
+      "epoch": 2.0673431734317345,
+      "grad_norm": 0.41251392187694436,
+      "learning_rate": 1.3361814578525922e-05,
+      "loss": 0.0753,
+      "step": 2241
+    },
+    {
+      "epoch": 2.0682656826568264,
+      "grad_norm": 0.339735749819185,
+      "learning_rate": 1.333806518557047e-05,
+      "loss": 0.0619,
+      "step": 2242
+    },
+    {
+      "epoch": 2.069188191881919,
+      "grad_norm": 0.41289076151051984,
+      "learning_rate": 1.331432923640345e-05,
+      "loss": 0.0723,
+      "step": 2243
+    },
+    {
+      "epoch": 2.0701107011070112,
+      "grad_norm": 0.3387208444780958,
+      "learning_rate": 1.3290606758387498e-05,
+      "loss": 0.0579,
+      "step": 2244
+    },
+    {
+      "epoch": 2.071033210332103,
+      "grad_norm": 0.36216492165089403,
+      "learning_rate": 1.3266897778869702e-05,
+      "loss": 0.0596,
+      "step": 2245
+    },
+    {
+      "epoch": 2.0719557195571956,
+      "grad_norm": 0.3870132395876351,
+      "learning_rate": 1.324320232518158e-05,
+      "loss": 0.0674,
+      "step": 2246
+    },
+    {
+      "epoch": 2.072878228782288,
+      "grad_norm": 0.35777903887924795,
+      "learning_rate": 1.3219520424639076e-05,
+      "loss": 0.0609,
+      "step": 2247
+    },
+    {
+      "epoch": 2.07380073800738,
+      "grad_norm": 0.3974356851698215,
+      "learning_rate": 1.3195852104542511e-05,
+      "loss": 0.0638,
+      "step": 2248
+    },
+    {
+      "epoch": 2.0747232472324724,
+      "grad_norm": 0.4139265534374722,
+      "learning_rate": 1.3172197392176525e-05,
+      "loss": 0.0675,
+      "step": 2249
+    },
+    {
+      "epoch": 2.0756457564575648,
+      "grad_norm": 0.4060327318740307,
+      "learning_rate": 1.3148556314810092e-05,
+      "loss": 0.0615,
+      "step": 2250
+    },
+    {
+      "epoch": 2.0765682656826567,
+      "grad_norm": 0.41270482151611887,
+      "learning_rate": 1.3124928899696476e-05,
+      "loss": 0.0618,
+      "step": 2251
+    },
+    {
+      "epoch": 2.077490774907749,
+      "grad_norm": 0.40572409594054104,
+      "learning_rate": 1.3101315174073162e-05,
+      "loss": 0.0667,
+      "step": 2252
+    },
+    {
+      "epoch": 2.0784132841328415,
+      "grad_norm": 0.3527089118513685,
+      "learning_rate": 1.3077715165161878e-05,
+      "loss": 0.0613,
+      "step": 2253
+    },
+    {
+      "epoch": 2.0793357933579335,
+      "grad_norm": 0.3815807082735544,
+      "learning_rate": 1.3054128900168538e-05,
+      "loss": 0.066,
+      "step": 2254
+    },
+    {
+      "epoch": 2.080258302583026,
+      "grad_norm": 0.3601800142960928,
+      "learning_rate": 1.3030556406283195e-05,
+      "loss": 0.0697,
+      "step": 2255
+    },
+    {
+      "epoch": 2.081180811808118,
+      "grad_norm": 0.3585956071637972,
+      "learning_rate": 1.3006997710680041e-05,
+      "loss": 0.0663,
+      "step": 2256
+    },
+    {
+      "epoch": 2.0821033210332103,
+      "grad_norm": 0.36790788748310854,
+      "learning_rate": 1.298345284051737e-05,
+      "loss": 0.0659,
+      "step": 2257
+    },
+    {
+      "epoch": 2.0830258302583027,
+      "grad_norm": 0.406472928802804,
+      "learning_rate": 1.295992182293751e-05,
+      "loss": 0.0612,
+      "step": 2258
+    },
+    {
+      "epoch": 2.0839483394833946,
+      "grad_norm": 0.4229844842336367,
+      "learning_rate": 1.2936404685066852e-05,
+      "loss": 0.0758,
+      "step": 2259
+    },
+    {
+      "epoch": 2.084870848708487,
+      "grad_norm": 0.4137305147479194,
+      "learning_rate": 1.2912901454015752e-05,
+      "loss": 0.0695,
+      "step": 2260
+    },
+    {
+      "epoch": 2.0857933579335795,
+      "grad_norm": 0.3884366301219784,
+      "learning_rate": 1.2889412156878566e-05,
+      "loss": 0.0639,
+      "step": 2261
+    },
+    {
+      "epoch": 2.0867158671586714,
+      "grad_norm": 0.37571829415748415,
+      "learning_rate": 1.2865936820733582e-05,
+      "loss": 0.0704,
+      "step": 2262
+    },
+    {
+      "epoch": 2.087638376383764,
+      "grad_norm": 0.4365705947838686,
+      "learning_rate": 1.2842475472642968e-05,
+      "loss": 0.0718,
+      "step": 2263
+    },
+    {
+      "epoch": 2.088560885608856,
+      "grad_norm": 0.3621254157807818,
+      "learning_rate": 1.2819028139652794e-05,
+      "loss": 0.0768,
+      "step": 2264
+    },
+    {
+      "epoch": 2.089483394833948,
+      "grad_norm": 0.3571045514206385,
+      "learning_rate": 1.2795594848792975e-05,
+      "loss": 0.0663,
+      "step": 2265
+    },
+    {
+      "epoch": 2.0904059040590406,
+      "grad_norm": 0.382496586818103,
+      "learning_rate": 1.2772175627077205e-05,
+      "loss": 0.0618,
+      "step": 2266
+    },
+    {
+      "epoch": 2.091328413284133,
+      "grad_norm": 0.35917726450691434,
+      "learning_rate": 1.2748770501502994e-05,
+      "loss": 0.0616,
+      "step": 2267
+    },
+    {
+      "epoch": 2.092250922509225,
+      "grad_norm": 0.35707437854815516,
+      "learning_rate": 1.2725379499051603e-05,
+      "loss": 0.0595,
+      "step": 2268
+    },
+    {
+      "epoch": 2.0931734317343174,
+      "grad_norm": 0.36622075267297094,
+      "learning_rate": 1.2702002646687976e-05,
+      "loss": 0.054,
+      "step": 2269
+    },
+    {
+      "epoch": 2.0940959409594098,
+      "grad_norm": 0.37514244833794425,
+      "learning_rate": 1.2678639971360778e-05,
+      "loss": 0.06,
+      "step": 2270
+    },
+    {
+      "epoch": 2.0950184501845017,
+      "grad_norm": 0.41975878863510097,
+      "learning_rate": 1.265529150000233e-05,
+      "loss": 0.0651,
+      "step": 2271
+    },
+    {
+      "epoch": 2.095940959409594,
+      "grad_norm": 0.42133748723146647,
+      "learning_rate": 1.2631957259528553e-05,
+      "loss": 0.0629,
+      "step": 2272
+    },
+    {
+      "epoch": 2.0968634686346865,
+      "grad_norm": 0.39875129353097116,
+      "learning_rate": 1.2608637276838986e-05,
+      "loss": 0.0725,
+      "step": 2273
+    },
+    {
+      "epoch": 2.0977859778597785,
+      "grad_norm": 0.40771910402112077,
+      "learning_rate": 1.2585331578816738e-05,
+      "loss": 0.0658,
+      "step": 2274
+    },
+    {
+      "epoch": 2.098708487084871,
+      "grad_norm": 0.34925279844290885,
+      "learning_rate": 1.2562040192328414e-05,
+      "loss": 0.0608,
+      "step": 2275
+    },
+    {
+      "epoch": 2.0996309963099633,
+      "grad_norm": 0.36084418302418064,
+      "learning_rate": 1.2538763144224157e-05,
+      "loss": 0.0624,
+      "step": 2276
+    },
+    {
+      "epoch": 2.1005535055350553,
+      "grad_norm": 0.3621477738896384,
+      "learning_rate": 1.2515500461337581e-05,
+      "loss": 0.0622,
+      "step": 2277
+    },
+    {
+      "epoch": 2.1014760147601477,
+      "grad_norm": 0.4166217819345527,
+      "learning_rate": 1.2492252170485702e-05,
+      "loss": 0.0609,
+      "step": 2278
+    },
+    {
+      "epoch": 2.10239852398524,
+      "grad_norm": 0.3963129875773945,
+      "learning_rate": 1.2469018298468982e-05,
+      "loss": 0.0735,
+      "step": 2279
+    },
+    {
+      "epoch": 2.103321033210332,
+      "grad_norm": 0.40031229458504575,
+      "learning_rate": 1.244579887207126e-05,
+      "loss": 0.0643,
+      "step": 2280
+    },
+    {
+      "epoch": 2.1042435424354244,
+      "grad_norm": 0.38710135688527125,
+      "learning_rate": 1.2422593918059702e-05,
+      "loss": 0.061,
+      "step": 2281
+    },
+    {
+      "epoch": 2.1051660516605164,
+      "grad_norm": 0.42932007526489624,
+      "learning_rate": 1.239940346318478e-05,
+      "loss": 0.065,
+      "step": 2282
+    },
+    {
+      "epoch": 2.106088560885609,
+      "grad_norm": 0.3549475954255283,
+      "learning_rate": 1.2376227534180309e-05,
+      "loss": 0.0691,
+      "step": 2283
+    },
+    {
+      "epoch": 2.107011070110701,
+      "grad_norm": 0.36893883002542316,
+      "learning_rate": 1.2353066157763304e-05,
+      "loss": 0.0647,
+      "step": 2284
+    },
+    {
+      "epoch": 2.107933579335793,
+      "grad_norm": 0.38400940343698575,
+      "learning_rate": 1.2329919360634002e-05,
+      "loss": 0.0772,
+      "step": 2285
+    },
+    {
+      "epoch": 2.1088560885608856,
+      "grad_norm": 0.41039114987252323,
+      "learning_rate": 1.2306787169475887e-05,
+      "loss": 0.0864,
+      "step": 2286
+    },
+    {
+      "epoch": 2.109778597785978,
+      "grad_norm": 0.37516276125973114,
+      "learning_rate": 1.2283669610955542e-05,
+      "loss": 0.0686,
+      "step": 2287
+    },
+    {
+      "epoch": 2.11070110701107,
+      "grad_norm": 0.4340295903624818,
+      "learning_rate": 1.2260566711722723e-05,
+      "loss": 0.0649,
+      "step": 2288
+    },
+    {
+      "epoch": 2.1116236162361623,
+      "grad_norm": 0.32110411537845807,
+      "learning_rate": 1.2237478498410282e-05,
+      "loss": 0.0527,
+      "step": 2289
+    },
+    {
+      "epoch": 2.1125461254612548,
+      "grad_norm": 0.3726146451240176,
+      "learning_rate": 1.2214404997634117e-05,
+      "loss": 0.0668,
+      "step": 2290
+    },
+    {
+      "epoch": 2.1134686346863467,
+      "grad_norm": 0.39411252860068463,
+      "learning_rate": 1.2191346235993185e-05,
+      "loss": 0.0611,
+      "step": 2291
+    },
+    {
+      "epoch": 2.114391143911439,
+      "grad_norm": 0.4210251988531999,
+      "learning_rate": 1.216830224006946e-05,
+      "loss": 0.0628,
+      "step": 2292
+    },
+    {
+      "epoch": 2.1153136531365315,
+      "grad_norm": 0.3555758988732575,
+      "learning_rate": 1.2145273036427865e-05,
+      "loss": 0.0596,
+      "step": 2293
+    },
+    {
+      "epoch": 2.1162361623616235,
+      "grad_norm": 0.4227876412490418,
+      "learning_rate": 1.2122258651616306e-05,
+      "loss": 0.06,
+      "step": 2294
+    },
+    {
+      "epoch": 2.117158671586716,
+      "grad_norm": 0.42060758733162873,
+      "learning_rate": 1.209925911216557e-05,
+      "loss": 0.0717,
+      "step": 2295
+    },
+    {
+      "epoch": 2.1180811808118083,
+      "grad_norm": 0.47720257761139323,
+      "learning_rate": 1.2076274444589361e-05,
+      "loss": 0.0752,
+      "step": 2296
+    },
+    {
+      "epoch": 2.1190036900369003,
+      "grad_norm": 0.3681611428166761,
+      "learning_rate": 1.205330467538423e-05,
+      "loss": 0.0678,
+      "step": 2297
+    },
+    {
+      "epoch": 2.1199261992619927,
+      "grad_norm": 0.4372909250762327,
+      "learning_rate": 1.2030349831029537e-05,
+      "loss": 0.0623,
+      "step": 2298
+    },
+    {
+      "epoch": 2.120848708487085,
+      "grad_norm": 0.3553890499364409,
+      "learning_rate": 1.2007409937987451e-05,
+      "loss": 0.0642,
+      "step": 2299
+    },
+    {
+      "epoch": 2.121771217712177,
+      "grad_norm": 0.38781139433456296,
+      "learning_rate": 1.1984485022702918e-05,
+      "loss": 0.0648,
+      "step": 2300
+    },
+    {
+      "epoch": 2.1226937269372694,
+      "grad_norm": 0.3891831431988466,
+      "learning_rate": 1.1961575111603588e-05,
+      "loss": 0.0643,
+      "step": 2301
+    },
+    {
+      "epoch": 2.123616236162362,
+      "grad_norm": 0.4103801961617336,
+      "learning_rate": 1.1938680231099833e-05,
+      "loss": 0.0624,
+      "step": 2302
+    },
+    {
+      "epoch": 2.124538745387454,
+      "grad_norm": 0.3456931795144724,
+      "learning_rate": 1.1915800407584704e-05,
+      "loss": 0.0583,
+      "step": 2303
+    },
+    {
+      "epoch": 2.125461254612546,
+      "grad_norm": 0.4040172263992759,
+      "learning_rate": 1.1892935667433871e-05,
+      "loss": 0.0625,
+      "step": 2304
+    },
+    {
+      "epoch": 2.126383763837638,
+      "grad_norm": 0.390510944677046,
+      "learning_rate": 1.1870086037005635e-05,
+      "loss": 0.0627,
+      "step": 2305
+    },
+    {
+      "epoch": 2.1273062730627306,
+      "grad_norm": 0.42092084980087036,
+      "learning_rate": 1.1847251542640885e-05,
+      "loss": 0.0758,
+      "step": 2306
+    },
+    {
+      "epoch": 2.128228782287823,
+      "grad_norm": 0.3625245206687934,
+      "learning_rate": 1.182443221066303e-05,
+      "loss": 0.0628,
+      "step": 2307
+    },
+    {
+      "epoch": 2.129151291512915,
+      "grad_norm": 0.3543920209441862,
+      "learning_rate": 1.1801628067378031e-05,
+      "loss": 0.0658,
+      "step": 2308
+    },
+    {
+      "epoch": 2.1300738007380073,
+      "grad_norm": 0.3692478597606682,
+      "learning_rate": 1.1778839139074338e-05,
+      "loss": 0.0661,
+      "step": 2309
+    },
+    {
+      "epoch": 2.1309963099630997,
+      "grad_norm": 0.3532929371171087,
+      "learning_rate": 1.175606545202283e-05,
+      "loss": 0.0656,
+      "step": 2310
+    },
+    {
+      "epoch": 2.1319188191881917,
+      "grad_norm": 0.39556690096193775,
+      "learning_rate": 1.1733307032476848e-05,
+      "loss": 0.0662,
+      "step": 2311
+    },
+    {
+      "epoch": 2.132841328413284,
+      "grad_norm": 0.398480730251781,
+      "learning_rate": 1.1710563906672134e-05,
+      "loss": 0.0585,
+      "step": 2312
+    },
+    {
+      "epoch": 2.1337638376383765,
+      "grad_norm": 0.34506871906767433,
+      "learning_rate": 1.1687836100826765e-05,
+      "loss": 0.0534,
+      "step": 2313
+    },
+    {
+      "epoch": 2.1346863468634685,
+      "grad_norm": 0.3339671541671619,
+      "learning_rate": 1.1665123641141194e-05,
+      "loss": 0.0541,
+      "step": 2314
+    },
+    {
+      "epoch": 2.135608856088561,
+      "grad_norm": 0.37804976582248867,
+      "learning_rate": 1.1642426553798174e-05,
+      "loss": 0.0575,
+      "step": 2315
+    },
+    {
+      "epoch": 2.1365313653136533,
+      "grad_norm": 0.3580525123782122,
+      "learning_rate": 1.1619744864962727e-05,
+      "loss": 0.0738,
+      "step": 2316
+    },
+    {
+      "epoch": 2.1374538745387452,
+      "grad_norm": 0.3593492691535271,
+      "learning_rate": 1.159707860078211e-05,
+      "loss": 0.0694,
+      "step": 2317
+    },
+    {
+      "epoch": 2.1383763837638377,
+      "grad_norm": 0.38956493884705623,
+      "learning_rate": 1.1574427787385852e-05,
+      "loss": 0.0661,
+      "step": 2318
+    },
+    {
+      "epoch": 2.13929889298893,
+      "grad_norm": 0.3601261188299134,
+      "learning_rate": 1.1551792450885617e-05,
+      "loss": 0.0593,
+      "step": 2319
+    },
+    {
+      "epoch": 2.140221402214022,
+      "grad_norm": 0.3725932028380041,
+      "learning_rate": 1.1529172617375234e-05,
+      "loss": 0.0646,
+      "step": 2320
+    },
+    {
+      "epoch": 2.1411439114391144,
+      "grad_norm": 0.5115553281623878,
+      "learning_rate": 1.1506568312930698e-05,
+      "loss": 0.0672,
+      "step": 2321
+    },
+    {
+      "epoch": 2.142066420664207,
+      "grad_norm": 0.413571928782127,
+      "learning_rate": 1.148397956361007e-05,
+      "loss": 0.059,
+      "step": 2322
+    },
+    {
+      "epoch": 2.142988929889299,
+      "grad_norm": 0.4884203786739079,
+      "learning_rate": 1.1461406395453459e-05,
+      "loss": 0.0786,
+      "step": 2323
+    },
+    {
+      "epoch": 2.143911439114391,
+      "grad_norm": 0.3768262843453861,
+      "learning_rate": 1.1438848834483081e-05,
+      "loss": 0.0557,
+      "step": 2324
+    },
+    {
+      "epoch": 2.1448339483394836,
+      "grad_norm": 0.39452290042589616,
+      "learning_rate": 1.1416306906703097e-05,
+      "loss": 0.0621,
+      "step": 2325
+    },
+    {
+      "epoch": 2.1457564575645756,
+      "grad_norm": 0.40143554816102517,
+      "learning_rate": 1.139378063809966e-05,
+      "loss": 0.0622,
+      "step": 2326
+    },
+    {
+      "epoch": 2.146678966789668,
+      "grad_norm": 0.40885176319407773,
+      "learning_rate": 1.1371270054640884e-05,
+      "loss": 0.0607,
+      "step": 2327
+    },
+    {
+      "epoch": 2.14760147601476,
+      "grad_norm": 0.39687419030518506,
+      "learning_rate": 1.1348775182276802e-05,
+      "loss": 0.0624,
+      "step": 2328
+    },
+    {
+      "epoch": 2.1485239852398523,
+      "grad_norm": 0.34413609071406964,
+      "learning_rate": 1.1326296046939333e-05,
+      "loss": 0.0598,
+      "step": 2329
+    },
+    {
+      "epoch": 2.1494464944649447,
+      "grad_norm": 0.33814121563544586,
+      "learning_rate": 1.1303832674542236e-05,
+      "loss": 0.0562,
+      "step": 2330
+    },
+    {
+      "epoch": 2.1503690036900367,
+      "grad_norm": 0.3825110840593142,
+      "learning_rate": 1.1281385090981119e-05,
+      "loss": 0.0619,
+      "step": 2331
+    },
+    {
+      "epoch": 2.151291512915129,
+      "grad_norm": 0.3458059332912318,
+      "learning_rate": 1.1258953322133398e-05,
+      "loss": 0.0552,
+      "step": 2332
+    },
+    {
+      "epoch": 2.1522140221402215,
+      "grad_norm": 0.383444247383961,
+      "learning_rate": 1.1236537393858216e-05,
+      "loss": 0.0607,
+      "step": 2333
+    },
+    {
+      "epoch": 2.1531365313653135,
+      "grad_norm": 0.40269803579481,
+      "learning_rate": 1.12141373319965e-05,
+      "loss": 0.0688,
+      "step": 2334
+    },
+    {
+      "epoch": 2.154059040590406,
+      "grad_norm": 0.36849070485549473,
+      "learning_rate": 1.1191753162370871e-05,
+      "loss": 0.063,
+      "step": 2335
+    },
+    {
+      "epoch": 2.1549815498154983,
+      "grad_norm": 0.3671164565183918,
+      "learning_rate": 1.1169384910785614e-05,
+      "loss": 0.061,
+      "step": 2336
+    },
+    {
+      "epoch": 2.1559040590405902,
+      "grad_norm": 0.38888831346308533,
+      "learning_rate": 1.114703260302668e-05,
+      "loss": 0.0656,
+      "step": 2337
+    },
+    {
+      "epoch": 2.1568265682656826,
+      "grad_norm": 0.37897801285010585,
+      "learning_rate": 1.1124696264861654e-05,
+      "loss": 0.0663,
+      "step": 2338
+    },
+    {
+      "epoch": 2.157749077490775,
+      "grad_norm": 0.3468797017958368,
+      "learning_rate": 1.1102375922039665e-05,
+      "loss": 0.0701,
+      "step": 2339
+    },
+    {
+      "epoch": 2.158671586715867,
+      "grad_norm": 0.37549681661540163,
+      "learning_rate": 1.1080071600291453e-05,
+      "loss": 0.0689,
+      "step": 2340
+    },
+    {
+      "epoch": 2.1595940959409594,
+      "grad_norm": 0.3838873479425886,
+      "learning_rate": 1.1057783325329268e-05,
+      "loss": 0.0571,
+      "step": 2341
+    },
+    {
+      "epoch": 2.160516605166052,
+      "grad_norm": 0.4474304276936308,
+      "learning_rate": 1.1035511122846848e-05,
+      "loss": 0.0697,
+      "step": 2342
+    },
+    {
+      "epoch": 2.161439114391144,
+      "grad_norm": 0.4840220730012607,
+      "learning_rate": 1.1013255018519425e-05,
+      "loss": 0.073,
+      "step": 2343
+    },
+    {
+      "epoch": 2.162361623616236,
+      "grad_norm": 0.3578996455173482,
+      "learning_rate": 1.099101503800367e-05,
+      "loss": 0.0613,
+      "step": 2344
+    },
+    {
+      "epoch": 2.1632841328413286,
+      "grad_norm": 0.39934092683814854,
+      "learning_rate": 1.0968791206937645e-05,
+      "loss": 0.0656,
+      "step": 2345
+    },
+    {
+      "epoch": 2.1642066420664205,
+      "grad_norm": 0.41187406060511095,
+      "learning_rate": 1.094658355094082e-05,
+      "loss": 0.0648,
+      "step": 2346
+    },
+    {
+      "epoch": 2.165129151291513,
+      "grad_norm": 0.4091523390014875,
+      "learning_rate": 1.0924392095614019e-05,
+      "loss": 0.0691,
+      "step": 2347
+    },
+    {
+      "epoch": 2.1660516605166054,
+      "grad_norm": 0.44492007884459395,
+      "learning_rate": 1.0902216866539363e-05,
+      "loss": 0.0639,
+      "step": 2348
+    },
+    {
+      "epoch": 2.1669741697416973,
+      "grad_norm": 0.38470641398628264,
+      "learning_rate": 1.088005788928029e-05,
+      "loss": 0.0617,
+      "step": 2349
+    },
+    {
+      "epoch": 2.1678966789667897,
+      "grad_norm": 0.4269943517763368,
+      "learning_rate": 1.0857915189381513e-05,
+      "loss": 0.0654,
+      "step": 2350
+    },
+    {
+      "epoch": 2.1688191881918817,
+      "grad_norm": 0.45813278300867916,
+      "learning_rate": 1.083578879236895e-05,
+      "loss": 0.0758,
+      "step": 2351
+    },
+    {
+      "epoch": 2.169741697416974,
+      "grad_norm": 0.4183955750091749,
+      "learning_rate": 1.0813678723749725e-05,
+      "loss": 0.0779,
+      "step": 2352
+    },
+    {
+      "epoch": 2.1706642066420665,
+      "grad_norm": 0.3627470265115012,
+      "learning_rate": 1.0791585009012196e-05,
+      "loss": 0.0546,
+      "step": 2353
+    },
+    {
+      "epoch": 2.171586715867159,
+      "grad_norm": 0.3546274555675706,
+      "learning_rate": 1.07695076736258e-05,
+      "loss": 0.0557,
+      "step": 2354
+    },
+    {
+      "epoch": 2.172509225092251,
+      "grad_norm": 0.41356451482436757,
+      "learning_rate": 1.0747446743041107e-05,
+      "loss": 0.0703,
+      "step": 2355
+    },
+    {
+      "epoch": 2.1734317343173433,
+      "grad_norm": 0.40438384325635135,
+      "learning_rate": 1.0725402242689823e-05,
+      "loss": 0.0673,
+      "step": 2356
+    },
+    {
+      "epoch": 2.1743542435424352,
+      "grad_norm": 0.3952222934802737,
+      "learning_rate": 1.0703374197984653e-05,
+      "loss": 0.0635,
+      "step": 2357
+    },
+    {
+      "epoch": 2.1752767527675276,
+      "grad_norm": 0.40404658865059395,
+      "learning_rate": 1.0681362634319347e-05,
+      "loss": 0.0695,
+      "step": 2358
+    },
+    {
+      "epoch": 2.17619926199262,
+      "grad_norm": 0.38566575571588047,
+      "learning_rate": 1.0659367577068702e-05,
+      "loss": 0.0626,
+      "step": 2359
+    },
+    {
+      "epoch": 2.177121771217712,
+      "grad_norm": 0.38281627800852175,
+      "learning_rate": 1.0637389051588426e-05,
+      "loss": 0.0698,
+      "step": 2360
+    },
+    {
+      "epoch": 2.1780442804428044,
+      "grad_norm": 0.38619544092330665,
+      "learning_rate": 1.0615427083215187e-05,
+      "loss": 0.0597,
+      "step": 2361
+    },
+    {
+      "epoch": 2.178966789667897,
+      "grad_norm": 0.40890299425675625,
+      "learning_rate": 1.0593481697266583e-05,
+      "loss": 0.0713,
+      "step": 2362
+    },
+    {
+      "epoch": 2.1798892988929888,
+      "grad_norm": 0.37383695344433404,
+      "learning_rate": 1.0571552919041094e-05,
+      "loss": 0.0629,
+      "step": 2363
+    },
+    {
+      "epoch": 2.180811808118081,
+      "grad_norm": 0.4037774375780884,
+      "learning_rate": 1.0549640773818029e-05,
+      "loss": 0.0784,
+      "step": 2364
+    },
+    {
+      "epoch": 2.1817343173431736,
+      "grad_norm": 0.37167875891239616,
+      "learning_rate": 1.0527745286857549e-05,
+      "loss": 0.0626,
+      "step": 2365
+    },
+    {
+      "epoch": 2.1826568265682655,
+      "grad_norm": 0.4258021500571581,
+      "learning_rate": 1.050586648340061e-05,
+      "loss": 0.0644,
+      "step": 2366
+    },
+    {
+      "epoch": 2.183579335793358,
+      "grad_norm": 0.399482560989266,
+      "learning_rate": 1.0484004388668909e-05,
+      "loss": 0.0684,
+      "step": 2367
+    },
+    {
+      "epoch": 2.1845018450184504,
+      "grad_norm": 0.35748327143410524,
+      "learning_rate": 1.046215902786491e-05,
+      "loss": 0.0579,
+      "step": 2368
+    },
+    {
+      "epoch": 2.1854243542435423,
+      "grad_norm": 0.3796983469724346,
+      "learning_rate": 1.0440330426171786e-05,
+      "loss": 0.0615,
+      "step": 2369
+    },
+    {
+      "epoch": 2.1863468634686347,
+      "grad_norm": 0.41091979832216075,
+      "learning_rate": 1.0418518608753361e-05,
+      "loss": 0.0585,
+      "step": 2370
+    },
+    {
+      "epoch": 2.187269372693727,
+      "grad_norm": 0.4071090560604256,
+      "learning_rate": 1.0396723600754143e-05,
+      "loss": 0.0699,
+      "step": 2371
+    },
+    {
+      "epoch": 2.188191881918819,
+      "grad_norm": 0.4126300802092573,
+      "learning_rate": 1.0374945427299242e-05,
+      "loss": 0.0679,
+      "step": 2372
+    },
+    {
+      "epoch": 2.1891143911439115,
+      "grad_norm": 0.37866397424997933,
+      "learning_rate": 1.0353184113494386e-05,
+      "loss": 0.0575,
+      "step": 2373
+    },
+    {
+      "epoch": 2.190036900369004,
+      "grad_norm": 0.44172252715452986,
+      "learning_rate": 1.0331439684425822e-05,
+      "loss": 0.0646,
+      "step": 2374
+    },
+    {
+      "epoch": 2.190959409594096,
+      "grad_norm": 0.33456826651019794,
+      "learning_rate": 1.0309712165160376e-05,
+      "loss": 0.0574,
+      "step": 2375
+    },
+    {
+      "epoch": 2.1918819188191883,
+      "grad_norm": 0.4184823796794353,
+      "learning_rate": 1.0288001580745372e-05,
+      "loss": 0.0627,
+      "step": 2376
+    },
+    {
+      "epoch": 2.1928044280442807,
+      "grad_norm": 0.3716937237706671,
+      "learning_rate": 1.0266307956208585e-05,
+      "loss": 0.0551,
+      "step": 2377
+    },
+    {
+      "epoch": 2.1937269372693726,
+      "grad_norm": 0.38175690518378075,
+      "learning_rate": 1.0244631316558267e-05,
+      "loss": 0.0629,
+      "step": 2378
+    },
+    {
+      "epoch": 2.194649446494465,
+      "grad_norm": 0.4593206627948996,
+      "learning_rate": 1.022297168678309e-05,
+      "loss": 0.0565,
+      "step": 2379
+    },
+    {
+      "epoch": 2.195571955719557,
+      "grad_norm": 0.38855685156802766,
+      "learning_rate": 1.0201329091852091e-05,
+      "loss": 0.0648,
+      "step": 2380
+    },
+    {
+      "epoch": 2.1964944649446494,
+      "grad_norm": 0.38557600892648936,
+      "learning_rate": 1.0179703556714693e-05,
+      "loss": 0.0656,
+      "step": 2381
+    },
+    {
+      "epoch": 2.197416974169742,
+      "grad_norm": 0.35698436477503104,
+      "learning_rate": 1.0158095106300658e-05,
+      "loss": 0.0573,
+      "step": 2382
+    },
+    {
+      "epoch": 2.1983394833948338,
+      "grad_norm": 0.4095033326518666,
+      "learning_rate": 1.0136503765520023e-05,
+      "loss": 0.0695,
+      "step": 2383
+    },
+    {
+      "epoch": 2.199261992619926,
+      "grad_norm": 0.32909955635448535,
+      "learning_rate": 1.0114929559263122e-05,
+      "loss": 0.0627,
+      "step": 2384
+    },
+    {
+      "epoch": 2.2001845018450186,
+      "grad_norm": 0.42170322247331393,
+      "learning_rate": 1.0093372512400551e-05,
+      "loss": 0.0729,
+      "step": 2385
+    },
+    {
+      "epoch": 2.2011070110701105,
+      "grad_norm": 0.4170031294399854,
+      "learning_rate": 1.0071832649783094e-05,
+      "loss": 0.0648,
+      "step": 2386
+    },
+    {
+      "epoch": 2.202029520295203,
+      "grad_norm": 0.42263067841693086,
+      "learning_rate": 1.005030999624172e-05,
+      "loss": 0.0712,
+      "step": 2387
+    },
+    {
+      "epoch": 2.2029520295202953,
+      "grad_norm": 0.39586244798219883,
+      "learning_rate": 1.0028804576587613e-05,
+      "loss": 0.0711,
+      "step": 2388
+    },
+    {
+      "epoch": 2.2038745387453873,
+      "grad_norm": 0.36370479167557207,
+      "learning_rate": 1.0007316415612039e-05,
+      "loss": 0.0666,
+      "step": 2389
+    },
+    {
+      "epoch": 2.2047970479704797,
+      "grad_norm": 0.43220718510201855,
+      "learning_rate": 9.985845538086367e-06,
+      "loss": 0.0632,
+      "step": 2390
+    },
+    {
+      "epoch": 2.205719557195572,
+      "grad_norm": 0.42938732546064245,
+      "learning_rate": 9.964391968762091e-06,
+      "loss": 0.0686,
+      "step": 2391
+    },
+    {
+      "epoch": 2.206642066420664,
+      "grad_norm": 0.3836159221445964,
+      "learning_rate": 9.942955732370707e-06,
+      "loss": 0.0572,
+      "step": 2392
+    },
+    {
+      "epoch": 2.2075645756457565,
+      "grad_norm": 0.4065015429719508,
+      "learning_rate": 9.921536853623719e-06,
+      "loss": 0.0642,
+      "step": 2393
+    },
+    {
+      "epoch": 2.208487084870849,
+      "grad_norm": 0.37698155144624346,
+      "learning_rate": 9.900135357212687e-06,
+      "loss": 0.0647,
+      "step": 2394
+    },
+    {
+      "epoch": 2.209409594095941,
+      "grad_norm": 0.37442380699492644,
+      "learning_rate": 9.878751267809069e-06,
+      "loss": 0.0628,
+      "step": 2395
+    },
+    {
+      "epoch": 2.2103321033210332,
+      "grad_norm": 0.38191813136680935,
+      "learning_rate": 9.857384610064272e-06,
+      "loss": 0.0595,
+      "step": 2396
+    },
+    {
+      "epoch": 2.2112546125461257,
+      "grad_norm": 0.3813478518906519,
+      "learning_rate": 9.83603540860962e-06,
+      "loss": 0.0637,
+      "step": 2397
+    },
+    {
+      "epoch": 2.2121771217712176,
+      "grad_norm": 0.3727803675915483,
+      "learning_rate": 9.814703688056321e-06,
+      "loss": 0.057,
+      "step": 2398
+    },
+    {
+      "epoch": 2.21309963099631,
+      "grad_norm": 0.39278714021478184,
+      "learning_rate": 9.793389472995393e-06,
+      "loss": 0.0615,
+      "step": 2399
+    },
+    {
+      "epoch": 2.2140221402214024,
+      "grad_norm": 0.39529353507293297,
+      "learning_rate": 9.772092787997714e-06,
+      "loss": 0.0649,
+      "step": 2400
+    },
+    {
+      "epoch": 2.2149446494464944,
+      "grad_norm": 0.4186754277263272,
+      "learning_rate": 9.750813657613944e-06,
+      "loss": 0.0722,
+      "step": 2401
+    },
+    {
+      "epoch": 2.215867158671587,
+      "grad_norm": 0.4482267232721739,
+      "learning_rate": 9.729552106374485e-06,
+      "loss": 0.0623,
+      "step": 2402
+    },
+    {
+      "epoch": 2.2167896678966788,
+      "grad_norm": 0.5059238984314811,
+      "learning_rate": 9.708308158789494e-06,
+      "loss": 0.0675,
+      "step": 2403
+    },
+    {
+      "epoch": 2.217712177121771,
+      "grad_norm": 0.421014149027054,
+      "learning_rate": 9.687081839348841e-06,
+      "loss": 0.0601,
+      "step": 2404
+    },
+    {
+      "epoch": 2.2186346863468636,
+      "grad_norm": 0.3930231949813996,
+      "learning_rate": 9.665873172522047e-06,
+      "loss": 0.0697,
+      "step": 2405
+    },
+    {
+      "epoch": 2.2195571955719555,
+      "grad_norm": 0.41585105554699614,
+      "learning_rate": 9.644682182758306e-06,
+      "loss": 0.0835,
+      "step": 2406
+    },
+    {
+      "epoch": 2.220479704797048,
+      "grad_norm": 0.37615047337233737,
+      "learning_rate": 9.623508894486435e-06,
+      "loss": 0.0619,
+      "step": 2407
+    },
+    {
+      "epoch": 2.2214022140221403,
+      "grad_norm": 0.3502035160808852,
+      "learning_rate": 9.602353332114825e-06,
+      "loss": 0.0608,
+      "step": 2408
+    },
+    {
+      "epoch": 2.2223247232472323,
+      "grad_norm": 0.408616407033454,
+      "learning_rate": 9.581215520031448e-06,
+      "loss": 0.058,
+      "step": 2409
+    },
+    {
+      "epoch": 2.2232472324723247,
+      "grad_norm": 0.4089092435670495,
+      "learning_rate": 9.560095482603823e-06,
+      "loss": 0.0617,
+      "step": 2410
+    },
+    {
+      "epoch": 2.224169741697417,
+      "grad_norm": 0.36470344619995576,
+      "learning_rate": 9.538993244178945e-06,
+      "loss": 0.0626,
+      "step": 2411
+    },
+    {
+      "epoch": 2.225092250922509,
+      "grad_norm": 0.37057796529587894,
+      "learning_rate": 9.517908829083324e-06,
+      "loss": 0.0677,
+      "step": 2412
+    },
+    {
+      "epoch": 2.2260147601476015,
+      "grad_norm": 0.3558886527371803,
+      "learning_rate": 9.496842261622921e-06,
+      "loss": 0.0567,
+      "step": 2413
+    },
+    {
+      "epoch": 2.226937269372694,
+      "grad_norm": 0.3742612831864683,
+      "learning_rate": 9.47579356608309e-06,
+      "loss": 0.057,
+      "step": 2414
+    },
+    {
+      "epoch": 2.227859778597786,
+      "grad_norm": 0.3952989125863061,
+      "learning_rate": 9.454762766728617e-06,
+      "loss": 0.0715,
+      "step": 2415
+    },
+    {
+      "epoch": 2.2287822878228782,
+      "grad_norm": 0.4586153160783086,
+      "learning_rate": 9.433749887803645e-06,
+      "loss": 0.0605,
+      "step": 2416
+    },
+    {
+      "epoch": 2.2297047970479706,
+      "grad_norm": 0.3772144695613314,
+      "learning_rate": 9.412754953531663e-06,
+      "loss": 0.0639,
+      "step": 2417
+    },
+    {
+      "epoch": 2.2306273062730626,
+      "grad_norm": 0.4140506144692939,
+      "learning_rate": 9.391777988115466e-06,
+      "loss": 0.0645,
+      "step": 2418
+    },
+    {
+      "epoch": 2.231549815498155,
+      "grad_norm": 0.3579186048846278,
+      "learning_rate": 9.37081901573712e-06,
+      "loss": 0.0621,
+      "step": 2419
+    },
+    {
+      "epoch": 2.2324723247232474,
+      "grad_norm": 0.3934103535754087,
+      "learning_rate": 9.349878060557999e-06,
+      "loss": 0.0609,
+      "step": 2420
+    },
+    {
+      "epoch": 2.2333948339483394,
+      "grad_norm": 0.38813103752980976,
+      "learning_rate": 9.328955146718655e-06,
+      "loss": 0.0626,
+      "step": 2421
+    },
+    {
+      "epoch": 2.234317343173432,
+      "grad_norm": 0.37357724828486133,
+      "learning_rate": 9.30805029833885e-06,
+      "loss": 0.0647,
+      "step": 2422
+    },
+    {
+      "epoch": 2.235239852398524,
+      "grad_norm": 0.36284855719708364,
+      "learning_rate": 9.28716353951756e-06,
+      "loss": 0.0629,
+      "step": 2423
+    },
+    {
+      "epoch": 2.236162361623616,
+      "grad_norm": 0.37294511156556737,
+      "learning_rate": 9.26629489433287e-06,
+      "loss": 0.0577,
+      "step": 2424
+    },
+    {
+      "epoch": 2.2370848708487086,
+      "grad_norm": 0.3842550911082737,
+      "learning_rate": 9.245444386841966e-06,
+      "loss": 0.0658,
+      "step": 2425
+    },
+    {
+      "epoch": 2.2380073800738005,
+      "grad_norm": 0.4017289531736733,
+      "learning_rate": 9.224612041081199e-06,
+      "loss": 0.0692,
+      "step": 2426
+    },
+    {
+      "epoch": 2.238929889298893,
+      "grad_norm": 0.36152473426005177,
+      "learning_rate": 9.203797881065906e-06,
+      "loss": 0.0666,
+      "step": 2427
+    },
+    {
+      "epoch": 2.2398523985239853,
+      "grad_norm": 0.39368555294441043,
+      "learning_rate": 9.183001930790483e-06,
+      "loss": 0.0615,
+      "step": 2428
+    },
+    {
+      "epoch": 2.2407749077490773,
+      "grad_norm": 0.34184825135839725,
+      "learning_rate": 9.16222421422837e-06,
+      "loss": 0.055,
+      "step": 2429
+    },
+    {
+      "epoch": 2.2416974169741697,
+      "grad_norm": 0.4035152367587123,
+      "learning_rate": 9.141464755331944e-06,
+      "loss": 0.0654,
+      "step": 2430
+    },
+    {
+      "epoch": 2.242619926199262,
+      "grad_norm": 0.3990062896703022,
+      "learning_rate": 9.120723578032536e-06,
+      "loss": 0.0683,
+      "step": 2431
+    },
+    {
+      "epoch": 2.243542435424354,
+      "grad_norm": 0.4154132915461665,
+      "learning_rate": 9.10000070624043e-06,
+      "loss": 0.0617,
+      "step": 2432
+    },
+    {
+      "epoch": 2.2444649446494465,
+      "grad_norm": 0.3608552225977703,
+      "learning_rate": 9.079296163844794e-06,
+      "loss": 0.0595,
+      "step": 2433
+    },
+    {
+      "epoch": 2.245387453874539,
+      "grad_norm": 0.42304298850095023,
+      "learning_rate": 9.058609974713655e-06,
+      "loss": 0.0714,
+      "step": 2434
+    },
+    {
+      "epoch": 2.246309963099631,
+      "grad_norm": 0.34194774227206476,
+      "learning_rate": 9.037942162693894e-06,
+      "loss": 0.0608,
+      "step": 2435
+    },
+    {
+      "epoch": 2.2472324723247232,
+      "grad_norm": 0.4286068577211741,
+      "learning_rate": 9.01729275161122e-06,
+      "loss": 0.0651,
+      "step": 2436
+    },
+    {
+      "epoch": 2.2481549815498156,
+      "grad_norm": 0.3957326714342427,
+      "learning_rate": 8.996661765270092e-06,
+      "loss": 0.0614,
+      "step": 2437
+    },
+    {
+      "epoch": 2.2490774907749076,
+      "grad_norm": 0.4157264538009682,
+      "learning_rate": 8.976049227453762e-06,
+      "loss": 0.064,
+      "step": 2438
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.3811542012344819,
+      "learning_rate": 8.955455161924217e-06,
+      "loss": 0.0596,
+      "step": 2439
+    },
+    {
+      "epoch": 2.2509225092250924,
+      "grad_norm": 0.43354843989295067,
+      "learning_rate": 8.934879592422113e-06,
+      "loss": 0.0685,
+      "step": 2440
+    },
+    {
+      "epoch": 2.2518450184501844,
+      "grad_norm": 0.3880171606411912,
+      "learning_rate": 8.914322542666822e-06,
+      "loss": 0.0634,
+      "step": 2441
+    },
+    {
+      "epoch": 2.2527675276752768,
+      "grad_norm": 0.3943942248307995,
+      "learning_rate": 8.893784036356359e-06,
+      "loss": 0.0595,
+      "step": 2442
+    },
+    {
+      "epoch": 2.253690036900369,
+      "grad_norm": 0.38605063977501863,
+      "learning_rate": 8.873264097167339e-06,
+      "loss": 0.0578,
+      "step": 2443
+    },
+    {
+      "epoch": 2.254612546125461,
+      "grad_norm": 0.3935636105100131,
+      "learning_rate": 8.852762748754994e-06,
+      "loss": 0.0713,
+      "step": 2444
+    },
+    {
+      "epoch": 2.2555350553505535,
+      "grad_norm": 0.3663058052845239,
+      "learning_rate": 8.832280014753132e-06,
+      "loss": 0.0564,
+      "step": 2445
+    },
+    {
+      "epoch": 2.256457564575646,
+      "grad_norm": 0.38903228868741724,
+      "learning_rate": 8.811815918774077e-06,
+      "loss": 0.0644,
+      "step": 2446
+    },
+    {
+      "epoch": 2.257380073800738,
+      "grad_norm": 0.3757239882689567,
+      "learning_rate": 8.791370484408684e-06,
+      "loss": 0.0694,
+      "step": 2447
+    },
+    {
+      "epoch": 2.2583025830258303,
+      "grad_norm": 0.4419107192723807,
+      "learning_rate": 8.770943735226303e-06,
+      "loss": 0.0756,
+      "step": 2448
+    },
+    {
+      "epoch": 2.2592250922509223,
+      "grad_norm": 0.4040877974243685,
+      "learning_rate": 8.750535694774714e-06,
+      "loss": 0.0703,
+      "step": 2449
+    },
+    {
+      "epoch": 2.2601476014760147,
+      "grad_norm": 0.4115440215289224,
+      "learning_rate": 8.730146386580157e-06,
+      "loss": 0.0608,
+      "step": 2450
+    },
+    {
+      "epoch": 2.261070110701107,
+      "grad_norm": 0.4137796143232803,
+      "learning_rate": 8.709775834147283e-06,
+      "loss": 0.066,
+      "step": 2451
+    },
+    {
+      "epoch": 2.2619926199261995,
+      "grad_norm": 0.35760666076703834,
+      "learning_rate": 8.689424060959082e-06,
+      "loss": 0.0529,
+      "step": 2452
+    },
+    {
+      "epoch": 2.2629151291512914,
+      "grad_norm": 0.38983243909343557,
+      "learning_rate": 8.669091090476944e-06,
+      "loss": 0.0738,
+      "step": 2453
+    },
+    {
+      "epoch": 2.263837638376384,
+      "grad_norm": 0.3870233680470246,
+      "learning_rate": 8.648776946140544e-06,
+      "loss": 0.0632,
+      "step": 2454
+    },
+    {
+      "epoch": 2.264760147601476,
+      "grad_norm": 0.40092473241835,
+      "learning_rate": 8.628481651367876e-06,
+      "loss": 0.0672,
+      "step": 2455
+    },
+    {
+      "epoch": 2.265682656826568,
+      "grad_norm": 0.37111043910186203,
+      "learning_rate": 8.608205229555207e-06,
+      "loss": 0.0598,
+      "step": 2456
+    },
+    {
+      "epoch": 2.2666051660516606,
+      "grad_norm": 0.36140785385982965,
+      "learning_rate": 8.587947704077018e-06,
+      "loss": 0.0531,
+      "step": 2457
+    },
+    {
+      "epoch": 2.2675276752767526,
+      "grad_norm": 0.359218678748364,
+      "learning_rate": 8.567709098286058e-06,
+      "loss": 0.0635,
+      "step": 2458
+    },
+    {
+      "epoch": 2.268450184501845,
+      "grad_norm": 0.3563479517059402,
+      "learning_rate": 8.547489435513222e-06,
+      "loss": 0.0585,
+      "step": 2459
+    },
+    {
+      "epoch": 2.2693726937269374,
+      "grad_norm": 0.35950359188192726,
+      "learning_rate": 8.527288739067562e-06,
+      "loss": 0.0659,
+      "step": 2460
+    },
+    {
+      "epoch": 2.2702952029520294,
+      "grad_norm": 0.36424993799359373,
+      "learning_rate": 8.507107032236322e-06,
+      "loss": 0.0663,
+      "step": 2461
+    },
+    {
+      "epoch": 2.2712177121771218,
+      "grad_norm": 0.42770263025619104,
+      "learning_rate": 8.486944338284797e-06,
+      "loss": 0.0637,
+      "step": 2462
+    },
+    {
+      "epoch": 2.272140221402214,
+      "grad_norm": 0.406461182290524,
+      "learning_rate": 8.46680068045637e-06,
+      "loss": 0.0657,
+      "step": 2463
+    },
+    {
+      "epoch": 2.273062730627306,
+      "grad_norm": 0.3550071200265757,
+      "learning_rate": 8.446676081972526e-06,
+      "loss": 0.0631,
+      "step": 2464
+    },
+    {
+      "epoch": 2.2739852398523985,
+      "grad_norm": 0.38647593018164184,
+      "learning_rate": 8.426570566032733e-06,
+      "loss": 0.0636,
+      "step": 2465
+    },
+    {
+      "epoch": 2.274907749077491,
+      "grad_norm": 0.40620402765224695,
+      "learning_rate": 8.406484155814465e-06,
+      "loss": 0.0686,
+      "step": 2466
+    },
+    {
+      "epoch": 2.275830258302583,
+      "grad_norm": 0.3637540363413214,
+      "learning_rate": 8.386416874473188e-06,
+      "loss": 0.0571,
+      "step": 2467
+    },
+    {
+      "epoch": 2.2767527675276753,
+      "grad_norm": 0.40426816905863777,
+      "learning_rate": 8.366368745142316e-06,
+      "loss": 0.0594,
+      "step": 2468
+    },
+    {
+      "epoch": 2.2776752767527677,
+      "grad_norm": 0.37627678017473065,
+      "learning_rate": 8.346339790933166e-06,
+      "loss": 0.0577,
+      "step": 2469
+    },
+    {
+      "epoch": 2.2785977859778597,
+      "grad_norm": 0.4329990706603838,
+      "learning_rate": 8.326330034934968e-06,
+      "loss": 0.0666,
+      "step": 2470
+    },
+    {
+      "epoch": 2.279520295202952,
+      "grad_norm": 0.37292130288204195,
+      "learning_rate": 8.306339500214821e-06,
+      "loss": 0.0549,
+      "step": 2471
+    },
+    {
+      "epoch": 2.280442804428044,
+      "grad_norm": 0.3779364943112543,
+      "learning_rate": 8.286368209817644e-06,
+      "loss": 0.0608,
+      "step": 2472
+    },
+    {
+      "epoch": 2.2813653136531364,
+      "grad_norm": 0.41512067511116424,
+      "learning_rate": 8.266416186766194e-06,
+      "loss": 0.0632,
+      "step": 2473
+    },
+    {
+      "epoch": 2.282287822878229,
+      "grad_norm": 0.397342143265139,
+      "learning_rate": 8.246483454061015e-06,
+      "loss": 0.0595,
+      "step": 2474
+    },
+    {
+      "epoch": 2.2832103321033212,
+      "grad_norm": 0.3750701551097111,
+      "learning_rate": 8.226570034680398e-06,
+      "loss": 0.0593,
+      "step": 2475
+    },
+    {
+      "epoch": 2.284132841328413,
+      "grad_norm": 0.37482356137981715,
+      "learning_rate": 8.206675951580381e-06,
+      "loss": 0.0536,
+      "step": 2476
+    },
+    {
+      "epoch": 2.2850553505535056,
+      "grad_norm": 0.4277169525656212,
+      "learning_rate": 8.186801227694722e-06,
+      "loss": 0.072,
+      "step": 2477
+    },
+    {
+      "epoch": 2.2859778597785976,
+      "grad_norm": 0.35039009884090044,
+      "learning_rate": 8.166945885934827e-06,
+      "loss": 0.0557,
+      "step": 2478
+    },
+    {
+      "epoch": 2.28690036900369,
+      "grad_norm": 0.37618948434110666,
+      "learning_rate": 8.147109949189793e-06,
+      "loss": 0.0587,
+      "step": 2479
+    },
+    {
+      "epoch": 2.2878228782287824,
+      "grad_norm": 0.34429121349112896,
+      "learning_rate": 8.127293440326344e-06,
+      "loss": 0.0524,
+      "step": 2480
+    },
+    {
+      "epoch": 2.2887453874538743,
+      "grad_norm": 0.41348640468203074,
+      "learning_rate": 8.107496382188781e-06,
+      "loss": 0.0683,
+      "step": 2481
+    },
+    {
+      "epoch": 2.2896678966789668,
+      "grad_norm": 0.35089151485209286,
+      "learning_rate": 8.087718797599006e-06,
+      "loss": 0.0613,
+      "step": 2482
+    },
+    {
+      "epoch": 2.290590405904059,
+      "grad_norm": 0.38263238067766436,
+      "learning_rate": 8.067960709356478e-06,
+      "loss": 0.0677,
+      "step": 2483
+    },
+    {
+      "epoch": 2.291512915129151,
+      "grad_norm": 0.4303658428177165,
+      "learning_rate": 8.048222140238148e-06,
+      "loss": 0.0801,
+      "step": 2484
+    },
+    {
+      "epoch": 2.2924354243542435,
+      "grad_norm": 0.3542670487397755,
+      "learning_rate": 8.028503112998496e-06,
+      "loss": 0.0567,
+      "step": 2485
+    },
+    {
+      "epoch": 2.293357933579336,
+      "grad_norm": 0.39400023544371354,
+      "learning_rate": 8.008803650369473e-06,
+      "loss": 0.0605,
+      "step": 2486
+    },
+    {
+      "epoch": 2.294280442804428,
+      "grad_norm": 0.3408886467122237,
+      "learning_rate": 7.989123775060453e-06,
+      "loss": 0.0605,
+      "step": 2487
+    },
+    {
+      "epoch": 2.2952029520295203,
+      "grad_norm": 0.40111276129887286,
+      "learning_rate": 7.969463509758254e-06,
+      "loss": 0.0661,
+      "step": 2488
+    },
+    {
+      "epoch": 2.2961254612546127,
+      "grad_norm": 0.4020667133890386,
+      "learning_rate": 7.949822877127072e-06,
+      "loss": 0.0605,
+      "step": 2489
+    },
+    {
+      "epoch": 2.2970479704797047,
+      "grad_norm": 0.34519922888637733,
+      "learning_rate": 7.930201899808475e-06,
+      "loss": 0.0562,
+      "step": 2490
+    },
+    {
+      "epoch": 2.297970479704797,
+      "grad_norm": 0.3623006619704617,
+      "learning_rate": 7.910600600421388e-06,
+      "loss": 0.0584,
+      "step": 2491
+    },
+    {
+      "epoch": 2.2988929889298895,
+      "grad_norm": 0.36246658853652025,
+      "learning_rate": 7.89101900156202e-06,
+      "loss": 0.062,
+      "step": 2492
+    },
+    {
+      "epoch": 2.2998154981549814,
+      "grad_norm": 0.4018204804803805,
+      "learning_rate": 7.871457125803896e-06,
+      "loss": 0.0643,
+      "step": 2493
+    },
+    {
+      "epoch": 2.300738007380074,
+      "grad_norm": 0.35305726451785513,
+      "learning_rate": 7.8519149956978e-06,
+      "loss": 0.0672,
+      "step": 2494
+    },
+    {
+      "epoch": 2.3016605166051662,
+      "grad_norm": 0.41273237368631244,
+      "learning_rate": 7.83239263377174e-06,
+      "loss": 0.0604,
+      "step": 2495
+    },
+    {
+      "epoch": 2.302583025830258,
+      "grad_norm": 0.45513339506555817,
+      "learning_rate": 7.812890062530942e-06,
+      "loss": 0.0664,
+      "step": 2496
+    },
+    {
+      "epoch": 2.3035055350553506,
+      "grad_norm": 0.4144182159517466,
+      "learning_rate": 7.793407304457836e-06,
+      "loss": 0.0661,
+      "step": 2497
+    },
+    {
+      "epoch": 2.304428044280443,
+      "grad_norm": 0.35440525157499714,
+      "learning_rate": 7.773944382011977e-06,
+      "loss": 0.0687,
+      "step": 2498
+    },
+    {
+      "epoch": 2.305350553505535,
+      "grad_norm": 0.397688879723002,
+      "learning_rate": 7.754501317630079e-06,
+      "loss": 0.0635,
+      "step": 2499
+    },
+    {
+      "epoch": 2.3062730627306274,
+      "grad_norm": 0.3950733114212165,
+      "learning_rate": 7.735078133725961e-06,
+      "loss": 0.0676,
+      "step": 2500
+    },
+    {
+      "epoch": 2.3071955719557193,
+      "grad_norm": 0.3390242865811373,
+      "learning_rate": 7.715674852690511e-06,
+      "loss": 0.0553,
+      "step": 2501
+    },
+    {
+      "epoch": 2.3081180811808117,
+      "grad_norm": 0.4163539201466788,
+      "learning_rate": 7.696291496891683e-06,
+      "loss": 0.0631,
+      "step": 2502
+    },
+    {
+      "epoch": 2.309040590405904,
+      "grad_norm": 0.40064556557483877,
+      "learning_rate": 7.67692808867447e-06,
+      "loss": 0.072,
+      "step": 2503
+    },
+    {
+      "epoch": 2.3099630996309966,
+      "grad_norm": 0.3977443109567292,
+      "learning_rate": 7.657584650360847e-06,
+      "loss": 0.0636,
+      "step": 2504
+    },
+    {
+      "epoch": 2.3108856088560885,
+      "grad_norm": 0.3558820806014279,
+      "learning_rate": 7.638261204249784e-06,
+      "loss": 0.0637,
+      "step": 2505
+    },
+    {
+      "epoch": 2.311808118081181,
+      "grad_norm": 0.40137296194830435,
+      "learning_rate": 7.618957772617211e-06,
+      "loss": 0.0639,
+      "step": 2506
+    },
+    {
+      "epoch": 2.312730627306273,
+      "grad_norm": 0.4098971801635019,
+      "learning_rate": 7.599674377715957e-06,
+      "loss": 0.0693,
+      "step": 2507
+    },
+    {
+      "epoch": 2.3136531365313653,
+      "grad_norm": 0.4300474601558293,
+      "learning_rate": 7.580411041775779e-06,
+      "loss": 0.0729,
+      "step": 2508
+    },
+    {
+      "epoch": 2.3145756457564577,
+      "grad_norm": 0.3720867720294639,
+      "learning_rate": 7.561167787003312e-06,
+      "loss": 0.0555,
+      "step": 2509
+    },
+    {
+      "epoch": 2.3154981549815496,
+      "grad_norm": 0.38623457849609255,
+      "learning_rate": 7.541944635582012e-06,
+      "loss": 0.0562,
+      "step": 2510
+    },
+    {
+      "epoch": 2.316420664206642,
+      "grad_norm": 0.4165238986568089,
+      "learning_rate": 7.522741609672193e-06,
+      "loss": 0.057,
+      "step": 2511
+    },
+    {
+      "epoch": 2.3173431734317345,
+      "grad_norm": 0.4637665834518347,
+      "learning_rate": 7.503558731410959e-06,
+      "loss": 0.0628,
+      "step": 2512
+    },
+    {
+      "epoch": 2.3182656826568264,
+      "grad_norm": 0.39517469580622855,
+      "learning_rate": 7.484396022912168e-06,
+      "loss": 0.0525,
+      "step": 2513
+    },
+    {
+      "epoch": 2.319188191881919,
+      "grad_norm": 0.4469403423279296,
+      "learning_rate": 7.465253506266454e-06,
+      "loss": 0.078,
+      "step": 2514
+    },
+    {
+      "epoch": 2.3201107011070112,
+      "grad_norm": 0.3546576408290238,
+      "learning_rate": 7.446131203541168e-06,
+      "loss": 0.0545,
+      "step": 2515
+    },
+    {
+      "epoch": 2.321033210332103,
+      "grad_norm": 0.35910317163594696,
+      "learning_rate": 7.427029136780333e-06,
+      "loss": 0.059,
+      "step": 2516
+    },
+    {
+      "epoch": 2.3219557195571956,
+      "grad_norm": 0.37292966043796955,
+      "learning_rate": 7.40794732800468e-06,
+      "loss": 0.07,
+      "step": 2517
+    },
+    {
+      "epoch": 2.322878228782288,
+      "grad_norm": 0.4253684481855254,
+      "learning_rate": 7.388885799211573e-06,
+      "loss": 0.0708,
+      "step": 2518
+    },
+    {
+      "epoch": 2.32380073800738,
+      "grad_norm": 0.34809959850854133,
+      "learning_rate": 7.369844572374981e-06,
+      "loss": 0.0588,
+      "step": 2519
+    },
+    {
+      "epoch": 2.3247232472324724,
+      "grad_norm": 0.3662135905594016,
+      "learning_rate": 7.350823669445495e-06,
+      "loss": 0.0582,
+      "step": 2520
+    },
+    {
+      "epoch": 2.3256457564575648,
+      "grad_norm": 0.3640604087471566,
+      "learning_rate": 7.3318231123502666e-06,
+      "loss": 0.0633,
+      "step": 2521
+    },
+    {
+      "epoch": 2.3265682656826567,
+      "grad_norm": 0.40467998469740746,
+      "learning_rate": 7.312842922992977e-06,
+      "loss": 0.0715,
+      "step": 2522
+    },
+    {
+      "epoch": 2.327490774907749,
+      "grad_norm": 0.36356399996561506,
+      "learning_rate": 7.293883123253861e-06,
+      "loss": 0.0536,
+      "step": 2523
+    },
+    {
+      "epoch": 2.328413284132841,
+      "grad_norm": 0.42665396685165385,
+      "learning_rate": 7.2749437349896115e-06,
+      "loss": 0.0629,
+      "step": 2524
+    },
+    {
+      "epoch": 2.3293357933579335,
+      "grad_norm": 0.40459888743575806,
+      "learning_rate": 7.256024780033418e-06,
+      "loss": 0.0791,
+      "step": 2525
+    },
+    {
+      "epoch": 2.330258302583026,
+      "grad_norm": 0.4017198772300197,
+      "learning_rate": 7.237126280194914e-06,
+      "loss": 0.0579,
+      "step": 2526
+    },
+    {
+      "epoch": 2.3311808118081183,
+      "grad_norm": 0.4056286111976866,
+      "learning_rate": 7.218248257260127e-06,
+      "loss": 0.0623,
+      "step": 2527
+    },
+    {
+      "epoch": 2.3321033210332103,
+      "grad_norm": 0.35597187276662773,
+      "learning_rate": 7.199390732991504e-06,
+      "loss": 0.0525,
+      "step": 2528
+    },
+    {
+      "epoch": 2.3330258302583027,
+      "grad_norm": 0.4092432831485503,
+      "learning_rate": 7.180553729127862e-06,
+      "loss": 0.0576,
+      "step": 2529
+    },
+    {
+      "epoch": 2.3339483394833946,
+      "grad_norm": 0.41150000904800804,
+      "learning_rate": 7.1617372673843354e-06,
+      "loss": 0.0628,
+      "step": 2530
+    },
+    {
+      "epoch": 2.334870848708487,
+      "grad_norm": 0.36975495289691623,
+      "learning_rate": 7.142941369452411e-06,
+      "loss": 0.0673,
+      "step": 2531
+    },
+    {
+      "epoch": 2.3357933579335795,
+      "grad_norm": 0.39441684219402856,
+      "learning_rate": 7.124166056999854e-06,
+      "loss": 0.062,
+      "step": 2532
+    },
+    {
+      "epoch": 2.3367158671586714,
+      "grad_norm": 0.4452527917351628,
+      "learning_rate": 7.105411351670691e-06,
+      "loss": 0.0764,
+      "step": 2533
+    },
+    {
+      "epoch": 2.337638376383764,
+      "grad_norm": 0.4404555798940548,
+      "learning_rate": 7.086677275085205e-06,
+      "loss": 0.0584,
+      "step": 2534
+    },
+    {
+      "epoch": 2.338560885608856,
+      "grad_norm": 0.39590190275406095,
+      "learning_rate": 7.0679638488399035e-06,
+      "loss": 0.0718,
+      "step": 2535
+    },
+    {
+      "epoch": 2.339483394833948,
+      "grad_norm": 0.3960244041804856,
+      "learning_rate": 7.049271094507465e-06,
+      "loss": 0.0686,
+      "step": 2536
+    },
+    {
+      "epoch": 2.3404059040590406,
+      "grad_norm": 0.49858709489915143,
+      "learning_rate": 7.030599033636759e-06,
+      "loss": 0.0671,
+      "step": 2537
+    },
+    {
+      "epoch": 2.341328413284133,
+      "grad_norm": 0.40343009520783757,
+      "learning_rate": 7.011947687752804e-06,
+      "loss": 0.0717,
+      "step": 2538
+    },
+    {
+      "epoch": 2.342250922509225,
+      "grad_norm": 0.36482715970289087,
+      "learning_rate": 6.993317078356709e-06,
+      "loss": 0.0589,
+      "step": 2539
+    },
+    {
+      "epoch": 2.3431734317343174,
+      "grad_norm": 0.3615894997654129,
+      "learning_rate": 6.9747072269257054e-06,
+      "loss": 0.0564,
+      "step": 2540
+    },
+    {
+      "epoch": 2.3440959409594098,
+      "grad_norm": 0.3889981031366702,
+      "learning_rate": 6.956118154913096e-06,
+      "loss": 0.0615,
+      "step": 2541
+    },
+    {
+      "epoch": 2.3450184501845017,
+      "grad_norm": 0.4066475196714116,
+      "learning_rate": 6.937549883748201e-06,
+      "loss": 0.0722,
+      "step": 2542
+    },
+    {
+      "epoch": 2.345940959409594,
+      "grad_norm": 0.3538875333416918,
+      "learning_rate": 6.919002434836389e-06,
+      "loss": 0.0576,
+      "step": 2543
+    },
+    {
+      "epoch": 2.3468634686346865,
+      "grad_norm": 0.4085945838979749,
+      "learning_rate": 6.900475829559022e-06,
+      "loss": 0.0638,
+      "step": 2544
+    },
+    {
+      "epoch": 2.3477859778597785,
+      "grad_norm": 0.3735933754064253,
+      "learning_rate": 6.881970089273418e-06,
+      "loss": 0.0601,
+      "step": 2545
+    },
+    {
+      "epoch": 2.348708487084871,
+      "grad_norm": 0.3962004713822194,
+      "learning_rate": 6.863485235312853e-06,
+      "loss": 0.0539,
+      "step": 2546
+    },
+    {
+      "epoch": 2.349630996309963,
+      "grad_norm": 0.3844502751625651,
+      "learning_rate": 6.845021288986531e-06,
+      "loss": 0.063,
+      "step": 2547
+    },
+    {
+      "epoch": 2.3505535055350553,
+      "grad_norm": 0.4153282393300872,
+      "learning_rate": 6.826578271579537e-06,
+      "loss": 0.0629,
+      "step": 2548
+    },
+    {
+      "epoch": 2.3514760147601477,
+      "grad_norm": 0.4150414688277971,
+      "learning_rate": 6.8081562043528445e-06,
+      "loss": 0.0643,
+      "step": 2549
+    },
+    {
+      "epoch": 2.35239852398524,
+      "grad_norm": 0.5368005897409298,
+      "learning_rate": 6.789755108543275e-06,
+      "loss": 0.0714,
+      "step": 2550
+    },
+    {
+      "epoch": 2.353321033210332,
+      "grad_norm": 0.4053169220112357,
+      "learning_rate": 6.771375005363459e-06,
+      "loss": 0.061,
+      "step": 2551
+    },
+    {
+      "epoch": 2.3542435424354244,
+      "grad_norm": 0.3974430160467046,
+      "learning_rate": 6.753015916001842e-06,
+      "loss": 0.0629,
+      "step": 2552
+    },
+    {
+      "epoch": 2.3551660516605164,
+      "grad_norm": 0.36042512610834304,
+      "learning_rate": 6.7346778616226515e-06,
+      "loss": 0.0547,
+      "step": 2553
+    },
+    {
+      "epoch": 2.356088560885609,
+      "grad_norm": 0.44978400097506505,
+      "learning_rate": 6.716360863365837e-06,
+      "loss": 0.062,
+      "step": 2554
+    },
+    {
+      "epoch": 2.357011070110701,
+      "grad_norm": 0.4613801197544788,
+      "learning_rate": 6.698064942347098e-06,
+      "loss": 0.0661,
+      "step": 2555
+    },
+    {
+      "epoch": 2.357933579335793,
+      "grad_norm": 0.42809481167463337,
+      "learning_rate": 6.6797901196578475e-06,
+      "loss": 0.0609,
+      "step": 2556
+    },
+    {
+      "epoch": 2.3588560885608856,
+      "grad_norm": 0.4222186759732147,
+      "learning_rate": 6.661536416365133e-06,
+      "loss": 0.062,
+      "step": 2557
+    },
+    {
+      "epoch": 2.359778597785978,
+      "grad_norm": 0.392109861648142,
+      "learning_rate": 6.643303853511707e-06,
+      "loss": 0.0656,
+      "step": 2558
+    },
+    {
+      "epoch": 2.36070110701107,
+      "grad_norm": 0.350509019582078,
+      "learning_rate": 6.625092452115908e-06,
+      "loss": 0.0599,
+      "step": 2559
+    },
+    {
+      "epoch": 2.3616236162361623,
+      "grad_norm": 0.37873421265389096,
+      "learning_rate": 6.606902233171711e-06,
+      "loss": 0.0636,
+      "step": 2560
+    },
+    {
+      "epoch": 2.3625461254612548,
+      "grad_norm": 0.46300069978001546,
+      "learning_rate": 6.58873321764866e-06,
+      "loss": 0.0634,
+      "step": 2561
+    },
+    {
+      "epoch": 2.3634686346863467,
+      "grad_norm": 0.4014439079622017,
+      "learning_rate": 6.570585426491846e-06,
+      "loss": 0.0645,
+      "step": 2562
+    },
+    {
+      "epoch": 2.364391143911439,
+      "grad_norm": 0.3568217003146156,
+      "learning_rate": 6.552458880621909e-06,
+      "loss": 0.0646,
+      "step": 2563
+    },
+    {
+      "epoch": 2.3653136531365315,
+      "grad_norm": 0.38169279440631276,
+      "learning_rate": 6.534353600934997e-06,
+      "loss": 0.0646,
+      "step": 2564
+    },
+    {
+      "epoch": 2.3662361623616235,
+      "grad_norm": 0.3614823307055291,
+      "learning_rate": 6.5162696083027275e-06,
+      "loss": 0.063,
+      "step": 2565
+    },
+    {
+      "epoch": 2.367158671586716,
+      "grad_norm": 0.41450615121302437,
+      "learning_rate": 6.498206923572189e-06,
+      "loss": 0.0604,
+      "step": 2566
+    },
+    {
+      "epoch": 2.3680811808118083,
+      "grad_norm": 0.3888929140268929,
+      "learning_rate": 6.480165567565913e-06,
+      "loss": 0.0636,
+      "step": 2567
+    },
+    {
+      "epoch": 2.3690036900369003,
+      "grad_norm": 0.36675927843802164,
+      "learning_rate": 6.4621455610818225e-06,
+      "loss": 0.0561,
+      "step": 2568
+    },
+    {
+      "epoch": 2.3699261992619927,
+      "grad_norm": 0.4407090305739821,
+      "learning_rate": 6.4441469248932515e-06,
+      "loss": 0.0663,
+      "step": 2569
+    },
+    {
+      "epoch": 2.3708487084870846,
+      "grad_norm": 0.37352908950202196,
+      "learning_rate": 6.426169679748892e-06,
+      "loss": 0.0726,
+      "step": 2570
+    },
+    {
+      "epoch": 2.371771217712177,
+      "grad_norm": 0.3407934217583342,
+      "learning_rate": 6.40821384637276e-06,
+      "loss": 0.0579,
+      "step": 2571
+    },
+    {
+      "epoch": 2.3726937269372694,
+      "grad_norm": 0.3496884380757048,
+      "learning_rate": 6.390279445464209e-06,
+      "loss": 0.0537,
+      "step": 2572
+    },
+    {
+      "epoch": 2.373616236162362,
+      "grad_norm": 0.4210055489309458,
+      "learning_rate": 6.3723664976978875e-06,
+      "loss": 0.0622,
+      "step": 2573
+    },
+    {
+      "epoch": 2.374538745387454,
+      "grad_norm": 0.41545445097543465,
+      "learning_rate": 6.354475023723686e-06,
+      "loss": 0.0672,
+      "step": 2574
+    },
+    {
+      "epoch": 2.375461254612546,
+      "grad_norm": 0.40184438229982183,
+      "learning_rate": 6.336605044166763e-06,
+      "loss": 0.0665,
+      "step": 2575
+    },
+    {
+      "epoch": 2.376383763837638,
+      "grad_norm": 0.35845198960242597,
+      "learning_rate": 6.318756579627508e-06,
+      "loss": 0.057,
+      "step": 2576
+    },
+    {
+      "epoch": 2.3773062730627306,
+      "grad_norm": 0.38322370969285097,
+      "learning_rate": 6.30092965068147e-06,
+      "loss": 0.0697,
+      "step": 2577
+    },
+    {
+      "epoch": 2.378228782287823,
+      "grad_norm": 0.4107658977093919,
+      "learning_rate": 6.283124277879407e-06,
+      "loss": 0.0614,
+      "step": 2578
+    },
+    {
+      "epoch": 2.3791512915129154,
+      "grad_norm": 0.3804930863220513,
+      "learning_rate": 6.2653404817472226e-06,
+      "loss": 0.065,
+      "step": 2579
+    },
+    {
+      "epoch": 2.3800738007380073,
+      "grad_norm": 0.3437078010893973,
+      "learning_rate": 6.247578282785929e-06,
+      "loss": 0.0594,
+      "step": 2580
+    },
+    {
+      "epoch": 2.3809963099630997,
+      "grad_norm": 0.4353303768580871,
+      "learning_rate": 6.229837701471644e-06,
+      "loss": 0.0671,
+      "step": 2581
+    },
+    {
+      "epoch": 2.3819188191881917,
+      "grad_norm": 0.385517329552986,
+      "learning_rate": 6.212118758255595e-06,
+      "loss": 0.054,
+      "step": 2582
+    },
+    {
+      "epoch": 2.382841328413284,
+      "grad_norm": 0.4092193165406145,
+      "learning_rate": 6.194421473564033e-06,
+      "loss": 0.0591,
+      "step": 2583
+    },
+    {
+      "epoch": 2.3837638376383765,
+      "grad_norm": 0.3979284458536561,
+      "learning_rate": 6.176745867798234e-06,
+      "loss": 0.0675,
+      "step": 2584
+    },
+    {
+      "epoch": 2.3846863468634685,
+      "grad_norm": 0.38250678890596823,
+      "learning_rate": 6.159091961334531e-06,
+      "loss": 0.0676,
+      "step": 2585
+    },
+    {
+      "epoch": 2.385608856088561,
+      "grad_norm": 0.41128989688310613,
+      "learning_rate": 6.141459774524194e-06,
+      "loss": 0.0622,
+      "step": 2586
+    },
+    {
+      "epoch": 2.3865313653136533,
+      "grad_norm": 0.3592014079130497,
+      "learning_rate": 6.123849327693462e-06,
+      "loss": 0.0534,
+      "step": 2587
+    },
+    {
+      "epoch": 2.3874538745387452,
+      "grad_norm": 0.4048581447812899,
+      "learning_rate": 6.106260641143546e-06,
+      "loss": 0.0662,
+      "step": 2588
+    },
+    {
+      "epoch": 2.3883763837638377,
+      "grad_norm": 0.3616056918329504,
+      "learning_rate": 6.0886937351505276e-06,
+      "loss": 0.0544,
+      "step": 2589
+    },
+    {
+      "epoch": 2.38929889298893,
+      "grad_norm": 0.3542718481167006,
+      "learning_rate": 6.0711486299654095e-06,
+      "loss": 0.0587,
+      "step": 2590
+    },
+    {
+      "epoch": 2.390221402214022,
+      "grad_norm": 0.43225454759792764,
+      "learning_rate": 6.053625345814062e-06,
+      "loss": 0.0551,
+      "step": 2591
+    },
+    {
+      "epoch": 2.3911439114391144,
+      "grad_norm": 0.34743105596036417,
+      "learning_rate": 6.036123902897172e-06,
+      "loss": 0.0569,
+      "step": 2592
+    },
+    {
+      "epoch": 2.392066420664207,
+      "grad_norm": 0.4161427779221691,
+      "learning_rate": 6.018644321390288e-06,
+      "loss": 0.0609,
+      "step": 2593
+    },
+    {
+      "epoch": 2.392988929889299,
+      "grad_norm": 0.3695331552011525,
+      "learning_rate": 6.001186621443719e-06,
+      "loss": 0.0563,
+      "step": 2594
+    },
+    {
+      "epoch": 2.393911439114391,
+      "grad_norm": 0.396624124983913,
+      "learning_rate": 5.983750823182574e-06,
+      "loss": 0.0635,
+      "step": 2595
+    },
+    {
+      "epoch": 2.3948339483394836,
+      "grad_norm": 0.3642112004788074,
+      "learning_rate": 5.966336946706716e-06,
+      "loss": 0.0614,
+      "step": 2596
+    },
+    {
+      "epoch": 2.3957564575645756,
+      "grad_norm": 0.3603512900272667,
+      "learning_rate": 5.948945012090709e-06,
+      "loss": 0.0542,
+      "step": 2597
+    },
+    {
+      "epoch": 2.396678966789668,
+      "grad_norm": 0.3975643668719549,
+      "learning_rate": 5.931575039383852e-06,
+      "loss": 0.07,
+      "step": 2598
+    },
+    {
+      "epoch": 2.39760147601476,
+      "grad_norm": 0.39638831804174796,
+      "learning_rate": 5.914227048610121e-06,
+      "loss": 0.0582,
+      "step": 2599
+    },
+    {
+      "epoch": 2.3985239852398523,
+      "grad_norm": 0.35663381843101466,
+      "learning_rate": 5.896901059768134e-06,
+      "loss": 0.0539,
+      "step": 2600
+    },
+    {
+      "epoch": 2.3994464944649447,
+      "grad_norm": 0.3962885681918028,
+      "learning_rate": 5.87959709283116e-06,
+      "loss": 0.0661,
+      "step": 2601
+    },
+    {
+      "epoch": 2.400369003690037,
+      "grad_norm": 0.3626899065335606,
+      "learning_rate": 5.86231516774709e-06,
+      "loss": 0.0616,
+      "step": 2602
+    },
+    {
+      "epoch": 2.401291512915129,
+      "grad_norm": 0.449109122858332,
+      "learning_rate": 5.845055304438377e-06,
+      "loss": 0.0694,
+      "step": 2603
+    },
+    {
+      "epoch": 2.4022140221402215,
+      "grad_norm": 0.37200127550749645,
+      "learning_rate": 5.827817522802065e-06,
+      "loss": 0.0636,
+      "step": 2604
+    },
+    {
+      "epoch": 2.4031365313653135,
+      "grad_norm": 0.3735178974919831,
+      "learning_rate": 5.810601842709743e-06,
+      "loss": 0.064,
+      "step": 2605
+    },
+    {
+      "epoch": 2.404059040590406,
+      "grad_norm": 0.38994941566994723,
+      "learning_rate": 5.793408284007501e-06,
+      "loss": 0.0603,
+      "step": 2606
+    },
+    {
+      "epoch": 2.4049815498154983,
+      "grad_norm": 0.3967380432626151,
+      "learning_rate": 5.776236866515947e-06,
+      "loss": 0.0685,
+      "step": 2607
+    },
+    {
+      "epoch": 2.4059040590405902,
+      "grad_norm": 0.3459901203489991,
+      "learning_rate": 5.759087610030167e-06,
+      "loss": 0.0621,
+      "step": 2608
+    },
+    {
+      "epoch": 2.4068265682656826,
+      "grad_norm": 0.4065639509219356,
+      "learning_rate": 5.741960534319677e-06,
+      "loss": 0.0578,
+      "step": 2609
+    },
+    {
+      "epoch": 2.407749077490775,
+      "grad_norm": 0.44235335716406815,
+      "learning_rate": 5.724855659128442e-06,
+      "loss": 0.062,
+      "step": 2610
+    },
+    {
+      "epoch": 2.408671586715867,
+      "grad_norm": 0.3263850063837015,
+      "learning_rate": 5.707773004174841e-06,
+      "loss": 0.0569,
+      "step": 2611
+    },
+    {
+      "epoch": 2.4095940959409594,
+      "grad_norm": 0.40541025521987767,
+      "learning_rate": 5.6907125891516115e-06,
+      "loss": 0.0598,
+      "step": 2612
+    },
+    {
+      "epoch": 2.410516605166052,
+      "grad_norm": 0.4199545942580867,
+      "learning_rate": 5.673674433725873e-06,
+      "loss": 0.0672,
+      "step": 2613
+    },
+    {
+      "epoch": 2.411439114391144,
+      "grad_norm": 0.3746463616462845,
+      "learning_rate": 5.656658557539091e-06,
+      "loss": 0.0649,
+      "step": 2614
+    },
+    {
+      "epoch": 2.412361623616236,
+      "grad_norm": 0.3579648795946127,
+      "learning_rate": 5.639664980207024e-06,
+      "loss": 0.0569,
+      "step": 2615
+    },
+    {
+      "epoch": 2.4132841328413286,
+      "grad_norm": 0.39194718088289654,
+      "learning_rate": 5.622693721319727e-06,
+      "loss": 0.0549,
+      "step": 2616
+    },
+    {
+      "epoch": 2.4142066420664205,
+      "grad_norm": 0.3700713222546607,
+      "learning_rate": 5.605744800441562e-06,
+      "loss": 0.062,
+      "step": 2617
+    },
+    {
+      "epoch": 2.415129151291513,
+      "grad_norm": 0.3816651720819482,
+      "learning_rate": 5.588818237111102e-06,
+      "loss": 0.0538,
+      "step": 2618
+    },
+    {
+      "epoch": 2.4160516605166054,
+      "grad_norm": 0.3697697674957319,
+      "learning_rate": 5.57191405084114e-06,
+      "loss": 0.0668,
+      "step": 2619
+    },
+    {
+      "epoch": 2.4169741697416973,
+      "grad_norm": 0.3807310616054102,
+      "learning_rate": 5.5550322611187254e-06,
+      "loss": 0.0616,
+      "step": 2620
+    },
+    {
+      "epoch": 2.4178966789667897,
+      "grad_norm": 0.47064848383115176,
+      "learning_rate": 5.538172887405038e-06,
+      "loss": 0.0625,
+      "step": 2621
+    },
+    {
+      "epoch": 2.4188191881918817,
+      "grad_norm": 0.44060259255703815,
+      "learning_rate": 5.52133594913542e-06,
+      "loss": 0.0678,
+      "step": 2622
+    },
+    {
+      "epoch": 2.419741697416974,
+      "grad_norm": 0.4099851460303192,
+      "learning_rate": 5.5045214657193925e-06,
+      "loss": 0.0661,
+      "step": 2623
+    },
+    {
+      "epoch": 2.4206642066420665,
+      "grad_norm": 0.401475210690167,
+      "learning_rate": 5.487729456540547e-06,
+      "loss": 0.059,
+      "step": 2624
+    },
+    {
+      "epoch": 2.421586715867159,
+      "grad_norm": 0.35716427098125575,
+      "learning_rate": 5.470959940956572e-06,
+      "loss": 0.0595,
+      "step": 2625
+    },
+    {
+      "epoch": 2.422509225092251,
+      "grad_norm": 0.31909506874099236,
+      "learning_rate": 5.454212938299255e-06,
+      "loss": 0.0514,
+      "step": 2626
+    },
+    {
+      "epoch": 2.4234317343173433,
+      "grad_norm": 0.38220159344635984,
+      "learning_rate": 5.437488467874407e-06,
+      "loss": 0.0542,
+      "step": 2627
+    },
+    {
+      "epoch": 2.4243542435424352,
+      "grad_norm": 0.3859726021013262,
+      "learning_rate": 5.4207865489618565e-06,
+      "loss": 0.0585,
+      "step": 2628
+    },
+    {
+      "epoch": 2.4252767527675276,
+      "grad_norm": 0.42450114803191763,
+      "learning_rate": 5.404107200815456e-06,
+      "loss": 0.0647,
+      "step": 2629
+    },
+    {
+      "epoch": 2.42619926199262,
+      "grad_norm": 0.3987226969379201,
+      "learning_rate": 5.387450442663025e-06,
+      "loss": 0.0538,
+      "step": 2630
+    },
+    {
+      "epoch": 2.427121771217712,
+      "grad_norm": 0.3968955400530196,
+      "learning_rate": 5.370816293706357e-06,
+      "loss": 0.0572,
+      "step": 2631
+    },
+    {
+      "epoch": 2.4280442804428044,
+      "grad_norm": 0.40159835731652965,
+      "learning_rate": 5.354204773121155e-06,
+      "loss": 0.0553,
+      "step": 2632
+    },
+    {
+      "epoch": 2.428966789667897,
+      "grad_norm": 0.35205301145237317,
+      "learning_rate": 5.337615900057058e-06,
+      "loss": 0.0586,
+      "step": 2633
+    },
+    {
+      "epoch": 2.4298892988929888,
+      "grad_norm": 0.37210584463265833,
+      "learning_rate": 5.3210496936376e-06,
+      "loss": 0.0543,
+      "step": 2634
+    },
+    {
+      "epoch": 2.430811808118081,
+      "grad_norm": 0.3927689300274618,
+      "learning_rate": 5.304506172960161e-06,
+      "loss": 0.0596,
+      "step": 2635
+    },
+    {
+      "epoch": 2.4317343173431736,
+      "grad_norm": 0.3976834621136443,
+      "learning_rate": 5.287985357095989e-06,
+      "loss": 0.0634,
+      "step": 2636
+    },
+    {
+      "epoch": 2.4326568265682655,
+      "grad_norm": 0.3850666745291003,
+      "learning_rate": 5.271487265090163e-06,
+      "loss": 0.0613,
+      "step": 2637
+    },
+    {
+      "epoch": 2.433579335793358,
+      "grad_norm": 0.3931373334443755,
+      "learning_rate": 5.2550119159615414e-06,
+      "loss": 0.055,
+      "step": 2638
+    },
+    {
+      "epoch": 2.4345018450184504,
+      "grad_norm": 0.4441330734470427,
+      "learning_rate": 5.238559328702783e-06,
+      "loss": 0.0607,
+      "step": 2639
+    },
+    {
+      "epoch": 2.4354243542435423,
+      "grad_norm": 0.43134887293841717,
+      "learning_rate": 5.222129522280314e-06,
+      "loss": 0.0659,
+      "step": 2640
+    },
+    {
+      "epoch": 2.4363468634686347,
+      "grad_norm": 0.42422088503610306,
+      "learning_rate": 5.205722515634276e-06,
+      "loss": 0.0685,
+      "step": 2641
+    },
+    {
+      "epoch": 2.437269372693727,
+      "grad_norm": 0.4305265594810753,
+      "learning_rate": 5.189338327678541e-06,
+      "loss": 0.0615,
+      "step": 2642
+    },
+    {
+      "epoch": 2.438191881918819,
+      "grad_norm": 0.42859758627537825,
+      "learning_rate": 5.172976977300687e-06,
+      "loss": 0.0666,
+      "step": 2643
+    },
+    {
+      "epoch": 2.4391143911439115,
+      "grad_norm": 0.3653846943927543,
+      "learning_rate": 5.156638483361934e-06,
+      "loss": 0.053,
+      "step": 2644
+    },
+    {
+      "epoch": 2.4400369003690034,
+      "grad_norm": 0.47522142375272125,
+      "learning_rate": 5.140322864697183e-06,
+      "loss": 0.069,
+      "step": 2645
+    },
+    {
+      "epoch": 2.440959409594096,
+      "grad_norm": 0.39587718312963355,
+      "learning_rate": 5.124030140114958e-06,
+      "loss": 0.055,
+      "step": 2646
+    },
+    {
+      "epoch": 2.4418819188191883,
+      "grad_norm": 0.39370537094439917,
+      "learning_rate": 5.107760328397371e-06,
+      "loss": 0.0589,
+      "step": 2647
+    },
+    {
+      "epoch": 2.4428044280442807,
+      "grad_norm": 0.42584974603072745,
+      "learning_rate": 5.091513448300142e-06,
+      "loss": 0.0717,
+      "step": 2648
+    },
+    {
+      "epoch": 2.4437269372693726,
+      "grad_norm": 0.4181594731122546,
+      "learning_rate": 5.075289518552562e-06,
+      "loss": 0.0665,
+      "step": 2649
+    },
+    {
+      "epoch": 2.444649446494465,
+      "grad_norm": 0.390316797744379,
+      "learning_rate": 5.059088557857436e-06,
+      "loss": 0.0607,
+      "step": 2650
+    },
+    {
+      "epoch": 2.445571955719557,
+      "grad_norm": 0.38943012941645383,
+      "learning_rate": 5.0429105848911e-06,
+      "loss": 0.0651,
+      "step": 2651
+    },
+    {
+      "epoch": 2.4464944649446494,
+      "grad_norm": 0.383197867209574,
+      "learning_rate": 5.0267556183034195e-06,
+      "loss": 0.0508,
+      "step": 2652
+    },
+    {
+      "epoch": 2.447416974169742,
+      "grad_norm": 0.3843119449443553,
+      "learning_rate": 5.010623676717704e-06,
+      "loss": 0.058,
+      "step": 2653
+    },
+    {
+      "epoch": 2.4483394833948338,
+      "grad_norm": 0.3707663558261996,
+      "learning_rate": 4.994514778730719e-06,
+      "loss": 0.06,
+      "step": 2654
+    },
+    {
+      "epoch": 2.449261992619926,
+      "grad_norm": 0.4021464290126931,
+      "learning_rate": 4.9784289429127e-06,
+      "loss": 0.0611,
+      "step": 2655
+    },
+    {
+      "epoch": 2.4501845018450186,
+      "grad_norm": 0.35954260644458425,
+      "learning_rate": 4.9623661878072635e-06,
+      "loss": 0.055,
+      "step": 2656
+    },
+    {
+      "epoch": 2.4511070110701105,
+      "grad_norm": 0.4048880210682503,
+      "learning_rate": 4.946326531931417e-06,
+      "loss": 0.0611,
+      "step": 2657
+    },
+    {
+      "epoch": 2.452029520295203,
+      "grad_norm": 0.3987346958997723,
+      "learning_rate": 4.930309993775578e-06,
+      "loss": 0.0654,
+      "step": 2658
+    },
+    {
+      "epoch": 2.4529520295202953,
+      "grad_norm": 0.38942327732157767,
+      "learning_rate": 4.914316591803475e-06,
+      "loss": 0.074,
+      "step": 2659
+    },
+    {
+      "epoch": 2.4538745387453873,
+      "grad_norm": 0.40193876135548995,
+      "learning_rate": 4.8983463444521705e-06,
+      "loss": 0.06,
+      "step": 2660
+    },
+    {
+      "epoch": 2.4547970479704797,
+      "grad_norm": 0.38751800666766895,
+      "learning_rate": 4.882399270132052e-06,
+      "loss": 0.0623,
+      "step": 2661
+    },
+    {
+      "epoch": 2.455719557195572,
+      "grad_norm": 0.34738897235755367,
+      "learning_rate": 4.866475387226788e-06,
+      "loss": 0.0597,
+      "step": 2662
+    },
+    {
+      "epoch": 2.456642066420664,
+      "grad_norm": 0.38054854983134784,
+      "learning_rate": 4.850574714093292e-06,
+      "loss": 0.0575,
+      "step": 2663
+    },
+    {
+      "epoch": 2.4575645756457565,
+      "grad_norm": 0.36201575679205095,
+      "learning_rate": 4.83469726906175e-06,
+      "loss": 0.0574,
+      "step": 2664
+    },
+    {
+      "epoch": 2.458487084870849,
+      "grad_norm": 0.35727369220907385,
+      "learning_rate": 4.8188430704355605e-06,
+      "loss": 0.056,
+      "step": 2665
+    },
+    {
+      "epoch": 2.459409594095941,
+      "grad_norm": 0.35982158666662806,
+      "learning_rate": 4.803012136491308e-06,
+      "loss": 0.0556,
+      "step": 2666
+    },
+    {
+      "epoch": 2.4603321033210332,
+      "grad_norm": 0.339409645190489,
+      "learning_rate": 4.78720448547878e-06,
+      "loss": 0.0566,
+      "step": 2667
+    },
+    {
+      "epoch": 2.4612546125461257,
+      "grad_norm": 0.3464214698695882,
+      "learning_rate": 4.771420135620919e-06,
+      "loss": 0.0539,
+      "step": 2668
+    },
+    {
+      "epoch": 2.4621771217712176,
+      "grad_norm": 0.36442133108465213,
+      "learning_rate": 4.755659105113788e-06,
+      "loss": 0.0552,
+      "step": 2669
+    },
+    {
+      "epoch": 2.46309963099631,
+      "grad_norm": 0.4399996919727431,
+      "learning_rate": 4.739921412126591e-06,
+      "loss": 0.0715,
+      "step": 2670
+    },
+    {
+      "epoch": 2.4640221402214024,
+      "grad_norm": 0.3852161273688586,
+      "learning_rate": 4.724207074801623e-06,
+      "loss": 0.0496,
+      "step": 2671
+    },
+    {
+      "epoch": 2.4649446494464944,
+      "grad_norm": 0.37967417283926874,
+      "learning_rate": 4.708516111254238e-06,
+      "loss": 0.0645,
+      "step": 2672
+    },
+    {
+      "epoch": 2.465867158671587,
+      "grad_norm": 0.35478180776500723,
+      "learning_rate": 4.692848539572866e-06,
+      "loss": 0.052,
+      "step": 2673
+    },
+    {
+      "epoch": 2.4667896678966788,
+      "grad_norm": 0.403051447224795,
+      "learning_rate": 4.677204377818961e-06,
+      "loss": 0.0608,
+      "step": 2674
+    },
+    {
+      "epoch": 2.467712177121771,
+      "grad_norm": 0.3751198699123934,
+      "learning_rate": 4.661583644026998e-06,
+      "loss": 0.0622,
+      "step": 2675
+    },
+    {
+      "epoch": 2.4686346863468636,
+      "grad_norm": 0.3808509667709742,
+      "learning_rate": 4.6459863562044264e-06,
+      "loss": 0.0626,
+      "step": 2676
+    },
+    {
+      "epoch": 2.469557195571956,
+      "grad_norm": 0.40468818913039606,
+      "learning_rate": 4.630412532331685e-06,
+      "loss": 0.0596,
+      "step": 2677
+    },
+    {
+      "epoch": 2.470479704797048,
+      "grad_norm": 0.3998975816295429,
+      "learning_rate": 4.614862190362165e-06,
+      "loss": 0.0657,
+      "step": 2678
+    },
+    {
+      "epoch": 2.4714022140221403,
+      "grad_norm": 0.3994385358961541,
+      "learning_rate": 4.59933534822217e-06,
+      "loss": 0.0601,
+      "step": 2679
+    },
+    {
+      "epoch": 2.4723247232472323,
+      "grad_norm": 0.348068920835927,
+      "learning_rate": 4.583832023810925e-06,
+      "loss": 0.0496,
+      "step": 2680
+    },
+    {
+      "epoch": 2.4732472324723247,
+      "grad_norm": 0.37236078398166217,
+      "learning_rate": 4.56835223500055e-06,
+      "loss": 0.0581,
+      "step": 2681
+    },
+    {
+      "epoch": 2.474169741697417,
+      "grad_norm": 0.4182456478205192,
+      "learning_rate": 4.55289599963602e-06,
+      "loss": 0.0597,
+      "step": 2682
+    },
+    {
+      "epoch": 2.475092250922509,
+      "grad_norm": 0.4290296224964595,
+      "learning_rate": 4.537463335535161e-06,
+      "loss": 0.0617,
+      "step": 2683
+    },
+    {
+      "epoch": 2.4760147601476015,
+      "grad_norm": 0.4288200041721056,
+      "learning_rate": 4.52205426048864e-06,
+      "loss": 0.0683,
+      "step": 2684
+    },
+    {
+      "epoch": 2.476937269372694,
+      "grad_norm": 0.38425239725616517,
+      "learning_rate": 4.506668792259914e-06,
+      "loss": 0.0658,
+      "step": 2685
+    },
+    {
+      "epoch": 2.477859778597786,
+      "grad_norm": 0.36189081453617344,
+      "learning_rate": 4.491306948585219e-06,
+      "loss": 0.0603,
+      "step": 2686
+    },
+    {
+      "epoch": 2.4787822878228782,
+      "grad_norm": 0.3661967722872685,
+      "learning_rate": 4.475968747173592e-06,
+      "loss": 0.0549,
+      "step": 2687
+    },
+    {
+      "epoch": 2.4797047970479706,
+      "grad_norm": 0.39386173428974636,
+      "learning_rate": 4.460654205706785e-06,
+      "loss": 0.0509,
+      "step": 2688
+    },
+    {
+      "epoch": 2.4806273062730626,
+      "grad_norm": 0.4086739092191323,
+      "learning_rate": 4.4453633418392705e-06,
+      "loss": 0.0668,
+      "step": 2689
+    },
+    {
+      "epoch": 2.481549815498155,
+      "grad_norm": 0.44226148669322224,
+      "learning_rate": 4.430096173198259e-06,
+      "loss": 0.0698,
+      "step": 2690
+    },
+    {
+      "epoch": 2.4824723247232474,
+      "grad_norm": 0.41815992188414736,
+      "learning_rate": 4.414852717383616e-06,
+      "loss": 0.0707,
+      "step": 2691
+    },
+    {
+      "epoch": 2.4833948339483394,
+      "grad_norm": 0.40710245431221515,
+      "learning_rate": 4.3996329919678666e-06,
+      "loss": 0.0611,
+      "step": 2692
+    },
+    {
+      "epoch": 2.484317343173432,
+      "grad_norm": 0.3866036235460156,
+      "learning_rate": 4.384437014496215e-06,
+      "loss": 0.0663,
+      "step": 2693
+    },
+    {
+      "epoch": 2.485239852398524,
+      "grad_norm": 0.3925558660095124,
+      "learning_rate": 4.3692648024864585e-06,
+      "loss": 0.0621,
+      "step": 2694
+    },
+    {
+      "epoch": 2.486162361623616,
+      "grad_norm": 0.3898878289789644,
+      "learning_rate": 4.3541163734289955e-06,
+      "loss": 0.0631,
+      "step": 2695
+    },
+    {
+      "epoch": 2.4870848708487086,
+      "grad_norm": 0.34515468988767445,
+      "learning_rate": 4.33899174478683e-06,
+      "loss": 0.0559,
+      "step": 2696
+    },
+    {
+      "epoch": 2.4880073800738005,
+      "grad_norm": 0.38991615115843276,
+      "learning_rate": 4.323890933995517e-06,
+      "loss": 0.0648,
+      "step": 2697
+    },
+    {
+      "epoch": 2.488929889298893,
+      "grad_norm": 0.3567292574951863,
+      "learning_rate": 4.308813958463145e-06,
+      "loss": 0.0621,
+      "step": 2698
+    },
+    {
+      "epoch": 2.4898523985239853,
+      "grad_norm": 0.3756121323612728,
+      "learning_rate": 4.293760835570343e-06,
+      "loss": 0.0648,
+      "step": 2699
+    },
+    {
+      "epoch": 2.4907749077490777,
+      "grad_norm": 0.43625351673224455,
+      "learning_rate": 4.278731582670239e-06,
+      "loss": 0.0604,
+      "step": 2700
+    },
+    {
+      "epoch": 2.4916974169741697,
+      "grad_norm": 0.3841168021584028,
+      "learning_rate": 4.263726217088429e-06,
+      "loss": 0.0541,
+      "step": 2701
+    },
+    {
+      "epoch": 2.492619926199262,
+      "grad_norm": 0.3654561694274849,
+      "learning_rate": 4.248744756122986e-06,
+      "loss": 0.0544,
+      "step": 2702
+    },
+    {
+      "epoch": 2.493542435424354,
+      "grad_norm": 0.4111746551123488,
+      "learning_rate": 4.23378721704443e-06,
+      "loss": 0.0648,
+      "step": 2703
+    },
+    {
+      "epoch": 2.4944649446494465,
+      "grad_norm": 0.40144075585536443,
+      "learning_rate": 4.218853617095686e-06,
+      "loss": 0.069,
+      "step": 2704
+    },
+    {
+      "epoch": 2.495387453874539,
+      "grad_norm": 0.40482568402285934,
+      "learning_rate": 4.203943973492097e-06,
+      "loss": 0.0632,
+      "step": 2705
+    },
+    {
+      "epoch": 2.496309963099631,
+      "grad_norm": 0.38685996424627717,
+      "learning_rate": 4.189058303421392e-06,
+      "loss": 0.0605,
+      "step": 2706
+    },
+    {
+      "epoch": 2.4972324723247232,
+      "grad_norm": 0.30920485496123884,
+      "learning_rate": 4.1741966240436445e-06,
+      "loss": 0.0594,
+      "step": 2707
+    },
+    {
+      "epoch": 2.4981549815498156,
+      "grad_norm": 0.4536039150701168,
+      "learning_rate": 4.159358952491288e-06,
+      "loss": 0.0747,
+      "step": 2708
+    },
+    {
+      "epoch": 2.4990774907749076,
+      "grad_norm": 0.3733730049529338,
+      "learning_rate": 4.144545305869086e-06,
+      "loss": 0.0564,
+      "step": 2709
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.3554560516800568,
+      "learning_rate": 4.129755701254076e-06,
+      "loss": 0.0588,
+      "step": 2710
+    },
+    {
+      "epoch": 2.5009225092250924,
+      "grad_norm": 0.36966687007450383,
+      "learning_rate": 4.114990155695617e-06,
+      "loss": 0.0515,
+      "step": 2711
+    },
+    {
+      "epoch": 2.5018450184501844,
+      "grad_norm": 0.37330996084536244,
+      "learning_rate": 4.100248686215313e-06,
+      "loss": 0.0547,
+      "step": 2712
+    },
+    {
+      "epoch": 2.5027675276752768,
+      "grad_norm": 0.38314398749030215,
+      "learning_rate": 4.085531309807009e-06,
+      "loss": 0.0507,
+      "step": 2713
+    },
+    {
+      "epoch": 2.5036900369003687,
+      "grad_norm": 0.3501755082301952,
+      "learning_rate": 4.070838043436786e-06,
+      "loss": 0.0541,
+      "step": 2714
+    },
+    {
+      "epoch": 2.504612546125461,
+      "grad_norm": 0.36500233722105163,
+      "learning_rate": 4.056168904042934e-06,
+      "loss": 0.0579,
+      "step": 2715
+    },
+    {
+      "epoch": 2.5055350553505535,
+      "grad_norm": 0.3663050507266439,
+      "learning_rate": 4.041523908535916e-06,
+      "loss": 0.0627,
+      "step": 2716
+    },
+    {
+      "epoch": 2.506457564575646,
+      "grad_norm": 0.3795564027623454,
+      "learning_rate": 4.026903073798372e-06,
+      "loss": 0.0578,
+      "step": 2717
+    },
+    {
+      "epoch": 2.507380073800738,
+      "grad_norm": 0.3327657664195737,
+      "learning_rate": 4.012306416685088e-06,
+      "loss": 0.0597,
+      "step": 2718
+    },
+    {
+      "epoch": 2.5083025830258303,
+      "grad_norm": 0.39368371129688207,
+      "learning_rate": 3.997733954022986e-06,
+      "loss": 0.055,
+      "step": 2719
+    },
+    {
+      "epoch": 2.5092250922509223,
+      "grad_norm": 0.3744728045239547,
+      "learning_rate": 3.983185702611078e-06,
+      "loss": 0.0614,
+      "step": 2720
+    },
+    {
+      "epoch": 2.5101476014760147,
+      "grad_norm": 0.4271760817934105,
+      "learning_rate": 3.968661679220468e-06,
+      "loss": 0.0627,
+      "step": 2721
+    },
+    {
+      "epoch": 2.511070110701107,
+      "grad_norm": 0.40795233030440275,
+      "learning_rate": 3.954161900594361e-06,
+      "loss": 0.0652,
+      "step": 2722
+    },
+    {
+      "epoch": 2.5119926199261995,
+      "grad_norm": 0.4370878780434889,
+      "learning_rate": 3.9396863834479745e-06,
+      "loss": 0.0644,
+      "step": 2723
+    },
+    {
+      "epoch": 2.5129151291512914,
+      "grad_norm": 0.4126537341299734,
+      "learning_rate": 3.925235144468567e-06,
+      "loss": 0.0585,
+      "step": 2724
+    },
+    {
+      "epoch": 2.513837638376384,
+      "grad_norm": 0.35351145353358715,
+      "learning_rate": 3.9108082003154325e-06,
+      "loss": 0.0681,
+      "step": 2725
+    },
+    {
+      "epoch": 2.514760147601476,
+      "grad_norm": 0.3705626923081052,
+      "learning_rate": 3.896405567619835e-06,
+      "loss": 0.0644,
+      "step": 2726
+    },
+    {
+      "epoch": 2.515682656826568,
+      "grad_norm": 0.38780181245816625,
+      "learning_rate": 3.8820272629850056e-06,
+      "loss": 0.0626,
+      "step": 2727
+    },
+    {
+      "epoch": 2.5166051660516606,
+      "grad_norm": 0.41241279771867406,
+      "learning_rate": 3.867673302986161e-06,
+      "loss": 0.0514,
+      "step": 2728
+    },
+    {
+      "epoch": 2.517527675276753,
+      "grad_norm": 0.37531282930150933,
+      "learning_rate": 3.853343704170431e-06,
+      "loss": 0.0594,
+      "step": 2729
+    },
+    {
+      "epoch": 2.518450184501845,
+      "grad_norm": 0.34799920877727736,
+      "learning_rate": 3.839038483056856e-06,
+      "loss": 0.0651,
+      "step": 2730
+    },
+    {
+      "epoch": 2.5193726937269374,
+      "grad_norm": 0.3273893877448635,
+      "learning_rate": 3.824757656136391e-06,
+      "loss": 0.0535,
+      "step": 2731
+    },
+    {
+      "epoch": 2.5202952029520294,
+      "grad_norm": 0.3935545895818458,
+      "learning_rate": 3.8105012398718694e-06,
+      "loss": 0.0628,
+      "step": 2732
+    },
+    {
+      "epoch": 2.5212177121771218,
+      "grad_norm": 0.3785337299041966,
+      "learning_rate": 3.7962692506979645e-06,
+      "loss": 0.0591,
+      "step": 2733
+    },
+    {
+      "epoch": 2.522140221402214,
+      "grad_norm": 0.37094130150458543,
+      "learning_rate": 3.7820617050212144e-06,
+      "loss": 0.061,
+      "step": 2734
+    },
+    {
+      "epoch": 2.523062730627306,
+      "grad_norm": 0.403354810381643,
+      "learning_rate": 3.7678786192199694e-06,
+      "loss": 0.0603,
+      "step": 2735
+    },
+    {
+      "epoch": 2.5239852398523985,
+      "grad_norm": 0.3642577217808013,
+      "learning_rate": 3.753720009644371e-06,
+      "loss": 0.0512,
+      "step": 2736
+    },
+    {
+      "epoch": 2.524907749077491,
+      "grad_norm": 0.3834507155368341,
+      "learning_rate": 3.7395858926163594e-06,
+      "loss": 0.0646,
+      "step": 2737
+    },
+    {
+      "epoch": 2.525830258302583,
+      "grad_norm": 0.361264708903815,
+      "learning_rate": 3.7254762844296436e-06,
+      "loss": 0.0537,
+      "step": 2738
+    },
+    {
+      "epoch": 2.5267527675276753,
+      "grad_norm": 0.3877982026840756,
+      "learning_rate": 3.7113912013496593e-06,
+      "loss": 0.0629,
+      "step": 2739
+    },
+    {
+      "epoch": 2.5276752767527677,
+      "grad_norm": 0.4006174457463105,
+      "learning_rate": 3.697330659613588e-06,
+      "loss": 0.0566,
+      "step": 2740
+    },
+    {
+      "epoch": 2.5285977859778597,
+      "grad_norm": 0.4005753817571909,
+      "learning_rate": 3.6832946754303154e-06,
+      "loss": 0.0518,
+      "step": 2741
+    },
+    {
+      "epoch": 2.529520295202952,
+      "grad_norm": 0.442537675269194,
+      "learning_rate": 3.669283264980408e-06,
+      "loss": 0.0656,
+      "step": 2742
+    },
+    {
+      "epoch": 2.530442804428044,
+      "grad_norm": 0.3689980439012489,
+      "learning_rate": 3.6552964444161174e-06,
+      "loss": 0.0569,
+      "step": 2743
+    },
+    {
+      "epoch": 2.5313653136531364,
+      "grad_norm": 0.5035218731923568,
+      "learning_rate": 3.641334229861346e-06,
+      "loss": 0.0605,
+      "step": 2744
+    },
+    {
+      "epoch": 2.532287822878229,
+      "grad_norm": 0.3206007580427925,
+      "learning_rate": 3.6273966374116175e-06,
+      "loss": 0.0548,
+      "step": 2745
+    },
+    {
+      "epoch": 2.5332103321033212,
+      "grad_norm": 0.4163887790182137,
+      "learning_rate": 3.6134836831340836e-06,
+      "loss": 0.0659,
+      "step": 2746
+    },
+    {
+      "epoch": 2.534132841328413,
+      "grad_norm": 0.3628238971354742,
+      "learning_rate": 3.5995953830675e-06,
+      "loss": 0.058,
+      "step": 2747
+    },
+    {
+      "epoch": 2.5350553505535056,
+      "grad_norm": 0.3837183227515017,
+      "learning_rate": 3.5857317532221794e-06,
+      "loss": 0.0504,
+      "step": 2748
+    },
+    {
+      "epoch": 2.5359778597785976,
+      "grad_norm": 0.37920100429254777,
+      "learning_rate": 3.571892809580013e-06,
+      "loss": 0.0445,
+      "step": 2749
+    },
+    {
+      "epoch": 2.53690036900369,
+      "grad_norm": 0.3447604406273274,
+      "learning_rate": 3.5580785680944307e-06,
+      "loss": 0.0544,
+      "step": 2750
+    },
+    {
+      "epoch": 2.5378228782287824,
+      "grad_norm": 0.38584226115149434,
+      "learning_rate": 3.544289044690377e-06,
+      "loss": 0.0545,
+      "step": 2751
+    },
+    {
+      "epoch": 2.538745387453875,
+      "grad_norm": 0.3936622525854521,
+      "learning_rate": 3.530524255264314e-06,
+      "loss": 0.0678,
+      "step": 2752
+    },
+    {
+      "epoch": 2.5396678966789668,
+      "grad_norm": 0.39453661403518914,
+      "learning_rate": 3.5167842156841794e-06,
+      "loss": 0.0611,
+      "step": 2753
+    },
+    {
+      "epoch": 2.540590405904059,
+      "grad_norm": 0.3764251705184539,
+      "learning_rate": 3.5030689417893863e-06,
+      "loss": 0.0609,
+      "step": 2754
+    },
+    {
+      "epoch": 2.541512915129151,
+      "grad_norm": 0.4021122465737506,
+      "learning_rate": 3.4893784493908067e-06,
+      "loss": 0.0562,
+      "step": 2755
+    },
+    {
+      "epoch": 2.5424354243542435,
+      "grad_norm": 0.3823096516327499,
+      "learning_rate": 3.475712754270716e-06,
+      "loss": 0.0579,
+      "step": 2756
+    },
+    {
+      "epoch": 2.543357933579336,
+      "grad_norm": 0.3957064764294518,
+      "learning_rate": 3.4620718721828345e-06,
+      "loss": 0.0561,
+      "step": 2757
+    },
+    {
+      "epoch": 2.544280442804428,
+      "grad_norm": 0.3446733865172133,
+      "learning_rate": 3.448455818852267e-06,
+      "loss": 0.0552,
+      "step": 2758
+    },
+    {
+      "epoch": 2.5452029520295203,
+      "grad_norm": 0.35795244377491925,
+      "learning_rate": 3.43486460997548e-06,
+      "loss": 0.0597,
+      "step": 2759
+    },
+    {
+      "epoch": 2.5461254612546127,
+      "grad_norm": 0.4484437769876359,
+      "learning_rate": 3.421298261220335e-06,
+      "loss": 0.0593,
+      "step": 2760
+    },
+    {
+      "epoch": 2.5470479704797047,
+      "grad_norm": 0.38968243346032577,
+      "learning_rate": 3.4077567882260047e-06,
+      "loss": 0.0723,
+      "step": 2761
+    },
+    {
+      "epoch": 2.547970479704797,
+      "grad_norm": 0.3582603007306023,
+      "learning_rate": 3.3942402066029833e-06,
+      "loss": 0.0496,
+      "step": 2762
+    },
+    {
+      "epoch": 2.5488929889298895,
+      "grad_norm": 0.4177692339230776,
+      "learning_rate": 3.3807485319331034e-06,
+      "loss": 0.0664,
+      "step": 2763
+    },
+    {
+      "epoch": 2.5498154981549814,
+      "grad_norm": 0.41201530708558287,
+      "learning_rate": 3.3672817797694545e-06,
+      "loss": 0.0667,
+      "step": 2764
+    },
+    {
+      "epoch": 2.550738007380074,
+      "grad_norm": 0.3489752270331511,
+      "learning_rate": 3.3538399656363932e-06,
+      "loss": 0.0602,
+      "step": 2765
+    },
+    {
+      "epoch": 2.551660516605166,
+      "grad_norm": 0.35229813019066014,
+      "learning_rate": 3.3404231050295526e-06,
+      "loss": 0.0514,
+      "step": 2766
+    },
+    {
+      "epoch": 2.552583025830258,
+      "grad_norm": 0.38769282705801417,
+      "learning_rate": 3.327031213415785e-06,
+      "loss": 0.065,
+      "step": 2767
+    },
+    {
+      "epoch": 2.5535055350553506,
+      "grad_norm": 0.3588379497054201,
+      "learning_rate": 3.3136643062331497e-06,
+      "loss": 0.0492,
+      "step": 2768
+    },
+    {
+      "epoch": 2.554428044280443,
+      "grad_norm": 0.36378035249904805,
+      "learning_rate": 3.3003223988909234e-06,
+      "loss": 0.0604,
+      "step": 2769
+    },
+    {
+      "epoch": 2.555350553505535,
+      "grad_norm": 0.4028390243773359,
+      "learning_rate": 3.2870055067695556e-06,
+      "loss": 0.0594,
+      "step": 2770
+    },
+    {
+      "epoch": 2.5562730627306274,
+      "grad_norm": 0.3927875562598424,
+      "learning_rate": 3.2737136452206495e-06,
+      "loss": 0.0561,
+      "step": 2771
+    },
+    {
+      "epoch": 2.5571955719557193,
+      "grad_norm": 0.3638532814167452,
+      "learning_rate": 3.260446829566963e-06,
+      "loss": 0.0581,
+      "step": 2772
+    },
+    {
+      "epoch": 2.5581180811808117,
+      "grad_norm": 0.3531343217297762,
+      "learning_rate": 3.247205075102383e-06,
+      "loss": 0.0598,
+      "step": 2773
+    },
+    {
+      "epoch": 2.559040590405904,
+      "grad_norm": 0.379302976113349,
+      "learning_rate": 3.233988397091894e-06,
+      "loss": 0.0575,
+      "step": 2774
+    },
+    {
+      "epoch": 2.5599630996309966,
+      "grad_norm": 0.419983118539953,
+      "learning_rate": 3.220796810771584e-06,
+      "loss": 0.067,
+      "step": 2775
+    },
+    {
+      "epoch": 2.5608856088560885,
+      "grad_norm": 0.4107394147825061,
+      "learning_rate": 3.2076303313486185e-06,
+      "loss": 0.0644,
+      "step": 2776
+    },
+    {
+      "epoch": 2.561808118081181,
+      "grad_norm": 0.3445990678509047,
+      "learning_rate": 3.194488974001203e-06,
+      "loss": 0.0559,
+      "step": 2777
+    },
+    {
+      "epoch": 2.562730627306273,
+      "grad_norm": 0.3570425194212317,
+      "learning_rate": 3.181372753878595e-06,
+      "loss": 0.0525,
+      "step": 2778
+    },
+    {
+      "epoch": 2.5636531365313653,
+      "grad_norm": 0.3767962842818916,
+      "learning_rate": 3.168281686101082e-06,
+      "loss": 0.061,
+      "step": 2779
+    },
+    {
+      "epoch": 2.5645756457564577,
+      "grad_norm": 0.35713584848760216,
+      "learning_rate": 3.1552157857599324e-06,
+      "loss": 0.0605,
+      "step": 2780
+    },
+    {
+      "epoch": 2.5654981549815496,
+      "grad_norm": 0.3981031620259749,
+      "learning_rate": 3.142175067917419e-06,
+      "loss": 0.0629,
+      "step": 2781
+    },
+    {
+      "epoch": 2.566420664206642,
+      "grad_norm": 0.33988134730388825,
+      "learning_rate": 3.1291595476067885e-06,
+      "loss": 0.0581,
+      "step": 2782
+    },
+    {
+      "epoch": 2.5673431734317345,
+      "grad_norm": 0.3960214847155239,
+      "learning_rate": 3.116169239832223e-06,
+      "loss": 0.0692,
+      "step": 2783
+    },
+    {
+      "epoch": 2.5682656826568264,
+      "grad_norm": 0.36151105722935056,
+      "learning_rate": 3.103204159568851e-06,
+      "loss": 0.0589,
+      "step": 2784
+    },
+    {
+      "epoch": 2.569188191881919,
+      "grad_norm": 0.41104212560282516,
+      "learning_rate": 3.090264321762723e-06,
+      "loss": 0.0605,
+      "step": 2785
+    },
+    {
+      "epoch": 2.5701107011070112,
+      "grad_norm": 0.39688685526652495,
+      "learning_rate": 3.077349741330776e-06,
+      "loss": 0.056,
+      "step": 2786
+    },
+    {
+      "epoch": 2.571033210332103,
+      "grad_norm": 0.4599531096694881,
+      "learning_rate": 3.0644604331608456e-06,
+      "loss": 0.0575,
+      "step": 2787
+    },
+    {
+      "epoch": 2.5719557195571956,
+      "grad_norm": 0.36391951034293263,
+      "learning_rate": 3.051596412111618e-06,
+      "loss": 0.0534,
+      "step": 2788
+    },
+    {
+      "epoch": 2.5728782287822876,
+      "grad_norm": 0.35370583130882044,
+      "learning_rate": 3.038757693012642e-06,
+      "loss": 0.0524,
+      "step": 2789
+    },
+    {
+      "epoch": 2.57380073800738,
+      "grad_norm": 0.376223235497432,
+      "learning_rate": 3.025944290664301e-06,
+      "loss": 0.0665,
+      "step": 2790
+    },
+    {
+      "epoch": 2.5747232472324724,
+      "grad_norm": 0.36305580286586053,
+      "learning_rate": 3.013156219837776e-06,
+      "loss": 0.0549,
+      "step": 2791
+    },
+    {
+      "epoch": 2.5756457564575648,
+      "grad_norm": 0.37243590037908647,
+      "learning_rate": 3.0003934952750586e-06,
+      "loss": 0.0579,
+      "step": 2792
+    },
+    {
+      "epoch": 2.5765682656826567,
+      "grad_norm": 0.4178900830973266,
+      "learning_rate": 2.987656131688926e-06,
+      "loss": 0.0668,
+      "step": 2793
+    },
+    {
+      "epoch": 2.577490774907749,
+      "grad_norm": 0.36250689707924433,
+      "learning_rate": 2.9749441437629033e-06,
+      "loss": 0.0459,
+      "step": 2794
+    },
+    {
+      "epoch": 2.578413284132841,
+      "grad_norm": 0.35145674746729727,
+      "learning_rate": 2.9622575461512733e-06,
+      "loss": 0.0556,
+      "step": 2795
+    },
+    {
+      "epoch": 2.5793357933579335,
+      "grad_norm": 0.3671933764673374,
+      "learning_rate": 2.949596353479059e-06,
+      "loss": 0.0523,
+      "step": 2796
+    },
+    {
+      "epoch": 2.580258302583026,
+      "grad_norm": 0.40725450604752705,
+      "learning_rate": 2.9369605803419715e-06,
+      "loss": 0.057,
+      "step": 2797
+    },
+    {
+      "epoch": 2.5811808118081183,
+      "grad_norm": 0.40376867023345175,
+      "learning_rate": 2.9243502413064368e-06,
+      "loss": 0.0645,
+      "step": 2798
+    },
+    {
+      "epoch": 2.5821033210332103,
+      "grad_norm": 0.3732775675465853,
+      "learning_rate": 2.911765350909565e-06,
+      "loss": 0.0624,
+      "step": 2799
+    },
+    {
+      "epoch": 2.5830258302583027,
+      "grad_norm": 0.414186439960148,
+      "learning_rate": 2.899205923659107e-06,
+      "loss": 0.0631,
+      "step": 2800
+    },
+    {
+      "epoch": 2.5839483394833946,
+      "grad_norm": 0.357872007207858,
+      "learning_rate": 2.8866719740334807e-06,
+      "loss": 0.0554,
+      "step": 2801
+    },
+    {
+      "epoch": 2.584870848708487,
+      "grad_norm": 0.40658643535825106,
+      "learning_rate": 2.8741635164817315e-06,
+      "loss": 0.0676,
+      "step": 2802
+    },
+    {
+      "epoch": 2.5857933579335795,
+      "grad_norm": 0.43182062585631215,
+      "learning_rate": 2.8616805654234997e-06,
+      "loss": 0.0612,
+      "step": 2803
+    },
+    {
+      "epoch": 2.586715867158672,
+      "grad_norm": 0.39240960661008817,
+      "learning_rate": 2.8492231352490463e-06,
+      "loss": 0.0549,
+      "step": 2804
+    },
+    {
+      "epoch": 2.587638376383764,
+      "grad_norm": 0.416096310635408,
+      "learning_rate": 2.8367912403191977e-06,
+      "loss": 0.0627,
+      "step": 2805
+    },
+    {
+      "epoch": 2.588560885608856,
+      "grad_norm": 0.4138551349403716,
+      "learning_rate": 2.8243848949653428e-06,
+      "loss": 0.0603,
+      "step": 2806
+    },
+    {
+      "epoch": 2.589483394833948,
+      "grad_norm": 0.39622879164551106,
+      "learning_rate": 2.812004113489425e-06,
+      "loss": 0.0528,
+      "step": 2807
+    },
+    {
+      "epoch": 2.5904059040590406,
+      "grad_norm": 0.3581763549159933,
+      "learning_rate": 2.7996489101639157e-06,
+      "loss": 0.0583,
+      "step": 2808
+    },
+    {
+      "epoch": 2.591328413284133,
+      "grad_norm": 0.40241494830455715,
+      "learning_rate": 2.7873192992317887e-06,
+      "loss": 0.0627,
+      "step": 2809
+    },
+    {
+      "epoch": 2.592250922509225,
+      "grad_norm": 0.44573546308002177,
+      "learning_rate": 2.77501529490653e-06,
+      "loss": 0.0673,
+      "step": 2810
+    },
+    {
+      "epoch": 2.5931734317343174,
+      "grad_norm": 0.3759217600883933,
+      "learning_rate": 2.7627369113721045e-06,
+      "loss": 0.0533,
+      "step": 2811
+    },
+    {
+      "epoch": 2.5940959409594093,
+      "grad_norm": 0.41081070022494737,
+      "learning_rate": 2.750484162782929e-06,
+      "loss": 0.0592,
+      "step": 2812
+    },
+    {
+      "epoch": 2.5950184501845017,
+      "grad_norm": 0.417495764332896,
+      "learning_rate": 2.7382570632638854e-06,
+      "loss": 0.0549,
+      "step": 2813
+    },
+    {
+      "epoch": 2.595940959409594,
+      "grad_norm": 0.4211177228837055,
+      "learning_rate": 2.7260556269102815e-06,
+      "loss": 0.0657,
+      "step": 2814
+    },
+    {
+      "epoch": 2.5968634686346865,
+      "grad_norm": 0.45843874502664883,
+      "learning_rate": 2.7138798677878273e-06,
+      "loss": 0.0609,
+      "step": 2815
+    },
+    {
+      "epoch": 2.5977859778597785,
+      "grad_norm": 0.39580377551510015,
+      "learning_rate": 2.7017297999326537e-06,
+      "loss": 0.0574,
+      "step": 2816
+    },
+    {
+      "epoch": 2.598708487084871,
+      "grad_norm": 0.3806490365199812,
+      "learning_rate": 2.689605437351267e-06,
+      "loss": 0.0598,
+      "step": 2817
+    },
+    {
+      "epoch": 2.599630996309963,
+      "grad_norm": 0.40152279558778614,
+      "learning_rate": 2.6775067940205288e-06,
+      "loss": 0.0606,
+      "step": 2818
+    },
+    {
+      "epoch": 2.6005535055350553,
+      "grad_norm": 0.3733066792037547,
+      "learning_rate": 2.6654338838876665e-06,
+      "loss": 0.0585,
+      "step": 2819
+    },
+    {
+      "epoch": 2.6014760147601477,
+      "grad_norm": 0.45095338506563537,
+      "learning_rate": 2.6533867208702433e-06,
+      "loss": 0.0654,
+      "step": 2820
+    },
+    {
+      "epoch": 2.60239852398524,
+      "grad_norm": 0.4069945137391312,
+      "learning_rate": 2.641365318856126e-06,
+      "loss": 0.0615,
+      "step": 2821
+    },
+    {
+      "epoch": 2.603321033210332,
+      "grad_norm": 0.520598095821637,
+      "learning_rate": 2.6293696917035066e-06,
+      "loss": 0.0713,
+      "step": 2822
+    },
+    {
+      "epoch": 2.6042435424354244,
+      "grad_norm": 0.35595572132161335,
+      "learning_rate": 2.6173998532408347e-06,
+      "loss": 0.0557,
+      "step": 2823
+    },
+    {
+      "epoch": 2.6051660516605164,
+      "grad_norm": 0.42742458159350666,
+      "learning_rate": 2.6054558172668607e-06,
+      "loss": 0.0587,
+      "step": 2824
+    },
+    {
+      "epoch": 2.606088560885609,
+      "grad_norm": 0.4849666629112722,
+      "learning_rate": 2.593537597550577e-06,
+      "loss": 0.0636,
+      "step": 2825
+    },
+    {
+      "epoch": 2.607011070110701,
+      "grad_norm": 0.40275280892370763,
+      "learning_rate": 2.581645207831204e-06,
+      "loss": 0.0592,
+      "step": 2826
+    },
+    {
+      "epoch": 2.6079335793357936,
+      "grad_norm": 0.37392792931653523,
+      "learning_rate": 2.569778661818209e-06,
+      "loss": 0.0613,
+      "step": 2827
+    },
+    {
+      "epoch": 2.6088560885608856,
+      "grad_norm": 0.4020759847705154,
+      "learning_rate": 2.5579379731912517e-06,
+      "loss": 0.0582,
+      "step": 2828
+    },
+    {
+      "epoch": 2.609778597785978,
+      "grad_norm": 0.38384509311715215,
+      "learning_rate": 2.5461231556001803e-06,
+      "loss": 0.0612,
+      "step": 2829
+    },
+    {
+      "epoch": 2.61070110701107,
+      "grad_norm": 0.38682217224419096,
+      "learning_rate": 2.53433422266503e-06,
+      "loss": 0.0649,
+      "step": 2830
+    },
+    {
+      "epoch": 2.6116236162361623,
+      "grad_norm": 0.3982787313816613,
+      "learning_rate": 2.522571187975997e-06,
+      "loss": 0.0588,
+      "step": 2831
+    },
+    {
+      "epoch": 2.6125461254612548,
+      "grad_norm": 0.39020127803349186,
+      "learning_rate": 2.5108340650934065e-06,
+      "loss": 0.0624,
+      "step": 2832
+    },
+    {
+      "epoch": 2.6134686346863467,
+      "grad_norm": 0.35400390451308544,
+      "learning_rate": 2.499122867547729e-06,
+      "loss": 0.0496,
+      "step": 2833
+    },
+    {
+      "epoch": 2.614391143911439,
+      "grad_norm": 0.39965035233977825,
+      "learning_rate": 2.487437608839546e-06,
+      "loss": 0.0632,
+      "step": 2834
+    },
+    {
+      "epoch": 2.6153136531365315,
+      "grad_norm": 0.3550887873627179,
+      "learning_rate": 2.475778302439524e-06,
+      "loss": 0.0483,
+      "step": 2835
+    },
+    {
+      "epoch": 2.6162361623616235,
+      "grad_norm": 0.40614520630097883,
+      "learning_rate": 2.4641449617884257e-06,
+      "loss": 0.0633,
+      "step": 2836
+    },
+    {
+      "epoch": 2.617158671586716,
+      "grad_norm": 0.40055305533410873,
+      "learning_rate": 2.4525376002970835e-06,
+      "loss": 0.0656,
+      "step": 2837
+    },
+    {
+      "epoch": 2.6180811808118083,
+      "grad_norm": 0.3505541228685418,
+      "learning_rate": 2.4409562313463642e-06,
+      "loss": 0.0558,
+      "step": 2838
+    },
+    {
+      "epoch": 2.6190036900369003,
+      "grad_norm": 0.4178637542478256,
+      "learning_rate": 2.429400868287182e-06,
+      "loss": 0.065,
+      "step": 2839
+    },
+    {
+      "epoch": 2.6199261992619927,
+      "grad_norm": 0.3666081677039269,
+      "learning_rate": 2.4178715244404794e-06,
+      "loss": 0.0525,
+      "step": 2840
+    },
+    {
+      "epoch": 2.6208487084870846,
+      "grad_norm": 0.3871509737468735,
+      "learning_rate": 2.406368213097185e-06,
+      "loss": 0.0607,
+      "step": 2841
+    },
+    {
+      "epoch": 2.621771217712177,
+      "grad_norm": 0.36603426397149014,
+      "learning_rate": 2.3948909475182275e-06,
+      "loss": 0.0578,
+      "step": 2842
+    },
+    {
+      "epoch": 2.6226937269372694,
+      "grad_norm": 0.39963824073936477,
+      "learning_rate": 2.3834397409345205e-06,
+      "loss": 0.0624,
+      "step": 2843
+    },
+    {
+      "epoch": 2.623616236162362,
+      "grad_norm": 0.4236075293068511,
+      "learning_rate": 2.372014606546913e-06,
+      "loss": 0.0645,
+      "step": 2844
+    },
+    {
+      "epoch": 2.624538745387454,
+      "grad_norm": 0.35510643010869397,
+      "learning_rate": 2.360615557526219e-06,
+      "loss": 0.0586,
+      "step": 2845
+    },
+    {
+      "epoch": 2.625461254612546,
+      "grad_norm": 0.40337374663560444,
+      "learning_rate": 2.3492426070131747e-06,
+      "loss": 0.058,
+      "step": 2846
+    },
+    {
+      "epoch": 2.626383763837638,
+      "grad_norm": 0.35887743196715344,
+      "learning_rate": 2.3378957681184283e-06,
+      "loss": 0.0548,
+      "step": 2847
+    },
+    {
+      "epoch": 2.6273062730627306,
+      "grad_norm": 0.40549551953993745,
+      "learning_rate": 2.326575053922525e-06,
+      "loss": 0.0691,
+      "step": 2848
+    },
+    {
+      "epoch": 2.628228782287823,
+      "grad_norm": 0.3410434892822295,
+      "learning_rate": 2.315280477475906e-06,
+      "loss": 0.0622,
+      "step": 2849
+    },
+    {
+      "epoch": 2.6291512915129154,
+      "grad_norm": 0.4234698785178658,
+      "learning_rate": 2.3040120517988593e-06,
+      "loss": 0.0553,
+      "step": 2850
+    },
+    {
+      "epoch": 2.6300738007380073,
+      "grad_norm": 0.4480502742154732,
+      "learning_rate": 2.2927697898815465e-06,
+      "loss": 0.0601,
+      "step": 2851
+    },
+    {
+      "epoch": 2.6309963099630997,
+      "grad_norm": 0.4183436153605026,
+      "learning_rate": 2.281553704683964e-06,
+      "loss": 0.0667,
+      "step": 2852
+    },
+    {
+      "epoch": 2.6319188191881917,
+      "grad_norm": 0.3268256625000969,
+      "learning_rate": 2.270363809135917e-06,
+      "loss": 0.0473,
+      "step": 2853
+    },
+    {
+      "epoch": 2.632841328413284,
+      "grad_norm": 0.4243308731347921,
+      "learning_rate": 2.2592001161370392e-06,
+      "loss": 0.0579,
+      "step": 2854
+    },
+    {
+      "epoch": 2.6337638376383765,
+      "grad_norm": 0.4114129868395799,
+      "learning_rate": 2.2480626385567525e-06,
+      "loss": 0.0525,
+      "step": 2855
+    },
+    {
+      "epoch": 2.6346863468634685,
+      "grad_norm": 0.36434385215921766,
+      "learning_rate": 2.2369513892342458e-06,
+      "loss": 0.0534,
+      "step": 2856
+    },
+    {
+      "epoch": 2.635608856088561,
+      "grad_norm": 0.4154254525152973,
+      "learning_rate": 2.2258663809784892e-06,
+      "loss": 0.0589,
+      "step": 2857
+    },
+    {
+      "epoch": 2.6365313653136533,
+      "grad_norm": 0.3769354100037087,
+      "learning_rate": 2.2148076265681883e-06,
+      "loss": 0.0533,
+      "step": 2858
+    },
+    {
+      "epoch": 2.6374538745387452,
+      "grad_norm": 0.5050526804534307,
+      "learning_rate": 2.2037751387517902e-06,
+      "loss": 0.0581,
+      "step": 2859
+    },
+    {
+      "epoch": 2.6383763837638377,
+      "grad_norm": 0.47037972364930525,
+      "learning_rate": 2.1927689302474714e-06,
+      "loss": 0.0718,
+      "step": 2860
+    },
+    {
+      "epoch": 2.63929889298893,
+      "grad_norm": 0.38484287178920423,
+      "learning_rate": 2.1817890137430934e-06,
+      "loss": 0.0656,
+      "step": 2861
+    },
+    {
+      "epoch": 2.640221402214022,
+      "grad_norm": 0.3731154212738967,
+      "learning_rate": 2.1708354018962236e-06,
+      "loss": 0.0614,
+      "step": 2862
+    },
+    {
+      "epoch": 2.6411439114391144,
+      "grad_norm": 0.3610043698344828,
+      "learning_rate": 2.159908107334102e-06,
+      "loss": 0.0572,
+      "step": 2863
+    },
+    {
+      "epoch": 2.6420664206642064,
+      "grad_norm": 0.435744217972833,
+      "learning_rate": 2.149007142653625e-06,
+      "loss": 0.059,
+      "step": 2864
+    },
+    {
+      "epoch": 2.642988929889299,
+      "grad_norm": 0.412794686187554,
+      "learning_rate": 2.138132520421346e-06,
+      "loss": 0.0542,
+      "step": 2865
+    },
+    {
+      "epoch": 2.643911439114391,
+      "grad_norm": 0.39129914778294933,
+      "learning_rate": 2.127284253173445e-06,
+      "loss": 0.0677,
+      "step": 2866
+    },
+    {
+      "epoch": 2.6448339483394836,
+      "grad_norm": 0.36160315196238885,
+      "learning_rate": 2.116462353415716e-06,
+      "loss": 0.057,
+      "step": 2867
+    },
+    {
+      "epoch": 2.6457564575645756,
+      "grad_norm": 0.485793283197896,
+      "learning_rate": 2.1056668336235622e-06,
+      "loss": 0.0693,
+      "step": 2868
+    },
+    {
+      "epoch": 2.646678966789668,
+      "grad_norm": 0.4181192945796221,
+      "learning_rate": 2.0948977062419854e-06,
+      "loss": 0.057,
+      "step": 2869
+    },
+    {
+      "epoch": 2.64760147601476,
+      "grad_norm": 0.3830742446174007,
+      "learning_rate": 2.084154983685538e-06,
+      "loss": 0.0592,
+      "step": 2870
+    },
+    {
+      "epoch": 2.6485239852398523,
+      "grad_norm": 0.44990348728191754,
+      "learning_rate": 2.0734386783383573e-06,
+      "loss": 0.0575,
+      "step": 2871
+    },
+    {
+      "epoch": 2.6494464944649447,
+      "grad_norm": 0.35535204353980193,
+      "learning_rate": 2.0627488025541154e-06,
+      "loss": 0.0541,
+      "step": 2872
+    },
+    {
+      "epoch": 2.650369003690037,
+      "grad_norm": 0.387494327233977,
+      "learning_rate": 2.0520853686560178e-06,
+      "loss": 0.0599,
+      "step": 2873
+    },
+    {
+      "epoch": 2.651291512915129,
+      "grad_norm": 0.392244538962278,
+      "learning_rate": 2.041448388936784e-06,
+      "loss": 0.055,
+      "step": 2874
+    },
+    {
+      "epoch": 2.6522140221402215,
+      "grad_norm": 0.45221744715193424,
+      "learning_rate": 2.030837875658656e-06,
+      "loss": 0.0694,
+      "step": 2875
+    },
+    {
+      "epoch": 2.6531365313653135,
+      "grad_norm": 0.40471237010919736,
+      "learning_rate": 2.0202538410533352e-06,
+      "loss": 0.0661,
+      "step": 2876
+    },
+    {
+      "epoch": 2.654059040590406,
+      "grad_norm": 0.3860743963032248,
+      "learning_rate": 2.0096962973220225e-06,
+      "loss": 0.0614,
+      "step": 2877
+    },
+    {
+      "epoch": 2.6549815498154983,
+      "grad_norm": 0.3652598004060305,
+      "learning_rate": 1.999165256635377e-06,
+      "loss": 0.0592,
+      "step": 2878
+    },
+    {
+      "epoch": 2.6559040590405907,
+      "grad_norm": 0.3903898261051929,
+      "learning_rate": 1.988660731133499e-06,
+      "loss": 0.0541,
+      "step": 2879
+    },
+    {
+      "epoch": 2.6568265682656826,
+      "grad_norm": 0.401467646881587,
+      "learning_rate": 1.9781827329259125e-06,
+      "loss": 0.0575,
+      "step": 2880
+    },
+    {
+      "epoch": 2.657749077490775,
+      "grad_norm": 0.36351105489318314,
+      "learning_rate": 1.9677312740915913e-06,
+      "loss": 0.052,
+      "step": 2881
+    },
+    {
+      "epoch": 2.658671586715867,
+      "grad_norm": 0.38888434186610843,
+      "learning_rate": 1.9573063666788875e-06,
+      "loss": 0.0627,
+      "step": 2882
+    },
+    {
+      "epoch": 2.6595940959409594,
+      "grad_norm": 0.4088245619118583,
+      "learning_rate": 1.946908022705546e-06,
+      "loss": 0.0727,
+      "step": 2883
+    },
+    {
+      "epoch": 2.660516605166052,
+      "grad_norm": 0.3508965388722947,
+      "learning_rate": 1.9365362541587132e-06,
+      "loss": 0.0535,
+      "step": 2884
+    },
+    {
+      "epoch": 2.661439114391144,
+      "grad_norm": 0.46244332242395547,
+      "learning_rate": 1.926191072994879e-06,
+      "loss": 0.0666,
+      "step": 2885
+    },
+    {
+      "epoch": 2.662361623616236,
+      "grad_norm": 0.4542887467397042,
+      "learning_rate": 1.915872491139875e-06,
+      "loss": 0.0646,
+      "step": 2886
+    },
+    {
+      "epoch": 2.663284132841328,
+      "grad_norm": 0.358909025960324,
+      "learning_rate": 1.9055805204889033e-06,
+      "loss": 0.0573,
+      "step": 2887
+    },
+    {
+      "epoch": 2.6642066420664205,
+      "grad_norm": 0.4069627859242188,
+      "learning_rate": 1.8953151729064532e-06,
+      "loss": 0.0622,
+      "step": 2888
+    },
+    {
+      "epoch": 2.665129151291513,
+      "grad_norm": 0.3507560973723166,
+      "learning_rate": 1.8850764602263426e-06,
+      "loss": 0.0548,
+      "step": 2889
+    },
+    {
+      "epoch": 2.6660516605166054,
+      "grad_norm": 0.3981380354869457,
+      "learning_rate": 1.874864394251688e-06,
+      "loss": 0.0525,
+      "step": 2890
+    },
+    {
+      "epoch": 2.6669741697416973,
+      "grad_norm": 0.395914315217764,
+      "learning_rate": 1.864678986754867e-06,
+      "loss": 0.0548,
+      "step": 2891
+    },
+    {
+      "epoch": 2.6678966789667897,
+      "grad_norm": 0.3588837691563503,
+      "learning_rate": 1.8545202494775509e-06,
+      "loss": 0.0492,
+      "step": 2892
+    },
+    {
+      "epoch": 2.6688191881918817,
+      "grad_norm": 0.4808091742446625,
+      "learning_rate": 1.8443881941306417e-06,
+      "loss": 0.0687,
+      "step": 2893
+    },
+    {
+      "epoch": 2.669741697416974,
+      "grad_norm": 0.4107640158220843,
+      "learning_rate": 1.8342828323943046e-06,
+      "loss": 0.0525,
+      "step": 2894
+    },
+    {
+      "epoch": 2.6706642066420665,
+      "grad_norm": 0.43318566507277145,
+      "learning_rate": 1.8242041759179208e-06,
+      "loss": 0.0581,
+      "step": 2895
+    },
+    {
+      "epoch": 2.671586715867159,
+      "grad_norm": 0.36854549918837504,
+      "learning_rate": 1.8141522363200797e-06,
+      "loss": 0.0631,
+      "step": 2896
+    },
+    {
+      "epoch": 2.672509225092251,
+      "grad_norm": 0.4240095530003623,
+      "learning_rate": 1.80412702518859e-06,
+      "loss": 0.0656,
+      "step": 2897
+    },
+    {
+      "epoch": 2.6734317343173433,
+      "grad_norm": 0.37489214791621156,
+      "learning_rate": 1.7941285540804348e-06,
+      "loss": 0.0644,
+      "step": 2898
+    },
+    {
+      "epoch": 2.6743542435424352,
+      "grad_norm": 0.35262465473014704,
+      "learning_rate": 1.784156834521769e-06,
+      "loss": 0.0535,
+      "step": 2899
+    },
+    {
+      "epoch": 2.6752767527675276,
+      "grad_norm": 0.4350832985499843,
+      "learning_rate": 1.7742118780079197e-06,
+      "loss": 0.0619,
+      "step": 2900
+    },
+    {
+      "epoch": 2.67619926199262,
+      "grad_norm": 0.42795596406317976,
+      "learning_rate": 1.7642936960033578e-06,
+      "loss": 0.0586,
+      "step": 2901
+    },
+    {
+      "epoch": 2.6771217712177124,
+      "grad_norm": 0.36301900739582676,
+      "learning_rate": 1.7544022999416792e-06,
+      "loss": 0.0527,
+      "step": 2902
+    },
+    {
+      "epoch": 2.6780442804428044,
+      "grad_norm": 0.43278406476884773,
+      "learning_rate": 1.7445377012256126e-06,
+      "loss": 0.0603,
+      "step": 2903
+    },
+    {
+      "epoch": 2.678966789667897,
+      "grad_norm": 0.37181142368750203,
+      "learning_rate": 1.7346999112269973e-06,
+      "loss": 0.0571,
+      "step": 2904
+    },
+    {
+      "epoch": 2.6798892988929888,
+      "grad_norm": 0.46085662250480114,
+      "learning_rate": 1.7248889412867507e-06,
+      "loss": 0.0659,
+      "step": 2905
+    },
+    {
+      "epoch": 2.680811808118081,
+      "grad_norm": 0.35249503133774546,
+      "learning_rate": 1.7151048027148896e-06,
+      "loss": 0.0596,
+      "step": 2906
+    },
+    {
+      "epoch": 2.6817343173431736,
+      "grad_norm": 0.3614844575676369,
+      "learning_rate": 1.7053475067904973e-06,
+      "loss": 0.063,
+      "step": 2907
+    },
+    {
+      "epoch": 2.6826568265682655,
+      "grad_norm": 0.3419122906463993,
+      "learning_rate": 1.6956170647616982e-06,
+      "loss": 0.0601,
+      "step": 2908
+    },
+    {
+      "epoch": 2.683579335793358,
+      "grad_norm": 0.3163027577396601,
+      "learning_rate": 1.6859134878456806e-06,
+      "loss": 0.0617,
+      "step": 2909
+    },
+    {
+      "epoch": 2.6845018450184504,
+      "grad_norm": 0.36376314325136916,
+      "learning_rate": 1.6762367872286522e-06,
+      "loss": 0.0582,
+      "step": 2910
+    },
+    {
+      "epoch": 2.6854243542435423,
+      "grad_norm": 0.41047964262592385,
+      "learning_rate": 1.6665869740658312e-06,
+      "loss": 0.0683,
+      "step": 2911
+    },
+    {
+      "epoch": 2.6863468634686347,
+      "grad_norm": 0.4044684801609656,
+      "learning_rate": 1.6569640594814528e-06,
+      "loss": 0.0466,
+      "step": 2912
+    },
+    {
+      "epoch": 2.687269372693727,
+      "grad_norm": 0.3363346378880635,
+      "learning_rate": 1.647368054568743e-06,
+      "loss": 0.0541,
+      "step": 2913
+    },
+    {
+      "epoch": 2.688191881918819,
+      "grad_norm": 0.3689758433425425,
+      "learning_rate": 1.6377989703899006e-06,
+      "loss": 0.0569,
+      "step": 2914
+    },
+    {
+      "epoch": 2.6891143911439115,
+      "grad_norm": 0.3465649965905109,
+      "learning_rate": 1.6282568179760787e-06,
+      "loss": 0.0512,
+      "step": 2915
+    },
+    {
+      "epoch": 2.6900369003690034,
+      "grad_norm": 0.46391072040527415,
+      "learning_rate": 1.6187416083274149e-06,
+      "loss": 0.0549,
+      "step": 2916
+    },
+    {
+      "epoch": 2.690959409594096,
+      "grad_norm": 0.38803061785917914,
+      "learning_rate": 1.6092533524129622e-06,
+      "loss": 0.0532,
+      "step": 2917
+    },
+    {
+      "epoch": 2.6918819188191883,
+      "grad_norm": 0.3955683583672091,
+      "learning_rate": 1.5997920611707017e-06,
+      "loss": 0.0594,
+      "step": 2918
+    },
+    {
+      "epoch": 2.6928044280442807,
+      "grad_norm": 0.37998147874811905,
+      "learning_rate": 1.5903577455075508e-06,
+      "loss": 0.0539,
+      "step": 2919
+    },
+    {
+      "epoch": 2.6937269372693726,
+      "grad_norm": 0.3900903992942932,
+      "learning_rate": 1.5809504162993094e-06,
+      "loss": 0.0518,
+      "step": 2920
+    },
+    {
+      "epoch": 2.694649446494465,
+      "grad_norm": 0.42179565305103933,
+      "learning_rate": 1.571570084390664e-06,
+      "loss": 0.061,
+      "step": 2921
+    },
+    {
+      "epoch": 2.695571955719557,
+      "grad_norm": 0.36424224717684905,
+      "learning_rate": 1.5622167605952086e-06,
+      "loss": 0.056,
+      "step": 2922
+    },
+    {
+      "epoch": 2.6964944649446494,
+      "grad_norm": 0.38416630537088486,
+      "learning_rate": 1.552890455695369e-06,
+      "loss": 0.0616,
+      "step": 2923
+    },
+    {
+      "epoch": 2.697416974169742,
+      "grad_norm": 0.4007798955554396,
+      "learning_rate": 1.5435911804424357e-06,
+      "loss": 0.0608,
+      "step": 2924
+    },
+    {
+      "epoch": 2.698339483394834,
+      "grad_norm": 0.39524893043698506,
+      "learning_rate": 1.5343189455565537e-06,
+      "loss": 0.0608,
+      "step": 2925
+    },
+    {
+      "epoch": 2.699261992619926,
+      "grad_norm": 0.3734287745141257,
+      "learning_rate": 1.5250737617266753e-06,
+      "loss": 0.052,
+      "step": 2926
+    },
+    {
+      "epoch": 2.7001845018450186,
+      "grad_norm": 0.4006702967524649,
+      "learning_rate": 1.5158556396105749e-06,
+      "loss": 0.0564,
+      "step": 2927
+    },
+    {
+      "epoch": 2.7011070110701105,
+      "grad_norm": 0.36161962739478526,
+      "learning_rate": 1.5066645898348385e-06,
+      "loss": 0.0606,
+      "step": 2928
+    },
+    {
+      "epoch": 2.702029520295203,
+      "grad_norm": 0.38078822448583644,
+      "learning_rate": 1.497500622994835e-06,
+      "loss": 0.0575,
+      "step": 2929
+    },
+    {
+      "epoch": 2.7029520295202953,
+      "grad_norm": 0.4449918524142434,
+      "learning_rate": 1.4883637496547142e-06,
+      "loss": 0.0651,
+      "step": 2930
+    },
+    {
+      "epoch": 2.7038745387453873,
+      "grad_norm": 0.384957097979171,
+      "learning_rate": 1.479253980347392e-06,
+      "loss": 0.0696,
+      "step": 2931
+    },
+    {
+      "epoch": 2.7047970479704797,
+      "grad_norm": 0.415133375131178,
+      "learning_rate": 1.4701713255745403e-06,
+      "loss": 0.0566,
+      "step": 2932
+    },
+    {
+      "epoch": 2.705719557195572,
+      "grad_norm": 0.38360118027466955,
+      "learning_rate": 1.4611157958065807e-06,
+      "loss": 0.0607,
+      "step": 2933
+    },
+    {
+      "epoch": 2.706642066420664,
+      "grad_norm": 0.34898906743695224,
+      "learning_rate": 1.4520874014826463e-06,
+      "loss": 0.0536,
+      "step": 2934
+    },
+    {
+      "epoch": 2.7075645756457565,
+      "grad_norm": 0.35629998158132603,
+      "learning_rate": 1.4430861530106087e-06,
+      "loss": 0.0542,
+      "step": 2935
+    },
+    {
+      "epoch": 2.708487084870849,
+      "grad_norm": 0.36544252176415665,
+      "learning_rate": 1.4341120607670371e-06,
+      "loss": 0.0591,
+      "step": 2936
+    },
+    {
+      "epoch": 2.709409594095941,
+      "grad_norm": 0.3524151942164848,
+      "learning_rate": 1.4251651350971896e-06,
+      "loss": 0.0493,
+      "step": 2937
+    },
+    {
+      "epoch": 2.7103321033210332,
+      "grad_norm": 0.335656337855431,
+      "learning_rate": 1.4162453863150183e-06,
+      "loss": 0.0537,
+      "step": 2938
+    },
+    {
+      "epoch": 2.711254612546125,
+      "grad_norm": 0.3865218249137243,
+      "learning_rate": 1.4073528247031426e-06,
+      "loss": 0.0557,
+      "step": 2939
+    },
+    {
+      "epoch": 2.7121771217712176,
+      "grad_norm": 0.3438854336540138,
+      "learning_rate": 1.3984874605128345e-06,
+      "loss": 0.054,
+      "step": 2940
+    },
+    {
+      "epoch": 2.71309963099631,
+      "grad_norm": 0.3648875446352297,
+      "learning_rate": 1.3896493039640163e-06,
+      "loss": 0.0589,
+      "step": 2941
+    },
+    {
+      "epoch": 2.7140221402214024,
+      "grad_norm": 0.36558746492823224,
+      "learning_rate": 1.3808383652452545e-06,
+      "loss": 0.0558,
+      "step": 2942
+    },
+    {
+      "epoch": 2.7149446494464944,
+      "grad_norm": 0.41658909272946937,
+      "learning_rate": 1.3720546545137215e-06,
+      "loss": 0.0566,
+      "step": 2943
+    },
+    {
+      "epoch": 2.715867158671587,
+      "grad_norm": 0.3751851162404485,
+      "learning_rate": 1.3632981818952145e-06,
+      "loss": 0.0639,
+      "step": 2944
+    },
+    {
+      "epoch": 2.7167896678966788,
+      "grad_norm": 0.3673279764095167,
+      "learning_rate": 1.3545689574841342e-06,
+      "loss": 0.0628,
+      "step": 2945
+    },
+    {
+      "epoch": 2.717712177121771,
+      "grad_norm": 0.4019450184845962,
+      "learning_rate": 1.345866991343453e-06,
+      "loss": 0.0587,
+      "step": 2946
+    },
+    {
+      "epoch": 2.7186346863468636,
+      "grad_norm": 0.3298962322639587,
+      "learning_rate": 1.3371922935047355e-06,
+      "loss": 0.0596,
+      "step": 2947
+    },
+    {
+      "epoch": 2.719557195571956,
+      "grad_norm": 0.4767254384847376,
+      "learning_rate": 1.32854487396811e-06,
+      "loss": 0.0608,
+      "step": 2948
+    },
+    {
+      "epoch": 2.720479704797048,
+      "grad_norm": 0.3442916421234061,
+      "learning_rate": 1.3199247427022526e-06,
+      "loss": 0.0545,
+      "step": 2949
+    },
+    {
+      "epoch": 2.7214022140221403,
+      "grad_norm": 0.4750998762930767,
+      "learning_rate": 1.3113319096443728e-06,
+      "loss": 0.0634,
+      "step": 2950
+    },
+    {
+      "epoch": 2.7223247232472323,
+      "grad_norm": 0.33294417897663703,
+      "learning_rate": 1.302766384700238e-06,
+      "loss": 0.0578,
+      "step": 2951
+    },
+    {
+      "epoch": 2.7232472324723247,
+      "grad_norm": 0.3918038809940905,
+      "learning_rate": 1.2942281777441168e-06,
+      "loss": 0.0608,
+      "step": 2952
+    },
+    {
+      "epoch": 2.724169741697417,
+      "grad_norm": 0.3904724565577004,
+      "learning_rate": 1.2857172986187744e-06,
+      "loss": 0.0544,
+      "step": 2953
+    },
+    {
+      "epoch": 2.725092250922509,
+      "grad_norm": 0.3792220771172754,
+      "learning_rate": 1.2772337571355043e-06,
+      "loss": 0.0671,
+      "step": 2954
+    },
+    {
+      "epoch": 2.7260147601476015,
+      "grad_norm": 0.3508443223549421,
+      "learning_rate": 1.2687775630740612e-06,
+      "loss": 0.0537,
+      "step": 2955
+    },
+    {
+      "epoch": 2.726937269372694,
+      "grad_norm": 0.3958974325498116,
+      "learning_rate": 1.2603487261826724e-06,
+      "loss": 0.0493,
+      "step": 2956
+    },
+    {
+      "epoch": 2.727859778597786,
+      "grad_norm": 0.4394079976531421,
+      "learning_rate": 1.2519472561780488e-06,
+      "loss": 0.0626,
+      "step": 2957
+    },
+    {
+      "epoch": 2.7287822878228782,
+      "grad_norm": 0.4059333547531129,
+      "learning_rate": 1.2435731627453345e-06,
+      "loss": 0.0617,
+      "step": 2958
+    },
+    {
+      "epoch": 2.7297047970479706,
+      "grad_norm": 0.3699666582867163,
+      "learning_rate": 1.2352264555381132e-06,
+      "loss": 0.061,
+      "step": 2959
+    },
+    {
+      "epoch": 2.7306273062730626,
+      "grad_norm": 0.3435908868137269,
+      "learning_rate": 1.2269071441784158e-06,
+      "loss": 0.0485,
+      "step": 2960
+    },
+    {
+      "epoch": 2.731549815498155,
+      "grad_norm": 0.35712773344744997,
+      "learning_rate": 1.2186152382566763e-06,
+      "loss": 0.0536,
+      "step": 2961
+    },
+    {
+      "epoch": 2.732472324723247,
+      "grad_norm": 0.35871029191430415,
+      "learning_rate": 1.2103507473317371e-06,
+      "loss": 0.0562,
+      "step": 2962
+    },
+    {
+      "epoch": 2.7333948339483394,
+      "grad_norm": 0.4105326655298472,
+      "learning_rate": 1.2021136809308386e-06,
+      "loss": 0.0673,
+      "step": 2963
+    },
+    {
+      "epoch": 2.734317343173432,
+      "grad_norm": 0.3629105146533966,
+      "learning_rate": 1.1939040485496155e-06,
+      "loss": 0.0597,
+      "step": 2964
+    },
+    {
+      "epoch": 2.735239852398524,
+      "grad_norm": 0.47648826594715293,
+      "learning_rate": 1.1857218596520586e-06,
+      "loss": 0.0645,
+      "step": 2965
+    },
+    {
+      "epoch": 2.736162361623616,
+      "grad_norm": 0.4057763435449541,
+      "learning_rate": 1.1775671236705365e-06,
+      "loss": 0.0644,
+      "step": 2966
+    },
+    {
+      "epoch": 2.7370848708487086,
+      "grad_norm": 0.41786394043315017,
+      "learning_rate": 1.1694398500057714e-06,
+      "loss": 0.0614,
+      "step": 2967
+    },
+    {
+      "epoch": 2.7380073800738005,
+      "grad_norm": 0.37695759953636465,
+      "learning_rate": 1.1613400480268099e-06,
+      "loss": 0.0545,
+      "step": 2968
+    },
+    {
+      "epoch": 2.738929889298893,
+      "grad_norm": 0.42107806972519235,
+      "learning_rate": 1.1532677270710501e-06,
+      "loss": 0.0579,
+      "step": 2969
+    },
+    {
+      "epoch": 2.7398523985239853,
+      "grad_norm": 0.36109942662296984,
+      "learning_rate": 1.1452228964442007e-06,
+      "loss": 0.0552,
+      "step": 2970
+    },
+    {
+      "epoch": 2.7407749077490777,
+      "grad_norm": 0.38851677479381025,
+      "learning_rate": 1.1372055654202768e-06,
+      "loss": 0.061,
+      "step": 2971
+    },
+    {
+      "epoch": 2.7416974169741697,
+      "grad_norm": 0.4020199566520895,
+      "learning_rate": 1.1292157432415962e-06,
+      "loss": 0.0538,
+      "step": 2972
+    },
+    {
+      "epoch": 2.742619926199262,
+      "grad_norm": 0.37460538612402966,
+      "learning_rate": 1.121253439118769e-06,
+      "loss": 0.0545,
+      "step": 2973
+    },
+    {
+      "epoch": 2.743542435424354,
+      "grad_norm": 0.3474712521462762,
+      "learning_rate": 1.1133186622306724e-06,
+      "loss": 0.0528,
+      "step": 2974
+    },
+    {
+      "epoch": 2.7444649446494465,
+      "grad_norm": 0.3499998332064438,
+      "learning_rate": 1.105411421724456e-06,
+      "loss": 0.0576,
+      "step": 2975
+    },
+    {
+      "epoch": 2.745387453874539,
+      "grad_norm": 0.3491895773043017,
+      "learning_rate": 1.0975317267155283e-06,
+      "loss": 0.0557,
+      "step": 2976
+    },
+    {
+      "epoch": 2.7463099630996313,
+      "grad_norm": 0.43970555215228807,
+      "learning_rate": 1.0896795862875425e-06,
+      "loss": 0.0649,
+      "step": 2977
+    },
+    {
+      "epoch": 2.7472324723247232,
+      "grad_norm": 0.35652163871423376,
+      "learning_rate": 1.081855009492383e-06,
+      "loss": 0.0535,
+      "step": 2978
+    },
+    {
+      "epoch": 2.7481549815498156,
+      "grad_norm": 0.3621997806387468,
+      "learning_rate": 1.0740580053501592e-06,
+      "loss": 0.0544,
+      "step": 2979
+    },
+    {
+      "epoch": 2.7490774907749076,
+      "grad_norm": 0.3697751229366689,
+      "learning_rate": 1.0662885828492036e-06,
+      "loss": 0.0529,
+      "step": 2980
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.39400988537913456,
+      "learning_rate": 1.0585467509460378e-06,
+      "loss": 0.0525,
+      "step": 2981
+    },
+    {
+      "epoch": 2.7509225092250924,
+      "grad_norm": 0.3465690456368866,
+      "learning_rate": 1.0508325185653921e-06,
+      "loss": 0.0551,
+      "step": 2982
+    },
+    {
+      "epoch": 2.7518450184501844,
+      "grad_norm": 0.4509089421040558,
+      "learning_rate": 1.0431458946001754e-06,
+      "loss": 0.0578,
+      "step": 2983
+    },
+    {
+      "epoch": 2.7527675276752768,
+      "grad_norm": 0.36956723380857287,
+      "learning_rate": 1.035486887911466e-06,
+      "loss": 0.051,
+      "step": 2984
+    },
+    {
+      "epoch": 2.7536900369003687,
+      "grad_norm": 0.4224287758338531,
+      "learning_rate": 1.027855507328504e-06,
+      "loss": 0.0603,
+      "step": 2985
+    },
+    {
+      "epoch": 2.754612546125461,
+      "grad_norm": 0.3681992863929429,
+      "learning_rate": 1.0202517616486911e-06,
+      "loss": 0.0589,
+      "step": 2986
+    },
+    {
+      "epoch": 2.7555350553505535,
+      "grad_norm": 0.3675789213367372,
+      "learning_rate": 1.0126756596375686e-06,
+      "loss": 0.0547,
+      "step": 2987
+    },
+    {
+      "epoch": 2.756457564575646,
+      "grad_norm": 0.4003914269444886,
+      "learning_rate": 1.0051272100287974e-06,
+      "loss": 0.0635,
+      "step": 2988
+    },
+    {
+      "epoch": 2.757380073800738,
+      "grad_norm": 0.3809235673066969,
+      "learning_rate": 9.97606421524186e-07,
+      "loss": 0.0597,
+      "step": 2989
+    },
+    {
+      "epoch": 2.7583025830258303,
+      "grad_norm": 0.34379817476614527,
+      "learning_rate": 9.901133027936326e-07,
+      "loss": 0.0539,
+      "step": 2990
+    },
+    {
+      "epoch": 2.7592250922509223,
+      "grad_norm": 0.36702117020991015,
+      "learning_rate": 9.826478624751445e-07,
+      "loss": 0.0532,
+      "step": 2991
+    },
+    {
+      "epoch": 2.7601476014760147,
+      "grad_norm": 0.44445769585444944,
+      "learning_rate": 9.752101091748345e-07,
+      "loss": 0.0621,
+      "step": 2992
+    },
+    {
+      "epoch": 2.761070110701107,
+      "grad_norm": 0.4537384075371174,
+      "learning_rate": 9.6780005146688e-07,
+      "loss": 0.066,
+      "step": 2993
+    },
+    {
+      "epoch": 2.7619926199261995,
+      "grad_norm": 0.39150259327252046,
+      "learning_rate": 9.604176978935343e-07,
+      "loss": 0.0604,
+      "step": 2994
+    },
+    {
+      "epoch": 2.7629151291512914,
+      "grad_norm": 0.40388872411212934,
+      "learning_rate": 9.530630569651255e-07,
+      "loss": 0.0618,
+      "step": 2995
+    },
+    {
+      "epoch": 2.763837638376384,
+      "grad_norm": 0.34396772103566386,
+      "learning_rate": 9.457361371600249e-07,
+      "loss": 0.0513,
+      "step": 2996
+    },
+    {
+      "epoch": 2.764760147601476,
+      "grad_norm": 0.4942270722001889,
+      "learning_rate": 9.384369469246452e-07,
+      "loss": 0.0646,
+      "step": 2997
+    },
+    {
+      "epoch": 2.765682656826568,
+      "grad_norm": 0.36904705364153034,
+      "learning_rate": 9.311654946734388e-07,
+      "loss": 0.0626,
+      "step": 2998
+    },
+    {
+      "epoch": 2.7666051660516606,
+      "grad_norm": 0.4068666991950199,
+      "learning_rate": 9.23921788788884e-07,
+      "loss": 0.0714,
+      "step": 2999
+    },
+    {
+      "epoch": 2.767527675276753,
+      "grad_norm": 0.43526029054214177,
+      "learning_rate": 9.167058376214621e-07,
+      "loss": 0.0593,
+      "step": 3000
+    },
+    {
+      "epoch": 2.768450184501845,
+      "grad_norm": 0.37761165148831355,
+      "learning_rate": 9.095176494896663e-07,
+      "loss": 0.056,
+      "step": 3001
+    },
+    {
+      "epoch": 2.7693726937269374,
+      "grad_norm": 0.3646292823550303,
+      "learning_rate": 9.023572326799929e-07,
+      "loss": 0.0539,
+      "step": 3002
+    },
+    {
+      "epoch": 2.7702952029520294,
+      "grad_norm": 0.3593927712199849,
+      "learning_rate": 8.952245954469057e-07,
+      "loss": 0.054,
+      "step": 3003
+    },
+    {
+      "epoch": 2.7712177121771218,
+      "grad_norm": 0.44990180566298993,
+      "learning_rate": 8.881197460128581e-07,
+      "loss": 0.0659,
+      "step": 3004
+    },
+    {
+      "epoch": 2.772140221402214,
+      "grad_norm": 0.3712113285117311,
+      "learning_rate": 8.81042692568268e-07,
+      "loss": 0.055,
+      "step": 3005
+    },
+    {
+      "epoch": 2.773062730627306,
+      "grad_norm": 0.3705332344243857,
+      "learning_rate": 8.739934432715035e-07,
+      "loss": 0.0632,
+      "step": 3006
+    },
+    {
+      "epoch": 2.7739852398523985,
+      "grad_norm": 0.3205372003448543,
+      "learning_rate": 8.66972006248884e-07,
+      "loss": 0.0575,
+      "step": 3007
+    },
+    {
+      "epoch": 2.774907749077491,
+      "grad_norm": 0.4018872116949969,
+      "learning_rate": 8.599783895946761e-07,
+      "loss": 0.0638,
+      "step": 3008
+    },
+    {
+      "epoch": 2.775830258302583,
+      "grad_norm": 0.36652579061351526,
+      "learning_rate": 8.53012601371056e-07,
+      "loss": 0.0578,
+      "step": 3009
+    },
+    {
+      "epoch": 2.7767527675276753,
+      "grad_norm": 0.3202958884783173,
+      "learning_rate": 8.460746496081362e-07,
+      "loss": 0.0498,
+      "step": 3010
+    },
+    {
+      "epoch": 2.7776752767527677,
+      "grad_norm": 0.37852921602778405,
+      "learning_rate": 8.391645423039357e-07,
+      "loss": 0.0557,
+      "step": 3011
+    },
+    {
+      "epoch": 2.7785977859778597,
+      "grad_norm": 0.35706112242489096,
+      "learning_rate": 8.322822874243686e-07,
+      "loss": 0.0565,
+      "step": 3012
+    },
+    {
+      "epoch": 2.779520295202952,
+      "grad_norm": 0.4014573659348028,
+      "learning_rate": 8.254278929032494e-07,
+      "loss": 0.0551,
+      "step": 3013
+    },
+    {
+      "epoch": 2.780442804428044,
+      "grad_norm": 0.37341034836485737,
+      "learning_rate": 8.186013666422687e-07,
+      "loss": 0.0553,
+      "step": 3014
+    },
+    {
+      "epoch": 2.7813653136531364,
+      "grad_norm": 0.3703888430668142,
+      "learning_rate": 8.118027165109926e-07,
+      "loss": 0.0567,
+      "step": 3015
+    },
+    {
+      "epoch": 2.782287822878229,
+      "grad_norm": 0.36963410147160064,
+      "learning_rate": 8.050319503468546e-07,
+      "loss": 0.0609,
+      "step": 3016
+    },
+    {
+      "epoch": 2.7832103321033212,
+      "grad_norm": 0.45553415284580095,
+      "learning_rate": 7.982890759551415e-07,
+      "loss": 0.0763,
+      "step": 3017
+    },
+    {
+      "epoch": 2.784132841328413,
+      "grad_norm": 0.3528517617568652,
+      "learning_rate": 7.915741011089855e-07,
+      "loss": 0.0559,
+      "step": 3018
+    },
+    {
+      "epoch": 2.7850553505535056,
+      "grad_norm": 0.39876071016487835,
+      "learning_rate": 7.848870335493613e-07,
+      "loss": 0.0682,
+      "step": 3019
+    },
+    {
+      "epoch": 2.7859778597785976,
+      "grad_norm": 0.344238782439648,
+      "learning_rate": 7.78227880985058e-07,
+      "loss": 0.052,
+      "step": 3020
+    },
+    {
+      "epoch": 2.78690036900369,
+      "grad_norm": 0.47872136961864875,
+      "learning_rate": 7.715966510927097e-07,
+      "loss": 0.0602,
+      "step": 3021
+    },
+    {
+      "epoch": 2.7878228782287824,
+      "grad_norm": 0.35265544529469695,
+      "learning_rate": 7.649933515167407e-07,
+      "loss": 0.0561,
+      "step": 3022
+    },
+    {
+      "epoch": 2.788745387453875,
+      "grad_norm": 0.348002235267397,
+      "learning_rate": 7.584179898693783e-07,
+      "loss": 0.0554,
+      "step": 3023
+    },
+    {
+      "epoch": 2.7896678966789668,
+      "grad_norm": 0.38996218600694976,
+      "learning_rate": 7.518705737306591e-07,
+      "loss": 0.0621,
+      "step": 3024
+    },
+    {
+      "epoch": 2.790590405904059,
+      "grad_norm": 0.35492438028981754,
+      "learning_rate": 7.453511106483902e-07,
+      "loss": 0.0535,
+      "step": 3025
+    },
+    {
+      "epoch": 2.791512915129151,
+      "grad_norm": 0.37061112978174804,
+      "learning_rate": 7.38859608138151e-07,
+      "loss": 0.0557,
+      "step": 3026
+    },
+    {
+      "epoch": 2.7924354243542435,
+      "grad_norm": 0.4094905475119617,
+      "learning_rate": 7.323960736833057e-07,
+      "loss": 0.0702,
+      "step": 3027
+    },
+    {
+      "epoch": 2.793357933579336,
+      "grad_norm": 0.3645171489168169,
+      "learning_rate": 7.259605147349608e-07,
+      "loss": 0.0578,
+      "step": 3028
+    },
+    {
+      "epoch": 2.794280442804428,
+      "grad_norm": 0.4538889363916288,
+      "learning_rate": 7.195529387119815e-07,
+      "loss": 0.0597,
+      "step": 3029
+    },
+    {
+      "epoch": 2.7952029520295203,
+      "grad_norm": 0.34334974001270563,
+      "learning_rate": 7.131733530009704e-07,
+      "loss": 0.0547,
+      "step": 3030
+    },
+    {
+      "epoch": 2.7961254612546127,
+      "grad_norm": 0.3713757367965027,
+      "learning_rate": 7.06821764956267e-07,
+      "loss": 0.0613,
+      "step": 3031
+    },
+    {
+      "epoch": 2.7970479704797047,
+      "grad_norm": 0.3095422789862229,
+      "learning_rate": 7.004981818999279e-07,
+      "loss": 0.0508,
+      "step": 3032
+    },
+    {
+      "epoch": 2.797970479704797,
+      "grad_norm": 0.3557566485745798,
+      "learning_rate": 6.942026111217359e-07,
+      "loss": 0.0509,
+      "step": 3033
+    },
+    {
+      "epoch": 2.7988929889298895,
+      "grad_norm": 0.412387856193388,
+      "learning_rate": 6.879350598791772e-07,
+      "loss": 0.0572,
+      "step": 3034
+    },
+    {
+      "epoch": 2.7998154981549814,
+      "grad_norm": 0.3548976624104618,
+      "learning_rate": 6.816955353974335e-07,
+      "loss": 0.0608,
+      "step": 3035
+    },
+    {
+      "epoch": 2.800738007380074,
+      "grad_norm": 0.38439505994481193,
+      "learning_rate": 6.754840448693789e-07,
+      "loss": 0.061,
+      "step": 3036
+    },
+    {
+      "epoch": 2.801660516605166,
+      "grad_norm": 0.33877961758384073,
+      "learning_rate": 6.69300595455577e-07,
+      "loss": 0.0543,
+      "step": 3037
+    },
+    {
+      "epoch": 2.802583025830258,
+      "grad_norm": 0.41338407939341637,
+      "learning_rate": 6.631451942842565e-07,
+      "loss": 0.0541,
+      "step": 3038
+    },
+    {
+      "epoch": 2.8035055350553506,
+      "grad_norm": 0.4324337931055607,
+      "learning_rate": 6.570178484513162e-07,
+      "loss": 0.0556,
+      "step": 3039
+    },
+    {
+      "epoch": 2.804428044280443,
+      "grad_norm": 0.3658053320531132,
+      "learning_rate": 6.50918565020317e-07,
+      "loss": 0.0607,
+      "step": 3040
+    },
+    {
+      "epoch": 2.805350553505535,
+      "grad_norm": 0.40511232536405956,
+      "learning_rate": 6.448473510224595e-07,
+      "loss": 0.065,
+      "step": 3041
+    },
+    {
+      "epoch": 2.8062730627306274,
+      "grad_norm": 0.3544204839763948,
+      "learning_rate": 6.388042134565953e-07,
+      "loss": 0.0555,
+      "step": 3042
+    },
+    {
+      "epoch": 2.8071955719557193,
+      "grad_norm": 0.37855795799058395,
+      "learning_rate": 6.327891592892127e-07,
+      "loss": 0.0515,
+      "step": 3043
+    },
+    {
+      "epoch": 2.8081180811808117,
+      "grad_norm": 0.4186095737100796,
+      "learning_rate": 6.268021954544096e-07,
+      "loss": 0.06,
+      "step": 3044
+    },
+    {
+      "epoch": 2.809040590405904,
+      "grad_norm": 0.34904205929911225,
+      "learning_rate": 6.208433288539178e-07,
+      "loss": 0.0546,
+      "step": 3045
+    },
+    {
+      "epoch": 2.8099630996309966,
+      "grad_norm": 0.38973795389147187,
+      "learning_rate": 6.149125663570732e-07,
+      "loss": 0.055,
+      "step": 3046
+    },
+    {
+      "epoch": 2.8108856088560885,
+      "grad_norm": 0.4023007894124464,
+      "learning_rate": 6.090099148008094e-07,
+      "loss": 0.0696,
+      "step": 3047
+    },
+    {
+      "epoch": 2.811808118081181,
+      "grad_norm": 0.3513069846733514,
+      "learning_rate": 6.031353809896611e-07,
+      "loss": 0.0491,
+      "step": 3048
+    },
+    {
+      "epoch": 2.812730627306273,
+      "grad_norm": 0.46942274505022347,
+      "learning_rate": 5.97288971695742e-07,
+      "loss": 0.0774,
+      "step": 3049
+    },
+    {
+      "epoch": 2.8136531365313653,
+      "grad_norm": 0.3746075770999299,
+      "learning_rate": 5.914706936587494e-07,
+      "loss": 0.067,
+      "step": 3050
+    },
+    {
+      "epoch": 2.8145756457564577,
+      "grad_norm": 0.4334892424883726,
+      "learning_rate": 5.856805535859516e-07,
+      "loss": 0.0543,
+      "step": 3051
+    },
+    {
+      "epoch": 2.8154981549815496,
+      "grad_norm": 0.36292956029933826,
+      "learning_rate": 5.799185581521732e-07,
+      "loss": 0.0546,
+      "step": 3052
+    },
+    {
+      "epoch": 2.816420664206642,
+      "grad_norm": 0.4310683091039621,
+      "learning_rate": 5.741847139998008e-07,
+      "loss": 0.0653,
+      "step": 3053
+    },
+    {
+      "epoch": 2.8173431734317345,
+      "grad_norm": 0.3617562819256405,
+      "learning_rate": 5.684790277387663e-07,
+      "loss": 0.0649,
+      "step": 3054
+    },
+    {
+      "epoch": 2.8182656826568264,
+      "grad_norm": 0.3570502861875989,
+      "learning_rate": 5.628015059465363e-07,
+      "loss": 0.053,
+      "step": 3055
+    },
+    {
+      "epoch": 2.819188191881919,
+      "grad_norm": 0.41306359141018206,
+      "learning_rate": 5.571521551681169e-07,
+      "loss": 0.0596,
+      "step": 3056
+    },
+    {
+      "epoch": 2.8201107011070112,
+      "grad_norm": 0.35171632738645336,
+      "learning_rate": 5.515309819160403e-07,
+      "loss": 0.0551,
+      "step": 3057
+    },
+    {
+      "epoch": 2.821033210332103,
+      "grad_norm": 0.38256128198924205,
+      "learning_rate": 5.45937992670345e-07,
+      "loss": 0.0587,
+      "step": 3058
+    },
+    {
+      "epoch": 2.8219557195571956,
+      "grad_norm": 0.35631455510809057,
+      "learning_rate": 5.403731938785878e-07,
+      "loss": 0.0534,
+      "step": 3059
+    },
+    {
+      "epoch": 2.8228782287822876,
+      "grad_norm": 0.3913909917635343,
+      "learning_rate": 5.348365919558285e-07,
+      "loss": 0.0665,
+      "step": 3060
+    },
+    {
+      "epoch": 2.82380073800738,
+      "grad_norm": 0.3976575641492326,
+      "learning_rate": 5.293281932846145e-07,
+      "loss": 0.067,
+      "step": 3061
+    },
+    {
+      "epoch": 2.8247232472324724,
+      "grad_norm": 0.3878091048370286,
+      "learning_rate": 5.238480042149913e-07,
+      "loss": 0.0611,
+      "step": 3062
+    },
+    {
+      "epoch": 2.8256457564575648,
+      "grad_norm": 0.44775564598488815,
+      "learning_rate": 5.183960310644748e-07,
+      "loss": 0.0642,
+      "step": 3063
+    },
+    {
+      "epoch": 2.8265682656826567,
+      "grad_norm": 0.39969690272794917,
+      "learning_rate": 5.129722801180542e-07,
+      "loss": 0.0624,
+      "step": 3064
+    },
+    {
+      "epoch": 2.827490774907749,
+      "grad_norm": 0.40145395859816396,
+      "learning_rate": 5.07576757628192e-07,
+      "loss": 0.06,
+      "step": 3065
+    },
+    {
+      "epoch": 2.828413284132841,
+      "grad_norm": 0.35038941278206576,
+      "learning_rate": 5.022094698148072e-07,
+      "loss": 0.0603,
+      "step": 3066
+    },
+    {
+      "epoch": 2.8293357933579335,
+      "grad_norm": 0.3912530660871608,
+      "learning_rate": 4.968704228652643e-07,
+      "loss": 0.0679,
+      "step": 3067
+    },
+    {
+      "epoch": 2.830258302583026,
+      "grad_norm": 0.4503668889861151,
+      "learning_rate": 4.915596229343733e-07,
+      "loss": 0.0713,
+      "step": 3068
+    },
+    {
+      "epoch": 2.8311808118081183,
+      "grad_norm": 0.40552994897158007,
+      "learning_rate": 4.862770761443896e-07,
+      "loss": 0.0529,
+      "step": 3069
+    },
+    {
+      "epoch": 2.8321033210332103,
+      "grad_norm": 0.3965084132876259,
+      "learning_rate": 4.810227885849866e-07,
+      "loss": 0.0586,
+      "step": 3070
+    },
+    {
+      "epoch": 2.8330258302583027,
+      "grad_norm": 0.46853938963440855,
+      "learning_rate": 4.75796766313269e-07,
+      "loss": 0.0713,
+      "step": 3071
+    },
+    {
+      "epoch": 2.8339483394833946,
+      "grad_norm": 0.37870988056907734,
+      "learning_rate": 4.705990153537565e-07,
+      "loss": 0.0585,
+      "step": 3072
+    },
+    {
+      "epoch": 2.834870848708487,
+      "grad_norm": 0.4048309388437559,
+      "learning_rate": 4.654295416983728e-07,
+      "loss": 0.0608,
+      "step": 3073
+    },
+    {
+      "epoch": 2.8357933579335795,
+      "grad_norm": 0.37403675447498214,
+      "learning_rate": 4.602883513064482e-07,
+      "loss": 0.0514,
+      "step": 3074
+    },
+    {
+      "epoch": 2.836715867158672,
+      "grad_norm": 0.4200123129227064,
+      "learning_rate": 4.551754501047084e-07,
+      "loss": 0.0607,
+      "step": 3075
+    },
+    {
+      "epoch": 2.837638376383764,
+      "grad_norm": 0.35394890850486577,
+      "learning_rate": 4.500908439872664e-07,
+      "loss": 0.0522,
+      "step": 3076
+    },
+    {
+      "epoch": 2.838560885608856,
+      "grad_norm": 0.4372989175172387,
+      "learning_rate": 4.4503453881561407e-07,
+      "loss": 0.0607,
+      "step": 3077
+    },
+    {
+      "epoch": 2.839483394833948,
+      "grad_norm": 0.4151231584289497,
+      "learning_rate": 4.4000654041862764e-07,
+      "loss": 0.0661,
+      "step": 3078
+    },
+    {
+      "epoch": 2.8404059040590406,
+      "grad_norm": 0.38509653531385357,
+      "learning_rate": 4.350068545925373e-07,
+      "loss": 0.0603,
+      "step": 3079
+    },
+    {
+      "epoch": 2.841328413284133,
+      "grad_norm": 0.3596935660857444,
+      "learning_rate": 4.300354871009465e-07,
+      "loss": 0.0637,
+      "step": 3080
+    },
+    {
+      "epoch": 2.842250922509225,
+      "grad_norm": 0.4119426560230551,
+      "learning_rate": 4.2509244367480994e-07,
+      "loss": 0.0618,
+      "step": 3081
+    },
+    {
+      "epoch": 2.8431734317343174,
+      "grad_norm": 0.34913464173804626,
+      "learning_rate": 4.201777300124249e-07,
+      "loss": 0.0519,
+      "step": 3082
+    },
+    {
+      "epoch": 2.8440959409594093,
+      "grad_norm": 0.39015850377702105,
+      "learning_rate": 4.152913517794399e-07,
+      "loss": 0.058,
+      "step": 3083
+    },
+    {
+      "epoch": 2.8450184501845017,
+      "grad_norm": 0.48286557292257426,
+      "learning_rate": 4.104333146088379e-07,
+      "loss": 0.0673,
+      "step": 3084
+    },
+    {
+      "epoch": 2.845940959409594,
+      "grad_norm": 0.4101184601988959,
+      "learning_rate": 4.0560362410091704e-07,
+      "loss": 0.0609,
+      "step": 3085
+    },
+    {
+      "epoch": 2.8468634686346865,
+      "grad_norm": 0.4152920275704722,
+      "learning_rate": 4.0080228582331234e-07,
+      "loss": 0.0611,
+      "step": 3086
+    },
+    {
+      "epoch": 2.8477859778597785,
+      "grad_norm": 0.3790564708187215,
+      "learning_rate": 3.960293053109687e-07,
+      "loss": 0.06,
+      "step": 3087
+    },
+    {
+      "epoch": 2.848708487084871,
+      "grad_norm": 0.4192336015241333,
+      "learning_rate": 3.9128468806614306e-07,
+      "loss": 0.0592,
+      "step": 3088
+    },
+    {
+      "epoch": 2.849630996309963,
+      "grad_norm": 0.3809704162794546,
+      "learning_rate": 3.8656843955839075e-07,
+      "loss": 0.0528,
+      "step": 3089
+    },
+    {
+      "epoch": 2.8505535055350553,
+      "grad_norm": 0.36785498805492256,
+      "learning_rate": 3.818805652245683e-07,
+      "loss": 0.0592,
+      "step": 3090
+    },
+    {
+      "epoch": 2.8514760147601477,
+      "grad_norm": 0.3940379277381955,
+      "learning_rate": 3.7722107046882226e-07,
+      "loss": 0.0511,
+      "step": 3091
+    },
+    {
+      "epoch": 2.85239852398524,
+      "grad_norm": 0.39381179296698055,
+      "learning_rate": 3.7258996066258103e-07,
+      "loss": 0.0614,
+      "step": 3092
+    },
+    {
+      "epoch": 2.853321033210332,
+      "grad_norm": 0.38861666965095293,
+      "learning_rate": 3.67987241144549e-07,
+      "loss": 0.0589,
+      "step": 3093
+    },
+    {
+      "epoch": 2.8542435424354244,
+      "grad_norm": 0.3693220330097373,
+      "learning_rate": 3.6341291722070956e-07,
+      "loss": 0.055,
+      "step": 3094
+    },
+    {
+      "epoch": 2.8551660516605164,
+      "grad_norm": 0.4015995540257464,
+      "learning_rate": 3.588669941643086e-07,
+      "loss": 0.0625,
+      "step": 3095
+    },
+    {
+      "epoch": 2.856088560885609,
+      "grad_norm": 0.3949330924315789,
+      "learning_rate": 3.5434947721584846e-07,
+      "loss": 0.0566,
+      "step": 3096
+    },
+    {
+      "epoch": 2.857011070110701,
+      "grad_norm": 0.41813829293464966,
+      "learning_rate": 3.498603715830884e-07,
+      "loss": 0.062,
+      "step": 3097
+    },
+    {
+      "epoch": 2.8579335793357936,
+      "grad_norm": 0.34104975646802455,
+      "learning_rate": 3.453996824410388e-07,
+      "loss": 0.0552,
+      "step": 3098
+    },
+    {
+      "epoch": 2.8588560885608856,
+      "grad_norm": 0.37425611677733905,
+      "learning_rate": 3.4096741493194197e-07,
+      "loss": 0.0543,
+      "step": 3099
+    },
+    {
+      "epoch": 2.859778597785978,
+      "grad_norm": 0.43738932187425617,
+      "learning_rate": 3.3656357416528285e-07,
+      "loss": 0.0549,
+      "step": 3100
+    },
+    {
+      "epoch": 2.86070110701107,
+      "grad_norm": 0.3731623029223221,
+      "learning_rate": 3.321881652177783e-07,
+      "loss": 0.051,
+      "step": 3101
+    },
+    {
+      "epoch": 2.8616236162361623,
+      "grad_norm": 0.4723027130587314,
+      "learning_rate": 3.2784119313336305e-07,
+      "loss": 0.0611,
+      "step": 3102
+    },
+    {
+      "epoch": 2.8625461254612548,
+      "grad_norm": 0.3851958403736879,
+      "learning_rate": 3.2352266292319243e-07,
+      "loss": 0.0696,
+      "step": 3103
+    },
+    {
+      "epoch": 2.8634686346863467,
+      "grad_norm": 0.36550780829170026,
+      "learning_rate": 3.1923257956563703e-07,
+      "loss": 0.0606,
+      "step": 3104
+    },
+    {
+      "epoch": 2.864391143911439,
+      "grad_norm": 0.3670853621802748,
+      "learning_rate": 3.1497094800627124e-07,
+      "loss": 0.0474,
+      "step": 3105
+    },
+    {
+      "epoch": 2.8653136531365315,
+      "grad_norm": 0.38481821579325054,
+      "learning_rate": 3.107377731578709e-07,
+      "loss": 0.0584,
+      "step": 3106
+    },
+    {
+      "epoch": 2.8662361623616235,
+      "grad_norm": 0.43307530254517385,
+      "learning_rate": 3.0653305990040736e-07,
+      "loss": 0.0648,
+      "step": 3107
+    },
+    {
+      "epoch": 2.867158671586716,
+      "grad_norm": 0.3870432354135556,
+      "learning_rate": 3.0235681308103945e-07,
+      "loss": 0.0547,
+      "step": 3108
+    },
+    {
+      "epoch": 2.8680811808118083,
+      "grad_norm": 0.38574379242952656,
+      "learning_rate": 2.982090375141161e-07,
+      "loss": 0.0494,
+      "step": 3109
+    },
+    {
+      "epoch": 2.8690036900369003,
+      "grad_norm": 0.4294302467000169,
+      "learning_rate": 2.9408973798115967e-07,
+      "loss": 0.0678,
+      "step": 3110
+    },
+    {
+      "epoch": 2.8699261992619927,
+      "grad_norm": 0.4068011754116402,
+      "learning_rate": 2.899989192308633e-07,
+      "loss": 0.0524,
+      "step": 3111
+    },
+    {
+      "epoch": 2.8708487084870846,
+      "grad_norm": 0.3873339654924171,
+      "learning_rate": 2.859365859790963e-07,
+      "loss": 0.0477,
+      "step": 3112
+    },
+    {
+      "epoch": 2.871771217712177,
+      "grad_norm": 0.3587163133267895,
+      "learning_rate": 2.819027429088822e-07,
+      "loss": 0.0594,
+      "step": 3113
+    },
+    {
+      "epoch": 2.8726937269372694,
+      "grad_norm": 0.377462742092774,
+      "learning_rate": 2.7789739467040666e-07,
+      "loss": 0.0587,
+      "step": 3114
+    },
+    {
+      "epoch": 2.873616236162362,
+      "grad_norm": 0.420661122891959,
+      "learning_rate": 2.7392054588100127e-07,
+      "loss": 0.0651,
+      "step": 3115
+    },
+    {
+      "epoch": 2.874538745387454,
+      "grad_norm": 0.35825539443343773,
+      "learning_rate": 2.6997220112514877e-07,
+      "loss": 0.065,
+      "step": 3116
+    },
+    {
+      "epoch": 2.875461254612546,
+      "grad_norm": 0.36394797721574657,
+      "learning_rate": 2.660523649544666e-07,
+      "loss": 0.0646,
+      "step": 3117
+    },
+    {
+      "epoch": 2.876383763837638,
+      "grad_norm": 0.36620511469972156,
+      "learning_rate": 2.6216104188771793e-07,
+      "loss": 0.0623,
+      "step": 3118
+    },
+    {
+      "epoch": 2.8773062730627306,
+      "grad_norm": 0.42566183031464044,
+      "learning_rate": 2.5829823641078386e-07,
+      "loss": 0.063,
+      "step": 3119
+    },
+    {
+      "epoch": 2.878228782287823,
+      "grad_norm": 0.40045449113591813,
+      "learning_rate": 2.544639529766829e-07,
+      "loss": 0.059,
+      "step": 3120
+    },
+    {
+      "epoch": 2.8791512915129154,
+      "grad_norm": 0.43855184624490645,
+      "learning_rate": 2.506581960055432e-07,
+      "loss": 0.0577,
+      "step": 3121
+    },
+    {
+      "epoch": 2.8800738007380073,
+      "grad_norm": 0.3590229799292661,
+      "learning_rate": 2.4688096988461084e-07,
+      "loss": 0.057,
+      "step": 3122
+    },
+    {
+      "epoch": 2.8809963099630997,
+      "grad_norm": 0.40603600606145884,
+      "learning_rate": 2.431322789682444e-07,
+      "loss": 0.058,
+      "step": 3123
+    },
+    {
+      "epoch": 2.8819188191881917,
+      "grad_norm": 0.37206285598652145,
+      "learning_rate": 2.3941212757790934e-07,
+      "loss": 0.0556,
+      "step": 3124
+    },
+    {
+      "epoch": 2.882841328413284,
+      "grad_norm": 0.3825211879106796,
+      "learning_rate": 2.3572052000216393e-07,
+      "loss": 0.0523,
+      "step": 3125
+    },
+    {
+      "epoch": 2.8837638376383765,
+      "grad_norm": 0.4043895899196592,
+      "learning_rate": 2.3205746049666243e-07,
+      "loss": 0.0503,
+      "step": 3126
+    },
+    {
+      "epoch": 2.8846863468634685,
+      "grad_norm": 0.37371240332215466,
+      "learning_rate": 2.284229532841603e-07,
+      "loss": 0.0635,
+      "step": 3127
+    },
+    {
+      "epoch": 2.885608856088561,
+      "grad_norm": 0.36449477350329723,
+      "learning_rate": 2.2481700255447825e-07,
+      "loss": 0.0624,
+      "step": 3128
+    },
+    {
+      "epoch": 2.8865313653136533,
+      "grad_norm": 0.4207640198889781,
+      "learning_rate": 2.212396124645355e-07,
+      "loss": 0.0639,
+      "step": 3129
+    },
+    {
+      "epoch": 2.8874538745387452,
+      "grad_norm": 0.35639617746732283,
+      "learning_rate": 2.1769078713831935e-07,
+      "loss": 0.0521,
+      "step": 3130
+    },
+    {
+      "epoch": 2.8883763837638377,
+      "grad_norm": 0.349152719089145,
+      "learning_rate": 2.1417053066688787e-07,
+      "loss": 0.0579,
+      "step": 3131
+    },
+    {
+      "epoch": 2.88929889298893,
+      "grad_norm": 0.4019981344649913,
+      "learning_rate": 2.106788471083615e-07,
+      "loss": 0.0535,
+      "step": 3132
+    },
+    {
+      "epoch": 2.890221402214022,
+      "grad_norm": 0.40152901733528595,
+      "learning_rate": 2.0721574048793159e-07,
+      "loss": 0.0666,
+      "step": 3133
+    },
+    {
+      "epoch": 2.8911439114391144,
+      "grad_norm": 0.36928366702351667,
+      "learning_rate": 2.0378121479783796e-07,
+      "loss": 0.0629,
+      "step": 3134
+    },
+    {
+      "epoch": 2.8920664206642064,
+      "grad_norm": 0.4494398328104406,
+      "learning_rate": 2.0037527399737466e-07,
+      "loss": 0.0617,
+      "step": 3135
+    },
+    {
+      "epoch": 2.892988929889299,
+      "grad_norm": 0.39146009669205106,
+      "learning_rate": 1.9699792201288703e-07,
+      "loss": 0.0579,
+      "step": 3136
+    },
+    {
+      "epoch": 2.893911439114391,
+      "grad_norm": 0.39069058809145935,
+      "learning_rate": 1.936491627377579e-07,
+      "loss": 0.0543,
+      "step": 3137
+    },
+    {
+      "epoch": 2.8948339483394836,
+      "grad_norm": 0.39218651628686196,
+      "learning_rate": 1.9032900003241315e-07,
+      "loss": 0.062,
+      "step": 3138
+    },
+    {
+      "epoch": 2.8957564575645756,
+      "grad_norm": 0.38327349989050385,
+      "learning_rate": 1.870374377243078e-07,
+      "loss": 0.0669,
+      "step": 3139
+    },
+    {
+      "epoch": 2.896678966789668,
+      "grad_norm": 0.38666177814617686,
+      "learning_rate": 1.837744796079288e-07,
+      "loss": 0.0521,
+      "step": 3140
+    },
+    {
+      "epoch": 2.89760147601476,
+      "grad_norm": 0.3599162209804183,
+      "learning_rate": 1.8054012944479227e-07,
+      "loss": 0.0538,
+      "step": 3141
+    },
+    {
+      "epoch": 2.8985239852398523,
+      "grad_norm": 0.3953192653422239,
+      "learning_rate": 1.7733439096343508e-07,
+      "loss": 0.057,
+      "step": 3142
+    },
+    {
+      "epoch": 2.8994464944649447,
+      "grad_norm": 0.3967137925693252,
+      "learning_rate": 1.7415726785939834e-07,
+      "loss": 0.0662,
+      "step": 3143
+    },
+    {
+      "epoch": 2.900369003690037,
+      "grad_norm": 0.4078272572296947,
+      "learning_rate": 1.7100876379525232e-07,
+      "loss": 0.0686,
+      "step": 3144
+    },
+    {
+      "epoch": 2.901291512915129,
+      "grad_norm": 0.39070239748765256,
+      "learning_rate": 1.6788888240056865e-07,
+      "loss": 0.0591,
+      "step": 3145
+    },
+    {
+      "epoch": 2.9022140221402215,
+      "grad_norm": 0.3726091345240525,
+      "learning_rate": 1.647976272719176e-07,
+      "loss": 0.0545,
+      "step": 3146
+    },
+    {
+      "epoch": 2.9031365313653135,
+      "grad_norm": 0.41481503787534957,
+      "learning_rate": 1.6173500197287638e-07,
+      "loss": 0.0594,
+      "step": 3147
+    },
+    {
+      "epoch": 2.904059040590406,
+      "grad_norm": 0.3794000246422108,
+      "learning_rate": 1.5870101003402084e-07,
+      "loss": 0.0597,
+      "step": 3148
+    },
+    {
+      "epoch": 2.9049815498154983,
+      "grad_norm": 0.3369263864135048,
+      "learning_rate": 1.5569565495290318e-07,
+      "loss": 0.0522,
+      "step": 3149
+    },
+    {
+      "epoch": 2.9059040590405907,
+      "grad_norm": 0.40960030015767807,
+      "learning_rate": 1.5271894019408262e-07,
+      "loss": 0.0596,
+      "step": 3150
+    },
+    {
+      "epoch": 2.9068265682656826,
+      "grad_norm": 0.3441664544752149,
+      "learning_rate": 1.4977086918908923e-07,
+      "loss": 0.0486,
+      "step": 3151
+    },
+    {
+      "epoch": 2.907749077490775,
+      "grad_norm": 0.4139275443474721,
+      "learning_rate": 1.4685144533643502e-07,
+      "loss": 0.058,
+      "step": 3152
+    },
+    {
+      "epoch": 2.908671586715867,
+      "grad_norm": 0.4035398923339222,
+      "learning_rate": 1.43960672001614e-07,
+      "loss": 0.0594,
+      "step": 3153
+    },
+    {
+      "epoch": 2.9095940959409594,
+      "grad_norm": 0.37766409290452785,
+      "learning_rate": 1.410985525170827e-07,
+      "loss": 0.0615,
+      "step": 3154
+    },
+    {
+      "epoch": 2.910516605166052,
+      "grad_norm": 0.4460254147855921,
+      "learning_rate": 1.3826509018227128e-07,
+      "loss": 0.0681,
+      "step": 3155
+    },
+    {
+      "epoch": 2.911439114391144,
+      "grad_norm": 0.385927928269677,
+      "learning_rate": 1.3546028826357527e-07,
+      "loss": 0.0576,
+      "step": 3156
+    },
+    {
+      "epoch": 2.912361623616236,
+      "grad_norm": 0.39903411242617903,
+      "learning_rate": 1.3268414999434985e-07,
+      "loss": 0.0577,
+      "step": 3157
+    },
+    {
+      "epoch": 2.913284132841328,
+      "grad_norm": 0.431053626264846,
+      "learning_rate": 1.29936678574899e-07,
+      "loss": 0.0684,
+      "step": 3158
+    },
+    {
+      "epoch": 2.9142066420664205,
+      "grad_norm": 0.35442180641360227,
+      "learning_rate": 1.2721787717249466e-07,
+      "loss": 0.0573,
+      "step": 3159
+    },
+    {
+      "epoch": 2.915129151291513,
+      "grad_norm": 0.3463713119365184,
+      "learning_rate": 1.2452774892134368e-07,
+      "loss": 0.0589,
+      "step": 3160
+    },
+    {
+      "epoch": 2.9160516605166054,
+      "grad_norm": 0.3801898835041715,
+      "learning_rate": 1.2186629692260976e-07,
+      "loss": 0.0582,
+      "step": 3161
+    },
+    {
+      "epoch": 2.9169741697416973,
+      "grad_norm": 0.3660826668707689,
+      "learning_rate": 1.192335242443915e-07,
+      "loss": 0.0524,
+      "step": 3162
+    },
+    {
+      "epoch": 2.9178966789667897,
+      "grad_norm": 0.3605059072006074,
+      "learning_rate": 1.1662943392173053e-07,
+      "loss": 0.0548,
+      "step": 3163
+    },
+    {
+      "epoch": 2.9188191881918817,
+      "grad_norm": 0.3751885832042241,
+      "learning_rate": 1.1405402895660056e-07,
+      "loss": 0.0472,
+      "step": 3164
+    },
+    {
+      "epoch": 2.919741697416974,
+      "grad_norm": 0.4226757153010532,
+      "learning_rate": 1.1150731231791279e-07,
+      "loss": 0.0552,
+      "step": 3165
+    },
+    {
+      "epoch": 2.9206642066420665,
+      "grad_norm": 0.3982715198752065,
+      "learning_rate": 1.089892869414938e-07,
+      "loss": 0.0624,
+      "step": 3166
+    },
+    {
+      "epoch": 2.921586715867159,
+      "grad_norm": 0.41570676216447927,
+      "learning_rate": 1.0649995573011329e-07,
+      "loss": 0.0634,
+      "step": 3167
+    },
+    {
+      "epoch": 2.922509225092251,
+      "grad_norm": 0.37943720435725786,
+      "learning_rate": 1.0403932155344798e-07,
+      "loss": 0.0503,
+      "step": 3168
+    },
+    {
+      "epoch": 2.9234317343173433,
+      "grad_norm": 0.46216363117558795,
+      "learning_rate": 1.0160738724809549e-07,
+      "loss": 0.0612,
+      "step": 3169
+    },
+    {
+      "epoch": 2.9243542435424352,
+      "grad_norm": 0.42033750230395484,
+      "learning_rate": 9.920415561757712e-08,
+      "loss": 0.0595,
+      "step": 3170
+    },
+    {
+      "epoch": 2.9252767527675276,
+      "grad_norm": 0.3556062994479878,
+      "learning_rate": 9.682962943231843e-08,
+      "loss": 0.06,
+      "step": 3171
+    },
+    {
+      "epoch": 2.92619926199262,
+      "grad_norm": 0.3139807524312909,
+      "learning_rate": 9.448381142965201e-08,
+      "loss": 0.0488,
+      "step": 3172
+    },
+    {
+      "epoch": 2.9271217712177124,
+      "grad_norm": 0.3881161090492111,
+      "learning_rate": 9.216670431381747e-08,
+      "loss": 0.0526,
+      "step": 3173
+    },
+    {
+      "epoch": 2.9280442804428044,
+      "grad_norm": 0.42774485088737857,
+      "learning_rate": 8.98783107559642e-08,
+      "loss": 0.0683,
+      "step": 3174
+    },
+    {
+      "epoch": 2.928966789667897,
+      "grad_norm": 0.36789045233922624,
+      "learning_rate": 8.761863339412924e-08,
+      "loss": 0.0567,
+      "step": 3175
+    },
+    {
+      "epoch": 2.9298892988929888,
+      "grad_norm": 0.3992924122347115,
+      "learning_rate": 8.538767483325383e-08,
+      "loss": 0.0587,
+      "step": 3176
+    },
+    {
+      "epoch": 2.930811808118081,
+      "grad_norm": 0.35639401224290446,
+      "learning_rate": 8.318543764516961e-08,
+      "loss": 0.0457,
+      "step": 3177
+    },
+    {
+      "epoch": 2.9317343173431736,
+      "grad_norm": 0.4049382588541959,
+      "learning_rate": 8.101192436859584e-08,
+      "loss": 0.0597,
+      "step": 3178
+    },
+    {
+      "epoch": 2.9326568265682655,
+      "grad_norm": 0.37824866132664375,
+      "learning_rate": 7.886713750914487e-08,
+      "loss": 0.0604,
+      "step": 3179
+    },
+    {
+      "epoch": 2.933579335793358,
+      "grad_norm": 0.3785839291989575,
+      "learning_rate": 7.675107953931115e-08,
+      "loss": 0.0617,
+      "step": 3180
+    },
+    {
+      "epoch": 2.9345018450184504,
+      "grad_norm": 0.4385737390468286,
+      "learning_rate": 7.466375289846839e-08,
+      "loss": 0.0606,
+      "step": 3181
+    },
+    {
+      "epoch": 2.9354243542435423,
+      "grad_norm": 0.385430962252876,
+      "learning_rate": 7.260515999286677e-08,
+      "loss": 0.0639,
+      "step": 3182
+    },
+    {
+      "epoch": 2.9363468634686347,
+      "grad_norm": 0.4059174582314441,
+      "learning_rate": 7.057530319564409e-08,
+      "loss": 0.0593,
+      "step": 3183
+    },
+    {
+      "epoch": 2.937269372693727,
+      "grad_norm": 0.4007079161368407,
+      "learning_rate": 6.857418484679801e-08,
+      "loss": 0.0589,
+      "step": 3184
+    },
+    {
+      "epoch": 2.938191881918819,
+      "grad_norm": 0.41371594833639513,
+      "learning_rate": 6.660180725320542e-08,
+      "loss": 0.0557,
+      "step": 3185
+    },
+    {
+      "epoch": 2.9391143911439115,
+      "grad_norm": 0.38720152945378655,
+      "learning_rate": 6.465817268860586e-08,
+      "loss": 0.0611,
+      "step": 3186
+    },
+    {
+      "epoch": 2.9400369003690034,
+      "grad_norm": 0.46058747436877173,
+      "learning_rate": 6.274328339360703e-08,
+      "loss": 0.0623,
+      "step": 3187
+    },
+    {
+      "epoch": 2.940959409594096,
+      "grad_norm": 0.427316595808526,
+      "learning_rate": 6.085714157567646e-08,
+      "loss": 0.0624,
+      "step": 3188
+    },
+    {
+      "epoch": 2.9418819188191883,
+      "grad_norm": 0.36151995421843597,
+      "learning_rate": 5.89997494091471e-08,
+      "loss": 0.0578,
+      "step": 3189
+    },
+    {
+      "epoch": 2.9428044280442807,
+      "grad_norm": 0.3929333257867398,
+      "learning_rate": 5.717110903520617e-08,
+      "loss": 0.0513,
+      "step": 3190
+    },
+    {
+      "epoch": 2.9437269372693726,
+      "grad_norm": 0.3583213372419836,
+      "learning_rate": 5.53712225618952e-08,
+      "loss": 0.0585,
+      "step": 3191
+    },
+    {
+      "epoch": 2.944649446494465,
+      "grad_norm": 0.33733747911177364,
+      "learning_rate": 5.360009206410721e-08,
+      "loss": 0.0576,
+      "step": 3192
+    },
+    {
+      "epoch": 2.945571955719557,
+      "grad_norm": 0.42567376818493746,
+      "learning_rate": 5.1857719583592314e-08,
+      "loss": 0.058,
+      "step": 3193
+    },
+    {
+      "epoch": 2.9464944649446494,
+      "grad_norm": 0.4124616322920309,
+      "learning_rate": 5.014410712893825e-08,
+      "loss": 0.0604,
+      "step": 3194
+    },
+    {
+      "epoch": 2.947416974169742,
+      "grad_norm": 0.39271782211136996,
+      "learning_rate": 4.845925667558981e-08,
+      "loss": 0.0646,
+      "step": 3195
+    },
+    {
+      "epoch": 2.948339483394834,
+      "grad_norm": 0.3985025384944136,
+      "learning_rate": 4.680317016582669e-08,
+      "loss": 0.0631,
+      "step": 3196
+    },
+    {
+      "epoch": 2.949261992619926,
+      "grad_norm": 0.41581860892776334,
+      "learning_rate": 4.517584950877452e-08,
+      "loss": 0.0581,
+      "step": 3197
+    },
+    {
+      "epoch": 2.9501845018450186,
+      "grad_norm": 0.46541482571202747,
+      "learning_rate": 4.357729658039378e-08,
+      "loss": 0.0658,
+      "step": 3198
+    },
+    {
+      "epoch": 2.9511070110701105,
+      "grad_norm": 0.3385333526417377,
+      "learning_rate": 4.2007513223485396e-08,
+      "loss": 0.0504,
+      "step": 3199
+    },
+    {
+      "epoch": 2.952029520295203,
+      "grad_norm": 0.4109038841918147,
+      "learning_rate": 4.0466501247685143e-08,
+      "loss": 0.0505,
+      "step": 3200
+    },
+    {
+      "epoch": 2.9529520295202953,
+      "grad_norm": 0.4007606060640004,
+      "learning_rate": 3.895426242945532e-08,
+      "loss": 0.0585,
+      "step": 3201
+    },
+    {
+      "epoch": 2.9538745387453873,
+      "grad_norm": 0.36225177120950675,
+      "learning_rate": 3.74707985120959e-08,
+      "loss": 0.056,
+      "step": 3202
+    },
+    {
+      "epoch": 2.9547970479704797,
+      "grad_norm": 0.3804162709721264,
+      "learning_rate": 3.601611120573056e-08,
+      "loss": 0.0588,
+      "step": 3203
+    },
+    {
+      "epoch": 2.955719557195572,
+      "grad_norm": 0.41534819889707814,
+      "learning_rate": 3.459020218731512e-08,
+      "loss": 0.0537,
+      "step": 3204
+    },
+    {
+      "epoch": 2.956642066420664,
+      "grad_norm": 0.4311040488800863,
+      "learning_rate": 3.319307310062358e-08,
+      "loss": 0.062,
+      "step": 3205
+    },
+    {
+      "epoch": 2.9575645756457565,
+      "grad_norm": 0.3028959344330278,
+      "learning_rate": 3.182472555625926e-08,
+      "loss": 0.0523,
+      "step": 3206
+    },
+    {
+      "epoch": 2.958487084870849,
+      "grad_norm": 0.4192421837220252,
+      "learning_rate": 3.048516113163813e-08,
+      "loss": 0.0645,
+      "step": 3207
+    },
+    {
+      "epoch": 2.959409594095941,
+      "grad_norm": 0.36431487381541733,
+      "learning_rate": 2.9174381370999927e-08,
+      "loss": 0.059,
+      "step": 3208
+    },
+    {
+      "epoch": 2.9603321033210332,
+      "grad_norm": 0.33647666805310933,
+      "learning_rate": 2.789238778540537e-08,
+      "loss": 0.0544,
+      "step": 3209
+    },
+    {
+      "epoch": 2.961254612546125,
+      "grad_norm": 0.43733262509769244,
+      "learning_rate": 2.663918185272507e-08,
+      "loss": 0.0651,
+      "step": 3210
+    },
+    {
+      "epoch": 2.9621771217712176,
+      "grad_norm": 0.38996048249617143,
+      "learning_rate": 2.5414765017642284e-08,
+      "loss": 0.0586,
+      "step": 3211
+    },
+    {
+      "epoch": 2.96309963099631,
+      "grad_norm": 0.3769281522187525,
+      "learning_rate": 2.4219138691658482e-08,
+      "loss": 0.0603,
+      "step": 3212
+    },
+    {
+      "epoch": 2.9640221402214024,
+      "grad_norm": 0.3847947330110199,
+      "learning_rate": 2.3052304253082246e-08,
+      "loss": 0.0582,
+      "step": 3213
+    },
+    {
+      "epoch": 2.9649446494464944,
+      "grad_norm": 0.3997571039214077,
+      "learning_rate": 2.191426304702926e-08,
+      "loss": 0.0695,
+      "step": 3214
+    },
+    {
+      "epoch": 2.965867158671587,
+      "grad_norm": 0.3528516194523642,
+      "learning_rate": 2.0805016385427865e-08,
+      "loss": 0.0528,
+      "step": 3215
+    },
+    {
+      "epoch": 2.9667896678966788,
+      "grad_norm": 0.36457562727966364,
+      "learning_rate": 1.9724565547007968e-08,
+      "loss": 0.0588,
+      "step": 3216
+    },
+    {
+      "epoch": 2.967712177121771,
+      "grad_norm": 0.401797472107959,
+      "learning_rate": 1.8672911777301015e-08,
+      "loss": 0.0589,
+      "step": 3217
+    },
+    {
+      "epoch": 2.9686346863468636,
+      "grad_norm": 0.4502992009118647,
+      "learning_rate": 1.7650056288651127e-08,
+      "loss": 0.0688,
+      "step": 3218
+    },
+    {
+      "epoch": 2.969557195571956,
+      "grad_norm": 0.3639762154572545,
+      "learning_rate": 1.6656000260195648e-08,
+      "loss": 0.059,
+      "step": 3219
+    },
+    {
+      "epoch": 2.970479704797048,
+      "grad_norm": 0.40979035018838866,
+      "learning_rate": 1.5690744837873473e-08,
+      "loss": 0.0635,
+      "step": 3220
+    },
+    {
+      "epoch": 2.9714022140221403,
+      "grad_norm": 0.38266671196509183,
+      "learning_rate": 1.4754291134425058e-08,
+      "loss": 0.0555,
+      "step": 3221
+    },
+    {
+      "epoch": 2.9723247232472323,
+      "grad_norm": 0.3802321127786767,
+      "learning_rate": 1.3846640229386864e-08,
+      "loss": 0.061,
+      "step": 3222
+    },
+    {
+      "epoch": 2.9732472324723247,
+      "grad_norm": 0.381698112696052,
+      "learning_rate": 1.2967793169091358e-08,
+      "loss": 0.0511,
+      "step": 3223
+    },
+    {
+      "epoch": 2.974169741697417,
+      "grad_norm": 0.3169210239387918,
+      "learning_rate": 1.2117750966667008e-08,
+      "loss": 0.0527,
+      "step": 3224
+    },
+    {
+      "epoch": 2.975092250922509,
+      "grad_norm": 0.36279637299517065,
+      "learning_rate": 1.1296514602038288e-08,
+      "loss": 0.0481,
+      "step": 3225
+    },
+    {
+      "epoch": 2.9760147601476015,
+      "grad_norm": 0.4224857235132395,
+      "learning_rate": 1.0504085021914579e-08,
+      "loss": 0.0628,
+      "step": 3226
+    },
+    {
+      "epoch": 2.976937269372694,
+      "grad_norm": 0.4217813645086133,
+      "learning_rate": 9.74046313980681e-09,
+      "loss": 0.0589,
+      "step": 3227
+    },
+    {
+      "epoch": 2.977859778597786,
+      "grad_norm": 0.33581347684941426,
+      "learning_rate": 9.005649836013597e-09,
+      "loss": 0.0493,
+      "step": 3228
+    },
+    {
+      "epoch": 2.9787822878228782,
+      "grad_norm": 0.38548294586419124,
+      "learning_rate": 8.299645957615675e-09,
+      "loss": 0.0622,
+      "step": 3229
+    },
+    {
+      "epoch": 2.9797047970479706,
+      "grad_norm": 0.41781985113979636,
+      "learning_rate": 7.622452318495344e-09,
+      "loss": 0.0594,
+      "step": 3230
+    },
+    {
+      "epoch": 2.9806273062730626,
+      "grad_norm": 0.4069508964434947,
+      "learning_rate": 6.974069699314245e-09,
+      "loss": 0.0633,
+      "step": 3231
+    },
+    {
+      "epoch": 2.981549815498155,
+      "grad_norm": 0.3580843212211675,
+      "learning_rate": 6.354498847521706e-09,
+      "loss": 0.0512,
+      "step": 3232
+    },
+    {
+      "epoch": 2.982472324723247,
+      "grad_norm": 0.417975915367572,
+      "learning_rate": 5.763740477357504e-09,
+      "loss": 0.0679,
+      "step": 3233
+    },
+    {
+      "epoch": 2.9833948339483394,
+      "grad_norm": 0.40074037578986116,
+      "learning_rate": 5.201795269837995e-09,
+      "loss": 0.0614,
+      "step": 3234
+    },
+    {
+      "epoch": 2.984317343173432,
+      "grad_norm": 0.4142629840074768,
+      "learning_rate": 4.66866387277276e-09,
+      "loss": 0.0603,
+      "step": 3235
+    },
+    {
+      "epoch": 2.985239852398524,
+      "grad_norm": 0.35881008152437754,
+      "learning_rate": 4.164346900750737e-09,
+      "loss": 0.044,
+      "step": 3236
+    },
+    {
+      "epoch": 2.986162361623616,
+      "grad_norm": 0.3612726682414584,
+      "learning_rate": 3.6888449351457633e-09,
+      "loss": 0.0552,
+      "step": 3237
+    },
+    {
+      "epoch": 2.9870848708487086,
+      "grad_norm": 0.35277935338398936,
+      "learning_rate": 3.2421585241110276e-09,
+      "loss": 0.0547,
+      "step": 3238
+    },
+    {
+      "epoch": 2.9880073800738005,
+      "grad_norm": 0.3235211391456132,
+      "learning_rate": 2.8242881825846223e-09,
+      "loss": 0.0474,
+      "step": 3239
+    },
+    {
+      "epoch": 2.988929889298893,
+      "grad_norm": 0.41428593730616464,
+      "learning_rate": 2.4352343922839917e-09,
+      "loss": 0.0629,
+      "step": 3240
+    },
+    {
+      "epoch": 2.9898523985239853,
+      "grad_norm": 0.38148708756384203,
+      "learning_rate": 2.07499760170593e-09,
+      "loss": 0.0576,
+      "step": 3241
+    },
+    {
+      "epoch": 2.9907749077490777,
+      "grad_norm": 0.36477370715888274,
+      "learning_rate": 1.743578226129361e-09,
+      "loss": 0.0548,
+      "step": 3242
+    },
+    {
+      "epoch": 2.9916974169741697,
+      "grad_norm": 0.3853039173542559,
+      "learning_rate": 1.4409766476125575e-09,
+      "loss": 0.0567,
+      "step": 3243
+    },
+    {
+      "epoch": 2.992619926199262,
+      "grad_norm": 0.47482229554396677,
+      "learning_rate": 1.1671932149931453e-09,
+      "loss": 0.074,
+      "step": 3244
+    },
+    {
+      "epoch": 2.993542435424354,
+      "grad_norm": 0.40637755989313035,
+      "learning_rate": 9.222282438853258e-10,
+      "loss": 0.0545,
+      "step": 3245
+    },
+    {
+      "epoch": 2.9944649446494465,
+      "grad_norm": 0.379512699850719,
+      "learning_rate": 7.06082016682652e-10,
+      "loss": 0.0629,
+      "step": 3246
+    },
+    {
+      "epoch": 2.995387453874539,
+      "grad_norm": 0.3485901343457483,
+      "learning_rate": 5.187547825580285e-10,
+      "loss": 0.0502,
+      "step": 3247
+    },
+    {
+      "epoch": 2.9963099630996313,
+      "grad_norm": 0.3860947708929262,
+      "learning_rate": 3.602467574581603e-10,
+      "loss": 0.0556,
+      "step": 3248
+    },
+    {
+      "epoch": 2.9972324723247232,
+      "grad_norm": 0.38267347060083073,
+      "learning_rate": 2.3055812411187926e-10,
+      "loss": 0.0586,
+      "step": 3249
+    },
+    {
+      "epoch": 2.9981549815498156,
+      "grad_norm": 0.36726464087860233,
+      "learning_rate": 1.2968903202459358e-10,
+      "loss": 0.0632,
+      "step": 3250
+    },
+    {
+      "epoch": 2.9990774907749076,
+      "grad_norm": 0.3714096301031751,
+      "learning_rate": 5.763959747551173e-11,
+      "loss": 0.0502,
+      "step": 3251
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2703555109604041,
+      "learning_rate": 1.4409903520418155e-11,
+      "loss": 0.0307,
+      "step": 3252
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3252,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3855346449776640.0,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}