diff --git "a/e1.0/trainer_state.json" "b/e1.0/trainer_state.json"
deleted file mode 100644--- "a/e1.0/trainer_state.json"
+++ /dev/null
@@ -1,11233 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.005163240866168,
-  "eval_steps": 500,
-  "global_step": 3200,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.000628227025541355,
-      "grad_norm": 37.5,
-      "learning_rate": 2.5000000000000004e-07,
-      "loss": 1.9002,
-      "step": 2
-    },
-    {
-      "epoch": 0.00125645405108271,
-      "grad_norm": 5.65625,
-      "learning_rate": 5.000000000000001e-07,
-      "loss": 1.6234,
-      "step": 4
-    },
-    {
-      "epoch": 0.001884681076624065,
-      "grad_norm": 5.15625,
-      "learning_rate": 7.5e-07,
-      "loss": 1.6223,
-      "step": 6
-    },
-    {
-      "epoch": 0.00251290810216542,
-      "grad_norm": 4.8125,
-      "learning_rate": 1.0000000000000002e-06,
-      "loss": 1.6607,
-      "step": 8
-    },
-    {
-      "epoch": 0.003141135127706775,
-      "grad_norm": 5.03125,
-      "learning_rate": 1.25e-06,
-      "loss": 1.6644,
-      "step": 10
-    },
-    {
-      "epoch": 0.00376936215324813,
-      "grad_norm": 5.71875,
-      "learning_rate": 1.5e-06,
-      "loss": 1.8307,
-      "step": 12
-    },
-    {
-      "epoch": 0.004397589178789485,
-      "grad_norm": 5.28125,
-      "learning_rate": 1.75e-06,
-      "loss": 1.623,
-      "step": 14
-    },
-    {
-      "epoch": 0.00502581620433084,
-      "grad_norm": 3.625,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 1.6915,
-      "step": 16
-    },
-    {
-      "epoch": 0.005654043229872195,
-      "grad_norm": 3.609375,
-      "learning_rate": 2.25e-06,
-      "loss": 1.8222,
-      "step": 18
-    },
-    {
-      "epoch": 0.00628227025541355,
-      "grad_norm": 3.21875,
-      "learning_rate": 2.5e-06,
-      "loss": 1.6702,
-      "step": 20
-    },
-    {
-      "epoch": 0.006910497280954905,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.7500000000000004e-06,
-      "loss": 1.6118,
-      "step": 22
-    },
-    {
-      "epoch": 0.00753872430649626,
-      "grad_norm": 2.890625,
-      "learning_rate": 3e-06,
-      "loss": 1.6499,
-      "step": 24
-    },
-    {
-      "epoch": 0.008166951332037615,
-      "grad_norm": 2.421875,
-      "learning_rate": 3.2500000000000002e-06,
-      "loss": 1.4785,
-      "step": 26
-    },
-    {
-      "epoch": 0.00879517835757897,
-      "grad_norm": 2.421875,
-      "learning_rate": 3.5e-06,
-      "loss": 1.7235,
-      "step": 28
-    },
-    {
-      "epoch": 0.009423405383120325,
-      "grad_norm": 2.25,
-      "learning_rate": 3.7500000000000005e-06,
-      "loss": 1.6245,
-      "step": 30
-    },
-    {
-      "epoch": 0.01005163240866168,
-      "grad_norm": 1.6640625,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 1.5486,
-      "step": 32
-    },
-    {
-      "epoch": 0.010679859434203035,
-      "grad_norm": 2.09375,
-      "learning_rate": 4.25e-06,
-      "loss": 1.4448,
-      "step": 34
-    },
-    {
-      "epoch": 0.01130808645974439,
-      "grad_norm": 1.5078125,
-      "learning_rate": 4.5e-06,
-      "loss": 1.48,
-      "step": 36
-    },
-    {
-      "epoch": 0.011936313485285744,
-      "grad_norm": 1.6796875,
-      "learning_rate": 4.75e-06,
-      "loss": 1.5128,
-      "step": 38
-    },
-    {
-      "epoch": 0.0125645405108271,
-      "grad_norm": 1.5234375,
-      "learning_rate": 5e-06,
-      "loss": 1.4774,
-      "step": 40
-    },
-    {
-      "epoch": 0.013192767536368456,
-      "grad_norm": 1.546875,
-      "learning_rate": 5.2500000000000006e-06,
-      "loss": 1.3692,
-      "step": 42
-    },
-    {
-      "epoch": 0.01382099456190981,
-      "grad_norm": 1.34375,
-      "learning_rate": 5.500000000000001e-06,
-      "loss": 1.5056,
-      "step": 44
-    },
-    {
-      "epoch": 0.014449221587451166,
-      "grad_norm": 1.21875,
-      "learning_rate": 5.75e-06,
-      "loss": 1.4744,
-      "step": 46
-    },
-    {
-      "epoch": 0.01507744861299252,
-      "grad_norm": 1.0078125,
-      "learning_rate": 6e-06,
-      "loss": 1.5776,
-      "step": 48
-    },
-    {
-      "epoch": 0.015705675638533874,
-      "grad_norm": 1.1015625,
-      "learning_rate": 6.25e-06,
-      "loss": 1.486,
-      "step": 50
-    },
-    {
-      "epoch": 0.01633390266407523,
-      "grad_norm": 0.85546875,
-      "learning_rate": 6.5000000000000004e-06,
-      "loss": 1.5541,
-      "step": 52
-    },
-    {
-      "epoch": 0.016962129689616585,
-      "grad_norm": 0.984375,
-      "learning_rate": 6.750000000000001e-06,
-      "loss": 1.3999,
-      "step": 54
-    },
-    {
-      "epoch": 0.01759035671515794,
-      "grad_norm": 0.96875,
-      "learning_rate": 7e-06,
-      "loss": 1.4455,
-      "step": 56
-    },
-    {
-      "epoch": 0.018218583740699294,
-      "grad_norm": 1.0546875,
-      "learning_rate": 7.25e-06,
-      "loss": 1.5221,
-      "step": 58
-    },
-    {
-      "epoch": 0.01884681076624065,
-      "grad_norm": 0.88671875,
-      "learning_rate": 7.500000000000001e-06,
-      "loss": 1.4798,
-      "step": 60
-    },
-    {
-      "epoch": 0.019475037791782005,
-      "grad_norm": 0.9140625,
-      "learning_rate": 7.75e-06,
-      "loss": 1.4334,
-      "step": 62
-    },
-    {
-      "epoch": 0.02010326481732336,
-      "grad_norm": 0.98046875,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 1.3293,
-      "step": 64
-    },
-    {
-      "epoch": 0.020731491842864717,
-      "grad_norm": 0.83984375,
-      "learning_rate": 8.25e-06,
-      "loss": 1.4174,
-      "step": 66
-    },
-    {
-      "epoch": 0.02135971886840607,
-      "grad_norm": 0.859375,
-      "learning_rate": 8.5e-06,
-      "loss": 1.4177,
-      "step": 68
-    },
-    {
-      "epoch": 0.021987945893947425,
-      "grad_norm": 0.84765625,
-      "learning_rate": 8.750000000000001e-06,
-      "loss": 1.3708,
-      "step": 70
-    },
-    {
-      "epoch": 0.02261617291948878,
-      "grad_norm": 0.90234375,
-      "learning_rate": 9e-06,
-      "loss": 1.4062,
-      "step": 72
-    },
-    {
-      "epoch": 0.023244399945030136,
-      "grad_norm": 0.83984375,
-      "learning_rate": 9.250000000000001e-06,
-      "loss": 1.3829,
-      "step": 74
-    },
-    {
-      "epoch": 0.02387262697057149,
-      "grad_norm": 0.9375,
-      "learning_rate": 9.5e-06,
-      "loss": 1.3441,
-      "step": 76
-    },
-    {
-      "epoch": 0.024500853996112845,
-      "grad_norm": 0.90234375,
-      "learning_rate": 9.75e-06,
-      "loss": 1.5625,
-      "step": 78
-    },
-    {
-      "epoch": 0.0251290810216542,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1e-05,
-      "loss": 1.4504,
-      "step": 80
-    },
-    {
-      "epoch": 0.025757308047195556,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.025e-05,
-      "loss": 1.4672,
-      "step": 82
-    },
-    {
-      "epoch": 0.026385535072736912,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.0500000000000001e-05,
-      "loss": 1.4405,
-      "step": 84
-    },
-    {
-      "epoch": 0.027013762098278264,
-      "grad_norm": 1.0546875,
-      "learning_rate": 1.075e-05,
-      "loss": 1.3557,
-      "step": 86
-    },
-    {
-      "epoch": 0.02764198912381962,
-      "grad_norm": 0.90234375,
-      "learning_rate": 1.1000000000000001e-05,
-      "loss": 1.4309,
-      "step": 88
-    },
-    {
-      "epoch": 0.028270216149360976,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.125e-05,
-      "loss": 1.3528,
-      "step": 90
-    },
-    {
-      "epoch": 0.02889844317490233,
-      "grad_norm": 0.96875,
-      "learning_rate": 1.15e-05,
-      "loss": 1.4093,
-      "step": 92
-    },
-    {
-      "epoch": 0.029526670200443684,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.1750000000000001e-05,
-      "loss": 1.4324,
-      "step": 94
-    },
-    {
-      "epoch": 0.03015489722598504,
-      "grad_norm": 0.875,
-      "learning_rate": 1.2e-05,
-      "loss": 1.4622,
-      "step": 96
-    },
-    {
-      "epoch": 0.030783124251526395,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.2250000000000001e-05,
-      "loss": 1.5166,
-      "step": 98
-    },
-    {
-      "epoch": 0.03141135127706775,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.25e-05,
-      "loss": 1.4729,
-      "step": 100
-    },
-    {
-      "epoch": 0.032039578302609104,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.275e-05,
-      "loss": 1.4201,
-      "step": 102
-    },
-    {
-      "epoch": 0.03266780532815046,
-      "grad_norm": 0.97265625,
-      "learning_rate": 1.3000000000000001e-05,
-      "loss": 1.3646,
-      "step": 104
-    },
-    {
-      "epoch": 0.033296032353691815,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.325e-05,
-      "loss": 1.3105,
-      "step": 106
-    },
-    {
-      "epoch": 0.03392425937923317,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.3500000000000001e-05,
-      "loss": 1.5302,
-      "step": 108
-    },
-    {
-      "epoch": 0.03455248640477453,
-      "grad_norm": 0.875,
-      "learning_rate": 1.375e-05,
-      "loss": 1.3979,
-      "step": 110
-    },
-    {
-      "epoch": 0.03518071343031588,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.4e-05,
-      "loss": 1.3961,
-      "step": 112
-    },
-    {
-      "epoch": 0.03580894045585724,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.425e-05,
-      "loss": 1.3645,
-      "step": 114
-    },
-    {
-      "epoch": 0.03643716748139859,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.45e-05,
-      "loss": 1.306,
-      "step": 116
-    },
-    {
-      "epoch": 0.03706539450693994,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.4750000000000003e-05,
-      "loss": 1.3799,
-      "step": 118
-    },
-    {
-      "epoch": 0.0376936215324813,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.5000000000000002e-05,
-      "loss": 1.3281,
-      "step": 120
-    },
-    {
-      "epoch": 0.038321848558022654,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.525e-05,
-      "loss": 1.4052,
-      "step": 122
-    },
-    {
-      "epoch": 0.03895007558356401,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.55e-05,
-      "loss": 1.4946,
-      "step": 124
-    },
-    {
-      "epoch": 0.039578302609105366,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.575e-05,
-      "loss": 1.4292,
-      "step": 126
-    },
-    {
-      "epoch": 0.04020652963464672,
-      "grad_norm": 1.0078125,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.4858,
-      "step": 128
-    },
-    {
-      "epoch": 0.04083475666018808,
-      "grad_norm": 0.9765625,
-      "learning_rate": 1.6250000000000002e-05,
-      "loss": 1.2745,
-      "step": 130
-    },
-    {
-      "epoch": 0.04146298368572943,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.65e-05,
-      "loss": 1.4684,
-      "step": 132
-    },
-    {
-      "epoch": 0.04209121071127078,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.675e-05,
-      "loss": 1.4275,
-      "step": 134
-    },
-    {
-      "epoch": 0.04271943773681214,
-      "grad_norm": 0.90625,
-      "learning_rate": 1.7e-05,
-      "loss": 1.2831,
-      "step": 136
-    },
-    {
-      "epoch": 0.043347664762353494,
-      "grad_norm": 0.953125,
-      "learning_rate": 1.7250000000000003e-05,
-      "loss": 1.445,
-      "step": 138
-    },
-    {
-      "epoch": 0.04397589178789485,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.7500000000000002e-05,
-      "loss": 1.3457,
-      "step": 140
-    },
-    {
-      "epoch": 0.044604118813436205,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.775e-05,
-      "loss": 1.3961,
-      "step": 142
-    },
-    {
-      "epoch": 0.04523234583897756,
-      "grad_norm": 0.75,
-      "learning_rate": 1.8e-05,
-      "loss": 1.2985,
-      "step": 144
-    },
-    {
-      "epoch": 0.04586057286451892,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.825e-05,
-      "loss": 1.3075,
-      "step": 146
-    },
-    {
-      "epoch": 0.04648879989006027,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.8500000000000002e-05,
-      "loss": 1.3602,
-      "step": 148
-    },
-    {
-      "epoch": 0.04711702691560163,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.8750000000000002e-05,
-      "loss": 1.4481,
-      "step": 150
-    },
-    {
-      "epoch": 0.04774525394114298,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.9e-05,
-      "loss": 1.409,
-      "step": 152
-    },
-    {
-      "epoch": 0.04837348096668433,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.925e-05,
-      "loss": 1.357,
-      "step": 154
-    },
-    {
-      "epoch": 0.04900170799222569,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.95e-05,
-      "loss": 1.2841,
-      "step": 156
-    },
-    {
-      "epoch": 0.049629935017767045,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.9750000000000002e-05,
-      "loss": 1.4336,
-      "step": 158
-    },
-    {
-      "epoch": 0.0502581620433084,
-      "grad_norm": 0.9609375,
-      "learning_rate": 2e-05,
-      "loss": 1.3853,
-      "step": 160
-    },
-    {
-      "epoch": 0.050886389068849756,
-      "grad_norm": 0.89453125,
-      "learning_rate": 1.9997461123452876e-05,
-      "loss": 1.3465,
-      "step": 162
-    },
-    {
-      "epoch": 0.05151461609439111,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.9994922246905744e-05,
-      "loss": 1.3669,
-      "step": 164
-    },
-    {
-      "epoch": 0.05214284311993247,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.999238337035862e-05,
-      "loss": 1.6046,
-      "step": 166
-    },
-    {
-      "epoch": 0.052771070145473824,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.998984449381149e-05,
-      "loss": 1.2993,
-      "step": 168
-    },
-    {
-      "epoch": 0.05339929717101517,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.9987305617264362e-05,
-      "loss": 1.4495,
-      "step": 170
-    },
-    {
-      "epoch": 0.05402752419655653,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.9984766740717233e-05,
-      "loss": 1.3141,
-      "step": 172
-    },
-    {
-      "epoch": 0.054655751222097884,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.9982227864170108e-05,
-      "loss": 1.4852,
-      "step": 174
-    },
-    {
-      "epoch": 0.05528397824763924,
-      "grad_norm": 0.875,
-      "learning_rate": 1.997968898762298e-05,
-      "loss": 1.4228,
-      "step": 176
-    },
-    {
-      "epoch": 0.055912205273180596,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.997715011107585e-05,
-      "loss": 1.5617,
-      "step": 178
-    },
-    {
-      "epoch": 0.05654043229872195,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.9974611234528722e-05,
-      "loss": 1.2938,
-      "step": 180
-    },
-    {
-      "epoch": 0.05716865932426331,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.9972072357981597e-05,
-      "loss": 1.469,
-      "step": 182
-    },
-    {
-      "epoch": 0.05779688634980466,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.9969533481434465e-05,
-      "loss": 1.41,
-      "step": 184
-    },
-    {
-      "epoch": 0.05842511337534602,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.996699460488734e-05,
-      "loss": 1.4414,
-      "step": 186
-    },
-    {
-      "epoch": 0.05905334040088737,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.996445572834021e-05,
-      "loss": 1.2717,
-      "step": 188
-    },
-    {
-      "epoch": 0.059681567426428724,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.9961916851793083e-05,
-      "loss": 1.3179,
-      "step": 190
-    },
-    {
-      "epoch": 0.06030979445197008,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.9959377975245954e-05,
-      "loss": 1.4353,
-      "step": 192
-    },
-    {
-      "epoch": 0.060938021477511435,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.995683909869883e-05,
-      "loss": 1.4721,
-      "step": 194
-    },
-    {
-      "epoch": 0.06156624850305279,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.99543002221517e-05,
-      "loss": 1.4394,
-      "step": 196
-    },
-    {
-      "epoch": 0.06219447552859415,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.995176134560457e-05,
-      "loss": 1.4004,
-      "step": 198
-    },
-    {
-      "epoch": 0.0628227025541355,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.9949222469057443e-05,
-      "loss": 1.3159,
-      "step": 200
-    },
-    {
-      "epoch": 0.06345092957967685,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.9946683592510318e-05,
-      "loss": 1.408,
-      "step": 202
-    },
-    {
-      "epoch": 0.06407915660521821,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.994414471596319e-05,
-      "loss": 1.362,
-      "step": 204
-    },
-    {
-      "epoch": 0.06470738363075956,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.994160583941606e-05,
-      "loss": 1.321,
-      "step": 206
-    },
-    {
-      "epoch": 0.06533561065630092,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.9939066962868932e-05,
-      "loss": 1.3576,
-      "step": 208
-    },
-    {
-      "epoch": 0.06596383768184227,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.9936528086321803e-05,
-      "loss": 1.4552,
-      "step": 210
-    },
-    {
-      "epoch": 0.06659206470738363,
-      "grad_norm": 0.875,
-      "learning_rate": 1.9933989209774675e-05,
-      "loss": 1.3144,
-      "step": 212
-    },
-    {
-      "epoch": 0.06722029173292499,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.993145033322755e-05,
-      "loss": 1.3575,
-      "step": 214
-    },
-    {
-      "epoch": 0.06784851875846634,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.992891145668042e-05,
-      "loss": 1.25,
-      "step": 216
-    },
-    {
-      "epoch": 0.0684767457840077,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9926372580133292e-05,
-      "loss": 1.4611,
-      "step": 218
-    },
-    {
-      "epoch": 0.06910497280954905,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.9923833703586163e-05,
-      "loss": 1.2994,
-      "step": 220
-    },
-    {
-      "epoch": 0.06973319983509041,
-      "grad_norm": 0.9375,
-      "learning_rate": 1.9921294827039038e-05,
-      "loss": 1.2697,
-      "step": 222
-    },
-    {
-      "epoch": 0.07036142686063176,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.991875595049191e-05,
-      "loss": 1.4699,
-      "step": 224
-    },
-    {
-      "epoch": 0.07098965388617312,
-      "grad_norm": 0.92578125,
-      "learning_rate": 1.991621707394478e-05,
-      "loss": 1.4276,
-      "step": 226
-    },
-    {
-      "epoch": 0.07161788091171448,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.9913678197397652e-05,
-      "loss": 1.4052,
-      "step": 228
-    },
-    {
-      "epoch": 0.07224610793725583,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.9911139320850527e-05,
-      "loss": 1.3744,
-      "step": 230
-    },
-    {
-      "epoch": 0.07287433496279717,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.9908600444303395e-05,
-      "loss": 1.363,
-      "step": 232
-    },
-    {
-      "epoch": 0.07350256198833853,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.990606156775627e-05,
-      "loss": 1.421,
-      "step": 234
-    },
-    {
-      "epoch": 0.07413078901387989,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.990352269120914e-05,
-      "loss": 1.5131,
-      "step": 236
-    },
-    {
-      "epoch": 0.07475901603942124,
-      "grad_norm": 1.5625,
-      "learning_rate": 1.9900983814662013e-05,
-      "loss": 1.3155,
-      "step": 238
-    },
-    {
-      "epoch": 0.0753872430649626,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.9898444938114884e-05,
-      "loss": 1.3595,
-      "step": 240
-    },
-    {
-      "epoch": 0.07601547009050395,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.989590606156776e-05,
-      "loss": 1.3706,
-      "step": 242
-    },
-    {
-      "epoch": 0.07664369711604531,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.989336718502063e-05,
-      "loss": 1.3058,
-      "step": 244
-    },
-    {
-      "epoch": 0.07727192414158666,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.98908283084735e-05,
-      "loss": 1.3404,
-      "step": 246
-    },
-    {
-      "epoch": 0.07790015116712802,
-      "grad_norm": 0.9453125,
-      "learning_rate": 1.9888289431926376e-05,
-      "loss": 1.235,
-      "step": 248
-    },
-    {
-      "epoch": 0.07852837819266938,
-      "grad_norm": 0.84765625,
-      "learning_rate": 1.9885750555379248e-05,
-      "loss": 1.3668,
-      "step": 250
-    },
-    {
-      "epoch": 0.07915660521821073,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.988321167883212e-05,
-      "loss": 1.3602,
-      "step": 252
-    },
-    {
-      "epoch": 0.07978483224375209,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.988067280228499e-05,
-      "loss": 1.3833,
-      "step": 254
-    },
-    {
-      "epoch": 0.08041305926929344,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9878133925737865e-05,
-      "loss": 1.4476,
-      "step": 256
-    },
-    {
-      "epoch": 0.0810412862948348,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.9875595049190733e-05,
-      "loss": 1.4111,
-      "step": 258
-    },
-    {
-      "epoch": 0.08166951332037616,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.9873056172643608e-05,
-      "loss": 1.3636,
-      "step": 260
-    },
-    {
-      "epoch": 0.08229774034591751,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.987051729609648e-05,
-      "loss": 1.2524,
-      "step": 262
-    },
-    {
-      "epoch": 0.08292596737145887,
-      "grad_norm": 0.9609375,
-      "learning_rate": 1.986797841954935e-05,
-      "loss": 1.4048,
-      "step": 264
-    },
-    {
-      "epoch": 0.08355419439700022,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.9865439543002222e-05,
-      "loss": 1.3619,
-      "step": 266
-    },
-    {
-      "epoch": 0.08418242142254156,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9862900666455097e-05,
-      "loss": 1.4125,
-      "step": 268
-    },
-    {
-      "epoch": 0.08481064844808292,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.986036178990797e-05,
-      "loss": 1.4536,
-      "step": 270
-    },
-    {
-      "epoch": 0.08543887547362428,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.985782291336084e-05,
-      "loss": 1.392,
-      "step": 272
-    },
-    {
-      "epoch": 0.08606710249916563,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.985528403681371e-05,
-      "loss": 1.393,
-      "step": 274
-    },
-    {
-      "epoch": 0.08669532952470699,
-      "grad_norm": 0.91015625,
-      "learning_rate": 1.9852745160266586e-05,
-      "loss": 1.3635,
-      "step": 276
-    },
-    {
-      "epoch": 0.08732355655024834,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9850206283719454e-05,
-      "loss": 1.4626,
-      "step": 278
-    },
-    {
-      "epoch": 0.0879517835757897,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.984766740717233e-05,
-      "loss": 1.3507,
-      "step": 280
-    },
-    {
-      "epoch": 0.08858001060133106,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.98451285306252e-05,
-      "loss": 1.4432,
-      "step": 282
-    },
-    {
-      "epoch": 0.08920823762687241,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.984258965407807e-05,
-      "loss": 1.3932,
-      "step": 284
-    },
-    {
-      "epoch": 0.08983646465241377,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.9840050777530943e-05,
-      "loss": 1.391,
-      "step": 286
-    },
-    {
-      "epoch": 0.09046469167795512,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.9837511900983818e-05,
-      "loss": 1.4163,
-      "step": 288
-    },
-    {
-      "epoch": 0.09109291870349648,
-      "grad_norm": 1.171875,
-      "learning_rate": 1.983497302443669e-05,
-      "loss": 1.4135,
-      "step": 290
-    },
-    {
-      "epoch": 0.09172114572903783,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.983243414788956e-05,
-      "loss": 1.4099,
-      "step": 292
-    },
-    {
-      "epoch": 0.09234937275457919,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.982989527134243e-05,
-      "loss": 1.2512,
-      "step": 294
-    },
-    {
-      "epoch": 0.09297759978012055,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.9827356394795306e-05,
-      "loss": 1.255,
-      "step": 296
-    },
-    {
-      "epoch": 0.0936058268056619,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.9824817518248174e-05,
-      "loss": 1.2295,
-      "step": 298
-    },
-    {
-      "epoch": 0.09423405383120326,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.982227864170105e-05,
-      "loss": 1.4514,
-      "step": 300
-    },
-    {
-      "epoch": 0.09486228085674461,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.981973976515392e-05,
-      "loss": 1.3137,
-      "step": 302
-    },
-    {
-      "epoch": 0.09549050788228595,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.9817200888606792e-05,
-      "loss": 1.3511,
-      "step": 304
-    },
-    {
-      "epoch": 0.09611873490782731,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.9814662012059663e-05,
-      "loss": 1.3035,
-      "step": 306
-    },
-    {
-      "epoch": 0.09674696193336867,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.9812123135512538e-05,
-      "loss": 1.4151,
-      "step": 308
-    },
-    {
-      "epoch": 0.09737518895891002,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.980958425896541e-05,
-      "loss": 1.3819,
-      "step": 310
-    },
-    {
-      "epoch": 0.09800341598445138,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.980704538241828e-05,
-      "loss": 1.3212,
-      "step": 312
-    },
-    {
-      "epoch": 0.09863164300999273,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.9804506505871152e-05,
-      "loss": 1.4313,
-      "step": 314
-    },
-    {
-      "epoch": 0.09925987003553409,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.9801967629324027e-05,
-      "loss": 1.4021,
-      "step": 316
-    },
-    {
-      "epoch": 0.09988809706107545,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.97994287527769e-05,
-      "loss": 1.3191,
-      "step": 318
-    },
-    {
-      "epoch": 0.1005163240866168,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.979688987622977e-05,
-      "loss": 1.3138,
-      "step": 320
-    },
-    {
-      "epoch": 0.10114455111215816,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.979435099968264e-05,
-      "loss": 1.3881,
-      "step": 322
-    },
-    {
-      "epoch": 0.10177277813769951,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.9791812123135513e-05,
-      "loss": 1.1786,
-      "step": 324
-    },
-    {
-      "epoch": 0.10240100516324087,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.9789273246588384e-05,
-      "loss": 1.246,
-      "step": 326
-    },
-    {
-      "epoch": 0.10302923218878222,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.978673437004126e-05,
-      "loss": 1.359,
-      "step": 328
-    },
-    {
-      "epoch": 0.10365745921432358,
-      "grad_norm": 0.91796875,
-      "learning_rate": 1.978419549349413e-05,
-      "loss": 1.271,
-      "step": 330
-    },
-    {
-      "epoch": 0.10428568623986494,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9781656616947e-05,
-      "loss": 1.4137,
-      "step": 332
-    },
-    {
-      "epoch": 0.10491391326540629,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.9779117740399876e-05,
-      "loss": 1.3471,
-      "step": 334
-    },
-    {
-      "epoch": 0.10554214029094765,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.9776578863852748e-05,
-      "loss": 1.4421,
-      "step": 336
-    },
-    {
-      "epoch": 0.10617036731648899,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.977403998730562e-05,
-      "loss": 1.2823,
-      "step": 338
-    },
-    {
-      "epoch": 0.10679859434203035,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.977150111075849e-05,
-      "loss": 1.463,
-      "step": 340
-    },
-    {
-      "epoch": 0.1074268213675717,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.9768962234211365e-05,
-      "loss": 1.2987,
-      "step": 342
-    },
-    {
-      "epoch": 0.10805504839311306,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.9766423357664237e-05,
-      "loss": 1.4113,
-      "step": 344
-    },
-    {
-      "epoch": 0.10868327541865441,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.9763884481117108e-05,
-      "loss": 1.4153,
-      "step": 346
-    },
-    {
-      "epoch": 0.10931150244419577,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.976134560456998e-05,
-      "loss": 1.3976,
-      "step": 348
-    },
-    {
-      "epoch": 0.10993972946973712,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.975880672802285e-05,
-      "loss": 1.395,
-      "step": 350
-    },
-    {
-      "epoch": 0.11056795649527848,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.9756267851475722e-05,
-      "loss": 1.533,
-      "step": 352
-    },
-    {
-      "epoch": 0.11119618352081984,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9753728974928597e-05,
-      "loss": 1.3265,
-      "step": 354
-    },
-    {
-      "epoch": 0.11182441054636119,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.9751190098381468e-05,
-      "loss": 1.4088,
-      "step": 356
-    },
-    {
-      "epoch": 0.11245263757190255,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.974865122183434e-05,
-      "loss": 1.4432,
-      "step": 358
-    },
-    {
-      "epoch": 0.1130808645974439,
-      "grad_norm": 0.984375,
-      "learning_rate": 1.974611234528721e-05,
-      "loss": 1.2292,
-      "step": 360
-    },
-    {
-      "epoch": 0.11370909162298526,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.9743573468740086e-05,
-      "loss": 1.3708,
-      "step": 362
-    },
-    {
-      "epoch": 0.11433731864852661,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.9741034592192957e-05,
-      "loss": 1.2918,
-      "step": 364
-    },
-    {
-      "epoch": 0.11496554567406797,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.973849571564583e-05,
-      "loss": 1.4335,
-      "step": 366
-    },
-    {
-      "epoch": 0.11559377269960933,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.97359568390987e-05,
-      "loss": 1.2187,
-      "step": 368
-    },
-    {
-      "epoch": 0.11622199972515068,
-      "grad_norm": 10.625,
-      "learning_rate": 1.9733417962551575e-05,
-      "loss": 1.2494,
-      "step": 370
-    },
-    {
-      "epoch": 0.11685022675069204,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9730879086004443e-05,
-      "loss": 1.2348,
-      "step": 372
-    },
-    {
-      "epoch": 0.11747845377623338,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.9728340209457317e-05,
-      "loss": 1.3933,
-      "step": 374
-    },
-    {
-      "epoch": 0.11810668080177474,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.972580133291019e-05,
-      "loss": 1.445,
-      "step": 376
-    },
-    {
-      "epoch": 0.11873490782731609,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.972326245636306e-05,
-      "loss": 1.3521,
-      "step": 378
-    },
-    {
-      "epoch": 0.11936313485285745,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.972072357981593e-05,
-      "loss": 1.4533,
-      "step": 380
-    },
-    {
-      "epoch": 0.1199913618783988,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.9718184703268806e-05,
-      "loss": 1.2909,
-      "step": 382
-    },
-    {
-      "epoch": 0.12061958890394016,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.9715645826721678e-05,
-      "loss": 1.5502,
-      "step": 384
-    },
-    {
-      "epoch": 0.12124781592948151,
-      "grad_norm": 0.84765625,
-      "learning_rate": 1.971310695017455e-05,
-      "loss": 1.3525,
-      "step": 386
-    },
-    {
-      "epoch": 0.12187604295502287,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.971056807362742e-05,
-      "loss": 1.5028,
-      "step": 388
-    },
-    {
-      "epoch": 0.12250426998056423,
-      "grad_norm": 6.21875,
-      "learning_rate": 1.9708029197080295e-05,
-      "loss": 1.3943,
-      "step": 390
-    },
-    {
-      "epoch": 0.12313249700610558,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.9705490320533163e-05,
-      "loss": 1.5042,
-      "step": 392
-    },
-    {
-      "epoch": 0.12376072403164694,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.9702951443986038e-05,
-      "loss": 1.3527,
-      "step": 394
-    },
-    {
-      "epoch": 0.1243889510571883,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.970041256743891e-05,
-      "loss": 1.5268,
-      "step": 396
-    },
-    {
-      "epoch": 0.12501717808272964,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.969787369089178e-05,
-      "loss": 1.2923,
-      "step": 398
-    },
-    {
-      "epoch": 0.125645405108271,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.9695334814344652e-05,
-      "loss": 1.4865,
-      "step": 400
-    },
-    {
-      "epoch": 0.12627363213381235,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.9692795937797527e-05,
-      "loss": 1.2926,
-      "step": 402
-    },
-    {
-      "epoch": 0.1269018591593537,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.96902570612504e-05,
-      "loss": 1.3763,
-      "step": 404
-    },
-    {
-      "epoch": 0.12753008618489506,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.968771818470327e-05,
-      "loss": 1.4208,
-      "step": 406
-    },
-    {
-      "epoch": 0.12815831321043641,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.968517930815614e-05,
-      "loss": 1.2802,
-      "step": 408
-    },
-    {
-      "epoch": 0.12878654023597777,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.9682640431609016e-05,
-      "loss": 1.3137,
-      "step": 410
-    },
-    {
-      "epoch": 0.12941476726151913,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.9680101555061887e-05,
-      "loss": 1.2313,
-      "step": 412
-    },
-    {
-      "epoch": 0.13004299428706048,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.967756267851476e-05,
-      "loss": 1.3286,
-      "step": 414
-    },
-    {
-      "epoch": 0.13067122131260184,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.967502380196763e-05,
-      "loss": 1.3544,
-      "step": 416
-    },
-    {
-      "epoch": 0.1312994483381432,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.96724849254205e-05,
-      "loss": 1.4726,
-      "step": 418
-    },
-    {
-      "epoch": 0.13192767536368455,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.9669946048873376e-05,
-      "loss": 1.3215,
-      "step": 420
-    },
-    {
-      "epoch": 0.1325559023892259,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.9667407172326248e-05,
-      "loss": 1.5521,
-      "step": 422
-    },
-    {
-      "epoch": 0.13318412941476726,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.966486829577912e-05,
-      "loss": 1.3127,
-      "step": 424
-    },
-    {
-      "epoch": 0.13381235644030862,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.966232941923199e-05,
-      "loss": 1.2696,
-      "step": 426
-    },
-    {
-      "epoch": 0.13444058346584997,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.9659790542684865e-05,
-      "loss": 1.2138,
-      "step": 428
-    },
-    {
-      "epoch": 0.13506881049139133,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.9657251666137736e-05,
-      "loss": 1.4204,
-      "step": 430
-    },
-    {
-      "epoch": 0.13569703751693268,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.9654712789590608e-05,
-      "loss": 1.2865,
-      "step": 432
-    },
-    {
-      "epoch": 0.13632526454247404,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.965217391304348e-05,
-      "loss": 1.2856,
-      "step": 434
-    },
-    {
-      "epoch": 0.1369534915680154,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.9649635036496354e-05,
-      "loss": 1.4284,
-      "step": 436
-    },
-    {
-      "epoch": 0.13758171859355675,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.9647096159949225e-05,
-      "loss": 1.3569,
-      "step": 438
-    },
-    {
-      "epoch": 0.1382099456190981,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.9644557283402097e-05,
-      "loss": 1.3295,
-      "step": 440
-    },
-    {
-      "epoch": 0.13883817264463946,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.9642018406854968e-05,
-      "loss": 1.2948,
-      "step": 442
-    },
-    {
-      "epoch": 0.13946639967018082,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.963947953030784e-05,
-      "loss": 1.4097,
-      "step": 444
-    },
-    {
-      "epoch": 0.14009462669572217,
-      "grad_norm": 1.0703125,
-      "learning_rate": 1.963694065376071e-05,
-      "loss": 1.3674,
-      "step": 446
-    },
-    {
-      "epoch": 0.14072285372126353,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.9634401777213586e-05,
-      "loss": 1.4544,
-      "step": 448
-    },
-    {
-      "epoch": 0.14135108074680489,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.9631862900666457e-05,
-      "loss": 1.3385,
-      "step": 450
-    },
-    {
-      "epoch": 0.14197930777234624,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.962932402411933e-05,
-      "loss": 1.3962,
-      "step": 452
-    },
-    {
-      "epoch": 0.1426075347978876,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.96267851475722e-05,
-      "loss": 1.3257,
-      "step": 454
-    },
-    {
-      "epoch": 0.14323576182342895,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.9624246271025075e-05,
-      "loss": 1.3572,
-      "step": 456
-    },
-    {
-      "epoch": 0.1438639888489703,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.9621707394477946e-05,
-      "loss": 1.5115,
-      "step": 458
-    },
-    {
-      "epoch": 0.14449221587451166,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.9619168517930817e-05,
-      "loss": 1.3532,
-      "step": 460
-    },
-    {
-      "epoch": 0.14512044290005302,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.961662964138369e-05,
-      "loss": 1.3612,
-      "step": 462
-    },
-    {
-      "epoch": 0.14574866992559435,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.9614090764836564e-05,
-      "loss": 1.3881,
-      "step": 464
-    },
-    {
-      "epoch": 0.1463768969511357,
-      "grad_norm": 0.94140625,
-      "learning_rate": 1.961155188828943e-05,
-      "loss": 1.534,
-      "step": 466
-    },
-    {
-      "epoch": 0.14700512397667706,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.9609013011742306e-05,
-      "loss": 1.4607,
-      "step": 468
-    },
-    {
-      "epoch": 0.14763335100221842,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.9606474135195178e-05,
-      "loss": 1.3466,
-      "step": 470
-    },
-    {
-      "epoch": 0.14826157802775977,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.960393525864805e-05,
-      "loss": 1.3187,
-      "step": 472
-    },
-    {
-      "epoch": 0.14888980505330113,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.960139638210092e-05,
-      "loss": 1.36,
-      "step": 474
-    },
-    {
-      "epoch": 0.14951803207884248,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.9598857505553795e-05,
-      "loss": 1.2185,
-      "step": 476
-    },
-    {
-      "epoch": 0.15014625910438384,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.9596318629006667e-05,
-      "loss": 1.4085,
-      "step": 478
-    },
-    {
-      "epoch": 0.1507744861299252,
-      "grad_norm": 0.98046875,
-      "learning_rate": 1.9593779752459538e-05,
-      "loss": 1.3917,
-      "step": 480
-    },
-    {
-      "epoch": 0.15140271315546655,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.959124087591241e-05,
-      "loss": 1.3497,
-      "step": 482
-    },
-    {
-      "epoch": 0.1520309401810079,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.9588701999365284e-05,
-      "loss": 1.3855,
-      "step": 484
-    },
-    {
-      "epoch": 0.15265916720654926,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.9586163122818152e-05,
-      "loss": 1.4071,
-      "step": 486
-    },
-    {
-      "epoch": 0.15328739423209062,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.9583624246271027e-05,
-      "loss": 1.2817,
-      "step": 488
-    },
-    {
-      "epoch": 0.15391562125763197,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9581085369723898e-05,
-      "loss": 1.3758,
-      "step": 490
-    },
-    {
-      "epoch": 0.15454384828317333,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.957854649317677e-05,
-      "loss": 1.4021,
-      "step": 492
-    },
-    {
-      "epoch": 0.15517207530871469,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.957600761662964e-05,
-      "loss": 1.4163,
-      "step": 494
-    },
-    {
-      "epoch": 0.15580030233425604,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.9573468740082516e-05,
-      "loss": 1.3127,
-      "step": 496
-    },
-    {
-      "epoch": 0.1564285293597974,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.9570929863535387e-05,
-      "loss": 1.2817,
-      "step": 498
-    },
-    {
-      "epoch": 0.15705675638533875,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.956839098698826e-05,
-      "loss": 1.3561,
-      "step": 500
-    },
-    {
-      "epoch": 0.1576849834108801,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.956585211044113e-05,
-      "loss": 1.317,
-      "step": 502
-    },
-    {
-      "epoch": 0.15831321043642146,
-      "grad_norm": 1.3671875,
-      "learning_rate": 1.9563313233894005e-05,
-      "loss": 1.4507,
-      "step": 504
-    },
-    {
-      "epoch": 0.15894143746196282,
-      "grad_norm": 1.0703125,
-      "learning_rate": 1.9560774357346876e-05,
-      "loss": 1.3715,
-      "step": 506
-    },
-    {
-      "epoch": 0.15956966448750418,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.9558235480799747e-05,
-      "loss": 1.3502,
-      "step": 508
-    },
-    {
-      "epoch": 0.16019789151304553,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.9555696604252622e-05,
-      "loss": 1.4822,
-      "step": 510
-    },
-    {
-      "epoch": 0.1608261185385869,
-      "grad_norm": 0.75,
-      "learning_rate": 1.955315772770549e-05,
-      "loss": 1.2966,
-      "step": 512
-    },
-    {
-      "epoch": 0.16145434556412824,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.9550618851158365e-05,
-      "loss": 1.2823,
-      "step": 514
-    },
-    {
-      "epoch": 0.1620825725896696,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.9548079974611236e-05,
-      "loss": 1.3542,
-      "step": 516
-    },
-    {
-      "epoch": 0.16271079961521095,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.9545541098064108e-05,
-      "loss": 1.2341,
-      "step": 518
-    },
-    {
-      "epoch": 0.1633390266407523,
-      "grad_norm": 0.90625,
-      "learning_rate": 1.954300222151698e-05,
-      "loss": 1.4864,
-      "step": 520
-    },
-    {
-      "epoch": 0.16396725366629367,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.9540463344969854e-05,
-      "loss": 1.4035,
-      "step": 522
-    },
-    {
-      "epoch": 0.16459548069183502,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.9537924468422725e-05,
-      "loss": 1.3945,
-      "step": 524
-    },
-    {
-      "epoch": 0.16522370771737638,
-      "grad_norm": 0.94921875,
-      "learning_rate": 1.9535385591875597e-05,
-      "loss": 1.3526,
-      "step": 526
-    },
-    {
-      "epoch": 0.16585193474291773,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.9532846715328468e-05,
-      "loss": 1.4006,
-      "step": 528
-    },
-    {
-      "epoch": 0.1664801617684591,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.9530307838781343e-05,
-      "loss": 1.3114,
-      "step": 530
-    },
-    {
-      "epoch": 0.16710838879400045,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.9527768962234214e-05,
-      "loss": 1.4793,
-      "step": 532
-    },
-    {
-      "epoch": 0.16773661581954177,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.9525230085687086e-05,
-      "loss": 1.3226,
-      "step": 534
-    },
-    {
-      "epoch": 0.16836484284508313,
-      "grad_norm": 0.90625,
-      "learning_rate": 1.9522691209139957e-05,
-      "loss": 1.3562,
-      "step": 536
-    },
-    {
-      "epoch": 0.16899306987062448,
-      "grad_norm": 0.91796875,
-      "learning_rate": 1.952015233259283e-05,
-      "loss": 1.3007,
-      "step": 538
-    },
-    {
-      "epoch": 0.16962129689616584,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.95176134560457e-05,
-      "loss": 1.4449,
-      "step": 540
-    },
-    {
-      "epoch": 0.1702495239217072,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.9515074579498575e-05,
-      "loss": 1.3835,
-      "step": 542
-    },
-    {
-      "epoch": 0.17087775094724855,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.9512535702951446e-05,
-      "loss": 1.4734,
-      "step": 544
-    },
-    {
-      "epoch": 0.1715059779727899,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.9509996826404317e-05,
-      "loss": 1.4759,
-      "step": 546
-    },
-    {
-      "epoch": 0.17213420499833126,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.950745794985719e-05,
-      "loss": 1.3622,
-      "step": 548
-    },
-    {
-      "epoch": 0.17276243202387262,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.9504919073310063e-05,
-      "loss": 1.5,
-      "step": 550
-    },
-    {
-      "epoch": 0.17339065904941398,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.9502380196762935e-05,
-      "loss": 1.4584,
-      "step": 552
-    },
-    {
-      "epoch": 0.17401888607495533,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.9499841320215806e-05,
-      "loss": 1.1847,
-      "step": 554
-    },
-    {
-      "epoch": 0.1746471131004967,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.9497302443668678e-05,
-      "loss": 1.4887,
-      "step": 556
-    },
-    {
-      "epoch": 0.17527534012603804,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.9494763567121552e-05,
-      "loss": 1.3411,
-      "step": 558
-    },
-    {
-      "epoch": 0.1759035671515794,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.949222469057442e-05,
-      "loss": 1.4525,
-      "step": 560
-    },
-    {
-      "epoch": 0.17653179417712075,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.9489685814027295e-05,
-      "loss": 1.2743,
-      "step": 562
-    },
-    {
-      "epoch": 0.1771600212026621,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.9487146937480167e-05,
-      "loss": 1.385,
-      "step": 564
-    },
-    {
-      "epoch": 0.17778824822820347,
-      "grad_norm": 0.89453125,
-      "learning_rate": 1.9484608060933038e-05,
-      "loss": 1.3988,
-      "step": 566
-    },
-    {
-      "epoch": 0.17841647525374482,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.948206918438591e-05,
-      "loss": 1.2637,
-      "step": 568
-    },
-    {
-      "epoch": 0.17904470227928618,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.9479530307838784e-05,
-      "loss": 1.3832,
-      "step": 570
-    },
-    {
-      "epoch": 0.17967292930482753,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.9476991431291655e-05,
-      "loss": 1.367,
-      "step": 572
-    },
-    {
-      "epoch": 0.1803011563303689,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.9474452554744527e-05,
-      "loss": 1.3174,
-      "step": 574
-    },
-    {
-      "epoch": 0.18092938335591024,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.9471913678197398e-05,
-      "loss": 1.2966,
-      "step": 576
-    },
-    {
-      "epoch": 0.1815576103814516,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9469374801650273e-05,
-      "loss": 1.4582,
-      "step": 578
-    },
-    {
-      "epoch": 0.18218583740699296,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.946683592510314e-05,
-      "loss": 1.3229,
-      "step": 580
-    },
-    {
-      "epoch": 0.1828140644325343,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.9464297048556016e-05,
-      "loss": 1.2895,
-      "step": 582
-    },
-    {
-      "epoch": 0.18344229145807567,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.9461758172008887e-05,
-      "loss": 1.5382,
-      "step": 584
-    },
-    {
-      "epoch": 0.18407051848361702,
-      "grad_norm": 0.92578125,
-      "learning_rate": 1.945921929546176e-05,
-      "loss": 1.4349,
-      "step": 586
-    },
-    {
-      "epoch": 0.18469874550915838,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.945668041891463e-05,
-      "loss": 1.3861,
-      "step": 588
-    },
-    {
-      "epoch": 0.18532697253469974,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.9454141542367505e-05,
-      "loss": 1.2897,
-      "step": 590
-    },
-    {
-      "epoch": 0.1859551995602411,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.9451602665820376e-05,
-      "loss": 1.3362,
-      "step": 592
-    },
-    {
-      "epoch": 0.18658342658578245,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.9449063789273247e-05,
-      "loss": 1.3954,
-      "step": 594
-    },
-    {
-      "epoch": 0.1872116536113238,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.9446524912726122e-05,
-      "loss": 1.3541,
-      "step": 596
-    },
-    {
-      "epoch": 0.18783988063686516,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.9443986036178994e-05,
-      "loss": 1.4498,
-      "step": 598
-    },
-    {
-      "epoch": 0.18846810766240651,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.9441447159631865e-05,
-      "loss": 1.3767,
-      "step": 600
-    },
-    {
-      "epoch": 0.18909633468794787,
-      "grad_norm": 0.9453125,
-      "learning_rate": 1.9438908283084736e-05,
-      "loss": 1.245,
-      "step": 602
-    },
-    {
-      "epoch": 0.18972456171348923,
-      "grad_norm": 0.66015625,
-      "learning_rate": 1.943636940653761e-05,
-      "loss": 1.4371,
-      "step": 604
-    },
-    {
-      "epoch": 0.19035278873903055,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.943383052999048e-05,
-      "loss": 1.3194,
-      "step": 606
-    },
-    {
-      "epoch": 0.1909810157645719,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.9431291653443354e-05,
-      "loss": 1.3339,
-      "step": 608
-    },
-    {
-      "epoch": 0.19160924279011327,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.9428752776896225e-05,
-      "loss": 1.3773,
-      "step": 610
-    },
-    {
-      "epoch": 0.19223746981565462,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.9426213900349097e-05,
-      "loss": 1.3456,
-      "step": 612
-    },
-    {
-      "epoch": 0.19286569684119598,
-      "grad_norm": 1.015625,
-      "learning_rate": 1.9423675023801968e-05,
-      "loss": 1.3713,
-      "step": 614
-    },
-    {
-      "epoch": 0.19349392386673733,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.9421136147254843e-05,
-      "loss": 1.4541,
-      "step": 616
-    },
-    {
-      "epoch": 0.1941221508922787,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.9418597270707714e-05,
-      "loss": 1.4132,
-      "step": 618
-    },
-    {
-      "epoch": 0.19475037791782004,
-      "grad_norm": 0.90234375,
-      "learning_rate": 1.9416058394160586e-05,
-      "loss": 1.461,
-      "step": 620
-    },
-    {
-      "epoch": 0.1953786049433614,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.9413519517613457e-05,
-      "loss": 1.2741,
-      "step": 622
-    },
-    {
-      "epoch": 0.19600683196890276,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9410980641066332e-05,
-      "loss": 1.4783,
-      "step": 624
-    },
-    {
-      "epoch": 0.1966350589944441,
-      "grad_norm": 0.9296875,
-      "learning_rate": 1.94084417645192e-05,
-      "loss": 1.3676,
-      "step": 626
-    },
-    {
-      "epoch": 0.19726328601998547,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.9405902887972074e-05,
-      "loss": 1.2958,
-      "step": 628
-    },
-    {
-      "epoch": 0.19789151304552682,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.9403364011424946e-05,
-      "loss": 1.3159,
-      "step": 630
-    },
-    {
-      "epoch": 0.19851974007106818,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.9400825134877817e-05,
-      "loss": 1.3835,
-      "step": 632
-    },
-    {
-      "epoch": 0.19914796709660953,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.939828625833069e-05,
-      "loss": 1.2554,
-      "step": 634
-    },
-    {
-      "epoch": 0.1997761941221509,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.9395747381783563e-05,
-      "loss": 1.3815,
-      "step": 636
-    },
-    {
-      "epoch": 0.20040442114769225,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.9393208505236435e-05,
-      "loss": 1.4323,
-      "step": 638
-    },
-    {
-      "epoch": 0.2010326481732336,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.9390669628689306e-05,
-      "loss": 1.3292,
-      "step": 640
-    },
-    {
-      "epoch": 0.20166087519877496,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.9388130752142178e-05,
-      "loss": 1.2865,
-      "step": 642
-    },
-    {
-      "epoch": 0.20228910222431631,
-      "grad_norm": 1.015625,
-      "learning_rate": 1.9385591875595052e-05,
-      "loss": 1.3822,
-      "step": 644
-    },
-    {
-      "epoch": 0.20291732924985767,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9383052999047924e-05,
-      "loss": 1.3657,
-      "step": 646
-    },
-    {
-      "epoch": 0.20354555627539903,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.9380514122500795e-05,
-      "loss": 1.3554,
-      "step": 648
-    },
-    {
-      "epoch": 0.20417378330094038,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9377975245953666e-05,
-      "loss": 1.331,
-      "step": 650
-    },
-    {
-      "epoch": 0.20480201032648174,
-      "grad_norm": 0.96484375,
-      "learning_rate": 1.9375436369406538e-05,
-      "loss": 1.3798,
-      "step": 652
-    },
-    {
-      "epoch": 0.2054302373520231,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.937289749285941e-05,
-      "loss": 1.4126,
-      "step": 654
-    },
-    {
-      "epoch": 0.20605846437756445,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.9370358616312284e-05,
-      "loss": 1.5412,
-      "step": 656
-    },
-    {
-      "epoch": 0.2066866914031058,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.9367819739765155e-05,
-      "loss": 1.4367,
-      "step": 658
-    },
-    {
-      "epoch": 0.20731491842864716,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.9365280863218027e-05,
-      "loss": 1.3944,
-      "step": 660
-    },
-    {
-      "epoch": 0.20794314545418852,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.9362741986670898e-05,
-      "loss": 1.4311,
-      "step": 662
-    },
-    {
-      "epoch": 0.20857137247972987,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.9360203110123773e-05,
-      "loss": 1.452,
-      "step": 664
-    },
-    {
-      "epoch": 0.20919959950527123,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.9357664233576644e-05,
-      "loss": 1.3529,
-      "step": 666
-    },
-    {
-      "epoch": 0.20982782653081258,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.9355125357029516e-05,
-      "loss": 1.3444,
-      "step": 668
-    },
-    {
-      "epoch": 0.21045605355635394,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.9352586480482387e-05,
-      "loss": 1.352,
-      "step": 670
-    },
-    {
-      "epoch": 0.2110842805818953,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.9350047603935262e-05,
-      "loss": 1.455,
-      "step": 672
-    },
-    {
-      "epoch": 0.21171250760743665,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.934750872738813e-05,
-      "loss": 1.2581,
-      "step": 674
-    },
-    {
-      "epoch": 0.21234073463297798,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.9344969850841005e-05,
-      "loss": 1.3224,
-      "step": 676
-    },
-    {
-      "epoch": 0.21296896165851933,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.9342430974293876e-05,
-      "loss": 1.4604,
-      "step": 678
-    },
-    {
-      "epoch": 0.2135971886840607,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9339892097746747e-05,
-      "loss": 1.2345,
-      "step": 680
-    },
-    {
-      "epoch": 0.21422541570960205,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.9337353221199622e-05,
-      "loss": 1.4289,
-      "step": 682
-    },
-    {
-      "epoch": 0.2148536427351434,
-      "grad_norm": 0.875,
-      "learning_rate": 1.9334814344652494e-05,
-      "loss": 1.4216,
-      "step": 684
-    },
-    {
-      "epoch": 0.21548186976068476,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9332275468105365e-05,
-      "loss": 1.4541,
-      "step": 686
-    },
-    {
-      "epoch": 0.2161100967862261,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.9329736591558236e-05,
-      "loss": 1.3089,
-      "step": 688
-    },
-    {
-      "epoch": 0.21673832381176747,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.932719771501111e-05,
-      "loss": 1.3822,
-      "step": 690
-    },
-    {
-      "epoch": 0.21736655083730883,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.9324658838463982e-05,
-      "loss": 1.2839,
-      "step": 692
-    },
-    {
-      "epoch": 0.21799477786285018,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.9322119961916854e-05,
-      "loss": 1.2813,
-      "step": 694
-    },
-    {
-      "epoch": 0.21862300488839154,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.9319581085369725e-05,
-      "loss": 1.3437,
-      "step": 696
-    },
-    {
-      "epoch": 0.2192512319139329,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.93170422088226e-05,
-      "loss": 1.3649,
-      "step": 698
-    },
-    {
-      "epoch": 0.21987945893947425,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.9314503332275468e-05,
-      "loss": 1.3949,
-      "step": 700
-    },
-    {
-      "epoch": 0.2205076859650156,
-      "grad_norm": 0.9921875,
-      "learning_rate": 1.9311964455728343e-05,
-      "loss": 1.3488,
-      "step": 702
-    },
-    {
-      "epoch": 0.22113591299055696,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.9309425579181214e-05,
-      "loss": 1.268,
-      "step": 704
-    },
-    {
-      "epoch": 0.22176414001609832,
-      "grad_norm": 0.875,
-      "learning_rate": 1.9306886702634085e-05,
-      "loss": 1.3855,
-      "step": 706
-    },
-    {
-      "epoch": 0.22239236704163967,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.9304347826086957e-05,
-      "loss": 1.2877,
-      "step": 708
-    },
-    {
-      "epoch": 0.22302059406718103,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.930180894953983e-05,
-      "loss": 1.3324,
-      "step": 710
-    },
-    {
-      "epoch": 0.22364882109272238,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.9299270072992703e-05,
-      "loss": 1.3602,
-      "step": 712
-    },
-    {
-      "epoch": 0.22427704811826374,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.9296731196445574e-05,
-      "loss": 1.3006,
-      "step": 714
-    },
-    {
-      "epoch": 0.2249052751438051,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.9294192319898446e-05,
-      "loss": 1.3667,
-      "step": 716
-    },
-    {
-      "epoch": 0.22553350216934645,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.929165344335132e-05,
-      "loss": 1.3536,
-      "step": 718
-    },
-    {
-      "epoch": 0.2261617291948878,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.928911456680419e-05,
-      "loss": 1.4343,
-      "step": 720
-    },
-    {
-      "epoch": 0.22678995622042916,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.9286575690257063e-05,
-      "loss": 1.355,
-      "step": 722
-    },
-    {
-      "epoch": 0.22741818324597052,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.9284036813709935e-05,
-      "loss": 1.3999,
-      "step": 724
-    },
-    {
-      "epoch": 0.22804641027151187,
-      "grad_norm": 0.9140625,
-      "learning_rate": 1.9281497937162806e-05,
-      "loss": 1.3638,
-      "step": 726
-    },
-    {
-      "epoch": 0.22867463729705323,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.9278959060615677e-05,
-      "loss": 1.2724,
-      "step": 728
-    },
-    {
-      "epoch": 0.22930286432259459,
-      "grad_norm": 1.0,
-      "learning_rate": 1.9276420184068552e-05,
-      "loss": 1.3783,
-      "step": 730
-    },
-    {
-      "epoch": 0.22993109134813594,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.9273881307521424e-05,
-      "loss": 1.2429,
-      "step": 732
-    },
-    {
-      "epoch": 0.2305593183736773,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.9271342430974295e-05,
-      "loss": 1.4618,
-      "step": 734
-    },
-    {
-      "epoch": 0.23118754539921865,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.9268803554427166e-05,
-      "loss": 1.3145,
-      "step": 736
-    },
-    {
-      "epoch": 0.23181577242476,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.926626467788004e-05,
-      "loss": 1.3562,
-      "step": 738
-    },
-    {
-      "epoch": 0.23244399945030136,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.9263725801332913e-05,
-      "loss": 1.3047,
-      "step": 740
-    },
-    {
-      "epoch": 0.23307222647584272,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.9261186924785784e-05,
-      "loss": 1.4534,
-      "step": 742
-    },
-    {
-      "epoch": 0.23370045350138408,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.9258648048238655e-05,
-      "loss": 1.3435,
-      "step": 744
-    },
-    {
-      "epoch": 0.23432868052692543,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9256109171691527e-05,
-      "loss": 1.3951,
-      "step": 746
-    },
-    {
-      "epoch": 0.23495690755246676,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.9253570295144398e-05,
-      "loss": 1.3799,
-      "step": 748
-    },
-    {
-      "epoch": 0.23558513457800812,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.9251031418597273e-05,
-      "loss": 1.5794,
-      "step": 750
-    },
-    {
-      "epoch": 0.23621336160354947,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.9248492542050144e-05,
-      "loss": 1.3543,
-      "step": 752
-    },
-    {
-      "epoch": 0.23684158862909083,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.9245953665503016e-05,
-      "loss": 1.2956,
-      "step": 754
-    },
-    {
-      "epoch": 0.23746981565463218,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.9243414788955887e-05,
-      "loss": 1.2537,
-      "step": 756
-    },
-    {
-      "epoch": 0.23809804268017354,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.9240875912408762e-05,
-      "loss": 1.3696,
-      "step": 758
-    },
-    {
-      "epoch": 0.2387262697057149,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.9238337035861633e-05,
-      "loss": 1.4097,
-      "step": 760
-    },
-    {
-      "epoch": 0.23935449673125625,
-      "grad_norm": 1.015625,
-      "learning_rate": 1.9235798159314505e-05,
-      "loss": 1.307,
-      "step": 762
-    },
-    {
-      "epoch": 0.2399827237567976,
-      "grad_norm": 0.91015625,
-      "learning_rate": 1.923325928276738e-05,
-      "loss": 1.2294,
-      "step": 764
-    },
-    {
-      "epoch": 0.24061095078233896,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.923072040622025e-05,
-      "loss": 1.3091,
-      "step": 766
-    },
-    {
-      "epoch": 0.24123917780788032,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.9228181529673122e-05,
-      "loss": 1.3432,
-      "step": 768
-    },
-    {
-      "epoch": 0.24186740483342167,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.9225642653125993e-05,
-      "loss": 1.3201,
-      "step": 770
-    },
-    {
-      "epoch": 0.24249563185896303,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.9223103776578865e-05,
-      "loss": 1.4521,
-      "step": 772
-    },
-    {
-      "epoch": 0.24312385888450438,
-      "grad_norm": 0.95703125,
-      "learning_rate": 1.9220564900031736e-05,
-      "loss": 1.4254,
-      "step": 774
-    },
-    {
-      "epoch": 0.24375208591004574,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.921802602348461e-05,
-      "loss": 1.3347,
-      "step": 776
-    },
-    {
-      "epoch": 0.2443803129355871,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.9215487146937482e-05,
-      "loss": 1.2956,
-      "step": 778
-    },
-    {
-      "epoch": 0.24500853996112845,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.9212948270390354e-05,
-      "loss": 1.3128,
-      "step": 780
-    },
-    {
-      "epoch": 0.2456367669866698,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.9210409393843225e-05,
-      "loss": 1.3065,
-      "step": 782
-    },
-    {
-      "epoch": 0.24626499401221116,
-      "grad_norm": 1.015625,
-      "learning_rate": 1.92078705172961e-05,
-      "loss": 1.2968,
-      "step": 784
-    },
-    {
-      "epoch": 0.24689322103775252,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.920533164074897e-05,
-      "loss": 1.3041,
-      "step": 786
-    },
-    {
-      "epoch": 0.24752144806329388,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.9202792764201843e-05,
-      "loss": 1.4266,
-      "step": 788
-    },
-    {
-      "epoch": 0.24814967508883523,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.9200253887654714e-05,
-      "loss": 1.4958,
-      "step": 790
-    },
-    {
-      "epoch": 0.2487779021143766,
-      "grad_norm": 0.9375,
-      "learning_rate": 1.919771501110759e-05,
-      "loss": 1.4217,
-      "step": 792
-    },
-    {
-      "epoch": 0.24940612913991794,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.9195176134560457e-05,
-      "loss": 1.3905,
-      "step": 794
-    },
-    {
-      "epoch": 0.25003435616545927,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.919263725801333e-05,
-      "loss": 1.3715,
-      "step": 796
-    },
-    {
-      "epoch": 0.25066258319100065,
-      "grad_norm": 0.9296875,
-      "learning_rate": 1.9190098381466203e-05,
-      "loss": 1.4086,
-      "step": 798
-    },
-    {
-      "epoch": 0.251290810216542,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.9187559504919074e-05,
-      "loss": 1.4157,
-      "step": 800
-    },
-    {
-      "epoch": 0.25191903724208337,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.9185020628371946e-05,
-      "loss": 1.2557,
-      "step": 802
-    },
-    {
-      "epoch": 0.2525472642676247,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.918248175182482e-05,
-      "loss": 1.3713,
-      "step": 804
-    },
-    {
-      "epoch": 0.2531754912931661,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.9179942875277692e-05,
-      "loss": 1.3549,
-      "step": 806
-    },
-    {
-      "epoch": 0.2538037183187074,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.9177403998730563e-05,
-      "loss": 1.4184,
-      "step": 808
-    },
-    {
-      "epoch": 0.2544319453442488,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.9174865122183435e-05,
-      "loss": 1.3112,
-      "step": 810
-    },
-    {
-      "epoch": 0.2550601723697901,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.917232624563631e-05,
-      "loss": 1.3818,
-      "step": 812
-    },
-    {
-      "epoch": 0.2556883993953315,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9169787369089177e-05,
-      "loss": 1.4245,
-      "step": 814
-    },
-    {
-      "epoch": 0.25631662642087283,
-      "grad_norm": 0.91015625,
-      "learning_rate": 1.9167248492542052e-05,
-      "loss": 1.3986,
-      "step": 816
-    },
-    {
-      "epoch": 0.2569448534464142,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.9164709615994924e-05,
-      "loss": 1.3054,
-      "step": 818
-    },
-    {
-      "epoch": 0.25757308047195554,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.9162170739447795e-05,
-      "loss": 1.3303,
-      "step": 820
-    },
-    {
-      "epoch": 0.2582013074974969,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.9159631862900666e-05,
-      "loss": 1.3877,
-      "step": 822
-    },
-    {
-      "epoch": 0.25882953452303825,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.915709298635354e-05,
-      "loss": 1.3464,
-      "step": 824
-    },
-    {
-      "epoch": 0.25945776154857964,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.9154554109806412e-05,
-      "loss": 1.4358,
-      "step": 826
-    },
-    {
-      "epoch": 0.26008598857412096,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.9152015233259284e-05,
-      "loss": 1.2982,
-      "step": 828
-    },
-    {
-      "epoch": 0.26071421559966235,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9149476356712155e-05,
-      "loss": 1.398,
-      "step": 830
-    },
-    {
-      "epoch": 0.2613424426252037,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.914693748016503e-05,
-      "loss": 1.2641,
-      "step": 832
-    },
-    {
-      "epoch": 0.26197066965074506,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.91443986036179e-05,
-      "loss": 1.3669,
-      "step": 834
-    },
-    {
-      "epoch": 0.2625988966762864,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9141859727070773e-05,
-      "loss": 1.3182,
-      "step": 836
-    },
-    {
-      "epoch": 0.26322712370182777,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.9139320850523644e-05,
-      "loss": 1.3939,
-      "step": 838
-    },
-    {
-      "epoch": 0.2638553507273691,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.9136781973976516e-05,
-      "loss": 1.4948,
-      "step": 840
-    },
-    {
-      "epoch": 0.2644835777529105,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.9134243097429387e-05,
-      "loss": 1.34,
-      "step": 842
-    },
-    {
-      "epoch": 0.2651118047784518,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.913170422088226e-05,
-      "loss": 1.3576,
-      "step": 844
-    },
-    {
-      "epoch": 0.2657400318039932,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.9129165344335133e-05,
-      "loss": 1.3366,
-      "step": 846
-    },
-    {
-      "epoch": 0.2663682588295345,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.9126626467788004e-05,
-      "loss": 1.4665,
-      "step": 848
-    },
-    {
-      "epoch": 0.2669964858550759,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.912408759124088e-05,
-      "loss": 1.4036,
-      "step": 850
-    },
-    {
-      "epoch": 0.26762471288061723,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.912154871469375e-05,
-      "loss": 1.2714,
-      "step": 852
-    },
-    {
-      "epoch": 0.26825293990615856,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.9119009838146622e-05,
-      "loss": 1.3858,
-      "step": 854
-    },
-    {
-      "epoch": 0.26888116693169994,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.9116470961599493e-05,
-      "loss": 1.4882,
-      "step": 856
-    },
-    {
-      "epoch": 0.26950939395724127,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.9113932085052368e-05,
-      "loss": 1.2592,
-      "step": 858
-    },
-    {
-      "epoch": 0.27013762098278266,
-      "grad_norm": 0.75,
-      "learning_rate": 1.911139320850524e-05,
-      "loss": 1.4349,
-      "step": 860
-    },
-    {
-      "epoch": 0.270765848008324,
-      "grad_norm": 0.921875,
-      "learning_rate": 1.910885433195811e-05,
-      "loss": 1.2003,
-      "step": 862
-    },
-    {
-      "epoch": 0.27139407503386537,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.9106315455410982e-05,
-      "loss": 1.4485,
-      "step": 864
-    },
-    {
-      "epoch": 0.2720223020594067,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.9103776578863854e-05,
-      "loss": 1.2389,
-      "step": 866
-    },
-    {
-      "epoch": 0.2726505290849481,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9101237702316725e-05,
-      "loss": 1.4348,
-      "step": 868
-    },
-    {
-      "epoch": 0.2732787561104894,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.90986988257696e-05,
-      "loss": 1.4559,
-      "step": 870
-    },
-    {
-      "epoch": 0.2739069831360308,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.909615994922247e-05,
-      "loss": 1.4004,
-      "step": 872
-    },
-    {
-      "epoch": 0.2745352101615721,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.9093621072675343e-05,
-      "loss": 1.3105,
-      "step": 874
-    },
-    {
-      "epoch": 0.2751634371871135,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.9091082196128214e-05,
-      "loss": 1.2796,
-      "step": 876
-    },
-    {
-      "epoch": 0.27579166421265483,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.908854331958109e-05,
-      "loss": 1.4628,
-      "step": 878
-    },
-    {
-      "epoch": 0.2764198912381962,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.908600444303396e-05,
-      "loss": 1.3618,
-      "step": 880
-    },
-    {
-      "epoch": 0.27704811826373754,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.908346556648683e-05,
-      "loss": 1.3635,
-      "step": 882
-    },
-    {
-      "epoch": 0.2776763452892789,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.9080926689939703e-05,
-      "loss": 1.3921,
-      "step": 884
-    },
-    {
-      "epoch": 0.27830457231482025,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.9078387813392578e-05,
-      "loss": 1.3431,
-      "step": 886
-    },
-    {
-      "epoch": 0.27893279934036164,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.9075848936845446e-05,
-      "loss": 1.3725,
-      "step": 888
-    },
-    {
-      "epoch": 0.27956102636590296,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.907331006029832e-05,
-      "loss": 1.2754,
-      "step": 890
-    },
-    {
-      "epoch": 0.28018925339144435,
-      "grad_norm": 0.99609375,
-      "learning_rate": 1.9070771183751192e-05,
-      "loss": 1.1762,
-      "step": 892
-    },
-    {
-      "epoch": 0.2808174804169857,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.9068232307204063e-05,
-      "loss": 1.301,
-      "step": 894
-    },
-    {
-      "epoch": 0.28144570744252706,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.9065693430656935e-05,
-      "loss": 1.2999,
-      "step": 896
-    },
-    {
-      "epoch": 0.2820739344680684,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.906315455410981e-05,
-      "loss": 1.355,
-      "step": 898
-    },
-    {
-      "epoch": 0.28270216149360977,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.906061567756268e-05,
-      "loss": 1.4332,
-      "step": 900
-    },
-    {
-      "epoch": 0.2833303885191511,
-      "grad_norm": 0.96875,
-      "learning_rate": 1.9058076801015552e-05,
-      "loss": 1.4116,
-      "step": 902
-    },
-    {
-      "epoch": 0.2839586155446925,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.9055537924468423e-05,
-      "loss": 1.3064,
-      "step": 904
-    },
-    {
-      "epoch": 0.2845868425702338,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.9052999047921298e-05,
-      "loss": 1.5111,
-      "step": 906
-    },
-    {
-      "epoch": 0.2852150695957752,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.9050460171374166e-05,
-      "loss": 1.2457,
-      "step": 908
-    },
-    {
-      "epoch": 0.2858432966213165,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.904792129482704e-05,
-      "loss": 1.346,
-      "step": 910
-    },
-    {
-      "epoch": 0.2864715236468579,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.9045382418279912e-05,
-      "loss": 1.3722,
-      "step": 912
-    },
-    {
-      "epoch": 0.28709975067239923,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.9042843541732784e-05,
-      "loss": 1.3245,
-      "step": 914
-    },
-    {
-      "epoch": 0.2877279776979406,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.9040304665185655e-05,
-      "loss": 1.42,
-      "step": 916
-    },
-    {
-      "epoch": 0.28835620472348195,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.903776578863853e-05,
-      "loss": 1.405,
-      "step": 918
-    },
-    {
-      "epoch": 0.28898443174902333,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.90352269120914e-05,
-      "loss": 1.3066,
-      "step": 920
-    },
-    {
-      "epoch": 0.28961265877456466,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.9032688035544273e-05,
-      "loss": 1.3226,
-      "step": 922
-    },
-    {
-      "epoch": 0.29024088580010604,
-      "grad_norm": 0.875,
-      "learning_rate": 1.9030149158997144e-05,
-      "loss": 1.1937,
-      "step": 924
-    },
-    {
-      "epoch": 0.29086911282564737,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.902761028245002e-05,
-      "loss": 1.3474,
-      "step": 926
-    },
-    {
-      "epoch": 0.2914973398511887,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.9025071405902887e-05,
-      "loss": 1.3306,
-      "step": 928
-    },
-    {
-      "epoch": 0.2921255668767301,
-      "grad_norm": 0.89453125,
-      "learning_rate": 1.902253252935576e-05,
-      "loss": 1.3498,
-      "step": 930
-    },
-    {
-      "epoch": 0.2927537939022714,
-      "grad_norm": 0.92578125,
-      "learning_rate": 1.9019993652808633e-05,
-      "loss": 1.4435,
-      "step": 932
-    },
-    {
-      "epoch": 0.2933820209278128,
-      "grad_norm": 0.75,
-      "learning_rate": 1.9017454776261504e-05,
-      "loss": 1.3682,
-      "step": 934
-    },
-    {
-      "epoch": 0.2940102479533541,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.901491589971438e-05,
-      "loss": 1.3391,
-      "step": 936
-    },
-    {
-      "epoch": 0.2946384749788955,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.901237702316725e-05,
-      "loss": 1.5098,
-      "step": 938
-    },
-    {
-      "epoch": 0.29526670200443683,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.9009838146620122e-05,
-      "loss": 1.4432,
-      "step": 940
-    },
-    {
-      "epoch": 0.2958949290299782,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.9007299270072993e-05,
-      "loss": 1.3789,
-      "step": 942
-    },
-    {
-      "epoch": 0.29652315605551954,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.9004760393525868e-05,
-      "loss": 1.2943,
-      "step": 944
-    },
-    {
-      "epoch": 0.2971513830810609,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.900222151697874e-05,
-      "loss": 1.3753,
-      "step": 946
-    },
-    {
-      "epoch": 0.29777961010660225,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.899968264043161e-05,
-      "loss": 1.3704,
-      "step": 948
-    },
-    {
-      "epoch": 0.29840783713214364,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.8997143763884482e-05,
-      "loss": 1.4176,
-      "step": 950
-    },
-    {
-      "epoch": 0.29903606415768497,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8994604887337357e-05,
-      "loss": 1.2448,
-      "step": 952
-    },
-    {
-      "epoch": 0.29966429118322635,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.8992066010790225e-05,
-      "loss": 1.2357,
-      "step": 954
-    },
-    {
-      "epoch": 0.3002925182087677,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.89895271342431e-05,
-      "loss": 1.4002,
-      "step": 956
-    },
-    {
-      "epoch": 0.30092074523430906,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.898698825769597e-05,
-      "loss": 1.3756,
-      "step": 958
-    },
-    {
-      "epoch": 0.3015489722598504,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8984449381148842e-05,
-      "loss": 1.2851,
-      "step": 960
-    },
-    {
-      "epoch": 0.3021771992853918,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.8981910504601714e-05,
-      "loss": 1.3339,
-      "step": 962
-    },
-    {
-      "epoch": 0.3028054263109331,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.897937162805459e-05,
-      "loss": 1.4284,
-      "step": 964
-    },
-    {
-      "epoch": 0.3034336533364745,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.897683275150746e-05,
-      "loss": 1.3142,
-      "step": 966
-    },
-    {
-      "epoch": 0.3040618803620158,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.897429387496033e-05,
-      "loss": 1.4217,
-      "step": 968
-    },
-    {
-      "epoch": 0.3046901073875572,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8971754998413203e-05,
-      "loss": 1.4308,
-      "step": 970
-    },
-    {
-      "epoch": 0.3053183344130985,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8969216121866078e-05,
-      "loss": 1.2463,
-      "step": 972
-    },
-    {
-      "epoch": 0.3059465614386399,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.896667724531895e-05,
-      "loss": 1.3149,
-      "step": 974
-    },
-    {
-      "epoch": 0.30657478846418124,
-      "grad_norm": 0.91796875,
-      "learning_rate": 1.896413836877182e-05,
-      "loss": 1.3623,
-      "step": 976
-    },
-    {
-      "epoch": 0.3072030154897226,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.8961599492224692e-05,
-      "loss": 1.5585,
-      "step": 978
-    },
-    {
-      "epoch": 0.30783124251526395,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8959060615677563e-05,
-      "loss": 1.2155,
-      "step": 980
-    },
-    {
-      "epoch": 0.30845946954080533,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.8956521739130434e-05,
-      "loss": 1.384,
-      "step": 982
-    },
-    {
-      "epoch": 0.30908769656634666,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.895398286258331e-05,
-      "loss": 1.2724,
-      "step": 984
-    },
-    {
-      "epoch": 0.30971592359188804,
-      "grad_norm": 0.75,
-      "learning_rate": 1.895144398603618e-05,
-      "loss": 1.3504,
-      "step": 986
-    },
-    {
-      "epoch": 0.31034415061742937,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.8948905109489052e-05,
-      "loss": 1.3144,
-      "step": 988
-    },
-    {
-      "epoch": 0.31097237764297075,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.8946366232941923e-05,
-      "loss": 1.3399,
-      "step": 990
-    },
-    {
-      "epoch": 0.3116006046685121,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.8943827356394798e-05,
-      "loss": 1.3355,
-      "step": 992
-    },
-    {
-      "epoch": 0.31222883169405347,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.894128847984767e-05,
-      "loss": 1.2823,
-      "step": 994
-    },
-    {
-      "epoch": 0.3128570587195948,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.893874960330054e-05,
-      "loss": 1.4969,
-      "step": 996
-    },
-    {
-      "epoch": 0.3134852857451361,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8936210726753412e-05,
-      "loss": 1.3046,
-      "step": 998
-    },
-    {
-      "epoch": 0.3141135127706775,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.8933671850206287e-05,
-      "loss": 1.4317,
-      "step": 1000
-    },
-    {
-      "epoch": 0.31474173979621883,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8931132973659155e-05,
-      "loss": 1.3786,
-      "step": 1002
-    },
-    {
-      "epoch": 0.3153699668217602,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.892859409711203e-05,
-      "loss": 1.3259,
-      "step": 1004
-    },
-    {
-      "epoch": 0.31599819384730155,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.89260552205649e-05,
-      "loss": 1.3619,
-      "step": 1006
-    },
-    {
-      "epoch": 0.31662642087284293,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.8923516344017773e-05,
-      "loss": 1.4299,
-      "step": 1008
-    },
-    {
-      "epoch": 0.31725464789838426,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.8920977467470644e-05,
-      "loss": 1.389,
-      "step": 1010
-    },
-    {
-      "epoch": 0.31788287492392564,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.891843859092352e-05,
-      "loss": 1.3459,
-      "step": 1012
-    },
-    {
-      "epoch": 0.31851110194946697,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.891589971437639e-05,
-      "loss": 1.4309,
-      "step": 1014
-    },
-    {
-      "epoch": 0.31913932897500835,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.891336083782926e-05,
-      "loss": 1.3712,
-      "step": 1016
-    },
-    {
-      "epoch": 0.3197675560005497,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.8910821961282133e-05,
-      "loss": 1.3044,
-      "step": 1018
-    },
-    {
-      "epoch": 0.32039578302609106,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.8908283084735008e-05,
-      "loss": 1.3589,
-      "step": 1020
-    },
-    {
-      "epoch": 0.3210240100516324,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.890574420818788e-05,
-      "loss": 1.2593,
-      "step": 1022
-    },
-    {
-      "epoch": 0.3216522370771738,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.890320533164075e-05,
-      "loss": 1.3657,
-      "step": 1024
-    },
-    {
-      "epoch": 0.3222804641027151,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.8900666455093625e-05,
-      "loss": 1.2129,
-      "step": 1026
-    },
-    {
-      "epoch": 0.3229086911282565,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8898127578546493e-05,
-      "loss": 1.09,
-      "step": 1028
-    },
-    {
-      "epoch": 0.3235369181537978,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.8895588701999368e-05,
-      "loss": 1.3569,
-      "step": 1030
-    },
-    {
-      "epoch": 0.3241651451793392,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.889304982545224e-05,
-      "loss": 1.4419,
-      "step": 1032
-    },
-    {
-      "epoch": 0.3247933722048805,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.889051094890511e-05,
-      "loss": 1.3802,
-      "step": 1034
-    },
-    {
-      "epoch": 0.3254215992304219,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8887972072357982e-05,
-      "loss": 1.312,
-      "step": 1036
-    },
-    {
-      "epoch": 0.32604982625596324,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.8885433195810857e-05,
-      "loss": 1.4378,
-      "step": 1038
-    },
-    {
-      "epoch": 0.3266780532815046,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.8882894319263728e-05,
-      "loss": 1.2541,
-      "step": 1040
-    },
-    {
-      "epoch": 0.32730628030704595,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.88803554427166e-05,
-      "loss": 1.3656,
-      "step": 1042
-    },
-    {
-      "epoch": 0.32793450733258733,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.887781656616947e-05,
-      "loss": 1.4039,
-      "step": 1044
-    },
-    {
-      "epoch": 0.32856273435812866,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8875277689622346e-05,
-      "loss": 1.3563,
-      "step": 1046
-    },
-    {
-      "epoch": 0.32919096138367004,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.8872738813075214e-05,
-      "loss": 1.285,
-      "step": 1048
-    },
-    {
-      "epoch": 0.3298191884092114,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.887019993652809e-05,
-      "loss": 1.2465,
-      "step": 1050
-    },
-    {
-      "epoch": 0.33044741543475276,
-      "grad_norm": 0.92578125,
-      "learning_rate": 1.886766105998096e-05,
-      "loss": 1.2184,
-      "step": 1052
-    },
-    {
-      "epoch": 0.3310756424602941,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.886512218343383e-05,
-      "loss": 1.3098,
-      "step": 1054
-    },
-    {
-      "epoch": 0.33170386948583547,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.8862583306886703e-05,
-      "loss": 1.318,
-      "step": 1056
-    },
-    {
-      "epoch": 0.3323320965113768,
-      "grad_norm": 0.91015625,
-      "learning_rate": 1.8860044430339577e-05,
-      "loss": 1.2984,
-      "step": 1058
-    },
-    {
-      "epoch": 0.3329603235369182,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.885750555379245e-05,
-      "loss": 1.4075,
-      "step": 1060
-    },
-    {
-      "epoch": 0.3335885505624595,
-      "grad_norm": 0.94140625,
-      "learning_rate": 1.885496667724532e-05,
-      "loss": 1.354,
-      "step": 1062
-    },
-    {
-      "epoch": 0.3342167775880009,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.885242780069819e-05,
-      "loss": 1.2434,
-      "step": 1064
-    },
-    {
-      "epoch": 0.3348450046135422,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.8849888924151066e-05,
-      "loss": 1.4308,
-      "step": 1066
-    },
-    {
-      "epoch": 0.33547323163908355,
-      "grad_norm": 0.8984375,
-      "learning_rate": 1.8847350047603938e-05,
-      "loss": 1.2561,
-      "step": 1068
-    },
-    {
-      "epoch": 0.33610145866462493,
-      "grad_norm": 0.875,
-      "learning_rate": 1.884481117105681e-05,
-      "loss": 1.4753,
-      "step": 1070
-    },
-    {
-      "epoch": 0.33672968569016626,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.884227229450968e-05,
-      "loss": 1.369,
-      "step": 1072
-    },
-    {
-      "epoch": 0.33735791271570764,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.8839733417962552e-05,
-      "loss": 1.4776,
-      "step": 1074
-    },
-    {
-      "epoch": 0.33798613974124897,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8837194541415423e-05,
-      "loss": 1.3619,
-      "step": 1076
-    },
-    {
-      "epoch": 0.33861436676679035,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.8834655664868298e-05,
-      "loss": 1.2684,
-      "step": 1078
-    },
-    {
-      "epoch": 0.3392425937923317,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.883211678832117e-05,
-      "loss": 1.4172,
-      "step": 1080
-    },
-    {
-      "epoch": 0.33987082081787306,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.882957791177404e-05,
-      "loss": 1.501,
-      "step": 1082
-    },
-    {
-      "epoch": 0.3404990478434144,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.8827039035226912e-05,
-      "loss": 1.4823,
-      "step": 1084
-    },
-    {
-      "epoch": 0.3411272748689558,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.8824500158679787e-05,
-      "loss": 1.2784,
-      "step": 1086
-    },
-    {
-      "epoch": 0.3417555018944971,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.882196128213266e-05,
-      "loss": 1.359,
-      "step": 1088
-    },
-    {
-      "epoch": 0.3423837289200385,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.881942240558553e-05,
-      "loss": 1.2725,
-      "step": 1090
-    },
-    {
-      "epoch": 0.3430119559455798,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.88168835290384e-05,
-      "loss": 1.2185,
-      "step": 1092
-    },
-    {
-      "epoch": 0.3436401829711212,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.8814344652491276e-05,
-      "loss": 1.3709,
-      "step": 1094
-    },
-    {
-      "epoch": 0.3442684099966625,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.8811805775944144e-05,
-      "loss": 1.4139,
-      "step": 1096
-    },
-    {
-      "epoch": 0.3448966370222039,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.880926689939702e-05,
-      "loss": 1.5253,
-      "step": 1098
-    },
-    {
-      "epoch": 0.34552486404774524,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.880672802284989e-05,
-      "loss": 1.2929,
-      "step": 1100
-    },
-    {
-      "epoch": 0.3461530910732866,
-      "grad_norm": 0.90625,
-      "learning_rate": 1.880418914630276e-05,
-      "loss": 1.3314,
-      "step": 1102
-    },
-    {
-      "epoch": 0.34678131809882795,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.8801650269755633e-05,
-      "loss": 1.1409,
-      "step": 1104
-    },
-    {
-      "epoch": 0.34740954512436933,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.8799111393208508e-05,
-      "loss": 1.4453,
-      "step": 1106
-    },
-    {
-      "epoch": 0.34803777214991066,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.879657251666138e-05,
-      "loss": 1.3495,
-      "step": 1108
-    },
-    {
-      "epoch": 0.34866599917545205,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.879403364011425e-05,
-      "loss": 1.3406,
-      "step": 1110
-    },
-    {
-      "epoch": 0.3492942262009934,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.8791494763567125e-05,
-      "loss": 1.2358,
-      "step": 1112
-    },
-    {
-      "epoch": 0.34992245322653476,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.8788955887019997e-05,
-      "loss": 1.3972,
-      "step": 1114
-    },
-    {
-      "epoch": 0.3505506802520761,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8786417010472868e-05,
-      "loss": 1.3597,
-      "step": 1116
-    },
-    {
-      "epoch": 0.35117890727761747,
-      "grad_norm": 0.66015625,
-      "learning_rate": 1.878387813392574e-05,
-      "loss": 1.3003,
-      "step": 1118
-    },
-    {
-      "epoch": 0.3518071343031588,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.8781339257378614e-05,
-      "loss": 1.2663,
-      "step": 1120
-    },
-    {
-      "epoch": 0.3524353613287002,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.8778800380831482e-05,
-      "loss": 1.4089,
-      "step": 1122
-    },
-    {
-      "epoch": 0.3530635883542415,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.8776261504284357e-05,
-      "loss": 1.3793,
-      "step": 1124
-    },
-    {
-      "epoch": 0.3536918153797829,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.8773722627737228e-05,
-      "loss": 1.4041,
-      "step": 1126
-    },
-    {
-      "epoch": 0.3543200424053242,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.87711837511901e-05,
-      "loss": 1.252,
-      "step": 1128
-    },
-    {
-      "epoch": 0.3549482694308656,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.876864487464297e-05,
-      "loss": 1.3771,
-      "step": 1130
-    },
-    {
-      "epoch": 0.35557649645640693,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.8766105998095846e-05,
-      "loss": 1.2952,
-      "step": 1132
-    },
-    {
-      "epoch": 0.3562047234819483,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.8763567121548717e-05,
-      "loss": 1.2377,
-      "step": 1134
-    },
-    {
-      "epoch": 0.35683295050748964,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.876102824500159e-05,
-      "loss": 1.429,
-      "step": 1136
-    },
-    {
-      "epoch": 0.35746117753303097,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.875848936845446e-05,
-      "loss": 1.3617,
-      "step": 1138
-    },
-    {
-      "epoch": 0.35808940455857236,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.8755950491907335e-05,
-      "loss": 1.4136,
-      "step": 1140
-    },
-    {
-      "epoch": 0.3587176315841137,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.8753411615360203e-05,
-      "loss": 1.2859,
-      "step": 1142
-    },
-    {
-      "epoch": 0.35934585860965507,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.8750872738813077e-05,
-      "loss": 1.2145,
-      "step": 1144
-    },
-    {
-      "epoch": 0.3599740856351964,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.874833386226595e-05,
-      "loss": 1.294,
-      "step": 1146
-    },
-    {
-      "epoch": 0.3606023126607378,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.874579498571882e-05,
-      "loss": 1.1749,
-      "step": 1148
-    },
-    {
-      "epoch": 0.3612305396862791,
-      "grad_norm": 0.75,
-      "learning_rate": 1.874325610917169e-05,
-      "loss": 1.2759,
-      "step": 1150
-    },
-    {
-      "epoch": 0.3618587667118205,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.8740717232624566e-05,
-      "loss": 1.2798,
-      "step": 1152
-    },
-    {
-      "epoch": 0.3624869937373618,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.8738178356077438e-05,
-      "loss": 1.3493,
-      "step": 1154
-    },
-    {
-      "epoch": 0.3631152207629032,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.873563947953031e-05,
-      "loss": 1.4311,
-      "step": 1156
-    },
-    {
-      "epoch": 0.36374344778844453,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.873310060298318e-05,
-      "loss": 1.2613,
-      "step": 1158
-    },
-    {
-      "epoch": 0.3643716748139859,
-      "grad_norm": 1.0078125,
-      "learning_rate": 1.8730561726436055e-05,
-      "loss": 1.3474,
-      "step": 1160
-    },
-    {
-      "epoch": 0.36499990183952724,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.8728022849888923e-05,
-      "loss": 1.4257,
-      "step": 1162
-    },
-    {
-      "epoch": 0.3656281288650686,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.8725483973341798e-05,
-      "loss": 1.3411,
-      "step": 1164
-    },
-    {
-      "epoch": 0.36625635589060995,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.872294509679467e-05,
-      "loss": 1.3309,
-      "step": 1166
-    },
-    {
-      "epoch": 0.36688458291615134,
-      "grad_norm": 0.8984375,
-      "learning_rate": 1.872040622024754e-05,
-      "loss": 1.4467,
-      "step": 1168
-    },
-    {
-      "epoch": 0.36751280994169266,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.8717867343700412e-05,
-      "loss": 1.2754,
-      "step": 1170
-    },
-    {
-      "epoch": 0.36814103696723405,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8715328467153287e-05,
-      "loss": 1.4556,
-      "step": 1172
-    },
-    {
-      "epoch": 0.3687692639927754,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.871278959060616e-05,
-      "loss": 1.3598,
-      "step": 1174
-    },
-    {
-      "epoch": 0.36939749101831676,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.871025071405903e-05,
-      "loss": 1.2428,
-      "step": 1176
-    },
-    {
-      "epoch": 0.3700257180438581,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.87077118375119e-05,
-      "loss": 1.3761,
-      "step": 1178
-    },
-    {
-      "epoch": 0.37065394506939947,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.8705172960964776e-05,
-      "loss": 1.3929,
-      "step": 1180
-    },
-    {
-      "epoch": 0.3712821720949408,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.8702634084417647e-05,
-      "loss": 1.2633,
-      "step": 1182
-    },
-    {
-      "epoch": 0.3719103991204822,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.870009520787052e-05,
-      "loss": 1.4286,
-      "step": 1184
-    },
-    {
-      "epoch": 0.3725386261460235,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.869755633132339e-05,
-      "loss": 1.2967,
-      "step": 1186
-    },
-    {
-      "epoch": 0.3731668531715649,
-      "grad_norm": 1.015625,
-      "learning_rate": 1.869501745477626e-05,
-      "loss": 1.3566,
-      "step": 1188
-    },
-    {
-      "epoch": 0.3737950801971062,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8692478578229133e-05,
-      "loss": 1.3837,
-      "step": 1190
-    },
-    {
-      "epoch": 0.3744233072226476,
-      "grad_norm": 0.9296875,
-      "learning_rate": 1.8689939701682008e-05,
-      "loss": 1.369,
-      "step": 1192
-    },
-    {
-      "epoch": 0.37505153424818893,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.868740082513488e-05,
-      "loss": 1.4206,
-      "step": 1194
-    },
-    {
-      "epoch": 0.3756797612737303,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.868486194858775e-05,
-      "loss": 1.3345,
-      "step": 1196
-    },
-    {
-      "epoch": 0.37630798829927165,
-      "grad_norm": 0.84765625,
-      "learning_rate": 1.8682323072040625e-05,
-      "loss": 1.3843,
-      "step": 1198
-    },
-    {
-      "epoch": 0.37693621532481303,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.8679784195493496e-05,
-      "loss": 1.4273,
-      "step": 1200
-    },
-    {
-      "epoch": 0.37756444235035436,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.8677245318946368e-05,
-      "loss": 1.3729,
-      "step": 1202
-    },
-    {
-      "epoch": 0.37819266937589574,
-      "grad_norm": 1.15625,
-      "learning_rate": 1.867470644239924e-05,
-      "loss": 1.1632,
-      "step": 1204
-    },
-    {
-      "epoch": 0.37882089640143707,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.8672167565852114e-05,
-      "loss": 1.3493,
-      "step": 1206
-    },
-    {
-      "epoch": 0.37944912342697845,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.8669628689304985e-05,
-      "loss": 1.3056,
-      "step": 1208
-    },
-    {
-      "epoch": 0.3800773504525198,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.8667089812757857e-05,
-      "loss": 1.414,
-      "step": 1210
-    },
-    {
-      "epoch": 0.3807055774780611,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.8664550936210728e-05,
-      "loss": 1.33,
-      "step": 1212
-    },
-    {
-      "epoch": 0.3813338045036025,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.86620120596636e-05,
-      "loss": 1.378,
-      "step": 1214
-    },
-    {
-      "epoch": 0.3819620315291438,
-      "grad_norm": 0.95703125,
-      "learning_rate": 1.865947318311647e-05,
-      "loss": 1.2628,
-      "step": 1216
-    },
-    {
-      "epoch": 0.3825902585546852,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8656934306569346e-05,
-      "loss": 1.2875,
-      "step": 1218
-    },
-    {
-      "epoch": 0.38321848558022653,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.8654395430022217e-05,
-      "loss": 1.3463,
-      "step": 1220
-    },
-    {
-      "epoch": 0.3838467126057679,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.865185655347509e-05,
-      "loss": 1.3272,
-      "step": 1222
-    },
-    {
-      "epoch": 0.38447493963130924,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.864931767692796e-05,
-      "loss": 1.3908,
-      "step": 1224
-    },
-    {
-      "epoch": 0.3851031666568506,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.8646778800380835e-05,
-      "loss": 1.3235,
-      "step": 1226
-    },
-    {
-      "epoch": 0.38573139368239195,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.8644239923833706e-05,
-      "loss": 1.2354,
-      "step": 1228
-    },
-    {
-      "epoch": 0.38635962070793334,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.8641701047286577e-05,
-      "loss": 1.2592,
-      "step": 1230
-    },
-    {
-      "epoch": 0.38698784773347467,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.863916217073945e-05,
-      "loss": 1.3272,
-      "step": 1232
-    },
-    {
-      "epoch": 0.38761607475901605,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.8636623294192323e-05,
-      "loss": 1.2147,
-      "step": 1234
-    },
-    {
-      "epoch": 0.3882443017845574,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.863408441764519e-05,
-      "loss": 1.3168,
-      "step": 1236
-    },
-    {
-      "epoch": 0.38887252881009876,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.8631545541098066e-05,
-      "loss": 1.2581,
-      "step": 1238
-    },
-    {
-      "epoch": 0.3895007558356401,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.8629006664550938e-05,
-      "loss": 1.404,
-      "step": 1240
-    },
-    {
-      "epoch": 0.3901289828611815,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.862646778800381e-05,
-      "loss": 1.3546,
-      "step": 1242
-    },
-    {
-      "epoch": 0.3907572098867228,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.862392891145668e-05,
-      "loss": 1.2896,
-      "step": 1244
-    },
-    {
-      "epoch": 0.3913854369122642,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.8621390034909555e-05,
-      "loss": 1.3196,
-      "step": 1246
-    },
-    {
-      "epoch": 0.3920136639378055,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8618851158362427e-05,
-      "loss": 1.3084,
-      "step": 1248
-    },
-    {
-      "epoch": 0.3926418909633469,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8616312281815298e-05,
-      "loss": 1.2459,
-      "step": 1250
-    },
-    {
-      "epoch": 0.3932701179888882,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.861377340526817e-05,
-      "loss": 1.3642,
-      "step": 1252
-    },
-    {
-      "epoch": 0.3938983450144296,
-      "grad_norm": 0.9140625,
-      "learning_rate": 1.8611234528721044e-05,
-      "loss": 1.2232,
-      "step": 1254
-    },
-    {
-      "epoch": 0.39452657203997094,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.8608695652173912e-05,
-      "loss": 1.2384,
-      "step": 1256
-    },
-    {
-      "epoch": 0.3951547990655123,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.8606156775626787e-05,
-      "loss": 1.3031,
-      "step": 1258
-    },
-    {
-      "epoch": 0.39578302609105365,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.8603617899079658e-05,
-      "loss": 1.3142,
-      "step": 1260
-    },
-    {
-      "epoch": 0.39641125311659503,
-      "grad_norm": 0.875,
-      "learning_rate": 1.860107902253253e-05,
-      "loss": 1.2851,
-      "step": 1262
-    },
-    {
-      "epoch": 0.39703948014213636,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.85985401459854e-05,
-      "loss": 1.3063,
-      "step": 1264
-    },
-    {
-      "epoch": 0.39766770716767774,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.8596001269438276e-05,
-      "loss": 1.4062,
-      "step": 1266
-    },
-    {
-      "epoch": 0.39829593419321907,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.8593462392891147e-05,
-      "loss": 1.2698,
-      "step": 1268
-    },
-    {
-      "epoch": 0.39892416121876045,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.859092351634402e-05,
-      "loss": 1.3242,
-      "step": 1270
-    },
-    {
-      "epoch": 0.3995523882443018,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.858838463979689e-05,
-      "loss": 1.3655,
-      "step": 1272
-    },
-    {
-      "epoch": 0.40018061526984317,
-      "grad_norm": 0.75,
-      "learning_rate": 1.8585845763249765e-05,
-      "loss": 1.259,
-      "step": 1274
-    },
-    {
-      "epoch": 0.4008088422953845,
-      "grad_norm": 0.8984375,
-      "learning_rate": 1.8583306886702636e-05,
-      "loss": 1.2373,
-      "step": 1276
-    },
-    {
-      "epoch": 0.4014370693209259,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8580768010155507e-05,
-      "loss": 1.3231,
-      "step": 1278
-    },
-    {
-      "epoch": 0.4020652963464672,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.8578229133608382e-05,
-      "loss": 1.3715,
-      "step": 1280
-    },
-    {
-      "epoch": 0.40269352337200853,
-      "grad_norm": 0.91015625,
-      "learning_rate": 1.857569025706125e-05,
-      "loss": 1.4227,
-      "step": 1282
-    },
-    {
-      "epoch": 0.4033217503975499,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8573151380514125e-05,
-      "loss": 1.4352,
-      "step": 1284
-    },
-    {
-      "epoch": 0.40394997742309124,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.8570612503966996e-05,
-      "loss": 1.3358,
-      "step": 1286
-    },
-    {
-      "epoch": 0.40457820444863263,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.8568073627419868e-05,
-      "loss": 1.3508,
-      "step": 1288
-    },
-    {
-      "epoch": 0.40520643147417396,
-      "grad_norm": 0.94921875,
-      "learning_rate": 1.856553475087274e-05,
-      "loss": 1.4527,
-      "step": 1290
-    },
-    {
-      "epoch": 0.40583465849971534,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.8562995874325614e-05,
-      "loss": 1.4456,
-      "step": 1292
-    },
-    {
-      "epoch": 0.40646288552525667,
-      "grad_norm": 0.90625,
-      "learning_rate": 1.8560456997778485e-05,
-      "loss": 1.3093,
-      "step": 1294
-    },
-    {
-      "epoch": 0.40709111255079805,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.8557918121231357e-05,
-      "loss": 1.4534,
-      "step": 1296
-    },
-    {
-      "epoch": 0.4077193395763394,
-      "grad_norm": 0.9609375,
-      "learning_rate": 1.8555379244684228e-05,
-      "loss": 1.2337,
-      "step": 1298
-    },
-    {
-      "epoch": 0.40834756660188076,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8552840368137103e-05,
-      "loss": 1.212,
-      "step": 1300
-    },
-    {
-      "epoch": 0.4089757936274221,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.8550301491589974e-05,
-      "loss": 1.3673,
-      "step": 1302
-    },
-    {
-      "epoch": 0.4096040206529635,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.8547762615042846e-05,
-      "loss": 1.3345,
-      "step": 1304
-    },
-    {
-      "epoch": 0.4102322476785048,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.8545223738495717e-05,
-      "loss": 1.3542,
-      "step": 1306
-    },
-    {
-      "epoch": 0.4108604747040462,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.854268486194859e-05,
-      "loss": 1.4953,
-      "step": 1308
-    },
-    {
-      "epoch": 0.4114887017295875,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.854014598540146e-05,
-      "loss": 1.4254,
-      "step": 1310
-    },
-    {
-      "epoch": 0.4121169287551289,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8537607108854335e-05,
-      "loss": 1.3089,
-      "step": 1312
-    },
-    {
-      "epoch": 0.4127451557806702,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8535068232307206e-05,
-      "loss": 1.3985,
-      "step": 1314
-    },
-    {
-      "epoch": 0.4133733828062116,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.8532529355760077e-05,
-      "loss": 1.45,
-      "step": 1316
-    },
-    {
-      "epoch": 0.41400160983175294,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.852999047921295e-05,
-      "loss": 1.472,
-      "step": 1318
-    },
-    {
-      "epoch": 0.4146298368572943,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.8527451602665823e-05,
-      "loss": 1.4135,
-      "step": 1320
-    },
-    {
-      "epoch": 0.41525806388283565,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.8524912726118695e-05,
-      "loss": 1.2985,
-      "step": 1322
-    },
-    {
-      "epoch": 0.41588629090837703,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.8522373849571566e-05,
-      "loss": 1.292,
-      "step": 1324
-    },
-    {
-      "epoch": 0.41651451793391836,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8519834973024438e-05,
-      "loss": 1.3459,
-      "step": 1326
-    },
-    {
-      "epoch": 0.41714274495945974,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8517296096477312e-05,
-      "loss": 1.3259,
-      "step": 1328
-    },
-    {
-      "epoch": 0.41777097198500107,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.851475721993018e-05,
-      "loss": 1.3027,
-      "step": 1330
-    },
-    {
-      "epoch": 0.41839919901054246,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.8512218343383055e-05,
-      "loss": 1.3385,
-      "step": 1332
-    },
-    {
-      "epoch": 0.4190274260360838,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.8509679466835926e-05,
-      "loss": 1.3775,
-      "step": 1334
-    },
-    {
-      "epoch": 0.41965565306162517,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.8507140590288798e-05,
-      "loss": 1.1561,
-      "step": 1336
-    },
-    {
-      "epoch": 0.4202838800871665,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.850460171374167e-05,
-      "loss": 1.2644,
-      "step": 1338
-    },
-    {
-      "epoch": 0.4209121071127079,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8502062837194544e-05,
-      "loss": 1.3686,
-      "step": 1340
-    },
-    {
-      "epoch": 0.4215403341382492,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.8499523960647415e-05,
-      "loss": 1.4161,
-      "step": 1342
-    },
-    {
-      "epoch": 0.4221685611637906,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.8496985084100287e-05,
-      "loss": 1.3431,
-      "step": 1344
-    },
-    {
-      "epoch": 0.4227967881893319,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.8494446207553158e-05,
-      "loss": 1.3203,
-      "step": 1346
-    },
-    {
-      "epoch": 0.4234250152148733,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.8491907331006033e-05,
-      "loss": 1.3866,
-      "step": 1348
-    },
-    {
-      "epoch": 0.42405324224041463,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.84893684544589e-05,
-      "loss": 1.307,
-      "step": 1350
-    },
-    {
-      "epoch": 0.42468146926595596,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8486829577911776e-05,
-      "loss": 1.3054,
-      "step": 1352
-    },
-    {
-      "epoch": 0.42530969629149734,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.8484290701364647e-05,
-      "loss": 1.263,
-      "step": 1354
-    },
-    {
-      "epoch": 0.42593792331703867,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.848175182481752e-05,
-      "loss": 1.2961,
-      "step": 1356
-    },
-    {
-      "epoch": 0.42656615034258005,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.847921294827039e-05,
-      "loss": 1.386,
-      "step": 1358
-    },
-    {
-      "epoch": 0.4271943773681214,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.8476674071723265e-05,
-      "loss": 1.2587,
-      "step": 1360
-    },
-    {
-      "epoch": 0.42782260439366276,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.8474135195176136e-05,
-      "loss": 1.3613,
-      "step": 1362
-    },
-    {
-      "epoch": 0.4284508314192041,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.8471596318629007e-05,
-      "loss": 1.4578,
-      "step": 1364
-    },
-    {
-      "epoch": 0.4290790584447455,
-      "grad_norm": 0.75,
-      "learning_rate": 1.8469057442081882e-05,
-      "loss": 1.4915,
-      "step": 1366
-    },
-    {
-      "epoch": 0.4297072854702868,
-      "grad_norm": 0.984375,
-      "learning_rate": 1.8466518565534754e-05,
-      "loss": 1.2513,
-      "step": 1368
-    },
-    {
-      "epoch": 0.4303355124958282,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.8463979688987625e-05,
-      "loss": 1.3317,
-      "step": 1370
-    },
-    {
-      "epoch": 0.4309637395213695,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.8461440812440496e-05,
-      "loss": 1.3281,
-      "step": 1372
-    },
-    {
-      "epoch": 0.4315919665469109,
-      "grad_norm": 0.89453125,
-      "learning_rate": 1.845890193589337e-05,
-      "loss": 1.2836,
-      "step": 1374
-    },
-    {
-      "epoch": 0.4322201935724522,
-      "grad_norm": 0.96875,
-      "learning_rate": 1.845636305934624e-05,
-      "loss": 1.3258,
-      "step": 1376
-    },
-    {
-      "epoch": 0.4328484205979936,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.8453824182799114e-05,
-      "loss": 1.3192,
-      "step": 1378
-    },
-    {
-      "epoch": 0.43347664762353494,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8451285306251985e-05,
-      "loss": 1.2383,
-      "step": 1380
-    },
-    {
-      "epoch": 0.4341048746490763,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.8448746429704857e-05,
-      "loss": 1.4198,
-      "step": 1382
-    },
-    {
-      "epoch": 0.43473310167461765,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.8446207553157728e-05,
-      "loss": 1.3262,
-      "step": 1384
-    },
-    {
-      "epoch": 0.43536132870015903,
-      "grad_norm": 0.90234375,
-      "learning_rate": 1.8443668676610603e-05,
-      "loss": 1.3783,
-      "step": 1386
-    },
-    {
-      "epoch": 0.43598955572570036,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.8441129800063474e-05,
-      "loss": 1.3803,
-      "step": 1388
-    },
-    {
-      "epoch": 0.43661778275124175,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.8438590923516346e-05,
-      "loss": 1.2537,
-      "step": 1390
-    },
-    {
-      "epoch": 0.4372460097767831,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.8436052046969217e-05,
-      "loss": 1.4251,
-      "step": 1392
-    },
-    {
-      "epoch": 0.43787423680232446,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.843351317042209e-05,
-      "loss": 1.3708,
-      "step": 1394
-    },
-    {
-      "epoch": 0.4385024638278658,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.8430974293874963e-05,
-      "loss": 1.3983,
-      "step": 1396
-    },
-    {
-      "epoch": 0.43913069085340717,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.8428435417327834e-05,
-      "loss": 1.3208,
-      "step": 1398
-    },
-    {
-      "epoch": 0.4397589178789485,
-      "grad_norm": 0.6484375,
-      "learning_rate": 1.8425896540780706e-05,
-      "loss": 1.2447,
-      "step": 1400
-    },
-    {
-      "epoch": 0.4403871449044899,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.8423357664233577e-05,
-      "loss": 1.4995,
-      "step": 1402
-    },
-    {
-      "epoch": 0.4410153719300312,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.842081878768645e-05,
-      "loss": 1.2333,
-      "step": 1404
-    },
-    {
-      "epoch": 0.4416435989555726,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8418279911139323e-05,
-      "loss": 1.438,
-      "step": 1406
-    },
-    {
-      "epoch": 0.4422718259811139,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.8415741034592195e-05,
-      "loss": 1.3648,
-      "step": 1408
-    },
-    {
-      "epoch": 0.4429000530066553,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.8413202158045066e-05,
-      "loss": 1.3982,
-      "step": 1410
-    },
-    {
-      "epoch": 0.44352828003219663,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.8410663281497937e-05,
-      "loss": 1.2714,
-      "step": 1412
-    },
-    {
-      "epoch": 0.444156507057738,
-      "grad_norm": 0.66015625,
-      "learning_rate": 1.8408124404950812e-05,
-      "loss": 1.3464,
-      "step": 1414
-    },
-    {
-      "epoch": 0.44478473408327934,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.8405585528403684e-05,
-      "loss": 1.3379,
-      "step": 1416
-    },
-    {
-      "epoch": 0.4454129611088207,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8403046651856555e-05,
-      "loss": 1.3022,
-      "step": 1418
-    },
-    {
-      "epoch": 0.44604118813436205,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.8400507775309426e-05,
-      "loss": 1.3677,
-      "step": 1420
-    },
-    {
-      "epoch": 0.4466694151599034,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.83979688987623e-05,
-      "loss": 1.3101,
-      "step": 1422
-    },
-    {
-      "epoch": 0.44729764218544477,
-      "grad_norm": 0.94140625,
-      "learning_rate": 1.839543002221517e-05,
-      "loss": 1.2118,
-      "step": 1424
-    },
-    {
-      "epoch": 0.4479258692109861,
-      "grad_norm": 2.84375,
-      "learning_rate": 1.8392891145668044e-05,
-      "loss": 1.2927,
-      "step": 1426
-    },
-    {
-      "epoch": 0.4485540962365275,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.8390352269120915e-05,
-      "loss": 1.4683,
-      "step": 1428
-    },
-    {
-      "epoch": 0.4491823232620688,
-      "grad_norm": 0.75,
-      "learning_rate": 1.8387813392573787e-05,
-      "loss": 1.2949,
-      "step": 1430
-    },
-    {
-      "epoch": 0.4498105502876102,
-      "grad_norm": 0.75,
-      "learning_rate": 1.8385274516026658e-05,
-      "loss": 1.3789,
-      "step": 1432
-    },
-    {
-      "epoch": 0.4504387773131515,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.8382735639479533e-05,
-      "loss": 1.3308,
-      "step": 1434
-    },
-    {
-      "epoch": 0.4510670043386929,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.8380196762932404e-05,
-      "loss": 1.3221,
-      "step": 1436
-    },
-    {
-      "epoch": 0.45169523136423423,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.8377657886385276e-05,
-      "loss": 1.353,
-      "step": 1438
-    },
-    {
-      "epoch": 0.4523234583897756,
-      "grad_norm": 0.84765625,
-      "learning_rate": 1.8375119009838147e-05,
-      "loss": 1.2386,
-      "step": 1440
-    },
-    {
-      "epoch": 0.45295168541531694,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.8372580133291022e-05,
-      "loss": 1.5192,
-      "step": 1442
-    },
-    {
-      "epoch": 0.4535799124408583,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.837004125674389e-05,
-      "loss": 1.4076,
-      "step": 1444
-    },
-    {
-      "epoch": 0.45420813946639965,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.8367502380196765e-05,
-      "loss": 1.4394,
-      "step": 1446
-    },
-    {
-      "epoch": 0.45483636649194104,
-      "grad_norm": 0.96484375,
-      "learning_rate": 1.836496350364964e-05,
-      "loss": 1.3238,
-      "step": 1448
-    },
-    {
-      "epoch": 0.45546459351748236,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8362424627102507e-05,
-      "loss": 1.2685,
-      "step": 1450
-    },
-    {
-      "epoch": 0.45609282054302375,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8359885750555382e-05,
-      "loss": 1.3883,
-      "step": 1452
-    },
-    {
-      "epoch": 0.4567210475685651,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.8357346874008253e-05,
-      "loss": 1.3735,
-      "step": 1454
-    },
-    {
-      "epoch": 0.45734927459410646,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.8354807997461125e-05,
-      "loss": 1.4432,
-      "step": 1456
-    },
-    {
-      "epoch": 0.4579775016196478,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.8352269120913996e-05,
-      "loss": 1.3395,
-      "step": 1458
-    },
-    {
-      "epoch": 0.45860572864518917,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.834973024436687e-05,
-      "loss": 1.2355,
-      "step": 1460
-    },
-    {
-      "epoch": 0.4592339556707305,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.8347191367819742e-05,
-      "loss": 1.4257,
-      "step": 1462
-    },
-    {
-      "epoch": 0.4598621826962719,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.8344652491272614e-05,
-      "loss": 1.4014,
-      "step": 1464
-    },
-    {
-      "epoch": 0.4604904097218132,
-      "grad_norm": 0.66015625,
-      "learning_rate": 1.8342113614725485e-05,
-      "loss": 1.4452,
-      "step": 1466
-    },
-    {
-      "epoch": 0.4611186367473546,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.833957473817836e-05,
-      "loss": 1.2609,
-      "step": 1468
-    },
-    {
-      "epoch": 0.4617468637728959,
-      "grad_norm": 3.109375,
-      "learning_rate": 1.8337035861631228e-05,
-      "loss": 1.3392,
-      "step": 1470
-    },
-    {
-      "epoch": 0.4623750907984373,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.8334496985084103e-05,
-      "loss": 1.4992,
-      "step": 1472
-    },
-    {
-      "epoch": 0.46300331782397863,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8331958108536974e-05,
-      "loss": 1.3606,
-      "step": 1474
-    },
-    {
-      "epoch": 0.46363154484952,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8329419231989845e-05,
-      "loss": 1.4007,
-      "step": 1476
-    },
-    {
-      "epoch": 0.46425977187506134,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.8326880355442717e-05,
-      "loss": 1.3096,
-      "step": 1478
-    },
-    {
-      "epoch": 0.46488799890060273,
-      "grad_norm": 0.75,
-      "learning_rate": 1.832434147889559e-05,
-      "loss": 1.3873,
-      "step": 1480
-    },
-    {
-      "epoch": 0.46551622592614406,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8321802602348463e-05,
-      "loss": 1.3962,
-      "step": 1482
-    },
-    {
-      "epoch": 0.46614445295168544,
-      "grad_norm": 1.1171875,
-      "learning_rate": 1.8319263725801334e-05,
-      "loss": 1.3512,
-      "step": 1484
-    },
-    {
-      "epoch": 0.46677267997722677,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.8316724849254206e-05,
-      "loss": 1.3546,
-      "step": 1486
-    },
-    {
-      "epoch": 0.46740090700276815,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.831418597270708e-05,
-      "loss": 1.3739,
-      "step": 1488
-    },
-    {
-      "epoch": 0.4680291340283095,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.831164709615995e-05,
-      "loss": 1.3045,
-      "step": 1490
-    },
-    {
-      "epoch": 0.46865736105385086,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8309108219612823e-05,
-      "loss": 1.385,
-      "step": 1492
-    },
-    {
-      "epoch": 0.4692855880793922,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.8306569343065695e-05,
-      "loss": 1.282,
-      "step": 1494
-    },
-    {
-      "epoch": 0.4699138151049335,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.8304030466518566e-05,
-      "loss": 1.5008,
-      "step": 1496
-    },
-    {
-      "epoch": 0.4705420421304749,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.8301491589971437e-05,
-      "loss": 1.422,
-      "step": 1498
-    },
-    {
-      "epoch": 0.47117026915601623,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.8298952713424312e-05,
-      "loss": 1.3419,
-      "step": 1500
-    },
-    {
-      "epoch": 0.4717984961815576,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8296413836877184e-05,
-      "loss": 1.3878,
-      "step": 1502
-    },
-    {
-      "epoch": 0.47242672320709894,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.8293874960330055e-05,
-      "loss": 1.4378,
-      "step": 1504
-    },
-    {
-      "epoch": 0.4730549502326403,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.8291336083782926e-05,
-      "loss": 1.4115,
-      "step": 1506
-    },
-    {
-      "epoch": 0.47368317725818165,
-      "grad_norm": 0.75,
-      "learning_rate": 1.82887972072358e-05,
-      "loss": 1.2909,
-      "step": 1508
-    },
-    {
-      "epoch": 0.47431140428372304,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.8286258330688672e-05,
-      "loss": 1.3813,
-      "step": 1510
-    },
-    {
-      "epoch": 0.47493963130926437,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8283719454141544e-05,
-      "loss": 1.3018,
-      "step": 1512
-    },
-    {
-      "epoch": 0.47556785833480575,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.8281180577594415e-05,
-      "loss": 1.228,
-      "step": 1514
-    },
-    {
-      "epoch": 0.4761960853603471,
-      "grad_norm": 0.65625,
-      "learning_rate": 1.8278641701047287e-05,
-      "loss": 1.3985,
-      "step": 1516
-    },
-    {
-      "epoch": 0.47682431238588846,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.8276102824500158e-05,
-      "loss": 1.4065,
-      "step": 1518
-    },
-    {
-      "epoch": 0.4774525394114298,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.8273563947953033e-05,
-      "loss": 1.34,
-      "step": 1520
-    },
-    {
-      "epoch": 0.47808076643697117,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8271025071405904e-05,
-      "loss": 1.3451,
-      "step": 1522
-    },
-    {
-      "epoch": 0.4787089934625125,
-      "grad_norm": 0.75,
-      "learning_rate": 1.8268486194858776e-05,
-      "loss": 1.3477,
-      "step": 1524
-    },
-    {
-      "epoch": 0.4793372204880539,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.8265947318311647e-05,
-      "loss": 1.3247,
-      "step": 1526
-    },
-    {
-      "epoch": 0.4799654475135952,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8263408441764522e-05,
-      "loss": 1.21,
-      "step": 1528
-    },
-    {
-      "epoch": 0.4805936745391366,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8260869565217393e-05,
-      "loss": 1.3398,
-      "step": 1530
-    },
-    {
-      "epoch": 0.4812219015646779,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.8258330688670264e-05,
-      "loss": 1.3262,
-      "step": 1532
-    },
-    {
-      "epoch": 0.4818501285902193,
-      "grad_norm": 0.75,
-      "learning_rate": 1.825579181212314e-05,
-      "loss": 1.4908,
-      "step": 1534
-    },
-    {
-      "epoch": 0.48247835561576063,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.825325293557601e-05,
-      "loss": 1.3113,
-      "step": 1536
-    },
-    {
-      "epoch": 0.483106582641302,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.8250714059028882e-05,
-      "loss": 1.2718,
-      "step": 1538
-    },
-    {
-      "epoch": 0.48373480966684335,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.8248175182481753e-05,
-      "loss": 1.406,
-      "step": 1540
-    },
-    {
-      "epoch": 0.48436303669238473,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.8245636305934625e-05,
-      "loss": 1.3577,
-      "step": 1542
-    },
-    {
-      "epoch": 0.48499126371792606,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.8243097429387496e-05,
-      "loss": 1.3054,
-      "step": 1544
-    },
-    {
-      "epoch": 0.48561949074346744,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.824055855284037e-05,
-      "loss": 1.3842,
-      "step": 1546
-    },
-    {
-      "epoch": 0.48624771776900877,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.8238019676293242e-05,
-      "loss": 1.3789,
-      "step": 1548
-    },
-    {
-      "epoch": 0.48687594479455015,
-      "grad_norm": 0.65625,
-      "learning_rate": 1.8235480799746114e-05,
-      "loss": 1.3439,
-      "step": 1550
-    },
-    {
-      "epoch": 0.4875041718200915,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.8232941923198985e-05,
-      "loss": 1.3715,
-      "step": 1552
-    },
-    {
-      "epoch": 0.48813239884563286,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.823040304665186e-05,
-      "loss": 1.4506,
-      "step": 1554
-    },
-    {
-      "epoch": 0.4887606258711742,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.822786417010473e-05,
-      "loss": 1.4064,
-      "step": 1556
-    },
-    {
-      "epoch": 0.4893888528967156,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.8225325293557603e-05,
-      "loss": 1.3322,
-      "step": 1558
-    },
-    {
-      "epoch": 0.4900170799222569,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.8222786417010474e-05,
-      "loss": 1.3075,
-      "step": 1560
-    },
-    {
-      "epoch": 0.4906453069477983,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.822024754046335e-05,
-      "loss": 1.4187,
-      "step": 1562
-    },
-    {
-      "epoch": 0.4912735339733396,
-      "grad_norm": 0.8984375,
-      "learning_rate": 1.8217708663916217e-05,
-      "loss": 1.3365,
-      "step": 1564
-    },
-    {
-      "epoch": 0.49190176099888094,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.821516978736909e-05,
-      "loss": 1.3593,
-      "step": 1566
-    },
-    {
-      "epoch": 0.4925299880244223,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.8212630910821963e-05,
-      "loss": 1.298,
-      "step": 1568
-    },
-    {
-      "epoch": 0.49315821504996366,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8210092034274834e-05,
-      "loss": 1.4256,
-      "step": 1570
-    },
-    {
-      "epoch": 0.49378644207550504,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8207553157727706e-05,
-      "loss": 1.4808,
-      "step": 1572
-    },
-    {
-      "epoch": 0.49441466910104637,
-      "grad_norm": 0.875,
-      "learning_rate": 1.820501428118058e-05,
-      "loss": 1.4233,
-      "step": 1574
-    },
-    {
-      "epoch": 0.49504289612658775,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8202475404633452e-05,
-      "loss": 1.3273,
-      "step": 1576
-    },
-    {
-      "epoch": 0.4956711231521291,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.8199936528086323e-05,
-      "loss": 1.3384,
-      "step": 1578
-    },
-    {
-      "epoch": 0.49629935017767046,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.8197397651539195e-05,
-      "loss": 1.4455,
-      "step": 1580
-    },
-    {
-      "epoch": 0.4969275772032118,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.819485877499207e-05,
-      "loss": 1.3235,
-      "step": 1582
-    },
-    {
-      "epoch": 0.4975558042287532,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.8192319898444937e-05,
-      "loss": 1.3602,
-      "step": 1584
-    },
-    {
-      "epoch": 0.4981840312542945,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.8189781021897812e-05,
-      "loss": 1.2892,
-      "step": 1586
-    },
-    {
-      "epoch": 0.4988122582798359,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.8187242145350684e-05,
-      "loss": 1.2329,
-      "step": 1588
-    },
-    {
-      "epoch": 0.4994404853053772,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.8184703268803555e-05,
-      "loss": 1.2801,
-      "step": 1590
-    },
-    {
-      "epoch": 0.5000687123309185,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.8182164392256426e-05,
-      "loss": 1.3857,
-      "step": 1592
-    },
-    {
-      "epoch": 0.50069693935646,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.81796255157093e-05,
-      "loss": 1.3113,
-      "step": 1594
-    },
-    {
-      "epoch": 0.5013251663820013,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.8177086639162172e-05,
-      "loss": 1.4523,
-      "step": 1596
-    },
-    {
-      "epoch": 0.5019533934075426,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.8174547762615044e-05,
-      "loss": 1.3314,
-      "step": 1598
-    },
-    {
-      "epoch": 0.502581620433084,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.8172008886067915e-05,
-      "loss": 1.3618,
-      "step": 1600
-    },
-    {
-      "epoch": 0.5032098474586254,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.816947000952079e-05,
-      "loss": 1.3553,
-      "step": 1602
-    },
-    {
-      "epoch": 0.5038380744841667,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.816693113297366e-05,
-      "loss": 1.4201,
-      "step": 1604
-    },
-    {
-      "epoch": 0.5044663015097081,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.8164392256426533e-05,
-      "loss": 1.309,
-      "step": 1606
-    },
-    {
-      "epoch": 0.5050945285352494,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.8161853379879404e-05,
-      "loss": 1.3145,
-      "step": 1608
-    },
-    {
-      "epoch": 0.5057227555607908,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.8159314503332275e-05,
-      "loss": 1.3546,
-      "step": 1610
-    },
-    {
-      "epoch": 0.5063509825863322,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.8156775626785147e-05,
-      "loss": 1.2818,
-      "step": 1612
-    },
-    {
-      "epoch": 0.5069792096118735,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.815423675023802e-05,
-      "loss": 1.4176,
-      "step": 1614
-    },
-    {
-      "epoch": 0.5076074366374148,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.8151697873690893e-05,
-      "loss": 1.3501,
-      "step": 1616
-    },
-    {
-      "epoch": 0.5082356636629562,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.8149158997143764e-05,
-      "loss": 1.3265,
-      "step": 1618
-    },
-    {
-      "epoch": 0.5088638906884976,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.814662012059664e-05,
-      "loss": 1.333,
-      "step": 1620
-    },
-    {
-      "epoch": 0.5094921177140389,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.814408124404951e-05,
-      "loss": 1.2086,
-      "step": 1622
-    },
-    {
-      "epoch": 0.5101203447395802,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.8141542367502382e-05,
-      "loss": 1.2181,
-      "step": 1624
-    },
-    {
-      "epoch": 0.5107485717651217,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8139003490955253e-05,
-      "loss": 1.3269,
-      "step": 1626
-    },
-    {
-      "epoch": 0.511376798790663,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.8136464614408128e-05,
-      "loss": 1.2733,
-      "step": 1628
-    },
-    {
-      "epoch": 0.5120050258162043,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.8133925737861e-05,
-      "loss": 1.253,
-      "step": 1630
-    },
-    {
-      "epoch": 0.5126332528417457,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.813138686131387e-05,
-      "loss": 1.511,
-      "step": 1632
-    },
-    {
-      "epoch": 0.5132614798672871,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.8128847984766742e-05,
-      "loss": 1.2451,
-      "step": 1634
-    },
-    {
-      "epoch": 0.5138897068928284,
-      "grad_norm": 0.66015625,
-      "learning_rate": 1.8126309108219614e-05,
-      "loss": 1.2587,
-      "step": 1636
-    },
-    {
-      "epoch": 0.5145179339183698,
-      "grad_norm": 0.875,
-      "learning_rate": 1.8123770231672485e-05,
-      "loss": 1.301,
-      "step": 1638
-    },
-    {
-      "epoch": 0.5151461609439111,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.812123135512536e-05,
-      "loss": 1.4174,
-      "step": 1640
-    },
-    {
-      "epoch": 0.5157743879694524,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.811869247857823e-05,
-      "loss": 1.2725,
-      "step": 1642
-    },
-    {
-      "epoch": 0.5164026149949938,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.8116153602031103e-05,
-      "loss": 1.3744,
-      "step": 1644
-    },
-    {
-      "epoch": 0.5170308420205352,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.8113614725483974e-05,
-      "loss": 1.2455,
-      "step": 1646
-    },
-    {
-      "epoch": 0.5176590690460765,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.811107584893685e-05,
-      "loss": 1.4318,
-      "step": 1648
-    },
-    {
-      "epoch": 0.5182872960716178,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.810853697238972e-05,
-      "loss": 1.3426,
-      "step": 1650
-    },
-    {
-      "epoch": 0.5189155230971593,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.810599809584259e-05,
-      "loss": 1.1767,
-      "step": 1652
-    },
-    {
-      "epoch": 0.5195437501227006,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.8103459219295463e-05,
-      "loss": 1.2447,
-      "step": 1654
-    },
-    {
-      "epoch": 0.5201719771482419,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.8100920342748338e-05,
-      "loss": 1.3263,
-      "step": 1656
-    },
-    {
-      "epoch": 0.5208002041737833,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.8098381466201206e-05,
-      "loss": 1.3857,
-      "step": 1658
-    },
-    {
-      "epoch": 0.5214284311993247,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.809584258965408e-05,
-      "loss": 1.4577,
-      "step": 1660
-    },
-    {
-      "epoch": 0.522056658224866,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.8093303713106952e-05,
-      "loss": 1.4192,
-      "step": 1662
-    },
-    {
-      "epoch": 0.5226848852504073,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.8090764836559823e-05,
-      "loss": 1.2375,
-      "step": 1664
-    },
-    {
-      "epoch": 0.5233131122759487,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.8088225960012695e-05,
-      "loss": 1.3199,
-      "step": 1666
-    },
-    {
-      "epoch": 0.5239413393014901,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.808568708346557e-05,
-      "loss": 1.3239,
-      "step": 1668
-    },
-    {
-      "epoch": 0.5245695663270314,
-      "grad_norm": 0.63671875,
-      "learning_rate": 1.808314820691844e-05,
-      "loss": 1.4124,
-      "step": 1670
-    },
-    {
-      "epoch": 0.5251977933525728,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.8080609330371312e-05,
-      "loss": 1.3539,
-      "step": 1672
-    },
-    {
-      "epoch": 0.5258260203781141,
-      "grad_norm": 0.75,
-      "learning_rate": 1.8078070453824183e-05,
-      "loss": 1.2979,
-      "step": 1674
-    },
-    {
-      "epoch": 0.5264542474036555,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.8075531577277058e-05,
-      "loss": 1.3582,
-      "step": 1676
-    },
-    {
-      "epoch": 0.5270824744291969,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.8072992700729926e-05,
-      "loss": 1.2866,
-      "step": 1678
-    },
-    {
-      "epoch": 0.5277107014547382,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.80704538241828e-05,
-      "loss": 1.1929,
-      "step": 1680
-    },
-    {
-      "epoch": 0.5283389284802795,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.8067914947635672e-05,
-      "loss": 1.2081,
-      "step": 1682
-    },
-    {
-      "epoch": 0.528967155505821,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.8065376071088544e-05,
-      "loss": 1.4058,
-      "step": 1684
-    },
-    {
-      "epoch": 0.5295953825313623,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.8062837194541415e-05,
-      "loss": 1.3689,
-      "step": 1686
-    },
-    {
-      "epoch": 0.5302236095569036,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.806029831799429e-05,
-      "loss": 1.2963,
-      "step": 1688
-    },
-    {
-      "epoch": 0.530851836582445,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.805775944144716e-05,
-      "loss": 1.3622,
-      "step": 1690
-    },
-    {
-      "epoch": 0.5314800636079864,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.8055220564900033e-05,
-      "loss": 1.2601,
-      "step": 1692
-    },
-    {
-      "epoch": 0.5321082906335277,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.8052681688352904e-05,
-      "loss": 1.2963,
-      "step": 1694
-    },
-    {
-      "epoch": 0.532736517659069,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.805014281180578e-05,
-      "loss": 1.3702,
-      "step": 1696
-    },
-    {
-      "epoch": 0.5333647446846104,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.804760393525865e-05,
-      "loss": 1.2802,
-      "step": 1698
-    },
-    {
-      "epoch": 0.5339929717101518,
-      "grad_norm": 1.0078125,
-      "learning_rate": 1.804506505871152e-05,
-      "loss": 1.2703,
-      "step": 1700
-    },
-    {
-      "epoch": 0.5346211987356931,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.8042526182164393e-05,
-      "loss": 1.2858,
-      "step": 1702
-    },
-    {
-      "epoch": 0.5352494257612345,
-      "grad_norm": 1.03125,
-      "learning_rate": 1.8039987305617264e-05,
-      "loss": 1.2983,
-      "step": 1704
-    },
-    {
-      "epoch": 0.5358776527867758,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.803744842907014e-05,
-      "loss": 1.4192,
-      "step": 1706
-    },
-    {
-      "epoch": 0.5365058798123171,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.803490955252301e-05,
-      "loss": 1.4011,
-      "step": 1708
-    },
-    {
-      "epoch": 0.5371341068378586,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.8032370675975882e-05,
-      "loss": 1.3894,
-      "step": 1710
-    },
-    {
-      "epoch": 0.5377623338633999,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8029831799428753e-05,
-      "loss": 1.5168,
-      "step": 1712
-    },
-    {
-      "epoch": 0.5383905608889412,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.8027292922881628e-05,
-      "loss": 1.3708,
-      "step": 1714
-    },
-    {
-      "epoch": 0.5390187879144825,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.80247540463345e-05,
-      "loss": 1.3817,
-      "step": 1716
-    },
-    {
-      "epoch": 0.539647014940024,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.802221516978737e-05,
-      "loss": 1.3174,
-      "step": 1718
-    },
-    {
-      "epoch": 0.5402752419655653,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.8019676293240242e-05,
-      "loss": 1.3609,
-      "step": 1720
-    },
-    {
-      "epoch": 0.5409034689911066,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.8017137416693117e-05,
-      "loss": 1.4835,
-      "step": 1722
-    },
-    {
-      "epoch": 0.541531696016648,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.801459854014599e-05,
-      "loss": 1.5052,
-      "step": 1724
-    },
-    {
-      "epoch": 0.5421599230421894,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.801205966359886e-05,
-      "loss": 1.3482,
-      "step": 1726
-    },
-    {
-      "epoch": 0.5427881500677307,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.800952078705173e-05,
-      "loss": 1.21,
-      "step": 1728
-    },
-    {
-      "epoch": 0.5434163770932721,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.8006981910504602e-05,
-      "loss": 1.3702,
-      "step": 1730
-    },
-    {
-      "epoch": 0.5440446041188134,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.8004443033957474e-05,
-      "loss": 1.4266,
-      "step": 1732
-    },
-    {
-      "epoch": 0.5446728311443548,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.800190415741035e-05,
-      "loss": 1.339,
-      "step": 1734
-    },
-    {
-      "epoch": 0.5453010581698962,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.799936528086322e-05,
-      "loss": 1.3851,
-      "step": 1736
-    },
-    {
-      "epoch": 0.5459292851954375,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.799682640431609e-05,
-      "loss": 1.4017,
-      "step": 1738
-    },
-    {
-      "epoch": 0.5465575122209788,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.7994287527768963e-05,
-      "loss": 1.2933,
-      "step": 1740
-    },
-    {
-      "epoch": 0.5471857392465203,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.7991748651221838e-05,
-      "loss": 1.3104,
-      "step": 1742
-    },
-    {
-      "epoch": 0.5478139662720616,
-      "grad_norm": 0.65625,
-      "learning_rate": 1.798920977467471e-05,
-      "loss": 1.231,
-      "step": 1744
-    },
-    {
-      "epoch": 0.5484421932976029,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.798667089812758e-05,
-      "loss": 1.4584,
-      "step": 1746
-    },
-    {
-      "epoch": 0.5490704203231442,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.798413202158045e-05,
-      "loss": 1.2988,
-      "step": 1748
-    },
-    {
-      "epoch": 0.5496986473486857,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.7981593145033326e-05,
-      "loss": 1.2553,
-      "step": 1750
-    },
-    {
-      "epoch": 0.550326874374227,
-      "grad_norm": 0.75,
-      "learning_rate": 1.7979054268486194e-05,
-      "loss": 1.3824,
-      "step": 1752
-    },
-    {
-      "epoch": 0.5509551013997683,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.797651539193907e-05,
-      "loss": 1.4831,
-      "step": 1754
-    },
-    {
-      "epoch": 0.5515833284253097,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.797397651539194e-05,
-      "loss": 1.3839,
-      "step": 1756
-    },
-    {
-      "epoch": 0.5522115554508511,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.7971437638844812e-05,
-      "loss": 1.4556,
-      "step": 1758
-    },
-    {
-      "epoch": 0.5528397824763924,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.7968898762297683e-05,
-      "loss": 1.3564,
-      "step": 1760
-    },
-    {
-      "epoch": 0.5534680095019338,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.7966359885750558e-05,
-      "loss": 1.4027,
-      "step": 1762
-    },
-    {
-      "epoch": 0.5540962365274751,
-      "grad_norm": 0.93359375,
-      "learning_rate": 1.796382100920343e-05,
-      "loss": 1.3738,
-      "step": 1764
-    },
-    {
-      "epoch": 0.5547244635530165,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.79612821326563e-05,
-      "loss": 1.4116,
-      "step": 1766
-    },
-    {
-      "epoch": 0.5553526905785579,
-      "grad_norm": 0.9140625,
-      "learning_rate": 1.7958743256109172e-05,
-      "loss": 1.4383,
-      "step": 1768
-    },
-    {
-      "epoch": 0.5559809176040992,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.7956204379562047e-05,
-      "loss": 1.2174,
-      "step": 1770
-    },
-    {
-      "epoch": 0.5566091446296405,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.7953665503014915e-05,
-      "loss": 1.2893,
-      "step": 1772
-    },
-    {
-      "epoch": 0.557237371655182,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.795112662646779e-05,
-      "loss": 1.291,
-      "step": 1774
-    },
-    {
-      "epoch": 0.5578655986807233,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.794858774992066e-05,
-      "loss": 1.4798,
-      "step": 1776
-    },
-    {
-      "epoch": 0.5584938257062646,
-      "grad_norm": 0.9296875,
-      "learning_rate": 1.7946048873373533e-05,
-      "loss": 1.2961,
-      "step": 1778
-    },
-    {
-      "epoch": 0.5591220527318059,
-      "grad_norm": 2.1875,
-      "learning_rate": 1.7943509996826404e-05,
-      "loss": 1.3342,
-      "step": 1780
-    },
-    {
-      "epoch": 0.5597502797573473,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.794097112027928e-05,
-      "loss": 1.2582,
-      "step": 1782
-    },
-    {
-      "epoch": 0.5603785067828887,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.793843224373215e-05,
-      "loss": 1.3106,
-      "step": 1784
-    },
-    {
-      "epoch": 0.56100673380843,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.793589336718502e-05,
-      "loss": 1.3369,
-      "step": 1786
-    },
-    {
-      "epoch": 0.5616349608339714,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.7933354490637893e-05,
-      "loss": 1.2346,
-      "step": 1788
-    },
-    {
-      "epoch": 0.5622631878595127,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.7930815614090768e-05,
-      "loss": 1.2644,
-      "step": 1790
-    },
-    {
-      "epoch": 0.5628914148850541,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.792827673754364e-05,
-      "loss": 1.3247,
-      "step": 1792
-    },
-    {
-      "epoch": 0.5635196419105954,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.792573786099651e-05,
-      "loss": 1.2764,
-      "step": 1794
-    },
-    {
-      "epoch": 0.5641478689361368,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7923198984449385e-05,
-      "loss": 1.2428,
-      "step": 1796
-    },
-    {
-      "epoch": 0.5647760959616781,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.7920660107902253e-05,
-      "loss": 1.4744,
-      "step": 1798
-    },
-    {
-      "epoch": 0.5654043229872195,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.7918121231355128e-05,
-      "loss": 1.3754,
-      "step": 1800
-    },
-    {
-      "epoch": 0.5660325500127609,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.7915582354808e-05,
-      "loss": 1.3084,
-      "step": 1802
-    },
-    {
-      "epoch": 0.5666607770383022,
-      "grad_norm": 0.75,
-      "learning_rate": 1.791304347826087e-05,
-      "loss": 1.3269,
-      "step": 1804
-    },
-    {
-      "epoch": 0.5672890040638435,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7910504601713742e-05,
-      "loss": 1.3141,
-      "step": 1806
-    },
-    {
-      "epoch": 0.567917231089385,
-      "grad_norm": 0.90234375,
-      "learning_rate": 1.7907965725166617e-05,
-      "loss": 1.1482,
-      "step": 1808
-    },
-    {
-      "epoch": 0.5685454581149263,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.7905426848619488e-05,
-      "loss": 1.3093,
-      "step": 1810
-    },
-    {
-      "epoch": 0.5691736851404676,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.790288797207236e-05,
-      "loss": 1.4742,
-      "step": 1812
-    },
-    {
-      "epoch": 0.569801912166009,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.790034909552523e-05,
-      "loss": 1.3429,
-      "step": 1814
-    },
-    {
-      "epoch": 0.5704301391915504,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7897810218978106e-05,
-      "loss": 1.3247,
-      "step": 1816
-    },
-    {
-      "epoch": 0.5710583662170917,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.7895271342430974e-05,
-      "loss": 1.386,
-      "step": 1818
-    },
-    {
-      "epoch": 0.571686593242633,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.789273246588385e-05,
-      "loss": 1.3501,
-      "step": 1820
-    },
-    {
-      "epoch": 0.5723148202681744,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.789019358933672e-05,
-      "loss": 1.2759,
-      "step": 1822
-    },
-    {
-      "epoch": 0.5729430472937158,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.788765471278959e-05,
-      "loss": 1.2834,
-      "step": 1824
-    },
-    {
-      "epoch": 0.5735712743192571,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.7885115836242463e-05,
-      "loss": 1.3764,
-      "step": 1826
-    },
-    {
-      "epoch": 0.5741995013447985,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.7882576959695337e-05,
-      "loss": 1.2428,
-      "step": 1828
-    },
-    {
-      "epoch": 0.5748277283703398,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.788003808314821e-05,
-      "loss": 1.3577,
-      "step": 1830
-    },
-    {
-      "epoch": 0.5754559553958812,
-      "grad_norm": 0.94921875,
-      "learning_rate": 1.787749920660108e-05,
-      "loss": 1.2091,
-      "step": 1832
-    },
-    {
-      "epoch": 0.5760841824214226,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.787496033005395e-05,
-      "loss": 1.4,
-      "step": 1834
-    },
-    {
-      "epoch": 0.5767124094469639,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.7872421453506826e-05,
-      "loss": 1.3621,
-      "step": 1836
-    },
-    {
-      "epoch": 0.5773406364725052,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.7869882576959698e-05,
-      "loss": 1.3756,
-      "step": 1838
-    },
-    {
-      "epoch": 0.5779688634980467,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.786734370041257e-05,
-      "loss": 1.3658,
-      "step": 1840
-    },
-    {
-      "epoch": 0.578597090523588,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.786480482386544e-05,
-      "loss": 1.2812,
-      "step": 1842
-    },
-    {
-      "epoch": 0.5792253175491293,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7862265947318312e-05,
-      "loss": 1.4921,
-      "step": 1844
-    },
-    {
-      "epoch": 0.5798535445746706,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7859727070771183e-05,
-      "loss": 1.3042,
-      "step": 1846
-    },
-    {
-      "epoch": 0.5804817716002121,
-      "grad_norm": 1.203125,
-      "learning_rate": 1.7857188194224058e-05,
-      "loss": 1.1429,
-      "step": 1848
-    },
-    {
-      "epoch": 0.5811099986257534,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.785464931767693e-05,
-      "loss": 1.4471,
-      "step": 1850
-    },
-    {
-      "epoch": 0.5817382256512947,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.78521104411298e-05,
-      "loss": 1.3808,
-      "step": 1852
-    },
-    {
-      "epoch": 0.5823664526768361,
-      "grad_norm": 0.94140625,
-      "learning_rate": 1.7849571564582672e-05,
-      "loss": 1.3266,
-      "step": 1854
-    },
-    {
-      "epoch": 0.5829946797023774,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.7847032688035547e-05,
-      "loss": 1.4399,
-      "step": 1856
-    },
-    {
-      "epoch": 0.5836229067279188,
-      "grad_norm": 0.75,
-      "learning_rate": 1.784449381148842e-05,
-      "loss": 1.2884,
-      "step": 1858
-    },
-    {
-      "epoch": 0.5842511337534602,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.784195493494129e-05,
-      "loss": 1.3308,
-      "step": 1860
-    },
-    {
-      "epoch": 0.5848793607790015,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.783941605839416e-05,
-      "loss": 1.3215,
-      "step": 1862
-    },
-    {
-      "epoch": 0.5855075878045428,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.7836877181847036e-05,
-      "loss": 1.4684,
-      "step": 1864
-    },
-    {
-      "epoch": 0.5861358148300843,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.7834338305299904e-05,
-      "loss": 1.293,
-      "step": 1866
-    },
-    {
-      "epoch": 0.5867640418556256,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.783179942875278e-05,
-      "loss": 1.2667,
-      "step": 1868
-    },
-    {
-      "epoch": 0.5873922688811669,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.782926055220565e-05,
-      "loss": 1.3243,
-      "step": 1870
-    },
-    {
-      "epoch": 0.5880204959067082,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.782672167565852e-05,
-      "loss": 1.2651,
-      "step": 1872
-    },
-    {
-      "epoch": 0.5886487229322497,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.7824182799111393e-05,
-      "loss": 1.2973,
-      "step": 1874
-    },
-    {
-      "epoch": 0.589276949957791,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7821643922564268e-05,
-      "loss": 1.2823,
-      "step": 1876
-    },
-    {
-      "epoch": 0.5899051769833323,
-      "grad_norm": 0.94921875,
-      "learning_rate": 1.781910504601714e-05,
-      "loss": 1.415,
-      "step": 1878
-    },
-    {
-      "epoch": 0.5905334040088737,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.781656616947001e-05,
-      "loss": 1.2477,
-      "step": 1880
-    },
-    {
-      "epoch": 0.5911616310344151,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.7814027292922885e-05,
-      "loss": 1.2649,
-      "step": 1882
-    },
-    {
-      "epoch": 0.5917898580599564,
-      "grad_norm": 0.64453125,
-      "learning_rate": 1.7811488416375756e-05,
-      "loss": 1.3797,
-      "step": 1884
-    },
-    {
-      "epoch": 0.5924180850854978,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.7808949539828628e-05,
-      "loss": 1.3717,
-      "step": 1886
-    },
-    {
-      "epoch": 0.5930463121110391,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.78064106632815e-05,
-      "loss": 1.2677,
-      "step": 1888
-    },
-    {
-      "epoch": 0.5936745391365805,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7803871786734374e-05,
-      "loss": 1.4157,
-      "step": 1890
-    },
-    {
-      "epoch": 0.5943027661621219,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.7801332910187242e-05,
-      "loss": 1.2478,
-      "step": 1892
-    },
-    {
-      "epoch": 0.5949309931876632,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.7798794033640117e-05,
-      "loss": 1.3108,
-      "step": 1894
-    },
-    {
-      "epoch": 0.5955592202132045,
-      "grad_norm": 0.75,
-      "learning_rate": 1.7796255157092988e-05,
-      "loss": 1.3043,
-      "step": 1896
-    },
-    {
-      "epoch": 0.596187447238746,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.779371628054586e-05,
-      "loss": 1.259,
-      "step": 1898
-    },
-    {
-      "epoch": 0.5968156742642873,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.779117740399873e-05,
-      "loss": 1.2671,
-      "step": 1900
-    },
-    {
-      "epoch": 0.5974439012898286,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.7788638527451606e-05,
-      "loss": 1.3808,
-      "step": 1902
-    },
-    {
-      "epoch": 0.5980721283153699,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.7786099650904477e-05,
-      "loss": 1.3359,
-      "step": 1904
-    },
-    {
-      "epoch": 0.5987003553409114,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.778356077435735e-05,
-      "loss": 1.3205,
-      "step": 1906
-    },
-    {
-      "epoch": 0.5993285823664527,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.778102189781022e-05,
-      "loss": 1.3357,
-      "step": 1908
-    },
-    {
-      "epoch": 0.599956809391994,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.7778483021263095e-05,
-      "loss": 1.3952,
-      "step": 1910
-    },
-    {
-      "epoch": 0.6005850364175354,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.7775944144715963e-05,
-      "loss": 1.2332,
-      "step": 1912
-    },
-    {
-      "epoch": 0.6012132634430768,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.7773405268168837e-05,
-      "loss": 1.3406,
-      "step": 1914
-    },
-    {
-      "epoch": 0.6018414904686181,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.777086639162171e-05,
-      "loss": 1.3423,
-      "step": 1916
-    },
-    {
-      "epoch": 0.6024697174941595,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.776832751507458e-05,
-      "loss": 1.3578,
-      "step": 1918
-    },
-    {
-      "epoch": 0.6030979445197008,
-      "grad_norm": 0.65625,
-      "learning_rate": 1.776578863852745e-05,
-      "loss": 1.3513,
-      "step": 1920
-    },
-    {
-      "epoch": 0.6037261715452421,
-      "grad_norm": 0.8203125,
-      "learning_rate": 1.7763249761980326e-05,
-      "loss": 1.2171,
-      "step": 1922
-    },
-    {
-      "epoch": 0.6043543985707835,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.7760710885433198e-05,
-      "loss": 1.3335,
-      "step": 1924
-    },
-    {
-      "epoch": 0.6049826255963249,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.775817200888607e-05,
-      "loss": 1.3946,
-      "step": 1926
-    },
-    {
-      "epoch": 0.6056108526218662,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.775563313233894e-05,
-      "loss": 1.2521,
-      "step": 1928
-    },
-    {
-      "epoch": 0.6062390796474075,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.7753094255791815e-05,
-      "loss": 1.4663,
-      "step": 1930
-    },
-    {
-      "epoch": 0.606867306672949,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7750555379244687e-05,
-      "loss": 1.1201,
-      "step": 1932
-    },
-    {
-      "epoch": 0.6074955336984903,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7748016502697558e-05,
-      "loss": 1.4028,
-      "step": 1934
-    },
-    {
-      "epoch": 0.6081237607240316,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.774547762615043e-05,
-      "loss": 1.2642,
-      "step": 1936
-    },
-    {
-      "epoch": 0.608751987749573,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.77429387496033e-05,
-      "loss": 1.2945,
-      "step": 1938
-    },
-    {
-      "epoch": 0.6093802147751144,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.7740399873056172e-05,
-      "loss": 1.265,
-      "step": 1940
-    },
-    {
-      "epoch": 0.6100084418006557,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.7737860996509047e-05,
-      "loss": 1.4148,
-      "step": 1942
-    },
-    {
-      "epoch": 0.610636668826197,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.7735322119961918e-05,
-      "loss": 1.2506,
-      "step": 1944
-    },
-    {
-      "epoch": 0.6112648958517384,
-      "grad_norm": 0.90234375,
-      "learning_rate": 1.773278324341479e-05,
-      "loss": 1.3281,
-      "step": 1946
-    },
-    {
-      "epoch": 0.6118931228772798,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.773024436686766e-05,
-      "loss": 1.273,
-      "step": 1948
-    },
-    {
-      "epoch": 0.6125213499028211,
-      "grad_norm": 0.75,
-      "learning_rate": 1.7727705490320536e-05,
-      "loss": 1.2533,
-      "step": 1950
-    },
-    {
-      "epoch": 0.6131495769283625,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7725166613773407e-05,
-      "loss": 1.2262,
-      "step": 1952
-    },
-    {
-      "epoch": 0.6137778039539038,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.772262773722628e-05,
-      "loss": 1.2834,
-      "step": 1954
-    },
-    {
-      "epoch": 0.6144060309794452,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.772008886067915e-05,
-      "loss": 1.286,
-      "step": 1956
-    },
-    {
-      "epoch": 0.6150342580049866,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7717549984132025e-05,
-      "loss": 1.2474,
-      "step": 1958
-    },
-    {
-      "epoch": 0.6156624850305279,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7715011107584893e-05,
-      "loss": 1.345,
-      "step": 1960
-    },
-    {
-      "epoch": 0.6162907120560692,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.7712472231037767e-05,
-      "loss": 1.2253,
-      "step": 1962
-    },
-    {
-      "epoch": 0.6169189390816107,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.770993335449064e-05,
-      "loss": 1.3628,
-      "step": 1964
-    },
-    {
-      "epoch": 0.617547166107152,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.770739447794351e-05,
-      "loss": 1.2676,
-      "step": 1966
-    },
-    {
-      "epoch": 0.6181753931326933,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.7704855601396385e-05,
-      "loss": 1.2463,
-      "step": 1968
-    },
-    {
-      "epoch": 0.6188036201582346,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.7702316724849256e-05,
-      "loss": 1.3617,
-      "step": 1970
-    },
-    {
-      "epoch": 0.6194318471837761,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.7699777848302128e-05,
-      "loss": 1.4785,
-      "step": 1972
-    },
-    {
-      "epoch": 0.6200600742093174,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.7697238971755e-05,
-      "loss": 1.2933,
-      "step": 1974
-    },
-    {
-      "epoch": 0.6206883012348587,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.7694700095207874e-05,
-      "loss": 1.4211,
-      "step": 1976
-    },
-    {
-      "epoch": 0.6213165282604001,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.7692161218660745e-05,
-      "loss": 1.4411,
-      "step": 1978
-    },
-    {
-      "epoch": 0.6219447552859415,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.7689622342113617e-05,
-      "loss": 1.2598,
-      "step": 1980
-    },
-    {
-      "epoch": 0.6225729823114828,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7687083465566488e-05,
-      "loss": 1.2098,
-      "step": 1982
-    },
-    {
-      "epoch": 0.6232012093370242,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.7684544589019363e-05,
-      "loss": 1.3236,
-      "step": 1984
-    },
-    {
-      "epoch": 0.6238294363625655,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.768200571247223e-05,
-      "loss": 1.3541,
-      "step": 1986
-    },
-    {
-      "epoch": 0.6244576633881069,
-      "grad_norm": 0.84765625,
-      "learning_rate": 1.7679466835925106e-05,
-      "loss": 1.2746,
-      "step": 1988
-    },
-    {
-      "epoch": 0.6250858904136483,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.7676927959377977e-05,
-      "loss": 1.3703,
-      "step": 1990
-    },
-    {
-      "epoch": 0.6257141174391896,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.767438908283085e-05,
-      "loss": 1.2673,
-      "step": 1992
-    },
-    {
-      "epoch": 0.6263423444647309,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.767185020628372e-05,
-      "loss": 1.2734,
-      "step": 1994
-    },
-    {
-      "epoch": 0.6269705714902722,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.7669311329736595e-05,
-      "loss": 1.2994,
-      "step": 1996
-    },
-    {
-      "epoch": 0.6275987985158137,
-      "grad_norm": 0.84765625,
-      "learning_rate": 1.7666772453189466e-05,
-      "loss": 1.2314,
-      "step": 1998
-    },
-    {
-      "epoch": 0.628227025541355,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7664233576642337e-05,
-      "loss": 1.3692,
-      "step": 2000
-    },
-    {
-      "epoch": 0.6288552525668963,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.766169470009521e-05,
-      "loss": 1.1083,
-      "step": 2002
-    },
-    {
-      "epoch": 0.6294834795924377,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7659155823548083e-05,
-      "loss": 1.3513,
-      "step": 2004
-    },
-    {
-      "epoch": 0.6301117066179791,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.765661694700095e-05,
-      "loss": 1.2768,
-      "step": 2006
-    },
-    {
-      "epoch": 0.6307399336435204,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7654078070453826e-05,
-      "loss": 1.399,
-      "step": 2008
-    },
-    {
-      "epoch": 0.6313681606690618,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.7651539193906698e-05,
-      "loss": 1.3596,
-      "step": 2010
-    },
-    {
-      "epoch": 0.6319963876946031,
-      "grad_norm": 0.99609375,
-      "learning_rate": 1.764900031735957e-05,
-      "loss": 1.3298,
-      "step": 2012
-    },
-    {
-      "epoch": 0.6326246147201445,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.764646144081244e-05,
-      "loss": 1.3194,
-      "step": 2014
-    },
-    {
-      "epoch": 0.6332528417456859,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.7643922564265315e-05,
-      "loss": 1.2478,
-      "step": 2016
-    },
-    {
-      "epoch": 0.6338810687712272,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.7641383687718187e-05,
-      "loss": 1.285,
-      "step": 2018
-    },
-    {
-      "epoch": 0.6345092957967685,
-      "grad_norm": 0.75,
-      "learning_rate": 1.7638844811171058e-05,
-      "loss": 1.4251,
-      "step": 2020
-    },
-    {
-      "epoch": 0.63513752282231,
-      "grad_norm": 0.97265625,
-      "learning_rate": 1.763630593462393e-05,
-      "loss": 1.281,
-      "step": 2022
-    },
-    {
-      "epoch": 0.6357657498478513,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.7633767058076804e-05,
-      "loss": 1.3546,
-      "step": 2024
-    },
-    {
-      "epoch": 0.6363939768733926,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.7631228181529672e-05,
-      "loss": 1.2134,
-      "step": 2026
-    },
-    {
-      "epoch": 0.6370222038989339,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.7628689304982547e-05,
-      "loss": 1.352,
-      "step": 2028
-    },
-    {
-      "epoch": 0.6376504309244754,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.7626150428435418e-05,
-      "loss": 1.3923,
-      "step": 2030
-    },
-    {
-      "epoch": 0.6382786579500167,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.762361155188829e-05,
-      "loss": 1.3658,
-      "step": 2032
-    },
-    {
-      "epoch": 0.638906884975558,
-      "grad_norm": 0.96875,
-      "learning_rate": 1.762107267534116e-05,
-      "loss": 1.2421,
-      "step": 2034
-    },
-    {
-      "epoch": 0.6395351120010994,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7618533798794036e-05,
-      "loss": 1.3427,
-      "step": 2036
-    },
-    {
-      "epoch": 0.6401633390266408,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.7615994922246907e-05,
-      "loss": 1.3342,
-      "step": 2038
-    },
-    {
-      "epoch": 0.6407915660521821,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.761345604569978e-05,
-      "loss": 1.4221,
-      "step": 2040
-    },
-    {
-      "epoch": 0.6414197930777235,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.761091716915265e-05,
-      "loss": 1.3898,
-      "step": 2042
-    },
-    {
-      "epoch": 0.6420480201032648,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.7608378292605525e-05,
-      "loss": 1.5576,
-      "step": 2044
-    },
-    {
-      "epoch": 0.6426762471288062,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.7605839416058396e-05,
-      "loss": 1.3117,
-      "step": 2046
-    },
-    {
-      "epoch": 0.6433044741543475,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.7603300539511267e-05,
-      "loss": 1.2932,
-      "step": 2048
-    },
-    {
-      "epoch": 0.6439327011798889,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.7600761662964142e-05,
-      "loss": 1.2463,
-      "step": 2050
-    },
-    {
-      "epoch": 0.6445609282054302,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.759822278641701e-05,
-      "loss": 1.3657,
-      "step": 2052
-    },
-    {
-      "epoch": 0.6451891552309716,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.7595683909869885e-05,
-      "loss": 1.4386,
-      "step": 2054
-    },
-    {
-      "epoch": 0.645817382256513,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.7593145033322756e-05,
-      "loss": 1.3022,
-      "step": 2056
-    },
-    {
-      "epoch": 0.6464456092820543,
-      "grad_norm": 1.0546875,
-      "learning_rate": 1.7590606156775628e-05,
-      "loss": 1.5185,
-      "step": 2058
-    },
-    {
-      "epoch": 0.6470738363075956,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.75880672802285e-05,
-      "loss": 1.1322,
-      "step": 2060
-    },
-    {
-      "epoch": 0.647702063333137,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.7585528403681374e-05,
-      "loss": 1.3653,
-      "step": 2062
-    },
-    {
-      "epoch": 0.6483302903586784,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.7582989527134245e-05,
-      "loss": 1.2649,
-      "step": 2064
-    },
-    {
-      "epoch": 0.6489585173842197,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.7580450650587117e-05,
-      "loss": 1.4218,
-      "step": 2066
-    },
-    {
-      "epoch": 0.649586744409761,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.7577911774039988e-05,
-      "loss": 1.42,
-      "step": 2068
-    },
-    {
-      "epoch": 0.6502149714353024,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.7575372897492863e-05,
-      "loss": 1.2207,
-      "step": 2070
-    },
-    {
-      "epoch": 0.6508431984608438,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.7572834020945734e-05,
-      "loss": 1.3358,
-      "step": 2072
-    },
-    {
-      "epoch": 0.6514714254863851,
-      "grad_norm": 0.921875,
-      "learning_rate": 1.7570295144398606e-05,
-      "loss": 1.2739,
-      "step": 2074
-    },
-    {
-      "epoch": 0.6520996525119265,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.7567756267851477e-05,
-      "loss": 1.3296,
-      "step": 2076
-    },
-    {
-      "epoch": 0.6527278795374678,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.756521739130435e-05,
-      "loss": 1.3499,
-      "step": 2078
-    },
-    {
-      "epoch": 0.6533561065630092,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.756267851475722e-05,
-      "loss": 1.3431,
-      "step": 2080
-    },
-    {
-      "epoch": 0.6539843335885506,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.7560139638210094e-05,
-      "loss": 1.4784,
-      "step": 2082
-    },
-    {
-      "epoch": 0.6546125606140919,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7557600761662966e-05,
-      "loss": 1.2925,
-      "step": 2084
-    },
-    {
-      "epoch": 0.6552407876396332,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.7555061885115837e-05,
-      "loss": 1.2877,
-      "step": 2086
-    },
-    {
-      "epoch": 0.6558690146651747,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.755252300856871e-05,
-      "loss": 1.3971,
-      "step": 2088
-    },
-    {
-      "epoch": 0.656497241690716,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.7549984132021583e-05,
-      "loss": 1.2842,
-      "step": 2090
-    },
-    {
-      "epoch": 0.6571254687162573,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.7547445255474455e-05,
-      "loss": 1.1632,
-      "step": 2092
-    },
-    {
-      "epoch": 0.6577536957417986,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.7544906378927326e-05,
-      "loss": 1.3903,
-      "step": 2094
-    },
-    {
-      "epoch": 0.6583819227673401,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.7542367502380198e-05,
-      "loss": 1.4009,
-      "step": 2096
-    },
-    {
-      "epoch": 0.6590101497928814,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.7539828625833072e-05,
-      "loss": 1.353,
-      "step": 2098
-    },
-    {
-      "epoch": 0.6596383768184227,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.753728974928594e-05,
-      "loss": 1.4069,
-      "step": 2100
-    },
-    {
-      "epoch": 0.6602666038439641,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.7534750872738815e-05,
-      "loss": 1.2694,
-      "step": 2102
-    },
-    {
-      "epoch": 0.6608948308695055,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.7532211996191686e-05,
-      "loss": 1.2923,
-      "step": 2104
-    },
-    {
-      "epoch": 0.6615230578950468,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.7529673119644558e-05,
-      "loss": 1.2756,
-      "step": 2106
-    },
-    {
-      "epoch": 0.6621512849205882,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.752713424309743e-05,
-      "loss": 1.4151,
-      "step": 2108
-    },
-    {
-      "epoch": 0.6627795119461295,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7524595366550304e-05,
-      "loss": 1.3067,
-      "step": 2110
-    },
-    {
-      "epoch": 0.6634077389716709,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.7522056490003175e-05,
-      "loss": 1.3295,
-      "step": 2112
-    },
-    {
-      "epoch": 0.6640359659972123,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.7519517613456047e-05,
-      "loss": 1.3994,
-      "step": 2114
-    },
-    {
-      "epoch": 0.6646641930227536,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.7516978736908918e-05,
-      "loss": 1.3512,
-      "step": 2116
-    },
-    {
-      "epoch": 0.6652924200482949,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7514439860361793e-05,
-      "loss": 1.2393,
-      "step": 2118
-    },
-    {
-      "epoch": 0.6659206470738364,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.751190098381466e-05,
-      "loss": 1.2977,
-      "step": 2120
-    },
-    {
-      "epoch": 0.6665488740993777,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7509362107267536e-05,
-      "loss": 1.4039,
-      "step": 2122
-    },
-    {
-      "epoch": 0.667177101124919,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7506823230720407e-05,
-      "loss": 1.3294,
-      "step": 2124
-    },
-    {
-      "epoch": 0.6678053281504603,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.750428435417328e-05,
-      "loss": 1.2816,
-      "step": 2126
-    },
-    {
-      "epoch": 0.6684335551760018,
-      "grad_norm": 0.75,
-      "learning_rate": 1.750174547762615e-05,
-      "loss": 1.3298,
-      "step": 2128
-    },
-    {
-      "epoch": 0.6690617822015431,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.7499206601079025e-05,
-      "loss": 1.3823,
-      "step": 2130
-    },
-    {
-      "epoch": 0.6696900092270844,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.7496667724531896e-05,
-      "loss": 1.2973,
-      "step": 2132
-    },
-    {
-      "epoch": 0.6703182362526258,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.7494128847984767e-05,
-      "loss": 1.3873,
-      "step": 2134
-    },
-    {
-      "epoch": 0.6709464632781671,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7491589971437642e-05,
-      "loss": 1.3746,
-      "step": 2136
-    },
-    {
-      "epoch": 0.6715746903037085,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7489051094890514e-05,
-      "loss": 1.2803,
-      "step": 2138
-    },
-    {
-      "epoch": 0.6722029173292499,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7486512218343385e-05,
-      "loss": 1.3632,
-      "step": 2140
-    },
-    {
-      "epoch": 0.6728311443547912,
-      "grad_norm": 0.75,
-      "learning_rate": 1.7483973341796256e-05,
-      "loss": 1.3377,
-      "step": 2142
-    },
-    {
-      "epoch": 0.6734593713803325,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.748143446524913e-05,
-      "loss": 1.2896,
-      "step": 2144
-    },
-    {
-      "epoch": 0.674087598405874,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.7478895588702e-05,
-      "loss": 1.2543,
-      "step": 2146
-    },
-    {
-      "epoch": 0.6747158254314153,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.7476356712154874e-05,
-      "loss": 1.2882,
-      "step": 2148
-    },
-    {
-      "epoch": 0.6753440524569566,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.7473817835607745e-05,
-      "loss": 1.3234,
-      "step": 2150
-    },
-    {
-      "epoch": 0.6759722794824979,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.7471278959060617e-05,
-      "loss": 1.2965,
-      "step": 2152
-    },
-    {
-      "epoch": 0.6766005065080394,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.7468740082513488e-05,
-      "loss": 1.4172,
-      "step": 2154
-    },
-    {
-      "epoch": 0.6772287335335807,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7466201205966363e-05,
-      "loss": 1.3369,
-      "step": 2156
-    },
-    {
-      "epoch": 0.677856960559122,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7463662329419234e-05,
-      "loss": 1.3239,
-      "step": 2158
-    },
-    {
-      "epoch": 0.6784851875846634,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7461123452872105e-05,
-      "loss": 1.2926,
-      "step": 2160
-    },
-    {
-      "epoch": 0.6791134146102048,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.7458584576324977e-05,
-      "loss": 1.3588,
-      "step": 2162
-    },
-    {
-      "epoch": 0.6797416416357461,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.745604569977785e-05,
-      "loss": 1.1972,
-      "step": 2164
-    },
-    {
-      "epoch": 0.6803698686612875,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7453506823230723e-05,
-      "loss": 1.2391,
-      "step": 2166
-    },
-    {
-      "epoch": 0.6809980956868288,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.7450967946683594e-05,
-      "loss": 1.3438,
-      "step": 2168
-    },
-    {
-      "epoch": 0.6816263227123702,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.7448429070136466e-05,
-      "loss": 1.4117,
-      "step": 2170
-    },
-    {
-      "epoch": 0.6822545497379116,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.7445890193589337e-05,
-      "loss": 1.388,
-      "step": 2172
-    },
-    {
-      "epoch": 0.6828827767634529,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.744335131704221e-05,
-      "loss": 1.3602,
-      "step": 2174
-    },
-    {
-      "epoch": 0.6835110037889942,
-      "grad_norm": 1.0546875,
-      "learning_rate": 1.7440812440495083e-05,
-      "loss": 1.2999,
-      "step": 2176
-    },
-    {
-      "epoch": 0.6841392308145356,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.7438273563947955e-05,
-      "loss": 1.3296,
-      "step": 2178
-    },
-    {
-      "epoch": 0.684767457840077,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.7435734687400826e-05,
-      "loss": 1.302,
-      "step": 2180
-    },
-    {
-      "epoch": 0.6853956848656183,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.7433195810853697e-05,
-      "loss": 1.321,
-      "step": 2182
-    },
-    {
-      "epoch": 0.6860239118911596,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7430656934306572e-05,
-      "loss": 1.3628,
-      "step": 2184
-    },
-    {
-      "epoch": 0.6866521389167011,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.7428118057759444e-05,
-      "loss": 1.3183,
-      "step": 2186
-    },
-    {
-      "epoch": 0.6872803659422424,
-      "grad_norm": 0.921875,
-      "learning_rate": 1.7425579181212315e-05,
-      "loss": 1.4956,
-      "step": 2188
-    },
-    {
-      "epoch": 0.6879085929677837,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7423040304665186e-05,
-      "loss": 1.4264,
-      "step": 2190
-    },
-    {
-      "epoch": 0.688536819993325,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.742050142811806e-05,
-      "loss": 1.3176,
-      "step": 2192
-    },
-    {
-      "epoch": 0.6891650470188665,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.741796255157093e-05,
-      "loss": 1.3268,
-      "step": 2194
-    },
-    {
-      "epoch": 0.6897932740444078,
-      "grad_norm": 1.046875,
-      "learning_rate": 1.7415423675023804e-05,
-      "loss": 1.346,
-      "step": 2196
-    },
-    {
-      "epoch": 0.6904215010699492,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.7412884798476675e-05,
-      "loss": 1.3614,
-      "step": 2198
-    },
-    {
-      "epoch": 0.6910497280954905,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7410345921929547e-05,
-      "loss": 1.2779,
-      "step": 2200
-    },
-    {
-      "epoch": 0.6916779551210319,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7407807045382418e-05,
-      "loss": 1.2913,
-      "step": 2202
-    },
-    {
-      "epoch": 0.6923061821465732,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.7405268168835293e-05,
-      "loss": 1.3669,
-      "step": 2204
-    },
-    {
-      "epoch": 0.6929344091721146,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7402729292288164e-05,
-      "loss": 1.2992,
-      "step": 2206
-    },
-    {
-      "epoch": 0.6935626361976559,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.7400190415741036e-05,
-      "loss": 1.3157,
-      "step": 2208
-    },
-    {
-      "epoch": 0.6941908632231972,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.7397651539193907e-05,
-      "loss": 1.1712,
-      "step": 2210
-    },
-    {
-      "epoch": 0.6948190902487387,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7395112662646782e-05,
-      "loss": 1.1813,
-      "step": 2212
-    },
-    {
-      "epoch": 0.69544731727428,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.739257378609965e-05,
-      "loss": 1.3688,
-      "step": 2214
-    },
-    {
-      "epoch": 0.6960755442998213,
-      "grad_norm": 0.9453125,
-      "learning_rate": 1.7390034909552525e-05,
-      "loss": 1.3554,
-      "step": 2216
-    },
-    {
-      "epoch": 0.6967037713253627,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.7387496033005396e-05,
-      "loss": 1.3605,
-      "step": 2218
-    },
-    {
-      "epoch": 0.6973319983509041,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.7384957156458267e-05,
-      "loss": 1.2138,
-      "step": 2220
-    },
-    {
-      "epoch": 0.6979602253764454,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.7382418279911142e-05,
-      "loss": 1.309,
-      "step": 2222
-    },
-    {
-      "epoch": 0.6985884524019867,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7379879403364013e-05,
-      "loss": 1.2578,
-      "step": 2224
-    },
-    {
-      "epoch": 0.6992166794275281,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7377340526816885e-05,
-      "loss": 1.3457,
-      "step": 2226
-    },
-    {
-      "epoch": 0.6998449064530695,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.7374801650269756e-05,
-      "loss": 1.2938,
-      "step": 2228
-    },
-    {
-      "epoch": 0.7004731334786108,
-      "grad_norm": 0.8984375,
-      "learning_rate": 1.737226277372263e-05,
-      "loss": 1.284,
-      "step": 2230
-    },
-    {
-      "epoch": 0.7011013605041522,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.7369723897175502e-05,
-      "loss": 1.415,
-      "step": 2232
-    },
-    {
-      "epoch": 0.7017295875296935,
-      "grad_norm": 0.94921875,
-      "learning_rate": 1.7367185020628374e-05,
-      "loss": 1.2291,
-      "step": 2234
-    },
-    {
-      "epoch": 0.7023578145552349,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.7364646144081245e-05,
-      "loss": 1.374,
-      "step": 2236
-    },
-    {
-      "epoch": 0.7029860415807763,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.736210726753412e-05,
-      "loss": 1.4697,
-      "step": 2238
-    },
-    {
-      "epoch": 0.7036142686063176,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7359568390986988e-05,
-      "loss": 1.2784,
-      "step": 2240
-    },
-    {
-      "epoch": 0.7042424956318589,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7357029514439863e-05,
-      "loss": 1.2381,
-      "step": 2242
-    },
-    {
-      "epoch": 0.7048707226574004,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.7354490637892734e-05,
-      "loss": 1.2173,
-      "step": 2244
-    },
-    {
-      "epoch": 0.7054989496829417,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7351951761345605e-05,
-      "loss": 1.2839,
-      "step": 2246
-    },
-    {
-      "epoch": 0.706127176708483,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.7349412884798477e-05,
-      "loss": 1.3768,
-      "step": 2248
-    },
-    {
-      "epoch": 0.7067554037340243,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.734687400825135e-05,
-      "loss": 1.3607,
-      "step": 2250
-    },
-    {
-      "epoch": 0.7073836307595658,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.7344335131704223e-05,
-      "loss": 1.3943,
-      "step": 2252
-    },
-    {
-      "epoch": 0.7080118577851071,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.7341796255157094e-05,
-      "loss": 1.4092,
-      "step": 2254
-    },
-    {
-      "epoch": 0.7086400848106484,
-      "grad_norm": 0.75,
-      "learning_rate": 1.7339257378609966e-05,
-      "loss": 1.2429,
-      "step": 2256
-    },
-    {
-      "epoch": 0.7092683118361898,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.733671850206284e-05,
-      "loss": 1.458,
-      "step": 2258
-    },
-    {
-      "epoch": 0.7098965388617312,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.7334179625515712e-05,
-      "loss": 1.3227,
-      "step": 2260
-    },
-    {
-      "epoch": 0.7105247658872725,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.7331640748968583e-05,
-      "loss": 1.4453,
-      "step": 2262
-    },
-    {
-      "epoch": 0.7111529929128139,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.7329101872421455e-05,
-      "loss": 1.2725,
-      "step": 2264
-    },
-    {
-      "epoch": 0.7117812199383552,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.7326562995874326e-05,
-      "loss": 1.2165,
-      "step": 2266
-    },
-    {
-      "epoch": 0.7124094469638966,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.7324024119327197e-05,
-      "loss": 1.4287,
-      "step": 2268
-    },
-    {
-      "epoch": 0.713037673989438,
-      "grad_norm": 0.6484375,
-      "learning_rate": 1.7321485242780072e-05,
-      "loss": 1.2783,
-      "step": 2270
-    },
-    {
-      "epoch": 0.7136659010149793,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.7318946366232944e-05,
-      "loss": 1.3641,
-      "step": 2272
-    },
-    {
-      "epoch": 0.7142941280405206,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.7316407489685815e-05,
-      "loss": 1.3313,
-      "step": 2274
-    },
-    {
-      "epoch": 0.7149223550660619,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.7313868613138686e-05,
-      "loss": 1.3461,
-      "step": 2276
-    },
-    {
-      "epoch": 0.7155505820916034,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.731132973659156e-05,
-      "loss": 1.3159,
-      "step": 2278
-    },
-    {
-      "epoch": 0.7161788091171447,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.7308790860044432e-05,
-      "loss": 1.2575,
-      "step": 2280
-    },
-    {
-      "epoch": 0.716807036142686,
-      "grad_norm": 0.65625,
-      "learning_rate": 1.7306251983497304e-05,
-      "loss": 1.3816,
-      "step": 2282
-    },
-    {
-      "epoch": 0.7174352631682274,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7303713106950175e-05,
-      "loss": 1.4548,
-      "step": 2284
-    },
-    {
-      "epoch": 0.7180634901937688,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.730117423040305e-05,
-      "loss": 1.2777,
-      "step": 2286
-    },
-    {
-      "epoch": 0.7186917172193101,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.7298635353855918e-05,
-      "loss": 1.3142,
-      "step": 2288
-    },
-    {
-      "epoch": 0.7193199442448515,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.7296096477308793e-05,
-      "loss": 1.2618,
-      "step": 2290
-    },
-    {
-      "epoch": 0.7199481712703928,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.7293557600761664e-05,
-      "loss": 1.3586,
-      "step": 2292
-    },
-    {
-      "epoch": 0.7205763982959342,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7291018724214536e-05,
-      "loss": 1.2284,
-      "step": 2294
-    },
-    {
-      "epoch": 0.7212046253214756,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.7288479847667407e-05,
-      "loss": 1.3143,
-      "step": 2296
-    },
-    {
-      "epoch": 0.7218328523470169,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.728594097112028e-05,
-      "loss": 1.2988,
-      "step": 2298
-    },
-    {
-      "epoch": 0.7224610793725582,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.7283402094573153e-05,
-      "loss": 1.3754,
-      "step": 2300
-    },
-    {
-      "epoch": 0.7230893063980997,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.7280863218026024e-05,
-      "loss": 1.3633,
-      "step": 2302
-    },
-    {
-      "epoch": 0.723717533423641,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7278324341478896e-05,
-      "loss": 1.4773,
-      "step": 2304
-    },
-    {
-      "epoch": 0.7243457604491823,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.727578546493177e-05,
-      "loss": 1.351,
-      "step": 2306
-    },
-    {
-      "epoch": 0.7249739874747236,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7273246588384642e-05,
-      "loss": 1.3577,
-      "step": 2308
-    },
-    {
-      "epoch": 0.7256022145002651,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.7270707711837513e-05,
-      "loss": 1.2883,
-      "step": 2310
-    },
-    {
-      "epoch": 0.7262304415258064,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.7268168835290388e-05,
-      "loss": 1.4049,
-      "step": 2312
-    },
-    {
-      "epoch": 0.7268586685513477,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.7265629958743256e-05,
-      "loss": 1.3443,
-      "step": 2314
-    },
-    {
-      "epoch": 0.7274868955768891,
-      "grad_norm": 0.90234375,
-      "learning_rate": 1.726309108219613e-05,
-      "loss": 1.2131,
-      "step": 2316
-    },
-    {
-      "epoch": 0.7281151226024305,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7260552205649002e-05,
-      "loss": 1.353,
-      "step": 2318
-    },
-    {
-      "epoch": 0.7287433496279718,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7258013329101874e-05,
-      "loss": 1.2911,
-      "step": 2320
-    },
-    {
-      "epoch": 0.7293715766535132,
-      "grad_norm": 0.90625,
-      "learning_rate": 1.7255474452554745e-05,
-      "loss": 1.3567,
-      "step": 2322
-    },
-    {
-      "epoch": 0.7299998036790545,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.725293557600762e-05,
-      "loss": 1.3589,
-      "step": 2324
-    },
-    {
-      "epoch": 0.7306280307045959,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.725039669946049e-05,
-      "loss": 1.3734,
-      "step": 2326
-    },
-    {
-      "epoch": 0.7312562577301372,
-      "grad_norm": 0.6484375,
-      "learning_rate": 1.7247857822913363e-05,
-      "loss": 1.3007,
-      "step": 2328
-    },
-    {
-      "epoch": 0.7318844847556786,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.7245318946366234e-05,
-      "loss": 1.3652,
-      "step": 2330
-    },
-    {
-      "epoch": 0.7325127117812199,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.724278006981911e-05,
-      "loss": 1.3929,
-      "step": 2332
-    },
-    {
-      "epoch": 0.7331409388067613,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.7240241193271977e-05,
-      "loss": 1.2893,
-      "step": 2334
-    },
-    {
-      "epoch": 0.7337691658323027,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.723770231672485e-05,
-      "loss": 1.4986,
-      "step": 2336
-    },
-    {
-      "epoch": 0.734397392857844,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.7235163440177723e-05,
-      "loss": 1.3224,
-      "step": 2338
-    },
-    {
-      "epoch": 0.7350256198833853,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7232624563630594e-05,
-      "loss": 1.422,
-      "step": 2340
-    },
-    {
-      "epoch": 0.7356538469089268,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.7230085687083466e-05,
-      "loss": 1.4021,
-      "step": 2342
-    },
-    {
-      "epoch": 0.7362820739344681,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.722754681053634e-05,
-      "loss": 1.3948,
-      "step": 2344
-    },
-    {
-      "epoch": 0.7369103009600094,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.7225007933989212e-05,
-      "loss": 1.2958,
-      "step": 2346
-    },
-    {
-      "epoch": 0.7375385279855508,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.7222469057442083e-05,
-      "loss": 1.2972,
-      "step": 2348
-    },
-    {
-      "epoch": 0.7381667550110921,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.7219930180894955e-05,
-      "loss": 1.3356,
-      "step": 2350
-    },
-    {
-      "epoch": 0.7387949820366335,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.721739130434783e-05,
-      "loss": 1.2247,
-      "step": 2352
-    },
-    {
-      "epoch": 0.7394232090621748,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.7214852427800697e-05,
-      "loss": 1.3243,
-      "step": 2354
-    },
-    {
-      "epoch": 0.7400514360877162,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7212313551253572e-05,
-      "loss": 1.4064,
-      "step": 2356
-    },
-    {
-      "epoch": 0.7406796631132575,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7209774674706443e-05,
-      "loss": 1.4806,
-      "step": 2358
-    },
-    {
-      "epoch": 0.7413078901387989,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.7207235798159315e-05,
-      "loss": 1.3769,
-      "step": 2360
-    },
-    {
-      "epoch": 0.7419361171643403,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.7204696921612186e-05,
-      "loss": 1.2256,
-      "step": 2362
-    },
-    {
-      "epoch": 0.7425643441898816,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.720215804506506e-05,
-      "loss": 1.389,
-      "step": 2364
-    },
-    {
-      "epoch": 0.7431925712154229,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.7199619168517932e-05,
-      "loss": 1.4362,
-      "step": 2366
-    },
-    {
-      "epoch": 0.7438207982409644,
-      "grad_norm": 0.8984375,
-      "learning_rate": 1.7197080291970804e-05,
-      "loss": 1.4191,
-      "step": 2368
-    },
-    {
-      "epoch": 0.7444490252665057,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7194541415423675e-05,
-      "loss": 1.3115,
-      "step": 2370
-    },
-    {
-      "epoch": 0.745077252292047,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.719200253887655e-05,
-      "loss": 1.4019,
-      "step": 2372
-    },
-    {
-      "epoch": 0.7457054793175883,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.718946366232942e-05,
-      "loss": 1.3587,
-      "step": 2374
-    },
-    {
-      "epoch": 0.7463337063431298,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.7186924785782293e-05,
-      "loss": 1.3749,
-      "step": 2376
-    },
-    {
-      "epoch": 0.7469619333686711,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.7184385909235164e-05,
-      "loss": 1.3042,
-      "step": 2378
-    },
-    {
-      "epoch": 0.7475901603942124,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7181847032688035e-05,
-      "loss": 1.2356,
-      "step": 2380
-    },
-    {
-      "epoch": 0.7482183874197538,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.7179308156140907e-05,
-      "loss": 1.2864,
-      "step": 2382
-    },
-    {
-      "epoch": 0.7488466144452952,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.717676927959378e-05,
-      "loss": 1.3995,
-      "step": 2384
-    },
-    {
-      "epoch": 0.7494748414708365,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.7174230403046653e-05,
-      "loss": 1.2924,
-      "step": 2386
-    },
-    {
-      "epoch": 0.7501030684963779,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.7171691526499524e-05,
-      "loss": 1.2801,
-      "step": 2388
-    },
-    {
-      "epoch": 0.7507312955219192,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.7169152649952396e-05,
-      "loss": 1.2726,
-      "step": 2390
-    },
-    {
-      "epoch": 0.7513595225474606,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.716661377340527e-05,
-      "loss": 1.35,
-      "step": 2392
-    },
-    {
-      "epoch": 0.751987749573002,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.7164074896858142e-05,
-      "loss": 1.2783,
-      "step": 2394
-    },
-    {
-      "epoch": 0.7526159765985433,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7161536020311013e-05,
-      "loss": 1.3665,
-      "step": 2396
-    },
-    {
-      "epoch": 0.7532442036240846,
-      "grad_norm": 0.98046875,
-      "learning_rate": 1.7158997143763888e-05,
-      "loss": 1.3679,
-      "step": 2398
-    },
-    {
-      "epoch": 0.7538724306496261,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.715645826721676e-05,
-      "loss": 1.3874,
-      "step": 2400
-    },
-    {
-      "epoch": 0.7545006576751674,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.715391939066963e-05,
-      "loss": 1.2631,
-      "step": 2402
-    },
-    {
-      "epoch": 0.7551288847007087,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.7151380514122502e-05,
-      "loss": 1.2159,
-      "step": 2404
-    },
-    {
-      "epoch": 0.75575711172625,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.7148841637575374e-05,
-      "loss": 1.3067,
-      "step": 2406
-    },
-    {
-      "epoch": 0.7563853387517915,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.7146302761028245e-05,
-      "loss": 1.3503,
-      "step": 2408
-    },
-    {
-      "epoch": 0.7570135657773328,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.714376388448112e-05,
-      "loss": 1.369,
-      "step": 2410
-    },
-    {
-      "epoch": 0.7576417928028741,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.714122500793399e-05,
-      "loss": 1.2634,
-      "step": 2412
-    },
-    {
-      "epoch": 0.7582700198284155,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.7138686131386862e-05,
-      "loss": 1.2631,
-      "step": 2414
-    },
-    {
-      "epoch": 0.7588982468539569,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.7136147254839734e-05,
-      "loss": 1.33,
-      "step": 2416
-    },
-    {
-      "epoch": 0.7595264738794982,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.713360837829261e-05,
-      "loss": 1.3229,
-      "step": 2418
-    },
-    {
-      "epoch": 0.7601547009050396,
-      "grad_norm": 1.1640625,
-      "learning_rate": 1.713106950174548e-05,
-      "loss": 1.2857,
-      "step": 2420
-    },
-    {
-      "epoch": 0.7607829279305809,
-      "grad_norm": 0.875,
-      "learning_rate": 1.712853062519835e-05,
-      "loss": 1.3812,
-      "step": 2422
-    },
-    {
-      "epoch": 0.7614111549561222,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.7125991748651223e-05,
-      "loss": 1.3809,
-      "step": 2424
-    },
-    {
-      "epoch": 0.7620393819816637,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.7123452872104098e-05,
-      "loss": 1.3366,
-      "step": 2426
-    },
-    {
-      "epoch": 0.762667609007205,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.7120913995556966e-05,
-      "loss": 1.3628,
-      "step": 2428
-    },
-    {
-      "epoch": 0.7632958360327463,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.711837511900984e-05,
-      "loss": 1.3394,
-      "step": 2430
-    },
-    {
-      "epoch": 0.7639240630582876,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.7115836242462712e-05,
-      "loss": 1.4378,
-      "step": 2432
-    },
-    {
-      "epoch": 0.7645522900838291,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.7113297365915583e-05,
-      "loss": 1.1978,
-      "step": 2434
-    },
-    {
-      "epoch": 0.7651805171093704,
-      "grad_norm": 0.75,
-      "learning_rate": 1.7110758489368454e-05,
-      "loss": 1.2939,
-      "step": 2436
-    },
-    {
-      "epoch": 0.7658087441349117,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.710821961282133e-05,
-      "loss": 1.3248,
-      "step": 2438
-    },
-    {
-      "epoch": 0.7664369711604531,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.71056807362742e-05,
-      "loss": 1.2087,
-      "step": 2440
-    },
-    {
-      "epoch": 0.7670651981859945,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.7103141859727072e-05,
-      "loss": 1.1633,
-      "step": 2442
-    },
-    {
-      "epoch": 0.7676934252115358,
-      "grad_norm": 1.078125,
-      "learning_rate": 1.7100602983179943e-05,
-      "loss": 1.2432,
-      "step": 2444
-    },
-    {
-      "epoch": 0.7683216522370772,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.7098064106632818e-05,
-      "loss": 1.3272,
-      "step": 2446
-    },
-    {
-      "epoch": 0.7689498792626185,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.7095525230085686e-05,
-      "loss": 1.2589,
-      "step": 2448
-    },
-    {
-      "epoch": 0.7695781062881599,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.709298635353856e-05,
-      "loss": 1.375,
-      "step": 2450
-    },
-    {
-      "epoch": 0.7702063333137013,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.7090447476991432e-05,
-      "loss": 1.2817,
-      "step": 2452
-    },
-    {
-      "epoch": 0.7708345603392426,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.7087908600444304e-05,
-      "loss": 1.2879,
-      "step": 2454
-    },
-    {
-      "epoch": 0.7714627873647839,
-      "grad_norm": 1.234375,
-      "learning_rate": 1.7085369723897175e-05,
-      "loss": 1.2573,
-      "step": 2456
-    },
-    {
-      "epoch": 0.7720910143903253,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.708283084735005e-05,
-      "loss": 1.343,
-      "step": 2458
-    },
-    {
-      "epoch": 0.7727192414158667,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.708029197080292e-05,
-      "loss": 1.357,
-      "step": 2460
-    },
-    {
-      "epoch": 0.773347468441408,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.7077753094255793e-05,
-      "loss": 1.3493,
-      "step": 2462
-    },
-    {
-      "epoch": 0.7739756954669493,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7075214217708664e-05,
-      "loss": 1.2568,
-      "step": 2464
-    },
-    {
-      "epoch": 0.7746039224924908,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.707267534116154e-05,
-      "loss": 1.3476,
-      "step": 2466
-    },
-    {
-      "epoch": 0.7752321495180321,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.707013646461441e-05,
-      "loss": 1.2797,
-      "step": 2468
-    },
-    {
-      "epoch": 0.7758603765435734,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.706759758806728e-05,
-      "loss": 1.3368,
-      "step": 2470
-    },
-    {
-      "epoch": 0.7764886035691148,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.7065058711520153e-05,
-      "loss": 1.2807,
-      "step": 2472
-    },
-    {
-      "epoch": 0.7771168305946562,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.7062519834973024e-05,
-      "loss": 1.5012,
-      "step": 2474
-    },
-    {
-      "epoch": 0.7777450576201975,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.7059980958425896e-05,
-      "loss": 1.3204,
-      "step": 2476
-    },
-    {
-      "epoch": 0.7783732846457388,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.705744208187877e-05,
-      "loss": 1.3475,
-      "step": 2478
-    },
-    {
-      "epoch": 0.7790015116712802,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.7054903205331642e-05,
-      "loss": 1.2051,
-      "step": 2480
-    },
-    {
-      "epoch": 0.7796297386968216,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.7052364328784513e-05,
-      "loss": 1.3502,
-      "step": 2482
-    },
-    {
-      "epoch": 0.780257965722363,
-      "grad_norm": 0.9140625,
-      "learning_rate": 1.7049825452237388e-05,
-      "loss": 1.2337,
-      "step": 2484
-    },
-    {
-      "epoch": 0.7808861927479043,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.704728657569026e-05,
-      "loss": 1.3524,
-      "step": 2486
-    },
-    {
-      "epoch": 0.7815144197734456,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.704474769914313e-05,
-      "loss": 1.3843,
-      "step": 2488
-    },
-    {
-      "epoch": 0.7821426467989869,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.7042208822596002e-05,
-      "loss": 1.3905,
-      "step": 2490
-    },
-    {
-      "epoch": 0.7827708738245284,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.7039669946048877e-05,
-      "loss": 1.3168,
-      "step": 2492
-    },
-    {
-      "epoch": 0.7833991008500697,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.7037131069501748e-05,
-      "loss": 1.233,
-      "step": 2494
-    },
-    {
-      "epoch": 0.784027327875611,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.703459219295462e-05,
-      "loss": 1.3278,
-      "step": 2496
-    },
-    {
-      "epoch": 0.7846555549011524,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.703205331640749e-05,
-      "loss": 1.2751,
-      "step": 2498
-    },
-    {
-      "epoch": 0.7852837819266938,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.7029514439860362e-05,
-      "loss": 1.3463,
-      "step": 2500
-    },
-    {
-      "epoch": 0.7859120089522351,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.7026975563313234e-05,
-      "loss": 1.2921,
-      "step": 2502
-    },
-    {
-      "epoch": 0.7865402359777764,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.702443668676611e-05,
-      "loss": 1.1402,
-      "step": 2504
-    },
-    {
-      "epoch": 0.7871684630033178,
-      "grad_norm": 1.125,
-      "learning_rate": 1.702189781021898e-05,
-      "loss": 1.2382,
-      "step": 2506
-    },
-    {
-      "epoch": 0.7877966900288592,
-      "grad_norm": 0.63671875,
-      "learning_rate": 1.701935893367185e-05,
-      "loss": 1.3848,
-      "step": 2508
-    },
-    {
-      "epoch": 0.7884249170544005,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.7016820057124723e-05,
-      "loss": 1.2577,
-      "step": 2510
-    },
-    {
-      "epoch": 0.7890531440799419,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.7014281180577597e-05,
-      "loss": 1.4976,
-      "step": 2512
-    },
-    {
-      "epoch": 0.7896813711054832,
-      "grad_norm": 0.65234375,
-      "learning_rate": 1.701174230403047e-05,
-      "loss": 1.3051,
-      "step": 2514
-    },
-    {
-      "epoch": 0.7903095981310246,
-      "grad_norm": 0.75,
-      "learning_rate": 1.700920342748334e-05,
-      "loss": 1.3637,
-      "step": 2516
-    },
-    {
-      "epoch": 0.790937825156566,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.700666455093621e-05,
-      "loss": 1.2335,
-      "step": 2518
-    },
-    {
-      "epoch": 0.7915660521821073,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.7004125674389086e-05,
-      "loss": 1.2534,
-      "step": 2520
-    },
-    {
-      "epoch": 0.7921942792076486,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.7001586797841954e-05,
-      "loss": 1.4272,
-      "step": 2522
-    },
-    {
-      "epoch": 0.7928225062331901,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.699904792129483e-05,
-      "loss": 1.2296,
-      "step": 2524
-    },
-    {
-      "epoch": 0.7934507332587314,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.69965090447477e-05,
-      "loss": 1.3799,
-      "step": 2526
-    },
-    {
-      "epoch": 0.7940789602842727,
-      "grad_norm": 0.625,
-      "learning_rate": 1.6993970168200572e-05,
-      "loss": 1.4241,
-      "step": 2528
-    },
-    {
-      "epoch": 0.794707187309814,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.6991431291653443e-05,
-      "loss": 1.2411,
-      "step": 2530
-    },
-    {
-      "epoch": 0.7953354143353555,
-      "grad_norm": 1.078125,
-      "learning_rate": 1.6988892415106318e-05,
-      "loss": 1.3962,
-      "step": 2532
-    },
-    {
-      "epoch": 0.7959636413608968,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.698635353855919e-05,
-      "loss": 1.3154,
-      "step": 2534
-    },
-    {
-      "epoch": 0.7965918683864381,
-      "grad_norm": 0.62890625,
-      "learning_rate": 1.698381466201206e-05,
-      "loss": 1.3236,
-      "step": 2536
-    },
-    {
-      "epoch": 0.7972200954119795,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.6981275785464932e-05,
-      "loss": 1.2605,
-      "step": 2538
-    },
-    {
-      "epoch": 0.7978483224375209,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.6978736908917807e-05,
-      "loss": 1.2216,
-      "step": 2540
-    },
-    {
-      "epoch": 0.7984765494630622,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.6976198032370675e-05,
-      "loss": 1.3394,
-      "step": 2542
-    },
-    {
-      "epoch": 0.7991047764886036,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.697365915582355e-05,
-      "loss": 1.331,
-      "step": 2544
-    },
-    {
-      "epoch": 0.7997330035141449,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.697112027927642e-05,
-      "loss": 1.3703,
-      "step": 2546
-    },
-    {
-      "epoch": 0.8003612305396863,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.6968581402729293e-05,
-      "loss": 1.3128,
-      "step": 2548
-    },
-    {
-      "epoch": 0.8009894575652277,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.6966042526182164e-05,
-      "loss": 1.278,
-      "step": 2550
-    },
-    {
-      "epoch": 0.801617684590769,
-      "grad_norm": 0.65625,
-      "learning_rate": 1.696350364963504e-05,
-      "loss": 1.3876,
-      "step": 2552
-    },
-    {
-      "epoch": 0.8022459116163103,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.696096477308791e-05,
-      "loss": 1.2858,
-      "step": 2554
-    },
-    {
-      "epoch": 0.8028741386418518,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.695842589654078e-05,
-      "loss": 1.412,
-      "step": 2556
-    },
-    {
-      "epoch": 0.8035023656673931,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.6955887019993653e-05,
-      "loss": 1.4499,
-      "step": 2558
-    },
-    {
-      "epoch": 0.8041305926929344,
-      "grad_norm": 0.9140625,
-      "learning_rate": 1.6953348143446528e-05,
-      "loss": 1.291,
-      "step": 2560
-    },
-    {
-      "epoch": 0.8047588197184757,
-      "grad_norm": 0.90625,
-      "learning_rate": 1.69508092668994e-05,
-      "loss": 1.4154,
-      "step": 2562
-    },
-    {
-      "epoch": 0.8053870467440171,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.694827039035227e-05,
-      "loss": 1.4474,
-      "step": 2564
-    },
-    {
-      "epoch": 0.8060152737695585,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.6945731513805145e-05,
-      "loss": 1.3263,
-      "step": 2566
-    },
-    {
-      "epoch": 0.8066435007950998,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.6943192637258013e-05,
-      "loss": 1.3238,
-      "step": 2568
-    },
-    {
-      "epoch": 0.8072717278206412,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.6940653760710888e-05,
-      "loss": 1.4225,
-      "step": 2570
-    },
-    {
-      "epoch": 0.8078999548461825,
-      "grad_norm": 0.70703125,
-      "learning_rate": 1.693811488416376e-05,
-      "loss": 1.2038,
-      "step": 2572
-    },
-    {
-      "epoch": 0.8085281818717239,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.693557600761663e-05,
-      "loss": 1.1913,
-      "step": 2574
-    },
-    {
-      "epoch": 0.8091564088972653,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.6933037131069502e-05,
-      "loss": 1.3431,
-      "step": 2576
-    },
-    {
-      "epoch": 0.8097846359228066,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.6930498254522377e-05,
-      "loss": 1.3336,
-      "step": 2578
-    },
-    {
-      "epoch": 0.8104128629483479,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.6927959377975248e-05,
-      "loss": 1.2205,
-      "step": 2580
-    },
-    {
-      "epoch": 0.8110410899738894,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.692542050142812e-05,
-      "loss": 1.3004,
-      "step": 2582
-    },
-    {
-      "epoch": 0.8116693169994307,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.692288162488099e-05,
-      "loss": 1.3125,
-      "step": 2584
-    },
-    {
-      "epoch": 0.812297544024972,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6920342748333866e-05,
-      "loss": 1.4572,
-      "step": 2586
-    },
-    {
-      "epoch": 0.8129257710505133,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.6917803871786737e-05,
-      "loss": 1.2809,
-      "step": 2588
-    },
-    {
-      "epoch": 0.8135539980760548,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.691526499523961e-05,
-      "loss": 1.2979,
-      "step": 2590
-    },
-    {
-      "epoch": 0.8141822251015961,
-      "grad_norm": 0.890625,
-      "learning_rate": 1.691272611869248e-05,
-      "loss": 1.3751,
-      "step": 2592
-    },
-    {
-      "epoch": 0.8148104521271374,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.691018724214535e-05,
-      "loss": 1.3556,
-      "step": 2594
-    },
-    {
-      "epoch": 0.8154386791526788,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.6907648365598223e-05,
-      "loss": 1.2648,
-      "step": 2596
-    },
-    {
-      "epoch": 0.8160669061782202,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.6905109489051097e-05,
-      "loss": 1.3499,
-      "step": 2598
-    },
-    {
-      "epoch": 0.8166951332037615,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.690257061250397e-05,
-      "loss": 1.3424,
-      "step": 2600
-    },
-    {
-      "epoch": 0.8173233602293029,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.690003173595684e-05,
-      "loss": 1.3746,
-      "step": 2602
-    },
-    {
-      "epoch": 0.8179515872548442,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.689749285940971e-05,
-      "loss": 1.3152,
-      "step": 2604
-    },
-    {
-      "epoch": 0.8185798142803856,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.6894953982862586e-05,
-      "loss": 1.3755,
-      "step": 2606
-    },
-    {
-      "epoch": 0.819208041305927,
-      "grad_norm": 0.84765625,
-      "learning_rate": 1.6892415106315458e-05,
-      "loss": 1.2247,
-      "step": 2608
-    },
-    {
-      "epoch": 0.8198362683314683,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.688987622976833e-05,
-      "loss": 1.4328,
-      "step": 2610
-    },
-    {
-      "epoch": 0.8204644953570096,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.68873373532212e-05,
-      "loss": 1.2965,
-      "step": 2612
-    },
-    {
-      "epoch": 0.821092722382551,
-      "grad_norm": 0.91015625,
-      "learning_rate": 1.6884798476674075e-05,
-      "loss": 1.2175,
-      "step": 2614
-    },
-    {
-      "epoch": 0.8217209494080924,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.6882259600126943e-05,
-      "loss": 1.1868,
-      "step": 2616
-    },
-    {
-      "epoch": 0.8223491764336337,
-      "grad_norm": 0.9296875,
-      "learning_rate": 1.6879720723579818e-05,
-      "loss": 1.331,
-      "step": 2618
-    },
-    {
-      "epoch": 0.822977403459175,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.687718184703269e-05,
-      "loss": 1.3342,
-      "step": 2620
-    },
-    {
-      "epoch": 0.8236056304847165,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.687464297048556e-05,
-      "loss": 1.3036,
-      "step": 2622
-    },
-    {
-      "epoch": 0.8242338575102578,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.6872104093938432e-05,
-      "loss": 1.2481,
-      "step": 2624
-    },
-    {
-      "epoch": 0.8248620845357991,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6869565217391307e-05,
-      "loss": 1.3175,
-      "step": 2626
-    },
-    {
-      "epoch": 0.8254903115613405,
-      "grad_norm": 0.97265625,
-      "learning_rate": 1.686702634084418e-05,
-      "loss": 1.3181,
-      "step": 2628
-    },
-    {
-      "epoch": 0.8261185385868819,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.686448746429705e-05,
-      "loss": 1.3106,
-      "step": 2630
-    },
-    {
-      "epoch": 0.8267467656124232,
-      "grad_norm": 0.82421875,
-      "learning_rate": 1.686194858774992e-05,
-      "loss": 1.3216,
-      "step": 2632
-    },
-    {
-      "epoch": 0.8273749926379645,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.6859409711202796e-05,
-      "loss": 1.3221,
-      "step": 2634
-    },
-    {
-      "epoch": 0.8280032196635059,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.6856870834655664e-05,
-      "loss": 1.3614,
-      "step": 2636
-    },
-    {
-      "epoch": 0.8286314466890472,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.685433195810854e-05,
-      "loss": 1.3956,
-      "step": 2638
-    },
-    {
-      "epoch": 0.8292596737145886,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.685179308156141e-05,
-      "loss": 1.1662,
-      "step": 2640
-    },
-    {
-      "epoch": 0.82988790074013,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.684925420501428e-05,
-      "loss": 1.2505,
-      "step": 2642
-    },
-    {
-      "epoch": 0.8305161277656713,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.6846715328467153e-05,
-      "loss": 1.251,
-      "step": 2644
-    },
-    {
-      "epoch": 0.8311443547912126,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.6844176451920028e-05,
-      "loss": 1.398,
-      "step": 2646
-    },
-    {
-      "epoch": 0.8317725818167541,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.68416375753729e-05,
-      "loss": 1.2618,
-      "step": 2648
-    },
-    {
-      "epoch": 0.8324008088422954,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.683909869882577e-05,
-      "loss": 1.3516,
-      "step": 2650
-    },
-    {
-      "epoch": 0.8330290358678367,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.6836559822278645e-05,
-      "loss": 1.4359,
-      "step": 2652
-    },
-    {
-      "epoch": 0.833657262893378,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6834020945731516e-05,
-      "loss": 1.3158,
-      "step": 2654
-    },
-    {
-      "epoch": 0.8342854899189195,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.6831482069184388e-05,
-      "loss": 1.2849,
-      "step": 2656
-    },
-    {
-      "epoch": 0.8349137169444608,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.682894319263726e-05,
-      "loss": 1.4921,
-      "step": 2658
-    },
-    {
-      "epoch": 0.8355419439700021,
-      "grad_norm": 0.94921875,
-      "learning_rate": 1.6826404316090134e-05,
-      "loss": 1.2774,
-      "step": 2660
-    },
-    {
-      "epoch": 0.8361701709955435,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.6823865439543002e-05,
-      "loss": 1.3282,
-      "step": 2662
-    },
-    {
-      "epoch": 0.8367983980210849,
-      "grad_norm": 0.75,
-      "learning_rate": 1.6821326562995877e-05,
-      "loss": 1.2604,
-      "step": 2664
-    },
-    {
-      "epoch": 0.8374266250466262,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.6818787686448748e-05,
-      "loss": 1.2322,
-      "step": 2666
-    },
-    {
-      "epoch": 0.8380548520721676,
-      "grad_norm": 0.75,
-      "learning_rate": 1.681624880990162e-05,
-      "loss": 1.3847,
-      "step": 2668
-    },
-    {
-      "epoch": 0.8386830790977089,
-      "grad_norm": 1.0,
-      "learning_rate": 1.681370993335449e-05,
-      "loss": 1.2521,
-      "step": 2670
-    },
-    {
-      "epoch": 0.8393113061232503,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.6811171056807366e-05,
-      "loss": 1.4187,
-      "step": 2672
-    },
-    {
-      "epoch": 0.8399395331487917,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.6808632180260237e-05,
-      "loss": 1.195,
-      "step": 2674
-    },
-    {
-      "epoch": 0.840567760174333,
-      "grad_norm": 1.015625,
-      "learning_rate": 1.680609330371311e-05,
-      "loss": 1.3454,
-      "step": 2676
-    },
-    {
-      "epoch": 0.8411959871998743,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.680355442716598e-05,
-      "loss": 1.453,
-      "step": 2678
-    },
-    {
-      "epoch": 0.8418242142254158,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.6801015550618855e-05,
-      "loss": 1.4218,
-      "step": 2680
-    },
-    {
-      "epoch": 0.8424524412509571,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.6798476674071723e-05,
-      "loss": 1.4194,
-      "step": 2682
-    },
-    {
-      "epoch": 0.8430806682764984,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.6795937797524597e-05,
-      "loss": 1.3225,
-      "step": 2684
-    },
-    {
-      "epoch": 0.8437088953020397,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.679339892097747e-05,
-      "loss": 1.205,
-      "step": 2686
-    },
-    {
-      "epoch": 0.8443371223275812,
-      "grad_norm": 1.046875,
-      "learning_rate": 1.679086004443034e-05,
-      "loss": 1.425,
-      "step": 2688
-    },
-    {
-      "epoch": 0.8449653493531225,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.678832116788321e-05,
-      "loss": 1.3743,
-      "step": 2690
-    },
-    {
-      "epoch": 0.8455935763786638,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.6785782291336086e-05,
-      "loss": 1.2462,
-      "step": 2692
-    },
-    {
-      "epoch": 0.8462218034042052,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.6783243414788958e-05,
-      "loss": 1.3989,
-      "step": 2694
-    },
-    {
-      "epoch": 0.8468500304297466,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.678070453824183e-05,
-      "loss": 1.4101,
-      "step": 2696
-    },
-    {
-      "epoch": 0.8474782574552879,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.67781656616947e-05,
-      "loss": 1.2639,
-      "step": 2698
-    },
-    {
-      "epoch": 0.8481064844808293,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.6775626785147575e-05,
-      "loss": 1.3388,
-      "step": 2700
-    },
-    {
-      "epoch": 0.8487347115063706,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.6773087908600447e-05,
-      "loss": 1.363,
-      "step": 2702
-    },
-    {
-      "epoch": 0.8493629385319119,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.6770549032053318e-05,
-      "loss": 1.2831,
-      "step": 2704
-    },
-    {
-      "epoch": 0.8499911655574534,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.676801015550619e-05,
-      "loss": 1.2638,
-      "step": 2706
-    },
-    {
-      "epoch": 0.8506193925829947,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.676547127895906e-05,
-      "loss": 1.3733,
-      "step": 2708
-    },
-    {
-      "epoch": 0.851247619608536,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.6762932402411932e-05,
-      "loss": 1.3726,
-      "step": 2710
-    },
-    {
-      "epoch": 0.8518758466340773,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.6760393525864807e-05,
-      "loss": 1.3406,
-      "step": 2712
-    },
-    {
-      "epoch": 0.8525040736596188,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.6757854649317678e-05,
-      "loss": 1.4331,
-      "step": 2714
-    },
-    {
-      "epoch": 0.8531323006851601,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.675531577277055e-05,
-      "loss": 1.302,
-      "step": 2716
-    },
-    {
-      "epoch": 0.8537605277107014,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.675277689622342e-05,
-      "loss": 1.3428,
-      "step": 2718
-    },
-    {
-      "epoch": 0.8543887547362428,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.6750238019676296e-05,
-      "loss": 1.2827,
-      "step": 2720
-    },
-    {
-      "epoch": 0.8550169817617842,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.6747699143129167e-05,
-      "loss": 1.2744,
-      "step": 2722
-    },
-    {
-      "epoch": 0.8556452087873255,
-      "grad_norm": 0.75,
-      "learning_rate": 1.674516026658204e-05,
-      "loss": 1.2999,
-      "step": 2724
-    },
-    {
-      "epoch": 0.8562734358128669,
-      "grad_norm": 0.9765625,
-      "learning_rate": 1.674262139003491e-05,
-      "loss": 1.2288,
-      "step": 2726
-    },
-    {
-      "epoch": 0.8569016628384082,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.6740082513487785e-05,
-      "loss": 1.3101,
-      "step": 2728
-    },
-    {
-      "epoch": 0.8575298898639496,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.6737543636940653e-05,
-      "loss": 1.309,
-      "step": 2730
-    },
-    {
-      "epoch": 0.858158116889491,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.6735004760393527e-05,
-      "loss": 1.3683,
-      "step": 2732
-    },
-    {
-      "epoch": 0.8587863439150323,
-      "grad_norm": 1.015625,
-      "learning_rate": 1.67324658838464e-05,
-      "loss": 1.2708,
-      "step": 2734
-    },
-    {
-      "epoch": 0.8594145709405736,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.672992700729927e-05,
-      "loss": 1.5443,
-      "step": 2736
-    },
-    {
-      "epoch": 0.860042797966115,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.6727388130752145e-05,
-      "loss": 1.3305,
-      "step": 2738
-    },
-    {
-      "epoch": 0.8606710249916564,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.6724849254205016e-05,
-      "loss": 1.3512,
-      "step": 2740
-    },
-    {
-      "epoch": 0.8612992520171977,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.6722310377657888e-05,
-      "loss": 1.3854,
-      "step": 2742
-    },
-    {
-      "epoch": 0.861927479042739,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.671977150111076e-05,
-      "loss": 1.2901,
-      "step": 2744
-    },
-    {
-      "epoch": 0.8625557060682805,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.6717232624563634e-05,
-      "loss": 1.3502,
-      "step": 2746
-    },
-    {
-      "epoch": 0.8631839330938218,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.6714693748016505e-05,
-      "loss": 1.1293,
-      "step": 2748
-    },
-    {
-      "epoch": 0.8638121601193631,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.6712154871469377e-05,
-      "loss": 1.3325,
-      "step": 2750
-    },
-    {
-      "epoch": 0.8644403871449045,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.6709615994922248e-05,
-      "loss": 1.4138,
-      "step": 2752
-    },
-    {
-      "epoch": 0.8650686141704459,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.6707077118375123e-05,
-      "loss": 1.2818,
-      "step": 2754
-    },
-    {
-      "epoch": 0.8656968411959872,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.670453824182799e-05,
-      "loss": 1.2926,
-      "step": 2756
-    },
-    {
-      "epoch": 0.8663250682215285,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6701999365280866e-05,
-      "loss": 1.3686,
-      "step": 2758
-    },
-    {
-      "epoch": 0.8669532952470699,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.6699460488733737e-05,
-      "loss": 1.2924,
-      "step": 2760
-    },
-    {
-      "epoch": 0.8675815222726113,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.669692161218661e-05,
-      "loss": 1.4022,
-      "step": 2762
-    },
-    {
-      "epoch": 0.8682097492981526,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.669438273563948e-05,
-      "loss": 1.429,
-      "step": 2764
-    },
-    {
-      "epoch": 0.868837976323694,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.6691843859092355e-05,
-      "loss": 1.2911,
-      "step": 2766
-    },
-    {
-      "epoch": 0.8694662033492353,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.6689304982545226e-05,
-      "loss": 1.4,
-      "step": 2768
-    },
-    {
-      "epoch": 0.8700944303747767,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.6686766105998097e-05,
-      "loss": 1.3409,
-      "step": 2770
-    },
-    {
-      "epoch": 0.8707226574003181,
-      "grad_norm": 1.0390625,
-      "learning_rate": 1.668422722945097e-05,
-      "loss": 1.2781,
-      "step": 2772
-    },
-    {
-      "epoch": 0.8713508844258594,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.6681688352903843e-05,
-      "loss": 1.3083,
-      "step": 2774
-    },
-    {
-      "epoch": 0.8719791114514007,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.667914947635671e-05,
-      "loss": 1.2491,
-      "step": 2776
-    },
-    {
-      "epoch": 0.872607338476942,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.6676610599809586e-05,
-      "loss": 1.3156,
-      "step": 2778
-    },
-    {
-      "epoch": 0.8732355655024835,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.6674071723262458e-05,
-      "loss": 1.2403,
-      "step": 2780
-    },
-    {
-      "epoch": 0.8738637925280248,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.667153284671533e-05,
-      "loss": 1.4226,
-      "step": 2782
-    },
-    {
-      "epoch": 0.8744920195535661,
-      "grad_norm": 0.84765625,
-      "learning_rate": 1.66689939701682e-05,
-      "loss": 1.2981,
-      "step": 2784
-    },
-    {
-      "epoch": 0.8751202465791075,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6666455093621075e-05,
-      "loss": 1.256,
-      "step": 2786
-    },
-    {
-      "epoch": 0.8757484736046489,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.6663916217073946e-05,
-      "loss": 1.255,
-      "step": 2788
-    },
-    {
-      "epoch": 0.8763767006301902,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.6661377340526818e-05,
-      "loss": 1.2185,
-      "step": 2790
-    },
-    {
-      "epoch": 0.8770049276557316,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.665883846397969e-05,
-      "loss": 1.4315,
-      "step": 2792
-    },
-    {
-      "epoch": 0.8776331546812729,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6656299587432564e-05,
-      "loss": 1.4531,
-      "step": 2794
-    },
-    {
-      "epoch": 0.8782613817068143,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.6653760710885435e-05,
-      "loss": 1.2937,
-      "step": 2796
-    },
-    {
-      "epoch": 0.8788896087323557,
-      "grad_norm": 0.9375,
-      "learning_rate": 1.6651221834338307e-05,
-      "loss": 1.2382,
-      "step": 2798
-    },
-    {
-      "epoch": 0.879517835757897,
-      "grad_norm": 0.8515625,
-      "learning_rate": 1.6648682957791178e-05,
-      "loss": 1.2398,
-      "step": 2800
-    },
-    {
-      "epoch": 0.8801460627834383,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.664614408124405e-05,
-      "loss": 1.3117,
-      "step": 2802
-    },
-    {
-      "epoch": 0.8807742898089798,
-      "grad_norm": 0.88671875,
-      "learning_rate": 1.664360520469692e-05,
-      "loss": 1.35,
-      "step": 2804
-    },
-    {
-      "epoch": 0.8814025168345211,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.6641066328149796e-05,
-      "loss": 1.4186,
-      "step": 2806
-    },
-    {
-      "epoch": 0.8820307438600624,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.6638527451602667e-05,
-      "loss": 1.2733,
-      "step": 2808
-    },
-    {
-      "epoch": 0.8826589708856037,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.663598857505554e-05,
-      "loss": 1.312,
-      "step": 2810
-    },
-    {
-      "epoch": 0.8832871979111452,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.663344969850841e-05,
-      "loss": 1.2711,
-      "step": 2812
-    },
-    {
-      "epoch": 0.8839154249366865,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.6630910821961285e-05,
-      "loss": 1.3649,
-      "step": 2814
-    },
-    {
-      "epoch": 0.8845436519622278,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.6628371945414156e-05,
-      "loss": 1.5247,
-      "step": 2816
-    },
-    {
-      "epoch": 0.8851718789877692,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.6625833068867027e-05,
-      "loss": 1.2794,
-      "step": 2818
-    },
-    {
-      "epoch": 0.8858001060133106,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.66232941923199e-05,
-      "loss": 1.2475,
-      "step": 2820
-    },
-    {
-      "epoch": 0.8864283330388519,
-      "grad_norm": 0.8359375,
-      "learning_rate": 1.6620755315772774e-05,
-      "loss": 1.1927,
-      "step": 2822
-    },
-    {
-      "epoch": 0.8870565600643933,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.6618216439225645e-05,
-      "loss": 1.2817,
-      "step": 2824
-    },
-    {
-      "epoch": 0.8876847870899346,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.6615677562678516e-05,
-      "loss": 1.3958,
-      "step": 2826
-    },
-    {
-      "epoch": 0.888313014115476,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.6613138686131388e-05,
-      "loss": 1.4729,
-      "step": 2828
-    },
-    {
-      "epoch": 0.8889412411410174,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.661059980958426e-05,
-      "loss": 1.2705,
-      "step": 2830
-    },
-    {
-      "epoch": 0.8895694681665587,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.6608060933037134e-05,
-      "loss": 1.4104,
-      "step": 2832
-    },
-    {
-      "epoch": 0.8901976951921,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.6605522056490005e-05,
-      "loss": 1.3625,
-      "step": 2834
-    },
-    {
-      "epoch": 0.8908259222176415,
-      "grad_norm": 0.8046875,
-      "learning_rate": 1.6602983179942877e-05,
-      "loss": 1.3754,
-      "step": 2836
-    },
-    {
-      "epoch": 0.8914541492431828,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.6600444303395748e-05,
-      "loss": 1.2763,
-      "step": 2838
-    },
-    {
-      "epoch": 0.8920823762687241,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6597905426848623e-05,
-      "loss": 1.1745,
-      "step": 2840
-    },
-    {
-      "epoch": 0.8927106032942654,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.6595366550301494e-05,
-      "loss": 1.2782,
-      "step": 2842
-    },
-    {
-      "epoch": 0.8933388303198068,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.6592827673754366e-05,
-      "loss": 1.3032,
-      "step": 2844
-    },
-    {
-      "epoch": 0.8939670573453482,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.6590288797207237e-05,
-      "loss": 1.4176,
-      "step": 2846
-    },
-    {
-      "epoch": 0.8945952843708895,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.658774992066011e-05,
-      "loss": 1.3804,
-      "step": 2848
-    },
-    {
-      "epoch": 0.8952235113964309,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.658521104411298e-05,
-      "loss": 1.3173,
-      "step": 2850
-    },
-    {
-      "epoch": 0.8958517384219722,
-      "grad_norm": 0.86328125,
-      "learning_rate": 1.6582672167565854e-05,
-      "loss": 1.1673,
-      "step": 2852
-    },
-    {
-      "epoch": 0.8964799654475136,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.6580133291018726e-05,
-      "loss": 1.3201,
-      "step": 2854
-    },
-    {
-      "epoch": 0.897108192473055,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6577594414471597e-05,
-      "loss": 1.3234,
-      "step": 2856
-    },
-    {
-      "epoch": 0.8977364194985963,
-      "grad_norm": 2.359375,
-      "learning_rate": 1.657505553792447e-05,
-      "loss": 1.4672,
-      "step": 2858
-    },
-    {
-      "epoch": 0.8983646465241376,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.6572516661377343e-05,
-      "loss": 1.3377,
-      "step": 2860
-    },
-    {
-      "epoch": 0.898992873549679,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.6569977784830215e-05,
-      "loss": 1.2545,
-      "step": 2862
-    },
-    {
-      "epoch": 0.8996211005752204,
-      "grad_norm": 0.8984375,
-      "learning_rate": 1.6567438908283086e-05,
-      "loss": 1.2684,
-      "step": 2864
-    },
-    {
-      "epoch": 0.9002493276007617,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.6564900031735957e-05,
-      "loss": 1.2329,
-      "step": 2866
-    },
-    {
-      "epoch": 0.900877554626303,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.6562361155188832e-05,
-      "loss": 1.4101,
-      "step": 2868
-    },
-    {
-      "epoch": 0.9015057816518445,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.65598222786417e-05,
-      "loss": 1.2965,
-      "step": 2870
-    },
-    {
-      "epoch": 0.9021340086773858,
-      "grad_norm": 0.90234375,
-      "learning_rate": 1.6557283402094575e-05,
-      "loss": 1.331,
-      "step": 2872
-    },
-    {
-      "epoch": 0.9027622357029271,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.6554744525547446e-05,
-      "loss": 1.4061,
-      "step": 2874
-    },
-    {
-      "epoch": 0.9033904627284685,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.6552205649000318e-05,
-      "loss": 1.3482,
-      "step": 2876
-    },
-    {
-      "epoch": 0.9040186897540099,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.654966677245319e-05,
-      "loss": 1.3822,
-      "step": 2878
-    },
-    {
-      "epoch": 0.9046469167795512,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.6547127895906064e-05,
-      "loss": 1.2013,
-      "step": 2880
-    },
-    {
-      "epoch": 0.9052751438050926,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.6544589019358935e-05,
-      "loss": 1.2415,
-      "step": 2882
-    },
-    {
-      "epoch": 0.9059033708306339,
-      "grad_norm": 0.8984375,
-      "learning_rate": 1.6542050142811807e-05,
-      "loss": 1.3142,
-      "step": 2884
-    },
-    {
-      "epoch": 0.9065315978561753,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.6539511266264678e-05,
-      "loss": 1.3292,
-      "step": 2886
-    },
-    {
-      "epoch": 0.9071598248817166,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.6536972389717553e-05,
-      "loss": 1.3243,
-      "step": 2888
-    },
-    {
-      "epoch": 0.907788051907258,
-      "grad_norm": 0.75,
-      "learning_rate": 1.6534433513170424e-05,
-      "loss": 1.2548,
-      "step": 2890
-    },
-    {
-      "epoch": 0.9084162789327993,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.6531894636623296e-05,
-      "loss": 1.3526,
-      "step": 2892
-    },
-    {
-      "epoch": 0.9090445059583407,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.6529355760076167e-05,
-      "loss": 1.3198,
-      "step": 2894
-    },
-    {
-      "epoch": 0.9096727329838821,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.652681688352904e-05,
-      "loss": 1.0987,
-      "step": 2896
-    },
-    {
-      "epoch": 0.9103009600094234,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.652427800698191e-05,
-      "loss": 1.2387,
-      "step": 2898
-    },
-    {
-      "epoch": 0.9109291870349647,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.6521739130434785e-05,
-      "loss": 1.1774,
-      "step": 2900
-    },
-    {
-      "epoch": 0.9115574140605062,
-      "grad_norm": 0.78515625,
-      "learning_rate": 1.6519200253887656e-05,
-      "loss": 1.2341,
-      "step": 2902
-    },
-    {
-      "epoch": 0.9121856410860475,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.6516661377340527e-05,
-      "loss": 1.2046,
-      "step": 2904
-    },
-    {
-      "epoch": 0.9128138681115888,
-      "grad_norm": 0.7890625,
-      "learning_rate": 1.65141225007934e-05,
-      "loss": 1.477,
-      "step": 2906
-    },
-    {
-      "epoch": 0.9134420951371302,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.6511583624246273e-05,
-      "loss": 1.4045,
-      "step": 2908
-    },
-    {
-      "epoch": 0.9140703221626716,
-      "grad_norm": 0.7734375,
-      "learning_rate": 1.6509044747699145e-05,
-      "loss": 1.2798,
-      "step": 2910
-    },
-    {
-      "epoch": 0.9146985491882129,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6506505871152016e-05,
-      "loss": 1.3729,
-      "step": 2912
-    },
-    {
-      "epoch": 0.9153267762137542,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.650396699460489e-05,
-      "loss": 1.3434,
-      "step": 2914
-    },
-    {
-      "epoch": 0.9159550032392956,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.650142811805776e-05,
-      "loss": 1.3866,
-      "step": 2916
-    },
-    {
-      "epoch": 0.9165832302648369,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.6498889241510634e-05,
-      "loss": 1.2233,
-      "step": 2918
-    },
-    {
-      "epoch": 0.9172114572903783,
-      "grad_norm": 0.95703125,
-      "learning_rate": 1.6496350364963505e-05,
-      "loss": 1.2631,
-      "step": 2920
-    },
-    {
-      "epoch": 0.9178396843159197,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.6493811488416377e-05,
-      "loss": 1.3768,
-      "step": 2922
-    },
-    {
-      "epoch": 0.918467911341461,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.6491272611869248e-05,
-      "loss": 1.2722,
-      "step": 2924
-    },
-    {
-      "epoch": 0.9190961383670023,
-      "grad_norm": 0.83203125,
-      "learning_rate": 1.6488733735322123e-05,
-      "loss": 1.2799,
-      "step": 2926
-    },
-    {
-      "epoch": 0.9197243653925438,
-      "grad_norm": 0.859375,
-      "learning_rate": 1.6486194858774994e-05,
-      "loss": 1.2571,
-      "step": 2928
-    },
-    {
-      "epoch": 0.9203525924180851,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.6483655982227865e-05,
-      "loss": 1.2148,
-      "step": 2930
-    },
-    {
-      "epoch": 0.9209808194436264,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.6481117105680737e-05,
-      "loss": 1.3129,
-      "step": 2932
-    },
-    {
-      "epoch": 0.9216090464691677,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.647857822913361e-05,
-      "loss": 1.2683,
-      "step": 2934
-    },
-    {
-      "epoch": 0.9222372734947092,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6476039352586483e-05,
-      "loss": 1.356,
-      "step": 2936
-    },
-    {
-      "epoch": 0.9228655005202505,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.6473500476039354e-05,
-      "loss": 1.2901,
-      "step": 2938
-    },
-    {
-      "epoch": 0.9234937275457918,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.6470961599492226e-05,
-      "loss": 1.4158,
-      "step": 2940
-    },
-    {
-      "epoch": 0.9241219545713332,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.6468422722945097e-05,
-      "loss": 1.2391,
-      "step": 2942
-    },
-    {
-      "epoch": 0.9247501815968746,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.646588384639797e-05,
-      "loss": 1.3964,
-      "step": 2944
-    },
-    {
-      "epoch": 0.9253784086224159,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.6463344969850843e-05,
-      "loss": 1.3187,
-      "step": 2946
-    },
-    {
-      "epoch": 0.9260066356479573,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.6460806093303715e-05,
-      "loss": 1.3794,
-      "step": 2948
-    },
-    {
-      "epoch": 0.9266348626734986,
-      "grad_norm": 0.69140625,
-      "learning_rate": 1.6458267216756586e-05,
-      "loss": 1.3897,
-      "step": 2950
-    },
-    {
-      "epoch": 0.92726308969904,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.6455728340209457e-05,
-      "loss": 1.2514,
-      "step": 2952
-    },
-    {
-      "epoch": 0.9278913167245814,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.6453189463662332e-05,
-      "loss": 1.2275,
-      "step": 2954
-    },
-    {
-      "epoch": 0.9285195437501227,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.6450650587115204e-05,
-      "loss": 1.3661,
-      "step": 2956
-    },
-    {
-      "epoch": 0.929147770775664,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6448111710568075e-05,
-      "loss": 1.3095,
-      "step": 2958
-    },
-    {
-      "epoch": 0.9297759978012055,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.6445572834020946e-05,
-      "loss": 1.4244,
-      "step": 2960
-    },
-    {
-      "epoch": 0.9304042248267468,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.644303395747382e-05,
-      "loss": 1.3683,
-      "step": 2962
-    },
-    {
-      "epoch": 0.9310324518522881,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.644049508092669e-05,
-      "loss": 1.512,
-      "step": 2964
-    },
-    {
-      "epoch": 0.9316606788778294,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.6437956204379564e-05,
-      "loss": 1.4732,
-      "step": 2966
-    },
-    {
-      "epoch": 0.9322889059033709,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.6435417327832435e-05,
-      "loss": 1.34,
-      "step": 2968
-    },
-    {
-      "epoch": 0.9329171329289122,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.6432878451285307e-05,
-      "loss": 1.2436,
-      "step": 2970
-    },
-    {
-      "epoch": 0.9335453599544535,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.6430339574738178e-05,
-      "loss": 1.3719,
-      "step": 2972
-    },
-    {
-      "epoch": 0.9341735869799949,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.6427800698191053e-05,
-      "loss": 1.3081,
-      "step": 2974
-    },
-    {
-      "epoch": 0.9348018140055363,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.6425261821643924e-05,
-      "loss": 1.3141,
-      "step": 2976
-    },
-    {
-      "epoch": 0.9354300410310776,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.6422722945096796e-05,
-      "loss": 1.3385,
-      "step": 2978
-    },
-    {
-      "epoch": 0.936058268056619,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.6420184068549667e-05,
-      "loss": 1.3452,
-      "step": 2980
-    },
-    {
-      "epoch": 0.9366864950821603,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.6417645192002542e-05,
-      "loss": 1.2058,
-      "step": 2982
-    },
-    {
-      "epoch": 0.9373147221077017,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.641510631545541e-05,
-      "loss": 1.3226,
-      "step": 2984
-    },
-    {
-      "epoch": 0.937942949133243,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.6412567438908284e-05,
-      "loss": 1.362,
-      "step": 2986
-    },
-    {
-      "epoch": 0.9385711761587844,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.6410028562361156e-05,
-      "loss": 1.2904,
-      "step": 2988
-    },
-    {
-      "epoch": 0.9391994031843257,
-      "grad_norm": 0.875,
-      "learning_rate": 1.6407489685814027e-05,
-      "loss": 1.3076,
-      "step": 2990
-    },
-    {
-      "epoch": 0.939827630209867,
-      "grad_norm": 1.2890625,
-      "learning_rate": 1.64049508092669e-05,
-      "loss": 1.2634,
-      "step": 2992
-    },
-    {
-      "epoch": 0.9404558572354085,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.6402411932719773e-05,
-      "loss": 1.3545,
-      "step": 2994
-    },
-    {
-      "epoch": 0.9410840842609498,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.6399873056172645e-05,
-      "loss": 1.3319,
-      "step": 2996
-    },
-    {
-      "epoch": 0.9417123112864911,
-      "grad_norm": 0.75,
-      "learning_rate": 1.6397334179625516e-05,
-      "loss": 1.3534,
-      "step": 2998
-    },
-    {
-      "epoch": 0.9423405383120325,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.639479530307839e-05,
-      "loss": 1.2353,
-      "step": 3000
-    },
-    {
-      "epoch": 0.9429687653375739,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.6392256426531262e-05,
-      "loss": 1.2108,
-      "step": 3002
-    },
-    {
-      "epoch": 0.9435969923631152,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.6389717549984134e-05,
-      "loss": 1.2961,
-      "step": 3004
-    },
-    {
-      "epoch": 0.9442252193886566,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.6387178673437005e-05,
-      "loss": 1.2746,
-      "step": 3006
-    },
-    {
-      "epoch": 0.9448534464141979,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.638463979688988e-05,
-      "loss": 1.2231,
-      "step": 3008
-    },
-    {
-      "epoch": 0.9454816734397393,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.6382100920342748e-05,
-      "loss": 1.4304,
-      "step": 3010
-    },
-    {
-      "epoch": 0.9461099004652807,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.6379562043795623e-05,
-      "loss": 1.336,
-      "step": 3012
-    },
-    {
-      "epoch": 0.946738127490822,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.6377023167248494e-05,
-      "loss": 1.2532,
-      "step": 3014
-    },
-    {
-      "epoch": 0.9473663545163633,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.6374484290701365e-05,
-      "loss": 1.3438,
-      "step": 3016
-    },
-    {
-      "epoch": 0.9479945815419047,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.6371945414154237e-05,
-      "loss": 1.3412,
-      "step": 3018
-    },
-    {
-      "epoch": 0.9486228085674461,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.636940653760711e-05,
-      "loss": 1.2943,
-      "step": 3020
-    },
-    {
-      "epoch": 0.9492510355929874,
-      "grad_norm": 0.71484375,
-      "learning_rate": 1.6366867661059983e-05,
-      "loss": 1.3679,
-      "step": 3022
-    },
-    {
-      "epoch": 0.9498792626185287,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6364328784512854e-05,
-      "loss": 1.2899,
-      "step": 3024
-    },
-    {
-      "epoch": 0.9505074896440702,
-      "grad_norm": 0.87109375,
-      "learning_rate": 1.6361789907965726e-05,
-      "loss": 1.4329,
-      "step": 3026
-    },
-    {
-      "epoch": 0.9511357166696115,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.63592510314186e-05,
-      "loss": 1.2883,
-      "step": 3028
-    },
-    {
-      "epoch": 0.9517639436951528,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.6356712154871472e-05,
-      "loss": 1.2225,
-      "step": 3030
-    },
-    {
-      "epoch": 0.9523921707206942,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.6354173278324343e-05,
-      "loss": 1.3128,
-      "step": 3032
-    },
-    {
-      "epoch": 0.9530203977462356,
-      "grad_norm": 0.8828125,
-      "learning_rate": 1.6351634401777215e-05,
-      "loss": 1.3125,
-      "step": 3034
-    },
-    {
-      "epoch": 0.9536486247717769,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.6349095525230086e-05,
-      "loss": 1.3294,
-      "step": 3036
-    },
-    {
-      "epoch": 0.9542768517973182,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.6346556648682957e-05,
-      "loss": 1.2799,
-      "step": 3038
-    },
-    {
-      "epoch": 0.9549050788228596,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6344017772135832e-05,
-      "loss": 1.3259,
-      "step": 3040
-    },
-    {
-      "epoch": 0.955533305848401,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.6341478895588704e-05,
-      "loss": 1.2925,
-      "step": 3042
-    },
-    {
-      "epoch": 0.9561615328739423,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.6338940019041575e-05,
-      "loss": 1.4347,
-      "step": 3044
-    },
-    {
-      "epoch": 0.9567897598994837,
-      "grad_norm": 0.71875,
-      "learning_rate": 1.6336401142494446e-05,
-      "loss": 1.1746,
-      "step": 3046
-    },
-    {
-      "epoch": 0.957417986925025,
-      "grad_norm": 0.8671875,
-      "learning_rate": 1.633386226594732e-05,
-      "loss": 1.3328,
-      "step": 3048
-    },
-    {
-      "epoch": 0.9580462139505664,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6331323389400192e-05,
-      "loss": 1.3283,
-      "step": 3050
-    },
-    {
-      "epoch": 0.9586744409761078,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.6328784512853064e-05,
-      "loss": 1.4036,
-      "step": 3052
-    },
-    {
-      "epoch": 0.9593026680016491,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.6326245636305935e-05,
-      "loss": 1.2255,
-      "step": 3054
-    },
-    {
-      "epoch": 0.9599308950271904,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.632370675975881e-05,
-      "loss": 1.2382,
-      "step": 3056
-    },
-    {
-      "epoch": 0.9605591220527318,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.6321167883211678e-05,
-      "loss": 1.311,
-      "step": 3058
-    },
-    {
-      "epoch": 0.9611873490782732,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.6318629006664553e-05,
-      "loss": 1.2916,
-      "step": 3060
-    },
-    {
-      "epoch": 0.9618155761038145,
-      "grad_norm": 0.734375,
-      "learning_rate": 1.6316090130117424e-05,
-      "loss": 1.2996,
-      "step": 3062
-    },
-    {
-      "epoch": 0.9624438031293558,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.6313551253570295e-05,
-      "loss": 1.3253,
-      "step": 3064
-    },
-    {
-      "epoch": 0.9630720301548972,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.6311012377023167e-05,
-      "loss": 1.3582,
-      "step": 3066
-    },
-    {
-      "epoch": 0.9637002571804386,
-      "grad_norm": 0.671875,
-      "learning_rate": 1.630847350047604e-05,
-      "loss": 1.3637,
-      "step": 3068
-    },
-    {
-      "epoch": 0.9643284842059799,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.6305934623928913e-05,
-      "loss": 1.2882,
-      "step": 3070
-    },
-    {
-      "epoch": 0.9649567112315213,
-      "grad_norm": 1.1171875,
-      "learning_rate": 1.6303395747381784e-05,
-      "loss": 1.2281,
-      "step": 3072
-    },
-    {
-      "epoch": 0.9655849382570626,
-      "grad_norm": 0.80078125,
-      "learning_rate": 1.6300856870834656e-05,
-      "loss": 1.3915,
-      "step": 3074
-    },
-    {
-      "epoch": 0.966213165282604,
-      "grad_norm": 0.75,
-      "learning_rate": 1.629831799428753e-05,
-      "loss": 1.401,
-      "step": 3076
-    },
-    {
-      "epoch": 0.9668413923081454,
-      "grad_norm": 0.73828125,
-      "learning_rate": 1.62957791177404e-05,
-      "loss": 1.2698,
-      "step": 3078
-    },
-    {
-      "epoch": 0.9674696193336867,
-      "grad_norm": 0.75,
-      "learning_rate": 1.6293240241193273e-05,
-      "loss": 1.3833,
-      "step": 3080
-    },
-    {
-      "epoch": 0.968097846359228,
-      "grad_norm": 0.84375,
-      "learning_rate": 1.6290701364646148e-05,
-      "loss": 1.2167,
-      "step": 3082
-    },
-    {
-      "epoch": 0.9687260733847695,
-      "grad_norm": 0.6875,
-      "learning_rate": 1.6288162488099016e-05,
-      "loss": 1.3872,
-      "step": 3084
-    },
-    {
-      "epoch": 0.9693543004103108,
-      "grad_norm": 0.65625,
-      "learning_rate": 1.628562361155189e-05,
-      "loss": 1.1942,
-      "step": 3086
-    },
-    {
-      "epoch": 0.9699825274358521,
-      "grad_norm": 0.91015625,
-      "learning_rate": 1.6283084735004762e-05,
-      "loss": 1.3442,
-      "step": 3088
-    },
-    {
-      "epoch": 0.9706107544613934,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.6280545858457634e-05,
-      "loss": 1.1895,
-      "step": 3090
-    },
-    {
-      "epoch": 0.9712389814869349,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.6278006981910505e-05,
-      "loss": 1.2817,
-      "step": 3092
-    },
-    {
-      "epoch": 0.9718672085124762,
-      "grad_norm": 0.85546875,
-      "learning_rate": 1.627546810536338e-05,
-      "loss": 1.3469,
-      "step": 3094
-    },
-    {
-      "epoch": 0.9724954355380175,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.627292922881625e-05,
-      "loss": 1.3717,
-      "step": 3096
-    },
-    {
-      "epoch": 0.9731236625635589,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6270390352269123e-05,
-      "loss": 1.3015,
-      "step": 3098
-    },
-    {
-      "epoch": 0.9737518895891003,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6267851475721994e-05,
-      "loss": 1.4265,
-      "step": 3100
-    },
-    {
-      "epoch": 0.9743801166146416,
-      "grad_norm": 0.80859375,
-      "learning_rate": 1.626531259917487e-05,
-      "loss": 1.2424,
-      "step": 3102
-    },
-    {
-      "epoch": 0.975008343640183,
-      "grad_norm": 0.8125,
-      "learning_rate": 1.6262773722627737e-05,
-      "loss": 1.3252,
-      "step": 3104
-    },
-    {
-      "epoch": 0.9756365706657243,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.626023484608061e-05,
-      "loss": 1.3464,
-      "step": 3106
-    },
-    {
-      "epoch": 0.9762647976912657,
-      "grad_norm": 0.81640625,
-      "learning_rate": 1.6257695969533483e-05,
-      "loss": 1.3007,
-      "step": 3108
-    },
-    {
-      "epoch": 0.9768930247168071,
-      "grad_norm": 0.90234375,
-      "learning_rate": 1.6255157092986354e-05,
-      "loss": 1.2549,
-      "step": 3110
-    },
-    {
-      "epoch": 0.9775212517423484,
-      "grad_norm": 0.76171875,
-      "learning_rate": 1.6252618216439226e-05,
-      "loss": 1.3299,
-      "step": 3112
-    },
-    {
-      "epoch": 0.9781494787678897,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.62500793398921e-05,
-      "loss": 1.3118,
-      "step": 3114
-    },
-    {
-      "epoch": 0.9787777057934312,
-      "grad_norm": 0.765625,
-      "learning_rate": 1.6247540463344972e-05,
-      "loss": 1.3114,
-      "step": 3116
-    },
-    {
-      "epoch": 0.9794059328189725,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.6245001586797843e-05,
-      "loss": 1.2729,
-      "step": 3118
-    },
-    {
-      "epoch": 0.9800341598445138,
-      "grad_norm": 0.69921875,
-      "learning_rate": 1.6242462710250715e-05,
-      "loss": 1.3769,
-      "step": 3120
-    },
-    {
-      "epoch": 0.9806623868700551,
-      "grad_norm": 0.65234375,
-      "learning_rate": 1.623992383370359e-05,
-      "loss": 1.2581,
-      "step": 3122
-    },
-    {
-      "epoch": 0.9812906138955966,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.623738495715646e-05,
-      "loss": 1.3578,
-      "step": 3124
-    },
-    {
-      "epoch": 0.9819188409211379,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.6234846080609332e-05,
-      "loss": 1.3055,
-      "step": 3126
-    },
-    {
-      "epoch": 0.9825470679466792,
-      "grad_norm": 1.3125,
-      "learning_rate": 1.6232307204062203e-05,
-      "loss": 1.2103,
-      "step": 3128
-    },
-    {
-      "epoch": 0.9831752949722206,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.6229768327515075e-05,
-      "loss": 1.3818,
-      "step": 3130
-    },
-    {
-      "epoch": 0.9838035219977619,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.6227229450967946e-05,
-      "loss": 1.3531,
-      "step": 3132
-    },
-    {
-      "epoch": 0.9844317490233033,
-      "grad_norm": 0.73046875,
-      "learning_rate": 1.622469057442082e-05,
-      "loss": 1.2813,
-      "step": 3134
-    },
-    {
-      "epoch": 0.9850599760488447,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.6222151697873692e-05,
-      "loss": 1.3331,
-      "step": 3136
-    },
-    {
-      "epoch": 0.985688203074386,
-      "grad_norm": 0.72265625,
-      "learning_rate": 1.6219612821326564e-05,
-      "loss": 1.3681,
-      "step": 3138
-    },
-    {
-      "epoch": 0.9863164300999273,
-      "grad_norm": 0.7421875,
-      "learning_rate": 1.6217073944779435e-05,
-      "loss": 1.3545,
-      "step": 3140
-    },
-    {
-      "epoch": 0.9869446571254687,
-      "grad_norm": 0.703125,
-      "learning_rate": 1.621453506823231e-05,
-      "loss": 1.392,
-      "step": 3142
-    },
-    {
-      "epoch": 0.9875728841510101,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.621199619168518e-05,
-      "loss": 1.3232,
-      "step": 3144
-    },
-    {
-      "epoch": 0.9882011111765514,
-      "grad_norm": 0.7578125,
-      "learning_rate": 1.6209457315138053e-05,
-      "loss": 1.2144,
-      "step": 3146
-    },
-    {
-      "epoch": 0.9888293382020927,
-      "grad_norm": 0.6796875,
-      "learning_rate": 1.6206918438590924e-05,
-      "loss": 1.3129,
-      "step": 3148
-    },
-    {
-      "epoch": 0.9894575652276342,
-      "grad_norm": 0.74609375,
-      "learning_rate": 1.62043795620438e-05,
-      "loss": 1.3398,
-      "step": 3150
-    },
-    {
-      "epoch": 0.9900857922531755,
-      "grad_norm": 0.79296875,
-      "learning_rate": 1.6201840685496667e-05,
-      "loss": 1.3094,
-      "step": 3152
-    },
-    {
-      "epoch": 0.9907140192787168,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.619930180894954e-05,
-      "loss": 1.436,
-      "step": 3154
-    },
-    {
-      "epoch": 0.9913422463042582,
-      "grad_norm": 0.828125,
-      "learning_rate": 1.6196762932402413e-05,
-      "loss": 1.4225,
-      "step": 3156
-    },
-    {
-      "epoch": 0.9919704733297996,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.6194224055855284e-05,
-      "loss": 1.2521,
-      "step": 3158
-    },
-    {
-      "epoch": 0.9925987003553409,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.6191685179308156e-05,
-      "loss": 1.2926,
-      "step": 3160
-    },
-    {
-      "epoch": 0.9932269273808823,
-      "grad_norm": 0.66796875,
-      "learning_rate": 1.618914630276103e-05,
-      "loss": 1.222,
-      "step": 3162
-    },
-    {
-      "epoch": 0.9938551544064236,
-      "grad_norm": 0.87890625,
-      "learning_rate": 1.6186607426213902e-05,
-      "loss": 1.3083,
-      "step": 3164
-    },
-    {
-      "epoch": 0.994483381431965,
-      "grad_norm": 0.66015625,
-      "learning_rate": 1.6184068549666773e-05,
-      "loss": 1.3349,
-      "step": 3166
-    },
-    {
-      "epoch": 0.9951116084575063,
-      "grad_norm": 0.66015625,
-      "learning_rate": 1.6181529673119648e-05,
-      "loss": 1.306,
-      "step": 3168
-    },
-    {
-      "epoch": 0.9957398354830477,
-      "grad_norm": 0.796875,
-      "learning_rate": 1.617899079657252e-05,
-      "loss": 1.2053,
-      "step": 3170
-    },
-    {
-      "epoch": 0.996368062508589,
-      "grad_norm": 0.91015625,
-      "learning_rate": 1.617645192002539e-05,
-      "loss": 1.2241,
-      "step": 3172
-    },
-    {
-      "epoch": 0.9969962895341304,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.6173913043478262e-05,
-      "loss": 1.1861,
-      "step": 3174
-    },
-    {
-      "epoch": 0.9976245165596718,
-      "grad_norm": 0.6953125,
-      "learning_rate": 1.6171374166931137e-05,
-      "loss": 1.3181,
-      "step": 3176
-    },
-    {
-      "epoch": 0.9982527435852131,
-      "grad_norm": 0.7265625,
-      "learning_rate": 1.6168835290384005e-05,
-      "loss": 1.2654,
-      "step": 3178
-    },
-    {
-      "epoch": 0.9988809706107544,
-      "grad_norm": 0.77734375,
-      "learning_rate": 1.616629641383688e-05,
-      "loss": 1.435,
-      "step": 3180
-    },
-    {
-      "epoch": 0.9995091976362959,
-      "grad_norm": 0.75,
-      "learning_rate": 1.616375753728975e-05,
-      "loss": 1.4596,
-      "step": 3182
-    },
-    {
-      "epoch": 1.000137424661837,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.6161218660742622e-05,
-      "loss": 1.3658,
-      "step": 3184
-    },
-    {
-      "epoch": 1.0007656516873786,
-      "grad_norm": 0.67578125,
-      "learning_rate": 1.6158679784195494e-05,
-      "loss": 1.278,
-      "step": 3186
-    },
-    {
-      "epoch": 1.00139387871292,
-      "grad_norm": 0.6640625,
-      "learning_rate": 1.615614090764837e-05,
-      "loss": 1.3165,
-      "step": 3188
-    },
-    {
-      "epoch": 1.0020221057384613,
-      "grad_norm": 0.68359375,
-      "learning_rate": 1.615360203110124e-05,
-      "loss": 1.292,
-      "step": 3190
-    },
-    {
-      "epoch": 1.0026503327640026,
-      "grad_norm": 0.7109375,
-      "learning_rate": 1.615106315455411e-05,
-      "loss": 1.2257,
-      "step": 3192
-    },
-    {
-      "epoch": 1.003278559789544,
-      "grad_norm": 0.78125,
-      "learning_rate": 1.6148524278006983e-05,
-      "loss": 1.256,
-      "step": 3194
-    },
-    {
-      "epoch": 1.0039067868150853,
-      "grad_norm": 0.83984375,
-      "learning_rate": 1.6145985401459858e-05,
-      "loss": 1.2126,
-      "step": 3196
-    },
-    {
-      "epoch": 1.0045350138406266,
-      "grad_norm": 0.75390625,
-      "learning_rate": 1.6143446524912726e-05,
-      "loss": 1.2712,
-      "step": 3198
-    },
-    {
-      "epoch": 1.005163240866168,
-      "grad_norm": 0.76953125,
-      "learning_rate": 1.61409076483656e-05,
-      "loss": 1.1845,
-      "step": 3200
-    }
-  ],
-  "logging_steps": 2,
-  "max_steps": 15915,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
-  "save_steps": 100,
-  "stateful_callbacks": {
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": false
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 8.666866260954317e+18,
-  "train_batch_size": 1,
-  "trial_name": null,
-  "trial_params": null
-}