diff --git "a/e3.0/trainer_state.json" "b/e3.0/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/e3.0/trainer_state.json"
@@ -0,0 +1,33283 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.984078371321436,
+  "eval_steps": 500,
+  "global_step": 9500,
+  "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
+    },
+    {
+      "epoch": 1.0057914678917093,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.613836877181847e-05,
+      "loss": 1.2542,
+      "step": 3202
+    },
+    {
+      "epoch": 1.0064196949172508,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.6135829895271343e-05,
+      "loss": 1.2691,
+      "step": 3204
+    },
+    {
+      "epoch": 1.0070479219427921,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6133291018724214e-05,
+      "loss": 1.1913,
+      "step": 3206
+    },
+    {
+      "epoch": 1.0076761489683335,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.613075214217709e-05,
+      "loss": 1.1205,
+      "step": 3208
+    },
+    {
+      "epoch": 1.0083043759938748,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.612821326562996e-05,
+      "loss": 1.2168,
+      "step": 3210
+    },
+    {
+      "epoch": 1.0089326030194161,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6125674389082832e-05,
+      "loss": 1.0832,
+      "step": 3212
+    },
+    {
+      "epoch": 1.0095608300449574,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6123135512535703e-05,
+      "loss": 1.2185,
+      "step": 3214
+    },
+    {
+      "epoch": 1.0101890570704988,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6120596635988578e-05,
+      "loss": 1.3084,
+      "step": 3216
+    },
+    {
+      "epoch": 1.01081728409604,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6118057759441446e-05,
+      "loss": 1.1159,
+      "step": 3218
+    },
+    {
+      "epoch": 1.0114455111215817,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.611551888289432e-05,
+      "loss": 1.165,
+      "step": 3220
+    },
+    {
+      "epoch": 1.012073738147123,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6112980006347192e-05,
+      "loss": 1.1485,
+      "step": 3222
+    },
+    {
+      "epoch": 1.0127019651726643,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.6110441129800064e-05,
+      "loss": 1.23,
+      "step": 3224
+    },
+    {
+      "epoch": 1.0133301921982056,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.6107902253252935e-05,
+      "loss": 1.1168,
+      "step": 3226
+    },
+    {
+      "epoch": 1.013958419223747,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.610536337670581e-05,
+      "loss": 1.0645,
+      "step": 3228
+    },
+    {
+      "epoch": 1.0145866462492883,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.610282450015868e-05,
+      "loss": 1.2309,
+      "step": 3230
+    },
+    {
+      "epoch": 1.0152148732748296,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6100285623611553e-05,
+      "loss": 1.2314,
+      "step": 3232
+    },
+    {
+      "epoch": 1.015843100300371,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6097746747064424e-05,
+      "loss": 1.3229,
+      "step": 3234
+    },
+    {
+      "epoch": 1.0164713273259125,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.60952078705173e-05,
+      "loss": 1.1837,
+      "step": 3236
+    },
+    {
+      "epoch": 1.0170995543514538,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.609266899397017e-05,
+      "loss": 1.2404,
+      "step": 3238
+    },
+    {
+      "epoch": 1.0177277813769952,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.609013011742304e-05,
+      "loss": 1.344,
+      "step": 3240
+    },
+    {
+      "epoch": 1.0183560084025365,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6087591240875913e-05,
+      "loss": 1.3881,
+      "step": 3242
+    },
+    {
+      "epoch": 1.0189842354280778,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.6085052364328784e-05,
+      "loss": 1.3539,
+      "step": 3244
+    },
+    {
+      "epoch": 1.0196124624536191,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6082513487781656e-05,
+      "loss": 1.2862,
+      "step": 3246
+    },
+    {
+      "epoch": 1.0202406894791605,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.607997461123453e-05,
+      "loss": 1.3068,
+      "step": 3248
+    },
+    {
+      "epoch": 1.0208689165047018,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.6077435734687402e-05,
+      "loss": 1.2403,
+      "step": 3250
+    },
+    {
+      "epoch": 1.0214971435302433,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.6074896858140273e-05,
+      "loss": 1.2603,
+      "step": 3252
+    },
+    {
+      "epoch": 1.0221253705557847,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6072357981593148e-05,
+      "loss": 1.1728,
+      "step": 3254
+    },
+    {
+      "epoch": 1.022753597581326,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.606981910504602e-05,
+      "loss": 1.1989,
+      "step": 3256
+    },
+    {
+      "epoch": 1.0233818246068673,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.606728022849889e-05,
+      "loss": 1.2041,
+      "step": 3258
+    },
+    {
+      "epoch": 1.0240100516324087,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.6064741351951762e-05,
+      "loss": 1.2197,
+      "step": 3260
+    },
+    {
+      "epoch": 1.02463827865795,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.6062202475404637e-05,
+      "loss": 1.1039,
+      "step": 3262
+    },
+    {
+      "epoch": 1.0252665056834913,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6059663598857508e-05,
+      "loss": 1.2697,
+      "step": 3264
+    },
+    {
+      "epoch": 1.0258947327090326,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.605712472231038e-05,
+      "loss": 1.152,
+      "step": 3266
+    },
+    {
+      "epoch": 1.026522959734574,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.605458584576325e-05,
+      "loss": 1.2728,
+      "step": 3268
+    },
+    {
+      "epoch": 1.0271511867601155,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.6052046969216122e-05,
+      "loss": 1.1901,
+      "step": 3270
+    },
+    {
+      "epoch": 1.0277794137856568,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6049508092668994e-05,
+      "loss": 1.2918,
+      "step": 3272
+    },
+    {
+      "epoch": 1.0284076408111982,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.604696921612187e-05,
+      "loss": 1.2328,
+      "step": 3274
+    },
+    {
+      "epoch": 1.0290358678367395,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.604443033957474e-05,
+      "loss": 1.2688,
+      "step": 3276
+    },
+    {
+      "epoch": 1.0296640948622808,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.604189146302761e-05,
+      "loss": 1.1562,
+      "step": 3278
+    },
+    {
+      "epoch": 1.0302923218878222,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6039352586480483e-05,
+      "loss": 1.2546,
+      "step": 3280
+    },
+    {
+      "epoch": 1.0309205489133635,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6036813709933357e-05,
+      "loss": 1.1542,
+      "step": 3282
+    },
+    {
+      "epoch": 1.0315487759389048,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.603427483338623e-05,
+      "loss": 1.2115,
+      "step": 3284
+    },
+    {
+      "epoch": 1.0321770029644464,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.60317359568391e-05,
+      "loss": 1.3227,
+      "step": 3286
+    },
+    {
+      "epoch": 1.0328052299899877,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.602919708029197e-05,
+      "loss": 1.2416,
+      "step": 3288
+    },
+    {
+      "epoch": 1.033433457015529,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.6026658203744846e-05,
+      "loss": 1.2146,
+      "step": 3290
+    },
+    {
+      "epoch": 1.0340616840410704,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.6024119327197714e-05,
+      "loss": 1.2237,
+      "step": 3292
+    },
+    {
+      "epoch": 1.0346899110666117,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.602158045065059e-05,
+      "loss": 1.1345,
+      "step": 3294
+    },
+    {
+      "epoch": 1.035318138092153,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.601904157410346e-05,
+      "loss": 1.1062,
+      "step": 3296
+    },
+    {
+      "epoch": 1.0359463651176943,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6016502697556332e-05,
+      "loss": 1.2563,
+      "step": 3298
+    },
+    {
+      "epoch": 1.0365745921432357,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.6013963821009203e-05,
+      "loss": 1.1443,
+      "step": 3300
+    },
+    {
+      "epoch": 1.0372028191687772,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.6011424944462078e-05,
+      "loss": 1.3163,
+      "step": 3302
+    },
+    {
+      "epoch": 1.0378310461943185,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.600888606791495e-05,
+      "loss": 1.2362,
+      "step": 3304
+    },
+    {
+      "epoch": 1.0384592732198599,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.600634719136782e-05,
+      "loss": 1.2361,
+      "step": 3306
+    },
+    {
+      "epoch": 1.0390875002454012,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.6003808314820692e-05,
+      "loss": 1.2125,
+      "step": 3308
+    },
+    {
+      "epoch": 1.0397157272709425,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.6001269438273567e-05,
+      "loss": 1.1605,
+      "step": 3310
+    },
+    {
+      "epoch": 1.0403439542964839,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.5998730561726435e-05,
+      "loss": 1.1425,
+      "step": 3312
+    },
+    {
+      "epoch": 1.0409721813220252,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.599619168517931e-05,
+      "loss": 1.1267,
+      "step": 3314
+    },
+    {
+      "epoch": 1.0416004083475665,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.599365280863218e-05,
+      "loss": 1.3268,
+      "step": 3316
+    },
+    {
+      "epoch": 1.042228635373108,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.5991113932085053e-05,
+      "loss": 1.2864,
+      "step": 3318
+    },
+    {
+      "epoch": 1.0428568623986494,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5988575055537924e-05,
+      "loss": 1.3753,
+      "step": 3320
+    },
+    {
+      "epoch": 1.0434850894241907,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.59860361789908e-05,
+      "loss": 1.279,
+      "step": 3322
+    },
+    {
+      "epoch": 1.044113316449732,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.598349730244367e-05,
+      "loss": 1.2186,
+      "step": 3324
+    },
+    {
+      "epoch": 1.0447415434752734,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.598095842589654e-05,
+      "loss": 1.2752,
+      "step": 3326
+    },
+    {
+      "epoch": 1.0453697705008147,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.5978419549349413e-05,
+      "loss": 1.0901,
+      "step": 3328
+    },
+    {
+      "epoch": 1.045997997526356,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5975880672802288e-05,
+      "loss": 1.2462,
+      "step": 3330
+    },
+    {
+      "epoch": 1.0466262245518974,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.597334179625516e-05,
+      "loss": 1.2196,
+      "step": 3332
+    },
+    {
+      "epoch": 1.0472544515774387,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.597080291970803e-05,
+      "loss": 1.181,
+      "step": 3334
+    },
+    {
+      "epoch": 1.0478826786029802,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5968264043160902e-05,
+      "loss": 1.2889,
+      "step": 3336
+    },
+    {
+      "epoch": 1.0485109056285216,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5965725166613773e-05,
+      "loss": 1.2918,
+      "step": 3338
+    },
+    {
+      "epoch": 1.049139132654063,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5963186290066648e-05,
+      "loss": 1.2074,
+      "step": 3340
+    },
+    {
+      "epoch": 1.0497673596796042,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.596064741351952e-05,
+      "loss": 1.0973,
+      "step": 3342
+    },
+    {
+      "epoch": 1.0503955867051455,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.595810853697239e-05,
+      "loss": 1.2965,
+      "step": 3344
+    },
+    {
+      "epoch": 1.0510238137306869,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5955569660425262e-05,
+      "loss": 1.1753,
+      "step": 3346
+    },
+    {
+      "epoch": 1.0516520407562282,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.5953030783878137e-05,
+      "loss": 1.1732,
+      "step": 3348
+    },
+    {
+      "epoch": 1.0522802677817698,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5950491907331008e-05,
+      "loss": 1.2432,
+      "step": 3350
+    },
+    {
+      "epoch": 1.052908494807311,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.594795303078388e-05,
+      "loss": 1.3405,
+      "step": 3352
+    },
+    {
+      "epoch": 1.0535367218328524,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.594541415423675e-05,
+      "loss": 1.2937,
+      "step": 3354
+    },
+    {
+      "epoch": 1.0541649488583937,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5942875277689626e-05,
+      "loss": 1.2403,
+      "step": 3356
+    },
+    {
+      "epoch": 1.054793175883935,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5940336401142497e-05,
+      "loss": 1.1935,
+      "step": 3358
+    },
+    {
+      "epoch": 1.0554214029094764,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.593779752459537e-05,
+      "loss": 1.1554,
+      "step": 3360
+    },
+    {
+      "epoch": 1.0560496299350177,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.593525864804824e-05,
+      "loss": 1.2863,
+      "step": 3362
+    },
+    {
+      "epoch": 1.056677856960559,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.593271977150111e-05,
+      "loss": 1.3232,
+      "step": 3364
+    },
+    {
+      "epoch": 1.0573060839861004,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.5930180894953983e-05,
+      "loss": 1.165,
+      "step": 3366
+    },
+    {
+      "epoch": 1.057934311011642,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5927642018406857e-05,
+      "loss": 1.2739,
+      "step": 3368
+    },
+    {
+      "epoch": 1.0585625380371833,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.592510314185973e-05,
+      "loss": 1.1609,
+      "step": 3370
+    },
+    {
+      "epoch": 1.0591907650627246,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.59225642653126e-05,
+      "loss": 1.3293,
+      "step": 3372
+    },
+    {
+      "epoch": 1.059818992088266,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.592002538876547e-05,
+      "loss": 1.2502,
+      "step": 3374
+    },
+    {
+      "epoch": 1.0604472191138072,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5917486512218346e-05,
+      "loss": 1.1905,
+      "step": 3376
+    },
+    {
+      "epoch": 1.0610754461393486,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5914947635671218e-05,
+      "loss": 1.3572,
+      "step": 3378
+    },
+    {
+      "epoch": 1.06170367316489,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.591240875912409e-05,
+      "loss": 1.2005,
+      "step": 3380
+    },
+    {
+      "epoch": 1.0623319001904312,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.590986988257696e-05,
+      "loss": 1.2663,
+      "step": 3382
+    },
+    {
+      "epoch": 1.0629601272159728,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5907331006029835e-05,
+      "loss": 1.2143,
+      "step": 3384
+    },
+    {
+      "epoch": 1.063588354241514,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5904792129482703e-05,
+      "loss": 1.2849,
+      "step": 3386
+    },
+    {
+      "epoch": 1.0642165812670554,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5902253252935578e-05,
+      "loss": 1.2787,
+      "step": 3388
+    },
+    {
+      "epoch": 1.0648448082925968,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.589971437638845e-05,
+      "loss": 1.1279,
+      "step": 3390
+    },
+    {
+      "epoch": 1.065473035318138,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.589717549984132e-05,
+      "loss": 1.2399,
+      "step": 3392
+    },
+    {
+      "epoch": 1.0661012623436794,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5894636623294192e-05,
+      "loss": 1.2965,
+      "step": 3394
+    },
+    {
+      "epoch": 1.0667294893692207,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5892097746747067e-05,
+      "loss": 1.1753,
+      "step": 3396
+    },
+    {
+      "epoch": 1.067357716394762,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5889558870199938e-05,
+      "loss": 1.229,
+      "step": 3398
+    },
+    {
+      "epoch": 1.0679859434203034,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.588701999365281e-05,
+      "loss": 1.2592,
+      "step": 3400
+    },
+    {
+      "epoch": 1.068614170445845,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.588448111710568e-05,
+      "loss": 1.2941,
+      "step": 3402
+    },
+    {
+      "epoch": 1.0692423974713863,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5881942240558556e-05,
+      "loss": 1.2962,
+      "step": 3404
+    },
+    {
+      "epoch": 1.0698706244969276,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5879403364011424e-05,
+      "loss": 1.2558,
+      "step": 3406
+    },
+    {
+      "epoch": 1.070498851522469,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.58768644874643e-05,
+      "loss": 1.3021,
+      "step": 3408
+    },
+    {
+      "epoch": 1.0711270785480103,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.587432561091717e-05,
+      "loss": 1.2742,
+      "step": 3410
+    },
+    {
+      "epoch": 1.0717553055735516,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.587178673437004e-05,
+      "loss": 1.2945,
+      "step": 3412
+    },
+    {
+      "epoch": 1.072383532599093,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5869247857822913e-05,
+      "loss": 1.323,
+      "step": 3414
+    },
+    {
+      "epoch": 1.0730117596246345,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5866708981275787e-05,
+      "loss": 1.2498,
+      "step": 3416
+    },
+    {
+      "epoch": 1.0736399866501758,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.586417010472866e-05,
+      "loss": 1.0935,
+      "step": 3418
+    },
+    {
+      "epoch": 1.0742682136757171,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.586163122818153e-05,
+      "loss": 1.1667,
+      "step": 3420
+    },
+    {
+      "epoch": 1.0748964407012584,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.58590923516344e-05,
+      "loss": 1.2579,
+      "step": 3422
+    },
+    {
+      "epoch": 1.0755246677267998,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5856553475087276e-05,
+      "loss": 1.1673,
+      "step": 3424
+    },
+    {
+      "epoch": 1.076152894752341,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5854014598540148e-05,
+      "loss": 1.411,
+      "step": 3426
+    },
+    {
+      "epoch": 1.0767811217778824,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.585147572199302e-05,
+      "loss": 1.1398,
+      "step": 3428
+    },
+    {
+      "epoch": 1.0774093488034238,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5848936845445894e-05,
+      "loss": 1.2636,
+      "step": 3430
+    },
+    {
+      "epoch": 1.078037575828965,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5846397968898762e-05,
+      "loss": 1.2364,
+      "step": 3432
+    },
+    {
+      "epoch": 1.0786658028545066,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5843859092351637e-05,
+      "loss": 1.4491,
+      "step": 3434
+    },
+    {
+      "epoch": 1.079294029880048,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5841320215804508e-05,
+      "loss": 1.1533,
+      "step": 3436
+    },
+    {
+      "epoch": 1.0799222569055893,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.583878133925738e-05,
+      "loss": 1.229,
+      "step": 3438
+    },
+    {
+      "epoch": 1.0805504839311306,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.583624246271025e-05,
+      "loss": 1.4361,
+      "step": 3440
+    },
+    {
+      "epoch": 1.081178710956672,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5833703586163126e-05,
+      "loss": 1.3107,
+      "step": 3442
+    },
+    {
+      "epoch": 1.0818069379822133,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.5831164709615997e-05,
+      "loss": 1.1652,
+      "step": 3444
+    },
+    {
+      "epoch": 1.0824351650077546,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.582862583306887e-05,
+      "loss": 1.2367,
+      "step": 3446
+    },
+    {
+      "epoch": 1.083063392033296,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.582608695652174e-05,
+      "loss": 1.178,
+      "step": 3448
+    },
+    {
+      "epoch": 1.0836916190588375,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.5823548079974615e-05,
+      "loss": 1.2211,
+      "step": 3450
+    },
+    {
+      "epoch": 1.0843198460843788,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.5821009203427486e-05,
+      "loss": 1.2025,
+      "step": 3452
+    },
+    {
+      "epoch": 1.0849480731099201,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5818470326880357e-05,
+      "loss": 1.2151,
+      "step": 3454
+    },
+    {
+      "epoch": 1.0855763001354615,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.581593145033323e-05,
+      "loss": 1.2384,
+      "step": 3456
+    },
+    {
+      "epoch": 1.0862045271610028,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.58133925737861e-05,
+      "loss": 1.2462,
+      "step": 3458
+    },
+    {
+      "epoch": 1.0868327541865441,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.581085369723897e-05,
+      "loss": 1.1743,
+      "step": 3460
+    },
+    {
+      "epoch": 1.0874609812120855,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5808314820691846e-05,
+      "loss": 1.2484,
+      "step": 3462
+    },
+    {
+      "epoch": 1.0880892082376268,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5805775944144718e-05,
+      "loss": 1.1292,
+      "step": 3464
+    },
+    {
+      "epoch": 1.088717435263168,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.580323706759759e-05,
+      "loss": 1.3436,
+      "step": 3466
+    },
+    {
+      "epoch": 1.0893456622887097,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.580069819105046e-05,
+      "loss": 1.106,
+      "step": 3468
+    },
+    {
+      "epoch": 1.089973889314251,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5798159314503335e-05,
+      "loss": 1.1988,
+      "step": 3470
+    },
+    {
+      "epoch": 1.0906021163397923,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.5795620437956207e-05,
+      "loss": 1.4247,
+      "step": 3472
+    },
+    {
+      "epoch": 1.0912303433653336,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5793081561409078e-05,
+      "loss": 1.2297,
+      "step": 3474
+    },
+    {
+      "epoch": 1.091858570390875,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.579054268486195e-05,
+      "loss": 1.1387,
+      "step": 3476
+    },
+    {
+      "epoch": 1.0924867974164163,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5788003808314824e-05,
+      "loss": 1.3101,
+      "step": 3478
+    },
+    {
+      "epoch": 1.0931150244419576,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.5785464931767692e-05,
+      "loss": 1.2612,
+      "step": 3480
+    },
+    {
+      "epoch": 1.0937432514674992,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5782926055220567e-05,
+      "loss": 1.1812,
+      "step": 3482
+    },
+    {
+      "epoch": 1.0943714784930405,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5780387178673438e-05,
+      "loss": 1.2531,
+      "step": 3484
+    },
+    {
+      "epoch": 1.0949997055185818,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.577784830212631e-05,
+      "loss": 1.1636,
+      "step": 3486
+    },
+    {
+      "epoch": 1.0956279325441232,
+      "grad_norm": 0.75,
+      "learning_rate": 1.577530942557918e-05,
+      "loss": 1.2318,
+      "step": 3488
+    },
+    {
+      "epoch": 1.0962561595696645,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5772770549032056e-05,
+      "loss": 1.337,
+      "step": 3490
+    },
+    {
+      "epoch": 1.0968843865952058,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.5770231672484927e-05,
+      "loss": 1.2689,
+      "step": 3492
+    },
+    {
+      "epoch": 1.0975126136207471,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.57676927959378e-05,
+      "loss": 1.2185,
+      "step": 3494
+    },
+    {
+      "epoch": 1.0981408406462885,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.576515391939067e-05,
+      "loss": 1.3053,
+      "step": 3496
+    },
+    {
+      "epoch": 1.0987690676718298,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5762615042843545e-05,
+      "loss": 1.1346,
+      "step": 3498
+    },
+    {
+      "epoch": 1.0993972946973714,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5760076166296413e-05,
+      "loss": 1.2331,
+      "step": 3500
+    },
+    {
+      "epoch": 1.1000255217229127,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5757537289749287e-05,
+      "loss": 1.2669,
+      "step": 3502
+    },
+    {
+      "epoch": 1.100653748748454,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.575499841320216e-05,
+      "loss": 1.2222,
+      "step": 3504
+    },
+    {
+      "epoch": 1.1012819757739953,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.575245953665503e-05,
+      "loss": 1.1517,
+      "step": 3506
+    },
+    {
+      "epoch": 1.1019102027995367,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.57499206601079e-05,
+      "loss": 1.2206,
+      "step": 3508
+    },
+    {
+      "epoch": 1.102538429825078,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5747381783560776e-05,
+      "loss": 1.2261,
+      "step": 3510
+    },
+    {
+      "epoch": 1.1031666568506193,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5744842907013648e-05,
+      "loss": 1.0953,
+      "step": 3512
+    },
+    {
+      "epoch": 1.1037948838761606,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.574230403046652e-05,
+      "loss": 1.3215,
+      "step": 3514
+    },
+    {
+      "epoch": 1.1044231109017022,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5739765153919394e-05,
+      "loss": 1.3369,
+      "step": 3516
+    },
+    {
+      "epoch": 1.1050513379272435,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5737226277372265e-05,
+      "loss": 1.2915,
+      "step": 3518
+    },
+    {
+      "epoch": 1.1056795649527849,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.5734687400825137e-05,
+      "loss": 1.3005,
+      "step": 3520
+    },
+    {
+      "epoch": 1.1063077919783262,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5732148524278008e-05,
+      "loss": 1.2166,
+      "step": 3522
+    },
+    {
+      "epoch": 1.1069360190038675,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.5729609647730883e-05,
+      "loss": 1.1558,
+      "step": 3524
+    },
+    {
+      "epoch": 1.1075642460294088,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.572707077118375e-05,
+      "loss": 1.2904,
+      "step": 3526
+    },
+    {
+      "epoch": 1.1081924730549502,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5724531894636626e-05,
+      "loss": 1.1995,
+      "step": 3528
+    },
+    {
+      "epoch": 1.1088207000804915,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5721993018089497e-05,
+      "loss": 1.3063,
+      "step": 3530
+    },
+    {
+      "epoch": 1.109448927106033,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.571945414154237e-05,
+      "loss": 1.3033,
+      "step": 3532
+    },
+    {
+      "epoch": 1.1100771541315744,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.571691526499524e-05,
+      "loss": 1.2816,
+      "step": 3534
+    },
+    {
+      "epoch": 1.1107053811571157,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5714376388448114e-05,
+      "loss": 1.3709,
+      "step": 3536
+    },
+    {
+      "epoch": 1.111333608182657,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5711837511900986e-05,
+      "loss": 1.1248,
+      "step": 3538
+    },
+    {
+      "epoch": 1.1119618352081984,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5709298635353857e-05,
+      "loss": 1.2373,
+      "step": 3540
+    },
+    {
+      "epoch": 1.1125900622337397,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.570675975880673e-05,
+      "loss": 1.3827,
+      "step": 3542
+    },
+    {
+      "epoch": 1.113218289259281,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.5704220882259603e-05,
+      "loss": 1.2102,
+      "step": 3544
+    },
+    {
+      "epoch": 1.1138465162848223,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.570168200571247e-05,
+      "loss": 1.1883,
+      "step": 3546
+    },
+    {
+      "epoch": 1.114474743310364,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5699143129165346e-05,
+      "loss": 1.1861,
+      "step": 3548
+    },
+    {
+      "epoch": 1.1151029703359052,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5696604252618218e-05,
+      "loss": 1.3095,
+      "step": 3550
+    },
+    {
+      "epoch": 1.1157311973614465,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.569406537607109e-05,
+      "loss": 1.1794,
+      "step": 3552
+    },
+    {
+      "epoch": 1.1163594243869879,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.569152649952396e-05,
+      "loss": 1.2195,
+      "step": 3554
+    },
+    {
+      "epoch": 1.1169876514125292,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5688987622976835e-05,
+      "loss": 1.3026,
+      "step": 3556
+    },
+    {
+      "epoch": 1.1176158784380705,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5686448746429706e-05,
+      "loss": 1.308,
+      "step": 3558
+    },
+    {
+      "epoch": 1.1182441054636119,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5683909869882578e-05,
+      "loss": 1.192,
+      "step": 3560
+    },
+    {
+      "epoch": 1.1188723324891532,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.568137099333545e-05,
+      "loss": 1.1283,
+      "step": 3562
+    },
+    {
+      "epoch": 1.1195005595146945,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5678832116788324e-05,
+      "loss": 1.2358,
+      "step": 3564
+    },
+    {
+      "epoch": 1.120128786540236,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5676293240241195e-05,
+      "loss": 1.1856,
+      "step": 3566
+    },
+    {
+      "epoch": 1.1207570135657774,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5673754363694067e-05,
+      "loss": 1.2214,
+      "step": 3568
+    },
+    {
+      "epoch": 1.1213852405913187,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5671215487146938e-05,
+      "loss": 1.1597,
+      "step": 3570
+    },
+    {
+      "epoch": 1.12201346761686,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.566867661059981e-05,
+      "loss": 1.1691,
+      "step": 3572
+    },
+    {
+      "epoch": 1.1226416946424014,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.566613773405268e-05,
+      "loss": 1.3458,
+      "step": 3574
+    },
+    {
+      "epoch": 1.1232699216679427,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5663598857505556e-05,
+      "loss": 1.0832,
+      "step": 3576
+    },
+    {
+      "epoch": 1.123898148693484,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.5661059980958427e-05,
+      "loss": 1.2571,
+      "step": 3578
+    },
+    {
+      "epoch": 1.1245263757190254,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.56585211044113e-05,
+      "loss": 1.3284,
+      "step": 3580
+    },
+    {
+      "epoch": 1.125154602744567,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.565598222786417e-05,
+      "loss": 1.1614,
+      "step": 3582
+    },
+    {
+      "epoch": 1.1257828297701082,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5653443351317045e-05,
+      "loss": 1.208,
+      "step": 3584
+    },
+    {
+      "epoch": 1.1264110567956496,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5650904474769916e-05,
+      "loss": 1.1761,
+      "step": 3586
+    },
+    {
+      "epoch": 1.127039283821191,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5648365598222787e-05,
+      "loss": 1.3036,
+      "step": 3588
+    },
+    {
+      "epoch": 1.1276675108467322,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.564582672167566e-05,
+      "loss": 1.1291,
+      "step": 3590
+    },
+    {
+      "epoch": 1.1282957378722736,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5643287845128534e-05,
+      "loss": 1.2135,
+      "step": 3592
+    },
+    {
+      "epoch": 1.1289239648978149,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.56407489685814e-05,
+      "loss": 1.2834,
+      "step": 3594
+    },
+    {
+      "epoch": 1.1295521919233562,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5638210092034276e-05,
+      "loss": 1.328,
+      "step": 3596
+    },
+    {
+      "epoch": 1.1301804189488975,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5635671215487148e-05,
+      "loss": 1.346,
+      "step": 3598
+    },
+    {
+      "epoch": 1.130808645974439,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.563313233894002e-05,
+      "loss": 1.0939,
+      "step": 3600
+    },
+    {
+      "epoch": 1.1314368729999804,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5630593462392894e-05,
+      "loss": 1.243,
+      "step": 3602
+    },
+    {
+      "epoch": 1.1320651000255217,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.5628054585845765e-05,
+      "loss": 1.2313,
+      "step": 3604
+    },
+    {
+      "epoch": 1.132693327051063,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5625515709298637e-05,
+      "loss": 1.2229,
+      "step": 3606
+    },
+    {
+      "epoch": 1.1333215540766044,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5622976832751508e-05,
+      "loss": 1.3329,
+      "step": 3608
+    },
+    {
+      "epoch": 1.1339497811021457,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.5620437956204383e-05,
+      "loss": 1.1156,
+      "step": 3610
+    },
+    {
+      "epoch": 1.134578008127687,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5617899079657254e-05,
+      "loss": 1.1354,
+      "step": 3612
+    },
+    {
+      "epoch": 1.1352062351532286,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.5615360203110125e-05,
+      "loss": 1.2974,
+      "step": 3614
+    },
+    {
+      "epoch": 1.13583446217877,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5612821326562997e-05,
+      "loss": 1.3404,
+      "step": 3616
+    },
+    {
+      "epoch": 1.1364626892043113,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.561028245001587e-05,
+      "loss": 1.2677,
+      "step": 3618
+    },
+    {
+      "epoch": 1.1370909162298526,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.560774357346874e-05,
+      "loss": 1.1016,
+      "step": 3620
+    },
+    {
+      "epoch": 1.137719143255394,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5605204696921614e-05,
+      "loss": 1.2943,
+      "step": 3622
+    },
+    {
+      "epoch": 1.1383473702809352,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5602665820374486e-05,
+      "loss": 1.196,
+      "step": 3624
+    },
+    {
+      "epoch": 1.1389755973064766,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5600126943827357e-05,
+      "loss": 1.1846,
+      "step": 3626
+    },
+    {
+      "epoch": 1.139603824332018,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.559758806728023e-05,
+      "loss": 1.2626,
+      "step": 3628
+    },
+    {
+      "epoch": 1.1402320513575592,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5595049190733103e-05,
+      "loss": 1.3086,
+      "step": 3630
+    },
+    {
+      "epoch": 1.1408602783831008,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.5592510314185975e-05,
+      "loss": 1.2576,
+      "step": 3632
+    },
+    {
+      "epoch": 1.141488505408642,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5589971437638846e-05,
+      "loss": 1.3336,
+      "step": 3634
+    },
+    {
+      "epoch": 1.1421167324341834,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5587432561091717e-05,
+      "loss": 1.1999,
+      "step": 3636
+    },
+    {
+      "epoch": 1.1427449594597248,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5584893684544592e-05,
+      "loss": 1.2824,
+      "step": 3638
+    },
+    {
+      "epoch": 1.143373186485266,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.558235480799746e-05,
+      "loss": 1.3522,
+      "step": 3640
+    },
+    {
+      "epoch": 1.1440014135108074,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.5579815931450335e-05,
+      "loss": 1.2944,
+      "step": 3642
+    },
+    {
+      "epoch": 1.1446296405363487,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5577277054903206e-05,
+      "loss": 1.2327,
+      "step": 3644
+    },
+    {
+      "epoch": 1.1452578675618903,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5574738178356078e-05,
+      "loss": 1.3088,
+      "step": 3646
+    },
+    {
+      "epoch": 1.1458860945874316,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.557219930180895e-05,
+      "loss": 1.2625,
+      "step": 3648
+    },
+    {
+      "epoch": 1.146514321612973,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5569660425261824e-05,
+      "loss": 1.3164,
+      "step": 3650
+    },
+    {
+      "epoch": 1.1471425486385143,
+      "grad_norm": 0.875,
+      "learning_rate": 1.5567121548714695e-05,
+      "loss": 1.1474,
+      "step": 3652
+    },
+    {
+      "epoch": 1.1477707756640556,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5564582672167567e-05,
+      "loss": 1.3335,
+      "step": 3654
+    },
+    {
+      "epoch": 1.148399002689597,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5562043795620438e-05,
+      "loss": 1.2914,
+      "step": 3656
+    },
+    {
+      "epoch": 1.1490272297151383,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5559504919073313e-05,
+      "loss": 1.2283,
+      "step": 3658
+    },
+    {
+      "epoch": 1.1496554567406796,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5556966042526184e-05,
+      "loss": 1.2828,
+      "step": 3660
+    },
+    {
+      "epoch": 1.150283683766221,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.5554427165979056e-05,
+      "loss": 1.2971,
+      "step": 3662
+    },
+    {
+      "epoch": 1.1509119107917622,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5551888289431927e-05,
+      "loss": 1.1939,
+      "step": 3664
+    },
+    {
+      "epoch": 1.1515401378173038,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.55493494128848e-05,
+      "loss": 1.2101,
+      "step": 3666
+    },
+    {
+      "epoch": 1.1521683648428451,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.554681053633767e-05,
+      "loss": 1.3979,
+      "step": 3668
+    },
+    {
+      "epoch": 1.1527965918683865,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5544271659790545e-05,
+      "loss": 1.1002,
+      "step": 3670
+    },
+    {
+      "epoch": 1.1534248188939278,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5541732783243416e-05,
+      "loss": 1.2375,
+      "step": 3672
+    },
+    {
+      "epoch": 1.1540530459194691,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5539193906696287e-05,
+      "loss": 1.315,
+      "step": 3674
+    },
+    {
+      "epoch": 1.1546812729450104,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.553665503014916e-05,
+      "loss": 1.3209,
+      "step": 3676
+    },
+    {
+      "epoch": 1.1553094999705518,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5534116153602033e-05,
+      "loss": 1.2149,
+      "step": 3678
+    },
+    {
+      "epoch": 1.1559377269960933,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5531577277054905e-05,
+      "loss": 1.2292,
+      "step": 3680
+    },
+    {
+      "epoch": 1.1565659540216346,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5529038400507776e-05,
+      "loss": 1.148,
+      "step": 3682
+    },
+    {
+      "epoch": 1.157194181047176,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.552649952396065e-05,
+      "loss": 1.2629,
+      "step": 3684
+    },
+    {
+      "epoch": 1.1578224080727173,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5523960647413522e-05,
+      "loss": 1.13,
+      "step": 3686
+    },
+    {
+      "epoch": 1.1584506350982586,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5521421770866394e-05,
+      "loss": 1.2941,
+      "step": 3688
+    },
+    {
+      "epoch": 1.1590788621238,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5518882894319265e-05,
+      "loss": 1.2437,
+      "step": 3690
+    },
+    {
+      "epoch": 1.1597070891493413,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.5516344017772136e-05,
+      "loss": 1.3109,
+      "step": 3692
+    },
+    {
+      "epoch": 1.1603353161748826,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5513805141225008e-05,
+      "loss": 1.0916,
+      "step": 3694
+    },
+    {
+      "epoch": 1.160963543200424,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5511266264677883e-05,
+      "loss": 1.1996,
+      "step": 3696
+    },
+    {
+      "epoch": 1.1615917702259655,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.5508727388130754e-05,
+      "loss": 1.2492,
+      "step": 3698
+    },
+    {
+      "epoch": 1.1622199972515068,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5506188511583625e-05,
+      "loss": 1.2623,
+      "step": 3700
+    },
+    {
+      "epoch": 1.1628482242770481,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5503649635036497e-05,
+      "loss": 1.4083,
+      "step": 3702
+    },
+    {
+      "epoch": 1.1634764513025895,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.550111075848937e-05,
+      "loss": 1.2405,
+      "step": 3704
+    },
+    {
+      "epoch": 1.1641046783281308,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.5498571881942243e-05,
+      "loss": 1.0913,
+      "step": 3706
+    },
+    {
+      "epoch": 1.1647329053536721,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5496033005395114e-05,
+      "loss": 1.1841,
+      "step": 3708
+    },
+    {
+      "epoch": 1.1653611323792135,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5493494128847986e-05,
+      "loss": 1.3798,
+      "step": 3710
+    },
+    {
+      "epoch": 1.165989359404755,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.549095525230086e-05,
+      "loss": 1.3218,
+      "step": 3712
+    },
+    {
+      "epoch": 1.1666175864302963,
+      "grad_norm": 1.0,
+      "learning_rate": 1.548841637575373e-05,
+      "loss": 1.3271,
+      "step": 3714
+    },
+    {
+      "epoch": 1.1672458134558377,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5485877499206603e-05,
+      "loss": 1.3283,
+      "step": 3716
+    },
+    {
+      "epoch": 1.167874040481379,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.5483338622659475e-05,
+      "loss": 1.1618,
+      "step": 3718
+    },
+    {
+      "epoch": 1.1685022675069203,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5480799746112346e-05,
+      "loss": 1.2939,
+      "step": 3720
+    },
+    {
+      "epoch": 1.1691304945324617,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5478260869565217e-05,
+      "loss": 1.2001,
+      "step": 3722
+    },
+    {
+      "epoch": 1.169758721558003,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5475721993018092e-05,
+      "loss": 1.2964,
+      "step": 3724
+    },
+    {
+      "epoch": 1.1703869485835443,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5473183116470964e-05,
+      "loss": 1.2945,
+      "step": 3726
+    },
+    {
+      "epoch": 1.1710151756090856,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5470644239923835e-05,
+      "loss": 1.3001,
+      "step": 3728
+    },
+    {
+      "epoch": 1.1716434026346272,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5468105363376706e-05,
+      "loss": 1.0717,
+      "step": 3730
+    },
+    {
+      "epoch": 1.1722716296601685,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.546556648682958e-05,
+      "loss": 1.1697,
+      "step": 3732
+    },
+    {
+      "epoch": 1.1728998566857098,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.546302761028245e-05,
+      "loss": 1.1945,
+      "step": 3734
+    },
+    {
+      "epoch": 1.1735280837112512,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5460488733735324e-05,
+      "loss": 1.1745,
+      "step": 3736
+    },
+    {
+      "epoch": 1.1741563107367925,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5457949857188195e-05,
+      "loss": 1.2856,
+      "step": 3738
+    },
+    {
+      "epoch": 1.1747845377623338,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5455410980641067e-05,
+      "loss": 1.2289,
+      "step": 3740
+    },
+    {
+      "epoch": 1.1754127647878752,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.5452872104093938e-05,
+      "loss": 1.1443,
+      "step": 3742
+    },
+    {
+      "epoch": 1.1760409918134165,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5450333227546813e-05,
+      "loss": 1.2261,
+      "step": 3744
+    },
+    {
+      "epoch": 1.176669218838958,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5447794350999684e-05,
+      "loss": 1.2703,
+      "step": 3746
+    },
+    {
+      "epoch": 1.1772974458644994,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5445255474452556e-05,
+      "loss": 1.2568,
+      "step": 3748
+    },
+    {
+      "epoch": 1.1779256728900407,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5442716597905427e-05,
+      "loss": 1.2741,
+      "step": 3750
+    },
+    {
+      "epoch": 1.178553899915582,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.54401777213583e-05,
+      "loss": 1.2063,
+      "step": 3752
+    },
+    {
+      "epoch": 1.1791821269411233,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5437638844811173e-05,
+      "loss": 1.2187,
+      "step": 3754
+    },
+    {
+      "epoch": 1.1798103539666647,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5435099968264044e-05,
+      "loss": 1.2115,
+      "step": 3756
+    },
+    {
+      "epoch": 1.180438580992206,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5432561091716916e-05,
+      "loss": 1.2467,
+      "step": 3758
+    },
+    {
+      "epoch": 1.1810668080177473,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5430022215169787e-05,
+      "loss": 1.3909,
+      "step": 3760
+    },
+    {
+      "epoch": 1.1816950350432887,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.542748333862266e-05,
+      "loss": 1.3045,
+      "step": 3762
+    },
+    {
+      "epoch": 1.1823232620688302,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5424944462075533e-05,
+      "loss": 1.2189,
+      "step": 3764
+    },
+    {
+      "epoch": 1.1829514890943715,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5422405585528405e-05,
+      "loss": 1.0881,
+      "step": 3766
+    },
+    {
+      "epoch": 1.1835797161199129,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5419866708981276e-05,
+      "loss": 1.2768,
+      "step": 3768
+    },
+    {
+      "epoch": 1.1842079431454542,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.541732783243415e-05,
+      "loss": 1.3732,
+      "step": 3770
+    },
+    {
+      "epoch": 1.1848361701709955,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5414788955887022e-05,
+      "loss": 1.3118,
+      "step": 3772
+    },
+    {
+      "epoch": 1.1854643971965368,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.5412250079339894e-05,
+      "loss": 1.3224,
+      "step": 3774
+    },
+    {
+      "epoch": 1.1860926242220782,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5409711202792765e-05,
+      "loss": 1.1572,
+      "step": 3776
+    },
+    {
+      "epoch": 1.1867208512476197,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.540717232624564e-05,
+      "loss": 1.2863,
+      "step": 3778
+    },
+    {
+      "epoch": 1.187349078273161,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5404633449698508e-05,
+      "loss": 1.1624,
+      "step": 3780
+    },
+    {
+      "epoch": 1.1879773052987024,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5402094573151383e-05,
+      "loss": 1.3295,
+      "step": 3782
+    },
+    {
+      "epoch": 1.1886055323242437,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5399555696604254e-05,
+      "loss": 1.3428,
+      "step": 3784
+    },
+    {
+      "epoch": 1.189233759349785,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5397016820057125e-05,
+      "loss": 1.2549,
+      "step": 3786
+    },
+    {
+      "epoch": 1.1898619863753264,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5394477943509997e-05,
+      "loss": 1.2031,
+      "step": 3788
+    },
+    {
+      "epoch": 1.1904902134008677,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.539193906696287e-05,
+      "loss": 1.13,
+      "step": 3790
+    },
+    {
+      "epoch": 1.191118440426409,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.5389400190415743e-05,
+      "loss": 1.2701,
+      "step": 3792
+    },
+    {
+      "epoch": 1.1917466674519503,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5386861313868614e-05,
+      "loss": 1.208,
+      "step": 3794
+    },
+    {
+      "epoch": 1.192374894477492,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5384322437321486e-05,
+      "loss": 1.2884,
+      "step": 3796
+    },
+    {
+      "epoch": 1.1930031215030332,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.538178356077436e-05,
+      "loss": 1.1525,
+      "step": 3798
+    },
+    {
+      "epoch": 1.1936313485285746,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5379244684227232e-05,
+      "loss": 1.3432,
+      "step": 3800
+    },
+    {
+      "epoch": 1.1942595755541159,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5376705807680103e-05,
+      "loss": 1.1941,
+      "step": 3802
+    },
+    {
+      "epoch": 1.1948878025796572,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5374166931132975e-05,
+      "loss": 1.1627,
+      "step": 3804
+    },
+    {
+      "epoch": 1.1955160296051985,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5371628054585846e-05,
+      "loss": 1.1927,
+      "step": 3806
+    },
+    {
+      "epoch": 1.1961442566307399,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5369089178038717e-05,
+      "loss": 1.2496,
+      "step": 3808
+    },
+    {
+      "epoch": 1.1967724836562812,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5366550301491592e-05,
+      "loss": 1.3009,
+      "step": 3810
+    },
+    {
+      "epoch": 1.1974007106818227,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5364011424944463e-05,
+      "loss": 1.2321,
+      "step": 3812
+    },
+    {
+      "epoch": 1.198028937707364,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5361472548397335e-05,
+      "loss": 1.3292,
+      "step": 3814
+    },
+    {
+      "epoch": 1.1986571647329054,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5358933671850206e-05,
+      "loss": 1.1956,
+      "step": 3816
+    },
+    {
+      "epoch": 1.1992853917584467,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.535639479530308e-05,
+      "loss": 1.1833,
+      "step": 3818
+    },
+    {
+      "epoch": 1.199913618783988,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5353855918755952e-05,
+      "loss": 1.3083,
+      "step": 3820
+    },
+    {
+      "epoch": 1.2005418458095294,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5351317042208824e-05,
+      "loss": 1.2926,
+      "step": 3822
+    },
+    {
+      "epoch": 1.2011700728350707,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5348778165661695e-05,
+      "loss": 1.2154,
+      "step": 3824
+    },
+    {
+      "epoch": 1.201798299860612,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.534623928911457e-05,
+      "loss": 1.2833,
+      "step": 3826
+    },
+    {
+      "epoch": 1.2024265268861534,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.5343700412567438e-05,
+      "loss": 1.2619,
+      "step": 3828
+    },
+    {
+      "epoch": 1.203054753911695,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.5341161536020313e-05,
+      "loss": 1.1693,
+      "step": 3830
+    },
+    {
+      "epoch": 1.2036829809372362,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.5338622659473184e-05,
+      "loss": 1.388,
+      "step": 3832
+    },
+    {
+      "epoch": 1.2043112079627776,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5336083782926055e-05,
+      "loss": 1.1663,
+      "step": 3834
+    },
+    {
+      "epoch": 1.204939434988319,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5333544906378927e-05,
+      "loss": 1.2422,
+      "step": 3836
+    },
+    {
+      "epoch": 1.2055676620138602,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.53310060298318e-05,
+      "loss": 1.3582,
+      "step": 3838
+    },
+    {
+      "epoch": 1.2061958890394016,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5328467153284673e-05,
+      "loss": 1.279,
+      "step": 3840
+    },
+    {
+      "epoch": 1.2068241160649429,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5325928276737544e-05,
+      "loss": 1.1224,
+      "step": 3842
+    },
+    {
+      "epoch": 1.2074523430904844,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5323389400190416e-05,
+      "loss": 1.164,
+      "step": 3844
+    },
+    {
+      "epoch": 1.2080805701160258,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.532085052364329e-05,
+      "loss": 1.2088,
+      "step": 3846
+    },
+    {
+      "epoch": 1.208708797141567,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.531831164709616e-05,
+      "loss": 1.2329,
+      "step": 3848
+    },
+    {
+      "epoch": 1.2093370241671084,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5315772770549033e-05,
+      "loss": 1.251,
+      "step": 3850
+    },
+    {
+      "epoch": 1.2099652511926497,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.5313233894001908e-05,
+      "loss": 1.2042,
+      "step": 3852
+    },
+    {
+      "epoch": 1.210593478218191,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5310695017454776e-05,
+      "loss": 1.143,
+      "step": 3854
+    },
+    {
+      "epoch": 1.2112217052437324,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.530815614090765e-05,
+      "loss": 1.319,
+      "step": 3856
+    },
+    {
+      "epoch": 1.2118499322692737,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5305617264360522e-05,
+      "loss": 1.2545,
+      "step": 3858
+    },
+    {
+      "epoch": 1.212478159294815,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5303078387813394e-05,
+      "loss": 1.2631,
+      "step": 3860
+    },
+    {
+      "epoch": 1.2131063863203566,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5300539511266265e-05,
+      "loss": 1.2877,
+      "step": 3862
+    },
+    {
+      "epoch": 1.213734613345898,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.529800063471914e-05,
+      "loss": 1.2363,
+      "step": 3864
+    },
+    {
+      "epoch": 1.2143628403714393,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.529546175817201e-05,
+      "loss": 1.1702,
+      "step": 3866
+    },
+    {
+      "epoch": 1.2149910673969806,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5292922881624882e-05,
+      "loss": 1.25,
+      "step": 3868
+    },
+    {
+      "epoch": 1.215619294422522,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5290384005077754e-05,
+      "loss": 1.3495,
+      "step": 3870
+    },
+    {
+      "epoch": 1.2162475214480633,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.528784512853063e-05,
+      "loss": 1.2843,
+      "step": 3872
+    },
+    {
+      "epoch": 1.2168757484736046,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5285306251983497e-05,
+      "loss": 1.2542,
+      "step": 3874
+    },
+    {
+      "epoch": 1.217503975499146,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.528276737543637e-05,
+      "loss": 1.3101,
+      "step": 3876
+    },
+    {
+      "epoch": 1.2181322025246875,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5280228498889243e-05,
+      "loss": 1.28,
+      "step": 3878
+    },
+    {
+      "epoch": 1.2187604295502288,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5277689622342114e-05,
+      "loss": 1.2766,
+      "step": 3880
+    },
+    {
+      "epoch": 1.2193886565757701,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5275150745794986e-05,
+      "loss": 1.236,
+      "step": 3882
+    },
+    {
+      "epoch": 1.2200168836013114,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.527261186924786e-05,
+      "loss": 1.1882,
+      "step": 3884
+    },
+    {
+      "epoch": 1.2206451106268528,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.5270072992700732e-05,
+      "loss": 1.1035,
+      "step": 3886
+    },
+    {
+      "epoch": 1.221273337652394,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5267534116153603e-05,
+      "loss": 1.1533,
+      "step": 3888
+    },
+    {
+      "epoch": 1.2219015646779354,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5264995239606474e-05,
+      "loss": 1.1649,
+      "step": 3890
+    },
+    {
+      "epoch": 1.2225297917034768,
+      "grad_norm": 0.875,
+      "learning_rate": 1.526245636305935e-05,
+      "loss": 1.2608,
+      "step": 3892
+    },
+    {
+      "epoch": 1.223158018729018,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.525991748651222e-05,
+      "loss": 1.4731,
+      "step": 3894
+    },
+    {
+      "epoch": 1.2237862457545596,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5257378609965092e-05,
+      "loss": 1.276,
+      "step": 3896
+    },
+    {
+      "epoch": 1.224414472780101,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5254839733417963e-05,
+      "loss": 1.1836,
+      "step": 3898
+    },
+    {
+      "epoch": 1.2250426998056423,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5252300856870836e-05,
+      "loss": 1.1804,
+      "step": 3900
+    },
+    {
+      "epoch": 1.2256709268311836,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5249761980323708e-05,
+      "loss": 1.3125,
+      "step": 3902
+    },
+    {
+      "epoch": 1.226299153856725,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5247223103776581e-05,
+      "loss": 1.2143,
+      "step": 3904
+    },
+    {
+      "epoch": 1.2269273808822663,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5244684227229452e-05,
+      "loss": 1.2005,
+      "step": 3906
+    },
+    {
+      "epoch": 1.2275556079078076,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.5242145350682325e-05,
+      "loss": 1.2187,
+      "step": 3908
+    },
+    {
+      "epoch": 1.2281838349333492,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5239606474135195e-05,
+      "loss": 1.3649,
+      "step": 3910
+    },
+    {
+      "epoch": 1.2288120619588905,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.5237067597588068e-05,
+      "loss": 1.2188,
+      "step": 3912
+    },
+    {
+      "epoch": 1.2294402889844318,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.523452872104094e-05,
+      "loss": 1.1949,
+      "step": 3914
+    },
+    {
+      "epoch": 1.2300685160099731,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5231989844493813e-05,
+      "loss": 1.2312,
+      "step": 3916
+    },
+    {
+      "epoch": 1.2306967430355145,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5229450967946684e-05,
+      "loss": 1.1944,
+      "step": 3918
+    },
+    {
+      "epoch": 1.2313249700610558,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5226912091399557e-05,
+      "loss": 1.2278,
+      "step": 3920
+    },
+    {
+      "epoch": 1.2319531970865971,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5224373214852428e-05,
+      "loss": 1.2428,
+      "step": 3922
+    },
+    {
+      "epoch": 1.2325814241121384,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5221834338305302e-05,
+      "loss": 1.3336,
+      "step": 3924
+    },
+    {
+      "epoch": 1.2332096511376798,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5219295461758173e-05,
+      "loss": 1.3555,
+      "step": 3926
+    },
+    {
+      "epoch": 1.2338378781632213,
+      "grad_norm": 0.875,
+      "learning_rate": 1.5216756585211046e-05,
+      "loss": 1.1514,
+      "step": 3928
+    },
+    {
+      "epoch": 1.2344661051887627,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5214217708663916e-05,
+      "loss": 1.3315,
+      "step": 3930
+    },
+    {
+      "epoch": 1.235094332214304,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.521167883211679e-05,
+      "loss": 1.2199,
+      "step": 3932
+    },
+    {
+      "epoch": 1.2357225592398453,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.520913995556966e-05,
+      "loss": 1.1983,
+      "step": 3934
+    },
+    {
+      "epoch": 1.2363507862653866,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5206601079022533e-05,
+      "loss": 1.2382,
+      "step": 3936
+    },
+    {
+      "epoch": 1.236979013290928,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.5204062202475406e-05,
+      "loss": 1.1452,
+      "step": 3938
+    },
+    {
+      "epoch": 1.2376072403164693,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.5201523325928278e-05,
+      "loss": 1.2165,
+      "step": 3940
+    },
+    {
+      "epoch": 1.2382354673420108,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.519898444938115e-05,
+      "loss": 1.1627,
+      "step": 3942
+    },
+    {
+      "epoch": 1.2388636943675522,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5196445572834022e-05,
+      "loss": 1.1965,
+      "step": 3944
+    },
+    {
+      "epoch": 1.2394919213930935,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.5193906696286895e-05,
+      "loss": 1.19,
+      "step": 3946
+    },
+    {
+      "epoch": 1.2401201484186348,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5191367819739767e-05,
+      "loss": 1.2672,
+      "step": 3948
+    },
+    {
+      "epoch": 1.2407483754441762,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.518882894319264e-05,
+      "loss": 1.2684,
+      "step": 3950
+    },
+    {
+      "epoch": 1.2413766024697175,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5186290066645511e-05,
+      "loss": 1.3026,
+      "step": 3952
+    },
+    {
+      "epoch": 1.2420048294952588,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5183751190098384e-05,
+      "loss": 1.2142,
+      "step": 3954
+    },
+    {
+      "epoch": 1.2426330565208001,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5181212313551254e-05,
+      "loss": 1.2394,
+      "step": 3956
+    },
+    {
+      "epoch": 1.2432612835463415,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.5178673437004129e-05,
+      "loss": 1.1269,
+      "step": 3958
+    },
+    {
+      "epoch": 1.2438895105718828,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5176134560456998e-05,
+      "loss": 1.2404,
+      "step": 3960
+    },
+    {
+      "epoch": 1.2445177375974243,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.5173595683909871e-05,
+      "loss": 1.2585,
+      "step": 3962
+    },
+    {
+      "epoch": 1.2451459646229657,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5171056807362743e-05,
+      "loss": 1.2264,
+      "step": 3964
+    },
+    {
+      "epoch": 1.245774191648507,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5168517930815616e-05,
+      "loss": 1.3453,
+      "step": 3966
+    },
+    {
+      "epoch": 1.2464024186740483,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5165979054268487e-05,
+      "loss": 1.2773,
+      "step": 3968
+    },
+    {
+      "epoch": 1.2470306456995897,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.516344017772136e-05,
+      "loss": 1.15,
+      "step": 3970
+    },
+    {
+      "epoch": 1.247658872725131,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5160901301174232e-05,
+      "loss": 1.3222,
+      "step": 3972
+    },
+    {
+      "epoch": 1.2482870997506723,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5158362424627105e-05,
+      "loss": 1.1521,
+      "step": 3974
+    },
+    {
+      "epoch": 1.2489153267762139,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.5155823548079976e-05,
+      "loss": 1.2704,
+      "step": 3976
+    },
+    {
+      "epoch": 1.2495435538017552,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.515328467153285e-05,
+      "loss": 1.3046,
+      "step": 3978
+    },
+    {
+      "epoch": 1.2501717808272965,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5150745794985719e-05,
+      "loss": 1.1866,
+      "step": 3980
+    },
+    {
+      "epoch": 1.2508000078528378,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5148206918438592e-05,
+      "loss": 1.3763,
+      "step": 3982
+    },
+    {
+      "epoch": 1.2514282348783792,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5145668041891463e-05,
+      "loss": 1.3093,
+      "step": 3984
+    },
+    {
+      "epoch": 1.2520564619039205,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5143129165344336e-05,
+      "loss": 1.1302,
+      "step": 3986
+    },
+    {
+      "epoch": 1.2526846889294618,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5140590288797208e-05,
+      "loss": 1.2198,
+      "step": 3988
+    },
+    {
+      "epoch": 1.2533129159550032,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5138051412250081e-05,
+      "loss": 1.2666,
+      "step": 3990
+    },
+    {
+      "epoch": 1.2539411429805445,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.5135512535702952e-05,
+      "loss": 1.0958,
+      "step": 3992
+    },
+    {
+      "epoch": 1.2545693700060858,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5132973659155825e-05,
+      "loss": 1.2114,
+      "step": 3994
+    },
+    {
+      "epoch": 1.2551975970316274,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5130434782608697e-05,
+      "loss": 1.3153,
+      "step": 3996
+    },
+    {
+      "epoch": 1.2558258240571687,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.512789590606157e-05,
+      "loss": 1.2129,
+      "step": 3998
+    },
+    {
+      "epoch": 1.25645405108271,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.512535702951444e-05,
+      "loss": 1.2922,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2570822781082513,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5122818152967314e-05,
+      "loss": 1.1252,
+      "step": 4002
+    },
+    {
+      "epoch": 1.2577105051337927,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5120279276420184e-05,
+      "loss": 1.1488,
+      "step": 4004
+    },
+    {
+      "epoch": 1.258338732159334,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5117740399873057e-05,
+      "loss": 1.2191,
+      "step": 4006
+    },
+    {
+      "epoch": 1.2589669591848756,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5115201523325928e-05,
+      "loss": 1.345,
+      "step": 4008
+    },
+    {
+      "epoch": 1.2595951862104169,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5112662646778801e-05,
+      "loss": 1.2722,
+      "step": 4010
+    },
+    {
+      "epoch": 1.2602234132359582,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5110123770231673e-05,
+      "loss": 1.2438,
+      "step": 4012
+    },
+    {
+      "epoch": 1.2608516402614995,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.5107584893684546e-05,
+      "loss": 1.2488,
+      "step": 4014
+    },
+    {
+      "epoch": 1.2614798672870409,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5105046017137417e-05,
+      "loss": 1.1744,
+      "step": 4016
+    },
+    {
+      "epoch": 1.2621080943125822,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.510250714059029e-05,
+      "loss": 1.2071,
+      "step": 4018
+    },
+    {
+      "epoch": 1.2627363213381235,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5099968264043162e-05,
+      "loss": 1.3156,
+      "step": 4020
+    },
+    {
+      "epoch": 1.2633645483636649,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5097429387496035e-05,
+      "loss": 1.1231,
+      "step": 4022
+    },
+    {
+      "epoch": 1.2639927753892062,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5094890510948908e-05,
+      "loss": 1.151,
+      "step": 4024
+    },
+    {
+      "epoch": 1.2646210024147475,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5092351634401778e-05,
+      "loss": 1.3427,
+      "step": 4026
+    },
+    {
+      "epoch": 1.265249229440289,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5089812757854652e-05,
+      "loss": 1.2093,
+      "step": 4028
+    },
+    {
+      "epoch": 1.2658774564658304,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5087273881307522e-05,
+      "loss": 1.179,
+      "step": 4030
+    },
+    {
+      "epoch": 1.2665056834913717,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5084735004760395e-05,
+      "loss": 1.2487,
+      "step": 4032
+    },
+    {
+      "epoch": 1.267133910516913,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5082196128213267e-05,
+      "loss": 1.2788,
+      "step": 4034
+    },
+    {
+      "epoch": 1.2677621375424544,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.507965725166614e-05,
+      "loss": 1.2915,
+      "step": 4036
+    },
+    {
+      "epoch": 1.2683903645679957,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5077118375119011e-05,
+      "loss": 1.2832,
+      "step": 4038
+    },
+    {
+      "epoch": 1.2690185915935372,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5074579498571884e-05,
+      "loss": 1.2789,
+      "step": 4040
+    },
+    {
+      "epoch": 1.2696468186190786,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5072040622024755e-05,
+      "loss": 1.0852,
+      "step": 4042
+    },
+    {
+      "epoch": 1.27027504564462,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.5069501745477629e-05,
+      "loss": 1.2983,
+      "step": 4044
+    },
+    {
+      "epoch": 1.2709032726701612,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.50669628689305e-05,
+      "loss": 1.2256,
+      "step": 4046
+    },
+    {
+      "epoch": 1.2715314996957026,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.5064423992383373e-05,
+      "loss": 1.2011,
+      "step": 4048
+    },
+    {
+      "epoch": 1.2721597267212439,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5061885115836243e-05,
+      "loss": 1.2055,
+      "step": 4050
+    },
+    {
+      "epoch": 1.2727879537467852,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5059346239289116e-05,
+      "loss": 1.1325,
+      "step": 4052
+    },
+    {
+      "epoch": 1.2734161807723265,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5056807362741987e-05,
+      "loss": 1.182,
+      "step": 4054
+    },
+    {
+      "epoch": 1.2740444077978679,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.505426848619486e-05,
+      "loss": 1.2868,
+      "step": 4056
+    },
+    {
+      "epoch": 1.2746726348234092,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5051729609647732e-05,
+      "loss": 1.3885,
+      "step": 4058
+    },
+    {
+      "epoch": 1.2753008618489505,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5049190733100605e-05,
+      "loss": 1.2637,
+      "step": 4060
+    },
+    {
+      "epoch": 1.275929088874492,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5046651856553476e-05,
+      "loss": 1.2665,
+      "step": 4062
+    },
+    {
+      "epoch": 1.2765573159000334,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5044112980006349e-05,
+      "loss": 1.3012,
+      "step": 4064
+    },
+    {
+      "epoch": 1.2771855429255747,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.504157410345922e-05,
+      "loss": 1.2597,
+      "step": 4066
+    },
+    {
+      "epoch": 1.277813769951116,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5039035226912094e-05,
+      "loss": 1.2633,
+      "step": 4068
+    },
+    {
+      "epoch": 1.2784419969766574,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5036496350364965e-05,
+      "loss": 1.2377,
+      "step": 4070
+    },
+    {
+      "epoch": 1.2790702240021987,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5033957473817838e-05,
+      "loss": 1.4411,
+      "step": 4072
+    },
+    {
+      "epoch": 1.2796984510277403,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5031418597270708e-05,
+      "loss": 1.1997,
+      "step": 4074
+    },
+    {
+      "epoch": 1.2803266780532816,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.502887972072358e-05,
+      "loss": 1.2864,
+      "step": 4076
+    },
+    {
+      "epoch": 1.280954905078823,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5026340844176452e-05,
+      "loss": 1.293,
+      "step": 4078
+    },
+    {
+      "epoch": 1.2815831321043643,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5023801967629325e-05,
+      "loss": 1.2866,
+      "step": 4080
+    },
+    {
+      "epoch": 1.2822113591299056,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5021263091082197e-05,
+      "loss": 1.2788,
+      "step": 4082
+    },
+    {
+      "epoch": 1.282839586155447,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.501872421453507e-05,
+      "loss": 1.1705,
+      "step": 4084
+    },
+    {
+      "epoch": 1.2834678131809882,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.5016185337987941e-05,
+      "loss": 1.1257,
+      "step": 4086
+    },
+    {
+      "epoch": 1.2840960402065296,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5013646461440814e-05,
+      "loss": 1.2574,
+      "step": 4088
+    },
+    {
+      "epoch": 1.284724267232071,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5011107584893686e-05,
+      "loss": 1.2126,
+      "step": 4090
+    },
+    {
+      "epoch": 1.2853524942576122,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5008568708346559e-05,
+      "loss": 1.281,
+      "step": 4092
+    },
+    {
+      "epoch": 1.2859807212831538,
+      "grad_norm": 2.875,
+      "learning_rate": 1.5006029831799428e-05,
+      "loss": 1.2376,
+      "step": 4094
+    },
+    {
+      "epoch": 1.286608948308695,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5003490955252303e-05,
+      "loss": 1.2656,
+      "step": 4096
+    },
+    {
+      "epoch": 1.2872371753342364,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.5000952078705173e-05,
+      "loss": 1.3466,
+      "step": 4098
+    },
+    {
+      "epoch": 1.2878654023597778,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4998413202158046e-05,
+      "loss": 1.266,
+      "step": 4100
+    },
+    {
+      "epoch": 1.288493629385319,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4995874325610917e-05,
+      "loss": 1.283,
+      "step": 4102
+    },
+    {
+      "epoch": 1.2891218564108604,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.499333544906379e-05,
+      "loss": 1.2892,
+      "step": 4104
+    },
+    {
+      "epoch": 1.289750083436402,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4990796572516662e-05,
+      "loss": 1.2792,
+      "step": 4106
+    },
+    {
+      "epoch": 1.2903783104619433,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4988257695969535e-05,
+      "loss": 1.3327,
+      "step": 4108
+    },
+    {
+      "epoch": 1.2910065374874846,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4985718819422408e-05,
+      "loss": 1.1793,
+      "step": 4110
+    },
+    {
+      "epoch": 1.291634764513026,
+      "grad_norm": 0.75,
+      "learning_rate": 1.498317994287528e-05,
+      "loss": 1.3966,
+      "step": 4112
+    },
+    {
+      "epoch": 1.2922629915385673,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4980641066328152e-05,
+      "loss": 1.235,
+      "step": 4114
+    },
+    {
+      "epoch": 1.2928912185641086,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.4978102189781024e-05,
+      "loss": 1.1285,
+      "step": 4116
+    },
+    {
+      "epoch": 1.29351944558965,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4975563313233897e-05,
+      "loss": 1.1071,
+      "step": 4118
+    },
+    {
+      "epoch": 1.2941476726151913,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4973024436686766e-05,
+      "loss": 1.3497,
+      "step": 4120
+    },
+    {
+      "epoch": 1.2947758996407326,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4970485560139641e-05,
+      "loss": 1.2041,
+      "step": 4122
+    },
+    {
+      "epoch": 1.295404126666274,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4967946683592511e-05,
+      "loss": 1.2132,
+      "step": 4124
+    },
+    {
+      "epoch": 1.2960323536918152,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.4965407807045384e-05,
+      "loss": 1.2878,
+      "step": 4126
+    },
+    {
+      "epoch": 1.2966605807173568,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4962868930498255e-05,
+      "loss": 1.337,
+      "step": 4128
+    },
+    {
+      "epoch": 1.2972888077428981,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4960330053951128e-05,
+      "loss": 1.2482,
+      "step": 4130
+    },
+    {
+      "epoch": 1.2979170347684394,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4957791177404e-05,
+      "loss": 1.3392,
+      "step": 4132
+    },
+    {
+      "epoch": 1.2985452617939808,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4955252300856873e-05,
+      "loss": 1.3088,
+      "step": 4134
+    },
+    {
+      "epoch": 1.299173488819522,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4952713424309744e-05,
+      "loss": 1.3334,
+      "step": 4136
+    },
+    {
+      "epoch": 1.2998017158450634,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4950174547762617e-05,
+      "loss": 1.1723,
+      "step": 4138
+    },
+    {
+      "epoch": 1.300429942870605,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4947635671215489e-05,
+      "loss": 1.3257,
+      "step": 4140
+    },
+    {
+      "epoch": 1.3010581698961463,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4945096794668362e-05,
+      "loss": 1.3229,
+      "step": 4142
+    },
+    {
+      "epoch": 1.3016863969216876,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4942557918121231e-05,
+      "loss": 1.3304,
+      "step": 4144
+    },
+    {
+      "epoch": 1.302314623947229,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4940019041574105e-05,
+      "loss": 1.3206,
+      "step": 4146
+    },
+    {
+      "epoch": 1.3029428509727703,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4937480165026976e-05,
+      "loss": 1.1349,
+      "step": 4148
+    },
+    {
+      "epoch": 1.3035710779983116,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4934941288479849e-05,
+      "loss": 1.2338,
+      "step": 4150
+    },
+    {
+      "epoch": 1.304199305023853,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.493240241193272e-05,
+      "loss": 1.1981,
+      "step": 4152
+    },
+    {
+      "epoch": 1.3048275320493943,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.4929863535385593e-05,
+      "loss": 1.2488,
+      "step": 4154
+    },
+    {
+      "epoch": 1.3054557590749356,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4927324658838465e-05,
+      "loss": 1.1739,
+      "step": 4156
+    },
+    {
+      "epoch": 1.306083986100477,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4924785782291338e-05,
+      "loss": 1.1962,
+      "step": 4158
+    },
+    {
+      "epoch": 1.3067122131260185,
+      "grad_norm": 0.875,
+      "learning_rate": 1.492224690574421e-05,
+      "loss": 1.2288,
+      "step": 4160
+    },
+    {
+      "epoch": 1.3073404401515598,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4919708029197082e-05,
+      "loss": 1.2643,
+      "step": 4162
+    },
+    {
+      "epoch": 1.3079686671771011,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4917169152649952e-05,
+      "loss": 1.3054,
+      "step": 4164
+    },
+    {
+      "epoch": 1.3085968942026425,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4914630276102827e-05,
+      "loss": 1.135,
+      "step": 4166
+    },
+    {
+      "epoch": 1.3092251212281838,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.4912091399555697e-05,
+      "loss": 1.1145,
+      "step": 4168
+    },
+    {
+      "epoch": 1.3098533482537251,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.490955252300857e-05,
+      "loss": 1.3165,
+      "step": 4170
+    },
+    {
+      "epoch": 1.3104815752792667,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.4907013646461441e-05,
+      "loss": 1.2578,
+      "step": 4172
+    },
+    {
+      "epoch": 1.311109802304808,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4904474769914314e-05,
+      "loss": 1.3692,
+      "step": 4174
+    },
+    {
+      "epoch": 1.3117380293303493,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4901935893367185e-05,
+      "loss": 1.3078,
+      "step": 4176
+    },
+    {
+      "epoch": 1.3123662563558907,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.4899397016820059e-05,
+      "loss": 1.2137,
+      "step": 4178
+    },
+    {
+      "epoch": 1.312994483381432,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.489685814027293e-05,
+      "loss": 1.2295,
+      "step": 4180
+    },
+    {
+      "epoch": 1.3136227104069733,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4894319263725803e-05,
+      "loss": 1.1217,
+      "step": 4182
+    },
+    {
+      "epoch": 1.3142509374325146,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4891780387178674e-05,
+      "loss": 1.2411,
+      "step": 4184
+    },
+    {
+      "epoch": 1.314879164458056,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4889241510631547e-05,
+      "loss": 1.3081,
+      "step": 4186
+    },
+    {
+      "epoch": 1.3155073914835973,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4886702634084417e-05,
+      "loss": 1.3283,
+      "step": 4188
+    },
+    {
+      "epoch": 1.3161356185091386,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.488416375753729e-05,
+      "loss": 1.2081,
+      "step": 4190
+    },
+    {
+      "epoch": 1.3167638455346802,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4881624880990162e-05,
+      "loss": 1.2085,
+      "step": 4192
+    },
+    {
+      "epoch": 1.3173920725602215,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4879086004443035e-05,
+      "loss": 1.3019,
+      "step": 4194
+    },
+    {
+      "epoch": 1.3180202995857628,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4876547127895908e-05,
+      "loss": 1.2424,
+      "step": 4196
+    },
+    {
+      "epoch": 1.3186485266113042,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4874008251348779e-05,
+      "loss": 1.2922,
+      "step": 4198
+    },
+    {
+      "epoch": 1.3192767536368455,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4871469374801652e-05,
+      "loss": 1.262,
+      "step": 4200
+    },
+    {
+      "epoch": 1.3199049806623868,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4868930498254524e-05,
+      "loss": 1.2194,
+      "step": 4202
+    },
+    {
+      "epoch": 1.3205332076879281,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4866391621707397e-05,
+      "loss": 1.1209,
+      "step": 4204
+    },
+    {
+      "epoch": 1.3211614347134697,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4863852745160268e-05,
+      "loss": 1.3312,
+      "step": 4206
+    },
+    {
+      "epoch": 1.321789661739011,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.4861313868613141e-05,
+      "loss": 1.2484,
+      "step": 4208
+    },
+    {
+      "epoch": 1.3224178887645524,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.4858774992066013e-05,
+      "loss": 1.2436,
+      "step": 4210
+    },
+    {
+      "epoch": 1.3230461157900937,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4856236115518886e-05,
+      "loss": 1.2811,
+      "step": 4212
+    },
+    {
+      "epoch": 1.323674342815635,
+      "grad_norm": 2.375,
+      "learning_rate": 1.4853697238971755e-05,
+      "loss": 1.0794,
+      "step": 4214
+    },
+    {
+      "epoch": 1.3243025698411763,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4851158362424628e-05,
+      "loss": 1.2972,
+      "step": 4216
+    },
+    {
+      "epoch": 1.3249307968667177,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.48486194858775e-05,
+      "loss": 1.2615,
+      "step": 4218
+    },
+    {
+      "epoch": 1.325559023892259,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4846080609330373e-05,
+      "loss": 1.2853,
+      "step": 4220
+    },
+    {
+      "epoch": 1.3261872509178003,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4843541732783244e-05,
+      "loss": 1.2105,
+      "step": 4222
+    },
+    {
+      "epoch": 1.3268154779433416,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.4841002856236117e-05,
+      "loss": 1.0318,
+      "step": 4224
+    },
+    {
+      "epoch": 1.3274437049688832,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4838463979688989e-05,
+      "loss": 1.2111,
+      "step": 4226
+    },
+    {
+      "epoch": 1.3280719319944245,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.4835925103141862e-05,
+      "loss": 1.0525,
+      "step": 4228
+    },
+    {
+      "epoch": 1.3287001590199659,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4833386226594733e-05,
+      "loss": 1.2753,
+      "step": 4230
+    },
+    {
+      "epoch": 1.3293283860455072,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4830847350047606e-05,
+      "loss": 1.1766,
+      "step": 4232
+    },
+    {
+      "epoch": 1.3299566130710485,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4828308473500478e-05,
+      "loss": 1.3159,
+      "step": 4234
+    },
+    {
+      "epoch": 1.3305848400965898,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.482576959695335e-05,
+      "loss": 1.1587,
+      "step": 4236
+    },
+    {
+      "epoch": 1.3312130671221314,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.482323072040622e-05,
+      "loss": 1.1362,
+      "step": 4238
+    },
+    {
+      "epoch": 1.3318412941476727,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.4820691843859093e-05,
+      "loss": 1.2305,
+      "step": 4240
+    },
+    {
+      "epoch": 1.332469521173214,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4818152967311965e-05,
+      "loss": 1.163,
+      "step": 4242
+    },
+    {
+      "epoch": 1.3330977481987554,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4815614090764838e-05,
+      "loss": 1.2654,
+      "step": 4244
+    },
+    {
+      "epoch": 1.3337259752242967,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.481307521421771e-05,
+      "loss": 1.1764,
+      "step": 4246
+    },
+    {
+      "epoch": 1.334354202249838,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4810536337670582e-05,
+      "loss": 1.2239,
+      "step": 4248
+    },
+    {
+      "epoch": 1.3349824292753794,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4807997461123454e-05,
+      "loss": 1.2153,
+      "step": 4250
+    },
+    {
+      "epoch": 1.3356106563009207,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4805458584576327e-05,
+      "loss": 1.1689,
+      "step": 4252
+    },
+    {
+      "epoch": 1.336238883326462,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4802919708029198e-05,
+      "loss": 1.2603,
+      "step": 4254
+    },
+    {
+      "epoch": 1.3368671103520033,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4800380831482071e-05,
+      "loss": 1.1904,
+      "step": 4256
+    },
+    {
+      "epoch": 1.337495337377545,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4797841954934941e-05,
+      "loss": 1.4012,
+      "step": 4258
+    },
+    {
+      "epoch": 1.3381235644030862,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.4795303078387816e-05,
+      "loss": 1.2366,
+      "step": 4260
+    },
+    {
+      "epoch": 1.3387517914286275,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4792764201840685e-05,
+      "loss": 1.2833,
+      "step": 4262
+    },
+    {
+      "epoch": 1.3393800184541689,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4790225325293558e-05,
+      "loss": 1.3163,
+      "step": 4264
+    },
+    {
+      "epoch": 1.3400082454797102,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.478768644874643e-05,
+      "loss": 1.2823,
+      "step": 4266
+    },
+    {
+      "epoch": 1.3406364725052515,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4785147572199303e-05,
+      "loss": 1.1118,
+      "step": 4268
+    },
+    {
+      "epoch": 1.3412646995307929,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4782608695652174e-05,
+      "loss": 1.1947,
+      "step": 4270
+    },
+    {
+      "epoch": 1.3418929265563344,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4780069819105047e-05,
+      "loss": 1.3296,
+      "step": 4272
+    },
+    {
+      "epoch": 1.3425211535818757,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4777530942557919e-05,
+      "loss": 1.3039,
+      "step": 4274
+    },
+    {
+      "epoch": 1.343149380607417,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.4774992066010792e-05,
+      "loss": 1.1281,
+      "step": 4276
+    },
+    {
+      "epoch": 1.3437776076329584,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4772453189463663e-05,
+      "loss": 1.2095,
+      "step": 4278
+    },
+    {
+      "epoch": 1.3444058346584997,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4769914312916536e-05,
+      "loss": 1.2037,
+      "step": 4280
+    },
+    {
+      "epoch": 1.345034061684041,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.476737543636941e-05,
+      "loss": 1.2451,
+      "step": 4282
+    },
+    {
+      "epoch": 1.3456622887095824,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4764836559822279e-05,
+      "loss": 1.2812,
+      "step": 4284
+    },
+    {
+      "epoch": 1.3462905157351237,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4762297683275154e-05,
+      "loss": 1.1563,
+      "step": 4286
+    },
+    {
+      "epoch": 1.346918742760665,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4759758806728024e-05,
+      "loss": 1.2046,
+      "step": 4288
+    },
+    {
+      "epoch": 1.3475469697862064,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4757219930180897e-05,
+      "loss": 1.1475,
+      "step": 4290
+    },
+    {
+      "epoch": 1.348175196811748,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4754681053633768e-05,
+      "loss": 1.19,
+      "step": 4292
+    },
+    {
+      "epoch": 1.3488034238372892,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4752142177086641e-05,
+      "loss": 1.1384,
+      "step": 4294
+    },
+    {
+      "epoch": 1.3494316508628306,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4749603300539512e-05,
+      "loss": 1.3681,
+      "step": 4296
+    },
+    {
+      "epoch": 1.350059877888372,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4747064423992386e-05,
+      "loss": 1.4912,
+      "step": 4298
+    },
+    {
+      "epoch": 1.3506881049139132,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4744525547445257e-05,
+      "loss": 1.2292,
+      "step": 4300
+    },
+    {
+      "epoch": 1.3513163319394546,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.474198667089813e-05,
+      "loss": 1.2261,
+      "step": 4302
+    },
+    {
+      "epoch": 1.351944558964996,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4739447794351001e-05,
+      "loss": 1.3621,
+      "step": 4304
+    },
+    {
+      "epoch": 1.3525727859905374,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4736908917803874e-05,
+      "loss": 1.1424,
+      "step": 4306
+    },
+    {
+      "epoch": 1.3532010130160788,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4734370041256744e-05,
+      "loss": 1.1803,
+      "step": 4308
+    },
+    {
+      "epoch": 1.35382924004162,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4731831164709617e-05,
+      "loss": 1.2921,
+      "step": 4310
+    },
+    {
+      "epoch": 1.3544574670671614,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4729292288162489e-05,
+      "loss": 1.359,
+      "step": 4312
+    },
+    {
+      "epoch": 1.3550856940927027,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4726753411615362e-05,
+      "loss": 1.2099,
+      "step": 4314
+    },
+    {
+      "epoch": 1.355713921118244,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4724214535068233e-05,
+      "loss": 1.3183,
+      "step": 4316
+    },
+    {
+      "epoch": 1.3563421481437854,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4721675658521106e-05,
+      "loss": 1.3491,
+      "step": 4318
+    },
+    {
+      "epoch": 1.3569703751693267,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4719136781973978e-05,
+      "loss": 1.239,
+      "step": 4320
+    },
+    {
+      "epoch": 1.357598602194868,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.471659790542685e-05,
+      "loss": 1.2861,
+      "step": 4322
+    },
+    {
+      "epoch": 1.3582268292204096,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4714059028879722e-05,
+      "loss": 1.2679,
+      "step": 4324
+    },
+    {
+      "epoch": 1.358855056245951,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4711520152332595e-05,
+      "loss": 1.2569,
+      "step": 4326
+    },
+    {
+      "epoch": 1.3594832832714923,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4708981275785465e-05,
+      "loss": 1.2359,
+      "step": 4328
+    },
+    {
+      "epoch": 1.3601115102970336,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.470644239923834e-05,
+      "loss": 1.2838,
+      "step": 4330
+    },
+    {
+      "epoch": 1.360739737322575,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.470390352269121e-05,
+      "loss": 1.302,
+      "step": 4332
+    },
+    {
+      "epoch": 1.3613679643481162,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4701364646144082e-05,
+      "loss": 1.3001,
+      "step": 4334
+    },
+    {
+      "epoch": 1.3619961913736578,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4698825769596954e-05,
+      "loss": 1.2303,
+      "step": 4336
+    },
+    {
+      "epoch": 1.3626244183991991,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4696286893049827e-05,
+      "loss": 1.2391,
+      "step": 4338
+    },
+    {
+      "epoch": 1.3632526454247405,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4693748016502698e-05,
+      "loss": 1.2481,
+      "step": 4340
+    },
+    {
+      "epoch": 1.3638808724502818,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4691209139955571e-05,
+      "loss": 1.2994,
+      "step": 4342
+    },
+    {
+      "epoch": 1.364509099475823,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.4688670263408443e-05,
+      "loss": 1.392,
+      "step": 4344
+    },
+    {
+      "epoch": 1.3651373265013644,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4686131386861316e-05,
+      "loss": 1.3212,
+      "step": 4346
+    },
+    {
+      "epoch": 1.3657655535269058,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4683592510314187e-05,
+      "loss": 1.3631,
+      "step": 4348
+    },
+    {
+      "epoch": 1.366393780552447,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.468105363376706e-05,
+      "loss": 1.2164,
+      "step": 4350
+    },
+    {
+      "epoch": 1.3670220075779884,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.467851475721993e-05,
+      "loss": 1.1514,
+      "step": 4352
+    },
+    {
+      "epoch": 1.3676502346035297,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4675975880672803e-05,
+      "loss": 1.3862,
+      "step": 4354
+    },
+    {
+      "epoch": 1.368278461629071,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.4673437004125674e-05,
+      "loss": 1.2115,
+      "step": 4356
+    },
+    {
+      "epoch": 1.3689066886546126,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4670898127578547e-05,
+      "loss": 1.1759,
+      "step": 4358
+    },
+    {
+      "epoch": 1.369534915680154,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4668359251031419e-05,
+      "loss": 1.2677,
+      "step": 4360
+    },
+    {
+      "epoch": 1.3701631427056953,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4665820374484292e-05,
+      "loss": 1.287,
+      "step": 4362
+    },
+    {
+      "epoch": 1.3707913697312366,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4663281497937163e-05,
+      "loss": 1.3592,
+      "step": 4364
+    },
+    {
+      "epoch": 1.371419596756778,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4660742621390036e-05,
+      "loss": 1.2563,
+      "step": 4366
+    },
+    {
+      "epoch": 1.3720478237823193,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.465820374484291e-05,
+      "loss": 1.1968,
+      "step": 4368
+    },
+    {
+      "epoch": 1.3726760508078608,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.465566486829578e-05,
+      "loss": 1.3271,
+      "step": 4370
+    },
+    {
+      "epoch": 1.3733042778334021,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4653125991748654e-05,
+      "loss": 1.304,
+      "step": 4372
+    },
+    {
+      "epoch": 1.3739325048589435,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4650587115201525e-05,
+      "loss": 1.3881,
+      "step": 4374
+    },
+    {
+      "epoch": 1.3745607318844848,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4648048238654398e-05,
+      "loss": 1.2676,
+      "step": 4376
+    },
+    {
+      "epoch": 1.3751889589100261,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4645509362107268e-05,
+      "loss": 1.2412,
+      "step": 4378
+    },
+    {
+      "epoch": 1.3758171859355675,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4642970485560141e-05,
+      "loss": 1.1142,
+      "step": 4380
+    },
+    {
+      "epoch": 1.3764454129611088,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4640431609013012e-05,
+      "loss": 1.2977,
+      "step": 4382
+    },
+    {
+      "epoch": 1.37707363998665,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4637892732465885e-05,
+      "loss": 1.1754,
+      "step": 4384
+    },
+    {
+      "epoch": 1.3777018670121914,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4635353855918757e-05,
+      "loss": 1.3088,
+      "step": 4386
+    },
+    {
+      "epoch": 1.3783300940377328,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.463281497937163e-05,
+      "loss": 1.359,
+      "step": 4388
+    },
+    {
+      "epoch": 1.3789583210632743,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4630276102824501e-05,
+      "loss": 1.1869,
+      "step": 4390
+    },
+    {
+      "epoch": 1.3795865480888156,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4627737226277374e-05,
+      "loss": 1.3206,
+      "step": 4392
+    },
+    {
+      "epoch": 1.380214775114357,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4625198349730246e-05,
+      "loss": 1.308,
+      "step": 4394
+    },
+    {
+      "epoch": 1.3808430021398983,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4622659473183119e-05,
+      "loss": 1.2614,
+      "step": 4396
+    },
+    {
+      "epoch": 1.3814712291654396,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.462012059663599e-05,
+      "loss": 1.1872,
+      "step": 4398
+    },
+    {
+      "epoch": 1.382099456190981,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4617581720088863e-05,
+      "loss": 1.2658,
+      "step": 4400
+    },
+    {
+      "epoch": 1.3827276832165225,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4615042843541733e-05,
+      "loss": 1.3195,
+      "step": 4402
+    },
+    {
+      "epoch": 1.3833559102420638,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4612503966994606e-05,
+      "loss": 1.1573,
+      "step": 4404
+    },
+    {
+      "epoch": 1.3839841372676052,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4609965090447477e-05,
+      "loss": 1.2768,
+      "step": 4406
+    },
+    {
+      "epoch": 1.3846123642931465,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.460742621390035e-05,
+      "loss": 1.2157,
+      "step": 4408
+    },
+    {
+      "epoch": 1.3852405913186878,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4604887337353222e-05,
+      "loss": 1.1608,
+      "step": 4410
+    },
+    {
+      "epoch": 1.3858688183442291,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.4602348460806095e-05,
+      "loss": 1.1619,
+      "step": 4412
+    },
+    {
+      "epoch": 1.3864970453697705,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4599809584258966e-05,
+      "loss": 1.1692,
+      "step": 4414
+    },
+    {
+      "epoch": 1.3871252723953118,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.459727070771184e-05,
+      "loss": 1.3201,
+      "step": 4416
+    },
+    {
+      "epoch": 1.3877534994208531,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.459473183116471e-05,
+      "loss": 1.2667,
+      "step": 4418
+    },
+    {
+      "epoch": 1.3883817264463945,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4592192954617584e-05,
+      "loss": 1.3936,
+      "step": 4420
+    },
+    {
+      "epoch": 1.3890099534719358,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4589654078070454e-05,
+      "loss": 1.271,
+      "step": 4422
+    },
+    {
+      "epoch": 1.3896381804974773,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4587115201523328e-05,
+      "loss": 1.2298,
+      "step": 4424
+    },
+    {
+      "epoch": 1.3902664075230187,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4584576324976198e-05,
+      "loss": 1.3023,
+      "step": 4426
+    },
+    {
+      "epoch": 1.39089463454856,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4582037448429071e-05,
+      "loss": 1.0285,
+      "step": 4428
+    },
+    {
+      "epoch": 1.3915228615741013,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4579498571881942e-05,
+      "loss": 1.345,
+      "step": 4430
+    },
+    {
+      "epoch": 1.3921510885996426,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4576959695334816e-05,
+      "loss": 1.3171,
+      "step": 4432
+    },
+    {
+      "epoch": 1.392779315625184,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4574420818787687e-05,
+      "loss": 1.1312,
+      "step": 4434
+    },
+    {
+      "epoch": 1.3934075426507255,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.457188194224056e-05,
+      "loss": 1.281,
+      "step": 4436
+    },
+    {
+      "epoch": 1.3940357696762669,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4569343065693431e-05,
+      "loss": 1.3061,
+      "step": 4438
+    },
+    {
+      "epoch": 1.3946639967018082,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.4566804189146304e-05,
+      "loss": 1.2167,
+      "step": 4440
+    },
+    {
+      "epoch": 1.3952922237273495,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4564265312599176e-05,
+      "loss": 1.2837,
+      "step": 4442
+    },
+    {
+      "epoch": 1.3959204507528908,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.4561726436052049e-05,
+      "loss": 1.1497,
+      "step": 4444
+    },
+    {
+      "epoch": 1.3965486777784322,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.4559187559504919e-05,
+      "loss": 1.2668,
+      "step": 4446
+    },
+    {
+      "epoch": 1.3971769048039735,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4556648682957792e-05,
+      "loss": 1.2495,
+      "step": 4448
+    },
+    {
+      "epoch": 1.3978051318295148,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4554109806410663e-05,
+      "loss": 1.3183,
+      "step": 4450
+    },
+    {
+      "epoch": 1.3984333588550562,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4551570929863536e-05,
+      "loss": 1.2387,
+      "step": 4452
+    },
+    {
+      "epoch": 1.3990615858805975,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.454903205331641e-05,
+      "loss": 1.3808,
+      "step": 4454
+    },
+    {
+      "epoch": 1.399689812906139,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.454649317676928e-05,
+      "loss": 1.2578,
+      "step": 4456
+    },
+    {
+      "epoch": 1.4003180399316804,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4543954300222154e-05,
+      "loss": 1.3209,
+      "step": 4458
+    },
+    {
+      "epoch": 1.4009462669572217,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4541415423675025e-05,
+      "loss": 1.2369,
+      "step": 4460
+    },
+    {
+      "epoch": 1.401574493982763,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4538876547127898e-05,
+      "loss": 1.2486,
+      "step": 4462
+    },
+    {
+      "epoch": 1.4022027210083043,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.453633767058077e-05,
+      "loss": 1.1907,
+      "step": 4464
+    },
+    {
+      "epoch": 1.4028309480338457,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4533798794033643e-05,
+      "loss": 1.1616,
+      "step": 4466
+    },
+    {
+      "epoch": 1.4034591750593872,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4531259917486514e-05,
+      "loss": 1.1429,
+      "step": 4468
+    },
+    {
+      "epoch": 1.4040874020849285,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4528721040939387e-05,
+      "loss": 1.2914,
+      "step": 4470
+    },
+    {
+      "epoch": 1.4047156291104699,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4526182164392257e-05,
+      "loss": 1.3573,
+      "step": 4472
+    },
+    {
+      "epoch": 1.4053438561360112,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.452364328784513e-05,
+      "loss": 1.258,
+      "step": 4474
+    },
+    {
+      "epoch": 1.4059720831615525,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4521104411298001e-05,
+      "loss": 1.3615,
+      "step": 4476
+    },
+    {
+      "epoch": 1.4066003101870939,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4518565534750874e-05,
+      "loss": 1.2793,
+      "step": 4478
+    },
+    {
+      "epoch": 1.4072285372126352,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4516026658203746e-05,
+      "loss": 1.2621,
+      "step": 4480
+    },
+    {
+      "epoch": 1.4078567642381765,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4513487781656619e-05,
+      "loss": 1.2854,
+      "step": 4482
+    },
+    {
+      "epoch": 1.4084849912637178,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.451094890510949e-05,
+      "loss": 1.2765,
+      "step": 4484
+    },
+    {
+      "epoch": 1.4091132182892592,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4508410028562363e-05,
+      "loss": 1.2565,
+      "step": 4486
+    },
+    {
+      "epoch": 1.4097414453148005,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4505871152015235e-05,
+      "loss": 1.1619,
+      "step": 4488
+    },
+    {
+      "epoch": 1.410369672340342,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4503332275468108e-05,
+      "loss": 1.2787,
+      "step": 4490
+    },
+    {
+      "epoch": 1.4109978993658834,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4500793398920977e-05,
+      "loss": 1.1383,
+      "step": 4492
+    },
+    {
+      "epoch": 1.4116261263914247,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4498254522373852e-05,
+      "loss": 1.2907,
+      "step": 4494
+    },
+    {
+      "epoch": 1.412254353416966,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4495715645826722e-05,
+      "loss": 1.2964,
+      "step": 4496
+    },
+    {
+      "epoch": 1.4128825804425074,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4493176769279595e-05,
+      "loss": 1.4016,
+      "step": 4498
+    },
+    {
+      "epoch": 1.4135108074680487,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4490637892732466e-05,
+      "loss": 1.1816,
+      "step": 4500
+    },
+    {
+      "epoch": 1.4141390344935902,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.448809901618534e-05,
+      "loss": 1.2369,
+      "step": 4502
+    },
+    {
+      "epoch": 1.4147672615191316,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.448556013963821e-05,
+      "loss": 1.2545,
+      "step": 4504
+    },
+    {
+      "epoch": 1.415395488544673,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4483021263091084e-05,
+      "loss": 1.3477,
+      "step": 4506
+    },
+    {
+      "epoch": 1.4160237155702142,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4480482386543955e-05,
+      "loss": 1.2051,
+      "step": 4508
+    },
+    {
+      "epoch": 1.4166519425957556,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.4477943509996828e-05,
+      "loss": 1.1868,
+      "step": 4510
+    },
+    {
+      "epoch": 1.4172801696212969,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.44754046334497e-05,
+      "loss": 1.2113,
+      "step": 4512
+    },
+    {
+      "epoch": 1.4179083966468382,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4472865756902573e-05,
+      "loss": 1.2932,
+      "step": 4514
+    },
+    {
+      "epoch": 1.4185366236723795,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4470326880355442e-05,
+      "loss": 1.2502,
+      "step": 4516
+    },
+    {
+      "epoch": 1.4191648506979209,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4467788003808315e-05,
+      "loss": 1.3597,
+      "step": 4518
+    },
+    {
+      "epoch": 1.4197930777234622,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4465249127261187e-05,
+      "loss": 1.2579,
+      "step": 4520
+    },
+    {
+      "epoch": 1.4204213047490037,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.446271025071406e-05,
+      "loss": 1.315,
+      "step": 4522
+    },
+    {
+      "epoch": 1.421049531774545,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4460171374166931e-05,
+      "loss": 1.1785,
+      "step": 4524
+    },
+    {
+      "epoch": 1.4216777588000864,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4457632497619804e-05,
+      "loss": 1.2612,
+      "step": 4526
+    },
+    {
+      "epoch": 1.4223059858256277,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4455093621072676e-05,
+      "loss": 1.2523,
+      "step": 4528
+    },
+    {
+      "epoch": 1.422934212851169,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4452554744525549e-05,
+      "loss": 1.288,
+      "step": 4530
+    },
+    {
+      "epoch": 1.4235624398767104,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.445001586797842e-05,
+      "loss": 1.2778,
+      "step": 4532
+    },
+    {
+      "epoch": 1.424190666902252,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4447476991431293e-05,
+      "loss": 1.3666,
+      "step": 4534
+    },
+    {
+      "epoch": 1.4248188939277933,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4444938114884165e-05,
+      "loss": 1.2365,
+      "step": 4536
+    },
+    {
+      "epoch": 1.4254471209533346,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4442399238337038e-05,
+      "loss": 1.4104,
+      "step": 4538
+    },
+    {
+      "epoch": 1.426075347978876,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4439860361789911e-05,
+      "loss": 1.3859,
+      "step": 4540
+    },
+    {
+      "epoch": 1.4267035750044172,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.443732148524278e-05,
+      "loss": 1.2868,
+      "step": 4542
+    },
+    {
+      "epoch": 1.4273318020299586,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4434782608695654e-05,
+      "loss": 1.2102,
+      "step": 4544
+    },
+    {
+      "epoch": 1.4279600290555,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4432243732148525e-05,
+      "loss": 1.4068,
+      "step": 4546
+    },
+    {
+      "epoch": 1.4285882560810412,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4429704855601398e-05,
+      "loss": 1.1468,
+      "step": 4548
+    },
+    {
+      "epoch": 1.4292164831065826,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.442716597905427e-05,
+      "loss": 1.114,
+      "step": 4550
+    },
+    {
+      "epoch": 1.4298447101321239,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4424627102507143e-05,
+      "loss": 1.1786,
+      "step": 4552
+    },
+    {
+      "epoch": 1.4304729371576652,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4422088225960014e-05,
+      "loss": 1.2882,
+      "step": 4554
+    },
+    {
+      "epoch": 1.4311011641832068,
+      "grad_norm": 0.875,
+      "learning_rate": 1.4419549349412887e-05,
+      "loss": 1.2154,
+      "step": 4556
+    },
+    {
+      "epoch": 1.431729391208748,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4417010472865758e-05,
+      "loss": 1.3404,
+      "step": 4558
+    },
+    {
+      "epoch": 1.4323576182342894,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4414471596318631e-05,
+      "loss": 1.17,
+      "step": 4560
+    },
+    {
+      "epoch": 1.4329858452598307,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4411932719771503e-05,
+      "loss": 1.2197,
+      "step": 4562
+    },
+    {
+      "epoch": 1.433614072285372,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4409393843224376e-05,
+      "loss": 1.2704,
+      "step": 4564
+    },
+    {
+      "epoch": 1.4342422993109134,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4406854966677246e-05,
+      "loss": 1.217,
+      "step": 4566
+    },
+    {
+      "epoch": 1.434870526336455,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4404316090130119e-05,
+      "loss": 1.158,
+      "step": 4568
+    },
+    {
+      "epoch": 1.4354987533619963,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.440177721358299e-05,
+      "loss": 1.2378,
+      "step": 4570
+    },
+    {
+      "epoch": 1.4361269803875376,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4399238337035863e-05,
+      "loss": 1.1346,
+      "step": 4572
+    },
+    {
+      "epoch": 1.436755207413079,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4396699460488735e-05,
+      "loss": 1.3427,
+      "step": 4574
+    },
+    {
+      "epoch": 1.4373834344386203,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4394160583941608e-05,
+      "loss": 1.2679,
+      "step": 4576
+    },
+    {
+      "epoch": 1.4380116614641616,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4391621707394479e-05,
+      "loss": 1.1892,
+      "step": 4578
+    },
+    {
+      "epoch": 1.438639888489703,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4389082830847352e-05,
+      "loss": 1.3267,
+      "step": 4580
+    },
+    {
+      "epoch": 1.4392681155152443,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4386543954300223e-05,
+      "loss": 1.2524,
+      "step": 4582
+    },
+    {
+      "epoch": 1.4398963425407856,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4384005077753097e-05,
+      "loss": 1.1367,
+      "step": 4584
+    },
+    {
+      "epoch": 1.440524569566327,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4381466201205966e-05,
+      "loss": 1.3201,
+      "step": 4586
+    },
+    {
+      "epoch": 1.4411527965918685,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.437892732465884e-05,
+      "loss": 1.2344,
+      "step": 4588
+    },
+    {
+      "epoch": 1.4417810236174098,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.437638844811171e-05,
+      "loss": 1.1569,
+      "step": 4590
+    },
+    {
+      "epoch": 1.4424092506429511,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4373849571564584e-05,
+      "loss": 1.1568,
+      "step": 4592
+    },
+    {
+      "epoch": 1.4430374776684924,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4371310695017455e-05,
+      "loss": 1.1668,
+      "step": 4594
+    },
+    {
+      "epoch": 1.4436657046940338,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4368771818470328e-05,
+      "loss": 1.318,
+      "step": 4596
+    },
+    {
+      "epoch": 1.444293931719575,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.43662329419232e-05,
+      "loss": 1.3324,
+      "step": 4598
+    },
+    {
+      "epoch": 1.4449221587451166,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4363694065376073e-05,
+      "loss": 1.2083,
+      "step": 4600
+    },
+    {
+      "epoch": 1.445550385770658,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4361155188828944e-05,
+      "loss": 1.1944,
+      "step": 4602
+    },
+    {
+      "epoch": 1.4461786127961993,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4358616312281817e-05,
+      "loss": 1.2482,
+      "step": 4604
+    },
+    {
+      "epoch": 1.4468068398217406,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4356077435734688e-05,
+      "loss": 1.162,
+      "step": 4606
+    },
+    {
+      "epoch": 1.447435066847282,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4353538559187562e-05,
+      "loss": 1.1867,
+      "step": 4608
+    },
+    {
+      "epoch": 1.4480632938728233,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4350999682640431e-05,
+      "loss": 1.2864,
+      "step": 4610
+    },
+    {
+      "epoch": 1.4486915208983646,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.4348460806093304e-05,
+      "loss": 1.1464,
+      "step": 4612
+    },
+    {
+      "epoch": 1.449319747923906,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4345921929546176e-05,
+      "loss": 1.1811,
+      "step": 4614
+    },
+    {
+      "epoch": 1.4499479749494473,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4343383052999049e-05,
+      "loss": 1.2498,
+      "step": 4616
+    },
+    {
+      "epoch": 1.4505762019749886,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.434084417645192e-05,
+      "loss": 1.192,
+      "step": 4618
+    },
+    {
+      "epoch": 1.4512044290005301,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4338305299904793e-05,
+      "loss": 1.3238,
+      "step": 4620
+    },
+    {
+      "epoch": 1.4518326560260715,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4335766423357665e-05,
+      "loss": 1.1391,
+      "step": 4622
+    },
+    {
+      "epoch": 1.4524608830516128,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4333227546810538e-05,
+      "loss": 1.2083,
+      "step": 4624
+    },
+    {
+      "epoch": 1.4530891100771541,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.433068867026341e-05,
+      "loss": 1.2331,
+      "step": 4626
+    },
+    {
+      "epoch": 1.4537173371026955,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4328149793716282e-05,
+      "loss": 1.2771,
+      "step": 4628
+    },
+    {
+      "epoch": 1.4543455641282368,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4325610917169155e-05,
+      "loss": 1.1523,
+      "step": 4630
+    },
+    {
+      "epoch": 1.4549737911537781,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4323072040622027e-05,
+      "loss": 1.3079,
+      "step": 4632
+    },
+    {
+      "epoch": 1.4556020181793197,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.43205331640749e-05,
+      "loss": 1.1524,
+      "step": 4634
+    },
+    {
+      "epoch": 1.456230245204861,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.431799428752777e-05,
+      "loss": 1.1978,
+      "step": 4636
+    },
+    {
+      "epoch": 1.4568584722304023,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4315455410980642e-05,
+      "loss": 1.2417,
+      "step": 4638
+    },
+    {
+      "epoch": 1.4574866992559437,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4312916534433514e-05,
+      "loss": 1.3005,
+      "step": 4640
+    },
+    {
+      "epoch": 1.458114926281485,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4310377657886387e-05,
+      "loss": 1.2385,
+      "step": 4642
+    },
+    {
+      "epoch": 1.4587431533070263,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4307838781339258e-05,
+      "loss": 1.324,
+      "step": 4644
+    },
+    {
+      "epoch": 1.4593713803325676,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4305299904792131e-05,
+      "loss": 1.3922,
+      "step": 4646
+    },
+    {
+      "epoch": 1.459999607358109,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4302761028245003e-05,
+      "loss": 1.3316,
+      "step": 4648
+    },
+    {
+      "epoch": 1.4606278343836503,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4300222151697876e-05,
+      "loss": 1.2378,
+      "step": 4650
+    },
+    {
+      "epoch": 1.4612560614091916,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4297683275150747e-05,
+      "loss": 1.2244,
+      "step": 4652
+    },
+    {
+      "epoch": 1.4618842884347332,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.429514439860362e-05,
+      "loss": 1.2384,
+      "step": 4654
+    },
+    {
+      "epoch": 1.4625125154602745,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.429260552205649e-05,
+      "loss": 1.39,
+      "step": 4656
+    },
+    {
+      "epoch": 1.4631407424858158,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4290066645509365e-05,
+      "loss": 1.1775,
+      "step": 4658
+    },
+    {
+      "epoch": 1.4637689695113572,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4287527768962234e-05,
+      "loss": 1.2629,
+      "step": 4660
+    },
+    {
+      "epoch": 1.4643971965368985,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4284988892415108e-05,
+      "loss": 1.1894,
+      "step": 4662
+    },
+    {
+      "epoch": 1.4650254235624398,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4282450015867979e-05,
+      "loss": 1.4307,
+      "step": 4664
+    },
+    {
+      "epoch": 1.4656536505879814,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4279911139320852e-05,
+      "loss": 1.2757,
+      "step": 4666
+    },
+    {
+      "epoch": 1.4662818776135227,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4277372262773723e-05,
+      "loss": 1.1575,
+      "step": 4668
+    },
+    {
+      "epoch": 1.466910104639064,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4274833386226596e-05,
+      "loss": 1.2476,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4675383316646053,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4272294509679468e-05,
+      "loss": 1.1947,
+      "step": 4672
+    },
+    {
+      "epoch": 1.4681665586901467,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4269755633132341e-05,
+      "loss": 1.2767,
+      "step": 4674
+    },
+    {
+      "epoch": 1.468794785715688,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4267216756585212e-05,
+      "loss": 1.1294,
+      "step": 4676
+    },
+    {
+      "epoch": 1.4694230127412293,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4264677880038085e-05,
+      "loss": 1.1821,
+      "step": 4678
+    },
+    {
+      "epoch": 1.4700512397667707,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4262139003490955e-05,
+      "loss": 1.1981,
+      "step": 4680
+    },
+    {
+      "epoch": 1.470679466792312,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4259600126943828e-05,
+      "loss": 1.2921,
+      "step": 4682
+    },
+    {
+      "epoch": 1.4713076938178533,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.42570612503967e-05,
+      "loss": 1.2258,
+      "step": 4684
+    },
+    {
+      "epoch": 1.4719359208433949,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4254522373849573e-05,
+      "loss": 1.1364,
+      "step": 4686
+    },
+    {
+      "epoch": 1.4725641478689362,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4251983497302444e-05,
+      "loss": 1.2497,
+      "step": 4688
+    },
+    {
+      "epoch": 1.4731923748944775,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4249444620755317e-05,
+      "loss": 1.1485,
+      "step": 4690
+    },
+    {
+      "epoch": 1.4738206019200188,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4246905744208188e-05,
+      "loss": 1.1257,
+      "step": 4692
+    },
+    {
+      "epoch": 1.4744488289455602,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4244366867661061e-05,
+      "loss": 1.2718,
+      "step": 4694
+    },
+    {
+      "epoch": 1.4750770559711015,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4241827991113933e-05,
+      "loss": 1.0906,
+      "step": 4696
+    },
+    {
+      "epoch": 1.4757052829966428,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4239289114566806e-05,
+      "loss": 1.2542,
+      "step": 4698
+    },
+    {
+      "epoch": 1.4763335100221844,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4236750238019676e-05,
+      "loss": 1.2242,
+      "step": 4700
+    },
+    {
+      "epoch": 1.4769617370477257,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.423421136147255e-05,
+      "loss": 1.1202,
+      "step": 4702
+    },
+    {
+      "epoch": 1.477589964073267,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.423167248492542e-05,
+      "loss": 1.2829,
+      "step": 4704
+    },
+    {
+      "epoch": 1.4782181910988084,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4229133608378293e-05,
+      "loss": 1.2276,
+      "step": 4706
+    },
+    {
+      "epoch": 1.4788464181243497,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4226594731831165e-05,
+      "loss": 1.0959,
+      "step": 4708
+    },
+    {
+      "epoch": 1.479474645149891,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4224055855284038e-05,
+      "loss": 1.2402,
+      "step": 4710
+    },
+    {
+      "epoch": 1.4801028721754323,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.422151697873691e-05,
+      "loss": 1.41,
+      "step": 4712
+    },
+    {
+      "epoch": 1.4807310992009737,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4218978102189782e-05,
+      "loss": 1.2751,
+      "step": 4714
+    },
+    {
+      "epoch": 1.481359326226515,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4216439225642655e-05,
+      "loss": 1.1848,
+      "step": 4716
+    },
+    {
+      "epoch": 1.4819875532520563,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4213900349095527e-05,
+      "loss": 1.1541,
+      "step": 4718
+    },
+    {
+      "epoch": 1.4826157802775979,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.42113614725484e-05,
+      "loss": 1.248,
+      "step": 4720
+    },
+    {
+      "epoch": 1.4832440073031392,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4208822596001271e-05,
+      "loss": 1.0783,
+      "step": 4722
+    },
+    {
+      "epoch": 1.4838722343286805,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4206283719454144e-05,
+      "loss": 1.3559,
+      "step": 4724
+    },
+    {
+      "epoch": 1.4845004613542219,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4203744842907014e-05,
+      "loss": 1.2525,
+      "step": 4726
+    },
+    {
+      "epoch": 1.4851286883797632,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4201205966359889e-05,
+      "loss": 1.3513,
+      "step": 4728
+    },
+    {
+      "epoch": 1.4857569154053045,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.4198667089812758e-05,
+      "loss": 1.1958,
+      "step": 4730
+    },
+    {
+      "epoch": 1.486385142430846,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4196128213265631e-05,
+      "loss": 1.2256,
+      "step": 4732
+    },
+    {
+      "epoch": 1.4870133694563874,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4193589336718503e-05,
+      "loss": 1.2719,
+      "step": 4734
+    },
+    {
+      "epoch": 1.4876415964819287,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.4191050460171376e-05,
+      "loss": 1.186,
+      "step": 4736
+    },
+    {
+      "epoch": 1.48826982350747,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4188511583624247e-05,
+      "loss": 1.3374,
+      "step": 4738
+    },
+    {
+      "epoch": 1.4888980505330114,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.418597270707712e-05,
+      "loss": 1.139,
+      "step": 4740
+    },
+    {
+      "epoch": 1.4895262775585527,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4183433830529992e-05,
+      "loss": 1.269,
+      "step": 4742
+    },
+    {
+      "epoch": 1.490154504584094,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4180894953982865e-05,
+      "loss": 1.1428,
+      "step": 4744
+    },
+    {
+      "epoch": 1.4907827316096354,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4178356077435736e-05,
+      "loss": 1.304,
+      "step": 4746
+    },
+    {
+      "epoch": 1.4914109586351767,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4175817200888609e-05,
+      "loss": 1.4207,
+      "step": 4748
+    },
+    {
+      "epoch": 1.492039185660718,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4173278324341479e-05,
+      "loss": 1.2348,
+      "step": 4750
+    },
+    {
+      "epoch": 1.4926674126862596,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4170739447794352e-05,
+      "loss": 1.2315,
+      "step": 4752
+    },
+    {
+      "epoch": 1.493295639711801,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4168200571247223e-05,
+      "loss": 1.408,
+      "step": 4754
+    },
+    {
+      "epoch": 1.4939238667373422,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4165661694700096e-05,
+      "loss": 1.2126,
+      "step": 4756
+    },
+    {
+      "epoch": 1.4945520937628836,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4163122818152968e-05,
+      "loss": 1.3488,
+      "step": 4758
+    },
+    {
+      "epoch": 1.4951803207884249,
+      "grad_norm": 0.75,
+      "learning_rate": 1.416058394160584e-05,
+      "loss": 1.2919,
+      "step": 4760
+    },
+    {
+      "epoch": 1.4958085478139662,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4158045065058712e-05,
+      "loss": 1.2386,
+      "step": 4762
+    },
+    {
+      "epoch": 1.4964367748395075,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.4155506188511585e-05,
+      "loss": 1.1495,
+      "step": 4764
+    },
+    {
+      "epoch": 1.497065001865049,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4152967311964457e-05,
+      "loss": 1.0899,
+      "step": 4766
+    },
+    {
+      "epoch": 1.4976932288905904,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.415042843541733e-05,
+      "loss": 1.1979,
+      "step": 4768
+    },
+    {
+      "epoch": 1.4983214559161318,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4147889558870201e-05,
+      "loss": 1.2938,
+      "step": 4770
+    },
+    {
+      "epoch": 1.498949682941673,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4145350682323074e-05,
+      "loss": 1.1939,
+      "step": 4772
+    },
+    {
+      "epoch": 1.4995779099672144,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4142811805775944e-05,
+      "loss": 1.2171,
+      "step": 4774
+    },
+    {
+      "epoch": 1.5002061369927557,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4140272929228817e-05,
+      "loss": 1.202,
+      "step": 4776
+    },
+    {
+      "epoch": 1.500834364018297,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4137734052681688e-05,
+      "loss": 1.2284,
+      "step": 4778
+    },
+    {
+      "epoch": 1.5014625910438384,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4135195176134561e-05,
+      "loss": 1.2907,
+      "step": 4780
+    },
+    {
+      "epoch": 1.5020908180693797,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.4132656299587433e-05,
+      "loss": 1.304,
+      "step": 4782
+    },
+    {
+      "epoch": 1.502719045094921,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4130117423040306e-05,
+      "loss": 1.3084,
+      "step": 4784
+    },
+    {
+      "epoch": 1.5033472721204624,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4127578546493177e-05,
+      "loss": 1.277,
+      "step": 4786
+    },
+    {
+      "epoch": 1.503975499146004,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.412503966994605e-05,
+      "loss": 1.2962,
+      "step": 4788
+    },
+    {
+      "epoch": 1.5046037261715453,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4122500793398922e-05,
+      "loss": 1.1253,
+      "step": 4790
+    },
+    {
+      "epoch": 1.5052319531970866,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.4119961916851795e-05,
+      "loss": 1.1104,
+      "step": 4792
+    },
+    {
+      "epoch": 1.505860180222628,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4117423040304664e-05,
+      "loss": 1.2353,
+      "step": 4794
+    },
+    {
+      "epoch": 1.5064884072481695,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.411488416375754e-05,
+      "loss": 1.2573,
+      "step": 4796
+    },
+    {
+      "epoch": 1.5071166342737108,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4112345287210412e-05,
+      "loss": 1.2517,
+      "step": 4798
+    },
+    {
+      "epoch": 1.5077448612992521,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4109806410663282e-05,
+      "loss": 1.3036,
+      "step": 4800
+    },
+    {
+      "epoch": 1.5083730883247934,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4107267534116155e-05,
+      "loss": 1.364,
+      "step": 4802
+    },
+    {
+      "epoch": 1.5090013153503348,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.4104728657569026e-05,
+      "loss": 1.1784,
+      "step": 4804
+    },
+    {
+      "epoch": 1.509629542375876,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.41021897810219e-05,
+      "loss": 1.1693,
+      "step": 4806
+    },
+    {
+      "epoch": 1.5102577694014174,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4099650904474771e-05,
+      "loss": 1.2911,
+      "step": 4808
+    },
+    {
+      "epoch": 1.5108859964269588,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4097112027927644e-05,
+      "loss": 1.1448,
+      "step": 4810
+    },
+    {
+      "epoch": 1.5115142234525,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4094573151380515e-05,
+      "loss": 1.3268,
+      "step": 4812
+    },
+    {
+      "epoch": 1.5121424504780414,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4092034274833388e-05,
+      "loss": 1.3592,
+      "step": 4814
+    },
+    {
+      "epoch": 1.5127706775035827,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.408949539828626e-05,
+      "loss": 1.2284,
+      "step": 4816
+    },
+    {
+      "epoch": 1.513398904529124,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4086956521739133e-05,
+      "loss": 1.3063,
+      "step": 4818
+    },
+    {
+      "epoch": 1.5140271315546656,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4084417645192003e-05,
+      "loss": 1.291,
+      "step": 4820
+    },
+    {
+      "epoch": 1.514655358580207,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4081878768644877e-05,
+      "loss": 1.1867,
+      "step": 4822
+    },
+    {
+      "epoch": 1.5152835856057483,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4079339892097747e-05,
+      "loss": 1.1726,
+      "step": 4824
+    },
+    {
+      "epoch": 1.5159118126312896,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.407680101555062e-05,
+      "loss": 1.2812,
+      "step": 4826
+    },
+    {
+      "epoch": 1.5165400396568312,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4074262139003492e-05,
+      "loss": 1.2503,
+      "step": 4828
+    },
+    {
+      "epoch": 1.5171682666823725,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4071723262456365e-05,
+      "loss": 1.1713,
+      "step": 4830
+    },
+    {
+      "epoch": 1.5177964937079138,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4069184385909236e-05,
+      "loss": 1.2758,
+      "step": 4832
+    },
+    {
+      "epoch": 1.5184247207334551,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4066645509362109e-05,
+      "loss": 1.4859,
+      "step": 4834
+    },
+    {
+      "epoch": 1.5190529477589965,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.406410663281498e-05,
+      "loss": 1.1555,
+      "step": 4836
+    },
+    {
+      "epoch": 1.5196811747845378,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.4061567756267854e-05,
+      "loss": 1.211,
+      "step": 4838
+    },
+    {
+      "epoch": 1.5203094018100791,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4059028879720725e-05,
+      "loss": 1.1481,
+      "step": 4840
+    },
+    {
+      "epoch": 1.5209376288356204,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.4056490003173598e-05,
+      "loss": 1.1373,
+      "step": 4842
+    },
+    {
+      "epoch": 1.5215658558611618,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4053951126626468e-05,
+      "loss": 1.2099,
+      "step": 4844
+    },
+    {
+      "epoch": 1.522194082886703,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.405141225007934e-05,
+      "loss": 1.331,
+      "step": 4846
+    },
+    {
+      "epoch": 1.5228223099122444,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4048873373532212e-05,
+      "loss": 1.2244,
+      "step": 4848
+    },
+    {
+      "epoch": 1.5234505369377858,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4046334496985085e-05,
+      "loss": 1.301,
+      "step": 4850
+    },
+    {
+      "epoch": 1.524078763963327,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4043795620437957e-05,
+      "loss": 1.2133,
+      "step": 4852
+    },
+    {
+      "epoch": 1.5247069909888686,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.404125674389083e-05,
+      "loss": 1.1242,
+      "step": 4854
+    },
+    {
+      "epoch": 1.52533521801441,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4038717867343701e-05,
+      "loss": 1.2627,
+      "step": 4856
+    },
+    {
+      "epoch": 1.5259634450399513,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4036178990796574e-05,
+      "loss": 1.3087,
+      "step": 4858
+    },
+    {
+      "epoch": 1.5265916720654926,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4033640114249446e-05,
+      "loss": 1.1323,
+      "step": 4860
+    },
+    {
+      "epoch": 1.5272198990910342,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4031101237702319e-05,
+      "loss": 1.2379,
+      "step": 4862
+    },
+    {
+      "epoch": 1.5278481261165755,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4028562361155188e-05,
+      "loss": 1.2671,
+      "step": 4864
+    },
+    {
+      "epoch": 1.5284763531421168,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4026023484608063e-05,
+      "loss": 1.2881,
+      "step": 4866
+    },
+    {
+      "epoch": 1.5291045801676582,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4023484608060933e-05,
+      "loss": 1.2102,
+      "step": 4868
+    },
+    {
+      "epoch": 1.5297328071931995,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4020945731513806e-05,
+      "loss": 1.2949,
+      "step": 4870
+    },
+    {
+      "epoch": 1.5303610342187408,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4018406854966677e-05,
+      "loss": 1.2383,
+      "step": 4872
+    },
+    {
+      "epoch": 1.5309892612442821,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.401586797841955e-05,
+      "loss": 1.1815,
+      "step": 4874
+    },
+    {
+      "epoch": 1.5316174882698235,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4013329101872422e-05,
+      "loss": 1.2497,
+      "step": 4876
+    },
+    {
+      "epoch": 1.5322457152953648,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4010790225325295e-05,
+      "loss": 1.1109,
+      "step": 4878
+    },
+    {
+      "epoch": 1.5328739423209061,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4008251348778166e-05,
+      "loss": 1.2732,
+      "step": 4880
+    },
+    {
+      "epoch": 1.5335021693464475,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.400571247223104e-05,
+      "loss": 1.2428,
+      "step": 4882
+    },
+    {
+      "epoch": 1.5341303963719888,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.4003173595683912e-05,
+      "loss": 1.2501,
+      "step": 4884
+    },
+    {
+      "epoch": 1.5347586233975303,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.4000634719136784e-05,
+      "loss": 1.2886,
+      "step": 4886
+    },
+    {
+      "epoch": 1.5353868504230717,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3998095842589657e-05,
+      "loss": 1.2391,
+      "step": 4888
+    },
+    {
+      "epoch": 1.536015077448613,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3995556966042526e-05,
+      "loss": 1.2547,
+      "step": 4890
+    },
+    {
+      "epoch": 1.5366433044741543,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3993018089495401e-05,
+      "loss": 1.2538,
+      "step": 4892
+    },
+    {
+      "epoch": 1.5372715314996959,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3990479212948271e-05,
+      "loss": 1.2798,
+      "step": 4894
+    },
+    {
+      "epoch": 1.5378997585252372,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3987940336401144e-05,
+      "loss": 1.2126,
+      "step": 4896
+    },
+    {
+      "epoch": 1.5385279855507785,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3985401459854015e-05,
+      "loss": 1.2583,
+      "step": 4898
+    },
+    {
+      "epoch": 1.5391562125763198,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3982862583306888e-05,
+      "loss": 1.2099,
+      "step": 4900
+    },
+    {
+      "epoch": 1.5397844396018612,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.398032370675976e-05,
+      "loss": 1.3456,
+      "step": 4902
+    },
+    {
+      "epoch": 1.5404126666274025,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3977784830212633e-05,
+      "loss": 1.2055,
+      "step": 4904
+    },
+    {
+      "epoch": 1.5410408936529438,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3975245953665504e-05,
+      "loss": 1.2495,
+      "step": 4906
+    },
+    {
+      "epoch": 1.5416691206784852,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3972707077118377e-05,
+      "loss": 1.2034,
+      "step": 4908
+    },
+    {
+      "epoch": 1.5422973477040265,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.3970168200571249e-05,
+      "loss": 1.2081,
+      "step": 4910
+    },
+    {
+      "epoch": 1.5429255747295678,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3967629324024122e-05,
+      "loss": 1.0711,
+      "step": 4912
+    },
+    {
+      "epoch": 1.5435538017551091,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3965090447476991e-05,
+      "loss": 1.233,
+      "step": 4914
+    },
+    {
+      "epoch": 1.5441820287806505,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3962551570929865e-05,
+      "loss": 1.2307,
+      "step": 4916
+    },
+    {
+      "epoch": 1.5448102558061918,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3960012694382736e-05,
+      "loss": 1.3425,
+      "step": 4918
+    },
+    {
+      "epoch": 1.5454384828317334,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3957473817835609e-05,
+      "loss": 1.2415,
+      "step": 4920
+    },
+    {
+      "epoch": 1.5460667098572747,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.395493494128848e-05,
+      "loss": 1.1973,
+      "step": 4922
+    },
+    {
+      "epoch": 1.546694936882816,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3952396064741353e-05,
+      "loss": 1.2275,
+      "step": 4924
+    },
+    {
+      "epoch": 1.5473231639083573,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3949857188194225e-05,
+      "loss": 1.416,
+      "step": 4926
+    },
+    {
+      "epoch": 1.5479513909338989,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3947318311647098e-05,
+      "loss": 1.262,
+      "step": 4928
+    },
+    {
+      "epoch": 1.5485796179594402,
+      "grad_norm": 0.75,
+      "learning_rate": 1.394477943509997e-05,
+      "loss": 1.0942,
+      "step": 4930
+    },
+    {
+      "epoch": 1.5492078449849815,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3942240558552842e-05,
+      "loss": 1.2883,
+      "step": 4932
+    },
+    {
+      "epoch": 1.5498360720105229,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3939701682005714e-05,
+      "loss": 1.2058,
+      "step": 4934
+    },
+    {
+      "epoch": 1.5504642990360642,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3937162805458587e-05,
+      "loss": 1.2756,
+      "step": 4936
+    },
+    {
+      "epoch": 1.5510925260616055,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3934623928911457e-05,
+      "loss": 1.2265,
+      "step": 4938
+    },
+    {
+      "epoch": 1.5517207530871469,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.393208505236433e-05,
+      "loss": 1.0057,
+      "step": 4940
+    },
+    {
+      "epoch": 1.5523489801126882,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3929546175817201e-05,
+      "loss": 1.1519,
+      "step": 4942
+    },
+    {
+      "epoch": 1.5529772071382295,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3927007299270074e-05,
+      "loss": 1.3281,
+      "step": 4944
+    },
+    {
+      "epoch": 1.5536054341637708,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3924468422722945e-05,
+      "loss": 1.267,
+      "step": 4946
+    },
+    {
+      "epoch": 1.5542336611893122,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3921929546175819e-05,
+      "loss": 1.2482,
+      "step": 4948
+    },
+    {
+      "epoch": 1.5548618882148535,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.391939066962869e-05,
+      "loss": 1.3847,
+      "step": 4950
+    },
+    {
+      "epoch": 1.555490115240395,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3916851793081563e-05,
+      "loss": 1.4112,
+      "step": 4952
+    },
+    {
+      "epoch": 1.5561183422659364,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3914312916534434e-05,
+      "loss": 1.2111,
+      "step": 4954
+    },
+    {
+      "epoch": 1.5567465692914777,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3911774039987307e-05,
+      "loss": 1.4313,
+      "step": 4956
+    },
+    {
+      "epoch": 1.557374796317019,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3909235163440177e-05,
+      "loss": 1.2122,
+      "step": 4958
+    },
+    {
+      "epoch": 1.5580030233425606,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3906696286893052e-05,
+      "loss": 1.2285,
+      "step": 4960
+    },
+    {
+      "epoch": 1.558631250368102,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.3904157410345922e-05,
+      "loss": 1.2074,
+      "step": 4962
+    },
+    {
+      "epoch": 1.5592594773936432,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3901618533798795e-05,
+      "loss": 1.3096,
+      "step": 4964
+    },
+    {
+      "epoch": 1.5598877044191846,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3899079657251666e-05,
+      "loss": 1.1488,
+      "step": 4966
+    },
+    {
+      "epoch": 1.560515931444726,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3896540780704539e-05,
+      "loss": 1.358,
+      "step": 4968
+    },
+    {
+      "epoch": 1.5611441584702672,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3894001904157412e-05,
+      "loss": 1.2134,
+      "step": 4970
+    },
+    {
+      "epoch": 1.5617723854958085,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3891463027610284e-05,
+      "loss": 1.3555,
+      "step": 4972
+    },
+    {
+      "epoch": 1.5624006125213499,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.3888924151063157e-05,
+      "loss": 1.2697,
+      "step": 4974
+    },
+    {
+      "epoch": 1.5630288395468912,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.3886385274516028e-05,
+      "loss": 1.1357,
+      "step": 4976
+    },
+    {
+      "epoch": 1.5636570665724325,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3883846397968901e-05,
+      "loss": 1.1122,
+      "step": 4978
+    },
+    {
+      "epoch": 1.5642852935979739,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.3881307521421772e-05,
+      "loss": 1.3238,
+      "step": 4980
+    },
+    {
+      "epoch": 1.5649135206235152,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3878768644874646e-05,
+      "loss": 1.1624,
+      "step": 4982
+    },
+    {
+      "epoch": 1.5655417476490565,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3876229768327515e-05,
+      "loss": 1.1995,
+      "step": 4984
+    },
+    {
+      "epoch": 1.566169974674598,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.387369089178039e-05,
+      "loss": 1.433,
+      "step": 4986
+    },
+    {
+      "epoch": 1.5667982017001394,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.387115201523326e-05,
+      "loss": 1.2223,
+      "step": 4988
+    },
+    {
+      "epoch": 1.5674264287256807,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3868613138686133e-05,
+      "loss": 1.2335,
+      "step": 4990
+    },
+    {
+      "epoch": 1.568054655751222,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3866074262139004e-05,
+      "loss": 1.2387,
+      "step": 4992
+    },
+    {
+      "epoch": 1.5686828827767636,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3863535385591877e-05,
+      "loss": 1.2488,
+      "step": 4994
+    },
+    {
+      "epoch": 1.569311109802305,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3860996509044749e-05,
+      "loss": 1.3808,
+      "step": 4996
+    },
+    {
+      "epoch": 1.5699393368278463,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3858457632497622e-05,
+      "loss": 1.4113,
+      "step": 4998
+    },
+    {
+      "epoch": 1.5705675638533876,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3855918755950493e-05,
+      "loss": 1.2078,
+      "step": 5000
+    },
+    {
+      "epoch": 1.571195790878929,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3853379879403366e-05,
+      "loss": 1.2806,
+      "step": 5002
+    },
+    {
+      "epoch": 1.5718240179044702,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3850841002856238e-05,
+      "loss": 1.2592,
+      "step": 5004
+    },
+    {
+      "epoch": 1.5724522449300116,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.384830212630911e-05,
+      "loss": 1.2786,
+      "step": 5006
+    },
+    {
+      "epoch": 1.573080471955553,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.384576324976198e-05,
+      "loss": 1.2668,
+      "step": 5008
+    },
+    {
+      "epoch": 1.5737086989810942,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3843224373214853e-05,
+      "loss": 1.1614,
+      "step": 5010
+    },
+    {
+      "epoch": 1.5743369260066356,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3840685496667725e-05,
+      "loss": 1.1552,
+      "step": 5012
+    },
+    {
+      "epoch": 1.5749651530321769,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3838146620120598e-05,
+      "loss": 1.295,
+      "step": 5014
+    },
+    {
+      "epoch": 1.5755933800577182,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.383560774357347e-05,
+      "loss": 1.2958,
+      "step": 5016
+    },
+    {
+      "epoch": 1.5762216070832598,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3833068867026342e-05,
+      "loss": 1.3272,
+      "step": 5018
+    },
+    {
+      "epoch": 1.576849834108801,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.3830529990479214e-05,
+      "loss": 1.2085,
+      "step": 5020
+    },
+    {
+      "epoch": 1.5774780611343424,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3827991113932087e-05,
+      "loss": 1.28,
+      "step": 5022
+    },
+    {
+      "epoch": 1.5781062881598837,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3825452237384958e-05,
+      "loss": 1.2884,
+      "step": 5024
+    },
+    {
+      "epoch": 1.5787345151854253,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3822913360837831e-05,
+      "loss": 1.3882,
+      "step": 5026
+    },
+    {
+      "epoch": 1.5793627422109666,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3820374484290701e-05,
+      "loss": 1.3267,
+      "step": 5028
+    },
+    {
+      "epoch": 1.579990969236508,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3817835607743576e-05,
+      "loss": 1.2475,
+      "step": 5030
+    },
+    {
+      "epoch": 1.5806191962620493,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3815296731196445e-05,
+      "loss": 1.2973,
+      "step": 5032
+    },
+    {
+      "epoch": 1.5812474232875906,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3812757854649318e-05,
+      "loss": 1.1669,
+      "step": 5034
+    },
+    {
+      "epoch": 1.581875650313132,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.381021897810219e-05,
+      "loss": 1.2783,
+      "step": 5036
+    },
+    {
+      "epoch": 1.5825038773386733,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3807680101555063e-05,
+      "loss": 1.1044,
+      "step": 5038
+    },
+    {
+      "epoch": 1.5831321043642146,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3805141225007934e-05,
+      "loss": 1.2039,
+      "step": 5040
+    },
+    {
+      "epoch": 1.583760331389756,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3802602348460807e-05,
+      "loss": 1.3272,
+      "step": 5042
+    },
+    {
+      "epoch": 1.5843885584152972,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3800063471913679e-05,
+      "loss": 1.1171,
+      "step": 5044
+    },
+    {
+      "epoch": 1.5850167854408386,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3797524595366552e-05,
+      "loss": 1.085,
+      "step": 5046
+    },
+    {
+      "epoch": 1.58564501246638,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.3794985718819423e-05,
+      "loss": 1.1656,
+      "step": 5048
+    },
+    {
+      "epoch": 1.5862732394919212,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3792446842272296e-05,
+      "loss": 1.2341,
+      "step": 5050
+    },
+    {
+      "epoch": 1.5869014665174628,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3789907965725166e-05,
+      "loss": 1.2506,
+      "step": 5052
+    },
+    {
+      "epoch": 1.587529693543004,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3787369089178039e-05,
+      "loss": 1.2872,
+      "step": 5054
+    },
+    {
+      "epoch": 1.5881579205685454,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3784830212630914e-05,
+      "loss": 1.1942,
+      "step": 5056
+    },
+    {
+      "epoch": 1.5887861475940868,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3782291336083783e-05,
+      "loss": 1.2892,
+      "step": 5058
+    },
+    {
+      "epoch": 1.5894143746196283,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3779752459536657e-05,
+      "loss": 1.2254,
+      "step": 5060
+    },
+    {
+      "epoch": 1.5900426016451696,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3777213582989528e-05,
+      "loss": 1.2721,
+      "step": 5062
+    },
+    {
+      "epoch": 1.590670828670711,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3774674706442401e-05,
+      "loss": 1.2791,
+      "step": 5064
+    },
+    {
+      "epoch": 1.5912990556962523,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3772135829895272e-05,
+      "loss": 1.2474,
+      "step": 5066
+    },
+    {
+      "epoch": 1.5919272827217936,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3769596953348145e-05,
+      "loss": 1.2996,
+      "step": 5068
+    },
+    {
+      "epoch": 1.592555509747335,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3767058076801017e-05,
+      "loss": 1.1624,
+      "step": 5070
+    },
+    {
+      "epoch": 1.5931837367728763,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.376451920025389e-05,
+      "loss": 1.1621,
+      "step": 5072
+    },
+    {
+      "epoch": 1.5938119637984176,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3761980323706761e-05,
+      "loss": 1.1595,
+      "step": 5074
+    },
+    {
+      "epoch": 1.594440190823959,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3759441447159634e-05,
+      "loss": 1.2585,
+      "step": 5076
+    },
+    {
+      "epoch": 1.5950684178495003,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3756902570612504e-05,
+      "loss": 1.1875,
+      "step": 5078
+    },
+    {
+      "epoch": 1.5956966448750416,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3754363694065377e-05,
+      "loss": 1.286,
+      "step": 5080
+    },
+    {
+      "epoch": 1.596324871900583,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.3751824817518249e-05,
+      "loss": 1.313,
+      "step": 5082
+    },
+    {
+      "epoch": 1.5969530989261245,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.3749285940971122e-05,
+      "loss": 1.2761,
+      "step": 5084
+    },
+    {
+      "epoch": 1.5975813259516658,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3746747064423993e-05,
+      "loss": 1.2133,
+      "step": 5086
+    },
+    {
+      "epoch": 1.5982095529772071,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3744208187876866e-05,
+      "loss": 1.2465,
+      "step": 5088
+    },
+    {
+      "epoch": 1.5988377800027485,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3741669311329737e-05,
+      "loss": 1.3734,
+      "step": 5090
+    },
+    {
+      "epoch": 1.59946600702829,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.373913043478261e-05,
+      "loss": 1.4389,
+      "step": 5092
+    },
+    {
+      "epoch": 1.6000942340538313,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3736591558235482e-05,
+      "loss": 1.2194,
+      "step": 5094
+    },
+    {
+      "epoch": 1.6007224610793727,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3734052681688355e-05,
+      "loss": 1.1488,
+      "step": 5096
+    },
+    {
+      "epoch": 1.601350688104914,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.3731513805141226e-05,
+      "loss": 1.324,
+      "step": 5098
+    },
+    {
+      "epoch": 1.6019789151304553,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.37289749285941e-05,
+      "loss": 1.2289,
+      "step": 5100
+    },
+    {
+      "epoch": 1.6026071421559966,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3726436052046969e-05,
+      "loss": 1.1883,
+      "step": 5102
+    },
+    {
+      "epoch": 1.603235369181538,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3723897175499842e-05,
+      "loss": 1.2215,
+      "step": 5104
+    },
+    {
+      "epoch": 1.6038635962070793,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3721358298952714e-05,
+      "loss": 1.2423,
+      "step": 5106
+    },
+    {
+      "epoch": 1.6044918232326206,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3718819422405587e-05,
+      "loss": 1.1615,
+      "step": 5108
+    },
+    {
+      "epoch": 1.605120050258162,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3716280545858458e-05,
+      "loss": 1.1432,
+      "step": 5110
+    },
+    {
+      "epoch": 1.6057482772837033,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3713741669311331e-05,
+      "loss": 1.3432,
+      "step": 5112
+    },
+    {
+      "epoch": 1.6063765043092446,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.3711202792764203e-05,
+      "loss": 1.2572,
+      "step": 5114
+    },
+    {
+      "epoch": 1.6070047313347862,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3708663916217076e-05,
+      "loss": 1.141,
+      "step": 5116
+    },
+    {
+      "epoch": 1.6076329583603275,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3706125039669947e-05,
+      "loss": 1.2724,
+      "step": 5118
+    },
+    {
+      "epoch": 1.6082611853858688,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.370358616312282e-05,
+      "loss": 1.2636,
+      "step": 5120
+    },
+    {
+      "epoch": 1.6088894124114101,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.370104728657569e-05,
+      "loss": 1.2074,
+      "step": 5122
+    },
+    {
+      "epoch": 1.6095176394369515,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3698508410028565e-05,
+      "loss": 1.3499,
+      "step": 5124
+    },
+    {
+      "epoch": 1.610145866462493,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3695969533481434e-05,
+      "loss": 1.3866,
+      "step": 5126
+    },
+    {
+      "epoch": 1.6107740934880344,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3693430656934307e-05,
+      "loss": 1.2968,
+      "step": 5128
+    },
+    {
+      "epoch": 1.6114023205135757,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3690891780387179e-05,
+      "loss": 1.4032,
+      "step": 5130
+    },
+    {
+      "epoch": 1.612030547539117,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3688352903840052e-05,
+      "loss": 1.3089,
+      "step": 5132
+    },
+    {
+      "epoch": 1.6126587745646583,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3685814027292923e-05,
+      "loss": 1.332,
+      "step": 5134
+    },
+    {
+      "epoch": 1.6132870015901997,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3683275150745796e-05,
+      "loss": 1.1898,
+      "step": 5136
+    },
+    {
+      "epoch": 1.613915228615741,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3680736274198668e-05,
+      "loss": 1.2405,
+      "step": 5138
+    },
+    {
+      "epoch": 1.6145434556412823,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.367819739765154e-05,
+      "loss": 1.2174,
+      "step": 5140
+    },
+    {
+      "epoch": 1.6151716826668236,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3675658521104414e-05,
+      "loss": 1.0467,
+      "step": 5142
+    },
+    {
+      "epoch": 1.615799909692365,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3673119644557285e-05,
+      "loss": 1.2141,
+      "step": 5144
+    },
+    {
+      "epoch": 1.6164281367179063,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3670580768010158e-05,
+      "loss": 1.362,
+      "step": 5146
+    },
+    {
+      "epoch": 1.6170563637434476,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3668041891463028e-05,
+      "loss": 1.239,
+      "step": 5148
+    },
+    {
+      "epoch": 1.6176845907689892,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.3665503014915903e-05,
+      "loss": 1.3551,
+      "step": 5150
+    },
+    {
+      "epoch": 1.6183128177945305,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.3662964138368772e-05,
+      "loss": 1.3027,
+      "step": 5152
+    },
+    {
+      "epoch": 1.6189410448200718,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3660425261821645e-05,
+      "loss": 1.262,
+      "step": 5154
+    },
+    {
+      "epoch": 1.6195692718456132,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3657886385274517e-05,
+      "loss": 1.3053,
+      "step": 5156
+    },
+    {
+      "epoch": 1.6201974988711547,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.365534750872739e-05,
+      "loss": 1.3218,
+      "step": 5158
+    },
+    {
+      "epoch": 1.620825725896696,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3652808632180261e-05,
+      "loss": 1.2845,
+      "step": 5160
+    },
+    {
+      "epoch": 1.6214539529222374,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.3650269755633134e-05,
+      "loss": 1.1991,
+      "step": 5162
+    },
+    {
+      "epoch": 1.6220821799477787,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.3647730879086006e-05,
+      "loss": 1.3062,
+      "step": 5164
+    },
+    {
+      "epoch": 1.62271040697332,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3645192002538879e-05,
+      "loss": 1.0548,
+      "step": 5166
+    },
+    {
+      "epoch": 1.6233386339988614,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.364265312599175e-05,
+      "loss": 1.1821,
+      "step": 5168
+    },
+    {
+      "epoch": 1.6239668610244027,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3640114249444623e-05,
+      "loss": 1.1822,
+      "step": 5170
+    },
+    {
+      "epoch": 1.624595088049944,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3637575372897493e-05,
+      "loss": 1.2382,
+      "step": 5172
+    },
+    {
+      "epoch": 1.6252233150754853,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3635036496350366e-05,
+      "loss": 1.2857,
+      "step": 5174
+    },
+    {
+      "epoch": 1.6258515421010267,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3632497619803237e-05,
+      "loss": 1.0598,
+      "step": 5176
+    },
+    {
+      "epoch": 1.626479769126568,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.362995874325611e-05,
+      "loss": 1.2989,
+      "step": 5178
+    },
+    {
+      "epoch": 1.6271079961521093,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3627419866708982e-05,
+      "loss": 1.2732,
+      "step": 5180
+    },
+    {
+      "epoch": 1.6277362231776509,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3624880990161855e-05,
+      "loss": 1.1808,
+      "step": 5182
+    },
+    {
+      "epoch": 1.6283644502031922,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3622342113614726e-05,
+      "loss": 1.2011,
+      "step": 5184
+    },
+    {
+      "epoch": 1.6289926772287335,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.36198032370676e-05,
+      "loss": 1.2063,
+      "step": 5186
+    },
+    {
+      "epoch": 1.6296209042542749,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.361726436052047e-05,
+      "loss": 1.243,
+      "step": 5188
+    },
+    {
+      "epoch": 1.6302491312798162,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3614725483973344e-05,
+      "loss": 1.3291,
+      "step": 5190
+    },
+    {
+      "epoch": 1.6308773583053577,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3612186607426214e-05,
+      "loss": 1.2388,
+      "step": 5192
+    },
+    {
+      "epoch": 1.631505585330899,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3609647730879088e-05,
+      "loss": 1.2186,
+      "step": 5194
+    },
+    {
+      "epoch": 1.6321338123564404,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3607108854331958e-05,
+      "loss": 1.3375,
+      "step": 5196
+    },
+    {
+      "epoch": 1.6327620393819817,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3604569977784831e-05,
+      "loss": 1.1837,
+      "step": 5198
+    },
+    {
+      "epoch": 1.633390266407523,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3602031101237702e-05,
+      "loss": 1.1972,
+      "step": 5200
+    },
+    {
+      "epoch": 1.6340184934330644,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3599492224690576e-05,
+      "loss": 1.303,
+      "step": 5202
+    },
+    {
+      "epoch": 1.6346467204586057,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.3596953348143447e-05,
+      "loss": 1.2091,
+      "step": 5204
+    },
+    {
+      "epoch": 1.635274947484147,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.359441447159632e-05,
+      "loss": 1.2598,
+      "step": 5206
+    },
+    {
+      "epoch": 1.6359031745096884,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3591875595049191e-05,
+      "loss": 1.0553,
+      "step": 5208
+    },
+    {
+      "epoch": 1.6365314015352297,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3589336718502064e-05,
+      "loss": 1.2889,
+      "step": 5210
+    },
+    {
+      "epoch": 1.637159628560771,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3586797841954936e-05,
+      "loss": 1.1413,
+      "step": 5212
+    },
+    {
+      "epoch": 1.6377878555863123,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3584258965407809e-05,
+      "loss": 1.1636,
+      "step": 5214
+    },
+    {
+      "epoch": 1.638416082611854,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.3581720088860679e-05,
+      "loss": 1.276,
+      "step": 5216
+    },
+    {
+      "epoch": 1.6390443096373952,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3579181212313552e-05,
+      "loss": 1.305,
+      "step": 5218
+    },
+    {
+      "epoch": 1.6396725366629366,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3576642335766423e-05,
+      "loss": 1.201,
+      "step": 5220
+    },
+    {
+      "epoch": 1.6403007636884779,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3574103459219296e-05,
+      "loss": 1.4048,
+      "step": 5222
+    },
+    {
+      "epoch": 1.6409289907140194,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3571564582672168e-05,
+      "loss": 1.2899,
+      "step": 5224
+    },
+    {
+      "epoch": 1.6415572177395608,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.356902570612504e-05,
+      "loss": 1.2957,
+      "step": 5226
+    },
+    {
+      "epoch": 1.642185444765102,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3566486829577914e-05,
+      "loss": 1.1528,
+      "step": 5228
+    },
+    {
+      "epoch": 1.6428136717906434,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3563947953030785e-05,
+      "loss": 1.3055,
+      "step": 5230
+    },
+    {
+      "epoch": 1.6434418988161847,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.3561409076483658e-05,
+      "loss": 1.0612,
+      "step": 5232
+    },
+    {
+      "epoch": 1.644070125841726,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.355887019993653e-05,
+      "loss": 1.1704,
+      "step": 5234
+    },
+    {
+      "epoch": 1.6446983528672674,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.3556331323389403e-05,
+      "loss": 1.1888,
+      "step": 5236
+    },
+    {
+      "epoch": 1.6453265798928087,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3553792446842274e-05,
+      "loss": 1.2332,
+      "step": 5238
+    },
+    {
+      "epoch": 1.64595480691835,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3551253570295147e-05,
+      "loss": 1.2469,
+      "step": 5240
+    },
+    {
+      "epoch": 1.6465830339438914,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3548714693748017e-05,
+      "loss": 1.2556,
+      "step": 5242
+    },
+    {
+      "epoch": 1.6472112609694327,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.354617581720089e-05,
+      "loss": 1.2354,
+      "step": 5244
+    },
+    {
+      "epoch": 1.647839487994974,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3543636940653761e-05,
+      "loss": 1.4014,
+      "step": 5246
+    },
+    {
+      "epoch": 1.6484677150205156,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3541098064106634e-05,
+      "loss": 1.3373,
+      "step": 5248
+    },
+    {
+      "epoch": 1.649095942046057,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.3538559187559506e-05,
+      "loss": 1.3094,
+      "step": 5250
+    },
+    {
+      "epoch": 1.6497241690715982,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3536020311012379e-05,
+      "loss": 1.2154,
+      "step": 5252
+    },
+    {
+      "epoch": 1.6503523960971396,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.353348143446525e-05,
+      "loss": 1.1919,
+      "step": 5254
+    },
+    {
+      "epoch": 1.6509806231226811,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3530942557918123e-05,
+      "loss": 1.2142,
+      "step": 5256
+    },
+    {
+      "epoch": 1.6516088501482225,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3528403681370995e-05,
+      "loss": 1.2202,
+      "step": 5258
+    },
+    {
+      "epoch": 1.6522370771737638,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3525864804823868e-05,
+      "loss": 1.3794,
+      "step": 5260
+    },
+    {
+      "epoch": 1.652865304199305,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3523325928276739e-05,
+      "loss": 1.1983,
+      "step": 5262
+    },
+    {
+      "epoch": 1.6534935312248464,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3520787051729612e-05,
+      "loss": 1.3406,
+      "step": 5264
+    },
+    {
+      "epoch": 1.6541217582503878,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.3518248175182482e-05,
+      "loss": 1.1838,
+      "step": 5266
+    },
+    {
+      "epoch": 1.654749985275929,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3515709298635355e-05,
+      "loss": 1.3651,
+      "step": 5268
+    },
+    {
+      "epoch": 1.6553782123014704,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3513170422088226e-05,
+      "loss": 1.2919,
+      "step": 5270
+    },
+    {
+      "epoch": 1.6560064393270117,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.35106315455411e-05,
+      "loss": 1.3154,
+      "step": 5272
+    },
+    {
+      "epoch": 1.656634666352553,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.350809266899397e-05,
+      "loss": 1.2643,
+      "step": 5274
+    },
+    {
+      "epoch": 1.6572628933780944,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3505553792446844e-05,
+      "loss": 1.3239,
+      "step": 5276
+    },
+    {
+      "epoch": 1.6578911204036357,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.3503014915899715e-05,
+      "loss": 1.2137,
+      "step": 5278
+    },
+    {
+      "epoch": 1.658519347429177,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3500476039352588e-05,
+      "loss": 1.2215,
+      "step": 5280
+    },
+    {
+      "epoch": 1.6591475744547186,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.349793716280546e-05,
+      "loss": 1.1606,
+      "step": 5282
+    },
+    {
+      "epoch": 1.65977580148026,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3495398286258333e-05,
+      "loss": 1.2496,
+      "step": 5284
+    },
+    {
+      "epoch": 1.6604040285058013,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3492859409711202e-05,
+      "loss": 1.1227,
+      "step": 5286
+    },
+    {
+      "epoch": 1.6610322555313426,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3490320533164077e-05,
+      "loss": 1.2112,
+      "step": 5288
+    },
+    {
+      "epoch": 1.6616604825568841,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3487781656616947e-05,
+      "loss": 1.206,
+      "step": 5290
+    },
+    {
+      "epoch": 1.6622887095824255,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.348524278006982e-05,
+      "loss": 1.2691,
+      "step": 5292
+    },
+    {
+      "epoch": 1.6629169366079668,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3482703903522691e-05,
+      "loss": 1.2359,
+      "step": 5294
+    },
+    {
+      "epoch": 1.6635451636335081,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3480165026975564e-05,
+      "loss": 1.2491,
+      "step": 5296
+    },
+    {
+      "epoch": 1.6641733906590495,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.3477626150428436e-05,
+      "loss": 1.1253,
+      "step": 5298
+    },
+    {
+      "epoch": 1.6648016176845908,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3475087273881309e-05,
+      "loss": 1.2954,
+      "step": 5300
+    },
+    {
+      "epoch": 1.6654298447101321,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.347254839733418e-05,
+      "loss": 1.178,
+      "step": 5302
+    },
+    {
+      "epoch": 1.6660580717356734,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3470009520787053e-05,
+      "loss": 1.2116,
+      "step": 5304
+    },
+    {
+      "epoch": 1.6666862987612148,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3467470644239925e-05,
+      "loss": 1.2565,
+      "step": 5306
+    },
+    {
+      "epoch": 1.667314525786756,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.3464931767692798e-05,
+      "loss": 1.0026,
+      "step": 5308
+    },
+    {
+      "epoch": 1.6679427528122974,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3462392891145667e-05,
+      "loss": 1.292,
+      "step": 5310
+    },
+    {
+      "epoch": 1.6685709798378388,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.345985401459854e-05,
+      "loss": 1.104,
+      "step": 5312
+    },
+    {
+      "epoch": 1.6691992068633803,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3457315138051414e-05,
+      "loss": 1.2361,
+      "step": 5314
+    },
+    {
+      "epoch": 1.6698274338889216,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3454776261504285e-05,
+      "loss": 1.2665,
+      "step": 5316
+    },
+    {
+      "epoch": 1.670455660914463,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3452237384957158e-05,
+      "loss": 1.2824,
+      "step": 5318
+    },
+    {
+      "epoch": 1.6710838879400043,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.344969850841003e-05,
+      "loss": 1.1185,
+      "step": 5320
+    },
+    {
+      "epoch": 1.6717121149655458,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3447159631862903e-05,
+      "loss": 1.2738,
+      "step": 5322
+    },
+    {
+      "epoch": 1.6723403419910872,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3444620755315774e-05,
+      "loss": 1.2713,
+      "step": 5324
+    },
+    {
+      "epoch": 1.6729685690166285,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.3442081878768647e-05,
+      "loss": 1.1093,
+      "step": 5326
+    },
+    {
+      "epoch": 1.6735967960421698,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3439543002221518e-05,
+      "loss": 1.3263,
+      "step": 5328
+    },
+    {
+      "epoch": 1.6742250230677111,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3437004125674391e-05,
+      "loss": 1.3777,
+      "step": 5330
+    },
+    {
+      "epoch": 1.6748532500932525,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.3434465249127263e-05,
+      "loss": 1.1151,
+      "step": 5332
+    },
+    {
+      "epoch": 1.6754814771187938,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3431926372580136e-05,
+      "loss": 1.5733,
+      "step": 5334
+    },
+    {
+      "epoch": 1.6761097041443351,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3429387496033006e-05,
+      "loss": 1.1432,
+      "step": 5336
+    },
+    {
+      "epoch": 1.6767379311698765,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.3426848619485879e-05,
+      "loss": 1.3853,
+      "step": 5338
+    },
+    {
+      "epoch": 1.6773661581954178,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.342430974293875e-05,
+      "loss": 1.3414,
+      "step": 5340
+    },
+    {
+      "epoch": 1.6779943852209591,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3421770866391623e-05,
+      "loss": 1.2539,
+      "step": 5342
+    },
+    {
+      "epoch": 1.6786226122465004,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.3419231989844494e-05,
+      "loss": 1.2596,
+      "step": 5344
+    },
+    {
+      "epoch": 1.6792508392720418,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.3416693113297368e-05,
+      "loss": 1.1853,
+      "step": 5346
+    },
+    {
+      "epoch": 1.6798790662975833,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3414154236750239e-05,
+      "loss": 1.3223,
+      "step": 5348
+    },
+    {
+      "epoch": 1.6805072933231247,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3411615360203112e-05,
+      "loss": 1.3358,
+      "step": 5350
+    },
+    {
+      "epoch": 1.681135520348666,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3409076483655983e-05,
+      "loss": 1.1666,
+      "step": 5352
+    },
+    {
+      "epoch": 1.6817637473742073,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.3406537607108856e-05,
+      "loss": 1.2604,
+      "step": 5354
+    },
+    {
+      "epoch": 1.6823919743997489,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3403998730561726e-05,
+      "loss": 1.2772,
+      "step": 5356
+    },
+    {
+      "epoch": 1.6830202014252902,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.3401459854014601e-05,
+      "loss": 1.2507,
+      "step": 5358
+    },
+    {
+      "epoch": 1.6836484284508315,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.339892097746747e-05,
+      "loss": 1.1198,
+      "step": 5360
+    },
+    {
+      "epoch": 1.6842766554763728,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3396382100920344e-05,
+      "loss": 1.3049,
+      "step": 5362
+    },
+    {
+      "epoch": 1.6849048825019142,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3393843224373215e-05,
+      "loss": 1.1426,
+      "step": 5364
+    },
+    {
+      "epoch": 1.6855331095274555,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3391304347826088e-05,
+      "loss": 1.1442,
+      "step": 5366
+    },
+    {
+      "epoch": 1.6861613365529968,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.338876547127896e-05,
+      "loss": 1.2436,
+      "step": 5368
+    },
+    {
+      "epoch": 1.6867895635785382,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3386226594731833e-05,
+      "loss": 1.2264,
+      "step": 5370
+    },
+    {
+      "epoch": 1.6874177906040795,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3383687718184704e-05,
+      "loss": 1.2157,
+      "step": 5372
+    },
+    {
+      "epoch": 1.6880460176296208,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3381148841637577e-05,
+      "loss": 1.2976,
+      "step": 5374
+    },
+    {
+      "epoch": 1.6886742446551621,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3378609965090448e-05,
+      "loss": 1.2154,
+      "step": 5376
+    },
+    {
+      "epoch": 1.6893024716807035,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3376071088543322e-05,
+      "loss": 1.1889,
+      "step": 5378
+    },
+    {
+      "epoch": 1.689930698706245,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3373532211996191e-05,
+      "loss": 1.2453,
+      "step": 5380
+    },
+    {
+      "epoch": 1.6905589257317863,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3370993335449064e-05,
+      "loss": 1.1871,
+      "step": 5382
+    },
+    {
+      "epoch": 1.6911871527573277,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3368454458901936e-05,
+      "loss": 1.3467,
+      "step": 5384
+    },
+    {
+      "epoch": 1.691815379782869,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3365915582354809e-05,
+      "loss": 1.3004,
+      "step": 5386
+    },
+    {
+      "epoch": 1.6924436068084106,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.336337670580768e-05,
+      "loss": 1.3259,
+      "step": 5388
+    },
+    {
+      "epoch": 1.6930718338339519,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3360837829260553e-05,
+      "loss": 1.2515,
+      "step": 5390
+    },
+    {
+      "epoch": 1.6937000608594932,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3358298952713425e-05,
+      "loss": 1.2712,
+      "step": 5392
+    },
+    {
+      "epoch": 1.6943282878850345,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3355760076166298e-05,
+      "loss": 1.2064,
+      "step": 5394
+    },
+    {
+      "epoch": 1.6949565149105759,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3353221199619169e-05,
+      "loss": 1.3884,
+      "step": 5396
+    },
+    {
+      "epoch": 1.6955847419361172,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3350682323072042e-05,
+      "loss": 1.24,
+      "step": 5398
+    },
+    {
+      "epoch": 1.6962129689616585,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3348143446524915e-05,
+      "loss": 1.2544,
+      "step": 5400
+    },
+    {
+      "epoch": 1.6968411959871998,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3345604569977787e-05,
+      "loss": 1.2522,
+      "step": 5402
+    },
+    {
+      "epoch": 1.6974694230127412,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.334306569343066e-05,
+      "loss": 1.2362,
+      "step": 5404
+    },
+    {
+      "epoch": 1.6980976500382825,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.334052681688353e-05,
+      "loss": 1.3574,
+      "step": 5406
+    },
+    {
+      "epoch": 1.6987258770638238,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3337987940336402e-05,
+      "loss": 1.274,
+      "step": 5408
+    },
+    {
+      "epoch": 1.6993541040893652,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3335449063789274e-05,
+      "loss": 1.2817,
+      "step": 5410
+    },
+    {
+      "epoch": 1.6999823311149065,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3332910187242147e-05,
+      "loss": 1.3324,
+      "step": 5412
+    },
+    {
+      "epoch": 1.700610558140448,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3330371310695018e-05,
+      "loss": 1.2094,
+      "step": 5414
+    },
+    {
+      "epoch": 1.7012387851659894,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3327832434147891e-05,
+      "loss": 1.1927,
+      "step": 5416
+    },
+    {
+      "epoch": 1.7018670121915307,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3325293557600763e-05,
+      "loss": 1.166,
+      "step": 5418
+    },
+    {
+      "epoch": 1.702495239217072,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3322754681053636e-05,
+      "loss": 1.3582,
+      "step": 5420
+    },
+    {
+      "epoch": 1.7031234662426136,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3320215804506507e-05,
+      "loss": 1.3184,
+      "step": 5422
+    },
+    {
+      "epoch": 1.703751693268155,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.331767692795938e-05,
+      "loss": 1.2835,
+      "step": 5424
+    },
+    {
+      "epoch": 1.7043799202936962,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3315138051412252e-05,
+      "loss": 1.2446,
+      "step": 5426
+    },
+    {
+      "epoch": 1.7050081473192376,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3312599174865125e-05,
+      "loss": 1.2357,
+      "step": 5428
+    },
+    {
+      "epoch": 1.7056363743447789,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.3310060298317994e-05,
+      "loss": 1.2273,
+      "step": 5430
+    },
+    {
+      "epoch": 1.7062646013703202,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3307521421770867e-05,
+      "loss": 1.2684,
+      "step": 5432
+    },
+    {
+      "epoch": 1.7068928283958615,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3304982545223739e-05,
+      "loss": 1.2107,
+      "step": 5434
+    },
+    {
+      "epoch": 1.7075210554214029,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3302443668676612e-05,
+      "loss": 1.3441,
+      "step": 5436
+    },
+    {
+      "epoch": 1.7081492824469442,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3299904792129483e-05,
+      "loss": 1.2601,
+      "step": 5438
+    },
+    {
+      "epoch": 1.7087775094724855,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3297365915582356e-05,
+      "loss": 1.2648,
+      "step": 5440
+    },
+    {
+      "epoch": 1.7094057364980269,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3294827039035228e-05,
+      "loss": 1.1744,
+      "step": 5442
+    },
+    {
+      "epoch": 1.7100339635235682,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3292288162488101e-05,
+      "loss": 1.0949,
+      "step": 5444
+    },
+    {
+      "epoch": 1.7106621905491097,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3289749285940972e-05,
+      "loss": 1.1723,
+      "step": 5446
+    },
+    {
+      "epoch": 1.711290417574651,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3287210409393845e-05,
+      "loss": 1.2938,
+      "step": 5448
+    },
+    {
+      "epoch": 1.7119186446001924,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3284671532846715e-05,
+      "loss": 1.3799,
+      "step": 5450
+    },
+    {
+      "epoch": 1.7125468716257337,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3282132656299588e-05,
+      "loss": 1.0393,
+      "step": 5452
+    },
+    {
+      "epoch": 1.7131750986512753,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.327959377975246e-05,
+      "loss": 1.0883,
+      "step": 5454
+    },
+    {
+      "epoch": 1.7138033256768166,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3277054903205333e-05,
+      "loss": 1.1746,
+      "step": 5456
+    },
+    {
+      "epoch": 1.714431552702358,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.3274516026658204e-05,
+      "loss": 1.2404,
+      "step": 5458
+    },
+    {
+      "epoch": 1.7150597797278992,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3271977150111077e-05,
+      "loss": 1.1545,
+      "step": 5460
+    },
+    {
+      "epoch": 1.7156880067534406,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.3269438273563948e-05,
+      "loss": 1.3406,
+      "step": 5462
+    },
+    {
+      "epoch": 1.716316233778982,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3266899397016821e-05,
+      "loss": 1.2526,
+      "step": 5464
+    },
+    {
+      "epoch": 1.7169444608045232,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3264360520469693e-05,
+      "loss": 1.2369,
+      "step": 5466
+    },
+    {
+      "epoch": 1.7175726878300646,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3261821643922566e-05,
+      "loss": 1.2515,
+      "step": 5468
+    },
+    {
+      "epoch": 1.7182009148556059,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3259282767375437e-05,
+      "loss": 1.1968,
+      "step": 5470
+    },
+    {
+      "epoch": 1.7188291418811472,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.325674389082831e-05,
+      "loss": 1.2784,
+      "step": 5472
+    },
+    {
+      "epoch": 1.7194573689066885,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.325420501428118e-05,
+      "loss": 1.2446,
+      "step": 5474
+    },
+    {
+      "epoch": 1.7200855959322299,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.3251666137734053e-05,
+      "loss": 1.1592,
+      "step": 5476
+    },
+    {
+      "epoch": 1.7207138229577712,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3249127261186925e-05,
+      "loss": 1.2955,
+      "step": 5478
+    },
+    {
+      "epoch": 1.7213420499833127,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3246588384639798e-05,
+      "loss": 1.334,
+      "step": 5480
+    },
+    {
+      "epoch": 1.721970277008854,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3244049508092669e-05,
+      "loss": 1.114,
+      "step": 5482
+    },
+    {
+      "epoch": 1.7225985040343954,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3241510631545542e-05,
+      "loss": 1.0863,
+      "step": 5484
+    },
+    {
+      "epoch": 1.7232267310599367,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3238971754998415e-05,
+      "loss": 1.3088,
+      "step": 5486
+    },
+    {
+      "epoch": 1.7238549580854783,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3236432878451287e-05,
+      "loss": 1.0641,
+      "step": 5488
+    },
+    {
+      "epoch": 1.7244831851110196,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.323389400190416e-05,
+      "loss": 1.3067,
+      "step": 5490
+    },
+    {
+      "epoch": 1.725111412136561,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3231355125357031e-05,
+      "loss": 1.2372,
+      "step": 5492
+    },
+    {
+      "epoch": 1.7257396391621023,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3228816248809904e-05,
+      "loss": 1.0898,
+      "step": 5494
+    },
+    {
+      "epoch": 1.7263678661876436,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3226277372262775e-05,
+      "loss": 1.297,
+      "step": 5496
+    },
+    {
+      "epoch": 1.726996093213185,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.3223738495715649e-05,
+      "loss": 1.2627,
+      "step": 5498
+    },
+    {
+      "epoch": 1.7276243202387263,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.3221199619168518e-05,
+      "loss": 1.3876,
+      "step": 5500
+    },
+    {
+      "epoch": 1.7282525472642676,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3218660742621391e-05,
+      "loss": 1.1941,
+      "step": 5502
+    },
+    {
+      "epoch": 1.728880774289809,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3216121866074263e-05,
+      "loss": 1.2344,
+      "step": 5504
+    },
+    {
+      "epoch": 1.7295090013153502,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3213582989527136e-05,
+      "loss": 1.1713,
+      "step": 5506
+    },
+    {
+      "epoch": 1.7301372283408916,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3211044112980007e-05,
+      "loss": 1.242,
+      "step": 5508
+    },
+    {
+      "epoch": 1.730765455366433,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.320850523643288e-05,
+      "loss": 1.1983,
+      "step": 5510
+    },
+    {
+      "epoch": 1.7313936823919744,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3205966359885752e-05,
+      "loss": 1.2646,
+      "step": 5512
+    },
+    {
+      "epoch": 1.7320219094175158,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3203427483338625e-05,
+      "loss": 1.3675,
+      "step": 5514
+    },
+    {
+      "epoch": 1.732650136443057,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3200888606791496e-05,
+      "loss": 1.2915,
+      "step": 5516
+    },
+    {
+      "epoch": 1.7332783634685984,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3198349730244369e-05,
+      "loss": 1.3231,
+      "step": 5518
+    },
+    {
+      "epoch": 1.73390659049414,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3195810853697239e-05,
+      "loss": 1.228,
+      "step": 5520
+    },
+    {
+      "epoch": 1.7345348175196813,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.3193271977150114e-05,
+      "loss": 1.154,
+      "step": 5522
+    },
+    {
+      "epoch": 1.7351630445452226,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3190733100602983e-05,
+      "loss": 1.2426,
+      "step": 5524
+    },
+    {
+      "epoch": 1.735791271570764,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3188194224055856e-05,
+      "loss": 1.1921,
+      "step": 5526
+    },
+    {
+      "epoch": 1.7364194985963053,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3185655347508728e-05,
+      "loss": 1.2466,
+      "step": 5528
+    },
+    {
+      "epoch": 1.7370477256218466,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.31831164709616e-05,
+      "loss": 1.2553,
+      "step": 5530
+    },
+    {
+      "epoch": 1.737675952647388,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3180577594414472e-05,
+      "loss": 1.2312,
+      "step": 5532
+    },
+    {
+      "epoch": 1.7383041796729293,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3178038717867345e-05,
+      "loss": 1.2277,
+      "step": 5534
+    },
+    {
+      "epoch": 1.7389324066984706,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3175499841320217e-05,
+      "loss": 1.2192,
+      "step": 5536
+    },
+    {
+      "epoch": 1.739560633724012,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.317296096477309e-05,
+      "loss": 1.1536,
+      "step": 5538
+    },
+    {
+      "epoch": 1.7401888607495533,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.3170422088225961e-05,
+      "loss": 1.16,
+      "step": 5540
+    },
+    {
+      "epoch": 1.7408170877750946,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3167883211678834e-05,
+      "loss": 1.472,
+      "step": 5542
+    },
+    {
+      "epoch": 1.7414453148006361,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3165344335131704e-05,
+      "loss": 1.3138,
+      "step": 5544
+    },
+    {
+      "epoch": 1.7420735418261775,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3162805458584577e-05,
+      "loss": 1.2316,
+      "step": 5546
+    },
+    {
+      "epoch": 1.7427017688517188,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3160266582037448e-05,
+      "loss": 1.2371,
+      "step": 5548
+    },
+    {
+      "epoch": 1.7433299958772601,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3157727705490321e-05,
+      "loss": 1.302,
+      "step": 5550
+    },
+    {
+      "epoch": 1.7439582229028014,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3155188828943193e-05,
+      "loss": 1.1977,
+      "step": 5552
+    },
+    {
+      "epoch": 1.744586449928343,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3152649952396066e-05,
+      "loss": 1.234,
+      "step": 5554
+    },
+    {
+      "epoch": 1.7452146769538843,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3150111075848937e-05,
+      "loss": 1.2929,
+      "step": 5556
+    },
+    {
+      "epoch": 1.7458429039794257,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.314757219930181e-05,
+      "loss": 1.2919,
+      "step": 5558
+    },
+    {
+      "epoch": 1.746471131004967,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3145033322754682e-05,
+      "loss": 1.3428,
+      "step": 5560
+    },
+    {
+      "epoch": 1.7470993580305083,
+      "grad_norm": 0.875,
+      "learning_rate": 1.3142494446207555e-05,
+      "loss": 1.1509,
+      "step": 5562
+    },
+    {
+      "epoch": 1.7477275850560496,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.3139955569660424e-05,
+      "loss": 1.1809,
+      "step": 5564
+    },
+    {
+      "epoch": 1.748355812081591,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.31374166931133e-05,
+      "loss": 1.2189,
+      "step": 5566
+    },
+    {
+      "epoch": 1.7489840391071323,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3134877816566169e-05,
+      "loss": 1.1861,
+      "step": 5568
+    },
+    {
+      "epoch": 1.7496122661326736,
+      "grad_norm": 0.875,
+      "learning_rate": 1.3132338940019042e-05,
+      "loss": 1.1364,
+      "step": 5570
+    },
+    {
+      "epoch": 1.750240493158215,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3129800063471915e-05,
+      "loss": 1.1878,
+      "step": 5572
+    },
+    {
+      "epoch": 1.7508687201837563,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3127261186924786e-05,
+      "loss": 1.1927,
+      "step": 5574
+    },
+    {
+      "epoch": 1.7514969472092976,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.312472231037766e-05,
+      "loss": 1.1183,
+      "step": 5576
+    },
+    {
+      "epoch": 1.7521251742348392,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3122183433830531e-05,
+      "loss": 1.293,
+      "step": 5578
+    },
+    {
+      "epoch": 1.7527534012603805,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3119644557283404e-05,
+      "loss": 1.2526,
+      "step": 5580
+    },
+    {
+      "epoch": 1.7533816282859218,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3117105680736275e-05,
+      "loss": 1.1872,
+      "step": 5582
+    },
+    {
+      "epoch": 1.7540098553114631,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.3114566804189148e-05,
+      "loss": 1.2216,
+      "step": 5584
+    },
+    {
+      "epoch": 1.7546380823370047,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.311202792764202e-05,
+      "loss": 1.1968,
+      "step": 5586
+    },
+    {
+      "epoch": 1.755266309362546,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3109489051094893e-05,
+      "loss": 1.3127,
+      "step": 5588
+    },
+    {
+      "epoch": 1.7558945363880873,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3106950174547763e-05,
+      "loss": 1.143,
+      "step": 5590
+    },
+    {
+      "epoch": 1.7565227634136287,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.3104411298000637e-05,
+      "loss": 1.2036,
+      "step": 5592
+    },
+    {
+      "epoch": 1.75715099043917,
+      "grad_norm": 0.875,
+      "learning_rate": 1.3101872421453507e-05,
+      "loss": 1.3434,
+      "step": 5594
+    },
+    {
+      "epoch": 1.7577792174647113,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.309933354490638e-05,
+      "loss": 1.1842,
+      "step": 5596
+    },
+    {
+      "epoch": 1.7584074444902527,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3096794668359251e-05,
+      "loss": 1.2543,
+      "step": 5598
+    },
+    {
+      "epoch": 1.759035671515794,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.3094255791812125e-05,
+      "loss": 1.2244,
+      "step": 5600
+    },
+    {
+      "epoch": 1.7596638985413353,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3091716915264996e-05,
+      "loss": 1.1647,
+      "step": 5602
+    },
+    {
+      "epoch": 1.7602921255668766,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3089178038717869e-05,
+      "loss": 1.3046,
+      "step": 5604
+    },
+    {
+      "epoch": 1.760920352592418,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.308663916217074e-05,
+      "loss": 1.1823,
+      "step": 5606
+    },
+    {
+      "epoch": 1.7615485796179593,
+      "grad_norm": 0.875,
+      "learning_rate": 1.3084100285623613e-05,
+      "loss": 1.3261,
+      "step": 5608
+    },
+    {
+      "epoch": 1.7621768066435008,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3081561409076485e-05,
+      "loss": 1.2197,
+      "step": 5610
+    },
+    {
+      "epoch": 1.7628050336690422,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3079022532529358e-05,
+      "loss": 1.3214,
+      "step": 5612
+    },
+    {
+      "epoch": 1.7634332606945835,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3076483655982228e-05,
+      "loss": 1.3303,
+      "step": 5614
+    },
+    {
+      "epoch": 1.7640614877201248,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.30739447794351e-05,
+      "loss": 1.2344,
+      "step": 5616
+    },
+    {
+      "epoch": 1.7646897147456662,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3071405902887972e-05,
+      "loss": 1.2934,
+      "step": 5618
+    },
+    {
+      "epoch": 1.7653179417712077,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.3068867026340845e-05,
+      "loss": 1.1482,
+      "step": 5620
+    },
+    {
+      "epoch": 1.765946168796749,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3066328149793717e-05,
+      "loss": 1.2069,
+      "step": 5622
+    },
+    {
+      "epoch": 1.7665743958222904,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.306378927324659e-05,
+      "loss": 1.1891,
+      "step": 5624
+    },
+    {
+      "epoch": 1.7672026228478317,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3061250396699461e-05,
+      "loss": 1.3099,
+      "step": 5626
+    },
+    {
+      "epoch": 1.767830849873373,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3058711520152334e-05,
+      "loss": 1.2631,
+      "step": 5628
+    },
+    {
+      "epoch": 1.7684590768989144,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3056172643605205e-05,
+      "loss": 1.2516,
+      "step": 5630
+    },
+    {
+      "epoch": 1.7690873039244557,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3053633767058079e-05,
+      "loss": 1.2557,
+      "step": 5632
+    },
+    {
+      "epoch": 1.769715530949997,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.305109489051095e-05,
+      "loss": 1.2499,
+      "step": 5634
+    },
+    {
+      "epoch": 1.7703437579755383,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3048556013963823e-05,
+      "loss": 1.3002,
+      "step": 5636
+    },
+    {
+      "epoch": 1.7709719850010797,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3046017137416693e-05,
+      "loss": 1.2968,
+      "step": 5638
+    },
+    {
+      "epoch": 1.771600212026621,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 1.3062,
+      "step": 5640
+    },
+    {
+      "epoch": 1.7722284390521623,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3040939384322437e-05,
+      "loss": 1.1898,
+      "step": 5642
+    },
+    {
+      "epoch": 1.7728566660777039,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.303840050777531e-05,
+      "loss": 1.3012,
+      "step": 5644
+    },
+    {
+      "epoch": 1.7734848931032452,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3035861631228182e-05,
+      "loss": 1.1432,
+      "step": 5646
+    },
+    {
+      "epoch": 1.7741131201287865,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3033322754681055e-05,
+      "loss": 1.2746,
+      "step": 5648
+    },
+    {
+      "epoch": 1.7747413471543279,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3030783878133926e-05,
+      "loss": 1.2069,
+      "step": 5650
+    },
+    {
+      "epoch": 1.7753695741798694,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3028245001586799e-05,
+      "loss": 1.4756,
+      "step": 5652
+    },
+    {
+      "epoch": 1.7759978012054107,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.302570612503967e-05,
+      "loss": 1.2794,
+      "step": 5654
+    },
+    {
+      "epoch": 1.776626028230952,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.3023167248492544e-05,
+      "loss": 1.4529,
+      "step": 5656
+    },
+    {
+      "epoch": 1.7772542552564934,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3020628371945417e-05,
+      "loss": 1.2728,
+      "step": 5658
+    },
+    {
+      "epoch": 1.7778824822820347,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3018089495398288e-05,
+      "loss": 1.2532,
+      "step": 5660
+    },
+    {
+      "epoch": 1.778510709307576,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3015550618851161e-05,
+      "loss": 1.1706,
+      "step": 5662
+    },
+    {
+      "epoch": 1.7791389363331174,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.301301174230403e-05,
+      "loss": 1.3033,
+      "step": 5664
+    },
+    {
+      "epoch": 1.7797671633586587,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3010472865756904e-05,
+      "loss": 1.2695,
+      "step": 5666
+    },
+    {
+      "epoch": 1.7803953903842,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3007933989209775e-05,
+      "loss": 1.1627,
+      "step": 5668
+    },
+    {
+      "epoch": 1.7810236174097414,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.3005395112662648e-05,
+      "loss": 1.2485,
+      "step": 5670
+    },
+    {
+      "epoch": 1.7816518444352827,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.300285623611552e-05,
+      "loss": 1.2374,
+      "step": 5672
+    },
+    {
+      "epoch": 1.782280071460824,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3000317359568393e-05,
+      "loss": 1.302,
+      "step": 5674
+    },
+    {
+      "epoch": 1.7829082984863656,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.2997778483021264e-05,
+      "loss": 1.3329,
+      "step": 5676
+    },
+    {
+      "epoch": 1.783536525511907,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2995239606474137e-05,
+      "loss": 1.2451,
+      "step": 5678
+    },
+    {
+      "epoch": 1.7841647525374482,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2992700729927009e-05,
+      "loss": 1.2296,
+      "step": 5680
+    },
+    {
+      "epoch": 1.7847929795629895,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2990161853379882e-05,
+      "loss": 1.2764,
+      "step": 5682
+    },
+    {
+      "epoch": 1.785421206588531,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2987622976832751e-05,
+      "loss": 1.3022,
+      "step": 5684
+    },
+    {
+      "epoch": 1.7860494336140724,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2985084100285626e-05,
+      "loss": 1.1862,
+      "step": 5686
+    },
+    {
+      "epoch": 1.7866776606396138,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.2982545223738496e-05,
+      "loss": 1.097,
+      "step": 5688
+    },
+    {
+      "epoch": 1.787305887665155,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2980006347191369e-05,
+      "loss": 1.3138,
+      "step": 5690
+    },
+    {
+      "epoch": 1.7879341146906964,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.297746747064424e-05,
+      "loss": 1.2888,
+      "step": 5692
+    },
+    {
+      "epoch": 1.7885623417162377,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2974928594097113e-05,
+      "loss": 1.3605,
+      "step": 5694
+    },
+    {
+      "epoch": 1.789190568741779,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2972389717549985e-05,
+      "loss": 1.2925,
+      "step": 5696
+    },
+    {
+      "epoch": 1.7898187957673204,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.2969850841002858e-05,
+      "loss": 1.0879,
+      "step": 5698
+    },
+    {
+      "epoch": 1.7904470227928617,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.296731196445573e-05,
+      "loss": 1.1732,
+      "step": 5700
+    },
+    {
+      "epoch": 1.791075249818403,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2964773087908602e-05,
+      "loss": 1.2151,
+      "step": 5702
+    },
+    {
+      "epoch": 1.7917034768439444,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2962234211361474e-05,
+      "loss": 1.1779,
+      "step": 5704
+    },
+    {
+      "epoch": 1.7923317038694857,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2959695334814347e-05,
+      "loss": 1.2626,
+      "step": 5706
+    },
+    {
+      "epoch": 1.792959930895027,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2957156458267216e-05,
+      "loss": 1.1991,
+      "step": 5708
+    },
+    {
+      "epoch": 1.7935881579205686,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.295461758172009e-05,
+      "loss": 1.3407,
+      "step": 5710
+    },
+    {
+      "epoch": 1.79421638494611,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2952078705172961e-05,
+      "loss": 1.2508,
+      "step": 5712
+    },
+    {
+      "epoch": 1.7948446119716512,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2949539828625834e-05,
+      "loss": 1.4018,
+      "step": 5714
+    },
+    {
+      "epoch": 1.7954728389971926,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2947000952078705e-05,
+      "loss": 1.1721,
+      "step": 5716
+    },
+    {
+      "epoch": 1.7961010660227341,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2944462075531578e-05,
+      "loss": 1.2232,
+      "step": 5718
+    },
+    {
+      "epoch": 1.7967292930482754,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.294192319898445e-05,
+      "loss": 1.2397,
+      "step": 5720
+    },
+    {
+      "epoch": 1.7973575200738168,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2939384322437323e-05,
+      "loss": 1.172,
+      "step": 5722
+    },
+    {
+      "epoch": 1.797985747099358,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2936845445890194e-05,
+      "loss": 1.3208,
+      "step": 5724
+    },
+    {
+      "epoch": 1.7986139741248994,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2934306569343067e-05,
+      "loss": 1.2101,
+      "step": 5726
+    },
+    {
+      "epoch": 1.7992422011504408,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2931767692795937e-05,
+      "loss": 1.2136,
+      "step": 5728
+    },
+    {
+      "epoch": 1.799870428175982,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2929228816248812e-05,
+      "loss": 1.3315,
+      "step": 5730
+    },
+    {
+      "epoch": 1.8004986552015234,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2926689939701682e-05,
+      "loss": 1.1851,
+      "step": 5732
+    },
+    {
+      "epoch": 1.8011268822270647,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2924151063154555e-05,
+      "loss": 1.1484,
+      "step": 5734
+    },
+    {
+      "epoch": 1.801755109252606,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2921612186607426e-05,
+      "loss": 1.1852,
+      "step": 5736
+    },
+    {
+      "epoch": 1.8023833362781474,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2919073310060299e-05,
+      "loss": 1.2899,
+      "step": 5738
+    },
+    {
+      "epoch": 1.8030115633036887,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.291653443351317e-05,
+      "loss": 1.2624,
+      "step": 5740
+    },
+    {
+      "epoch": 1.8036397903292303,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2913995556966044e-05,
+      "loss": 1.1763,
+      "step": 5742
+    },
+    {
+      "epoch": 1.8042680173547716,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2911456680418917e-05,
+      "loss": 1.384,
+      "step": 5744
+    },
+    {
+      "epoch": 1.804896244380313,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2908917803871788e-05,
+      "loss": 1.241,
+      "step": 5746
+    },
+    {
+      "epoch": 1.8055244714058543,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2906378927324661e-05,
+      "loss": 1.3487,
+      "step": 5748
+    },
+    {
+      "epoch": 1.8061526984313958,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2903840050777532e-05,
+      "loss": 1.2097,
+      "step": 5750
+    },
+    {
+      "epoch": 1.8067809254569371,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2901301174230406e-05,
+      "loss": 1.2788,
+      "step": 5752
+    },
+    {
+      "epoch": 1.8074091524824785,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2898762297683275e-05,
+      "loss": 1.2052,
+      "step": 5754
+    },
+    {
+      "epoch": 1.8080373795080198,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.289622342113615e-05,
+      "loss": 1.3531,
+      "step": 5756
+    },
+    {
+      "epoch": 1.8086656065335611,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.289368454458902e-05,
+      "loss": 1.2512,
+      "step": 5758
+    },
+    {
+      "epoch": 1.8092938335591024,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2891145668041893e-05,
+      "loss": 1.2617,
+      "step": 5760
+    },
+    {
+      "epoch": 1.8099220605846438,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2888606791494764e-05,
+      "loss": 1.2172,
+      "step": 5762
+    },
+    {
+      "epoch": 1.810550287610185,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2886067914947637e-05,
+      "loss": 1.3569,
+      "step": 5764
+    },
+    {
+      "epoch": 1.8111785146357264,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.2883529038400509e-05,
+      "loss": 1.1964,
+      "step": 5766
+    },
+    {
+      "epoch": 1.8118067416612678,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2880990161853382e-05,
+      "loss": 1.1786,
+      "step": 5768
+    },
+    {
+      "epoch": 1.812434968686809,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2878451285306253e-05,
+      "loss": 1.2148,
+      "step": 5770
+    },
+    {
+      "epoch": 1.8130631957123504,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.2875912408759126e-05,
+      "loss": 1.1685,
+      "step": 5772
+    },
+    {
+      "epoch": 1.8136914227378917,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2873373532211998e-05,
+      "loss": 1.1826,
+      "step": 5774
+    },
+    {
+      "epoch": 1.8143196497634333,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.287083465566487e-05,
+      "loss": 1.2594,
+      "step": 5776
+    },
+    {
+      "epoch": 1.8149478767889746,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.286829577911774e-05,
+      "loss": 1.209,
+      "step": 5778
+    },
+    {
+      "epoch": 1.815576103814516,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2865756902570613e-05,
+      "loss": 1.2065,
+      "step": 5780
+    },
+    {
+      "epoch": 1.8162043308400573,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2863218026023485e-05,
+      "loss": 1.1886,
+      "step": 5782
+    },
+    {
+      "epoch": 1.8168325578655988,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2860679149476358e-05,
+      "loss": 1.2196,
+      "step": 5784
+    },
+    {
+      "epoch": 1.8174607848911402,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.285814027292923e-05,
+      "loss": 1.1996,
+      "step": 5786
+    },
+    {
+      "epoch": 1.8180890119166815,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2855601396382102e-05,
+      "loss": 1.1933,
+      "step": 5788
+    },
+    {
+      "epoch": 1.8187172389422228,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2853062519834974e-05,
+      "loss": 1.2438,
+      "step": 5790
+    },
+    {
+      "epoch": 1.8193454659677641,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2850523643287847e-05,
+      "loss": 1.2136,
+      "step": 5792
+    },
+    {
+      "epoch": 1.8199736929933055,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2847984766740718e-05,
+      "loss": 1.3761,
+      "step": 5794
+    },
+    {
+      "epoch": 1.8206019200188468,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2845445890193591e-05,
+      "loss": 1.1976,
+      "step": 5796
+    },
+    {
+      "epoch": 1.8212301470443881,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2842907013646463e-05,
+      "loss": 1.3462,
+      "step": 5798
+    },
+    {
+      "epoch": 1.8218583740699295,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2840368137099336e-05,
+      "loss": 1.2293,
+      "step": 5800
+    },
+    {
+      "epoch": 1.8224866010954708,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.2837829260552205e-05,
+      "loss": 1.2233,
+      "step": 5802
+    },
+    {
+      "epoch": 1.823114828121012,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2835290384005078e-05,
+      "loss": 1.2824,
+      "step": 5804
+    },
+    {
+      "epoch": 1.8237430551465534,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.283275150745795e-05,
+      "loss": 1.2794,
+      "step": 5806
+    },
+    {
+      "epoch": 1.824371282172095,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2830212630910823e-05,
+      "loss": 1.1825,
+      "step": 5808
+    },
+    {
+      "epoch": 1.8249995091976363,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.2827673754363694e-05,
+      "loss": 1.1168,
+      "step": 5810
+    },
+    {
+      "epoch": 1.8256277362231776,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2825134877816567e-05,
+      "loss": 1.2021,
+      "step": 5812
+    },
+    {
+      "epoch": 1.826255963248719,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2822596001269439e-05,
+      "loss": 1.3024,
+      "step": 5814
+    },
+    {
+      "epoch": 1.8268841902742605,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2820057124722312e-05,
+      "loss": 1.2621,
+      "step": 5816
+    },
+    {
+      "epoch": 1.8275124172998019,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2817518248175183e-05,
+      "loss": 1.0934,
+      "step": 5818
+    },
+    {
+      "epoch": 1.8281406443253432,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2814979371628056e-05,
+      "loss": 1.3914,
+      "step": 5820
+    },
+    {
+      "epoch": 1.8287688713508845,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.2812440495080926e-05,
+      "loss": 1.1312,
+      "step": 5822
+    },
+    {
+      "epoch": 1.8293970983764258,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.28099016185338e-05,
+      "loss": 1.1222,
+      "step": 5824
+    },
+    {
+      "epoch": 1.8300253254019672,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.280736274198667e-05,
+      "loss": 1.2303,
+      "step": 5826
+    },
+    {
+      "epoch": 1.8306535524275085,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2804823865439543e-05,
+      "loss": 1.2426,
+      "step": 5828
+    },
+    {
+      "epoch": 1.8312817794530498,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2802284988892417e-05,
+      "loss": 1.1024,
+      "step": 5830
+    },
+    {
+      "epoch": 1.8319100064785911,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2799746112345288e-05,
+      "loss": 1.3141,
+      "step": 5832
+    },
+    {
+      "epoch": 1.8325382335041325,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2797207235798161e-05,
+      "loss": 1.1605,
+      "step": 5834
+    },
+    {
+      "epoch": 1.8331664605296738,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2794668359251032e-05,
+      "loss": 1.1718,
+      "step": 5836
+    },
+    {
+      "epoch": 1.8337946875552151,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2792129482703905e-05,
+      "loss": 1.1882,
+      "step": 5838
+    },
+    {
+      "epoch": 1.8344229145807565,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2789590606156777e-05,
+      "loss": 1.3604,
+      "step": 5840
+    },
+    {
+      "epoch": 1.835051141606298,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.278705172960965e-05,
+      "loss": 1.1974,
+      "step": 5842
+    },
+    {
+      "epoch": 1.8356793686318393,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.2784512853062521e-05,
+      "loss": 1.1672,
+      "step": 5844
+    },
+    {
+      "epoch": 1.8363075956573807,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2781973976515394e-05,
+      "loss": 1.2177,
+      "step": 5846
+    },
+    {
+      "epoch": 1.836935822682922,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.2779435099968264e-05,
+      "loss": 1.2313,
+      "step": 5848
+    },
+    {
+      "epoch": 1.8375640497084635,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2776896223421139e-05,
+      "loss": 1.3003,
+      "step": 5850
+    },
+    {
+      "epoch": 1.8381922767340049,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2774357346874009e-05,
+      "loss": 1.192,
+      "step": 5852
+    },
+    {
+      "epoch": 1.8388205037595462,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.2771818470326882e-05,
+      "loss": 1.2617,
+      "step": 5854
+    },
+    {
+      "epoch": 1.8394487307850875,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2769279593779753e-05,
+      "loss": 1.0411,
+      "step": 5856
+    },
+    {
+      "epoch": 1.8400769578106289,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2766740717232626e-05,
+      "loss": 1.2904,
+      "step": 5858
+    },
+    {
+      "epoch": 1.8407051848361702,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2764201840685497e-05,
+      "loss": 1.3183,
+      "step": 5860
+    },
+    {
+      "epoch": 1.8413334118617115,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.276166296413837e-05,
+      "loss": 1.3205,
+      "step": 5862
+    },
+    {
+      "epoch": 1.8419616388872528,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2759124087591242e-05,
+      "loss": 1.2196,
+      "step": 5864
+    },
+    {
+      "epoch": 1.8425898659127942,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2756585211044115e-05,
+      "loss": 1.243,
+      "step": 5866
+    },
+    {
+      "epoch": 1.8432180929383355,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2754046334496986e-05,
+      "loss": 1.1317,
+      "step": 5868
+    },
+    {
+      "epoch": 1.8438463199638768,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.275150745794986e-05,
+      "loss": 1.2339,
+      "step": 5870
+    },
+    {
+      "epoch": 1.8444745469894182,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2748968581402729e-05,
+      "loss": 1.1604,
+      "step": 5872
+    },
+    {
+      "epoch": 1.8451027740149597,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2746429704855602e-05,
+      "loss": 1.1184,
+      "step": 5874
+    },
+    {
+      "epoch": 1.845731001040501,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2743890828308474e-05,
+      "loss": 1.2737,
+      "step": 5876
+    },
+    {
+      "epoch": 1.8463592280660424,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2741351951761347e-05,
+      "loss": 1.2273,
+      "step": 5878
+    },
+    {
+      "epoch": 1.8469874550915837,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2738813075214218e-05,
+      "loss": 1.2768,
+      "step": 5880
+    },
+    {
+      "epoch": 1.8476156821171252,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2736274198667091e-05,
+      "loss": 1.3388,
+      "step": 5882
+    },
+    {
+      "epoch": 1.8482439091426666,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2733735322119962e-05,
+      "loss": 1.2175,
+      "step": 5884
+    },
+    {
+      "epoch": 1.848872136168208,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2731196445572836e-05,
+      "loss": 1.2659,
+      "step": 5886
+    },
+    {
+      "epoch": 1.8495003631937492,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2728657569025707e-05,
+      "loss": 1.2642,
+      "step": 5888
+    },
+    {
+      "epoch": 1.8501285902192905,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.272611869247858e-05,
+      "loss": 1.0971,
+      "step": 5890
+    },
+    {
+      "epoch": 1.8507568172448319,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.272357981593145e-05,
+      "loss": 1.3001,
+      "step": 5892
+    },
+    {
+      "epoch": 1.8513850442703732,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2721040939384324e-05,
+      "loss": 1.457,
+      "step": 5894
+    },
+    {
+      "epoch": 1.8520132712959145,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2718502062837194e-05,
+      "loss": 1.3597,
+      "step": 5896
+    },
+    {
+      "epoch": 1.8526414983214559,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.2715963186290067e-05,
+      "loss": 1.2853,
+      "step": 5898
+    },
+    {
+      "epoch": 1.8532697253469972,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2713424309742939e-05,
+      "loss": 1.2468,
+      "step": 5900
+    },
+    {
+      "epoch": 1.8538979523725385,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2710885433195812e-05,
+      "loss": 1.3252,
+      "step": 5902
+    },
+    {
+      "epoch": 1.8545261793980798,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2708346556648683e-05,
+      "loss": 1.3121,
+      "step": 5904
+    },
+    {
+      "epoch": 1.8551544064236212,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2705807680101556e-05,
+      "loss": 1.1819,
+      "step": 5906
+    },
+    {
+      "epoch": 1.8557826334491627,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2703268803554428e-05,
+      "loss": 1.2418,
+      "step": 5908
+    },
+    {
+      "epoch": 1.856410860474704,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.27007299270073e-05,
+      "loss": 1.3177,
+      "step": 5910
+    },
+    {
+      "epoch": 1.8570390875002454,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2698191050460172e-05,
+      "loss": 1.2514,
+      "step": 5912
+    },
+    {
+      "epoch": 1.8576673145257867,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.2695652173913045e-05,
+      "loss": 1.2886,
+      "step": 5914
+    },
+    {
+      "epoch": 1.8582955415513283,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2693113297365918e-05,
+      "loss": 1.3111,
+      "step": 5916
+    },
+    {
+      "epoch": 1.8589237685768696,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2690574420818788e-05,
+      "loss": 1.1563,
+      "step": 5918
+    },
+    {
+      "epoch": 1.859551995602411,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.2688035544271663e-05,
+      "loss": 1.2238,
+      "step": 5920
+    },
+    {
+      "epoch": 1.8601802226279522,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2685496667724532e-05,
+      "loss": 1.2487,
+      "step": 5922
+    },
+    {
+      "epoch": 1.8608084496534936,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2682957791177405e-05,
+      "loss": 1.2178,
+      "step": 5924
+    },
+    {
+      "epoch": 1.861436676679035,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2680418914630277e-05,
+      "loss": 1.1426,
+      "step": 5926
+    },
+    {
+      "epoch": 1.8620649037045762,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.267788003808315e-05,
+      "loss": 1.2623,
+      "step": 5928
+    },
+    {
+      "epoch": 1.8626931307301176,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.2675341161536021e-05,
+      "loss": 1.0865,
+      "step": 5930
+    },
+    {
+      "epoch": 1.8633213577556589,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2672802284988894e-05,
+      "loss": 1.1582,
+      "step": 5932
+    },
+    {
+      "epoch": 1.8639495847812002,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2670263408441766e-05,
+      "loss": 0.9822,
+      "step": 5934
+    },
+    {
+      "epoch": 1.8645778118067415,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.2667724531894639e-05,
+      "loss": 1.1975,
+      "step": 5936
+    },
+    {
+      "epoch": 1.8652060388322829,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.266518565534751e-05,
+      "loss": 1.1873,
+      "step": 5938
+    },
+    {
+      "epoch": 1.8658342658578244,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2662646778800383e-05,
+      "loss": 1.2006,
+      "step": 5940
+    },
+    {
+      "epoch": 1.8664624928833657,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2660107902253253e-05,
+      "loss": 1.2259,
+      "step": 5942
+    },
+    {
+      "epoch": 1.867090719908907,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2657569025706126e-05,
+      "loss": 1.1884,
+      "step": 5944
+    },
+    {
+      "epoch": 1.8677189469344484,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2655030149158997e-05,
+      "loss": 1.1981,
+      "step": 5946
+    },
+    {
+      "epoch": 1.86834717395999,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.265249127261187e-05,
+      "loss": 1.1178,
+      "step": 5948
+    },
+    {
+      "epoch": 1.8689754009855313,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2649952396064742e-05,
+      "loss": 1.2778,
+      "step": 5950
+    },
+    {
+      "epoch": 1.8696036280110726,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2647413519517615e-05,
+      "loss": 1.171,
+      "step": 5952
+    },
+    {
+      "epoch": 1.870231855036614,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2644874642970486e-05,
+      "loss": 1.2154,
+      "step": 5954
+    },
+    {
+      "epoch": 1.8708600820621553,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.264233576642336e-05,
+      "loss": 1.2481,
+      "step": 5956
+    },
+    {
+      "epoch": 1.8714883090876966,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.263979688987623e-05,
+      "loss": 1.2464,
+      "step": 5958
+    },
+    {
+      "epoch": 1.872116536113238,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2637258013329104e-05,
+      "loss": 1.1656,
+      "step": 5960
+    },
+    {
+      "epoch": 1.8727447631387792,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2634719136781975e-05,
+      "loss": 1.3967,
+      "step": 5962
+    },
+    {
+      "epoch": 1.8733729901643206,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2632180260234848e-05,
+      "loss": 1.295,
+      "step": 5964
+    },
+    {
+      "epoch": 1.874001217189862,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2629641383687718e-05,
+      "loss": 1.1594,
+      "step": 5966
+    },
+    {
+      "epoch": 1.8746294442154032,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2627102507140591e-05,
+      "loss": 1.1973,
+      "step": 5968
+    },
+    {
+      "epoch": 1.8752576712409446,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.2624563630593462e-05,
+      "loss": 1.3587,
+      "step": 5970
+    },
+    {
+      "epoch": 1.8758858982664859,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2622024754046335e-05,
+      "loss": 1.3327,
+      "step": 5972
+    },
+    {
+      "epoch": 1.8765141252920274,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2619485877499207e-05,
+      "loss": 1.1387,
+      "step": 5974
+    },
+    {
+      "epoch": 1.8771423523175688,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.261694700095208e-05,
+      "loss": 1.1447,
+      "step": 5976
+    },
+    {
+      "epoch": 1.87777057934311,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2614408124404951e-05,
+      "loss": 1.1872,
+      "step": 5978
+    },
+    {
+      "epoch": 1.8783988063686514,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2611869247857824e-05,
+      "loss": 1.2919,
+      "step": 5980
+    },
+    {
+      "epoch": 1.879027033394193,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2609330371310696e-05,
+      "loss": 1.3053,
+      "step": 5982
+    },
+    {
+      "epoch": 1.8796552604197343,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2606791494763569e-05,
+      "loss": 1.1493,
+      "step": 5984
+    },
+    {
+      "epoch": 1.8802834874452756,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2604252618216439e-05,
+      "loss": 1.2792,
+      "step": 5986
+    },
+    {
+      "epoch": 1.880911714470817,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2601713741669313e-05,
+      "loss": 1.3208,
+      "step": 5988
+    },
+    {
+      "epoch": 1.8815399414963583,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2599174865122183e-05,
+      "loss": 1.1665,
+      "step": 5990
+    },
+    {
+      "epoch": 1.8821681685218996,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2596635988575056e-05,
+      "loss": 1.2419,
+      "step": 5992
+    },
+    {
+      "epoch": 1.882796395547441,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2594097112027927e-05,
+      "loss": 1.4174,
+      "step": 5994
+    },
+    {
+      "epoch": 1.8834246225729823,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.25915582354808e-05,
+      "loss": 1.2457,
+      "step": 5996
+    },
+    {
+      "epoch": 1.8840528495985236,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2589019358933672e-05,
+      "loss": 1.2441,
+      "step": 5998
+    },
+    {
+      "epoch": 1.884681076624065,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2586480482386545e-05,
+      "loss": 1.2268,
+      "step": 6000
+    },
+    {
+      "epoch": 1.8853093036496062,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.2583941605839418e-05,
+      "loss": 1.117,
+      "step": 6002
+    },
+    {
+      "epoch": 1.8859375306751476,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.258140272929229e-05,
+      "loss": 1.2595,
+      "step": 6004
+    },
+    {
+      "epoch": 1.8865657577006891,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2578863852745163e-05,
+      "loss": 1.2623,
+      "step": 6006
+    },
+    {
+      "epoch": 1.8871939847262305,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2576324976198034e-05,
+      "loss": 1.2805,
+      "step": 6008
+    },
+    {
+      "epoch": 1.8878222117517718,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2573786099650907e-05,
+      "loss": 1.1744,
+      "step": 6010
+    },
+    {
+      "epoch": 1.8884504387773131,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2571247223103777e-05,
+      "loss": 1.3327,
+      "step": 6012
+    },
+    {
+      "epoch": 1.8890786658028547,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2568708346556651e-05,
+      "loss": 1.1917,
+      "step": 6014
+    },
+    {
+      "epoch": 1.889706892828396,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2566169470009521e-05,
+      "loss": 1.4492,
+      "step": 6016
+    },
+    {
+      "epoch": 1.8903351198539373,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2563630593462394e-05,
+      "loss": 1.1507,
+      "step": 6018
+    },
+    {
+      "epoch": 1.8909633468794786,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2561091716915266e-05,
+      "loss": 1.2665,
+      "step": 6020
+    },
+    {
+      "epoch": 1.89159157390502,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2558552840368139e-05,
+      "loss": 1.2222,
+      "step": 6022
+    },
+    {
+      "epoch": 1.8922198009305613,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.255601396382101e-05,
+      "loss": 1.2067,
+      "step": 6024
+    },
+    {
+      "epoch": 1.8928480279561026,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2553475087273883e-05,
+      "loss": 1.1897,
+      "step": 6026
+    },
+    {
+      "epoch": 1.893476254981644,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.2550936210726755e-05,
+      "loss": 1.1622,
+      "step": 6028
+    },
+    {
+      "epoch": 1.8941044820071853,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.2548397334179628e-05,
+      "loss": 1.3068,
+      "step": 6030
+    },
+    {
+      "epoch": 1.8947327090327266,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.2545858457632499e-05,
+      "loss": 1.2051,
+      "step": 6032
+    },
+    {
+      "epoch": 1.895360936058268,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2543319581085372e-05,
+      "loss": 1.1956,
+      "step": 6034
+    },
+    {
+      "epoch": 1.8959891630838093,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2540780704538242e-05,
+      "loss": 1.3253,
+      "step": 6036
+    },
+    {
+      "epoch": 1.8966173901093508,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2538241827991115e-05,
+      "loss": 1.2558,
+      "step": 6038
+    },
+    {
+      "epoch": 1.8972456171348921,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2535702951443986e-05,
+      "loss": 1.2556,
+      "step": 6040
+    },
+    {
+      "epoch": 1.8978738441604335,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.253316407489686e-05,
+      "loss": 1.1596,
+      "step": 6042
+    },
+    {
+      "epoch": 1.8985020711859748,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.253062519834973e-05,
+      "loss": 1.2663,
+      "step": 6044
+    },
+    {
+      "epoch": 1.8991302982115161,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2528086321802604e-05,
+      "loss": 1.2155,
+      "step": 6046
+    },
+    {
+      "epoch": 1.8997585252370577,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2525547445255475e-05,
+      "loss": 1.272,
+      "step": 6048
+    },
+    {
+      "epoch": 1.900386752262599,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2523008568708348e-05,
+      "loss": 1.2012,
+      "step": 6050
+    },
+    {
+      "epoch": 1.9010149792881403,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.252046969216122e-05,
+      "loss": 1.2281,
+      "step": 6052
+    },
+    {
+      "epoch": 1.9016432063136817,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2517930815614093e-05,
+      "loss": 1.2045,
+      "step": 6054
+    },
+    {
+      "epoch": 1.902271433339223,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2515391939066962e-05,
+      "loss": 1.2902,
+      "step": 6056
+    },
+    {
+      "epoch": 1.9028996603647643,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2512853062519837e-05,
+      "loss": 1.2146,
+      "step": 6058
+    },
+    {
+      "epoch": 1.9035278873903057,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2510314185972707e-05,
+      "loss": 1.1062,
+      "step": 6060
+    },
+    {
+      "epoch": 1.904156114415847,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.250777530942558e-05,
+      "loss": 1.1466,
+      "step": 6062
+    },
+    {
+      "epoch": 1.9047843414413883,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2505236432878451e-05,
+      "loss": 1.2591,
+      "step": 6064
+    },
+    {
+      "epoch": 1.9054125684669296,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2502697556331324e-05,
+      "loss": 1.2363,
+      "step": 6066
+    },
+    {
+      "epoch": 1.906040795492471,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2500158679784196e-05,
+      "loss": 1.3046,
+      "step": 6068
+    },
+    {
+      "epoch": 1.9066690225180123,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2497619803237069e-05,
+      "loss": 1.3882,
+      "step": 6070
+    },
+    {
+      "epoch": 1.9072972495435538,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.249508092668994e-05,
+      "loss": 1.2277,
+      "step": 6072
+    },
+    {
+      "epoch": 1.9079254765690952,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2492542050142813e-05,
+      "loss": 1.2135,
+      "step": 6074
+    },
+    {
+      "epoch": 1.9085537035946365,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2490003173595685e-05,
+      "loss": 1.171,
+      "step": 6076
+    },
+    {
+      "epoch": 1.9091819306201778,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2487464297048558e-05,
+      "loss": 1.2015,
+      "step": 6078
+    },
+    {
+      "epoch": 1.9098101576457194,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2484925420501427e-05,
+      "loss": 1.3251,
+      "step": 6080
+    },
+    {
+      "epoch": 1.9104383846712607,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.24823865439543e-05,
+      "loss": 1.3621,
+      "step": 6082
+    },
+    {
+      "epoch": 1.911066611696802,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2479847667407172e-05,
+      "loss": 1.2972,
+      "step": 6084
+    },
+    {
+      "epoch": 1.9116948387223434,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2477308790860045e-05,
+      "loss": 1.1877,
+      "step": 6086
+    },
+    {
+      "epoch": 1.9123230657478847,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2474769914312918e-05,
+      "loss": 1.221,
+      "step": 6088
+    },
+    {
+      "epoch": 1.912951292773426,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.247223103776579e-05,
+      "loss": 1.3329,
+      "step": 6090
+    },
+    {
+      "epoch": 1.9135795197989673,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2469692161218662e-05,
+      "loss": 1.2478,
+      "step": 6092
+    },
+    {
+      "epoch": 1.9142077468245087,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.2467153284671534e-05,
+      "loss": 1.2821,
+      "step": 6094
+    },
+    {
+      "epoch": 1.91483597385005,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2464614408124407e-05,
+      "loss": 1.1848,
+      "step": 6096
+    },
+    {
+      "epoch": 1.9154642008755913,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2462075531577278e-05,
+      "loss": 1.2984,
+      "step": 6098
+    },
+    {
+      "epoch": 1.9160924279011327,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2459536655030151e-05,
+      "loss": 1.1486,
+      "step": 6100
+    },
+    {
+      "epoch": 1.916720654926674,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2456997778483023e-05,
+      "loss": 1.3557,
+      "step": 6102
+    },
+    {
+      "epoch": 1.9173488819522155,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2454458901935896e-05,
+      "loss": 1.2475,
+      "step": 6104
+    },
+    {
+      "epoch": 1.9179771089777569,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2451920025388766e-05,
+      "loss": 1.1533,
+      "step": 6106
+    },
+    {
+      "epoch": 1.9186053360032982,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2449381148841639e-05,
+      "loss": 1.2381,
+      "step": 6108
+    },
+    {
+      "epoch": 1.9192335630288395,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.244684227229451e-05,
+      "loss": 1.2304,
+      "step": 6110
+    },
+    {
+      "epoch": 1.919861790054381,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2444303395747383e-05,
+      "loss": 1.3272,
+      "step": 6112
+    },
+    {
+      "epoch": 1.9204900170799224,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2441764519200254e-05,
+      "loss": 1.1725,
+      "step": 6114
+    },
+    {
+      "epoch": 1.9211182441054637,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.2439225642653128e-05,
+      "loss": 1.0932,
+      "step": 6116
+    },
+    {
+      "epoch": 1.921746471131005,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2436686766105999e-05,
+      "loss": 1.2361,
+      "step": 6118
+    },
+    {
+      "epoch": 1.9223746981565464,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2434147889558872e-05,
+      "loss": 1.3047,
+      "step": 6120
+    },
+    {
+      "epoch": 1.9230029251820877,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2431609013011743e-05,
+      "loss": 1.3639,
+      "step": 6122
+    },
+    {
+      "epoch": 1.923631152207629,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.2429070136464616e-05,
+      "loss": 1.1849,
+      "step": 6124
+    },
+    {
+      "epoch": 1.9242593792331704,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2426531259917488e-05,
+      "loss": 1.2506,
+      "step": 6126
+    },
+    {
+      "epoch": 1.9248876062587117,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2423992383370361e-05,
+      "loss": 1.2748,
+      "step": 6128
+    },
+    {
+      "epoch": 1.925515833284253,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.242145350682323e-05,
+      "loss": 1.3513,
+      "step": 6130
+    },
+    {
+      "epoch": 1.9261440603097943,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2418914630276104e-05,
+      "loss": 1.3052,
+      "step": 6132
+    },
+    {
+      "epoch": 1.9267722873353357,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2416375753728975e-05,
+      "loss": 1.1691,
+      "step": 6134
+    },
+    {
+      "epoch": 1.927400514360877,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2413836877181848e-05,
+      "loss": 1.1503,
+      "step": 6136
+    },
+    {
+      "epoch": 1.9280287413864186,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.241129800063472e-05,
+      "loss": 1.218,
+      "step": 6138
+    },
+    {
+      "epoch": 1.9286569684119599,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.2408759124087593e-05,
+      "loss": 1.1987,
+      "step": 6140
+    },
+    {
+      "epoch": 1.9292851954375012,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2406220247540464e-05,
+      "loss": 1.1148,
+      "step": 6142
+    },
+    {
+      "epoch": 1.9299134224630425,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2403681370993337e-05,
+      "loss": 1.3232,
+      "step": 6144
+    },
+    {
+      "epoch": 1.930541649488584,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2401142494446208e-05,
+      "loss": 1.2638,
+      "step": 6146
+    },
+    {
+      "epoch": 1.9311698765141254,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2398603617899081e-05,
+      "loss": 1.282,
+      "step": 6148
+    },
+    {
+      "epoch": 1.9317981035396667,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2396064741351951e-05,
+      "loss": 1.292,
+      "step": 6150
+    },
+    {
+      "epoch": 1.932426330565208,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2393525864804826e-05,
+      "loss": 1.3647,
+      "step": 6152
+    },
+    {
+      "epoch": 1.9330545575907494,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2390986988257696e-05,
+      "loss": 1.1766,
+      "step": 6154
+    },
+    {
+      "epoch": 1.9336827846162907,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2388448111710569e-05,
+      "loss": 1.2779,
+      "step": 6156
+    },
+    {
+      "epoch": 1.934311011641832,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.238590923516344e-05,
+      "loss": 1.2979,
+      "step": 6158
+    },
+    {
+      "epoch": 1.9349392386673734,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2383370358616313e-05,
+      "loss": 1.3134,
+      "step": 6160
+    },
+    {
+      "epoch": 1.9355674656929147,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.2380831482069185e-05,
+      "loss": 1.1554,
+      "step": 6162
+    },
+    {
+      "epoch": 1.936195692718456,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2378292605522058e-05,
+      "loss": 1.3649,
+      "step": 6164
+    },
+    {
+      "epoch": 1.9368239197439974,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2375753728974929e-05,
+      "loss": 1.2077,
+      "step": 6166
+    },
+    {
+      "epoch": 1.9374521467695387,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2373214852427802e-05,
+      "loss": 1.288,
+      "step": 6168
+    },
+    {
+      "epoch": 1.9380803737950802,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2370675975880673e-05,
+      "loss": 1.3333,
+      "step": 6170
+    },
+    {
+      "epoch": 1.9387086008206216,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2368137099333547e-05,
+      "loss": 1.345,
+      "step": 6172
+    },
+    {
+      "epoch": 1.939336827846163,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.236559822278642e-05,
+      "loss": 1.3079,
+      "step": 6174
+    },
+    {
+      "epoch": 1.9399650548717042,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.236305934623929e-05,
+      "loss": 1.2672,
+      "step": 6176
+    },
+    {
+      "epoch": 1.9405932818972458,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2360520469692162e-05,
+      "loss": 1.2107,
+      "step": 6178
+    },
+    {
+      "epoch": 1.941221508922787,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2357981593145034e-05,
+      "loss": 1.2125,
+      "step": 6180
+    },
+    {
+      "epoch": 1.9418497359483284,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2355442716597907e-05,
+      "loss": 1.2236,
+      "step": 6182
+    },
+    {
+      "epoch": 1.9424779629738698,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2352903840050778e-05,
+      "loss": 1.1823,
+      "step": 6184
+    },
+    {
+      "epoch": 1.943106189999411,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.2350364963503651e-05,
+      "loss": 1.2182,
+      "step": 6186
+    },
+    {
+      "epoch": 1.9437344170249524,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2347826086956523e-05,
+      "loss": 1.2996,
+      "step": 6188
+    },
+    {
+      "epoch": 1.9443626440504937,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2345287210409396e-05,
+      "loss": 1.1519,
+      "step": 6190
+    },
+    {
+      "epoch": 1.944990871076035,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2342748333862267e-05,
+      "loss": 1.125,
+      "step": 6192
+    },
+    {
+      "epoch": 1.9456190981015764,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.234020945731514e-05,
+      "loss": 1.1524,
+      "step": 6194
+    },
+    {
+      "epoch": 1.9462473251271177,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2337670580768012e-05,
+      "loss": 1.2544,
+      "step": 6196
+    },
+    {
+      "epoch": 1.946875552152659,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2335131704220885e-05,
+      "loss": 1.1555,
+      "step": 6198
+    },
+    {
+      "epoch": 1.9475037791782004,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2332592827673754e-05,
+      "loss": 1.2127,
+      "step": 6200
+    },
+    {
+      "epoch": 1.9481320062037417,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.2330053951126627e-05,
+      "loss": 1.2985,
+      "step": 6202
+    },
+    {
+      "epoch": 1.9487602332292833,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2327515074579499e-05,
+      "loss": 1.3043,
+      "step": 6204
+    },
+    {
+      "epoch": 1.9493884602548246,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2324976198032372e-05,
+      "loss": 1.2911,
+      "step": 6206
+    },
+    {
+      "epoch": 1.950016687280366,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2322437321485243e-05,
+      "loss": 1.2217,
+      "step": 6208
+    },
+    {
+      "epoch": 1.9506449143059073,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2319898444938116e-05,
+      "loss": 1.2718,
+      "step": 6210
+    },
+    {
+      "epoch": 1.9512731413314488,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2317359568390988e-05,
+      "loss": 1.3247,
+      "step": 6212
+    },
+    {
+      "epoch": 1.9519013683569901,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.231482069184386e-05,
+      "loss": 1.31,
+      "step": 6214
+    },
+    {
+      "epoch": 1.9525295953825315,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2312281815296732e-05,
+      "loss": 1.2022,
+      "step": 6216
+    },
+    {
+      "epoch": 1.9531578224080728,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2309742938749605e-05,
+      "loss": 1.1422,
+      "step": 6218
+    },
+    {
+      "epoch": 1.9537860494336141,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2307204062202475e-05,
+      "loss": 1.1487,
+      "step": 6220
+    },
+    {
+      "epoch": 1.9544142764591554,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.230466518565535e-05,
+      "loss": 1.2259,
+      "step": 6222
+    },
+    {
+      "epoch": 1.9550425034846968,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.230212630910822e-05,
+      "loss": 1.1845,
+      "step": 6224
+    },
+    {
+      "epoch": 1.955670730510238,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2299587432561093e-05,
+      "loss": 1.2533,
+      "step": 6226
+    },
+    {
+      "epoch": 1.9562989575357794,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2297048556013964e-05,
+      "loss": 1.2527,
+      "step": 6228
+    },
+    {
+      "epoch": 1.9569271845613208,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2294509679466837e-05,
+      "loss": 1.2083,
+      "step": 6230
+    },
+    {
+      "epoch": 1.957555411586862,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2291970802919708e-05,
+      "loss": 1.2065,
+      "step": 6232
+    },
+    {
+      "epoch": 1.9581836386124034,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2289431926372581e-05,
+      "loss": 1.2186,
+      "step": 6234
+    },
+    {
+      "epoch": 1.958811865637945,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2286893049825453e-05,
+      "loss": 1.2687,
+      "step": 6236
+    },
+    {
+      "epoch": 1.9594400926634863,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2284354173278326e-05,
+      "loss": 1.2311,
+      "step": 6238
+    },
+    {
+      "epoch": 1.9600683196890276,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.2281815296731197e-05,
+      "loss": 1.1847,
+      "step": 6240
+    },
+    {
+      "epoch": 1.960696546714569,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.227927642018407e-05,
+      "loss": 1.2003,
+      "step": 6242
+    },
+    {
+      "epoch": 1.9613247737401105,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.227673754363694e-05,
+      "loss": 1.4039,
+      "step": 6244
+    },
+    {
+      "epoch": 1.9619530007656518,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2274198667089813e-05,
+      "loss": 1.2274,
+      "step": 6246
+    },
+    {
+      "epoch": 1.9625812277911932,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2271659790542684e-05,
+      "loss": 1.2743,
+      "step": 6248
+    },
+    {
+      "epoch": 1.9632094548167345,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2269120913995558e-05,
+      "loss": 1.414,
+      "step": 6250
+    },
+    {
+      "epoch": 1.9638376818422758,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2266582037448429e-05,
+      "loss": 1.2246,
+      "step": 6252
+    },
+    {
+      "epoch": 1.9644659088678171,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2264043160901302e-05,
+      "loss": 1.2005,
+      "step": 6254
+    },
+    {
+      "epoch": 1.9650941358933585,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2261504284354175e-05,
+      "loss": 1.2002,
+      "step": 6256
+    },
+    {
+      "epoch": 1.9657223629188998,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2258965407807046e-05,
+      "loss": 1.2792,
+      "step": 6258
+    },
+    {
+      "epoch": 1.9663505899444411,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.225642653125992e-05,
+      "loss": 1.0757,
+      "step": 6260
+    },
+    {
+      "epoch": 1.9669788169699824,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2253887654712791e-05,
+      "loss": 1.3089,
+      "step": 6262
+    },
+    {
+      "epoch": 1.9676070439955238,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2251348778165664e-05,
+      "loss": 1.2696,
+      "step": 6264
+    },
+    {
+      "epoch": 1.968235271021065,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2248809901618535e-05,
+      "loss": 1.2951,
+      "step": 6266
+    },
+    {
+      "epoch": 1.9688634980466064,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2246271025071408e-05,
+      "loss": 1.232,
+      "step": 6268
+    },
+    {
+      "epoch": 1.969491725072148,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2243732148524278e-05,
+      "loss": 1.3751,
+      "step": 6270
+    },
+    {
+      "epoch": 1.9701199520976893,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2241193271977151e-05,
+      "loss": 1.3549,
+      "step": 6272
+    },
+    {
+      "epoch": 1.9707481791232306,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.2238654395430023e-05,
+      "loss": 1.2652,
+      "step": 6274
+    },
+    {
+      "epoch": 1.971376406148772,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2236115518882896e-05,
+      "loss": 1.2687,
+      "step": 6276
+    },
+    {
+      "epoch": 1.9720046331743135,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2233576642335767e-05,
+      "loss": 1.4137,
+      "step": 6278
+    },
+    {
+      "epoch": 1.9726328601998548,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.223103776578864e-05,
+      "loss": 1.1847,
+      "step": 6280
+    },
+    {
+      "epoch": 1.9732610872253962,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2228498889241512e-05,
+      "loss": 1.3383,
+      "step": 6282
+    },
+    {
+      "epoch": 1.9738893142509375,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2225960012694385e-05,
+      "loss": 1.2893,
+      "step": 6284
+    },
+    {
+      "epoch": 1.9745175412764788,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2223421136147256e-05,
+      "loss": 1.2479,
+      "step": 6286
+    },
+    {
+      "epoch": 1.9751457683020202,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2220882259600129e-05,
+      "loss": 1.2674,
+      "step": 6288
+    },
+    {
+      "epoch": 1.9757739953275615,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2218343383053e-05,
+      "loss": 1.332,
+      "step": 6290
+    },
+    {
+      "epoch": 1.9764022223531028,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2215804506505874e-05,
+      "loss": 1.3086,
+      "step": 6292
+    },
+    {
+      "epoch": 1.9770304493786441,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2213265629958743e-05,
+      "loss": 1.3607,
+      "step": 6294
+    },
+    {
+      "epoch": 1.9776586764041855,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2210726753411616e-05,
+      "loss": 1.2653,
+      "step": 6296
+    },
+    {
+      "epoch": 1.9782869034297268,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2208187876864488e-05,
+      "loss": 1.3309,
+      "step": 6298
+    },
+    {
+      "epoch": 1.9789151304552681,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.220564900031736e-05,
+      "loss": 1.3084,
+      "step": 6300
+    },
+    {
+      "epoch": 1.9795433574808097,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2203110123770232e-05,
+      "loss": 1.2319,
+      "step": 6302
+    },
+    {
+      "epoch": 1.980171584506351,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2200571247223105e-05,
+      "loss": 1.193,
+      "step": 6304
+    },
+    {
+      "epoch": 1.9807998115318923,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2198032370675977e-05,
+      "loss": 1.1819,
+      "step": 6306
+    },
+    {
+      "epoch": 1.9814280385574337,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.219549349412885e-05,
+      "loss": 1.2694,
+      "step": 6308
+    },
+    {
+      "epoch": 1.9820562655829752,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2192954617581721e-05,
+      "loss": 1.2051,
+      "step": 6310
+    },
+    {
+      "epoch": 1.9826844926085165,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2190415741034594e-05,
+      "loss": 1.1432,
+      "step": 6312
+    },
+    {
+      "epoch": 1.9833127196340579,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.2187876864487464e-05,
+      "loss": 1.1966,
+      "step": 6314
+    },
+    {
+      "epoch": 1.9839409466595992,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2185337987940337e-05,
+      "loss": 1.1952,
+      "step": 6316
+    },
+    {
+      "epoch": 1.9845691736851405,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2182799111393208e-05,
+      "loss": 1.2597,
+      "step": 6318
+    },
+    {
+      "epoch": 1.9851974007106818,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.2180260234846081e-05,
+      "loss": 1.1049,
+      "step": 6320
+    },
+    {
+      "epoch": 1.9858256277362232,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2177721358298953e-05,
+      "loss": 1.2622,
+      "step": 6322
+    },
+    {
+      "epoch": 1.9864538547617645,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2175182481751826e-05,
+      "loss": 1.2554,
+      "step": 6324
+    },
+    {
+      "epoch": 1.9870820817873058,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2172643605204697e-05,
+      "loss": 1.4156,
+      "step": 6326
+    },
+    {
+      "epoch": 1.9877103088128472,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.217010472865757e-05,
+      "loss": 1.2517,
+      "step": 6328
+    },
+    {
+      "epoch": 1.9883385358383885,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2167565852110442e-05,
+      "loss": 1.3161,
+      "step": 6330
+    },
+    {
+      "epoch": 1.9889667628639298,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2165026975563315e-05,
+      "loss": 1.2247,
+      "step": 6332
+    },
+    {
+      "epoch": 1.9895949898894711,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2162488099016186e-05,
+      "loss": 1.1215,
+      "step": 6334
+    },
+    {
+      "epoch": 1.9902232169150127,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.215994922246906e-05,
+      "loss": 1.1817,
+      "step": 6336
+    },
+    {
+      "epoch": 1.990851443940554,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2157410345921929e-05,
+      "loss": 1.2248,
+      "step": 6338
+    },
+    {
+      "epoch": 1.9914796709660953,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2154871469374802e-05,
+      "loss": 1.3265,
+      "step": 6340
+    },
+    {
+      "epoch": 1.9921078979916367,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2152332592827675e-05,
+      "loss": 1.2027,
+      "step": 6342
+    },
+    {
+      "epoch": 1.9927361250171782,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.2149793716280546e-05,
+      "loss": 1.1223,
+      "step": 6344
+    },
+    {
+      "epoch": 1.9933643520427196,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.214725483973342e-05,
+      "loss": 1.2041,
+      "step": 6346
+    },
+    {
+      "epoch": 1.9939925790682609,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2144715963186291e-05,
+      "loss": 1.2209,
+      "step": 6348
+    },
+    {
+      "epoch": 1.9946208060938022,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2142177086639164e-05,
+      "loss": 1.3876,
+      "step": 6350
+    },
+    {
+      "epoch": 1.9952490331193435,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2139638210092035e-05,
+      "loss": 1.341,
+      "step": 6352
+    },
+    {
+      "epoch": 1.9958772601448849,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2137099333544908e-05,
+      "loss": 1.2775,
+      "step": 6354
+    },
+    {
+      "epoch": 1.9965054871704262,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.213456045699778e-05,
+      "loss": 1.2379,
+      "step": 6356
+    },
+    {
+      "epoch": 1.9971337141959675,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2132021580450653e-05,
+      "loss": 1.311,
+      "step": 6358
+    },
+    {
+      "epoch": 1.9977619412215089,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2129482703903524e-05,
+      "loss": 1.3089,
+      "step": 6360
+    },
+    {
+      "epoch": 1.9983901682470502,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2126943827356397e-05,
+      "loss": 1.1843,
+      "step": 6362
+    },
+    {
+      "epoch": 1.9990183952725915,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2124404950809267e-05,
+      "loss": 1.1485,
+      "step": 6364
+    },
+    {
+      "epoch": 1.9996466222981328,
+      "grad_norm": 0.875,
+      "learning_rate": 1.212186607426214e-05,
+      "loss": 1.0792,
+      "step": 6366
+    },
+    {
+      "epoch": 2.000274849323674,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2119327197715011e-05,
+      "loss": 1.3043,
+      "step": 6368
+    },
+    {
+      "epoch": 2.0009030763492155,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2116788321167885e-05,
+      "loss": 1.1913,
+      "step": 6370
+    },
+    {
+      "epoch": 2.0015313033747573,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2114249444620756e-05,
+      "loss": 1.1807,
+      "step": 6372
+    },
+    {
+      "epoch": 2.0021595304002986,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2111710568073629e-05,
+      "loss": 1.1107,
+      "step": 6374
+    },
+    {
+      "epoch": 2.00278775742584,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.21091716915265e-05,
+      "loss": 1.2526,
+      "step": 6376
+    },
+    {
+      "epoch": 2.0034159844513812,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2106632814979373e-05,
+      "loss": 1.1686,
+      "step": 6378
+    },
+    {
+      "epoch": 2.0040442114769226,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.2104093938432245e-05,
+      "loss": 1.0691,
+      "step": 6380
+    },
+    {
+      "epoch": 2.004672438502464,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.2101555061885118e-05,
+      "loss": 1.199,
+      "step": 6382
+    },
+    {
+      "epoch": 2.0053006655280052,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2099016185337988e-05,
+      "loss": 1.194,
+      "step": 6384
+    },
+    {
+      "epoch": 2.0059288925535466,
+      "grad_norm": 1.0,
+      "learning_rate": 1.2096477308790862e-05,
+      "loss": 1.2757,
+      "step": 6386
+    },
+    {
+      "epoch": 2.006557119579088,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2093938432243732e-05,
+      "loss": 1.2772,
+      "step": 6388
+    },
+    {
+      "epoch": 2.007185346604629,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.2091399555696605e-05,
+      "loss": 1.2131,
+      "step": 6390
+    },
+    {
+      "epoch": 2.0078135736301705,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2088860679149477e-05,
+      "loss": 1.0315,
+      "step": 6392
+    },
+    {
+      "epoch": 2.008441800655712,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.208632180260235e-05,
+      "loss": 1.207,
+      "step": 6394
+    },
+    {
+      "epoch": 2.009070027681253,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2083782926055221e-05,
+      "loss": 1.1659,
+      "step": 6396
+    },
+    {
+      "epoch": 2.0096982547067945,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.2081244049508094e-05,
+      "loss": 1.1712,
+      "step": 6398
+    },
+    {
+      "epoch": 2.010326481732336,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.2078705172960965e-05,
+      "loss": 1.235,
+      "step": 6400
+    },
+    {
+      "epoch": 2.010954708757877,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.2076166296413839e-05,
+      "loss": 1.1337,
+      "step": 6402
+    },
+    {
+      "epoch": 2.0115829357834185,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.207362741986671e-05,
+      "loss": 1.1306,
+      "step": 6404
+    },
+    {
+      "epoch": 2.0122111628089603,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.2071088543319583e-05,
+      "loss": 1.1945,
+      "step": 6406
+    },
+    {
+      "epoch": 2.0128393898345016,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.2068549666772453e-05,
+      "loss": 1.1251,
+      "step": 6408
+    },
+    {
+      "epoch": 2.013467616860043,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2066010790225326e-05,
+      "loss": 1.1349,
+      "step": 6410
+    },
+    {
+      "epoch": 2.0140958438855843,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.2063471913678197e-05,
+      "loss": 1.0783,
+      "step": 6412
+    },
+    {
+      "epoch": 2.0147240709111256,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.206093303713107e-05,
+      "loss": 1.0695,
+      "step": 6414
+    },
+    {
+      "epoch": 2.015352297936667,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.2058394160583942e-05,
+      "loss": 1.1497,
+      "step": 6416
+    },
+    {
+      "epoch": 2.0159805249622083,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2055855284036815e-05,
+      "loss": 1.3358,
+      "step": 6418
+    },
+    {
+      "epoch": 2.0166087519877496,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2053316407489686e-05,
+      "loss": 1.1467,
+      "step": 6420
+    },
+    {
+      "epoch": 2.017236979013291,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2050777530942559e-05,
+      "loss": 1.1837,
+      "step": 6422
+    },
+    {
+      "epoch": 2.0178652060388322,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.204823865439543e-05,
+      "loss": 1.1541,
+      "step": 6424
+    },
+    {
+      "epoch": 2.0184934330643736,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.2045699777848304e-05,
+      "loss": 1.2496,
+      "step": 6426
+    },
+    {
+      "epoch": 2.019121660089915,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2043160901301177e-05,
+      "loss": 1.1149,
+      "step": 6428
+    },
+    {
+      "epoch": 2.019749887115456,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2040622024754048e-05,
+      "loss": 1.236,
+      "step": 6430
+    },
+    {
+      "epoch": 2.0203781141409975,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2038083148206921e-05,
+      "loss": 1.1907,
+      "step": 6432
+    },
+    {
+      "epoch": 2.021006341166539,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.203554427165979e-05,
+      "loss": 1.1428,
+      "step": 6434
+    },
+    {
+      "epoch": 2.02163456819208,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2033005395112664e-05,
+      "loss": 1.2013,
+      "step": 6436
+    },
+    {
+      "epoch": 2.022262795217622,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.2030466518565535e-05,
+      "loss": 1.2182,
+      "step": 6438
+    },
+    {
+      "epoch": 2.0228910222431633,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.2027927642018408e-05,
+      "loss": 1.1166,
+      "step": 6440
+    },
+    {
+      "epoch": 2.0235192492687046,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.202538876547128e-05,
+      "loss": 1.1307,
+      "step": 6442
+    },
+    {
+      "epoch": 2.024147476294246,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.2022849888924153e-05,
+      "loss": 0.9865,
+      "step": 6444
+    },
+    {
+      "epoch": 2.0247757033197873,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2020311012377024e-05,
+      "loss": 1.0833,
+      "step": 6446
+    },
+    {
+      "epoch": 2.0254039303453286,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.2017772135829897e-05,
+      "loss": 1.1807,
+      "step": 6448
+    },
+    {
+      "epoch": 2.02603215737087,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.2015233259282769e-05,
+      "loss": 1.0344,
+      "step": 6450
+    },
+    {
+      "epoch": 2.0266603843964113,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2012694382735642e-05,
+      "loss": 1.0775,
+      "step": 6452
+    },
+    {
+      "epoch": 2.0272886114219526,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2010155506188511e-05,
+      "loss": 1.1049,
+      "step": 6454
+    },
+    {
+      "epoch": 2.027916838447494,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2007616629641386e-05,
+      "loss": 1.3446,
+      "step": 6456
+    },
+    {
+      "epoch": 2.0285450654730353,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2005077753094256e-05,
+      "loss": 1.3205,
+      "step": 6458
+    },
+    {
+      "epoch": 2.0291732924985766,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2002538876547129e-05,
+      "loss": 1.28,
+      "step": 6460
+    },
+    {
+      "epoch": 2.029801519524118,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2e-05,
+      "loss": 1.157,
+      "step": 6462
+    },
+    {
+      "epoch": 2.0304297465496592,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1997461123452873e-05,
+      "loss": 1.132,
+      "step": 6464
+    },
+    {
+      "epoch": 2.0310579735752006,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1994922246905745e-05,
+      "loss": 1.1373,
+      "step": 6466
+    },
+    {
+      "epoch": 2.031686200600742,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1992383370358618e-05,
+      "loss": 1.2225,
+      "step": 6468
+    },
+    {
+      "epoch": 2.0323144276262832,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.198984449381149e-05,
+      "loss": 1.21,
+      "step": 6470
+    },
+    {
+      "epoch": 2.032942654651825,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.1987305617264362e-05,
+      "loss": 1.1602,
+      "step": 6472
+    },
+    {
+      "epoch": 2.0335708816773663,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1984766740717234e-05,
+      "loss": 1.1242,
+      "step": 6474
+    },
+    {
+      "epoch": 2.0341991087029077,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1982227864170107e-05,
+      "loss": 1.1352,
+      "step": 6476
+    },
+    {
+      "epoch": 2.034827335728449,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1979688987622976e-05,
+      "loss": 1.1541,
+      "step": 6478
+    },
+    {
+      "epoch": 2.0354555627539903,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.197715011107585e-05,
+      "loss": 1.1356,
+      "step": 6480
+    },
+    {
+      "epoch": 2.0360837897795316,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1974611234528721e-05,
+      "loss": 1.3641,
+      "step": 6482
+    },
+    {
+      "epoch": 2.036712016805073,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.1972072357981594e-05,
+      "loss": 1.0959,
+      "step": 6484
+    },
+    {
+      "epoch": 2.0373402438306143,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1969533481434465e-05,
+      "loss": 1.2647,
+      "step": 6486
+    },
+    {
+      "epoch": 2.0379684708561556,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1966994604887338e-05,
+      "loss": 1.0983,
+      "step": 6488
+    },
+    {
+      "epoch": 2.038596697881697,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.196445572834021e-05,
+      "loss": 1.1839,
+      "step": 6490
+    },
+    {
+      "epoch": 2.0392249249072383,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1961916851793083e-05,
+      "loss": 1.2528,
+      "step": 6492
+    },
+    {
+      "epoch": 2.0398531519327796,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1959377975245954e-05,
+      "loss": 1.1767,
+      "step": 6494
+    },
+    {
+      "epoch": 2.040481378958321,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1956839098698827e-05,
+      "loss": 1.2149,
+      "step": 6496
+    },
+    {
+      "epoch": 2.0411096059838623,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1954300222151699e-05,
+      "loss": 1.1772,
+      "step": 6498
+    },
+    {
+      "epoch": 2.0417378330094036,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.1951761345604572e-05,
+      "loss": 1.1659,
+      "step": 6500
+    },
+    {
+      "epoch": 2.042366060034945,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.1949222469057442e-05,
+      "loss": 1.1775,
+      "step": 6502
+    },
+    {
+      "epoch": 2.0429942870604867,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.1946683592510315e-05,
+      "loss": 1.2286,
+      "step": 6504
+    },
+    {
+      "epoch": 2.043622514086028,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1944144715963186e-05,
+      "loss": 1.265,
+      "step": 6506
+    },
+    {
+      "epoch": 2.0442507411115693,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1941605839416059e-05,
+      "loss": 1.2133,
+      "step": 6508
+    },
+    {
+      "epoch": 2.0448789681371107,
+      "grad_norm": 1.0,
+      "learning_rate": 1.193906696286893e-05,
+      "loss": 1.0619,
+      "step": 6510
+    },
+    {
+      "epoch": 2.045507195162652,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1936528086321803e-05,
+      "loss": 1.292,
+      "step": 6512
+    },
+    {
+      "epoch": 2.0461354221881933,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.1933989209774677e-05,
+      "loss": 1.1345,
+      "step": 6514
+    },
+    {
+      "epoch": 2.0467636492137347,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1931450333227548e-05,
+      "loss": 1.0742,
+      "step": 6516
+    },
+    {
+      "epoch": 2.047391876239276,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1928911456680421e-05,
+      "loss": 1.0563,
+      "step": 6518
+    },
+    {
+      "epoch": 2.0480201032648173,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1926372580133292e-05,
+      "loss": 1.0728,
+      "step": 6520
+    },
+    {
+      "epoch": 2.0486483302903586,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1923833703586165e-05,
+      "loss": 1.2013,
+      "step": 6522
+    },
+    {
+      "epoch": 2.0492765573159,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1921294827039037e-05,
+      "loss": 1.2654,
+      "step": 6524
+    },
+    {
+      "epoch": 2.0499047843414413,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.191875595049191e-05,
+      "loss": 1.1785,
+      "step": 6526
+    },
+    {
+      "epoch": 2.0505330113669826,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.191621707394478e-05,
+      "loss": 1.166,
+      "step": 6528
+    },
+    {
+      "epoch": 2.051161238392524,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.1913678197397653e-05,
+      "loss": 1.0234,
+      "step": 6530
+    },
+    {
+      "epoch": 2.0517894654180653,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1911139320850524e-05,
+      "loss": 1.1007,
+      "step": 6532
+    },
+    {
+      "epoch": 2.0524176924436066,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1908600444303397e-05,
+      "loss": 1.2466,
+      "step": 6534
+    },
+    {
+      "epoch": 2.053045919469148,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.1906061567756269e-05,
+      "loss": 1.1554,
+      "step": 6536
+    },
+    {
+      "epoch": 2.0536741464946897,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1903522691209142e-05,
+      "loss": 1.293,
+      "step": 6538
+    },
+    {
+      "epoch": 2.054302373520231,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.1900983814662013e-05,
+      "loss": 1.1603,
+      "step": 6540
+    },
+    {
+      "epoch": 2.0549306005457724,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.1898444938114886e-05,
+      "loss": 1.1157,
+      "step": 6542
+    },
+    {
+      "epoch": 2.0555588275713137,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1895906061567757e-05,
+      "loss": 1.102,
+      "step": 6544
+    },
+    {
+      "epoch": 2.056187054596855,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.189336718502063e-05,
+      "loss": 1.1276,
+      "step": 6546
+    },
+    {
+      "epoch": 2.0568152816223964,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.18908283084735e-05,
+      "loss": 1.3061,
+      "step": 6548
+    },
+    {
+      "epoch": 2.0574435086479377,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1888289431926375e-05,
+      "loss": 1.2868,
+      "step": 6550
+    },
+    {
+      "epoch": 2.058071735673479,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1885750555379245e-05,
+      "loss": 1.2344,
+      "step": 6552
+    },
+    {
+      "epoch": 2.0586999626990203,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.1883211678832118e-05,
+      "loss": 1.1805,
+      "step": 6554
+    },
+    {
+      "epoch": 2.0593281897245617,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1880672802284989e-05,
+      "loss": 1.1347,
+      "step": 6556
+    },
+    {
+      "epoch": 2.059956416750103,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1878133925737862e-05,
+      "loss": 1.1707,
+      "step": 6558
+    },
+    {
+      "epoch": 2.0605846437756443,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1875595049190734e-05,
+      "loss": 1.2383,
+      "step": 6560
+    },
+    {
+      "epoch": 2.0612128708011856,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1873056172643607e-05,
+      "loss": 1.1534,
+      "step": 6562
+    },
+    {
+      "epoch": 2.061841097826727,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1870517296096478e-05,
+      "loss": 1.1963,
+      "step": 6564
+    },
+    {
+      "epoch": 2.0624693248522683,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1867978419549351e-05,
+      "loss": 1.2344,
+      "step": 6566
+    },
+    {
+      "epoch": 2.0630975518778096,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1865439543002223e-05,
+      "loss": 1.158,
+      "step": 6568
+    },
+    {
+      "epoch": 2.0637257789033514,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1862900666455096e-05,
+      "loss": 1.2401,
+      "step": 6570
+    },
+    {
+      "epoch": 2.0643540059288927,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1860361789907965e-05,
+      "loss": 1.1387,
+      "step": 6572
+    },
+    {
+      "epoch": 2.064982232954434,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1857822913360838e-05,
+      "loss": 1.1706,
+      "step": 6574
+    },
+    {
+      "epoch": 2.0656104599799754,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.185528403681371e-05,
+      "loss": 1.2223,
+      "step": 6576
+    },
+    {
+      "epoch": 2.0662386870055167,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1852745160266583e-05,
+      "loss": 1.1918,
+      "step": 6578
+    },
+    {
+      "epoch": 2.066866914031058,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1850206283719454e-05,
+      "loss": 1.1429,
+      "step": 6580
+    },
+    {
+      "epoch": 2.0674951410565994,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.1847667407172327e-05,
+      "loss": 1.1263,
+      "step": 6582
+    },
+    {
+      "epoch": 2.0681233680821407,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1845128530625199e-05,
+      "loss": 1.2082,
+      "step": 6584
+    },
+    {
+      "epoch": 2.068751595107682,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1842589654078072e-05,
+      "loss": 1.1829,
+      "step": 6586
+    },
+    {
+      "epoch": 2.0693798221332234,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.1840050777530943e-05,
+      "loss": 1.0186,
+      "step": 6588
+    },
+    {
+      "epoch": 2.0700080491587647,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1837511900983816e-05,
+      "loss": 1.2465,
+      "step": 6590
+    },
+    {
+      "epoch": 2.070636276184306,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1834973024436686e-05,
+      "loss": 1.1994,
+      "step": 6592
+    },
+    {
+      "epoch": 2.0712645032098473,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.183243414788956e-05,
+      "loss": 1.1866,
+      "step": 6594
+    },
+    {
+      "epoch": 2.0718927302353887,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.182989527134243e-05,
+      "loss": 1.2716,
+      "step": 6596
+    },
+    {
+      "epoch": 2.07252095726093,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.1827356394795303e-05,
+      "loss": 1.3113,
+      "step": 6598
+    },
+    {
+      "epoch": 2.0731491842864713,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1824817518248176e-05,
+      "loss": 1.1559,
+      "step": 6600
+    },
+    {
+      "epoch": 2.0737774113120127,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1822278641701048e-05,
+      "loss": 1.0895,
+      "step": 6602
+    },
+    {
+      "epoch": 2.0744056383375544,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1819739765153921e-05,
+      "loss": 1.2286,
+      "step": 6604
+    },
+    {
+      "epoch": 2.0750338653630958,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.1817200888606792e-05,
+      "loss": 1.1859,
+      "step": 6606
+    },
+    {
+      "epoch": 2.075662092388637,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1814662012059665e-05,
+      "loss": 1.0625,
+      "step": 6608
+    },
+    {
+      "epoch": 2.0762903194141784,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1812123135512537e-05,
+      "loss": 1.1966,
+      "step": 6610
+    },
+    {
+      "epoch": 2.0769185464397197,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.180958425896541e-05,
+      "loss": 1.2155,
+      "step": 6612
+    },
+    {
+      "epoch": 2.077546773465261,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1807045382418281e-05,
+      "loss": 1.2382,
+      "step": 6614
+    },
+    {
+      "epoch": 2.0781750004908024,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1804506505871154e-05,
+      "loss": 1.3356,
+      "step": 6616
+    },
+    {
+      "epoch": 2.0788032275163437,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1801967629324024e-05,
+      "loss": 1.1531,
+      "step": 6618
+    },
+    {
+      "epoch": 2.079431454541885,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1799428752776899e-05,
+      "loss": 1.1634,
+      "step": 6620
+    },
+    {
+      "epoch": 2.0800596815674264,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1796889876229768e-05,
+      "loss": 1.0372,
+      "step": 6622
+    },
+    {
+      "epoch": 2.0806879085929677,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1794350999682642e-05,
+      "loss": 1.1603,
+      "step": 6624
+    },
+    {
+      "epoch": 2.081316135618509,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1791812123135513e-05,
+      "loss": 1.2412,
+      "step": 6626
+    },
+    {
+      "epoch": 2.0819443626440504,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1789273246588386e-05,
+      "loss": 1.2592,
+      "step": 6628
+    },
+    {
+      "epoch": 2.0825725896695917,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1786734370041257e-05,
+      "loss": 1.197,
+      "step": 6630
+    },
+    {
+      "epoch": 2.083200816695133,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.178419549349413e-05,
+      "loss": 1.0592,
+      "step": 6632
+    },
+    {
+      "epoch": 2.0838290437206743,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1781656616947002e-05,
+      "loss": 1.1009,
+      "step": 6634
+    },
+    {
+      "epoch": 2.084457270746216,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1779117740399875e-05,
+      "loss": 1.1818,
+      "step": 6636
+    },
+    {
+      "epoch": 2.0850854977717574,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1776578863852746e-05,
+      "loss": 1.1779,
+      "step": 6638
+    },
+    {
+      "epoch": 2.0857137247972988,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.177403998730562e-05,
+      "loss": 1.1441,
+      "step": 6640
+    },
+    {
+      "epoch": 2.08634195182284,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.1771501110758489e-05,
+      "loss": 1.0996,
+      "step": 6642
+    },
+    {
+      "epoch": 2.0869701788483814,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1768962234211362e-05,
+      "loss": 1.0396,
+      "step": 6644
+    },
+    {
+      "epoch": 2.0875984058739228,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1766423357664234e-05,
+      "loss": 1.1889,
+      "step": 6646
+    },
+    {
+      "epoch": 2.088226632899464,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1763884481117107e-05,
+      "loss": 1.0357,
+      "step": 6648
+    },
+    {
+      "epoch": 2.0888548599250054,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1761345604569978e-05,
+      "loss": 1.2735,
+      "step": 6650
+    },
+    {
+      "epoch": 2.0894830869505467,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1758806728022851e-05,
+      "loss": 1.0615,
+      "step": 6652
+    },
+    {
+      "epoch": 2.090111313976088,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.1756267851475722e-05,
+      "loss": 1.2334,
+      "step": 6654
+    },
+    {
+      "epoch": 2.0907395410016294,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1753728974928596e-05,
+      "loss": 1.1619,
+      "step": 6656
+    },
+    {
+      "epoch": 2.0913677680271707,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.1751190098381467e-05,
+      "loss": 1.2739,
+      "step": 6658
+    },
+    {
+      "epoch": 2.091995995052712,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.174865122183434e-05,
+      "loss": 1.1238,
+      "step": 6660
+    },
+    {
+      "epoch": 2.0926242220782534,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1746112345287211e-05,
+      "loss": 1.2821,
+      "step": 6662
+    },
+    {
+      "epoch": 2.0932524491037947,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1743573468740084e-05,
+      "loss": 1.128,
+      "step": 6664
+    },
+    {
+      "epoch": 2.093880676129336,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1741034592192954e-05,
+      "loss": 1.2666,
+      "step": 6666
+    },
+    {
+      "epoch": 2.0945089031548774,
+      "grad_norm": 1.0,
+      "learning_rate": 1.1738495715645827e-05,
+      "loss": 1.2499,
+      "step": 6668
+    },
+    {
+      "epoch": 2.095137130180419,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1735956839098699e-05,
+      "loss": 1.1686,
+      "step": 6670
+    },
+    {
+      "epoch": 2.0957653572059605,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1733417962551572e-05,
+      "loss": 1.2688,
+      "step": 6672
+    },
+    {
+      "epoch": 2.096393584231502,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.1730879086004443e-05,
+      "loss": 1.1446,
+      "step": 6674
+    },
+    {
+      "epoch": 2.097021811257043,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.1728340209457316e-05,
+      "loss": 1.0533,
+      "step": 6676
+    },
+    {
+      "epoch": 2.0976500382825845,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1725801332910188e-05,
+      "loss": 1.1993,
+      "step": 6678
+    },
+    {
+      "epoch": 2.098278265308126,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.172326245636306e-05,
+      "loss": 1.2439,
+      "step": 6680
+    },
+    {
+      "epoch": 2.098906492333667,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1720723579815932e-05,
+      "loss": 1.1796,
+      "step": 6682
+    },
+    {
+      "epoch": 2.0995347193592084,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.1718184703268805e-05,
+      "loss": 1.1686,
+      "step": 6684
+    },
+    {
+      "epoch": 2.1001629463847498,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1715645826721678e-05,
+      "loss": 1.2925,
+      "step": 6686
+    },
+    {
+      "epoch": 2.100791173410291,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.171310695017455e-05,
+      "loss": 1.1615,
+      "step": 6688
+    },
+    {
+      "epoch": 2.1014194004358324,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.1710568073627423e-05,
+      "loss": 1.1193,
+      "step": 6690
+    },
+    {
+      "epoch": 2.1020476274613737,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1708029197080292e-05,
+      "loss": 1.1156,
+      "step": 6692
+    },
+    {
+      "epoch": 2.102675854486915,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1705490320533165e-05,
+      "loss": 1.3214,
+      "step": 6694
+    },
+    {
+      "epoch": 2.1033040815124564,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1702951443986037e-05,
+      "loss": 1.3833,
+      "step": 6696
+    },
+    {
+      "epoch": 2.1039323085379977,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.170041256743891e-05,
+      "loss": 1.1531,
+      "step": 6698
+    },
+    {
+      "epoch": 2.1045605355635395,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1697873690891781e-05,
+      "loss": 1.0297,
+      "step": 6700
+    },
+    {
+      "epoch": 2.105188762589081,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1695334814344654e-05,
+      "loss": 1.1562,
+      "step": 6702
+    },
+    {
+      "epoch": 2.105816989614622,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1692795937797526e-05,
+      "loss": 1.171,
+      "step": 6704
+    },
+    {
+      "epoch": 2.1064452166401635,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1690257061250399e-05,
+      "loss": 1.0549,
+      "step": 6706
+    },
+    {
+      "epoch": 2.107073443665705,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.168771818470327e-05,
+      "loss": 1.094,
+      "step": 6708
+    },
+    {
+      "epoch": 2.107701670691246,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1685179308156143e-05,
+      "loss": 1.1696,
+      "step": 6710
+    },
+    {
+      "epoch": 2.1083298977167875,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1682640431609013e-05,
+      "loss": 1.1725,
+      "step": 6712
+    },
+    {
+      "epoch": 2.108958124742329,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1680101555061888e-05,
+      "loss": 1.3241,
+      "step": 6714
+    },
+    {
+      "epoch": 2.10958635176787,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.1677562678514757e-05,
+      "loss": 1.0591,
+      "step": 6716
+    },
+    {
+      "epoch": 2.1102145787934115,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.167502380196763e-05,
+      "loss": 1.1485,
+      "step": 6718
+    },
+    {
+      "epoch": 2.110842805818953,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1672484925420502e-05,
+      "loss": 1.1271,
+      "step": 6720
+    },
+    {
+      "epoch": 2.111471032844494,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1669946048873375e-05,
+      "loss": 1.1642,
+      "step": 6722
+    },
+    {
+      "epoch": 2.1120992598700354,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1667407172326246e-05,
+      "loss": 1.2039,
+      "step": 6724
+    },
+    {
+      "epoch": 2.1127274868955768,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.166486829577912e-05,
+      "loss": 1.1125,
+      "step": 6726
+    },
+    {
+      "epoch": 2.113355713921118,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.166232941923199e-05,
+      "loss": 1.2027,
+      "step": 6728
+    },
+    {
+      "epoch": 2.1139839409466594,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.1659790542684864e-05,
+      "loss": 1.1512,
+      "step": 6730
+    },
+    {
+      "epoch": 2.1146121679722008,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.1657251666137735e-05,
+      "loss": 1.2524,
+      "step": 6732
+    },
+    {
+      "epoch": 2.1152403949977425,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1654712789590608e-05,
+      "loss": 1.1181,
+      "step": 6734
+    },
+    {
+      "epoch": 2.115868622023284,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.1652173913043478e-05,
+      "loss": 1.1363,
+      "step": 6736
+    },
+    {
+      "epoch": 2.116496849048825,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1649635036496351e-05,
+      "loss": 1.2479,
+      "step": 6738
+    },
+    {
+      "epoch": 2.1171250760743665,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.1647096159949222e-05,
+      "loss": 1.1432,
+      "step": 6740
+    },
+    {
+      "epoch": 2.117753303099908,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.1644557283402095e-05,
+      "loss": 1.2668,
+      "step": 6742
+    },
+    {
+      "epoch": 2.118381530125449,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1642018406854967e-05,
+      "loss": 1.1356,
+      "step": 6744
+    },
+    {
+      "epoch": 2.1190097571509905,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.163947953030784e-05,
+      "loss": 1.0993,
+      "step": 6746
+    },
+    {
+      "epoch": 2.119637984176532,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.1636940653760711e-05,
+      "loss": 1.0308,
+      "step": 6748
+    },
+    {
+      "epoch": 2.120266211202073,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.1634401777213584e-05,
+      "loss": 1.1342,
+      "step": 6750
+    },
+    {
+      "epoch": 2.1208944382276145,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.1631862900666456e-05,
+      "loss": 1.0846,
+      "step": 6752
+    },
+    {
+      "epoch": 2.121522665253156,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.1629324024119329e-05,
+      "loss": 1.1746,
+      "step": 6754
+    },
+    {
+      "epoch": 2.122150892278697,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1626785147572199e-05,
+      "loss": 1.114,
+      "step": 6756
+    },
+    {
+      "epoch": 2.1227791193042385,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1624246271025073e-05,
+      "loss": 1.123,
+      "step": 6758
+    },
+    {
+      "epoch": 2.12340734632978,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1621707394477943e-05,
+      "loss": 1.2326,
+      "step": 6760
+    },
+    {
+      "epoch": 2.124035573355321,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1619168517930816e-05,
+      "loss": 1.2209,
+      "step": 6762
+    },
+    {
+      "epoch": 2.1246638003808624,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.1616629641383687e-05,
+      "loss": 0.9569,
+      "step": 6764
+    },
+    {
+      "epoch": 2.125292027406404,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.161409076483656e-05,
+      "loss": 1.139,
+      "step": 6766
+    },
+    {
+      "epoch": 2.1259202544319455,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1611551888289432e-05,
+      "loss": 1.2186,
+      "step": 6768
+    },
+    {
+      "epoch": 2.126548481457487,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1609013011742305e-05,
+      "loss": 1.2514,
+      "step": 6770
+    },
+    {
+      "epoch": 2.127176708483028,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1606474135195178e-05,
+      "loss": 1.2252,
+      "step": 6772
+    },
+    {
+      "epoch": 2.1278049355085695,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.160393525864805e-05,
+      "loss": 1.1557,
+      "step": 6774
+    },
+    {
+      "epoch": 2.128433162534111,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1601396382100923e-05,
+      "loss": 1.2549,
+      "step": 6776
+    },
+    {
+      "epoch": 2.129061389559652,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.1598857505553794e-05,
+      "loss": 1.1201,
+      "step": 6778
+    },
+    {
+      "epoch": 2.1296896165851935,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.1596318629006667e-05,
+      "loss": 1.2005,
+      "step": 6780
+    },
+    {
+      "epoch": 2.130317843610735,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.1593779752459537e-05,
+      "loss": 1.1851,
+      "step": 6782
+    },
+    {
+      "epoch": 2.130946070636276,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1591240875912411e-05,
+      "loss": 1.0721,
+      "step": 6784
+    },
+    {
+      "epoch": 2.1315742976618175,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1588701999365281e-05,
+      "loss": 1.1945,
+      "step": 6786
+    },
+    {
+      "epoch": 2.132202524687359,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1586163122818154e-05,
+      "loss": 1.1655,
+      "step": 6788
+    },
+    {
+      "epoch": 2.1328307517129,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.1583624246271026e-05,
+      "loss": 1.0261,
+      "step": 6790
+    },
+    {
+      "epoch": 2.1334589787384415,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1581085369723899e-05,
+      "loss": 1.1892,
+      "step": 6792
+    },
+    {
+      "epoch": 2.134087205763983,
+      "grad_norm": 0.875,
+      "learning_rate": 1.157854649317677e-05,
+      "loss": 1.196,
+      "step": 6794
+    },
+    {
+      "epoch": 2.134715432789524,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.1576007616629643e-05,
+      "loss": 1.2396,
+      "step": 6796
+    },
+    {
+      "epoch": 2.1353436598150655,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1573468740082514e-05,
+      "loss": 1.2077,
+      "step": 6798
+    },
+    {
+      "epoch": 2.135971886840607,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.1570929863535388e-05,
+      "loss": 1.1368,
+      "step": 6800
+    },
+    {
+      "epoch": 2.1366001138661486,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1568390986988259e-05,
+      "loss": 1.1799,
+      "step": 6802
+    },
+    {
+      "epoch": 2.13722834089169,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1565852110441132e-05,
+      "loss": 1.1639,
+      "step": 6804
+    },
+    {
+      "epoch": 2.137856567917231,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1563313233894002e-05,
+      "loss": 1.207,
+      "step": 6806
+    },
+    {
+      "epoch": 2.1384847949427725,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1560774357346875e-05,
+      "loss": 1.0282,
+      "step": 6808
+    },
+    {
+      "epoch": 2.139113021968314,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1558235480799746e-05,
+      "loss": 1.095,
+      "step": 6810
+    },
+    {
+      "epoch": 2.139741248993855,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.155569660425262e-05,
+      "loss": 1.1784,
+      "step": 6812
+    },
+    {
+      "epoch": 2.1403694760193965,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.155315772770549e-05,
+      "loss": 1.2698,
+      "step": 6814
+    },
+    {
+      "epoch": 2.140997703044938,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1550618851158364e-05,
+      "loss": 1.1665,
+      "step": 6816
+    },
+    {
+      "epoch": 2.141625930070479,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.1548079974611235e-05,
+      "loss": 1.1065,
+      "step": 6818
+    },
+    {
+      "epoch": 2.1422541570960205,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.1545541098064108e-05,
+      "loss": 1.1438,
+      "step": 6820
+    },
+    {
+      "epoch": 2.142882384121562,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.154300222151698e-05,
+      "loss": 1.1842,
+      "step": 6822
+    },
+    {
+      "epoch": 2.143510611147103,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1540463344969853e-05,
+      "loss": 1.1727,
+      "step": 6824
+    },
+    {
+      "epoch": 2.1441388381726445,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1537924468422724e-05,
+      "loss": 1.2433,
+      "step": 6826
+    },
+    {
+      "epoch": 2.144767065198186,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1535385591875597e-05,
+      "loss": 1.1275,
+      "step": 6828
+    },
+    {
+      "epoch": 2.145395292223727,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1532846715328467e-05,
+      "loss": 1.1126,
+      "step": 6830
+    },
+    {
+      "epoch": 2.146023519249269,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.153030783878134e-05,
+      "loss": 1.0254,
+      "step": 6832
+    },
+    {
+      "epoch": 2.1466517462748103,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1527768962234211e-05,
+      "loss": 1.1911,
+      "step": 6834
+    },
+    {
+      "epoch": 2.1472799733003516,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1525230085687084e-05,
+      "loss": 1.2118,
+      "step": 6836
+    },
+    {
+      "epoch": 2.147908200325893,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.1522691209139956e-05,
+      "loss": 1.081,
+      "step": 6838
+    },
+    {
+      "epoch": 2.1485364273514342,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.1520152332592829e-05,
+      "loss": 1.2892,
+      "step": 6840
+    },
+    {
+      "epoch": 2.1491646543769756,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.15176134560457e-05,
+      "loss": 1.0985,
+      "step": 6842
+    },
+    {
+      "epoch": 2.149792881402517,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.1515074579498573e-05,
+      "loss": 1.0704,
+      "step": 6844
+    },
+    {
+      "epoch": 2.1504211084280582,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1512535702951445e-05,
+      "loss": 1.0915,
+      "step": 6846
+    },
+    {
+      "epoch": 2.1510493354535996,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1509996826404318e-05,
+      "loss": 1.1404,
+      "step": 6848
+    },
+    {
+      "epoch": 2.151677562479141,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1507457949857187e-05,
+      "loss": 1.0769,
+      "step": 6850
+    },
+    {
+      "epoch": 2.152305789504682,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1504919073310062e-05,
+      "loss": 1.0789,
+      "step": 6852
+    },
+    {
+      "epoch": 2.1529340165302235,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1502380196762932e-05,
+      "loss": 1.2021,
+      "step": 6854
+    },
+    {
+      "epoch": 2.153562243555765,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1499841320215805e-05,
+      "loss": 1.1089,
+      "step": 6856
+    },
+    {
+      "epoch": 2.154190470581306,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1497302443668678e-05,
+      "loss": 1.1592,
+      "step": 6858
+    },
+    {
+      "epoch": 2.1548186976068475,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.149476356712155e-05,
+      "loss": 1.2386,
+      "step": 6860
+    },
+    {
+      "epoch": 2.155446924632389,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1492224690574422e-05,
+      "loss": 1.1695,
+      "step": 6862
+    },
+    {
+      "epoch": 2.15607515165793,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1489685814027294e-05,
+      "loss": 1.1925,
+      "step": 6864
+    },
+    {
+      "epoch": 2.1567033786834715,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1487146937480167e-05,
+      "loss": 1.1248,
+      "step": 6866
+    },
+    {
+      "epoch": 2.1573316057090133,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1484608060933038e-05,
+      "loss": 1.1595,
+      "step": 6868
+    },
+    {
+      "epoch": 2.1579598327345546,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1482069184385911e-05,
+      "loss": 1.2658,
+      "step": 6870
+    },
+    {
+      "epoch": 2.158588059760096,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1479530307838783e-05,
+      "loss": 1.1204,
+      "step": 6872
+    },
+    {
+      "epoch": 2.1592162867856373,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1476991431291656e-05,
+      "loss": 1.2738,
+      "step": 6874
+    },
+    {
+      "epoch": 2.1598445138111786,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1474452554744525e-05,
+      "loss": 1.183,
+      "step": 6876
+    },
+    {
+      "epoch": 2.16047274083672,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.14719136781974e-05,
+      "loss": 1.2339,
+      "step": 6878
+    },
+    {
+      "epoch": 2.1611009678622612,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.146937480165027e-05,
+      "loss": 1.2331,
+      "step": 6880
+    },
+    {
+      "epoch": 2.1617291948878026,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.1466835925103143e-05,
+      "loss": 1.2599,
+      "step": 6882
+    },
+    {
+      "epoch": 2.162357421913344,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1464297048556014e-05,
+      "loss": 1.1194,
+      "step": 6884
+    },
+    {
+      "epoch": 2.1629856489388852,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1461758172008887e-05,
+      "loss": 1.1056,
+      "step": 6886
+    },
+    {
+      "epoch": 2.1636138759644266,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1459219295461759e-05,
+      "loss": 1.1664,
+      "step": 6888
+    },
+    {
+      "epoch": 2.164242102989968,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.1456680418914632e-05,
+      "loss": 1.3292,
+      "step": 6890
+    },
+    {
+      "epoch": 2.164870330015509,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1454141542367503e-05,
+      "loss": 1.1255,
+      "step": 6892
+    },
+    {
+      "epoch": 2.1654985570410505,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.1451602665820376e-05,
+      "loss": 1.193,
+      "step": 6894
+    },
+    {
+      "epoch": 2.166126784066592,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1449063789273248e-05,
+      "loss": 1.1852,
+      "step": 6896
+    },
+    {
+      "epoch": 2.1667550110921336,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.1446524912726121e-05,
+      "loss": 1.1961,
+      "step": 6898
+    },
+    {
+      "epoch": 2.167383238117675,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.144398603617899e-05,
+      "loss": 1.0907,
+      "step": 6900
+    },
+    {
+      "epoch": 2.1680114651432163,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1441447159631864e-05,
+      "loss": 1.3612,
+      "step": 6902
+    },
+    {
+      "epoch": 2.1686396921687576,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.1438908283084735e-05,
+      "loss": 1.1934,
+      "step": 6904
+    },
+    {
+      "epoch": 2.169267919194299,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1436369406537608e-05,
+      "loss": 1.2283,
+      "step": 6906
+    },
+    {
+      "epoch": 2.1698961462198403,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.143383052999048e-05,
+      "loss": 1.1552,
+      "step": 6908
+    },
+    {
+      "epoch": 2.1705243732453816,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1431291653443353e-05,
+      "loss": 1.173,
+      "step": 6910
+    },
+    {
+      "epoch": 2.171152600270923,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1428752776896224e-05,
+      "loss": 1.148,
+      "step": 6912
+    },
+    {
+      "epoch": 2.1717808272964643,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1426213900349097e-05,
+      "loss": 1.2096,
+      "step": 6914
+    },
+    {
+      "epoch": 2.1724090543220056,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1423675023801968e-05,
+      "loss": 1.193,
+      "step": 6916
+    },
+    {
+      "epoch": 2.173037281347547,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1421136147254841e-05,
+      "loss": 1.1284,
+      "step": 6918
+    },
+    {
+      "epoch": 2.1736655083730883,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1418597270707711e-05,
+      "loss": 1.1559,
+      "step": 6920
+    },
+    {
+      "epoch": 2.1742937353986296,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1416058394160586e-05,
+      "loss": 1.3299,
+      "step": 6922
+    },
+    {
+      "epoch": 2.174921962424171,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1413519517613456e-05,
+      "loss": 1.2832,
+      "step": 6924
+    },
+    {
+      "epoch": 2.1755501894497122,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1410980641066329e-05,
+      "loss": 1.1829,
+      "step": 6926
+    },
+    {
+      "epoch": 2.1761784164752536,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.14084417645192e-05,
+      "loss": 1.2522,
+      "step": 6928
+    },
+    {
+      "epoch": 2.176806643500795,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1405902887972073e-05,
+      "loss": 1.0834,
+      "step": 6930
+    },
+    {
+      "epoch": 2.177434870526336,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1403364011424945e-05,
+      "loss": 1.225,
+      "step": 6932
+    },
+    {
+      "epoch": 2.178063097551878,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1400825134877818e-05,
+      "loss": 1.0854,
+      "step": 6934
+    },
+    {
+      "epoch": 2.1786913245774193,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1398286258330689e-05,
+      "loss": 1.1618,
+      "step": 6936
+    },
+    {
+      "epoch": 2.1793195516029606,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1395747381783562e-05,
+      "loss": 1.2538,
+      "step": 6938
+    },
+    {
+      "epoch": 2.179947778628502,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1393208505236433e-05,
+      "loss": 1.2129,
+      "step": 6940
+    },
+    {
+      "epoch": 2.1805760056540433,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1390669628689307e-05,
+      "loss": 1.1104,
+      "step": 6942
+    },
+    {
+      "epoch": 2.1812042326795846,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.138813075214218e-05,
+      "loss": 1.1527,
+      "step": 6944
+    },
+    {
+      "epoch": 2.181832459705126,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.138559187559505e-05,
+      "loss": 1.1983,
+      "step": 6946
+    },
+    {
+      "epoch": 2.1824606867306673,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1383052999047924e-05,
+      "loss": 1.1743,
+      "step": 6948
+    },
+    {
+      "epoch": 2.1830889137562086,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.1380514122500794e-05,
+      "loss": 1.0212,
+      "step": 6950
+    },
+    {
+      "epoch": 2.18371714078175,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1377975245953667e-05,
+      "loss": 1.1853,
+      "step": 6952
+    },
+    {
+      "epoch": 2.1843453678072913,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1375436369406538e-05,
+      "loss": 1.1346,
+      "step": 6954
+    },
+    {
+      "epoch": 2.1849735948328326,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1372897492859411e-05,
+      "loss": 1.3125,
+      "step": 6956
+    },
+    {
+      "epoch": 2.185601821858374,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1370358616312283e-05,
+      "loss": 1.3227,
+      "step": 6958
+    },
+    {
+      "epoch": 2.1862300488839153,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1367819739765156e-05,
+      "loss": 1.2947,
+      "step": 6960
+    },
+    {
+      "epoch": 2.1868582759094566,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.1365280863218027e-05,
+      "loss": 1.1531,
+      "step": 6962
+    },
+    {
+      "epoch": 2.1874865029349984,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.13627419866709e-05,
+      "loss": 1.0524,
+      "step": 6964
+    },
+    {
+      "epoch": 2.1881147299605397,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1360203110123772e-05,
+      "loss": 1.1572,
+      "step": 6966
+    },
+    {
+      "epoch": 2.188742956986081,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.1357664233576645e-05,
+      "loss": 1.3277,
+      "step": 6968
+    },
+    {
+      "epoch": 2.1893711840116223,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.1355125357029514e-05,
+      "loss": 1.2082,
+      "step": 6970
+    },
+    {
+      "epoch": 2.1899994110371637,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1352586480482387e-05,
+      "loss": 1.2906,
+      "step": 6972
+    },
+    {
+      "epoch": 2.190627638062705,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1350047603935259e-05,
+      "loss": 1.1647,
+      "step": 6974
+    },
+    {
+      "epoch": 2.1912558650882463,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.1347508727388132e-05,
+      "loss": 1.2105,
+      "step": 6976
+    },
+    {
+      "epoch": 2.1918840921137877,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1344969850841003e-05,
+      "loss": 1.2325,
+      "step": 6978
+    },
+    {
+      "epoch": 2.192512319139329,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1342430974293876e-05,
+      "loss": 1.2232,
+      "step": 6980
+    },
+    {
+      "epoch": 2.1931405461648703,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1339892097746748e-05,
+      "loss": 1.2276,
+      "step": 6982
+    },
+    {
+      "epoch": 2.1937687731904116,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.133735322119962e-05,
+      "loss": 1.1263,
+      "step": 6984
+    },
+    {
+      "epoch": 2.194397000215953,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1334814344652492e-05,
+      "loss": 1.1551,
+      "step": 6986
+    },
+    {
+      "epoch": 2.1950252272414943,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1332275468105365e-05,
+      "loss": 1.1824,
+      "step": 6988
+    },
+    {
+      "epoch": 2.1956534542670356,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1329736591558237e-05,
+      "loss": 1.2166,
+      "step": 6990
+    },
+    {
+      "epoch": 2.196281681292577,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.132719771501111e-05,
+      "loss": 1.2635,
+      "step": 6992
+    },
+    {
+      "epoch": 2.1969099083181183,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.132465883846398e-05,
+      "loss": 1.2588,
+      "step": 6994
+    },
+    {
+      "epoch": 2.1975381353436596,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1322119961916852e-05,
+      "loss": 1.3271,
+      "step": 6996
+    },
+    {
+      "epoch": 2.1981663623692014,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1319581085369724e-05,
+      "loss": 1.1563,
+      "step": 6998
+    },
+    {
+      "epoch": 2.1987945893947427,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1317042208822597e-05,
+      "loss": 1.109,
+      "step": 7000
+    },
+    {
+      "epoch": 2.199422816420284,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.1314503332275468e-05,
+      "loss": 1.144,
+      "step": 7002
+    },
+    {
+      "epoch": 2.2000510434458254,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1311964455728341e-05,
+      "loss": 1.2422,
+      "step": 7004
+    },
+    {
+      "epoch": 2.2006792704713667,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1309425579181213e-05,
+      "loss": 1.2329,
+      "step": 7006
+    },
+    {
+      "epoch": 2.201307497496908,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.1306886702634086e-05,
+      "loss": 1.1037,
+      "step": 7008
+    },
+    {
+      "epoch": 2.2019357245224493,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 1.2786,
+      "step": 7010
+    },
+    {
+      "epoch": 2.2025639515479907,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.130180894953983e-05,
+      "loss": 1.3284,
+      "step": 7012
+    },
+    {
+      "epoch": 2.203192178573532,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.12992700729927e-05,
+      "loss": 1.1047,
+      "step": 7014
+    },
+    {
+      "epoch": 2.2038204055990733,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1296731196445575e-05,
+      "loss": 1.2299,
+      "step": 7016
+    },
+    {
+      "epoch": 2.2044486326246147,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1294192319898444e-05,
+      "loss": 1.1535,
+      "step": 7018
+    },
+    {
+      "epoch": 2.205076859650156,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.1291653443351318e-05,
+      "loss": 1.3391,
+      "step": 7020
+    },
+    {
+      "epoch": 2.2057050866756973,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1289114566804189e-05,
+      "loss": 1.092,
+      "step": 7022
+    },
+    {
+      "epoch": 2.2063333137012386,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.1286575690257062e-05,
+      "loss": 1.2479,
+      "step": 7024
+    },
+    {
+      "epoch": 2.20696154072678,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.1284036813709933e-05,
+      "loss": 1.0299,
+      "step": 7026
+    },
+    {
+      "epoch": 2.2075897677523213,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.1281497937162806e-05,
+      "loss": 1.1123,
+      "step": 7028
+    },
+    {
+      "epoch": 2.208217994777863,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.127895906061568e-05,
+      "loss": 1.1984,
+      "step": 7030
+    },
+    {
+      "epoch": 2.2088462218034044,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1276420184068551e-05,
+      "loss": 1.1251,
+      "step": 7032
+    },
+    {
+      "epoch": 2.2094744488289457,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.1273881307521424e-05,
+      "loss": 1.0623,
+      "step": 7034
+    },
+    {
+      "epoch": 2.210102675854487,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1271342430974295e-05,
+      "loss": 1.1495,
+      "step": 7036
+    },
+    {
+      "epoch": 2.2107309028800284,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1268803554427168e-05,
+      "loss": 1.1898,
+      "step": 7038
+    },
+    {
+      "epoch": 2.2113591299055697,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.1266264677880038e-05,
+      "loss": 1.1346,
+      "step": 7040
+    },
+    {
+      "epoch": 2.211987356931111,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.1263725801332913e-05,
+      "loss": 1.0782,
+      "step": 7042
+    },
+    {
+      "epoch": 2.2126155839566524,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.1261186924785783e-05,
+      "loss": 1.0988,
+      "step": 7044
+    },
+    {
+      "epoch": 2.2132438109821937,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1258648048238656e-05,
+      "loss": 1.2726,
+      "step": 7046
+    },
+    {
+      "epoch": 2.213872038007735,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1256109171691527e-05,
+      "loss": 1.3077,
+      "step": 7048
+    },
+    {
+      "epoch": 2.2145002650332763,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.12535702951444e-05,
+      "loss": 1.1586,
+      "step": 7050
+    },
+    {
+      "epoch": 2.2151284920588177,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1251031418597271e-05,
+      "loss": 1.1196,
+      "step": 7052
+    },
+    {
+      "epoch": 2.215756719084359,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1248492542050145e-05,
+      "loss": 1.3712,
+      "step": 7054
+    },
+    {
+      "epoch": 2.2163849461099003,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.1245953665503016e-05,
+      "loss": 1.1446,
+      "step": 7056
+    },
+    {
+      "epoch": 2.2170131731354417,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1243414788955889e-05,
+      "loss": 1.2527,
+      "step": 7058
+    },
+    {
+      "epoch": 2.217641400160983,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.124087591240876e-05,
+      "loss": 1.1417,
+      "step": 7060
+    },
+    {
+      "epoch": 2.2182696271865243,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1238337035861633e-05,
+      "loss": 1.1248,
+      "step": 7062
+    },
+    {
+      "epoch": 2.218897854212066,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.1235798159314503e-05,
+      "loss": 1.1408,
+      "step": 7064
+    },
+    {
+      "epoch": 2.2195260812376074,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.1233259282767376e-05,
+      "loss": 1.1888,
+      "step": 7066
+    },
+    {
+      "epoch": 2.2201543082631487,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1230720406220248e-05,
+      "loss": 1.3227,
+      "step": 7068
+    },
+    {
+      "epoch": 2.22078253528869,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.122818152967312e-05,
+      "loss": 1.2715,
+      "step": 7070
+    },
+    {
+      "epoch": 2.2214107623142314,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.1225642653125992e-05,
+      "loss": 1.1767,
+      "step": 7072
+    },
+    {
+      "epoch": 2.2220389893397727,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.1223103776578865e-05,
+      "loss": 1.2829,
+      "step": 7074
+    },
+    {
+      "epoch": 2.222667216365314,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1220564900031737e-05,
+      "loss": 1.1629,
+      "step": 7076
+    },
+    {
+      "epoch": 2.2232954433908554,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.121802602348461e-05,
+      "loss": 1.252,
+      "step": 7078
+    },
+    {
+      "epoch": 2.2239236704163967,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1215487146937481e-05,
+      "loss": 1.0698,
+      "step": 7080
+    },
+    {
+      "epoch": 2.224551897441938,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1212948270390354e-05,
+      "loss": 1.136,
+      "step": 7082
+    },
+    {
+      "epoch": 2.2251801244674794,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.1210409393843224e-05,
+      "loss": 1.1405,
+      "step": 7084
+    },
+    {
+      "epoch": 2.2258083514930207,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.1207870517296099e-05,
+      "loss": 1.2,
+      "step": 7086
+    },
+    {
+      "epoch": 2.226436578518562,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1205331640748968e-05,
+      "loss": 1.1154,
+      "step": 7088
+    },
+    {
+      "epoch": 2.2270648055441034,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1202792764201841e-05,
+      "loss": 1.1577,
+      "step": 7090
+    },
+    {
+      "epoch": 2.2276930325696447,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1200253887654713e-05,
+      "loss": 1.1263,
+      "step": 7092
+    },
+    {
+      "epoch": 2.228321259595186,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1197715011107586e-05,
+      "loss": 1.0956,
+      "step": 7094
+    },
+    {
+      "epoch": 2.228949486620728,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.1195176134560457e-05,
+      "loss": 1.0752,
+      "step": 7096
+    },
+    {
+      "epoch": 2.229577713646269,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.119263725801333e-05,
+      "loss": 1.08,
+      "step": 7098
+    },
+    {
+      "epoch": 2.2302059406718104,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.1190098381466202e-05,
+      "loss": 1.206,
+      "step": 7100
+    },
+    {
+      "epoch": 2.2308341676973518,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1187559504919075e-05,
+      "loss": 1.1989,
+      "step": 7102
+    },
+    {
+      "epoch": 2.231462394722893,
+      "grad_norm": 1.0,
+      "learning_rate": 1.1185020628371946e-05,
+      "loss": 1.2279,
+      "step": 7104
+    },
+    {
+      "epoch": 2.2320906217484344,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1182481751824819e-05,
+      "loss": 1.2506,
+      "step": 7106
+    },
+    {
+      "epoch": 2.2327188487739758,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.1179942875277689e-05,
+      "loss": 1.1432,
+      "step": 7108
+    },
+    {
+      "epoch": 2.233347075799517,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1177403998730562e-05,
+      "loss": 1.2258,
+      "step": 7110
+    },
+    {
+      "epoch": 2.2339753028250584,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1174865122183433e-05,
+      "loss": 1.1279,
+      "step": 7112
+    },
+    {
+      "epoch": 2.2346035298505997,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1172326245636306e-05,
+      "loss": 1.2938,
+      "step": 7114
+    },
+    {
+      "epoch": 2.235231756876141,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.116978736908918e-05,
+      "loss": 1.1135,
+      "step": 7116
+    },
+    {
+      "epoch": 2.2358599839016824,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.116724849254205e-05,
+      "loss": 1.1685,
+      "step": 7118
+    },
+    {
+      "epoch": 2.2364882109272237,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1164709615994924e-05,
+      "loss": 1.1768,
+      "step": 7120
+    },
+    {
+      "epoch": 2.237116437952765,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.1162170739447795e-05,
+      "loss": 1.118,
+      "step": 7122
+    },
+    {
+      "epoch": 2.2377446649783064,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1159631862900668e-05,
+      "loss": 1.1579,
+      "step": 7124
+    },
+    {
+      "epoch": 2.2383728920038477,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.115709298635354e-05,
+      "loss": 1.2208,
+      "step": 7126
+    },
+    {
+      "epoch": 2.239001119029389,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1154554109806413e-05,
+      "loss": 1.2217,
+      "step": 7128
+    },
+    {
+      "epoch": 2.239629346054931,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1152015233259284e-05,
+      "loss": 1.0771,
+      "step": 7130
+    },
+    {
+      "epoch": 2.240257573080472,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1149476356712157e-05,
+      "loss": 1.366,
+      "step": 7132
+    },
+    {
+      "epoch": 2.2408858001060135,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1146937480165027e-05,
+      "loss": 1.118,
+      "step": 7134
+    },
+    {
+      "epoch": 2.241514027131555,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.11443986036179e-05,
+      "loss": 1.1676,
+      "step": 7136
+    },
+    {
+      "epoch": 2.242142254157096,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1141859727070771e-05,
+      "loss": 1.2008,
+      "step": 7138
+    },
+    {
+      "epoch": 2.2427704811826374,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1139320850523645e-05,
+      "loss": 1.1776,
+      "step": 7140
+    },
+    {
+      "epoch": 2.2433987082081788,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1136781973976516e-05,
+      "loss": 1.2754,
+      "step": 7142
+    },
+    {
+      "epoch": 2.24402693523372,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.1134243097429389e-05,
+      "loss": 1.2551,
+      "step": 7144
+    },
+    {
+      "epoch": 2.2446551622592614,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.113170422088226e-05,
+      "loss": 1.1994,
+      "step": 7146
+    },
+    {
+      "epoch": 2.2452833892848028,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1129165344335133e-05,
+      "loss": 1.2192,
+      "step": 7148
+    },
+    {
+      "epoch": 2.245911616310344,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.1126626467788005e-05,
+      "loss": 1.2826,
+      "step": 7150
+    },
+    {
+      "epoch": 2.2465398433358854,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1124087591240878e-05,
+      "loss": 1.2019,
+      "step": 7152
+    },
+    {
+      "epoch": 2.2471680703614267,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.112154871469375e-05,
+      "loss": 1.1506,
+      "step": 7154
+    },
+    {
+      "epoch": 2.247796297386968,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.1119009838146622e-05,
+      "loss": 1.0458,
+      "step": 7156
+    },
+    {
+      "epoch": 2.2484245244125094,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1116470961599492e-05,
+      "loss": 1.1759,
+      "step": 7158
+    },
+    {
+      "epoch": 2.2490527514380507,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1113932085052365e-05,
+      "loss": 1.1225,
+      "step": 7160
+    },
+    {
+      "epoch": 2.2496809784635925,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1111393208505236e-05,
+      "loss": 1.1735,
+      "step": 7162
+    },
+    {
+      "epoch": 2.250309205489134,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.110885433195811e-05,
+      "loss": 1.0964,
+      "step": 7164
+    },
+    {
+      "epoch": 2.250937432514675,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.1106315455410981e-05,
+      "loss": 1.1575,
+      "step": 7166
+    },
+    {
+      "epoch": 2.2515656595402165,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1103776578863854e-05,
+      "loss": 1.308,
+      "step": 7168
+    },
+    {
+      "epoch": 2.252193886565758,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.1101237702316725e-05,
+      "loss": 1.2186,
+      "step": 7170
+    },
+    {
+      "epoch": 2.252822113591299,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1098698825769598e-05,
+      "loss": 1.2044,
+      "step": 7172
+    },
+    {
+      "epoch": 2.2534503406168405,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.109615994922247e-05,
+      "loss": 1.1716,
+      "step": 7174
+    },
+    {
+      "epoch": 2.254078567642382,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.1093621072675343e-05,
+      "loss": 1.1242,
+      "step": 7176
+    },
+    {
+      "epoch": 2.254706794667923,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1091082196128213e-05,
+      "loss": 1.254,
+      "step": 7178
+    },
+    {
+      "epoch": 2.2553350216934644,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1088543319581086e-05,
+      "loss": 1.2207,
+      "step": 7180
+    },
+    {
+      "epoch": 2.2559632487190058,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.1086004443033957e-05,
+      "loss": 1.2314,
+      "step": 7182
+    },
+    {
+      "epoch": 2.256591475744547,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.108346556648683e-05,
+      "loss": 1.1507,
+      "step": 7184
+    },
+    {
+      "epoch": 2.2572197027700884,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1080926689939702e-05,
+      "loss": 1.0976,
+      "step": 7186
+    },
+    {
+      "epoch": 2.2578479297956298,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1078387813392575e-05,
+      "loss": 1.2232,
+      "step": 7188
+    },
+    {
+      "epoch": 2.258476156821171,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1075848936845446e-05,
+      "loss": 1.3585,
+      "step": 7190
+    },
+    {
+      "epoch": 2.2591043838467124,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.1073310060298319e-05,
+      "loss": 1.2097,
+      "step": 7192
+    },
+    {
+      "epoch": 2.2597326108722537,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.107077118375119e-05,
+      "loss": 1.1563,
+      "step": 7194
+    },
+    {
+      "epoch": 2.260360837897795,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.1068232307204064e-05,
+      "loss": 1.1191,
+      "step": 7196
+    },
+    {
+      "epoch": 2.260989064923337,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.1065693430656935e-05,
+      "loss": 1.1386,
+      "step": 7198
+    },
+    {
+      "epoch": 2.261617291948878,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.1063154554109808e-05,
+      "loss": 1.0648,
+      "step": 7200
+    },
+    {
+      "epoch": 2.2622455189744195,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.1060615677562681e-05,
+      "loss": 1.2043,
+      "step": 7202
+    },
+    {
+      "epoch": 2.262873745999961,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.105807680101555e-05,
+      "loss": 1.1253,
+      "step": 7204
+    },
+    {
+      "epoch": 2.263501973025502,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.1055537924468424e-05,
+      "loss": 1.2436,
+      "step": 7206
+    },
+    {
+      "epoch": 2.2641302000510435,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.1052999047921295e-05,
+      "loss": 1.1287,
+      "step": 7208
+    },
+    {
+      "epoch": 2.264758427076585,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1050460171374168e-05,
+      "loss": 1.1132,
+      "step": 7210
+    },
+    {
+      "epoch": 2.265386654102126,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.104792129482704e-05,
+      "loss": 1.1685,
+      "step": 7212
+    },
+    {
+      "epoch": 2.2660148811276675,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.1045382418279913e-05,
+      "loss": 1.1164,
+      "step": 7214
+    },
+    {
+      "epoch": 2.266643108153209,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1042843541732784e-05,
+      "loss": 1.259,
+      "step": 7216
+    },
+    {
+      "epoch": 2.26727133517875,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.1040304665185657e-05,
+      "loss": 1.2067,
+      "step": 7218
+    },
+    {
+      "epoch": 2.2678995622042915,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.1037765788638529e-05,
+      "loss": 1.2234,
+      "step": 7220
+    },
+    {
+      "epoch": 2.268527789229833,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1035226912091402e-05,
+      "loss": 1.2389,
+      "step": 7222
+    },
+    {
+      "epoch": 2.269156016255374,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1032688035544273e-05,
+      "loss": 1.307,
+      "step": 7224
+    },
+    {
+      "epoch": 2.269784243280916,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.1030149158997146e-05,
+      "loss": 1.1486,
+      "step": 7226
+    },
+    {
+      "epoch": 2.270412470306457,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1027610282450016e-05,
+      "loss": 1.2867,
+      "step": 7228
+    },
+    {
+      "epoch": 2.2710406973319985,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.1025071405902889e-05,
+      "loss": 1.1398,
+      "step": 7230
+    },
+    {
+      "epoch": 2.27166892435754,
+      "grad_norm": 0.875,
+      "learning_rate": 1.102253252935576e-05,
+      "loss": 1.1913,
+      "step": 7232
+    },
+    {
+      "epoch": 2.272297151383081,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1019993652808633e-05,
+      "loss": 1.2196,
+      "step": 7234
+    },
+    {
+      "epoch": 2.2729253784086225,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1017454776261505e-05,
+      "loss": 1.209,
+      "step": 7236
+    },
+    {
+      "epoch": 2.273553605434164,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1014915899714378e-05,
+      "loss": 1.0517,
+      "step": 7238
+    },
+    {
+      "epoch": 2.274181832459705,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.101237702316725e-05,
+      "loss": 1.1559,
+      "step": 7240
+    },
+    {
+      "epoch": 2.2748100594852465,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1009838146620122e-05,
+      "loss": 1.2096,
+      "step": 7242
+    },
+    {
+      "epoch": 2.275438286510788,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1007299270072994e-05,
+      "loss": 1.0206,
+      "step": 7244
+    },
+    {
+      "epoch": 2.276066513536329,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1004760393525867e-05,
+      "loss": 1.2358,
+      "step": 7246
+    },
+    {
+      "epoch": 2.2766947405618705,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.1002221516978736e-05,
+      "loss": 1.1215,
+      "step": 7248
+    },
+    {
+      "epoch": 2.277322967587412,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.0999682640431611e-05,
+      "loss": 1.313,
+      "step": 7250
+    },
+    {
+      "epoch": 2.277951194612953,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0997143763884481e-05,
+      "loss": 1.1596,
+      "step": 7252
+    },
+    {
+      "epoch": 2.2785794216384945,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0994604887337354e-05,
+      "loss": 1.1222,
+      "step": 7254
+    },
+    {
+      "epoch": 2.279207648664036,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.0992066010790225e-05,
+      "loss": 1.116,
+      "step": 7256
+    },
+    {
+      "epoch": 2.279835875689577,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0989527134243098e-05,
+      "loss": 1.1827,
+      "step": 7258
+    },
+    {
+      "epoch": 2.2804641027151185,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.098698825769597e-05,
+      "loss": 1.1286,
+      "step": 7260
+    },
+    {
+      "epoch": 2.28109232974066,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0984449381148843e-05,
+      "loss": 0.9997,
+      "step": 7262
+    },
+    {
+      "epoch": 2.2817205567662016,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0981910504601714e-05,
+      "loss": 1.1823,
+      "step": 7264
+    },
+    {
+      "epoch": 2.282348783791743,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0979371628054587e-05,
+      "loss": 1.2145,
+      "step": 7266
+    },
+    {
+      "epoch": 2.282977010817284,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0976832751507459e-05,
+      "loss": 1.1597,
+      "step": 7268
+    },
+    {
+      "epoch": 2.2836052378428255,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.0974293874960332e-05,
+      "loss": 1.1206,
+      "step": 7270
+    },
+    {
+      "epoch": 2.284233464868367,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.0971754998413201e-05,
+      "loss": 1.1522,
+      "step": 7272
+    },
+    {
+      "epoch": 2.284861691893908,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0969216121866075e-05,
+      "loss": 1.2065,
+      "step": 7274
+    },
+    {
+      "epoch": 2.2854899189194495,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0966677245318946e-05,
+      "loss": 1.2534,
+      "step": 7276
+    },
+    {
+      "epoch": 2.286118145944991,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0964138368771819e-05,
+      "loss": 1.3111,
+      "step": 7278
+    },
+    {
+      "epoch": 2.286746372970532,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.096159949222469e-05,
+      "loss": 1.1712,
+      "step": 7280
+    },
+    {
+      "epoch": 2.2873745999960735,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0959060615677563e-05,
+      "loss": 1.2978,
+      "step": 7282
+    },
+    {
+      "epoch": 2.288002827021615,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.0956521739130435e-05,
+      "loss": 1.1384,
+      "step": 7284
+    },
+    {
+      "epoch": 2.288631054047156,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.0953982862583308e-05,
+      "loss": 1.1677,
+      "step": 7286
+    },
+    {
+      "epoch": 2.2892592810726975,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0951443986036181e-05,
+      "loss": 1.2128,
+      "step": 7288
+    },
+    {
+      "epoch": 2.289887508098239,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.0948905109489052e-05,
+      "loss": 1.274,
+      "step": 7290
+    },
+    {
+      "epoch": 2.2905157351237806,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0946366232941925e-05,
+      "loss": 1.1172,
+      "step": 7292
+    },
+    {
+      "epoch": 2.291143962149322,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0943827356394797e-05,
+      "loss": 1.2049,
+      "step": 7294
+    },
+    {
+      "epoch": 2.2917721891748633,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.094128847984767e-05,
+      "loss": 1.1945,
+      "step": 7296
+    },
+    {
+      "epoch": 2.2924004162004046,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.093874960330054e-05,
+      "loss": 1.0366,
+      "step": 7298
+    },
+    {
+      "epoch": 2.293028643225946,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0936210726753413e-05,
+      "loss": 1.1452,
+      "step": 7300
+    },
+    {
+      "epoch": 2.2936568702514872,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0933671850206284e-05,
+      "loss": 1.2821,
+      "step": 7302
+    },
+    {
+      "epoch": 2.2942850972770286,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0931132973659157e-05,
+      "loss": 1.0961,
+      "step": 7304
+    },
+    {
+      "epoch": 2.29491332430257,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.0928594097112029e-05,
+      "loss": 1.2104,
+      "step": 7306
+    },
+    {
+      "epoch": 2.295541551328111,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0926055220564902e-05,
+      "loss": 1.29,
+      "step": 7308
+    },
+    {
+      "epoch": 2.2961697783536525,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0923516344017773e-05,
+      "loss": 1.2525,
+      "step": 7310
+    },
+    {
+      "epoch": 2.296798005379194,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0920977467470646e-05,
+      "loss": 1.2847,
+      "step": 7312
+    },
+    {
+      "epoch": 2.297426232404735,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0918438590923517e-05,
+      "loss": 1.1948,
+      "step": 7314
+    },
+    {
+      "epoch": 2.2980544594302765,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.091589971437639e-05,
+      "loss": 1.1595,
+      "step": 7316
+    },
+    {
+      "epoch": 2.298682686455818,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.091336083782926e-05,
+      "loss": 1.2342,
+      "step": 7318
+    },
+    {
+      "epoch": 2.299310913481359,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0910821961282135e-05,
+      "loss": 1.1913,
+      "step": 7320
+    },
+    {
+      "epoch": 2.2999391405069005,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.0908283084735005e-05,
+      "loss": 1.1333,
+      "step": 7322
+    },
+    {
+      "epoch": 2.300567367532442,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0905744208187878e-05,
+      "loss": 1.1376,
+      "step": 7324
+    },
+    {
+      "epoch": 2.301195594557983,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0903205331640749e-05,
+      "loss": 1.1524,
+      "step": 7326
+    },
+    {
+      "epoch": 2.3018238215835245,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0900666455093622e-05,
+      "loss": 1.1174,
+      "step": 7328
+    },
+    {
+      "epoch": 2.3024520486090663,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0898127578546494e-05,
+      "loss": 1.1047,
+      "step": 7330
+    },
+    {
+      "epoch": 2.3030802756346076,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0895588701999367e-05,
+      "loss": 1.1898,
+      "step": 7332
+    },
+    {
+      "epoch": 2.303708502660149,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0893049825452238e-05,
+      "loss": 1.3392,
+      "step": 7334
+    },
+    {
+      "epoch": 2.3043367296856903,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0890510948905111e-05,
+      "loss": 1.2683,
+      "step": 7336
+    },
+    {
+      "epoch": 2.3049649567112316,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0887972072357982e-05,
+      "loss": 1.0879,
+      "step": 7338
+    },
+    {
+      "epoch": 2.305593183736773,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.0885433195810856e-05,
+      "loss": 1.095,
+      "step": 7340
+    },
+    {
+      "epoch": 2.3062214107623142,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0882894319263725e-05,
+      "loss": 1.3192,
+      "step": 7342
+    },
+    {
+      "epoch": 2.3068496377878556,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0880355442716598e-05,
+      "loss": 1.3008,
+      "step": 7344
+    },
+    {
+      "epoch": 2.307477864813397,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.087781656616947e-05,
+      "loss": 1.1895,
+      "step": 7346
+    },
+    {
+      "epoch": 2.3081060918389382,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0875277689622343e-05,
+      "loss": 1.1833,
+      "step": 7348
+    },
+    {
+      "epoch": 2.3087343188644796,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0872738813075214e-05,
+      "loss": 1.1241,
+      "step": 7350
+    },
+    {
+      "epoch": 2.309362545890021,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.0870199936528087e-05,
+      "loss": 1.1805,
+      "step": 7352
+    },
+    {
+      "epoch": 2.309990772915562,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0867661059980959e-05,
+      "loss": 1.0791,
+      "step": 7354
+    },
+    {
+      "epoch": 2.3106189999411035,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0865122183433832e-05,
+      "loss": 1.1263,
+      "step": 7356
+    },
+    {
+      "epoch": 2.3112472269666453,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0862583306886703e-05,
+      "loss": 1.1829,
+      "step": 7358
+    },
+    {
+      "epoch": 2.3118754539921866,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.0860044430339576e-05,
+      "loss": 1.2385,
+      "step": 7360
+    },
+    {
+      "epoch": 2.312503681017728,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0857505553792448e-05,
+      "loss": 1.1448,
+      "step": 7362
+    },
+    {
+      "epoch": 2.3131319080432693,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.085496667724532e-05,
+      "loss": 1.2102,
+      "step": 7364
+    },
+    {
+      "epoch": 2.3137601350688106,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.085242780069819e-05,
+      "loss": 1.2639,
+      "step": 7366
+    },
+    {
+      "epoch": 2.314388362094352,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0849888924151063e-05,
+      "loss": 1.1321,
+      "step": 7368
+    },
+    {
+      "epoch": 2.3150165891198933,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0847350047603935e-05,
+      "loss": 1.0745,
+      "step": 7370
+    },
+    {
+      "epoch": 2.3156448161454346,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.0844811171056808e-05,
+      "loss": 1.1193,
+      "step": 7372
+    },
+    {
+      "epoch": 2.316273043170976,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0842272294509681e-05,
+      "loss": 1.303,
+      "step": 7374
+    },
+    {
+      "epoch": 2.3169012701965173,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.0839733417962552e-05,
+      "loss": 1.1937,
+      "step": 7376
+    },
+    {
+      "epoch": 2.3175294972220586,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.0837194541415425e-05,
+      "loss": 1.229,
+      "step": 7378
+    },
+    {
+      "epoch": 2.3181577242476,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.0834655664868297e-05,
+      "loss": 1.0703,
+      "step": 7380
+    },
+    {
+      "epoch": 2.3187859512731412,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.083211678832117e-05,
+      "loss": 1.3122,
+      "step": 7382
+    },
+    {
+      "epoch": 2.3194141782986826,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0829577911774041e-05,
+      "loss": 1.208,
+      "step": 7384
+    },
+    {
+      "epoch": 2.320042405324224,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0827039035226914e-05,
+      "loss": 1.1552,
+      "step": 7386
+    },
+    {
+      "epoch": 2.3206706323497652,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.0824500158679786e-05,
+      "loss": 1.1077,
+      "step": 7388
+    },
+    {
+      "epoch": 2.3212988593753066,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0821961282132659e-05,
+      "loss": 1.2388,
+      "step": 7390
+    },
+    {
+      "epoch": 2.321927086400848,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.0819422405585528e-05,
+      "loss": 1.2115,
+      "step": 7392
+    },
+    {
+      "epoch": 2.322555313426389,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0816883529038402e-05,
+      "loss": 1.1479,
+      "step": 7394
+    },
+    {
+      "epoch": 2.323183540451931,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0814344652491273e-05,
+      "loss": 1.2548,
+      "step": 7396
+    },
+    {
+      "epoch": 2.3238117674774723,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0811805775944146e-05,
+      "loss": 1.1445,
+      "step": 7398
+    },
+    {
+      "epoch": 2.3244399945030136,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0809266899397017e-05,
+      "loss": 1.1688,
+      "step": 7400
+    },
+    {
+      "epoch": 2.325068221528555,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.080672802284989e-05,
+      "loss": 1.2903,
+      "step": 7402
+    },
+    {
+      "epoch": 2.3256964485540963,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0804189146302762e-05,
+      "loss": 1.2819,
+      "step": 7404
+    },
+    {
+      "epoch": 2.3263246755796376,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0801650269755635e-05,
+      "loss": 1.1805,
+      "step": 7406
+    },
+    {
+      "epoch": 2.326952902605179,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.0799111393208506e-05,
+      "loss": 0.9431,
+      "step": 7408
+    },
+    {
+      "epoch": 2.3275811296307203,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.079657251666138e-05,
+      "loss": 1.0793,
+      "step": 7410
+    },
+    {
+      "epoch": 2.3282093566562616,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.0794033640114249e-05,
+      "loss": 1.1367,
+      "step": 7412
+    },
+    {
+      "epoch": 2.328837583681803,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0791494763567124e-05,
+      "loss": 1.1915,
+      "step": 7414
+    },
+    {
+      "epoch": 2.3294658107073443,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0788955887019993e-05,
+      "loss": 1.1777,
+      "step": 7416
+    },
+    {
+      "epoch": 2.3300940377328856,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0786417010472867e-05,
+      "loss": 1.1717,
+      "step": 7418
+    },
+    {
+      "epoch": 2.330722264758427,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0783878133925738e-05,
+      "loss": 1.229,
+      "step": 7420
+    },
+    {
+      "epoch": 2.3313504917839682,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0781339257378611e-05,
+      "loss": 1.1072,
+      "step": 7422
+    },
+    {
+      "epoch": 2.33197871880951,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0778800380831482e-05,
+      "loss": 1.2157,
+      "step": 7424
+    },
+    {
+      "epoch": 2.3326069458350513,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0776261504284355e-05,
+      "loss": 1.1263,
+      "step": 7426
+    },
+    {
+      "epoch": 2.3332351728605927,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0773722627737227e-05,
+      "loss": 1.2423,
+      "step": 7428
+    },
+    {
+      "epoch": 2.333863399886134,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.07711837511901e-05,
+      "loss": 1.2594,
+      "step": 7430
+    },
+    {
+      "epoch": 2.3344916269116753,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0768644874642971e-05,
+      "loss": 1.2169,
+      "step": 7432
+    },
+    {
+      "epoch": 2.3351198539372167,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0766105998095844e-05,
+      "loss": 1.2165,
+      "step": 7434
+    },
+    {
+      "epoch": 2.335748080962758,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0763567121548714e-05,
+      "loss": 1.1919,
+      "step": 7436
+    },
+    {
+      "epoch": 2.3363763079882993,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.0761028245001587e-05,
+      "loss": 1.1077,
+      "step": 7438
+    },
+    {
+      "epoch": 2.3370045350138406,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0758489368454459e-05,
+      "loss": 1.1936,
+      "step": 7440
+    },
+    {
+      "epoch": 2.337632762039382,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0755950491907332e-05,
+      "loss": 1.088,
+      "step": 7442
+    },
+    {
+      "epoch": 2.3382609890649233,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.0753411615360203e-05,
+      "loss": 1.2247,
+      "step": 7444
+    },
+    {
+      "epoch": 2.3388892160904646,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0750872738813076e-05,
+      "loss": 1.1426,
+      "step": 7446
+    },
+    {
+      "epoch": 2.339517443116006,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.0748333862265947e-05,
+      "loss": 1.1249,
+      "step": 7448
+    },
+    {
+      "epoch": 2.3401456701415473,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.074579498571882e-05,
+      "loss": 0.98,
+      "step": 7450
+    },
+    {
+      "epoch": 2.3407738971670886,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.0743256109171692e-05,
+      "loss": 1.1628,
+      "step": 7452
+    },
+    {
+      "epoch": 2.34140212419263,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.0740717232624565e-05,
+      "loss": 1.067,
+      "step": 7454
+    },
+    {
+      "epoch": 2.3420303512181713,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0738178356077435e-05,
+      "loss": 1.1918,
+      "step": 7456
+    },
+    {
+      "epoch": 2.3426585782437126,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.073563947953031e-05,
+      "loss": 1.2056,
+      "step": 7458
+    },
+    {
+      "epoch": 2.3432868052692544,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0733100602983183e-05,
+      "loss": 1.1389,
+      "step": 7460
+    },
+    {
+      "epoch": 2.3439150322947957,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0730561726436052e-05,
+      "loss": 1.1479,
+      "step": 7462
+    },
+    {
+      "epoch": 2.344543259320337,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.0728022849888925e-05,
+      "loss": 1.119,
+      "step": 7464
+    },
+    {
+      "epoch": 2.3451714863458784,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0725483973341797e-05,
+      "loss": 1.0562,
+      "step": 7466
+    },
+    {
+      "epoch": 2.3457997133714197,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.072294509679467e-05,
+      "loss": 1.097,
+      "step": 7468
+    },
+    {
+      "epoch": 2.346427940396961,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0720406220247541e-05,
+      "loss": 1.2363,
+      "step": 7470
+    },
+    {
+      "epoch": 2.3470561674225023,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0717867343700414e-05,
+      "loss": 1.168,
+      "step": 7472
+    },
+    {
+      "epoch": 2.3476843944480437,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0715328467153286e-05,
+      "loss": 1.0952,
+      "step": 7474
+    },
+    {
+      "epoch": 2.348312621473585,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0712789590606159e-05,
+      "loss": 1.2332,
+      "step": 7476
+    },
+    {
+      "epoch": 2.3489408484991263,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.071025071405903e-05,
+      "loss": 1.1856,
+      "step": 7478
+    },
+    {
+      "epoch": 2.3495690755246676,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0707711837511903e-05,
+      "loss": 1.2385,
+      "step": 7480
+    },
+    {
+      "epoch": 2.350197302550209,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.0705172960964773e-05,
+      "loss": 1.2637,
+      "step": 7482
+    },
+    {
+      "epoch": 2.3508255295757503,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.0702634084417648e-05,
+      "loss": 1.0995,
+      "step": 7484
+    },
+    {
+      "epoch": 2.3514537566012916,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.0700095207870517e-05,
+      "loss": 1.2033,
+      "step": 7486
+    },
+    {
+      "epoch": 2.352081983626833,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.069755633132339e-05,
+      "loss": 1.2124,
+      "step": 7488
+    },
+    {
+      "epoch": 2.3527102106523747,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.0695017454776262e-05,
+      "loss": 1.2302,
+      "step": 7490
+    },
+    {
+      "epoch": 2.353338437677916,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.0692478578229135e-05,
+      "loss": 1.2885,
+      "step": 7492
+    },
+    {
+      "epoch": 2.3539666647034574,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.0689939701682006e-05,
+      "loss": 1.1231,
+      "step": 7494
+    },
+    {
+      "epoch": 2.3545948917289987,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.068740082513488e-05,
+      "loss": 1.1761,
+      "step": 7496
+    },
+    {
+      "epoch": 2.35522311875454,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.068486194858775e-05,
+      "loss": 1.2641,
+      "step": 7498
+    },
+    {
+      "epoch": 2.3558513457800814,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.0682323072040624e-05,
+      "loss": 1.1151,
+      "step": 7500
+    },
+    {
+      "epoch": 2.3564795728056227,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0679784195493495e-05,
+      "loss": 1.2444,
+      "step": 7502
+    },
+    {
+      "epoch": 2.357107799831164,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0677245318946368e-05,
+      "loss": 1.272,
+      "step": 7504
+    },
+    {
+      "epoch": 2.3577360268567054,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0674706442399238e-05,
+      "loss": 1.1819,
+      "step": 7506
+    },
+    {
+      "epoch": 2.3583642538822467,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0672167565852111e-05,
+      "loss": 1.2951,
+      "step": 7508
+    },
+    {
+      "epoch": 2.358992480907788,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0669628689304982e-05,
+      "loss": 1.1947,
+      "step": 7510
+    },
+    {
+      "epoch": 2.3596207079333293,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0667089812757855e-05,
+      "loss": 1.2004,
+      "step": 7512
+    },
+    {
+      "epoch": 2.3602489349588707,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.0664550936210727e-05,
+      "loss": 1.2573,
+      "step": 7514
+    },
+    {
+      "epoch": 2.360877161984412,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.06620120596636e-05,
+      "loss": 1.1788,
+      "step": 7516
+    },
+    {
+      "epoch": 2.3615053890099533,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0659473183116471e-05,
+      "loss": 1.3298,
+      "step": 7518
+    },
+    {
+      "epoch": 2.3621336160354947,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0656934306569344e-05,
+      "loss": 1.2565,
+      "step": 7520
+    },
+    {
+      "epoch": 2.362761843061036,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0654395430022216e-05,
+      "loss": 1.3338,
+      "step": 7522
+    },
+    {
+      "epoch": 2.3633900700865773,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.0651856553475089e-05,
+      "loss": 1.12,
+      "step": 7524
+    },
+    {
+      "epoch": 2.364018297112119,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.064931767692796e-05,
+      "loss": 0.9746,
+      "step": 7526
+    },
+    {
+      "epoch": 2.3646465241376604,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0646778800380833e-05,
+      "loss": 1.2273,
+      "step": 7528
+    },
+    {
+      "epoch": 2.3652747511632017,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0644239923833703e-05,
+      "loss": 1.2064,
+      "step": 7530
+    },
+    {
+      "epoch": 2.365902978188743,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0641701047286576e-05,
+      "loss": 1.118,
+      "step": 7532
+    },
+    {
+      "epoch": 2.3665312052142844,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.0639162170739447e-05,
+      "loss": 1.1464,
+      "step": 7534
+    },
+    {
+      "epoch": 2.3671594322398257,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.063662329419232e-05,
+      "loss": 1.2686,
+      "step": 7536
+    },
+    {
+      "epoch": 2.367787659265367,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.0634084417645192e-05,
+      "loss": 1.0966,
+      "step": 7538
+    },
+    {
+      "epoch": 2.3684158862909084,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0631545541098065e-05,
+      "loss": 1.0444,
+      "step": 7540
+    },
+    {
+      "epoch": 2.3690441133164497,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.0629006664550936e-05,
+      "loss": 1.1434,
+      "step": 7542
+    },
+    {
+      "epoch": 2.369672340341991,
+      "grad_norm": 0.875,
+      "learning_rate": 1.062646778800381e-05,
+      "loss": 1.2808,
+      "step": 7544
+    },
+    {
+      "epoch": 2.3703005673675324,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.0623928911456682e-05,
+      "loss": 1.207,
+      "step": 7546
+    },
+    {
+      "epoch": 2.3709287943930737,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0621390034909554e-05,
+      "loss": 1.2094,
+      "step": 7548
+    },
+    {
+      "epoch": 2.371557021418615,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0618851158362427e-05,
+      "loss": 1.1322,
+      "step": 7550
+    },
+    {
+      "epoch": 2.3721852484441563,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0616312281815298e-05,
+      "loss": 1.1648,
+      "step": 7552
+    },
+    {
+      "epoch": 2.3728134754696977,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0613773405268171e-05,
+      "loss": 1.1761,
+      "step": 7554
+    },
+    {
+      "epoch": 2.3734417024952394,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0611234528721041e-05,
+      "loss": 1.1397,
+      "step": 7556
+    },
+    {
+      "epoch": 2.3740699295207808,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0608695652173914e-05,
+      "loss": 1.0505,
+      "step": 7558
+    },
+    {
+      "epoch": 2.374698156546322,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0606156775626786e-05,
+      "loss": 1.3071,
+      "step": 7560
+    },
+    {
+      "epoch": 2.3753263835718634,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0603617899079659e-05,
+      "loss": 1.1697,
+      "step": 7562
+    },
+    {
+      "epoch": 2.3759546105974048,
+      "grad_norm": 0.875,
+      "learning_rate": 1.060107902253253e-05,
+      "loss": 1.3307,
+      "step": 7564
+    },
+    {
+      "epoch": 2.376582837622946,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.0598540145985403e-05,
+      "loss": 1.1086,
+      "step": 7566
+    },
+    {
+      "epoch": 2.3772110646484874,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0596001269438274e-05,
+      "loss": 1.3303,
+      "step": 7568
+    },
+    {
+      "epoch": 2.3778392916740287,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0593462392891148e-05,
+      "loss": 1.176,
+      "step": 7570
+    },
+    {
+      "epoch": 2.37846751869957,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0590923516344019e-05,
+      "loss": 1.3351,
+      "step": 7572
+    },
+    {
+      "epoch": 2.3790957457251114,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0588384639796892e-05,
+      "loss": 1.1274,
+      "step": 7574
+    },
+    {
+      "epoch": 2.3797239727506527,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0585845763249762e-05,
+      "loss": 1.0805,
+      "step": 7576
+    },
+    {
+      "epoch": 2.380352199776194,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.0583306886702636e-05,
+      "loss": 1.1506,
+      "step": 7578
+    },
+    {
+      "epoch": 2.3809804268017354,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0580768010155506e-05,
+      "loss": 1.1659,
+      "step": 7580
+    },
+    {
+      "epoch": 2.3816086538272767,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.057822913360838e-05,
+      "loss": 1.2277,
+      "step": 7582
+    },
+    {
+      "epoch": 2.382236880852818,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.057569025706125e-05,
+      "loss": 1.1857,
+      "step": 7584
+    },
+    {
+      "epoch": 2.3828651078783594,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0573151380514124e-05,
+      "loss": 1.1172,
+      "step": 7586
+    },
+    {
+      "epoch": 2.3834933349039007,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0570612503966995e-05,
+      "loss": 1.1305,
+      "step": 7588
+    },
+    {
+      "epoch": 2.384121561929442,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0568073627419868e-05,
+      "loss": 1.2697,
+      "step": 7590
+    },
+    {
+      "epoch": 2.384749788954984,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.056553475087274e-05,
+      "loss": 1.2452,
+      "step": 7592
+    },
+    {
+      "epoch": 2.385378015980525,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0562995874325613e-05,
+      "loss": 1.2094,
+      "step": 7594
+    },
+    {
+      "epoch": 2.3860062430060665,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0560456997778484e-05,
+      "loss": 1.1119,
+      "step": 7596
+    },
+    {
+      "epoch": 2.386634470031608,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.0557918121231357e-05,
+      "loss": 1.2432,
+      "step": 7598
+    },
+    {
+      "epoch": 2.387262697057149,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.0555379244684227e-05,
+      "loss": 1.35,
+      "step": 7600
+    },
+    {
+      "epoch": 2.3878909240826904,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.05528403681371e-05,
+      "loss": 1.2152,
+      "step": 7602
+    },
+    {
+      "epoch": 2.3885191511082318,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0550301491589971e-05,
+      "loss": 1.0723,
+      "step": 7604
+    },
+    {
+      "epoch": 2.389147378133773,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0547762615042844e-05,
+      "loss": 1.2081,
+      "step": 7606
+    },
+    {
+      "epoch": 2.3897756051593144,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0545223738495716e-05,
+      "loss": 1.1111,
+      "step": 7608
+    },
+    {
+      "epoch": 2.3904038321848557,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0542684861948589e-05,
+      "loss": 1.2082,
+      "step": 7610
+    },
+    {
+      "epoch": 2.391032059210397,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.054014598540146e-05,
+      "loss": 1.2288,
+      "step": 7612
+    },
+    {
+      "epoch": 2.3916602862359384,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0537607108854333e-05,
+      "loss": 1.181,
+      "step": 7614
+    },
+    {
+      "epoch": 2.3922885132614797,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0535068232307205e-05,
+      "loss": 1.1182,
+      "step": 7616
+    },
+    {
+      "epoch": 2.392916740287021,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0532529355760078e-05,
+      "loss": 1.2174,
+      "step": 7618
+    },
+    {
+      "epoch": 2.3935449673125624,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0529990479212947e-05,
+      "loss": 1.1535,
+      "step": 7620
+    },
+    {
+      "epoch": 2.394173194338104,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0527451602665822e-05,
+      "loss": 1.1007,
+      "step": 7622
+    },
+    {
+      "epoch": 2.3948014213636455,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.0524912726118692e-05,
+      "loss": 1.2071,
+      "step": 7624
+    },
+    {
+      "epoch": 2.395429648389187,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0522373849571565e-05,
+      "loss": 1.2248,
+      "step": 7626
+    },
+    {
+      "epoch": 2.396057875414728,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0519834973024436e-05,
+      "loss": 1.1908,
+      "step": 7628
+    },
+    {
+      "epoch": 2.3966861024402695,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.051729609647731e-05,
+      "loss": 1.2519,
+      "step": 7630
+    },
+    {
+      "epoch": 2.397314329465811,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0514757219930182e-05,
+      "loss": 1.1041,
+      "step": 7632
+    },
+    {
+      "epoch": 2.397942556491352,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0512218343383054e-05,
+      "loss": 1.1484,
+      "step": 7634
+    },
+    {
+      "epoch": 2.3985707835168935,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0509679466835927e-05,
+      "loss": 1.165,
+      "step": 7636
+    },
+    {
+      "epoch": 2.399199010542435,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0507140590288798e-05,
+      "loss": 1.2547,
+      "step": 7638
+    },
+    {
+      "epoch": 2.399827237567976,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.0504601713741671e-05,
+      "loss": 1.3175,
+      "step": 7640
+    },
+    {
+      "epoch": 2.4004554645935174,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.0502062837194543e-05,
+      "loss": 1.1751,
+      "step": 7642
+    },
+    {
+      "epoch": 2.4010836916190588,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0499523960647416e-05,
+      "loss": 1.1984,
+      "step": 7644
+    },
+    {
+      "epoch": 2.4017119186446,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0496985084100285e-05,
+      "loss": 1.036,
+      "step": 7646
+    },
+    {
+      "epoch": 2.4023401456701414,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.049444620755316e-05,
+      "loss": 1.3037,
+      "step": 7648
+    },
+    {
+      "epoch": 2.4029683726956828,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.049190733100603e-05,
+      "loss": 1.2453,
+      "step": 7650
+    },
+    {
+      "epoch": 2.403596599721224,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0489368454458903e-05,
+      "loss": 1.1922,
+      "step": 7652
+    },
+    {
+      "epoch": 2.4042248267467654,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0486829577911774e-05,
+      "loss": 1.282,
+      "step": 7654
+    },
+    {
+      "epoch": 2.4048530537723067,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0484290701364647e-05,
+      "loss": 1.1515,
+      "step": 7656
+    },
+    {
+      "epoch": 2.4054812807978485,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0481751824817519e-05,
+      "loss": 1.1316,
+      "step": 7658
+    },
+    {
+      "epoch": 2.40610950782339,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0479212948270392e-05,
+      "loss": 1.1016,
+      "step": 7660
+    },
+    {
+      "epoch": 2.406737734848931,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0476674071723263e-05,
+      "loss": 1.2394,
+      "step": 7662
+    },
+    {
+      "epoch": 2.4073659618744725,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.0474135195176136e-05,
+      "loss": 1.2902,
+      "step": 7664
+    },
+    {
+      "epoch": 2.407994188900014,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0471596318629008e-05,
+      "loss": 1.1772,
+      "step": 7666
+    },
+    {
+      "epoch": 2.408622415925555,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.046905744208188e-05,
+      "loss": 1.1476,
+      "step": 7668
+    },
+    {
+      "epoch": 2.4092506429510965,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.046651856553475e-05,
+      "loss": 1.1821,
+      "step": 7670
+    },
+    {
+      "epoch": 2.409878869976638,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0463979688987624e-05,
+      "loss": 1.1742,
+      "step": 7672
+    },
+    {
+      "epoch": 2.410507097002179,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0461440812440495e-05,
+      "loss": 1.2042,
+      "step": 7674
+    },
+    {
+      "epoch": 2.4111353240277205,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0458901935893368e-05,
+      "loss": 1.1789,
+      "step": 7676
+    },
+    {
+      "epoch": 2.411763551053262,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.045636305934624e-05,
+      "loss": 1.297,
+      "step": 7678
+    },
+    {
+      "epoch": 2.412391778078803,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0453824182799113e-05,
+      "loss": 1.3176,
+      "step": 7680
+    },
+    {
+      "epoch": 2.4130200051043444,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.0451285306251984e-05,
+      "loss": 1.2168,
+      "step": 7682
+    },
+    {
+      "epoch": 2.4136482321298858,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0448746429704857e-05,
+      "loss": 1.1846,
+      "step": 7684
+    },
+    {
+      "epoch": 2.414276459155427,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.0446207553157728e-05,
+      "loss": 1.0846,
+      "step": 7686
+    },
+    {
+      "epoch": 2.414904686180969,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0443668676610601e-05,
+      "loss": 1.4388,
+      "step": 7688
+    },
+    {
+      "epoch": 2.41553291320651,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.0441129800063473e-05,
+      "loss": 1.253,
+      "step": 7690
+    },
+    {
+      "epoch": 2.4161611402320515,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0438590923516346e-05,
+      "loss": 1.1133,
+      "step": 7692
+    },
+    {
+      "epoch": 2.416789367257593,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.0436052046969216e-05,
+      "loss": 1.1624,
+      "step": 7694
+    },
+    {
+      "epoch": 2.417417594283134,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0433513170422089e-05,
+      "loss": 1.256,
+      "step": 7696
+    },
+    {
+      "epoch": 2.4180458213086755,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.043097429387496e-05,
+      "loss": 1.1515,
+      "step": 7698
+    },
+    {
+      "epoch": 2.418674048334217,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0428435417327833e-05,
+      "loss": 1.2425,
+      "step": 7700
+    },
+    {
+      "epoch": 2.419302275359758,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0425896540780704e-05,
+      "loss": 1.1807,
+      "step": 7702
+    },
+    {
+      "epoch": 2.4199305023852995,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0423357664233578e-05,
+      "loss": 1.1746,
+      "step": 7704
+    },
+    {
+      "epoch": 2.420558729410841,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.0420818787686449e-05,
+      "loss": 1.2989,
+      "step": 7706
+    },
+    {
+      "epoch": 2.421186956436382,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0418279911139322e-05,
+      "loss": 1.1241,
+      "step": 7708
+    },
+    {
+      "epoch": 2.4218151834619235,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.0415741034592193e-05,
+      "loss": 1.1698,
+      "step": 7710
+    },
+    {
+      "epoch": 2.422443410487465,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0413202158045066e-05,
+      "loss": 1.2301,
+      "step": 7712
+    },
+    {
+      "epoch": 2.423071637513006,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0410663281497936e-05,
+      "loss": 1.2581,
+      "step": 7714
+    },
+    {
+      "epoch": 2.4236998645385475,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0408124404950811e-05,
+      "loss": 1.2973,
+      "step": 7716
+    },
+    {
+      "epoch": 2.424328091564089,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0405585528403684e-05,
+      "loss": 1.2082,
+      "step": 7718
+    },
+    {
+      "epoch": 2.42495631858963,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.0403046651856554e-05,
+      "loss": 1.1844,
+      "step": 7720
+    },
+    {
+      "epoch": 2.4255845456151715,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.0400507775309427e-05,
+      "loss": 1.2662,
+      "step": 7722
+    },
+    {
+      "epoch": 2.4262127726407132,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.0397968898762298e-05,
+      "loss": 1.2522,
+      "step": 7724
+    },
+    {
+      "epoch": 2.4268409996662546,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0395430022215171e-05,
+      "loss": 1.1265,
+      "step": 7726
+    },
+    {
+      "epoch": 2.427469226691796,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.0392891145668043e-05,
+      "loss": 1.1795,
+      "step": 7728
+    },
+    {
+      "epoch": 2.428097453717337,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0390352269120916e-05,
+      "loss": 1.1316,
+      "step": 7730
+    },
+    {
+      "epoch": 2.4287256807428785,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.0387813392573787e-05,
+      "loss": 1.1675,
+      "step": 7732
+    },
+    {
+      "epoch": 2.42935390776842,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.038527451602666e-05,
+      "loss": 1.1802,
+      "step": 7734
+    },
+    {
+      "epoch": 2.429982134793961,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0382735639479532e-05,
+      "loss": 1.216,
+      "step": 7736
+    },
+    {
+      "epoch": 2.4306103618195025,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0380196762932405e-05,
+      "loss": 1.059,
+      "step": 7738
+    },
+    {
+      "epoch": 2.431238588845044,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0377657886385274e-05,
+      "loss": 1.31,
+      "step": 7740
+    },
+    {
+      "epoch": 2.431866815870585,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.0375119009838149e-05,
+      "loss": 1.1443,
+      "step": 7742
+    },
+    {
+      "epoch": 2.4324950428961265,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0372580133291019e-05,
+      "loss": 1.0858,
+      "step": 7744
+    },
+    {
+      "epoch": 2.433123269921668,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0370041256743892e-05,
+      "loss": 1.0358,
+      "step": 7746
+    },
+    {
+      "epoch": 2.433751496947209,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.0367502380196763e-05,
+      "loss": 1.2427,
+      "step": 7748
+    },
+    {
+      "epoch": 2.4343797239727505,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0364963503649636e-05,
+      "loss": 1.265,
+      "step": 7750
+    },
+    {
+      "epoch": 2.435007950998292,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.0362424627102508e-05,
+      "loss": 1.2406,
+      "step": 7752
+    },
+    {
+      "epoch": 2.4356361780238336,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.035988575055538e-05,
+      "loss": 1.2867,
+      "step": 7754
+    },
+    {
+      "epoch": 2.436264405049375,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0357346874008252e-05,
+      "loss": 1.1434,
+      "step": 7756
+    },
+    {
+      "epoch": 2.4368926320749162,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0354807997461125e-05,
+      "loss": 1.1873,
+      "step": 7758
+    },
+    {
+      "epoch": 2.4375208591004576,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0352269120913997e-05,
+      "loss": 1.1372,
+      "step": 7760
+    },
+    {
+      "epoch": 2.438149086125999,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.034973024436687e-05,
+      "loss": 1.0095,
+      "step": 7762
+    },
+    {
+      "epoch": 2.4387773131515402,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.034719136781974e-05,
+      "loss": 1.2339,
+      "step": 7764
+    },
+    {
+      "epoch": 2.4394055401770816,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0344652491272612e-05,
+      "loss": 1.1116,
+      "step": 7766
+    },
+    {
+      "epoch": 2.440033767202623,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0342113614725484e-05,
+      "loss": 1.3111,
+      "step": 7768
+    },
+    {
+      "epoch": 2.440661994228164,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0339574738178357e-05,
+      "loss": 1.119,
+      "step": 7770
+    },
+    {
+      "epoch": 2.4412902212537055,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0337035861631228e-05,
+      "loss": 1.1471,
+      "step": 7772
+    },
+    {
+      "epoch": 2.441918448279247,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.0334496985084101e-05,
+      "loss": 1.1424,
+      "step": 7774
+    },
+    {
+      "epoch": 2.442546675304788,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0331958108536973e-05,
+      "loss": 1.2449,
+      "step": 7776
+    },
+    {
+      "epoch": 2.4431749023303295,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0329419231989846e-05,
+      "loss": 1.1482,
+      "step": 7778
+    },
+    {
+      "epoch": 2.443803129355871,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0326880355442717e-05,
+      "loss": 1.3518,
+      "step": 7780
+    },
+    {
+      "epoch": 2.444431356381412,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.032434147889559e-05,
+      "loss": 1.2404,
+      "step": 7782
+    },
+    {
+      "epoch": 2.4450595834069535,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.032180260234846e-05,
+      "loss": 1.1999,
+      "step": 7784
+    },
+    {
+      "epoch": 2.445687810432495,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.0319263725801335e-05,
+      "loss": 1.1479,
+      "step": 7786
+    },
+    {
+      "epoch": 2.446316037458036,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0316724849254204e-05,
+      "loss": 1.164,
+      "step": 7788
+    },
+    {
+      "epoch": 2.446944264483578,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0314185972707077e-05,
+      "loss": 1.2489,
+      "step": 7790
+    },
+    {
+      "epoch": 2.4475724915091193,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0311647096159949e-05,
+      "loss": 1.0448,
+      "step": 7792
+    },
+    {
+      "epoch": 2.4482007185346606,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.0309108219612822e-05,
+      "loss": 1.1342,
+      "step": 7794
+    },
+    {
+      "epoch": 2.448828945560202,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.0306569343065693e-05,
+      "loss": 1.1211,
+      "step": 7796
+    },
+    {
+      "epoch": 2.4494571725857432,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0304030466518566e-05,
+      "loss": 1.1104,
+      "step": 7798
+    },
+    {
+      "epoch": 2.4500853996112846,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.0301491589971438e-05,
+      "loss": 1.0956,
+      "step": 7800
+    },
+    {
+      "epoch": 2.450713626636826,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0298952713424311e-05,
+      "loss": 1.2627,
+      "step": 7802
+    },
+    {
+      "epoch": 2.4513418536623672,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.0296413836877184e-05,
+      "loss": 1.1842,
+      "step": 7804
+    },
+    {
+      "epoch": 2.4519700806879086,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0293874960330055e-05,
+      "loss": 1.2095,
+      "step": 7806
+    },
+    {
+      "epoch": 2.45259830771345,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0291336083782928e-05,
+      "loss": 1.2931,
+      "step": 7808
+    },
+    {
+      "epoch": 2.453226534738991,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0288797207235798e-05,
+      "loss": 1.1803,
+      "step": 7810
+    },
+    {
+      "epoch": 2.4538547617645325,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0286258330688673e-05,
+      "loss": 1.1717,
+      "step": 7812
+    },
+    {
+      "epoch": 2.454482988790074,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0283719454141543e-05,
+      "loss": 0.999,
+      "step": 7814
+    },
+    {
+      "epoch": 2.455111215815615,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.0281180577594416e-05,
+      "loss": 1.103,
+      "step": 7816
+    },
+    {
+      "epoch": 2.455739442841157,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0278641701047287e-05,
+      "loss": 1.2353,
+      "step": 7818
+    },
+    {
+      "epoch": 2.4563676698666983,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.027610282450016e-05,
+      "loss": 1.1041,
+      "step": 7820
+    },
+    {
+      "epoch": 2.4569958968922396,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0273563947953031e-05,
+      "loss": 1.2442,
+      "step": 7822
+    },
+    {
+      "epoch": 2.457624123917781,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0271025071405905e-05,
+      "loss": 1.1837,
+      "step": 7824
+    },
+    {
+      "epoch": 2.4582523509433223,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.0268486194858776e-05,
+      "loss": 1.0694,
+      "step": 7826
+    },
+    {
+      "epoch": 2.4588805779688636,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0265947318311649e-05,
+      "loss": 1.1621,
+      "step": 7828
+    },
+    {
+      "epoch": 2.459508804994405,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.026340844176452e-05,
+      "loss": 1.2203,
+      "step": 7830
+    },
+    {
+      "epoch": 2.4601370320199463,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0260869565217393e-05,
+      "loss": 1.2414,
+      "step": 7832
+    },
+    {
+      "epoch": 2.4607652590454876,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0258330688670263e-05,
+      "loss": 1.298,
+      "step": 7834
+    },
+    {
+      "epoch": 2.461393486071029,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.0255791812123136e-05,
+      "loss": 1.0458,
+      "step": 7836
+    },
+    {
+      "epoch": 2.4620217130965703,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.0253252935576008e-05,
+      "loss": 1.2697,
+      "step": 7838
+    },
+    {
+      "epoch": 2.4626499401221116,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.025071405902888e-05,
+      "loss": 1.1559,
+      "step": 7840
+    },
+    {
+      "epoch": 2.463278167147653,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0248175182481752e-05,
+      "loss": 1.1766,
+      "step": 7842
+    },
+    {
+      "epoch": 2.4639063941731942,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0245636305934625e-05,
+      "loss": 1.1575,
+      "step": 7844
+    },
+    {
+      "epoch": 2.4645346211987356,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0243097429387497e-05,
+      "loss": 1.1253,
+      "step": 7846
+    },
+    {
+      "epoch": 2.465162848224277,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.024055855284037e-05,
+      "loss": 1.1804,
+      "step": 7848
+    },
+    {
+      "epoch": 2.465791075249818,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0238019676293241e-05,
+      "loss": 1.2267,
+      "step": 7850
+    },
+    {
+      "epoch": 2.4664193022753595,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.0235480799746114e-05,
+      "loss": 1.2627,
+      "step": 7852
+    },
+    {
+      "epoch": 2.467047529300901,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0232941923198985e-05,
+      "loss": 1.2447,
+      "step": 7854
+    },
+    {
+      "epoch": 2.4676757563264426,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0230403046651859e-05,
+      "loss": 1.1281,
+      "step": 7856
+    },
+    {
+      "epoch": 2.468303983351984,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0227864170104728e-05,
+      "loss": 1.1362,
+      "step": 7858
+    },
+    {
+      "epoch": 2.4689322103775253,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0225325293557601e-05,
+      "loss": 1.1659,
+      "step": 7860
+    },
+    {
+      "epoch": 2.4695604374030666,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0222786417010473e-05,
+      "loss": 1.1334,
+      "step": 7862
+    },
+    {
+      "epoch": 2.470188664428608,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.0220247540463346e-05,
+      "loss": 1.2789,
+      "step": 7864
+    },
+    {
+      "epoch": 2.4708168914541493,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0217708663916217e-05,
+      "loss": 1.2958,
+      "step": 7866
+    },
+    {
+      "epoch": 2.4714451184796906,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.021516978736909e-05,
+      "loss": 1.1327,
+      "step": 7868
+    },
+    {
+      "epoch": 2.472073345505232,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0212630910821962e-05,
+      "loss": 1.2687,
+      "step": 7870
+    },
+    {
+      "epoch": 2.4727015725307733,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.0210092034274835e-05,
+      "loss": 1.2305,
+      "step": 7872
+    },
+    {
+      "epoch": 2.4733297995563146,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0207553157727706e-05,
+      "loss": 1.3892,
+      "step": 7874
+    },
+    {
+      "epoch": 2.473958026581856,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0205014281180579e-05,
+      "loss": 1.1705,
+      "step": 7876
+    },
+    {
+      "epoch": 2.4745862536073973,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0202475404633449e-05,
+      "loss": 1.148,
+      "step": 7878
+    },
+    {
+      "epoch": 2.4752144806329386,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.0199936528086324e-05,
+      "loss": 1.3282,
+      "step": 7880
+    },
+    {
+      "epoch": 2.47584270765848,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0197397651539193e-05,
+      "loss": 1.1612,
+      "step": 7882
+    },
+    {
+      "epoch": 2.4764709346840217,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.0194858774992066e-05,
+      "loss": 1.2072,
+      "step": 7884
+    },
+    {
+      "epoch": 2.477099161709563,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0192319898444938e-05,
+      "loss": 1.3195,
+      "step": 7886
+    },
+    {
+      "epoch": 2.4777273887351043,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.018978102189781e-05,
+      "loss": 1.1765,
+      "step": 7888
+    },
+    {
+      "epoch": 2.4783556157606457,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0187242145350684e-05,
+      "loss": 1.1609,
+      "step": 7890
+    },
+    {
+      "epoch": 2.478983842786187,
+      "grad_norm": 1.0,
+      "learning_rate": 1.0184703268803555e-05,
+      "loss": 1.1592,
+      "step": 7892
+    },
+    {
+      "epoch": 2.4796120698117283,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0182164392256428e-05,
+      "loss": 1.1972,
+      "step": 7894
+    },
+    {
+      "epoch": 2.4802402968372697,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.01796255157093e-05,
+      "loss": 1.2579,
+      "step": 7896
+    },
+    {
+      "epoch": 2.480868523862811,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0177086639162173e-05,
+      "loss": 1.1676,
+      "step": 7898
+    },
+    {
+      "epoch": 2.4814967508883523,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0174547762615044e-05,
+      "loss": 1.299,
+      "step": 7900
+    },
+    {
+      "epoch": 2.4821249779138936,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0172008886067917e-05,
+      "loss": 1.2395,
+      "step": 7902
+    },
+    {
+      "epoch": 2.482753204939435,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0169470009520787e-05,
+      "loss": 1.2603,
+      "step": 7904
+    },
+    {
+      "epoch": 2.4833814319649763,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0166931132973662e-05,
+      "loss": 1.129,
+      "step": 7906
+    },
+    {
+      "epoch": 2.4840096589905176,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0164392256426531e-05,
+      "loss": 1.3505,
+      "step": 7908
+    },
+    {
+      "epoch": 2.484637886016059,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.0161853379879404e-05,
+      "loss": 1.1911,
+      "step": 7910
+    },
+    {
+      "epoch": 2.4852661130416003,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0159314503332276e-05,
+      "loss": 1.2497,
+      "step": 7912
+    },
+    {
+      "epoch": 2.4858943400671416,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0156775626785149e-05,
+      "loss": 1.2195,
+      "step": 7914
+    },
+    {
+      "epoch": 2.486522567092683,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.015423675023802e-05,
+      "loss": 1.2527,
+      "step": 7916
+    },
+    {
+      "epoch": 2.4871507941182243,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.0151697873690893e-05,
+      "loss": 1.2442,
+      "step": 7918
+    },
+    {
+      "epoch": 2.4877790211437656,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0149158997143765e-05,
+      "loss": 1.2383,
+      "step": 7920
+    },
+    {
+      "epoch": 2.4884072481693074,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0146620120596638e-05,
+      "loss": 1.1643,
+      "step": 7922
+    },
+    {
+      "epoch": 2.4890354751948487,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.014408124404951e-05,
+      "loss": 1.1913,
+      "step": 7924
+    },
+    {
+      "epoch": 2.48966370222039,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0141542367502382e-05,
+      "loss": 1.1843,
+      "step": 7926
+    },
+    {
+      "epoch": 2.4902919292459313,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.0139003490955252e-05,
+      "loss": 1.1378,
+      "step": 7928
+    },
+    {
+      "epoch": 2.4909201562714727,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.0136464614408125e-05,
+      "loss": 1.1175,
+      "step": 7930
+    },
+    {
+      "epoch": 2.491548383297014,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0133925737860996e-05,
+      "loss": 1.2275,
+      "step": 7932
+    },
+    {
+      "epoch": 2.4921766103225553,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.013138686131387e-05,
+      "loss": 1.0809,
+      "step": 7934
+    },
+    {
+      "epoch": 2.4928048373480967,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0128847984766741e-05,
+      "loss": 1.223,
+      "step": 7936
+    },
+    {
+      "epoch": 2.493433064373638,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0126309108219614e-05,
+      "loss": 1.2318,
+      "step": 7938
+    },
+    {
+      "epoch": 2.4940612913991793,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0123770231672485e-05,
+      "loss": 1.1458,
+      "step": 7940
+    },
+    {
+      "epoch": 2.4946895184247206,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.0121231355125358e-05,
+      "loss": 1.1603,
+      "step": 7942
+    },
+    {
+      "epoch": 2.495317745450262,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.011869247857823e-05,
+      "loss": 1.1126,
+      "step": 7944
+    },
+    {
+      "epoch": 2.4959459724758033,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0116153602031103e-05,
+      "loss": 1.2311,
+      "step": 7946
+    },
+    {
+      "epoch": 2.4965741995013446,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.0113614725483973e-05,
+      "loss": 1.2241,
+      "step": 7948
+    },
+    {
+      "epoch": 2.4972024265268864,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0111075848936847e-05,
+      "loss": 1.1663,
+      "step": 7950
+    },
+    {
+      "epoch": 2.4978306535524277,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0108536972389717e-05,
+      "loss": 1.1048,
+      "step": 7952
+    },
+    {
+      "epoch": 2.498458880577969,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.010599809584259e-05,
+      "loss": 1.1416,
+      "step": 7954
+    },
+    {
+      "epoch": 2.4990871076035104,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.0103459219295462e-05,
+      "loss": 1.2229,
+      "step": 7956
+    },
+    {
+      "epoch": 2.4997153346290517,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0100920342748335e-05,
+      "loss": 1.0936,
+      "step": 7958
+    },
+    {
+      "epoch": 2.500343561654593,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0098381466201206e-05,
+      "loss": 1.2353,
+      "step": 7960
+    },
+    {
+      "epoch": 2.5009717886801344,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0095842589654079e-05,
+      "loss": 1.1868,
+      "step": 7962
+    },
+    {
+      "epoch": 2.5016000157056757,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.009330371310695e-05,
+      "loss": 1.1775,
+      "step": 7964
+    },
+    {
+      "epoch": 2.502228242731217,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0090764836559823e-05,
+      "loss": 1.1873,
+      "step": 7966
+    },
+    {
+      "epoch": 2.5028564697567584,
+      "grad_norm": 1.0,
+      "learning_rate": 1.0088225960012695e-05,
+      "loss": 1.2451,
+      "step": 7968
+    },
+    {
+      "epoch": 2.5034846967822997,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0085687083465568e-05,
+      "loss": 1.1386,
+      "step": 7970
+    },
+    {
+      "epoch": 2.504112923807841,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0083148206918438e-05,
+      "loss": 1.2973,
+      "step": 7972
+    },
+    {
+      "epoch": 2.5047411508333823,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.008060933037131e-05,
+      "loss": 1.1667,
+      "step": 7974
+    },
+    {
+      "epoch": 2.5053693778589237,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0078070453824185e-05,
+      "loss": 1.3141,
+      "step": 7976
+    },
+    {
+      "epoch": 2.505997604884465,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0075531577277055e-05,
+      "loss": 1.1125,
+      "step": 7978
+    },
+    {
+      "epoch": 2.5066258319100063,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.0072992700729928e-05,
+      "loss": 1.2697,
+      "step": 7980
+    },
+    {
+      "epoch": 2.5072540589355476,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.00704538241828e-05,
+      "loss": 1.1035,
+      "step": 7982
+    },
+    {
+      "epoch": 2.507882285961089,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0067914947635673e-05,
+      "loss": 1.1884,
+      "step": 7984
+    },
+    {
+      "epoch": 2.5085105129866303,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.0065376071088544e-05,
+      "loss": 1.2257,
+      "step": 7986
+    },
+    {
+      "epoch": 2.5091387400121716,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0062837194541417e-05,
+      "loss": 1.2123,
+      "step": 7988
+    },
+    {
+      "epoch": 2.5097669670377134,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0060298317994289e-05,
+      "loss": 1.2626,
+      "step": 7990
+    },
+    {
+      "epoch": 2.5103951940632547,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.0057759441447162e-05,
+      "loss": 1.2488,
+      "step": 7992
+    },
+    {
+      "epoch": 2.511023421088796,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0055220564900033e-05,
+      "loss": 1.3123,
+      "step": 7994
+    },
+    {
+      "epoch": 2.5116516481143374,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0052681688352906e-05,
+      "loss": 1.204,
+      "step": 7996
+    },
+    {
+      "epoch": 2.5122798751398787,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.0050142811805776e-05,
+      "loss": 1.1576,
+      "step": 7998
+    },
+    {
+      "epoch": 2.51290810216542,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.0047603935258649e-05,
+      "loss": 1.1342,
+      "step": 8000
+    },
+    {
+      "epoch": 2.5135363291909614,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.004506505871152e-05,
+      "loss": 1.1785,
+      "step": 8002
+    },
+    {
+      "epoch": 2.5141645562165027,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0042526182164393e-05,
+      "loss": 1.1783,
+      "step": 8004
+    },
+    {
+      "epoch": 2.514792783242044,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0039987305617265e-05,
+      "loss": 1.2789,
+      "step": 8006
+    },
+    {
+      "epoch": 2.5154210102675854,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0037448429070138e-05,
+      "loss": 1.3081,
+      "step": 8008
+    },
+    {
+      "epoch": 2.5160492372931267,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.0034909552523009e-05,
+      "loss": 1.1454,
+      "step": 8010
+    },
+    {
+      "epoch": 2.516677464318668,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.0032370675975882e-05,
+      "loss": 1.1284,
+      "step": 8012
+    },
+    {
+      "epoch": 2.51730569134421,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.0029831799428754e-05,
+      "loss": 1.1488,
+      "step": 8014
+    },
+    {
+      "epoch": 2.517933918369751,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0027292922881627e-05,
+      "loss": 1.2606,
+      "step": 8016
+    },
+    {
+      "epoch": 2.5185621453952924,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0024754046334498e-05,
+      "loss": 1.2357,
+      "step": 8018
+    },
+    {
+      "epoch": 2.5191903724208338,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0022215169787371e-05,
+      "loss": 1.1624,
+      "step": 8020
+    },
+    {
+      "epoch": 2.519818599446375,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.001967629324024e-05,
+      "loss": 1.1266,
+      "step": 8022
+    },
+    {
+      "epoch": 2.5204468264719164,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.0017137416693114e-05,
+      "loss": 1.2335,
+      "step": 8024
+    },
+    {
+      "epoch": 2.5210750534974578,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0014598540145985e-05,
+      "loss": 1.1071,
+      "step": 8026
+    },
+    {
+      "epoch": 2.521703280522999,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0012059663598858e-05,
+      "loss": 1.1173,
+      "step": 8028
+    },
+    {
+      "epoch": 2.5223315075485404,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.000952078705173e-05,
+      "loss": 1.1481,
+      "step": 8030
+    },
+    {
+      "epoch": 2.5229597345740817,
+      "grad_norm": 0.875,
+      "learning_rate": 1.0006981910504603e-05,
+      "loss": 1.2871,
+      "step": 8032
+    },
+    {
+      "epoch": 2.523587961599623,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.0004443033957474e-05,
+      "loss": 1.1876,
+      "step": 8034
+    },
+    {
+      "epoch": 2.5242161886251644,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.0001904157410347e-05,
+      "loss": 1.22,
+      "step": 8036
+    },
+    {
+      "epoch": 2.5248444156507057,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.999365280863219e-06,
+      "loss": 1.0271,
+      "step": 8038
+    },
+    {
+      "epoch": 2.525472642676247,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.996826404316092e-06,
+      "loss": 1.2888,
+      "step": 8040
+    },
+    {
+      "epoch": 2.5261008697017884,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.994287527768963e-06,
+      "loss": 1.3042,
+      "step": 8042
+    },
+    {
+      "epoch": 2.5267290967273297,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.991748651221835e-06,
+      "loss": 1.2412,
+      "step": 8044
+    },
+    {
+      "epoch": 2.527357323752871,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.989209774674708e-06,
+      "loss": 1.2465,
+      "step": 8046
+    },
+    {
+      "epoch": 2.5279855507784124,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.986670898127579e-06,
+      "loss": 1.1235,
+      "step": 8048
+    },
+    {
+      "epoch": 2.5286137778039537,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.984132021580452e-06,
+      "loss": 1.1607,
+      "step": 8050
+    },
+    {
+      "epoch": 2.529242004829495,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.981593145033323e-06,
+      "loss": 1.1748,
+      "step": 8052
+    },
+    {
+      "epoch": 2.5298702318550363,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.979054268486196e-06,
+      "loss": 1.0597,
+      "step": 8054
+    },
+    {
+      "epoch": 2.530498458880578,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.976515391939068e-06,
+      "loss": 1.0744,
+      "step": 8056
+    },
+    {
+      "epoch": 2.5311266859061194,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.97397651539194e-06,
+      "loss": 1.2989,
+      "step": 8058
+    },
+    {
+      "epoch": 2.5317549129316608,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.971437638844812e-06,
+      "loss": 1.1052,
+      "step": 8060
+    },
+    {
+      "epoch": 2.532383139957202,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.968898762297684e-06,
+      "loss": 1.2326,
+      "step": 8062
+    },
+    {
+      "epoch": 2.5330113669827434,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.966359885750557e-06,
+      "loss": 1.258,
+      "step": 8064
+    },
+    {
+      "epoch": 2.5336395940082848,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.963821009203428e-06,
+      "loss": 1.3359,
+      "step": 8066
+    },
+    {
+      "epoch": 2.534267821033826,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.9612821326563e-06,
+      "loss": 1.1445,
+      "step": 8068
+    },
+    {
+      "epoch": 2.5348960480593674,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.958743256109173e-06,
+      "loss": 1.1916,
+      "step": 8070
+    },
+    {
+      "epoch": 2.5355242750849087,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.956204379562044e-06,
+      "loss": 1.2186,
+      "step": 8072
+    },
+    {
+      "epoch": 2.53615250211045,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.953665503014917e-06,
+      "loss": 1.1839,
+      "step": 8074
+    },
+    {
+      "epoch": 2.5367807291359914,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.951126626467788e-06,
+      "loss": 1.1174,
+      "step": 8076
+    },
+    {
+      "epoch": 2.5374089561615327,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.94858774992066e-06,
+      "loss": 1.3167,
+      "step": 8078
+    },
+    {
+      "epoch": 2.5380371831870745,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.946048873373533e-06,
+      "loss": 1.2239,
+      "step": 8080
+    },
+    {
+      "epoch": 2.538665410212616,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.943509996826404e-06,
+      "loss": 1.2857,
+      "step": 8082
+    },
+    {
+      "epoch": 2.539293637238157,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.940971120279277e-06,
+      "loss": 1.2533,
+      "step": 8084
+    },
+    {
+      "epoch": 2.5399218642636985,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.938432243732149e-06,
+      "loss": 1.2453,
+      "step": 8086
+    },
+    {
+      "epoch": 2.54055009128924,
+      "grad_norm": 1.4375,
+      "learning_rate": 9.935893367185022e-06,
+      "loss": 1.2452,
+      "step": 8088
+    },
+    {
+      "epoch": 2.541178318314781,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.933354490637893e-06,
+      "loss": 1.1738,
+      "step": 8090
+    },
+    {
+      "epoch": 2.5418065453403225,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.930815614090765e-06,
+      "loss": 1.2676,
+      "step": 8092
+    },
+    {
+      "epoch": 2.542434772365864,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.928276737543638e-06,
+      "loss": 1.1425,
+      "step": 8094
+    },
+    {
+      "epoch": 2.543062999391405,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.925737860996509e-06,
+      "loss": 1.134,
+      "step": 8096
+    },
+    {
+      "epoch": 2.5436912264169464,
+      "grad_norm": 0.875,
+      "learning_rate": 9.923198984449382e-06,
+      "loss": 1.1667,
+      "step": 8098
+    },
+    {
+      "epoch": 2.5443194534424878,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.920660107902254e-06,
+      "loss": 1.1229,
+      "step": 8100
+    },
+    {
+      "epoch": 2.544947680468029,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.918121231355127e-06,
+      "loss": 1.2614,
+      "step": 8102
+    },
+    {
+      "epoch": 2.5455759074935704,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.915582354807998e-06,
+      "loss": 1.2031,
+      "step": 8104
+    },
+    {
+      "epoch": 2.5462041345191118,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.913043478260871e-06,
+      "loss": 1.0952,
+      "step": 8106
+    },
+    {
+      "epoch": 2.546832361544653,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.910504601713742e-06,
+      "loss": 1.3312,
+      "step": 8108
+    },
+    {
+      "epoch": 2.5474605885701944,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.907965725166616e-06,
+      "loss": 1.1553,
+      "step": 8110
+    },
+    {
+      "epoch": 2.5480888155957357,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.905426848619487e-06,
+      "loss": 1.1358,
+      "step": 8112
+    },
+    {
+      "epoch": 2.548717042621277,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.90288797207236e-06,
+      "loss": 1.1889,
+      "step": 8114
+    },
+    {
+      "epoch": 2.5493452696468184,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.900349095525231e-06,
+      "loss": 1.3103,
+      "step": 8116
+    },
+    {
+      "epoch": 2.5499734966723597,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.897810218978103e-06,
+      "loss": 1.2243,
+      "step": 8118
+    },
+    {
+      "epoch": 2.550601723697901,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.895271342430976e-06,
+      "loss": 1.1031,
+      "step": 8120
+    },
+    {
+      "epoch": 2.551229950723443,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.892732465883847e-06,
+      "loss": 1.2186,
+      "step": 8122
+    },
+    {
+      "epoch": 2.551858177748984,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.89019358933672e-06,
+      "loss": 1.2599,
+      "step": 8124
+    },
+    {
+      "epoch": 2.5524864047745255,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.887654712789592e-06,
+      "loss": 1.2472,
+      "step": 8126
+    },
+    {
+      "epoch": 2.553114631800067,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.885115836242463e-06,
+      "loss": 1.1494,
+      "step": 8128
+    },
+    {
+      "epoch": 2.553742858825608,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.882576959695336e-06,
+      "loss": 1.0775,
+      "step": 8130
+    },
+    {
+      "epoch": 2.5543710858511495,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.880038083148208e-06,
+      "loss": 1.2127,
+      "step": 8132
+    },
+    {
+      "epoch": 2.554999312876691,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.87749920660108e-06,
+      "loss": 1.1725,
+      "step": 8134
+    },
+    {
+      "epoch": 2.555627539902232,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.874960330053952e-06,
+      "loss": 1.2109,
+      "step": 8136
+    },
+    {
+      "epoch": 2.5562557669277735,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.872421453506823e-06,
+      "loss": 1.1498,
+      "step": 8138
+    },
+    {
+      "epoch": 2.556883993953315,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.869882576959696e-06,
+      "loss": 1.2461,
+      "step": 8140
+    },
+    {
+      "epoch": 2.557512220978856,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.867343700412568e-06,
+      "loss": 1.1263,
+      "step": 8142
+    },
+    {
+      "epoch": 2.5581404480043974,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.864804823865441e-06,
+      "loss": 1.0624,
+      "step": 8144
+    },
+    {
+      "epoch": 2.558768675029939,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.862265947318312e-06,
+      "loss": 1.1112,
+      "step": 8146
+    },
+    {
+      "epoch": 2.5593969020554805,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.859727070771185e-06,
+      "loss": 1.1871,
+      "step": 8148
+    },
+    {
+      "epoch": 2.560025129081022,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.857188194224057e-06,
+      "loss": 1.2094,
+      "step": 8150
+    },
+    {
+      "epoch": 2.560653356106563,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.854649317676928e-06,
+      "loss": 1.0918,
+      "step": 8152
+    },
+    {
+      "epoch": 2.5612815831321045,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.852110441129801e-06,
+      "loss": 1.1145,
+      "step": 8154
+    },
+    {
+      "epoch": 2.561909810157646,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.849571564582673e-06,
+      "loss": 1.178,
+      "step": 8156
+    },
+    {
+      "epoch": 2.562538037183187,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.847032688035546e-06,
+      "loss": 1.2077,
+      "step": 8158
+    },
+    {
+      "epoch": 2.5631662642087285,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.844493811488417e-06,
+      "loss": 1.2177,
+      "step": 8160
+    },
+    {
+      "epoch": 2.56379449123427,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.841954934941288e-06,
+      "loss": 1.1381,
+      "step": 8162
+    },
+    {
+      "epoch": 2.564422718259811,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.839416058394161e-06,
+      "loss": 1.1781,
+      "step": 8164
+    },
+    {
+      "epoch": 2.5650509452853525,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.836877181847033e-06,
+      "loss": 1.1931,
+      "step": 8166
+    },
+    {
+      "epoch": 2.565679172310894,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.834338305299906e-06,
+      "loss": 1.203,
+      "step": 8168
+    },
+    {
+      "epoch": 2.566307399336435,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.831799428752777e-06,
+      "loss": 1.2147,
+      "step": 8170
+    },
+    {
+      "epoch": 2.5669356263619765,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.829260552205649e-06,
+      "loss": 1.2632,
+      "step": 8172
+    },
+    {
+      "epoch": 2.567563853387518,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.826721675658522e-06,
+      "loss": 1.2249,
+      "step": 8174
+    },
+    {
+      "epoch": 2.568192080413059,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.824182799111393e-06,
+      "loss": 1.1366,
+      "step": 8176
+    },
+    {
+      "epoch": 2.5688203074386005,
+      "grad_norm": 0.875,
+      "learning_rate": 9.821643922564266e-06,
+      "loss": 1.1498,
+      "step": 8178
+    },
+    {
+      "epoch": 2.569448534464142,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.819105046017138e-06,
+      "loss": 1.121,
+      "step": 8180
+    },
+    {
+      "epoch": 2.570076761489683,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.816566169470009e-06,
+      "loss": 1.1996,
+      "step": 8182
+    },
+    {
+      "epoch": 2.5707049885152244,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.814027292922882e-06,
+      "loss": 1.2784,
+      "step": 8184
+    },
+    {
+      "epoch": 2.5713332155407658,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.811488416375753e-06,
+      "loss": 1.145,
+      "step": 8186
+    },
+    {
+      "epoch": 2.5719614425663075,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.808949539828627e-06,
+      "loss": 1.1618,
+      "step": 8188
+    },
+    {
+      "epoch": 2.572589669591849,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.8064106632815e-06,
+      "loss": 1.2192,
+      "step": 8190
+    },
+    {
+      "epoch": 2.57321789661739,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.803871786734371e-06,
+      "loss": 1.166,
+      "step": 8192
+    },
+    {
+      "epoch": 2.5738461236429315,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.801332910187244e-06,
+      "loss": 1.3588,
+      "step": 8194
+    },
+    {
+      "epoch": 2.574474350668473,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.798794033640115e-06,
+      "loss": 1.0738,
+      "step": 8196
+    },
+    {
+      "epoch": 2.575102577694014,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.796255157092987e-06,
+      "loss": 1.3274,
+      "step": 8198
+    },
+    {
+      "epoch": 2.5757308047195555,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.79371628054586e-06,
+      "loss": 1.3462,
+      "step": 8200
+    },
+    {
+      "epoch": 2.576359031745097,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.791177403998731e-06,
+      "loss": 1.1931,
+      "step": 8202
+    },
+    {
+      "epoch": 2.576987258770638,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.788638527451604e-06,
+      "loss": 1.0762,
+      "step": 8204
+    },
+    {
+      "epoch": 2.5776154857961795,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.786099650904476e-06,
+      "loss": 1.0787,
+      "step": 8206
+    },
+    {
+      "epoch": 2.578243712821721,
+      "grad_norm": 0.875,
+      "learning_rate": 9.783560774357347e-06,
+      "loss": 1.1584,
+      "step": 8208
+    },
+    {
+      "epoch": 2.578871939847262,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.78102189781022e-06,
+      "loss": 1.1298,
+      "step": 8210
+    },
+    {
+      "epoch": 2.579500166872804,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.778483021263092e-06,
+      "loss": 1.1622,
+      "step": 8212
+    },
+    {
+      "epoch": 2.5801283938983453,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.775944144715965e-06,
+      "loss": 1.2961,
+      "step": 8214
+    },
+    {
+      "epoch": 2.5807566209238866,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.773405268168836e-06,
+      "loss": 1.1895,
+      "step": 8216
+    },
+    {
+      "epoch": 2.581384847949428,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.770866391621709e-06,
+      "loss": 1.1523,
+      "step": 8218
+    },
+    {
+      "epoch": 2.5820130749749692,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.76832751507458e-06,
+      "loss": 1.0955,
+      "step": 8220
+    },
+    {
+      "epoch": 2.5826413020005106,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.765788638527452e-06,
+      "loss": 1.1717,
+      "step": 8222
+    },
+    {
+      "epoch": 2.583269529026052,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.763249761980325e-06,
+      "loss": 1.1134,
+      "step": 8224
+    },
+    {
+      "epoch": 2.583897756051593,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.760710885433196e-06,
+      "loss": 1.1845,
+      "step": 8226
+    },
+    {
+      "epoch": 2.5845259830771345,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.75817200888607e-06,
+      "loss": 1.0873,
+      "step": 8228
+    },
+    {
+      "epoch": 2.585154210102676,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.75563313233894e-06,
+      "loss": 1.1864,
+      "step": 8230
+    },
+    {
+      "epoch": 2.585782437128217,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.753094255791812e-06,
+      "loss": 1.0575,
+      "step": 8232
+    },
+    {
+      "epoch": 2.5864106641537585,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.750555379244685e-06,
+      "loss": 1.2174,
+      "step": 8234
+    },
+    {
+      "epoch": 2.5870388911793,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.748016502697557e-06,
+      "loss": 1.0819,
+      "step": 8236
+    },
+    {
+      "epoch": 2.587667118204841,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.74547762615043e-06,
+      "loss": 1.3066,
+      "step": 8238
+    },
+    {
+      "epoch": 2.5882953452303825,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.742938749603301e-06,
+      "loss": 1.0422,
+      "step": 8240
+    },
+    {
+      "epoch": 2.588923572255924,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.740399873056172e-06,
+      "loss": 1.315,
+      "step": 8242
+    },
+    {
+      "epoch": 2.589551799281465,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.737860996509046e-06,
+      "loss": 1.1542,
+      "step": 8244
+    },
+    {
+      "epoch": 2.5901800263070065,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.735322119961917e-06,
+      "loss": 1.1142,
+      "step": 8246
+    },
+    {
+      "epoch": 2.590808253332548,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.73278324341479e-06,
+      "loss": 1.1555,
+      "step": 8248
+    },
+    {
+      "epoch": 2.591436480358089,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.730244366867661e-06,
+      "loss": 1.1318,
+      "step": 8250
+    },
+    {
+      "epoch": 2.5920647073836305,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.727705490320534e-06,
+      "loss": 1.1723,
+      "step": 8252
+    },
+    {
+      "epoch": 2.5926929344091723,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.725166613773406e-06,
+      "loss": 1.1697,
+      "step": 8254
+    },
+    {
+      "epoch": 2.5933211614347136,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.722627737226277e-06,
+      "loss": 1.142,
+      "step": 8256
+    },
+    {
+      "epoch": 2.593949388460255,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.72008886067915e-06,
+      "loss": 1.0324,
+      "step": 8258
+    },
+    {
+      "epoch": 2.5945776154857962,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.717549984132022e-06,
+      "loss": 1.0065,
+      "step": 8260
+    },
+    {
+      "epoch": 2.5952058425113376,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.715011107584895e-06,
+      "loss": 1.104,
+      "step": 8262
+    },
+    {
+      "epoch": 2.595834069536879,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.712472231037766e-06,
+      "loss": 1.1189,
+      "step": 8264
+    },
+    {
+      "epoch": 2.5964622965624202,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.709933354490638e-06,
+      "loss": 1.1932,
+      "step": 8266
+    },
+    {
+      "epoch": 2.5970905235879616,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.70739447794351e-06,
+      "loss": 1.2323,
+      "step": 8268
+    },
+    {
+      "epoch": 2.597718750613503,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.704855601396382e-06,
+      "loss": 1.1692,
+      "step": 8270
+    },
+    {
+      "epoch": 2.598346977639044,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.702316724849255e-06,
+      "loss": 1.2307,
+      "step": 8272
+    },
+    {
+      "epoch": 2.5989752046645855,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.699777848302128e-06,
+      "loss": 1.112,
+      "step": 8274
+    },
+    {
+      "epoch": 2.599603431690127,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.697238971755e-06,
+      "loss": 1.3361,
+      "step": 8276
+    },
+    {
+      "epoch": 2.6002316587156686,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.694700095207873e-06,
+      "loss": 0.9816,
+      "step": 8278
+    },
+    {
+      "epoch": 2.60085988574121,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.692161218660744e-06,
+      "loss": 1.2718,
+      "step": 8280
+    },
+    {
+      "epoch": 2.6014881127667513,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.689622342113615e-06,
+      "loss": 1.2261,
+      "step": 8282
+    },
+    {
+      "epoch": 2.6021163397922926,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.687083465566488e-06,
+      "loss": 1.1821,
+      "step": 8284
+    },
+    {
+      "epoch": 2.602744566817834,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.68454458901936e-06,
+      "loss": 1.1777,
+      "step": 8286
+    },
+    {
+      "epoch": 2.6033727938433753,
+      "grad_norm": 0.875,
+      "learning_rate": 9.682005712472233e-06,
+      "loss": 1.1155,
+      "step": 8288
+    },
+    {
+      "epoch": 2.6040010208689166,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.679466835925104e-06,
+      "loss": 1.1201,
+      "step": 8290
+    },
+    {
+      "epoch": 2.604629247894458,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.676927959377976e-06,
+      "loss": 1.0158,
+      "step": 8292
+    },
+    {
+      "epoch": 2.6052574749199993,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.674389082830849e-06,
+      "loss": 1.0516,
+      "step": 8294
+    },
+    {
+      "epoch": 2.6058857019455406,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.67185020628372e-06,
+      "loss": 1.3452,
+      "step": 8296
+    },
+    {
+      "epoch": 2.606513928971082,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.669311329736593e-06,
+      "loss": 1.2381,
+      "step": 8298
+    },
+    {
+      "epoch": 2.6071421559966232,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.666772453189465e-06,
+      "loss": 1.1524,
+      "step": 8300
+    },
+    {
+      "epoch": 2.6077703830221646,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.664233576642336e-06,
+      "loss": 1.0884,
+      "step": 8302
+    },
+    {
+      "epoch": 2.608398610047706,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.661694700095209e-06,
+      "loss": 1.2137,
+      "step": 8304
+    },
+    {
+      "epoch": 2.6090268370732472,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.65915582354808e-06,
+      "loss": 1.2906,
+      "step": 8306
+    },
+    {
+      "epoch": 2.6096550640987886,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.656616947000954e-06,
+      "loss": 1.1416,
+      "step": 8308
+    },
+    {
+      "epoch": 2.61028329112433,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.654078070453825e-06,
+      "loss": 1.2919,
+      "step": 8310
+    },
+    {
+      "epoch": 2.610911518149871,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.651539193906698e-06,
+      "loss": 1.3006,
+      "step": 8312
+    },
+    {
+      "epoch": 2.6115397451754125,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.64900031735957e-06,
+      "loss": 1.2853,
+      "step": 8314
+    },
+    {
+      "epoch": 2.612167972200954,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.64646144081244e-06,
+      "loss": 1.1489,
+      "step": 8316
+    },
+    {
+      "epoch": 2.612796199226495,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.643922564265314e-06,
+      "loss": 1.2045,
+      "step": 8318
+    },
+    {
+      "epoch": 2.613424426252037,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.641383687718185e-06,
+      "loss": 1.1971,
+      "step": 8320
+    },
+    {
+      "epoch": 2.6140526532775783,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.638844811171058e-06,
+      "loss": 1.1142,
+      "step": 8322
+    },
+    {
+      "epoch": 2.6146808803031196,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.63630593462393e-06,
+      "loss": 1.3033,
+      "step": 8324
+    },
+    {
+      "epoch": 2.615309107328661,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.633767058076801e-06,
+      "loss": 1.2125,
+      "step": 8326
+    },
+    {
+      "epoch": 2.6159373343542023,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.631228181529674e-06,
+      "loss": 1.0768,
+      "step": 8328
+    },
+    {
+      "epoch": 2.6165655613797436,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.628689304982545e-06,
+      "loss": 1.3004,
+      "step": 8330
+    },
+    {
+      "epoch": 2.617193788405285,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.626150428435419e-06,
+      "loss": 1.247,
+      "step": 8332
+    },
+    {
+      "epoch": 2.6178220154308263,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.62361155188829e-06,
+      "loss": 1.1625,
+      "step": 8334
+    },
+    {
+      "epoch": 2.6184502424563676,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.621072675341161e-06,
+      "loss": 1.1265,
+      "step": 8336
+    },
+    {
+      "epoch": 2.619078469481909,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.618533798794034e-06,
+      "loss": 1.2211,
+      "step": 8338
+    },
+    {
+      "epoch": 2.6197066965074502,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.615994922246906e-06,
+      "loss": 1.1047,
+      "step": 8340
+    },
+    {
+      "epoch": 2.6203349235329916,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.613456045699779e-06,
+      "loss": 1.2186,
+      "step": 8342
+    },
+    {
+      "epoch": 2.6209631505585333,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.61091716915265e-06,
+      "loss": 1.0722,
+      "step": 8344
+    },
+    {
+      "epoch": 2.6215913775840747,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.608378292605522e-06,
+      "loss": 1.2318,
+      "step": 8346
+    },
+    {
+      "epoch": 2.622219604609616,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.605839416058395e-06,
+      "loss": 1.1319,
+      "step": 8348
+    },
+    {
+      "epoch": 2.6228478316351573,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.603300539511266e-06,
+      "loss": 1.1703,
+      "step": 8350
+    },
+    {
+      "epoch": 2.6234760586606987,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.60076166296414e-06,
+      "loss": 1.2391,
+      "step": 8352
+    },
+    {
+      "epoch": 2.62410428568624,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.59822278641701e-06,
+      "loss": 1.2098,
+      "step": 8354
+    },
+    {
+      "epoch": 2.6247325127117813,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.595683909869884e-06,
+      "loss": 1.239,
+      "step": 8356
+    },
+    {
+      "epoch": 2.6253607397373226,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.593145033322755e-06,
+      "loss": 1.2125,
+      "step": 8358
+    },
+    {
+      "epoch": 2.625988966762864,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.590606156775628e-06,
+      "loss": 1.2088,
+      "step": 8360
+    },
+    {
+      "epoch": 2.6266171937884053,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.5880672802285e-06,
+      "loss": 1.166,
+      "step": 8362
+    },
+    {
+      "epoch": 2.6272454208139466,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.585528403681373e-06,
+      "loss": 1.204,
+      "step": 8364
+    },
+    {
+      "epoch": 2.627873647839488,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.582989527134244e-06,
+      "loss": 1.1946,
+      "step": 8366
+    },
+    {
+      "epoch": 2.6285018748650293,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.580450650587117e-06,
+      "loss": 1.2513,
+      "step": 8368
+    },
+    {
+      "epoch": 2.6291301018905706,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.577911774039988e-06,
+      "loss": 1.2714,
+      "step": 8370
+    },
+    {
+      "epoch": 2.629758328916112,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.57537289749286e-06,
+      "loss": 1.3492,
+      "step": 8372
+    },
+    {
+      "epoch": 2.6303865559416533,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.572834020945733e-06,
+      "loss": 1.2335,
+      "step": 8374
+    },
+    {
+      "epoch": 2.6310147829671946,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.570295144398604e-06,
+      "loss": 1.3027,
+      "step": 8376
+    },
+    {
+      "epoch": 2.631643009992736,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.567756267851477e-06,
+      "loss": 1.2272,
+      "step": 8378
+    },
+    {
+      "epoch": 2.6322712370182773,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 1.086,
+      "step": 8380
+    },
+    {
+      "epoch": 2.6328994640438186,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.562678514757222e-06,
+      "loss": 1.2643,
+      "step": 8382
+    },
+    {
+      "epoch": 2.6335276910693604,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.560139638210093e-06,
+      "loss": 1.1945,
+      "step": 8384
+    },
+    {
+      "epoch": 2.6341559180949017,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.557600761662965e-06,
+      "loss": 1.2233,
+      "step": 8386
+    },
+    {
+      "epoch": 2.634784145120443,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.555061885115838e-06,
+      "loss": 1.2933,
+      "step": 8388
+    },
+    {
+      "epoch": 2.6354123721459843,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.552523008568709e-06,
+      "loss": 1.2637,
+      "step": 8390
+    },
+    {
+      "epoch": 2.6360405991715257,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.549984132021582e-06,
+      "loss": 1.2201,
+      "step": 8392
+    },
+    {
+      "epoch": 2.636668826197067,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.547445255474453e-06,
+      "loss": 1.1775,
+      "step": 8394
+    },
+    {
+      "epoch": 2.6372970532226083,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.544906378927325e-06,
+      "loss": 1.1219,
+      "step": 8396
+    },
+    {
+      "epoch": 2.6379252802481497,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.542367502380198e-06,
+      "loss": 1.2112,
+      "step": 8398
+    },
+    {
+      "epoch": 2.638553507273691,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.53982862583307e-06,
+      "loss": 1.1167,
+      "step": 8400
+    },
+    {
+      "epoch": 2.6391817342992323,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.537289749285942e-06,
+      "loss": 1.2177,
+      "step": 8402
+    },
+    {
+      "epoch": 2.6398099613247736,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.534750872738814e-06,
+      "loss": 1.2147,
+      "step": 8404
+    },
+    {
+      "epoch": 2.640438188350315,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.532211996191685e-06,
+      "loss": 1.2503,
+      "step": 8406
+    },
+    {
+      "epoch": 2.6410664153758563,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.529673119644558e-06,
+      "loss": 1.2971,
+      "step": 8408
+    },
+    {
+      "epoch": 2.641694642401398,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.52713424309743e-06,
+      "loss": 1.1697,
+      "step": 8410
+    },
+    {
+      "epoch": 2.6423228694269394,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.524595366550303e-06,
+      "loss": 1.1083,
+      "step": 8412
+    },
+    {
+      "epoch": 2.6429510964524807,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.522056490003174e-06,
+      "loss": 1.2256,
+      "step": 8414
+    },
+    {
+      "epoch": 2.643579323478022,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.519517613456047e-06,
+      "loss": 1.1345,
+      "step": 8416
+    },
+    {
+      "epoch": 2.6442075505035634,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.516978736908918e-06,
+      "loss": 1.1583,
+      "step": 8418
+    },
+    {
+      "epoch": 2.6448357775291047,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.51443986036179e-06,
+      "loss": 1.1522,
+      "step": 8420
+    },
+    {
+      "epoch": 2.645464004554646,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.511900983814663e-06,
+      "loss": 1.1517,
+      "step": 8422
+    },
+    {
+      "epoch": 2.6460922315801874,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.509362107267534e-06,
+      "loss": 1.0531,
+      "step": 8424
+    },
+    {
+      "epoch": 2.6467204586057287,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.506823230720407e-06,
+      "loss": 1.177,
+      "step": 8426
+    },
+    {
+      "epoch": 2.64734868563127,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.504284354173279e-06,
+      "loss": 1.2406,
+      "step": 8428
+    },
+    {
+      "epoch": 2.6479769126568113,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.50174547762615e-06,
+      "loss": 1.2265,
+      "step": 8430
+    },
+    {
+      "epoch": 2.6486051396823527,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.499206601079023e-06,
+      "loss": 1.161,
+      "step": 8432
+    },
+    {
+      "epoch": 2.649233366707894,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.496667724531895e-06,
+      "loss": 1.1494,
+      "step": 8434
+    },
+    {
+      "epoch": 2.6498615937334353,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.494128847984768e-06,
+      "loss": 1.1465,
+      "step": 8436
+    },
+    {
+      "epoch": 2.6504898207589767,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.491589971437639e-06,
+      "loss": 1.1225,
+      "step": 8438
+    },
+    {
+      "epoch": 2.651118047784518,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.48905109489051e-06,
+      "loss": 1.094,
+      "step": 8440
+    },
+    {
+      "epoch": 2.6517462748100593,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.486512218343384e-06,
+      "loss": 1.0814,
+      "step": 8442
+    },
+    {
+      "epoch": 2.6523745018356006,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.483973341796255e-06,
+      "loss": 1.0285,
+      "step": 8444
+    },
+    {
+      "epoch": 2.653002728861142,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.481434465249128e-06,
+      "loss": 1.2455,
+      "step": 8446
+    },
+    {
+      "epoch": 2.6536309558866833,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.478895588702001e-06,
+      "loss": 1.1987,
+      "step": 8448
+    },
+    {
+      "epoch": 2.654259182912225,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.476356712154872e-06,
+      "loss": 1.2405,
+      "step": 8450
+    },
+    {
+      "epoch": 2.6548874099377664,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.473817835607746e-06,
+      "loss": 1.0695,
+      "step": 8452
+    },
+    {
+      "epoch": 2.6555156369633077,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.471278959060617e-06,
+      "loss": 1.0887,
+      "step": 8454
+    },
+    {
+      "epoch": 2.656143863988849,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.468740082513488e-06,
+      "loss": 1.1871,
+      "step": 8456
+    },
+    {
+      "epoch": 2.6567720910143904,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.466201205966361e-06,
+      "loss": 1.235,
+      "step": 8458
+    },
+    {
+      "epoch": 2.6574003180399317,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.463662329419233e-06,
+      "loss": 1.1305,
+      "step": 8460
+    },
+    {
+      "epoch": 2.658028545065473,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.461123452872106e-06,
+      "loss": 1.0736,
+      "step": 8462
+    },
+    {
+      "epoch": 2.6586567720910144,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.458584576324977e-06,
+      "loss": 1.1682,
+      "step": 8464
+    },
+    {
+      "epoch": 2.6592849991165557,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.456045699777849e-06,
+      "loss": 1.1639,
+      "step": 8466
+    },
+    {
+      "epoch": 2.659913226142097,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.453506823230722e-06,
+      "loss": 1.2223,
+      "step": 8468
+    },
+    {
+      "epoch": 2.6605414531676383,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.450967946683593e-06,
+      "loss": 1.1154,
+      "step": 8470
+    },
+    {
+      "epoch": 2.6611696801931797,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.448429070136466e-06,
+      "loss": 1.0826,
+      "step": 8472
+    },
+    {
+      "epoch": 2.661797907218721,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.445890193589338e-06,
+      "loss": 1.2074,
+      "step": 8474
+    },
+    {
+      "epoch": 2.6624261342442628,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.443351317042209e-06,
+      "loss": 1.1333,
+      "step": 8476
+    },
+    {
+      "epoch": 2.663054361269804,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.440812440495082e-06,
+      "loss": 1.2202,
+      "step": 8478
+    },
+    {
+      "epoch": 2.6636825882953454,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.438273563947953e-06,
+      "loss": 1.2403,
+      "step": 8480
+    },
+    {
+      "epoch": 2.6643108153208868,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.435734687400826e-06,
+      "loss": 1.1786,
+      "step": 8482
+    },
+    {
+      "epoch": 2.664939042346428,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.433195810853698e-06,
+      "loss": 1.1325,
+      "step": 8484
+    },
+    {
+      "epoch": 2.6655672693719694,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.430656934306571e-06,
+      "loss": 1.0782,
+      "step": 8486
+    },
+    {
+      "epoch": 2.6661954963975107,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.428118057759442e-06,
+      "loss": 1.217,
+      "step": 8488
+    },
+    {
+      "epoch": 2.666823723423052,
+      "grad_norm": 0.875,
+      "learning_rate": 9.425579181212314e-06,
+      "loss": 1.1414,
+      "step": 8490
+    },
+    {
+      "epoch": 2.6674519504485934,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.423040304665187e-06,
+      "loss": 1.2201,
+      "step": 8492
+    },
+    {
+      "epoch": 2.6680801774741347,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.420501428118058e-06,
+      "loss": 1.3012,
+      "step": 8494
+    },
+    {
+      "epoch": 2.668708404499676,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.417962551570931e-06,
+      "loss": 1.2001,
+      "step": 8496
+    },
+    {
+      "epoch": 2.6693366315252174,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.415423675023803e-06,
+      "loss": 1.2408,
+      "step": 8498
+    },
+    {
+      "epoch": 2.6699648585507587,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.412884798476674e-06,
+      "loss": 1.129,
+      "step": 8500
+    },
+    {
+      "epoch": 2.6705930855763,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.410345921929547e-06,
+      "loss": 1.3063,
+      "step": 8502
+    },
+    {
+      "epoch": 2.6712213126018414,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.407807045382418e-06,
+      "loss": 1.2523,
+      "step": 8504
+    },
+    {
+      "epoch": 2.6718495396273827,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.405268168835291e-06,
+      "loss": 1.1856,
+      "step": 8506
+    },
+    {
+      "epoch": 2.672477766652924,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.402729292288163e-06,
+      "loss": 1.1575,
+      "step": 8508
+    },
+    {
+      "epoch": 2.6731059936784654,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.400190415741034e-06,
+      "loss": 1.0457,
+      "step": 8510
+    },
+    {
+      "epoch": 2.6737342207040067,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.397651539193907e-06,
+      "loss": 1.1486,
+      "step": 8512
+    },
+    {
+      "epoch": 2.674362447729548,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.395112662646779e-06,
+      "loss": 1.094,
+      "step": 8514
+    },
+    {
+      "epoch": 2.67499067475509,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.392573786099652e-06,
+      "loss": 1.1416,
+      "step": 8516
+    },
+    {
+      "epoch": 2.675618901780631,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.390034909552523e-06,
+      "loss": 1.2959,
+      "step": 8518
+    },
+    {
+      "epoch": 2.6762471288061724,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.387496033005396e-06,
+      "loss": 1.0791,
+      "step": 8520
+    },
+    {
+      "epoch": 2.6768753558317138,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.384957156458268e-06,
+      "loss": 1.2186,
+      "step": 8522
+    },
+    {
+      "epoch": 2.677503582857255,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.382418279911139e-06,
+      "loss": 1.1701,
+      "step": 8524
+    },
+    {
+      "epoch": 2.6781318098827964,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.379879403364012e-06,
+      "loss": 1.0534,
+      "step": 8526
+    },
+    {
+      "epoch": 2.6787600369083377,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.377340526816883e-06,
+      "loss": 1.0403,
+      "step": 8528
+    },
+    {
+      "epoch": 2.679388263933879,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.374801650269757e-06,
+      "loss": 1.2593,
+      "step": 8530
+    },
+    {
+      "epoch": 2.6800164909594204,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.37226277372263e-06,
+      "loss": 1.401,
+      "step": 8532
+    },
+    {
+      "epoch": 2.6806447179849617,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.369723897175501e-06,
+      "loss": 1.113,
+      "step": 8534
+    },
+    {
+      "epoch": 2.681272945010503,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.367185020628372e-06,
+      "loss": 1.2193,
+      "step": 8536
+    },
+    {
+      "epoch": 2.6819011720360444,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.364646144081245e-06,
+      "loss": 1.2076,
+      "step": 8538
+    },
+    {
+      "epoch": 2.6825293990615857,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.362107267534117e-06,
+      "loss": 1.2201,
+      "step": 8540
+    },
+    {
+      "epoch": 2.6831576260871275,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.35956839098699e-06,
+      "loss": 1.2454,
+      "step": 8542
+    },
+    {
+      "epoch": 2.683785853112669,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.357029514439861e-06,
+      "loss": 1.1856,
+      "step": 8544
+    },
+    {
+      "epoch": 2.68441408013821,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.354490637892734e-06,
+      "loss": 1.1561,
+      "step": 8546
+    },
+    {
+      "epoch": 2.6850423071637515,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.351951761345606e-06,
+      "loss": 1.2809,
+      "step": 8548
+    },
+    {
+      "epoch": 2.685670534189293,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.349412884798477e-06,
+      "loss": 1.1202,
+      "step": 8550
+    },
+    {
+      "epoch": 2.686298761214834,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.34687400825135e-06,
+      "loss": 1.2268,
+      "step": 8552
+    },
+    {
+      "epoch": 2.6869269882403755,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.344335131704222e-06,
+      "loss": 1.3182,
+      "step": 8554
+    },
+    {
+      "epoch": 2.687555215265917,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.341796255157095e-06,
+      "loss": 1.1671,
+      "step": 8556
+    },
+    {
+      "epoch": 2.688183442291458,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.339257378609966e-06,
+      "loss": 1.1365,
+      "step": 8558
+    },
+    {
+      "epoch": 2.6888116693169994,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.336718502062837e-06,
+      "loss": 1.3515,
+      "step": 8560
+    },
+    {
+      "epoch": 2.6894398963425408,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.33417962551571e-06,
+      "loss": 1.2687,
+      "step": 8562
+    },
+    {
+      "epoch": 2.690068123368082,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.331640748968582e-06,
+      "loss": 1.3292,
+      "step": 8564
+    },
+    {
+      "epoch": 2.6906963503936234,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.329101872421455e-06,
+      "loss": 1.1909,
+      "step": 8566
+    },
+    {
+      "epoch": 2.6913245774191648,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.326562995874326e-06,
+      "loss": 1.1728,
+      "step": 8568
+    },
+    {
+      "epoch": 2.691952804444706,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.324024119327198e-06,
+      "loss": 1.129,
+      "step": 8570
+    },
+    {
+      "epoch": 2.6925810314702474,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.32148524278007e-06,
+      "loss": 1.1925,
+      "step": 8572
+    },
+    {
+      "epoch": 2.6932092584957887,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.318946366232942e-06,
+      "loss": 1.2486,
+      "step": 8574
+    },
+    {
+      "epoch": 2.69383748552133,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.316407489685815e-06,
+      "loss": 1.3359,
+      "step": 8576
+    },
+    {
+      "epoch": 2.6944657125468714,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.313868613138687e-06,
+      "loss": 1.1934,
+      "step": 8578
+    },
+    {
+      "epoch": 2.6950939395724127,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.31132973659156e-06,
+      "loss": 1.1376,
+      "step": 8580
+    },
+    {
+      "epoch": 2.6957221665979545,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.308790860044431e-06,
+      "loss": 1.1097,
+      "step": 8582
+    },
+    {
+      "epoch": 2.696350393623496,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.306251983497303e-06,
+      "loss": 1.1256,
+      "step": 8584
+    },
+    {
+      "epoch": 2.696978620649037,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.303713106950176e-06,
+      "loss": 1.1685,
+      "step": 8586
+    },
+    {
+      "epoch": 2.6976068476745785,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.301174230403047e-06,
+      "loss": 1.2755,
+      "step": 8588
+    },
+    {
+      "epoch": 2.69823507470012,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.29863535385592e-06,
+      "loss": 1.0006,
+      "step": 8590
+    },
+    {
+      "epoch": 2.698863301725661,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.296096477308791e-06,
+      "loss": 1.1454,
+      "step": 8592
+    },
+    {
+      "epoch": 2.6994915287512025,
+      "grad_norm": 0.875,
+      "learning_rate": 9.293557600761663e-06,
+      "loss": 1.1855,
+      "step": 8594
+    },
+    {
+      "epoch": 2.700119755776744,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.291018724214536e-06,
+      "loss": 1.0808,
+      "step": 8596
+    },
+    {
+      "epoch": 2.700747982802285,
+      "grad_norm": 0.875,
+      "learning_rate": 9.288479847667407e-06,
+      "loss": 1.2162,
+      "step": 8598
+    },
+    {
+      "epoch": 2.7013762098278264,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.28594097112028e-06,
+      "loss": 1.252,
+      "step": 8600
+    },
+    {
+      "epoch": 2.7020044368533678,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.283402094573152e-06,
+      "loss": 1.1132,
+      "step": 8602
+    },
+    {
+      "epoch": 2.702632663878909,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.280863218026023e-06,
+      "loss": 1.2573,
+      "step": 8604
+    },
+    {
+      "epoch": 2.7032608909044504,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.278324341478896e-06,
+      "loss": 0.9743,
+      "step": 8606
+    },
+    {
+      "epoch": 2.703889117929992,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.275785464931768e-06,
+      "loss": 1.2334,
+      "step": 8608
+    },
+    {
+      "epoch": 2.7045173449555335,
+      "grad_norm": 0.875,
+      "learning_rate": 9.27324658838464e-06,
+      "loss": 1.2431,
+      "step": 8610
+    },
+    {
+      "epoch": 2.705145571981075,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.270707711837512e-06,
+      "loss": 1.1449,
+      "step": 8612
+    },
+    {
+      "epoch": 2.705773799006616,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.268168835290383e-06,
+      "loss": 1.2655,
+      "step": 8614
+    },
+    {
+      "epoch": 2.7064020260321575,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.265629958743256e-06,
+      "loss": 1.0926,
+      "step": 8616
+    },
+    {
+      "epoch": 2.707030253057699,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.26309108219613e-06,
+      "loss": 1.2308,
+      "step": 8618
+    },
+    {
+      "epoch": 2.70765848008324,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.260552205649001e-06,
+      "loss": 1.1774,
+      "step": 8620
+    },
+    {
+      "epoch": 2.7082867071087815,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.258013329101874e-06,
+      "loss": 1.2488,
+      "step": 8622
+    },
+    {
+      "epoch": 2.708914934134323,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.255474452554745e-06,
+      "loss": 1.1772,
+      "step": 8624
+    },
+    {
+      "epoch": 2.709543161159864,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.252935576007618e-06,
+      "loss": 1.0649,
+      "step": 8626
+    },
+    {
+      "epoch": 2.7101713881854055,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.25039669946049e-06,
+      "loss": 1.1924,
+      "step": 8628
+    },
+    {
+      "epoch": 2.710799615210947,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.247857822913361e-06,
+      "loss": 1.2255,
+      "step": 8630
+    },
+    {
+      "epoch": 2.711427842236488,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.245318946366234e-06,
+      "loss": 1.2982,
+      "step": 8632
+    },
+    {
+      "epoch": 2.7120560692620295,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.242780069819106e-06,
+      "loss": 1.242,
+      "step": 8634
+    },
+    {
+      "epoch": 2.712684296287571,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.240241193271979e-06,
+      "loss": 1.2739,
+      "step": 8636
+    },
+    {
+      "epoch": 2.713312523313112,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.23770231672485e-06,
+      "loss": 1.1578,
+      "step": 8638
+    },
+    {
+      "epoch": 2.7139407503386535,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.235163440177722e-06,
+      "loss": 1.3106,
+      "step": 8640
+    },
+    {
+      "epoch": 2.714568977364195,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.232624563630595e-06,
+      "loss": 1.142,
+      "step": 8642
+    },
+    {
+      "epoch": 2.715197204389736,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.230085687083466e-06,
+      "loss": 1.1661,
+      "step": 8644
+    },
+    {
+      "epoch": 2.7158254314152774,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.227546810536339e-06,
+      "loss": 1.2508,
+      "step": 8646
+    },
+    {
+      "epoch": 2.716453658440819,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.22500793398921e-06,
+      "loss": 1.3473,
+      "step": 8648
+    },
+    {
+      "epoch": 2.7170818854663605,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.222469057442084e-06,
+      "loss": 1.2084,
+      "step": 8650
+    },
+    {
+      "epoch": 2.717710112491902,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.219930180894955e-06,
+      "loss": 1.1336,
+      "step": 8652
+    },
+    {
+      "epoch": 2.718338339517443,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.217391304347826e-06,
+      "loss": 1.2546,
+      "step": 8654
+    },
+    {
+      "epoch": 2.7189665665429845,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.2148524278007e-06,
+      "loss": 1.2444,
+      "step": 8656
+    },
+    {
+      "epoch": 2.719594793568526,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.21231355125357e-06,
+      "loss": 1.1496,
+      "step": 8658
+    },
+    {
+      "epoch": 2.720223020594067,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.209774674706444e-06,
+      "loss": 1.0408,
+      "step": 8660
+    },
+    {
+      "epoch": 2.7208512476196085,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.207235798159315e-06,
+      "loss": 1.1055,
+      "step": 8662
+    },
+    {
+      "epoch": 2.72147947464515,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.204696921612187e-06,
+      "loss": 1.203,
+      "step": 8664
+    },
+    {
+      "epoch": 2.722107701670691,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.20215804506506e-06,
+      "loss": 1.1595,
+      "step": 8666
+    },
+    {
+      "epoch": 2.7227359286962325,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.199619168517931e-06,
+      "loss": 1.4185,
+      "step": 8668
+    },
+    {
+      "epoch": 2.723364155721774,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.197080291970804e-06,
+      "loss": 1.1604,
+      "step": 8670
+    },
+    {
+      "epoch": 2.7239923827473156,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.194541415423676e-06,
+      "loss": 1.1948,
+      "step": 8672
+    },
+    {
+      "epoch": 2.724620609772857,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.192002538876547e-06,
+      "loss": 1.1164,
+      "step": 8674
+    },
+    {
+      "epoch": 2.7252488367983982,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.18946366232942e-06,
+      "loss": 1.1879,
+      "step": 8676
+    },
+    {
+      "epoch": 2.7258770638239396,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.186924785782291e-06,
+      "loss": 1.2257,
+      "step": 8678
+    },
+    {
+      "epoch": 2.726505290849481,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.184385909235164e-06,
+      "loss": 1.1763,
+      "step": 8680
+    },
+    {
+      "epoch": 2.7271335178750222,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.181847032688036e-06,
+      "loss": 1.1286,
+      "step": 8682
+    },
+    {
+      "epoch": 2.7277617449005636,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.179308156140909e-06,
+      "loss": 1.1175,
+      "step": 8684
+    },
+    {
+      "epoch": 2.728389971926105,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.17676927959378e-06,
+      "loss": 1.2022,
+      "step": 8686
+    },
+    {
+      "epoch": 2.729018198951646,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.174230403046652e-06,
+      "loss": 1.3269,
+      "step": 8688
+    },
+    {
+      "epoch": 2.7296464259771875,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.171691526499525e-06,
+      "loss": 1.131,
+      "step": 8690
+    },
+    {
+      "epoch": 2.730274653002729,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.169152649952396e-06,
+      "loss": 1.1066,
+      "step": 8692
+    },
+    {
+      "epoch": 2.73090288002827,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.16661377340527e-06,
+      "loss": 1.1499,
+      "step": 8694
+    },
+    {
+      "epoch": 2.7315311070538115,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.16407489685814e-06,
+      "loss": 1.1262,
+      "step": 8696
+    },
+    {
+      "epoch": 2.732159334079353,
+      "grad_norm": 1.0,
+      "learning_rate": 9.161536020311012e-06,
+      "loss": 1.1625,
+      "step": 8698
+    },
+    {
+      "epoch": 2.732787561104894,
+      "grad_norm": 1.0,
+      "learning_rate": 9.158997143763885e-06,
+      "loss": 1.1464,
+      "step": 8700
+    },
+    {
+      "epoch": 2.7334157881304355,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.156458267216756e-06,
+      "loss": 1.2327,
+      "step": 8702
+    },
+    {
+      "epoch": 2.734044015155977,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.15391939066963e-06,
+      "loss": 1.0969,
+      "step": 8704
+    },
+    {
+      "epoch": 2.734672242181518,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.151380514122503e-06,
+      "loss": 1.1112,
+      "step": 8706
+    },
+    {
+      "epoch": 2.7353004692070595,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.148841637575374e-06,
+      "loss": 1.3812,
+      "step": 8708
+    },
+    {
+      "epoch": 2.735928696232601,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.146302761028247e-06,
+      "loss": 1.2572,
+      "step": 8710
+    },
+    {
+      "epoch": 2.736556923258142,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.143763884481118e-06,
+      "loss": 1.142,
+      "step": 8712
+    },
+    {
+      "epoch": 2.737185150283684,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.14122500793399e-06,
+      "loss": 1.144,
+      "step": 8714
+    },
+    {
+      "epoch": 2.7378133773092252,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.138686131386863e-06,
+      "loss": 1.3231,
+      "step": 8716
+    },
+    {
+      "epoch": 2.7384416043347666,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.136147254839734e-06,
+      "loss": 1.1236,
+      "step": 8718
+    },
+    {
+      "epoch": 2.739069831360308,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.133608378292607e-06,
+      "loss": 1.1687,
+      "step": 8720
+    },
+    {
+      "epoch": 2.7396980583858492,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.131069501745479e-06,
+      "loss": 1.2533,
+      "step": 8722
+    },
+    {
+      "epoch": 2.7403262854113906,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.12853062519835e-06,
+      "loss": 1.0806,
+      "step": 8724
+    },
+    {
+      "epoch": 2.740954512436932,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.125991748651223e-06,
+      "loss": 1.2269,
+      "step": 8726
+    },
+    {
+      "epoch": 2.741582739462473,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.123452872104095e-06,
+      "loss": 1.0659,
+      "step": 8728
+    },
+    {
+      "epoch": 2.7422109664880145,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.120913995556968e-06,
+      "loss": 1.2846,
+      "step": 8730
+    },
+    {
+      "epoch": 2.742839193513556,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.118375119009839e-06,
+      "loss": 1.2342,
+      "step": 8732
+    },
+    {
+      "epoch": 2.743467420539097,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.11583624246271e-06,
+      "loss": 1.2884,
+      "step": 8734
+    },
+    {
+      "epoch": 2.7440956475646385,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.113297365915583e-06,
+      "loss": 1.2428,
+      "step": 8736
+    },
+    {
+      "epoch": 2.7447238745901803,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.110758489368455e-06,
+      "loss": 1.2098,
+      "step": 8738
+    },
+    {
+      "epoch": 2.7453521016157216,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.108219612821328e-06,
+      "loss": 1.2179,
+      "step": 8740
+    },
+    {
+      "epoch": 2.745980328641263,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.1056807362742e-06,
+      "loss": 1.3499,
+      "step": 8742
+    },
+    {
+      "epoch": 2.7466085556668043,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.103141859727072e-06,
+      "loss": 1.0601,
+      "step": 8744
+    },
+    {
+      "epoch": 2.7472367826923456,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.100602983179944e-06,
+      "loss": 1.2922,
+      "step": 8746
+    },
+    {
+      "epoch": 2.747865009717887,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.098064106632815e-06,
+      "loss": 1.1371,
+      "step": 8748
+    },
+    {
+      "epoch": 2.7484932367434283,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.095525230085688e-06,
+      "loss": 1.2278,
+      "step": 8750
+    },
+    {
+      "epoch": 2.7491214637689696,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.09298635353856e-06,
+      "loss": 1.1732,
+      "step": 8752
+    },
+    {
+      "epoch": 2.749749690794511,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.090447476991433e-06,
+      "loss": 1.1805,
+      "step": 8754
+    },
+    {
+      "epoch": 2.7503779178200523,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.087908600444304e-06,
+      "loss": 1.0657,
+      "step": 8756
+    },
+    {
+      "epoch": 2.7510061448455936,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.085369723897175e-06,
+      "loss": 1.1152,
+      "step": 8758
+    },
+    {
+      "epoch": 2.751634371871135,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.082830847350049e-06,
+      "loss": 1.1851,
+      "step": 8760
+    },
+    {
+      "epoch": 2.7522625988966762,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.08029197080292e-06,
+      "loss": 1.1882,
+      "step": 8762
+    },
+    {
+      "epoch": 2.7528908259222176,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.077753094255793e-06,
+      "loss": 1.1752,
+      "step": 8764
+    },
+    {
+      "epoch": 2.753519052947759,
+      "grad_norm": 1.625,
+      "learning_rate": 9.075214217708664e-06,
+      "loss": 1.2417,
+      "step": 8766
+    },
+    {
+      "epoch": 2.7541472799733,
+      "grad_norm": 0.84375,
+      "learning_rate": 9.072675341161536e-06,
+      "loss": 1.1084,
+      "step": 8768
+    },
+    {
+      "epoch": 2.7547755069988415,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.070136464614409e-06,
+      "loss": 1.1443,
+      "step": 8770
+    },
+    {
+      "epoch": 2.755403734024383,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.06759758806728e-06,
+      "loss": 1.193,
+      "step": 8772
+    },
+    {
+      "epoch": 2.756031961049924,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.065058711520153e-06,
+      "loss": 1.2437,
+      "step": 8774
+    },
+    {
+      "epoch": 2.7566601880754655,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.062519834973025e-06,
+      "loss": 1.2941,
+      "step": 8776
+    },
+    {
+      "epoch": 2.757288415101007,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.059980958425896e-06,
+      "loss": 1.0076,
+      "step": 8778
+    },
+    {
+      "epoch": 2.7579166421265486,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.057442081878769e-06,
+      "loss": 1.1461,
+      "step": 8780
+    },
+    {
+      "epoch": 2.75854486915209,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.05490320533164e-06,
+      "loss": 1.1561,
+      "step": 8782
+    },
+    {
+      "epoch": 2.7591730961776313,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.052364328784514e-06,
+      "loss": 1.104,
+      "step": 8784
+    },
+    {
+      "epoch": 2.7598013232031726,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.049825452237385e-06,
+      "loss": 1.1338,
+      "step": 8786
+    },
+    {
+      "epoch": 2.760429550228714,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.047286575690258e-06,
+      "loss": 1.2664,
+      "step": 8788
+    },
+    {
+      "epoch": 2.7610577772542553,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.044747699143131e-06,
+      "loss": 1.1987,
+      "step": 8790
+    },
+    {
+      "epoch": 2.7616860042797966,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.042208822596002e-06,
+      "loss": 1.1966,
+      "step": 8792
+    },
+    {
+      "epoch": 2.762314231305338,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.039669946048874e-06,
+      "loss": 1.3295,
+      "step": 8794
+    },
+    {
+      "epoch": 2.7629424583308793,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.037131069501747e-06,
+      "loss": 1.2477,
+      "step": 8796
+    },
+    {
+      "epoch": 2.7635706853564206,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.034592192954618e-06,
+      "loss": 1.1968,
+      "step": 8798
+    },
+    {
+      "epoch": 2.764198912381962,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.032053316407491e-06,
+      "loss": 1.2968,
+      "step": 8800
+    },
+    {
+      "epoch": 2.7648271394075032,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.029514439860363e-06,
+      "loss": 1.22,
+      "step": 8802
+    },
+    {
+      "epoch": 2.765455366433045,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.026975563313234e-06,
+      "loss": 1.184,
+      "step": 8804
+    },
+    {
+      "epoch": 2.7660835934585863,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.024436686766107e-06,
+      "loss": 1.2109,
+      "step": 8806
+    },
+    {
+      "epoch": 2.7667118204841277,
+      "grad_norm": 0.82421875,
+      "learning_rate": 9.021897810218979e-06,
+      "loss": 1.2232,
+      "step": 8808
+    },
+    {
+      "epoch": 2.767340047509669,
+      "grad_norm": 0.875,
+      "learning_rate": 9.019358933671852e-06,
+      "loss": 1.4004,
+      "step": 8810
+    },
+    {
+      "epoch": 2.7679682745352103,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.016820057124723e-06,
+      "loss": 1.2854,
+      "step": 8812
+    },
+    {
+      "epoch": 2.7685965015607517,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.014281180577596e-06,
+      "loss": 1.0666,
+      "step": 8814
+    },
+    {
+      "epoch": 2.769224728586293,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.011742304030468e-06,
+      "loss": 1.241,
+      "step": 8816
+    },
+    {
+      "epoch": 2.7698529556118343,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.009203427483339e-06,
+      "loss": 1.0892,
+      "step": 8818
+    },
+    {
+      "epoch": 2.7704811826373756,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.006664550936212e-06,
+      "loss": 1.1606,
+      "step": 8820
+    },
+    {
+      "epoch": 2.771109409662917,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.004125674389083e-06,
+      "loss": 1.2497,
+      "step": 8822
+    },
+    {
+      "epoch": 2.7717376366884583,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.001586797841956e-06,
+      "loss": 1.2447,
+      "step": 8824
+    },
+    {
+      "epoch": 2.7723658637139996,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.999047921294828e-06,
+      "loss": 1.1626,
+      "step": 8826
+    },
+    {
+      "epoch": 2.772994090739541,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.9965090447477e-06,
+      "loss": 1.0602,
+      "step": 8828
+    },
+    {
+      "epoch": 2.7736223177650823,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.993970168200572e-06,
+      "loss": 1.1906,
+      "step": 8830
+    },
+    {
+      "epoch": 2.7742505447906236,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.991431291653444e-06,
+      "loss": 1.1996,
+      "step": 8832
+    },
+    {
+      "epoch": 2.774878771816165,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.988892415106317e-06,
+      "loss": 1.2612,
+      "step": 8834
+    },
+    {
+      "epoch": 2.7755069988417063,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.986353538559188e-06,
+      "loss": 1.1974,
+      "step": 8836
+    },
+    {
+      "epoch": 2.7761352258672476,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.98381466201206e-06,
+      "loss": 1.1045,
+      "step": 8838
+    },
+    {
+      "epoch": 2.776763452892789,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.981275785464933e-06,
+      "loss": 1.1507,
+      "step": 8840
+    },
+    {
+      "epoch": 2.7773916799183302,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.978736908917804e-06,
+      "loss": 1.2665,
+      "step": 8842
+    },
+    {
+      "epoch": 2.7780199069438716,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.976198032370677e-06,
+      "loss": 1.1378,
+      "step": 8844
+    },
+    {
+      "epoch": 2.7786481339694133,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.973659155823548e-06,
+      "loss": 1.0703,
+      "step": 8846
+    },
+    {
+      "epoch": 2.7792763609949547,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.971120279276422e-06,
+      "loss": 1.0785,
+      "step": 8848
+    },
+    {
+      "epoch": 2.779904588020496,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.968581402729293e-06,
+      "loss": 1.2303,
+      "step": 8850
+    },
+    {
+      "epoch": 2.7805328150460373,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.966042526182164e-06,
+      "loss": 1.2005,
+      "step": 8852
+    },
+    {
+      "epoch": 2.7811610420715787,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.963503649635037e-06,
+      "loss": 1.2388,
+      "step": 8854
+    },
+    {
+      "epoch": 2.78178926909712,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.960964773087909e-06,
+      "loss": 1.1309,
+      "step": 8856
+    },
+    {
+      "epoch": 2.7824174961226613,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.958425896540782e-06,
+      "loss": 1.2428,
+      "step": 8858
+    },
+    {
+      "epoch": 2.7830457231482026,
+      "grad_norm": 0.875,
+      "learning_rate": 8.955887019993653e-06,
+      "loss": 1.1717,
+      "step": 8860
+    },
+    {
+      "epoch": 2.783673950173744,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.953348143446525e-06,
+      "loss": 1.2138,
+      "step": 8862
+    },
+    {
+      "epoch": 2.7843021771992853,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.950809266899398e-06,
+      "loss": 1.0967,
+      "step": 8864
+    },
+    {
+      "epoch": 2.7849304042248266,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.948270390352269e-06,
+      "loss": 1.164,
+      "step": 8866
+    },
+    {
+      "epoch": 2.785558631250368,
+      "grad_norm": 0.875,
+      "learning_rate": 8.945731513805142e-06,
+      "loss": 1.1851,
+      "step": 8868
+    },
+    {
+      "epoch": 2.7861868582759097,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.943192637258013e-06,
+      "loss": 1.09,
+      "step": 8870
+    },
+    {
+      "epoch": 2.786815085301451,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.940653760710885e-06,
+      "loss": 1.2916,
+      "step": 8872
+    },
+    {
+      "epoch": 2.7874433123269924,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.938114884163758e-06,
+      "loss": 1.2003,
+      "step": 8874
+    },
+    {
+      "epoch": 2.7880715393525337,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.935576007616631e-06,
+      "loss": 1.2229,
+      "step": 8876
+    },
+    {
+      "epoch": 2.788699766378075,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.933037131069502e-06,
+      "loss": 1.2452,
+      "step": 8878
+    },
+    {
+      "epoch": 2.7893279934036164,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.930498254522375e-06,
+      "loss": 1.2213,
+      "step": 8880
+    },
+    {
+      "epoch": 2.7899562204291577,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.927959377975247e-06,
+      "loss": 1.0626,
+      "step": 8882
+    },
+    {
+      "epoch": 2.790584447454699,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.92542050142812e-06,
+      "loss": 1.0885,
+      "step": 8884
+    },
+    {
+      "epoch": 2.7912126744802404,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.922881624880991e-06,
+      "loss": 1.1579,
+      "step": 8886
+    },
+    {
+      "epoch": 2.7918409015057817,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.920342748333863e-06,
+      "loss": 1.1024,
+      "step": 8888
+    },
+    {
+      "epoch": 2.792469128531323,
+      "grad_norm": 0.875,
+      "learning_rate": 8.917803871786736e-06,
+      "loss": 1.2193,
+      "step": 8890
+    },
+    {
+      "epoch": 2.7930973555568643,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.915264995239607e-06,
+      "loss": 1.0511,
+      "step": 8892
+    },
+    {
+      "epoch": 2.7937255825824057,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.91272611869248e-06,
+      "loss": 1.1799,
+      "step": 8894
+    },
+    {
+      "epoch": 2.794353809607947,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.910187242145352e-06,
+      "loss": 1.121,
+      "step": 8896
+    },
+    {
+      "epoch": 2.7949820366334883,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.907648365598223e-06,
+      "loss": 1.3075,
+      "step": 8898
+    },
+    {
+      "epoch": 2.7956102636590296,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.905109489051096e-06,
+      "loss": 1.0603,
+      "step": 8900
+    },
+    {
+      "epoch": 2.796238490684571,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.902570612503967e-06,
+      "loss": 1.1152,
+      "step": 8902
+    },
+    {
+      "epoch": 2.7968667177101123,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.90003173595684e-06,
+      "loss": 1.1665,
+      "step": 8904
+    },
+    {
+      "epoch": 2.7974949447356536,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.897492859409712e-06,
+      "loss": 1.1832,
+      "step": 8906
+    },
+    {
+      "epoch": 2.798123171761195,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.894953982862583e-06,
+      "loss": 1.162,
+      "step": 8908
+    },
+    {
+      "epoch": 2.7987513987867363,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.892415106315456e-06,
+      "loss": 1.2164,
+      "step": 8910
+    },
+    {
+      "epoch": 2.799379625812278,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.889876229768328e-06,
+      "loss": 1.1247,
+      "step": 8912
+    },
+    {
+      "epoch": 2.8000078528378194,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.8873373532212e-06,
+      "loss": 1.0951,
+      "step": 8914
+    },
+    {
+      "epoch": 2.8006360798633607,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.884798476674072e-06,
+      "loss": 1.1273,
+      "step": 8916
+    },
+    {
+      "epoch": 2.801264306888902,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.882259600126945e-06,
+      "loss": 1.0575,
+      "step": 8918
+    },
+    {
+      "epoch": 2.8018925339144434,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.879720723579817e-06,
+      "loss": 1.2221,
+      "step": 8920
+    },
+    {
+      "epoch": 2.8025207609399847,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.877181847032688e-06,
+      "loss": 1.2034,
+      "step": 8922
+    },
+    {
+      "epoch": 2.803148987965526,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.874642970485561e-06,
+      "loss": 1.1343,
+      "step": 8924
+    },
+    {
+      "epoch": 2.8037772149910674,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.872104093938433e-06,
+      "loss": 1.1435,
+      "step": 8926
+    },
+    {
+      "epoch": 2.8044054420166087,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.869565217391306e-06,
+      "loss": 1.1574,
+      "step": 8928
+    },
+    {
+      "epoch": 2.80503366904215,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.867026340844177e-06,
+      "loss": 1.1683,
+      "step": 8930
+    },
+    {
+      "epoch": 2.8056618960676913,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.864487464297048e-06,
+      "loss": 1.1728,
+      "step": 8932
+    },
+    {
+      "epoch": 2.8062901230932327,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.861948587749921e-06,
+      "loss": 1.3423,
+      "step": 8934
+    },
+    {
+      "epoch": 2.8069183501187744,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.859409711202793e-06,
+      "loss": 1.1885,
+      "step": 8936
+    },
+    {
+      "epoch": 2.8075465771443158,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.856870834655666e-06,
+      "loss": 1.2339,
+      "step": 8938
+    },
+    {
+      "epoch": 2.808174804169857,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.854331958108537e-06,
+      "loss": 1.1838,
+      "step": 8940
+    },
+    {
+      "epoch": 2.8088030311953984,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.851793081561409e-06,
+      "loss": 1.1717,
+      "step": 8942
+    },
+    {
+      "epoch": 2.8094312582209398,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.849254205014282e-06,
+      "loss": 1.1993,
+      "step": 8944
+    },
+    {
+      "epoch": 2.810059485246481,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.846715328467153e-06,
+      "loss": 1.2121,
+      "step": 8946
+    },
+    {
+      "epoch": 2.8106877122720224,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.844176451920026e-06,
+      "loss": 1.1136,
+      "step": 8948
+    },
+    {
+      "epoch": 2.8113159392975637,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.841637575372898e-06,
+      "loss": 1.0336,
+      "step": 8950
+    },
+    {
+      "epoch": 2.811944166323105,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.83909869882577e-06,
+      "loss": 1.2212,
+      "step": 8952
+    },
+    {
+      "epoch": 2.8125723933486464,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.836559822278642e-06,
+      "loss": 1.1801,
+      "step": 8954
+    },
+    {
+      "epoch": 2.8132006203741877,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.834020945731513e-06,
+      "loss": 1.2164,
+      "step": 8956
+    },
+    {
+      "epoch": 2.813828847399729,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.831482069184387e-06,
+      "loss": 1.1291,
+      "step": 8958
+    },
+    {
+      "epoch": 2.8144570744252704,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.828943192637258e-06,
+      "loss": 1.1178,
+      "step": 8960
+    },
+    {
+      "epoch": 2.8150853014508117,
+      "grad_norm": 1.0,
+      "learning_rate": 8.826404316090131e-06,
+      "loss": 1.0997,
+      "step": 8962
+    },
+    {
+      "epoch": 2.815713528476353,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.823865439543004e-06,
+      "loss": 1.2184,
+      "step": 8964
+    },
+    {
+      "epoch": 2.8163417555018944,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.821326562995875e-06,
+      "loss": 1.1807,
+      "step": 8966
+    },
+    {
+      "epoch": 2.8169699825274357,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.818787686448747e-06,
+      "loss": 1.2236,
+      "step": 8968
+    },
+    {
+      "epoch": 2.817598209552977,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.81624880990162e-06,
+      "loss": 1.2124,
+      "step": 8970
+    },
+    {
+      "epoch": 2.8182264365785183,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.813709933354491e-06,
+      "loss": 1.1715,
+      "step": 8972
+    },
+    {
+      "epoch": 2.8188546636040597,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.811171056807364e-06,
+      "loss": 1.1341,
+      "step": 8974
+    },
+    {
+      "epoch": 2.819482890629601,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.808632180260236e-06,
+      "loss": 1.1516,
+      "step": 8976
+    },
+    {
+      "epoch": 2.8201111176551428,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.806093303713109e-06,
+      "loss": 1.2244,
+      "step": 8978
+    },
+    {
+      "epoch": 2.820739344680684,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.80355442716598e-06,
+      "loss": 1.0713,
+      "step": 8980
+    },
+    {
+      "epoch": 2.8213675717062254,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.801015550618852e-06,
+      "loss": 1.2261,
+      "step": 8982
+    },
+    {
+      "epoch": 2.8219957987317668,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.798476674071725e-06,
+      "loss": 1.2586,
+      "step": 8984
+    },
+    {
+      "epoch": 2.822624025757308,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.795937797524596e-06,
+      "loss": 1.1711,
+      "step": 8986
+    },
+    {
+      "epoch": 2.8232522527828494,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.793398920977469e-06,
+      "loss": 1.2093,
+      "step": 8988
+    },
+    {
+      "epoch": 2.8238804798083907,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.79086004443034e-06,
+      "loss": 1.2378,
+      "step": 8990
+    },
+    {
+      "epoch": 2.824508706833932,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.788321167883212e-06,
+      "loss": 1.1352,
+      "step": 8992
+    },
+    {
+      "epoch": 2.8251369338594734,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.785782291336085e-06,
+      "loss": 1.2086,
+      "step": 8994
+    },
+    {
+      "epoch": 2.8257651608850147,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.783243414788956e-06,
+      "loss": 1.1629,
+      "step": 8996
+    },
+    {
+      "epoch": 2.826393387910556,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.78070453824183e-06,
+      "loss": 1.1229,
+      "step": 8998
+    },
+    {
+      "epoch": 2.8270216149360974,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.7781656616947e-06,
+      "loss": 1.2158,
+      "step": 9000
+    },
+    {
+      "epoch": 2.827649841961639,
+      "grad_norm": 0.875,
+      "learning_rate": 8.775626785147572e-06,
+      "loss": 1.2396,
+      "step": 9002
+    },
+    {
+      "epoch": 2.8282780689871805,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.773087908600445e-06,
+      "loss": 1.1963,
+      "step": 9004
+    },
+    {
+      "epoch": 2.828906296012722,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.770549032053317e-06,
+      "loss": 1.2174,
+      "step": 9006
+    },
+    {
+      "epoch": 2.829534523038263,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.76801015550619e-06,
+      "loss": 1.1699,
+      "step": 9008
+    },
+    {
+      "epoch": 2.8301627500638045,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.765471278959061e-06,
+      "loss": 1.1734,
+      "step": 9010
+    },
+    {
+      "epoch": 2.830790977089346,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.762932402411934e-06,
+      "loss": 1.1106,
+      "step": 9012
+    },
+    {
+      "epoch": 2.831419204114887,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.760393525864806e-06,
+      "loss": 1.228,
+      "step": 9014
+    },
+    {
+      "epoch": 2.8320474311404285,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.757854649317677e-06,
+      "loss": 1.2233,
+      "step": 9016
+    },
+    {
+      "epoch": 2.83267565816597,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.75531577277055e-06,
+      "loss": 1.265,
+      "step": 9018
+    },
+    {
+      "epoch": 2.833303885191511,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.752776896223421e-06,
+      "loss": 1.2456,
+      "step": 9020
+    },
+    {
+      "epoch": 2.8339321122170524,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.750238019676294e-06,
+      "loss": 1.2257,
+      "step": 9022
+    },
+    {
+      "epoch": 2.8345603392425938,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.747699143129166e-06,
+      "loss": 1.1829,
+      "step": 9024
+    },
+    {
+      "epoch": 2.835188566268135,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.745160266582037e-06,
+      "loss": 1.2734,
+      "step": 9026
+    },
+    {
+      "epoch": 2.8358167932936764,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.74262139003491e-06,
+      "loss": 1.1946,
+      "step": 9028
+    },
+    {
+      "epoch": 2.8364450203192177,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.740082513487782e-06,
+      "loss": 1.3757,
+      "step": 9030
+    },
+    {
+      "epoch": 2.837073247344759,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.737543636940655e-06,
+      "loss": 1.2224,
+      "step": 9032
+    },
+    {
+      "epoch": 2.8377014743703004,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.735004760393526e-06,
+      "loss": 1.1388,
+      "step": 9034
+    },
+    {
+      "epoch": 2.8383297013958417,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.732465883846398e-06,
+      "loss": 1.1863,
+      "step": 9036
+    },
+    {
+      "epoch": 2.838957928421383,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.72992700729927e-06,
+      "loss": 1.2225,
+      "step": 9038
+    },
+    {
+      "epoch": 2.8395861554469244,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.727388130752142e-06,
+      "loss": 1.2286,
+      "step": 9040
+    },
+    {
+      "epoch": 2.8402143824724657,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.724849254205015e-06,
+      "loss": 1.086,
+      "step": 9042
+    },
+    {
+      "epoch": 2.8408426094980075,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.722310377657886e-06,
+      "loss": 1.159,
+      "step": 9044
+    },
+    {
+      "epoch": 2.841470836523549,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.719771501110758e-06,
+      "loss": 1.2525,
+      "step": 9046
+    },
+    {
+      "epoch": 2.84209906354909,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.717232624563633e-06,
+      "loss": 1.0467,
+      "step": 9048
+    },
+    {
+      "epoch": 2.8427272905746315,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.714693748016504e-06,
+      "loss": 1.3227,
+      "step": 9050
+    },
+    {
+      "epoch": 2.843355517600173,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.712154871469375e-06,
+      "loss": 1.0992,
+      "step": 9052
+    },
+    {
+      "epoch": 2.843983744625714,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.709615994922248e-06,
+      "loss": 1.1781,
+      "step": 9054
+    },
+    {
+      "epoch": 2.8446119716512555,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.70707711837512e-06,
+      "loss": 1.2496,
+      "step": 9056
+    },
+    {
+      "epoch": 2.845240198676797,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.704538241827993e-06,
+      "loss": 1.2135,
+      "step": 9058
+    },
+    {
+      "epoch": 2.845868425702338,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.701999365280864e-06,
+      "loss": 1.0489,
+      "step": 9060
+    },
+    {
+      "epoch": 2.8464966527278794,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.699460488733736e-06,
+      "loss": 1.0931,
+      "step": 9062
+    },
+    {
+      "epoch": 2.8471248797534208,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.696921612186609e-06,
+      "loss": 1.1408,
+      "step": 9064
+    },
+    {
+      "epoch": 2.847753106778962,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.69438273563948e-06,
+      "loss": 1.2734,
+      "step": 9066
+    },
+    {
+      "epoch": 2.848381333804504,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.691843859092353e-06,
+      "loss": 1.1622,
+      "step": 9068
+    },
+    {
+      "epoch": 2.849009560830045,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.689304982545225e-06,
+      "loss": 1.1151,
+      "step": 9070
+    },
+    {
+      "epoch": 2.8496377878555865,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.686766105998096e-06,
+      "loss": 1.0864,
+      "step": 9072
+    },
+    {
+      "epoch": 2.850266014881128,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.684227229450969e-06,
+      "loss": 1.1366,
+      "step": 9074
+    },
+    {
+      "epoch": 2.850894241906669,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.68168835290384e-06,
+      "loss": 1.1638,
+      "step": 9076
+    },
+    {
+      "epoch": 2.8515224689322105,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.679149476356713e-06,
+      "loss": 1.2531,
+      "step": 9078
+    },
+    {
+      "epoch": 2.852150695957752,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.676610599809585e-06,
+      "loss": 1.2749,
+      "step": 9080
+    },
+    {
+      "epoch": 2.852778922983293,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.674071723262458e-06,
+      "loss": 1.1556,
+      "step": 9082
+    },
+    {
+      "epoch": 2.8534071500088345,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.67153284671533e-06,
+      "loss": 1.129,
+      "step": 9084
+    },
+    {
+      "epoch": 2.854035377034376,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.6689939701682e-06,
+      "loss": 1.2725,
+      "step": 9086
+    },
+    {
+      "epoch": 2.854663604059917,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.666455093621074e-06,
+      "loss": 1.3089,
+      "step": 9088
+    },
+    {
+      "epoch": 2.8552918310854585,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.663916217073945e-06,
+      "loss": 1.2405,
+      "step": 9090
+    },
+    {
+      "epoch": 2.855920058111,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.661377340526818e-06,
+      "loss": 1.2605,
+      "step": 9092
+    },
+    {
+      "epoch": 2.856548285136541,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.65883846397969e-06,
+      "loss": 1.1111,
+      "step": 9094
+    },
+    {
+      "epoch": 2.8571765121620825,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.656299587432561e-06,
+      "loss": 1.1104,
+      "step": 9096
+    },
+    {
+      "epoch": 2.857804739187624,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.653760710885434e-06,
+      "loss": 1.2555,
+      "step": 9098
+    },
+    {
+      "epoch": 2.858432966213165,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.651221834338305e-06,
+      "loss": 1.2309,
+      "step": 9100
+    },
+    {
+      "epoch": 2.8590611932387064,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.648682957791179e-06,
+      "loss": 1.1012,
+      "step": 9102
+    },
+    {
+      "epoch": 2.8596894202642478,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.64614408124405e-06,
+      "loss": 1.1028,
+      "step": 9104
+    },
+    {
+      "epoch": 2.860317647289789,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.643605204696921e-06,
+      "loss": 1.2502,
+      "step": 9106
+    },
+    {
+      "epoch": 2.8609458743153304,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.641066328149794e-06,
+      "loss": 1.2413,
+      "step": 9108
+    },
+    {
+      "epoch": 2.861574101340872,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.638527451602666e-06,
+      "loss": 1.1771,
+      "step": 9110
+    },
+    {
+      "epoch": 2.8622023283664135,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.635988575055539e-06,
+      "loss": 1.2355,
+      "step": 9112
+    },
+    {
+      "epoch": 2.862830555391955,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.63344969850841e-06,
+      "loss": 1.1476,
+      "step": 9114
+    },
+    {
+      "epoch": 2.863458782417496,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.630910821961283e-06,
+      "loss": 1.1657,
+      "step": 9116
+    },
+    {
+      "epoch": 2.8640870094430375,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.628371945414155e-06,
+      "loss": 1.2089,
+      "step": 9118
+    },
+    {
+      "epoch": 2.864715236468579,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.625833068867026e-06,
+      "loss": 1.173,
+      "step": 9120
+    },
+    {
+      "epoch": 2.86534346349412,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.623294192319899e-06,
+      "loss": 1.2437,
+      "step": 9122
+    },
+    {
+      "epoch": 2.8659716905196615,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.62075531577277e-06,
+      "loss": 1.1699,
+      "step": 9124
+    },
+    {
+      "epoch": 2.866599917545203,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.618216439225644e-06,
+      "loss": 1.2851,
+      "step": 9126
+    },
+    {
+      "epoch": 2.867228144570744,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.615677562678515e-06,
+      "loss": 1.0717,
+      "step": 9128
+    },
+    {
+      "epoch": 2.8678563715962855,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.613138686131386e-06,
+      "loss": 0.9805,
+      "step": 9130
+    },
+    {
+      "epoch": 2.868484598621827,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.61059980958426e-06,
+      "loss": 1.1905,
+      "step": 9132
+    },
+    {
+      "epoch": 2.8691128256473686,
+      "grad_norm": 0.875,
+      "learning_rate": 8.608060933037133e-06,
+      "loss": 1.3264,
+      "step": 9134
+    },
+    {
+      "epoch": 2.86974105267291,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.605522056490004e-06,
+      "loss": 1.2117,
+      "step": 9136
+    },
+    {
+      "epoch": 2.8703692796984512,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.602983179942877e-06,
+      "loss": 1.1902,
+      "step": 9138
+    },
+    {
+      "epoch": 2.8709975067239926,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.600444303395748e-06,
+      "loss": 1.2074,
+      "step": 9140
+    },
+    {
+      "epoch": 2.871625733749534,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.597905426848621e-06,
+      "loss": 1.2359,
+      "step": 9142
+    },
+    {
+      "epoch": 2.872253960775075,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.595366550301493e-06,
+      "loss": 1.0976,
+      "step": 9144
+    },
+    {
+      "epoch": 2.8728821878006165,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.592827673754364e-06,
+      "loss": 1.4076,
+      "step": 9146
+    },
+    {
+      "epoch": 2.873510414826158,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.590288797207237e-06,
+      "loss": 1.2328,
+      "step": 9148
+    },
+    {
+      "epoch": 2.874138641851699,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.587749920660109e-06,
+      "loss": 1.1426,
+      "step": 9150
+    },
+    {
+      "epoch": 2.8747668688772405,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.585211044112982e-06,
+      "loss": 1.1638,
+      "step": 9152
+    },
+    {
+      "epoch": 2.875395095902782,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.582672167565853e-06,
+      "loss": 1.3359,
+      "step": 9154
+    },
+    {
+      "epoch": 2.876023322928323,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.580133291018724e-06,
+      "loss": 1.0855,
+      "step": 9156
+    },
+    {
+      "epoch": 2.8766515499538645,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.577594414471598e-06,
+      "loss": 1.2933,
+      "step": 9158
+    },
+    {
+      "epoch": 2.877279776979406,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.575055537924469e-06,
+      "loss": 1.1769,
+      "step": 9160
+    },
+    {
+      "epoch": 2.877908004004947,
+      "grad_norm": 0.875,
+      "learning_rate": 8.572516661377342e-06,
+      "loss": 1.296,
+      "step": 9162
+    },
+    {
+      "epoch": 2.8785362310304885,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.569977784830213e-06,
+      "loss": 1.2159,
+      "step": 9164
+    },
+    {
+      "epoch": 2.87916445805603,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.567438908283085e-06,
+      "loss": 1.1394,
+      "step": 9166
+    },
+    {
+      "epoch": 2.879792685081571,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.564900031735958e-06,
+      "loss": 1.1248,
+      "step": 9168
+    },
+    {
+      "epoch": 2.8804209121071125,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.56236115518883e-06,
+      "loss": 1.2397,
+      "step": 9170
+    },
+    {
+      "epoch": 2.881049139132654,
+      "grad_norm": 0.875,
+      "learning_rate": 8.559822278641702e-06,
+      "loss": 1.2171,
+      "step": 9172
+    },
+    {
+      "epoch": 2.881677366158195,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.557283402094574e-06,
+      "loss": 1.1612,
+      "step": 9174
+    },
+    {
+      "epoch": 2.882305593183737,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.554744525547447e-06,
+      "loss": 1.1409,
+      "step": 9176
+    },
+    {
+      "epoch": 2.8829338202092782,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.552205649000318e-06,
+      "loss": 1.1146,
+      "step": 9178
+    },
+    {
+      "epoch": 2.8835620472348196,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.54966677245319e-06,
+      "loss": 1.1532,
+      "step": 9180
+    },
+    {
+      "epoch": 2.884190274260361,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.547127895906063e-06,
+      "loss": 1.1232,
+      "step": 9182
+    },
+    {
+      "epoch": 2.8848185012859022,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.544589019358934e-06,
+      "loss": 1.1765,
+      "step": 9184
+    },
+    {
+      "epoch": 2.8854467283114436,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.542050142811807e-06,
+      "loss": 1.1835,
+      "step": 9186
+    },
+    {
+      "epoch": 2.886074955336985,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.539511266264678e-06,
+      "loss": 1.1912,
+      "step": 9188
+    },
+    {
+      "epoch": 2.886703182362526,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.53697238971755e-06,
+      "loss": 1.2664,
+      "step": 9190
+    },
+    {
+      "epoch": 2.8873314093880675,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.534433513170423e-06,
+      "loss": 1.0205,
+      "step": 9192
+    },
+    {
+      "epoch": 2.887959636413609,
+      "grad_norm": 0.875,
+      "learning_rate": 8.531894636623294e-06,
+      "loss": 1.0527,
+      "step": 9194
+    },
+    {
+      "epoch": 2.88858786343915,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.529355760076167e-06,
+      "loss": 1.2545,
+      "step": 9196
+    },
+    {
+      "epoch": 2.8892160904646915,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.526816883529039e-06,
+      "loss": 1.0928,
+      "step": 9198
+    },
+    {
+      "epoch": 2.8898443174902333,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.52427800698191e-06,
+      "loss": 1.209,
+      "step": 9200
+    },
+    {
+      "epoch": 2.8904725445157746,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.521739130434783e-06,
+      "loss": 1.171,
+      "step": 9202
+    },
+    {
+      "epoch": 2.891100771541316,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.519200253887655e-06,
+      "loss": 1.2062,
+      "step": 9204
+    },
+    {
+      "epoch": 2.8917289985668573,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.516661377340528e-06,
+      "loss": 1.2552,
+      "step": 9206
+    },
+    {
+      "epoch": 2.8923572255923986,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.514122500793399e-06,
+      "loss": 1.1772,
+      "step": 9208
+    },
+    {
+      "epoch": 2.89298545261794,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.51158362424627e-06,
+      "loss": 1.2433,
+      "step": 9210
+    },
+    {
+      "epoch": 2.8936136796434813,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.509044747699144e-06,
+      "loss": 1.1172,
+      "step": 9212
+    },
+    {
+      "epoch": 2.8942419066690226,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.506505871152015e-06,
+      "loss": 1.224,
+      "step": 9214
+    },
+    {
+      "epoch": 2.894870133694564,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.503966994604888e-06,
+      "loss": 1.1046,
+      "step": 9216
+    },
+    {
+      "epoch": 2.8954983607201052,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.50142811805776e-06,
+      "loss": 1.0652,
+      "step": 9218
+    },
+    {
+      "epoch": 2.8961265877456466,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.498889241510632e-06,
+      "loss": 1.1309,
+      "step": 9220
+    },
+    {
+      "epoch": 2.896754814771188,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.496350364963506e-06,
+      "loss": 1.0387,
+      "step": 9222
+    },
+    {
+      "epoch": 2.8973830417967292,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.493811488416377e-06,
+      "loss": 1.0663,
+      "step": 9224
+    },
+    {
+      "epoch": 2.8980112688222706,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.491272611869248e-06,
+      "loss": 1.2568,
+      "step": 9226
+    },
+    {
+      "epoch": 2.898639495847812,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.488733735322121e-06,
+      "loss": 1.2742,
+      "step": 9228
+    },
+    {
+      "epoch": 2.899267722873353,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.486194858774993e-06,
+      "loss": 1.1714,
+      "step": 9230
+    },
+    {
+      "epoch": 2.8998959498988945,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.483655982227866e-06,
+      "loss": 1.1513,
+      "step": 9232
+    },
+    {
+      "epoch": 2.900524176924436,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.481117105680737e-06,
+      "loss": 1.1691,
+      "step": 9234
+    },
+    {
+      "epoch": 2.901152403949977,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.478578229133609e-06,
+      "loss": 1.3591,
+      "step": 9236
+    },
+    {
+      "epoch": 2.9017806309755185,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.476039352586482e-06,
+      "loss": 1.2139,
+      "step": 9238
+    },
+    {
+      "epoch": 2.9024088580010603,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.473500476039353e-06,
+      "loss": 1.1475,
+      "step": 9240
+    },
+    {
+      "epoch": 2.9030370850266016,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.470961599492226e-06,
+      "loss": 1.1807,
+      "step": 9242
+    },
+    {
+      "epoch": 2.903665312052143,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.468422722945097e-06,
+      "loss": 1.1678,
+      "step": 9244
+    },
+    {
+      "epoch": 2.9042935390776843,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.46588384639797e-06,
+      "loss": 1.3143,
+      "step": 9246
+    },
+    {
+      "epoch": 2.9049217661032256,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.463344969850842e-06,
+      "loss": 1.1935,
+      "step": 9248
+    },
+    {
+      "epoch": 2.905549993128767,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.460806093303713e-06,
+      "loss": 1.3033,
+      "step": 9250
+    },
+    {
+      "epoch": 2.9061782201543083,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.458267216756586e-06,
+      "loss": 1.2278,
+      "step": 9252
+    },
+    {
+      "epoch": 2.9068064471798496,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.455728340209458e-06,
+      "loss": 1.2614,
+      "step": 9254
+    },
+    {
+      "epoch": 2.907434674205391,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.453189463662331e-06,
+      "loss": 1.2318,
+      "step": 9256
+    },
+    {
+      "epoch": 2.9080629012309323,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.450650587115202e-06,
+      "loss": 1.1713,
+      "step": 9258
+    },
+    {
+      "epoch": 2.9086911282564736,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.448111710568074e-06,
+      "loss": 1.2783,
+      "step": 9260
+    },
+    {
+      "epoch": 2.909319355282015,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.445572834020947e-06,
+      "loss": 1.1782,
+      "step": 9262
+    },
+    {
+      "epoch": 2.9099475823075562,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.443033957473818e-06,
+      "loss": 1.2585,
+      "step": 9264
+    },
+    {
+      "epoch": 2.910575809333098,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.440495080926691e-06,
+      "loss": 1.0162,
+      "step": 9266
+    },
+    {
+      "epoch": 2.9112040363586393,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.437956204379563e-06,
+      "loss": 1.1789,
+      "step": 9268
+    },
+    {
+      "epoch": 2.9118322633841807,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.435417327832434e-06,
+      "loss": 1.1599,
+      "step": 9270
+    },
+    {
+      "epoch": 2.912460490409722,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.432878451285307e-06,
+      "loss": 1.2784,
+      "step": 9272
+    },
+    {
+      "epoch": 2.9130887174352633,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.430339574738178e-06,
+      "loss": 1.2503,
+      "step": 9274
+    },
+    {
+      "epoch": 2.9137169444608046,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.427800698191051e-06,
+      "loss": 1.2095,
+      "step": 9276
+    },
+    {
+      "epoch": 2.914345171486346,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.425261821643923e-06,
+      "loss": 1.1473,
+      "step": 9278
+    },
+    {
+      "epoch": 2.9149733985118873,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.422722945096796e-06,
+      "loss": 1.2989,
+      "step": 9280
+    },
+    {
+      "epoch": 2.9156016255374286,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.420184068549667e-06,
+      "loss": 1.1785,
+      "step": 9282
+    },
+    {
+      "epoch": 2.91622985256297,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.417645192002539e-06,
+      "loss": 1.1599,
+      "step": 9284
+    },
+    {
+      "epoch": 2.9168580795885113,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.415106315455412e-06,
+      "loss": 1.1402,
+      "step": 9286
+    },
+    {
+      "epoch": 2.9174863066140526,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.412567438908283e-06,
+      "loss": 1.056,
+      "step": 9288
+    },
+    {
+      "epoch": 2.918114533639594,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.410028562361156e-06,
+      "loss": 1.1068,
+      "step": 9290
+    },
+    {
+      "epoch": 2.9187427606651353,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.407489685814028e-06,
+      "loss": 1.1869,
+      "step": 9292
+    },
+    {
+      "epoch": 2.9193709876906766,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.404950809266899e-06,
+      "loss": 1.145,
+      "step": 9294
+    },
+    {
+      "epoch": 2.919999214716218,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.402411932719772e-06,
+      "loss": 1.1293,
+      "step": 9296
+    },
+    {
+      "epoch": 2.9206274417417593,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.399873056172643e-06,
+      "loss": 1.1936,
+      "step": 9298
+    },
+    {
+      "epoch": 2.9212556687673006,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.397334179625517e-06,
+      "loss": 1.1045,
+      "step": 9300
+    },
+    {
+      "epoch": 2.921883895792842,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.394795303078388e-06,
+      "loss": 1.2715,
+      "step": 9302
+    },
+    {
+      "epoch": 2.9225121228183832,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.392256426531261e-06,
+      "loss": 1.1606,
+      "step": 9304
+    },
+    {
+      "epoch": 2.923140349843925,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.389717549984134e-06,
+      "loss": 1.0997,
+      "step": 9306
+    },
+    {
+      "epoch": 2.9237685768694663,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.387178673437005e-06,
+      "loss": 1.0984,
+      "step": 9308
+    },
+    {
+      "epoch": 2.9243968038950077,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.384639796889877e-06,
+      "loss": 1.2038,
+      "step": 9310
+    },
+    {
+      "epoch": 2.925025030920549,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.38210092034275e-06,
+      "loss": 1.0816,
+      "step": 9312
+    },
+    {
+      "epoch": 2.9256532579460903,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.379562043795621e-06,
+      "loss": 1.2256,
+      "step": 9314
+    },
+    {
+      "epoch": 2.9262814849716317,
+      "grad_norm": 0.875,
+      "learning_rate": 8.377023167248494e-06,
+      "loss": 1.2699,
+      "step": 9316
+    },
+    {
+      "epoch": 2.926909711997173,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.374484290701366e-06,
+      "loss": 1.1746,
+      "step": 9318
+    },
+    {
+      "epoch": 2.9275379390227143,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.371945414154237e-06,
+      "loss": 1.1686,
+      "step": 9320
+    },
+    {
+      "epoch": 2.9281661660482556,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.36940653760711e-06,
+      "loss": 1.1485,
+      "step": 9322
+    },
+    {
+      "epoch": 2.928794393073797,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.366867661059982e-06,
+      "loss": 1.2353,
+      "step": 9324
+    },
+    {
+      "epoch": 2.9294226200993383,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.364328784512855e-06,
+      "loss": 1.2038,
+      "step": 9326
+    },
+    {
+      "epoch": 2.9300508471248796,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.361789907965726e-06,
+      "loss": 1.2617,
+      "step": 9328
+    },
+    {
+      "epoch": 2.930679074150421,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.359251031418597e-06,
+      "loss": 1.1843,
+      "step": 9330
+    },
+    {
+      "epoch": 2.9313073011759627,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.35671215487147e-06,
+      "loss": 1.217,
+      "step": 9332
+    },
+    {
+      "epoch": 2.931935528201504,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.354173278324342e-06,
+      "loss": 1.1822,
+      "step": 9334
+    },
+    {
+      "epoch": 2.9325637552270454,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.351634401777215e-06,
+      "loss": 1.1071,
+      "step": 9336
+    },
+    {
+      "epoch": 2.9331919822525867,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.349095525230086e-06,
+      "loss": 1.1019,
+      "step": 9338
+    },
+    {
+      "epoch": 2.933820209278128,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.346556648682958e-06,
+      "loss": 1.1753,
+      "step": 9340
+    },
+    {
+      "epoch": 2.9344484363036694,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.34401777213583e-06,
+      "loss": 1.1814,
+      "step": 9342
+    },
+    {
+      "epoch": 2.9350766633292107,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.341478895588702e-06,
+      "loss": 1.1398,
+      "step": 9344
+    },
+    {
+      "epoch": 2.935704890354752,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.338940019041575e-06,
+      "loss": 1.2175,
+      "step": 9346
+    },
+    {
+      "epoch": 2.9363331173802933,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.336401142494447e-06,
+      "loss": 1.057,
+      "step": 9348
+    },
+    {
+      "epoch": 2.9369613444058347,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.33386226594732e-06,
+      "loss": 1.2624,
+      "step": 9350
+    },
+    {
+      "epoch": 2.937589571431376,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.331323389400191e-06,
+      "loss": 1.1828,
+      "step": 9352
+    },
+    {
+      "epoch": 2.9382177984569173,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.328784512853062e-06,
+      "loss": 1.2289,
+      "step": 9354
+    },
+    {
+      "epoch": 2.9388460254824587,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.326245636305936e-06,
+      "loss": 1.0668,
+      "step": 9356
+    },
+    {
+      "epoch": 2.939474252508,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.323706759758807e-06,
+      "loss": 1.212,
+      "step": 9358
+    },
+    {
+      "epoch": 2.9401024795335413,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.32116788321168e-06,
+      "loss": 1.1254,
+      "step": 9360
+    },
+    {
+      "epoch": 2.9407307065590826,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.318629006664551e-06,
+      "loss": 1.048,
+      "step": 9362
+    },
+    {
+      "epoch": 2.941358933584624,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.316090130117423e-06,
+      "loss": 1.0829,
+      "step": 9364
+    },
+    {
+      "epoch": 2.9419871606101653,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.313551253570296e-06,
+      "loss": 1.163,
+      "step": 9366
+    },
+    {
+      "epoch": 2.9426153876357066,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.311012377023167e-06,
+      "loss": 1.3302,
+      "step": 9368
+    },
+    {
+      "epoch": 2.943243614661248,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.30847350047604e-06,
+      "loss": 1.115,
+      "step": 9370
+    },
+    {
+      "epoch": 2.9438718416867897,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.305934623928912e-06,
+      "loss": 1.218,
+      "step": 9372
+    },
+    {
+      "epoch": 2.944500068712331,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.303395747381783e-06,
+      "loss": 1.17,
+      "step": 9374
+    },
+    {
+      "epoch": 2.9451282957378724,
+      "grad_norm": 0.875,
+      "learning_rate": 8.300856870834656e-06,
+      "loss": 1.2241,
+      "step": 9376
+    },
+    {
+      "epoch": 2.9457565227634137,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.298317994287528e-06,
+      "loss": 1.1651,
+      "step": 9378
+    },
+    {
+      "epoch": 2.946384749788955,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.2957791177404e-06,
+      "loss": 1.1871,
+      "step": 9380
+    },
+    {
+      "epoch": 2.9470129768144964,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.293240241193272e-06,
+      "loss": 1.2128,
+      "step": 9382
+    },
+    {
+      "epoch": 2.9476412038400377,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.290701364646145e-06,
+      "loss": 1.2018,
+      "step": 9384
+    },
+    {
+      "epoch": 2.948269430865579,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.288162488099016e-06,
+      "loss": 1.2859,
+      "step": 9386
+    },
+    {
+      "epoch": 2.9488976578911203,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.285623611551888e-06,
+      "loss": 1.2422,
+      "step": 9388
+    },
+    {
+      "epoch": 2.9495258849166617,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.283084735004761e-06,
+      "loss": 1.1708,
+      "step": 9390
+    },
+    {
+      "epoch": 2.950154111942203,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.280545858457634e-06,
+      "loss": 1.1229,
+      "step": 9392
+    },
+    {
+      "epoch": 2.9507823389677443,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.278006981910505e-06,
+      "loss": 1.117,
+      "step": 9394
+    },
+    {
+      "epoch": 2.9514105659932857,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.275468105363378e-06,
+      "loss": 1.3889,
+      "step": 9396
+    },
+    {
+      "epoch": 2.9520387930188274,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.27292922881625e-06,
+      "loss": 1.1619,
+      "step": 9398
+    },
+    {
+      "epoch": 2.9526670200443688,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.270390352269121e-06,
+      "loss": 1.1436,
+      "step": 9400
+    },
+    {
+      "epoch": 2.95329524706991,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.267851475721994e-06,
+      "loss": 1.1739,
+      "step": 9402
+    },
+    {
+      "epoch": 2.9539234740954514,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.265312599174866e-06,
+      "loss": 1.1537,
+      "step": 9404
+    },
+    {
+      "epoch": 2.9545517011209927,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.262773722627739e-06,
+      "loss": 1.1615,
+      "step": 9406
+    },
+    {
+      "epoch": 2.955179928146534,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.26023484608061e-06,
+      "loss": 1.2431,
+      "step": 9408
+    },
+    {
+      "epoch": 2.9558081551720754,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.257695969533483e-06,
+      "loss": 1.1621,
+      "step": 9410
+    },
+    {
+      "epoch": 2.9564363821976167,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.255157092986355e-06,
+      "loss": 1.2451,
+      "step": 9412
+    },
+    {
+      "epoch": 2.957064609223158,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.252618216439226e-06,
+      "loss": 1.1724,
+      "step": 9414
+    },
+    {
+      "epoch": 2.9576928362486994,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.250079339892099e-06,
+      "loss": 1.095,
+      "step": 9416
+    },
+    {
+      "epoch": 2.9583210632742407,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.24754046334497e-06,
+      "loss": 1.1788,
+      "step": 9418
+    },
+    {
+      "epoch": 2.958949290299782,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.245001586797843e-06,
+      "loss": 1.1435,
+      "step": 9420
+    },
+    {
+      "epoch": 2.9595775173253234,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.242462710250715e-06,
+      "loss": 1.2214,
+      "step": 9422
+    },
+    {
+      "epoch": 2.9602057443508647,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.239923833703586e-06,
+      "loss": 1.1846,
+      "step": 9424
+    },
+    {
+      "epoch": 2.960833971376406,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.23738495715646e-06,
+      "loss": 1.1564,
+      "step": 9426
+    },
+    {
+      "epoch": 2.9614621984019474,
+      "grad_norm": 0.875,
+      "learning_rate": 8.23484608060933e-06,
+      "loss": 1.1448,
+      "step": 9428
+    },
+    {
+      "epoch": 2.9620904254274887,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.232307204062204e-06,
+      "loss": 1.3647,
+      "step": 9430
+    },
+    {
+      "epoch": 2.96271865245303,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.229768327515075e-06,
+      "loss": 1.221,
+      "step": 9432
+    },
+    {
+      "epoch": 2.9633468794785713,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.227229450967947e-06,
+      "loss": 1.2453,
+      "step": 9434
+    },
+    {
+      "epoch": 2.9639751065041127,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.22469057442082e-06,
+      "loss": 1.1119,
+      "step": 9436
+    },
+    {
+      "epoch": 2.9646033335296544,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.222151697873691e-06,
+      "loss": 1.2253,
+      "step": 9438
+    },
+    {
+      "epoch": 2.9652315605551958,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.219612821326564e-06,
+      "loss": 1.191,
+      "step": 9440
+    },
+    {
+      "epoch": 2.965859787580737,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.217073944779435e-06,
+      "loss": 1.1427,
+      "step": 9442
+    },
+    {
+      "epoch": 2.9664880146062784,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.214535068232309e-06,
+      "loss": 1.2054,
+      "step": 9444
+    },
+    {
+      "epoch": 2.9671162416318198,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.21199619168518e-06,
+      "loss": 1.2363,
+      "step": 9446
+    },
+    {
+      "epoch": 2.967744468657361,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.209457315138051e-06,
+      "loss": 1.3343,
+      "step": 9448
+    },
+    {
+      "epoch": 2.9683726956829024,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.206918438590924e-06,
+      "loss": 1.0634,
+      "step": 9450
+    },
+    {
+      "epoch": 2.9690009227084437,
+      "grad_norm": 0.8125,
+      "learning_rate": 8.204379562043796e-06,
+      "loss": 1.1639,
+      "step": 9452
+    },
+    {
+      "epoch": 2.969629149733985,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.201840685496669e-06,
+      "loss": 1.1921,
+      "step": 9454
+    },
+    {
+      "epoch": 2.9702573767595264,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.19930180894954e-06,
+      "loss": 1.3581,
+      "step": 9456
+    },
+    {
+      "epoch": 2.9708856037850677,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.196762932402412e-06,
+      "loss": 1.2369,
+      "step": 9458
+    },
+    {
+      "epoch": 2.971513830810609,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.194224055855285e-06,
+      "loss": 1.0998,
+      "step": 9460
+    },
+    {
+      "epoch": 2.9721420578361504,
+      "grad_norm": 1.375,
+      "learning_rate": 8.191685179308156e-06,
+      "loss": 1.0054,
+      "step": 9462
+    },
+    {
+      "epoch": 2.972770284861692,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.189146302761029e-06,
+      "loss": 1.2289,
+      "step": 9464
+    },
+    {
+      "epoch": 2.9733985118872335,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.1866074262139e-06,
+      "loss": 1.0944,
+      "step": 9466
+    },
+    {
+      "epoch": 2.974026738912775,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.184068549666772e-06,
+      "loss": 1.168,
+      "step": 9468
+    },
+    {
+      "epoch": 2.974654965938316,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.181529673119645e-06,
+      "loss": 1.2147,
+      "step": 9470
+    },
+    {
+      "epoch": 2.9752831929638575,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.178990796572516e-06,
+      "loss": 1.2055,
+      "step": 9472
+    },
+    {
+      "epoch": 2.975911419989399,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.17645192002539e-06,
+      "loss": 1.2641,
+      "step": 9474
+    },
+    {
+      "epoch": 2.97653964701494,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.173913043478263e-06,
+      "loss": 1.2016,
+      "step": 9476
+    },
+    {
+      "epoch": 2.9771678740404814,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.171374166931134e-06,
+      "loss": 1.2389,
+      "step": 9478
+    },
+    {
+      "epoch": 2.9777961010660228,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.168835290384007e-06,
+      "loss": 1.3398,
+      "step": 9480
+    },
+    {
+      "epoch": 2.978424328091564,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.166296413836878e-06,
+      "loss": 1.3117,
+      "step": 9482
+    },
+    {
+      "epoch": 2.9790525551171054,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.16375753728975e-06,
+      "loss": 1.1342,
+      "step": 9484
+    },
+    {
+      "epoch": 2.9796807821426468,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.161218660742623e-06,
+      "loss": 1.183,
+      "step": 9486
+    },
+    {
+      "epoch": 2.980309009168188,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.158679784195494e-06,
+      "loss": 1.1521,
+      "step": 9488
+    },
+    {
+      "epoch": 2.9809372361937294,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.156140907648367e-06,
+      "loss": 1.0698,
+      "step": 9490
+    },
+    {
+      "epoch": 2.9815654632192707,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.153602031101239e-06,
+      "loss": 1.1782,
+      "step": 9492
+    },
+    {
+      "epoch": 2.982193690244812,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.15106315455411e-06,
+      "loss": 1.2666,
+      "step": 9494
+    },
+    {
+      "epoch": 2.9828219172703534,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.148524278006983e-06,
+      "loss": 1.3233,
+      "step": 9496
+    },
+    {
+      "epoch": 2.9834501442958947,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.145985401459855e-06,
+      "loss": 1.1066,
+      "step": 9498
+    },
+    {
+      "epoch": 2.984078371321436,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.143446524912728e-06,
+      "loss": 1.1686,
+      "step": 9500
+    }
+  ],
+  "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": 2.572975921220813e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}