diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6082 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 0,
+  "global_step": 864,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011574074074074073,
+      "grad_norm": 0.22213494777679443,
+      "learning_rate": 1e-05,
+      "loss": 1.5076,
+      "step": 1
+    },
+    {
+      "epoch": 0.0023148148148148147,
+      "grad_norm": 0.22815562784671783,
+      "learning_rate": 9.988425925925927e-06,
+      "loss": 1.4784,
+      "step": 2
+    },
+    {
+      "epoch": 0.003472222222222222,
+      "grad_norm": 0.22811558842658997,
+      "learning_rate": 9.976851851851853e-06,
+      "loss": 1.4667,
+      "step": 3
+    },
+    {
+      "epoch": 0.004629629629629629,
+      "grad_norm": 0.2362716645002365,
+      "learning_rate": 9.965277777777778e-06,
+      "loss": 1.5482,
+      "step": 4
+    },
+    {
+      "epoch": 0.005787037037037037,
+      "grad_norm": 0.2334546148777008,
+      "learning_rate": 9.953703703703704e-06,
+      "loss": 1.5051,
+      "step": 5
+    },
+    {
+      "epoch": 0.006944444444444444,
+      "grad_norm": 0.224368616938591,
+      "learning_rate": 9.942129629629629e-06,
+      "loss": 1.4836,
+      "step": 6
+    },
+    {
+      "epoch": 0.008101851851851851,
+      "grad_norm": 0.24275268614292145,
+      "learning_rate": 9.930555555555557e-06,
+      "loss": 1.5993,
+      "step": 7
+    },
+    {
+      "epoch": 0.009259259259259259,
+      "grad_norm": 0.24044257402420044,
+      "learning_rate": 9.918981481481482e-06,
+      "loss": 1.4207,
+      "step": 8
+    },
+    {
+      "epoch": 0.010416666666666666,
+      "grad_norm": 0.2276177853345871,
+      "learning_rate": 9.907407407407408e-06,
+      "loss": 1.479,
+      "step": 9
+    },
+    {
+      "epoch": 0.011574074074074073,
+      "grad_norm": 0.23464104533195496,
+      "learning_rate": 9.895833333333334e-06,
+      "loss": 1.5192,
+      "step": 10
+    },
+    {
+      "epoch": 0.01273148148148148,
+      "grad_norm": 0.2245517075061798,
+      "learning_rate": 9.88425925925926e-06,
+      "loss": 1.4462,
+      "step": 11
+    },
+    {
+      "epoch": 0.013888888888888888,
+      "grad_norm": 0.4287811517715454,
+      "learning_rate": 9.872685185185185e-06,
+      "loss": 1.5524,
+      "step": 12
+    },
+    {
+      "epoch": 0.015046296296296295,
+      "grad_norm": 0.23612529039382935,
+      "learning_rate": 9.861111111111112e-06,
+      "loss": 1.4822,
+      "step": 13
+    },
+    {
+      "epoch": 0.016203703703703703,
+      "grad_norm": 0.19577904045581818,
+      "learning_rate": 9.849537037037038e-06,
+      "loss": 1.3985,
+      "step": 14
+    },
+    {
+      "epoch": 0.017361111111111112,
+      "grad_norm": 0.19019098579883575,
+      "learning_rate": 9.837962962962964e-06,
+      "loss": 1.3753,
+      "step": 15
+    },
+    {
+      "epoch": 0.018518518518518517,
+      "grad_norm": 0.21135130524635315,
+      "learning_rate": 9.826388888888889e-06,
+      "loss": 1.5148,
+      "step": 16
+    },
+    {
+      "epoch": 0.019675925925925927,
+      "grad_norm": 0.19131916761398315,
+      "learning_rate": 9.814814814814815e-06,
+      "loss": 1.5473,
+      "step": 17
+    },
+    {
+      "epoch": 0.020833333333333332,
+      "grad_norm": 0.1860252171754837,
+      "learning_rate": 9.803240740740742e-06,
+      "loss": 1.5053,
+      "step": 18
+    },
+    {
+      "epoch": 0.02199074074074074,
+      "grad_norm": 0.17170019447803497,
+      "learning_rate": 9.791666666666666e-06,
+      "loss": 1.5025,
+      "step": 19
+    },
+    {
+      "epoch": 0.023148148148148147,
+      "grad_norm": 0.16492699086666107,
+      "learning_rate": 9.780092592592594e-06,
+      "loss": 1.3922,
+      "step": 20
+    },
+    {
+      "epoch": 0.024305555555555556,
+      "grad_norm": 0.15029262006282806,
+      "learning_rate": 9.768518518518519e-06,
+      "loss": 1.44,
+      "step": 21
+    },
+    {
+      "epoch": 0.02546296296296296,
+      "grad_norm": 0.13995124399662018,
+      "learning_rate": 9.756944444444445e-06,
+      "loss": 1.4524,
+      "step": 22
+    },
+    {
+      "epoch": 0.02662037037037037,
+      "grad_norm": 0.14475427567958832,
+      "learning_rate": 9.745370370370372e-06,
+      "loss": 1.4576,
+      "step": 23
+    },
+    {
+      "epoch": 0.027777777777777776,
+      "grad_norm": 0.15014007687568665,
+      "learning_rate": 9.733796296296298e-06,
+      "loss": 1.4862,
+      "step": 24
+    },
+    {
+      "epoch": 0.028935185185185185,
+      "grad_norm": 0.13555286824703217,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 1.4582,
+      "step": 25
+    },
+    {
+      "epoch": 0.03009259259259259,
+      "grad_norm": 0.15213678777217865,
+      "learning_rate": 9.710648148148149e-06,
+      "loss": 1.4877,
+      "step": 26
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 0.13051781058311462,
+      "learning_rate": 9.699074074074075e-06,
+      "loss": 1.3628,
+      "step": 27
+    },
+    {
+      "epoch": 0.032407407407407406,
+      "grad_norm": 0.15080274641513824,
+      "learning_rate": 9.6875e-06,
+      "loss": 1.4834,
+      "step": 28
+    },
+    {
+      "epoch": 0.03356481481481482,
+      "grad_norm": 0.1410171389579773,
+      "learning_rate": 9.675925925925926e-06,
+      "loss": 1.5207,
+      "step": 29
+    },
+    {
+      "epoch": 0.034722222222222224,
+      "grad_norm": 0.14558039605617523,
+      "learning_rate": 9.664351851851853e-06,
+      "loss": 1.4201,
+      "step": 30
+    },
+    {
+      "epoch": 0.03587962962962963,
+      "grad_norm": 0.1254415363073349,
+      "learning_rate": 9.652777777777779e-06,
+      "loss": 1.3993,
+      "step": 31
+    },
+    {
+      "epoch": 0.037037037037037035,
+      "grad_norm": 0.14481398463249207,
+      "learning_rate": 9.641203703703704e-06,
+      "loss": 1.3278,
+      "step": 32
+    },
+    {
+      "epoch": 0.03819444444444445,
+      "grad_norm": 0.1267780363559723,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 1.3886,
+      "step": 33
+    },
+    {
+      "epoch": 0.03935185185185185,
+      "grad_norm": 0.14567571878433228,
+      "learning_rate": 9.618055555555556e-06,
+      "loss": 1.3612,
+      "step": 34
+    },
+    {
+      "epoch": 0.04050925925925926,
+      "grad_norm": 0.14164520800113678,
+      "learning_rate": 9.606481481481483e-06,
+      "loss": 1.339,
+      "step": 35
+    },
+    {
+      "epoch": 0.041666666666666664,
+      "grad_norm": 0.12852604687213898,
+      "learning_rate": 9.594907407407407e-06,
+      "loss": 1.4468,
+      "step": 36
+    },
+    {
+      "epoch": 0.04282407407407408,
+      "grad_norm": 0.14108137786388397,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 1.3848,
+      "step": 37
+    },
+    {
+      "epoch": 0.04398148148148148,
+      "grad_norm": 0.1143556609749794,
+      "learning_rate": 9.57175925925926e-06,
+      "loss": 1.344,
+      "step": 38
+    },
+    {
+      "epoch": 0.04513888888888889,
+      "grad_norm": 0.11572553217411041,
+      "learning_rate": 9.560185185185186e-06,
+      "loss": 1.331,
+      "step": 39
+    },
+    {
+      "epoch": 0.046296296296296294,
+      "grad_norm": 0.1106305718421936,
+      "learning_rate": 9.548611111111113e-06,
+      "loss": 1.4159,
+      "step": 40
+    },
+    {
+      "epoch": 0.047453703703703706,
+      "grad_norm": 0.11018543690443039,
+      "learning_rate": 9.537037037037037e-06,
+      "loss": 1.3817,
+      "step": 41
+    },
+    {
+      "epoch": 0.04861111111111111,
+      "grad_norm": 0.11123362183570862,
+      "learning_rate": 9.525462962962964e-06,
+      "loss": 1.3859,
+      "step": 42
+    },
+    {
+      "epoch": 0.04976851851851852,
+      "grad_norm": 0.11585158854722977,
+      "learning_rate": 9.51388888888889e-06,
+      "loss": 1.4172,
+      "step": 43
+    },
+    {
+      "epoch": 0.05092592592592592,
+      "grad_norm": 0.11234026402235031,
+      "learning_rate": 9.502314814814816e-06,
+      "loss": 1.4225,
+      "step": 44
+    },
+    {
+      "epoch": 0.052083333333333336,
+      "grad_norm": 0.17347489297389984,
+      "learning_rate": 9.490740740740741e-06,
+      "loss": 1.3333,
+      "step": 45
+    },
+    {
+      "epoch": 0.05324074074074074,
+      "grad_norm": 0.10686583071947098,
+      "learning_rate": 9.479166666666667e-06,
+      "loss": 1.3721,
+      "step": 46
+    },
+    {
+      "epoch": 0.05439814814814815,
+      "grad_norm": 0.11265669763088226,
+      "learning_rate": 9.467592592592594e-06,
+      "loss": 1.4224,
+      "step": 47
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 0.10087595134973526,
+      "learning_rate": 9.45601851851852e-06,
+      "loss": 1.3365,
+      "step": 48
+    },
+    {
+      "epoch": 0.056712962962962965,
+      "grad_norm": 0.10566269606351852,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 1.366,
+      "step": 49
+    },
+    {
+      "epoch": 0.05787037037037037,
+      "grad_norm": 0.11439696699380875,
+      "learning_rate": 9.432870370370371e-06,
+      "loss": 1.4139,
+      "step": 50
+    },
+    {
+      "epoch": 0.059027777777777776,
+      "grad_norm": 0.10534631460905075,
+      "learning_rate": 9.421296296296297e-06,
+      "loss": 1.3832,
+      "step": 51
+    },
+    {
+      "epoch": 0.06018518518518518,
+      "grad_norm": 0.09864930808544159,
+      "learning_rate": 9.409722222222224e-06,
+      "loss": 1.3692,
+      "step": 52
+    },
+    {
+      "epoch": 0.061342592592592594,
+      "grad_norm": 0.10979383438825607,
+      "learning_rate": 9.398148148148148e-06,
+      "loss": 1.4028,
+      "step": 53
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 0.11381463706493378,
+      "learning_rate": 9.386574074074075e-06,
+      "loss": 1.4229,
+      "step": 54
+    },
+    {
+      "epoch": 0.06365740740740741,
+      "grad_norm": 0.10466167330741882,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.3767,
+      "step": 55
+    },
+    {
+      "epoch": 0.06481481481481481,
+      "grad_norm": 0.10089318454265594,
+      "learning_rate": 9.363425925925927e-06,
+      "loss": 1.3755,
+      "step": 56
+    },
+    {
+      "epoch": 0.06597222222222222,
+      "grad_norm": 0.1125788539648056,
+      "learning_rate": 9.351851851851854e-06,
+      "loss": 1.3654,
+      "step": 57
+    },
+    {
+      "epoch": 0.06712962962962964,
+      "grad_norm": 0.10320808738470078,
+      "learning_rate": 9.340277777777778e-06,
+      "loss": 1.3788,
+      "step": 58
+    },
+    {
+      "epoch": 0.06828703703703703,
+      "grad_norm": 0.10044930130243301,
+      "learning_rate": 9.328703703703705e-06,
+      "loss": 1.3912,
+      "step": 59
+    },
+    {
+      "epoch": 0.06944444444444445,
+      "grad_norm": 0.10364623367786407,
+      "learning_rate": 9.31712962962963e-06,
+      "loss": 1.2838,
+      "step": 60
+    },
+    {
+      "epoch": 0.07060185185185185,
+      "grad_norm": 0.14089667797088623,
+      "learning_rate": 9.305555555555557e-06,
+      "loss": 1.3711,
+      "step": 61
+    },
+    {
+      "epoch": 0.07175925925925926,
+      "grad_norm": 0.10308589786291122,
+      "learning_rate": 9.293981481481482e-06,
+      "loss": 1.2889,
+      "step": 62
+    },
+    {
+      "epoch": 0.07291666666666667,
+      "grad_norm": 0.1022154688835144,
+      "learning_rate": 9.282407407407408e-06,
+      "loss": 1.3387,
+      "step": 63
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 0.09682654589414597,
+      "learning_rate": 9.270833333333334e-06,
+      "loss": 1.2755,
+      "step": 64
+    },
+    {
+      "epoch": 0.07523148148148148,
+      "grad_norm": 0.10220059752464294,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 1.2953,
+      "step": 65
+    },
+    {
+      "epoch": 0.0763888888888889,
+      "grad_norm": 0.09950669854879379,
+      "learning_rate": 9.247685185185185e-06,
+      "loss": 1.3319,
+      "step": 66
+    },
+    {
+      "epoch": 0.0775462962962963,
+      "grad_norm": 0.09631174057722092,
+      "learning_rate": 9.236111111111112e-06,
+      "loss": 1.3037,
+      "step": 67
+    },
+    {
+      "epoch": 0.0787037037037037,
+      "grad_norm": 0.10525327175855637,
+      "learning_rate": 9.224537037037038e-06,
+      "loss": 1.3669,
+      "step": 68
+    },
+    {
+      "epoch": 0.0798611111111111,
+      "grad_norm": 0.1023765280842781,
+      "learning_rate": 9.212962962962963e-06,
+      "loss": 1.3172,
+      "step": 69
+    },
+    {
+      "epoch": 0.08101851851851852,
+      "grad_norm": 0.09824973344802856,
+      "learning_rate": 9.201388888888889e-06,
+      "loss": 1.2471,
+      "step": 70
+    },
+    {
+      "epoch": 0.08217592592592593,
+      "grad_norm": 0.09957920759916306,
+      "learning_rate": 9.189814814814815e-06,
+      "loss": 1.2974,
+      "step": 71
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 0.10006606578826904,
+      "learning_rate": 9.178240740740742e-06,
+      "loss": 1.2478,
+      "step": 72
+    },
+    {
+      "epoch": 0.08449074074074074,
+      "grad_norm": 0.09730537980794907,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 1.2874,
+      "step": 73
+    },
+    {
+      "epoch": 0.08564814814814815,
+      "grad_norm": 0.10841472446918488,
+      "learning_rate": 9.155092592592594e-06,
+      "loss": 1.3142,
+      "step": 74
+    },
+    {
+      "epoch": 0.08680555555555555,
+      "grad_norm": 0.10033254325389862,
+      "learning_rate": 9.143518518518519e-06,
+      "loss": 1.3031,
+      "step": 75
+    },
+    {
+      "epoch": 0.08796296296296297,
+      "grad_norm": 0.10729570686817169,
+      "learning_rate": 9.131944444444445e-06,
+      "loss": 1.318,
+      "step": 76
+    },
+    {
+      "epoch": 0.08912037037037036,
+      "grad_norm": 0.10036814212799072,
+      "learning_rate": 9.120370370370372e-06,
+      "loss": 1.315,
+      "step": 77
+    },
+    {
+      "epoch": 0.09027777777777778,
+      "grad_norm": 0.09904678165912628,
+      "learning_rate": 9.108796296296296e-06,
+      "loss": 1.3105,
+      "step": 78
+    },
+    {
+      "epoch": 0.09143518518518519,
+      "grad_norm": 0.10250908136367798,
+      "learning_rate": 9.097222222222223e-06,
+      "loss": 1.3388,
+      "step": 79
+    },
+    {
+      "epoch": 0.09259259259259259,
+      "grad_norm": 0.10013310611248016,
+      "learning_rate": 9.085648148148149e-06,
+      "loss": 1.3081,
+      "step": 80
+    },
+    {
+      "epoch": 0.09375,
+      "grad_norm": 0.12040674686431885,
+      "learning_rate": 9.074074074074075e-06,
+      "loss": 1.29,
+      "step": 81
+    },
+    {
+      "epoch": 0.09490740740740741,
+      "grad_norm": 0.10173387080430984,
+      "learning_rate": 9.0625e-06,
+      "loss": 1.3216,
+      "step": 82
+    },
+    {
+      "epoch": 0.09606481481481481,
+      "grad_norm": 0.0995723232626915,
+      "learning_rate": 9.050925925925926e-06,
+      "loss": 1.2348,
+      "step": 83
+    },
+    {
+      "epoch": 0.09722222222222222,
+      "grad_norm": 0.09998712688684464,
+      "learning_rate": 9.039351851851853e-06,
+      "loss": 1.2941,
+      "step": 84
+    },
+    {
+      "epoch": 0.09837962962962964,
+      "grad_norm": 0.10181009769439697,
+      "learning_rate": 9.027777777777779e-06,
+      "loss": 1.3435,
+      "step": 85
+    },
+    {
+      "epoch": 0.09953703703703703,
+      "grad_norm": 0.10247822105884552,
+      "learning_rate": 9.016203703703704e-06,
+      "loss": 1.2227,
+      "step": 86
+    },
+    {
+      "epoch": 0.10069444444444445,
+      "grad_norm": 0.09980504959821701,
+      "learning_rate": 9.00462962962963e-06,
+      "loss": 1.2352,
+      "step": 87
+    },
+    {
+      "epoch": 0.10185185185185185,
+      "grad_norm": 0.09725635498762131,
+      "learning_rate": 8.993055555555556e-06,
+      "loss": 1.2665,
+      "step": 88
+    },
+    {
+      "epoch": 0.10300925925925926,
+      "grad_norm": 0.10134412348270416,
+      "learning_rate": 8.981481481481483e-06,
+      "loss": 1.2586,
+      "step": 89
+    },
+    {
+      "epoch": 0.10416666666666667,
+      "grad_norm": 0.10719292610883713,
+      "learning_rate": 8.969907407407407e-06,
+      "loss": 1.3019,
+      "step": 90
+    },
+    {
+      "epoch": 0.10532407407407407,
+      "grad_norm": 0.10486488789319992,
+      "learning_rate": 8.958333333333334e-06,
+      "loss": 1.2936,
+      "step": 91
+    },
+    {
+      "epoch": 0.10648148148148148,
+      "grad_norm": 0.10944393277168274,
+      "learning_rate": 8.94675925925926e-06,
+      "loss": 1.3363,
+      "step": 92
+    },
+    {
+      "epoch": 0.1076388888888889,
+      "grad_norm": 0.11079176515340805,
+      "learning_rate": 8.935185185185186e-06,
+      "loss": 1.2367,
+      "step": 93
+    },
+    {
+      "epoch": 0.1087962962962963,
+      "grad_norm": 0.10356653481721878,
+      "learning_rate": 8.923611111111113e-06,
+      "loss": 1.2405,
+      "step": 94
+    },
+    {
+      "epoch": 0.1099537037037037,
+      "grad_norm": 0.10176271945238113,
+      "learning_rate": 8.912037037037037e-06,
+      "loss": 1.2165,
+      "step": 95
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 0.10132142156362534,
+      "learning_rate": 8.900462962962964e-06,
+      "loss": 1.2244,
+      "step": 96
+    },
+    {
+      "epoch": 0.11226851851851852,
+      "grad_norm": 0.09974969923496246,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.2174,
+      "step": 97
+    },
+    {
+      "epoch": 0.11342592592592593,
+      "grad_norm": 0.10894669592380524,
+      "learning_rate": 8.877314814814816e-06,
+      "loss": 1.2788,
+      "step": 98
+    },
+    {
+      "epoch": 0.11458333333333333,
+      "grad_norm": 0.10840534418821335,
+      "learning_rate": 8.865740740740741e-06,
+      "loss": 1.3159,
+      "step": 99
+    },
+    {
+      "epoch": 0.11574074074074074,
+      "grad_norm": 0.10695167630910873,
+      "learning_rate": 8.854166666666667e-06,
+      "loss": 1.1849,
+      "step": 100
+    },
+    {
+      "epoch": 0.11689814814814815,
+      "grad_norm": 0.10731031000614166,
+      "learning_rate": 8.842592592592594e-06,
+      "loss": 1.2752,
+      "step": 101
+    },
+    {
+      "epoch": 0.11805555555555555,
+      "grad_norm": 0.10931055247783661,
+      "learning_rate": 8.83101851851852e-06,
+      "loss": 1.3104,
+      "step": 102
+    },
+    {
+      "epoch": 0.11921296296296297,
+      "grad_norm": 0.09977404028177261,
+      "learning_rate": 8.819444444444445e-06,
+      "loss": 1.2294,
+      "step": 103
+    },
+    {
+      "epoch": 0.12037037037037036,
+      "grad_norm": 0.10199018567800522,
+      "learning_rate": 8.807870370370371e-06,
+      "loss": 1.26,
+      "step": 104
+    },
+    {
+      "epoch": 0.12152777777777778,
+      "grad_norm": 0.10436324030160904,
+      "learning_rate": 8.796296296296297e-06,
+      "loss": 1.1939,
+      "step": 105
+    },
+    {
+      "epoch": 0.12268518518518519,
+      "grad_norm": 0.10257507115602493,
+      "learning_rate": 8.784722222222224e-06,
+      "loss": 1.2018,
+      "step": 106
+    },
+    {
+      "epoch": 0.12384259259259259,
+      "grad_norm": 0.10942145437002182,
+      "learning_rate": 8.773148148148148e-06,
+      "loss": 1.2486,
+      "step": 107
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.1106390580534935,
+      "learning_rate": 8.761574074074075e-06,
+      "loss": 1.2423,
+      "step": 108
+    },
+    {
+      "epoch": 0.1261574074074074,
+      "grad_norm": 0.10852088034152985,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.2164,
+      "step": 109
+    },
+    {
+      "epoch": 0.12731481481481483,
+      "grad_norm": 0.10899867117404938,
+      "learning_rate": 8.738425925925926e-06,
+      "loss": 1.1733,
+      "step": 110
+    },
+    {
+      "epoch": 0.1284722222222222,
+      "grad_norm": 0.1090475544333458,
+      "learning_rate": 8.726851851851854e-06,
+      "loss": 1.211,
+      "step": 111
+    },
+    {
+      "epoch": 0.12962962962962962,
+      "grad_norm": 0.10560119152069092,
+      "learning_rate": 8.715277777777778e-06,
+      "loss": 1.1627,
+      "step": 112
+    },
+    {
+      "epoch": 0.13078703703703703,
+      "grad_norm": 0.140657439827919,
+      "learning_rate": 8.703703703703705e-06,
+      "loss": 1.2528,
+      "step": 113
+    },
+    {
+      "epoch": 0.13194444444444445,
+      "grad_norm": 0.10568102449178696,
+      "learning_rate": 8.69212962962963e-06,
+      "loss": 1.1707,
+      "step": 114
+    },
+    {
+      "epoch": 0.13310185185185186,
+      "grad_norm": 0.10824406147003174,
+      "learning_rate": 8.680555555555557e-06,
+      "loss": 1.236,
+      "step": 115
+    },
+    {
+      "epoch": 0.13425925925925927,
+      "grad_norm": 0.13147617876529694,
+      "learning_rate": 8.668981481481482e-06,
+      "loss": 1.2215,
+      "step": 116
+    },
+    {
+      "epoch": 0.13541666666666666,
+      "grad_norm": 0.10489261895418167,
+      "learning_rate": 8.657407407407408e-06,
+      "loss": 1.1983,
+      "step": 117
+    },
+    {
+      "epoch": 0.13657407407407407,
+      "grad_norm": 0.10651307553052902,
+      "learning_rate": 8.645833333333335e-06,
+      "loss": 1.134,
+      "step": 118
+    },
+    {
+      "epoch": 0.13773148148148148,
+      "grad_norm": 0.10825668275356293,
+      "learning_rate": 8.63425925925926e-06,
+      "loss": 1.2174,
+      "step": 119
+    },
+    {
+      "epoch": 0.1388888888888889,
+      "grad_norm": 0.11069732904434204,
+      "learning_rate": 8.622685185185186e-06,
+      "loss": 1.1513,
+      "step": 120
+    },
+    {
+      "epoch": 0.1400462962962963,
+      "grad_norm": 0.10880957543849945,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.1804,
+      "step": 121
+    },
+    {
+      "epoch": 0.1412037037037037,
+      "grad_norm": 0.1930810511112213,
+      "learning_rate": 8.599537037037038e-06,
+      "loss": 1.258,
+      "step": 122
+    },
+    {
+      "epoch": 0.1423611111111111,
+      "grad_norm": 0.11332959681749344,
+      "learning_rate": 8.587962962962963e-06,
+      "loss": 1.1904,
+      "step": 123
+    },
+    {
+      "epoch": 0.14351851851851852,
+      "grad_norm": 0.10661832243204117,
+      "learning_rate": 8.57638888888889e-06,
+      "loss": 1.1768,
+      "step": 124
+    },
+    {
+      "epoch": 0.14467592592592593,
+      "grad_norm": 0.11842379719018936,
+      "learning_rate": 8.564814814814816e-06,
+      "loss": 1.262,
+      "step": 125
+    },
+    {
+      "epoch": 0.14583333333333334,
+      "grad_norm": 0.11844724416732788,
+      "learning_rate": 8.553240740740742e-06,
+      "loss": 1.2975,
+      "step": 126
+    },
+    {
+      "epoch": 0.14699074074074073,
+      "grad_norm": 0.11894673854112625,
+      "learning_rate": 8.541666666666666e-06,
+      "loss": 1.2188,
+      "step": 127
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 0.12049190700054169,
+      "learning_rate": 8.530092592592595e-06,
+      "loss": 1.2252,
+      "step": 128
+    },
+    {
+      "epoch": 0.14930555555555555,
+      "grad_norm": 0.1142427921295166,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 1.242,
+      "step": 129
+    },
+    {
+      "epoch": 0.15046296296296297,
+      "grad_norm": 0.1175389289855957,
+      "learning_rate": 8.506944444444445e-06,
+      "loss": 1.17,
+      "step": 130
+    },
+    {
+      "epoch": 0.15162037037037038,
+      "grad_norm": 0.11764886230230331,
+      "learning_rate": 8.495370370370372e-06,
+      "loss": 1.2249,
+      "step": 131
+    },
+    {
+      "epoch": 0.1527777777777778,
+      "grad_norm": 0.11646956950426102,
+      "learning_rate": 8.483796296296296e-06,
+      "loss": 1.2169,
+      "step": 132
+    },
+    {
+      "epoch": 0.15393518518518517,
+      "grad_norm": 0.11439011991024017,
+      "learning_rate": 8.472222222222223e-06,
+      "loss": 1.218,
+      "step": 133
+    },
+    {
+      "epoch": 0.1550925925925926,
+      "grad_norm": 0.12252136319875717,
+      "learning_rate": 8.460648148148149e-06,
+      "loss": 1.2023,
+      "step": 134
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 0.12436626106500626,
+      "learning_rate": 8.449074074074075e-06,
+      "loss": 1.2033,
+      "step": 135
+    },
+    {
+      "epoch": 0.1574074074074074,
+      "grad_norm": 0.11901386827230453,
+      "learning_rate": 8.4375e-06,
+      "loss": 1.272,
+      "step": 136
+    },
+    {
+      "epoch": 0.15856481481481483,
+      "grad_norm": 0.11729884147644043,
+      "learning_rate": 8.425925925925926e-06,
+      "loss": 1.2505,
+      "step": 137
+    },
+    {
+      "epoch": 0.1597222222222222,
+      "grad_norm": 0.13269205391407013,
+      "learning_rate": 8.414351851851853e-06,
+      "loss": 1.1483,
+      "step": 138
+    },
+    {
+      "epoch": 0.16087962962962962,
+      "grad_norm": 0.12206093221902847,
+      "learning_rate": 8.402777777777779e-06,
+      "loss": 1.1865,
+      "step": 139
+    },
+    {
+      "epoch": 0.16203703703703703,
+      "grad_norm": 0.11812015622854233,
+      "learning_rate": 8.391203703703704e-06,
+      "loss": 1.199,
+      "step": 140
+    },
+    {
+      "epoch": 0.16319444444444445,
+      "grad_norm": 0.11516984552145004,
+      "learning_rate": 8.37962962962963e-06,
+      "loss": 1.1904,
+      "step": 141
+    },
+    {
+      "epoch": 0.16435185185185186,
+      "grad_norm": 0.1210319846868515,
+      "learning_rate": 8.368055555555556e-06,
+      "loss": 1.178,
+      "step": 142
+    },
+    {
+      "epoch": 0.16550925925925927,
+      "grad_norm": 0.11981932818889618,
+      "learning_rate": 8.356481481481483e-06,
+      "loss": 1.1645,
+      "step": 143
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.1515599936246872,
+      "learning_rate": 8.344907407407407e-06,
+      "loss": 1.1536,
+      "step": 144
+    },
+    {
+      "epoch": 0.16782407407407407,
+      "grad_norm": 0.12906567752361298,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.2263,
+      "step": 145
+    },
+    {
+      "epoch": 0.16898148148148148,
+      "grad_norm": 0.13303443789482117,
+      "learning_rate": 8.32175925925926e-06,
+      "loss": 1.1575,
+      "step": 146
+    },
+    {
+      "epoch": 0.1701388888888889,
+      "grad_norm": 0.13370473682880402,
+      "learning_rate": 8.310185185185186e-06,
+      "loss": 1.1965,
+      "step": 147
+    },
+    {
+      "epoch": 0.1712962962962963,
+      "grad_norm": 0.13174037635326385,
+      "learning_rate": 8.298611111111113e-06,
+      "loss": 1.2192,
+      "step": 148
+    },
+    {
+      "epoch": 0.1724537037037037,
+      "grad_norm": 0.12153156101703644,
+      "learning_rate": 8.287037037037037e-06,
+      "loss": 1.2325,
+      "step": 149
+    },
+    {
+      "epoch": 0.1736111111111111,
+      "grad_norm": 0.12656794488430023,
+      "learning_rate": 8.275462962962964e-06,
+      "loss": 1.1419,
+      "step": 150
+    },
+    {
+      "epoch": 0.17476851851851852,
+      "grad_norm": 0.12459167093038559,
+      "learning_rate": 8.263888888888888e-06,
+      "loss": 1.1894,
+      "step": 151
+    },
+    {
+      "epoch": 0.17592592592592593,
+      "grad_norm": 0.1209847703576088,
+      "learning_rate": 8.252314814814816e-06,
+      "loss": 1.1009,
+      "step": 152
+    },
+    {
+      "epoch": 0.17708333333333334,
+      "grad_norm": 0.122232586145401,
+      "learning_rate": 8.240740740740741e-06,
+      "loss": 1.1768,
+      "step": 153
+    },
+    {
+      "epoch": 0.17824074074074073,
+      "grad_norm": 0.12607316672801971,
+      "learning_rate": 8.229166666666667e-06,
+      "loss": 1.1269,
+      "step": 154
+    },
+    {
+      "epoch": 0.17939814814814814,
+      "grad_norm": 0.12453708052635193,
+      "learning_rate": 8.217592592592594e-06,
+      "loss": 1.1205,
+      "step": 155
+    },
+    {
+      "epoch": 0.18055555555555555,
+      "grad_norm": 0.12600407004356384,
+      "learning_rate": 8.20601851851852e-06,
+      "loss": 1.1924,
+      "step": 156
+    },
+    {
+      "epoch": 0.18171296296296297,
+      "grad_norm": 0.13688451051712036,
+      "learning_rate": 8.194444444444445e-06,
+      "loss": 1.2399,
+      "step": 157
+    },
+    {
+      "epoch": 0.18287037037037038,
+      "grad_norm": 0.12164533883333206,
+      "learning_rate": 8.182870370370371e-06,
+      "loss": 1.119,
+      "step": 158
+    },
+    {
+      "epoch": 0.1840277777777778,
+      "grad_norm": 0.12914681434631348,
+      "learning_rate": 8.171296296296297e-06,
+      "loss": 1.2067,
+      "step": 159
+    },
+    {
+      "epoch": 0.18518518518518517,
+      "grad_norm": 0.13052506744861603,
+      "learning_rate": 8.159722222222222e-06,
+      "loss": 1.1148,
+      "step": 160
+    },
+    {
+      "epoch": 0.1863425925925926,
+      "grad_norm": 0.14109309017658234,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 1.1888,
+      "step": 161
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.1215931847691536,
+      "learning_rate": 8.136574074074075e-06,
+      "loss": 1.1696,
+      "step": 162
+    },
+    {
+      "epoch": 0.1886574074074074,
+      "grad_norm": 0.12615671753883362,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 1.0878,
+      "step": 163
+    },
+    {
+      "epoch": 0.18981481481481483,
+      "grad_norm": 0.12854883074760437,
+      "learning_rate": 8.113425925925926e-06,
+      "loss": 1.1693,
+      "step": 164
+    },
+    {
+      "epoch": 0.1909722222222222,
+      "grad_norm": 0.13595585525035858,
+      "learning_rate": 8.101851851851854e-06,
+      "loss": 1.2612,
+      "step": 165
+    },
+    {
+      "epoch": 0.19212962962962962,
+      "grad_norm": 0.125200554728508,
+      "learning_rate": 8.090277777777778e-06,
+      "loss": 1.1334,
+      "step": 166
+    },
+    {
+      "epoch": 0.19328703703703703,
+      "grad_norm": 0.12699203193187714,
+      "learning_rate": 8.078703703703705e-06,
+      "loss": 1.1895,
+      "step": 167
+    },
+    {
+      "epoch": 0.19444444444444445,
+      "grad_norm": 0.14522430300712585,
+      "learning_rate": 8.06712962962963e-06,
+      "loss": 1.2751,
+      "step": 168
+    },
+    {
+      "epoch": 0.19560185185185186,
+      "grad_norm": 0.12672430276870728,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 1.1263,
+      "step": 169
+    },
+    {
+      "epoch": 0.19675925925925927,
+      "grad_norm": 0.1354917585849762,
+      "learning_rate": 8.043981481481482e-06,
+      "loss": 1.1853,
+      "step": 170
+    },
+    {
+      "epoch": 0.19791666666666666,
+      "grad_norm": 0.12905584275722504,
+      "learning_rate": 8.032407407407408e-06,
+      "loss": 1.1671,
+      "step": 171
+    },
+    {
+      "epoch": 0.19907407407407407,
+      "grad_norm": 0.14876537024974823,
+      "learning_rate": 8.020833333333335e-06,
+      "loss": 1.192,
+      "step": 172
+    },
+    {
+      "epoch": 0.20023148148148148,
+      "grad_norm": 0.12572503089904785,
+      "learning_rate": 8.00925925925926e-06,
+      "loss": 1.1425,
+      "step": 173
+    },
+    {
+      "epoch": 0.2013888888888889,
+      "grad_norm": 0.1319337785243988,
+      "learning_rate": 7.997685185185186e-06,
+      "loss": 1.1048,
+      "step": 174
+    },
+    {
+      "epoch": 0.2025462962962963,
+      "grad_norm": 0.1327625960111618,
+      "learning_rate": 7.986111111111112e-06,
+      "loss": 1.1644,
+      "step": 175
+    },
+    {
+      "epoch": 0.2037037037037037,
+      "grad_norm": 0.15492317080497742,
+      "learning_rate": 7.974537037037038e-06,
+      "loss": 1.1716,
+      "step": 176
+    },
+    {
+      "epoch": 0.2048611111111111,
+      "grad_norm": 0.13174159824848175,
+      "learning_rate": 7.962962962962963e-06,
+      "loss": 1.17,
+      "step": 177
+    },
+    {
+      "epoch": 0.20601851851851852,
+      "grad_norm": 0.13209496438503265,
+      "learning_rate": 7.95138888888889e-06,
+      "loss": 1.1483,
+      "step": 178
+    },
+    {
+      "epoch": 0.20717592592592593,
+      "grad_norm": 0.13201004266738892,
+      "learning_rate": 7.939814814814816e-06,
+      "loss": 1.1383,
+      "step": 179
+    },
+    {
+      "epoch": 0.20833333333333334,
+      "grad_norm": 0.13409128785133362,
+      "learning_rate": 7.928240740740742e-06,
+      "loss": 1.2027,
+      "step": 180
+    },
+    {
+      "epoch": 0.20949074074074073,
+      "grad_norm": 0.12518879771232605,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 1.124,
+      "step": 181
+    },
+    {
+      "epoch": 0.21064814814814814,
+      "grad_norm": 0.13146327435970306,
+      "learning_rate": 7.905092592592593e-06,
+      "loss": 1.1499,
+      "step": 182
+    },
+    {
+      "epoch": 0.21180555555555555,
+      "grad_norm": 0.13177363574504852,
+      "learning_rate": 7.89351851851852e-06,
+      "loss": 1.1327,
+      "step": 183
+    },
+    {
+      "epoch": 0.21296296296296297,
+      "grad_norm": 0.1373285949230194,
+      "learning_rate": 7.881944444444446e-06,
+      "loss": 1.1803,
+      "step": 184
+    },
+    {
+      "epoch": 0.21412037037037038,
+      "grad_norm": 0.13236674666404724,
+      "learning_rate": 7.870370370370372e-06,
+      "loss": 1.1941,
+      "step": 185
+    },
+    {
+      "epoch": 0.2152777777777778,
+      "grad_norm": 0.14942052960395813,
+      "learning_rate": 7.858796296296297e-06,
+      "loss": 1.0679,
+      "step": 186
+    },
+    {
+      "epoch": 0.21643518518518517,
+      "grad_norm": 0.1368960291147232,
+      "learning_rate": 7.847222222222223e-06,
+      "loss": 1.2047,
+      "step": 187
+    },
+    {
+      "epoch": 0.2175925925925926,
+      "grad_norm": 0.13313372433185577,
+      "learning_rate": 7.835648148148147e-06,
+      "loss": 1.1604,
+      "step": 188
+    },
+    {
+      "epoch": 0.21875,
+      "grad_norm": 0.13543230295181274,
+      "learning_rate": 7.824074074074076e-06,
+      "loss": 1.1457,
+      "step": 189
+    },
+    {
+      "epoch": 0.2199074074074074,
+      "grad_norm": 0.13306613266468048,
+      "learning_rate": 7.8125e-06,
+      "loss": 1.1072,
+      "step": 190
+    },
+    {
+      "epoch": 0.22106481481481483,
+      "grad_norm": 0.1399572342634201,
+      "learning_rate": 7.800925925925926e-06,
+      "loss": 1.1941,
+      "step": 191
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.13952195644378662,
+      "learning_rate": 7.789351851851853e-06,
+      "loss": 1.1165,
+      "step": 192
+    },
+    {
+      "epoch": 0.22337962962962962,
+      "grad_norm": 0.1346345692873001,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.1592,
+      "step": 193
+    },
+    {
+      "epoch": 0.22453703703703703,
+      "grad_norm": 0.14998078346252441,
+      "learning_rate": 7.766203703703704e-06,
+      "loss": 1.0881,
+      "step": 194
+    },
+    {
+      "epoch": 0.22569444444444445,
+      "grad_norm": 0.1336279958486557,
+      "learning_rate": 7.75462962962963e-06,
+      "loss": 1.1719,
+      "step": 195
+    },
+    {
+      "epoch": 0.22685185185185186,
+      "grad_norm": 0.1357223093509674,
+      "learning_rate": 7.743055555555556e-06,
+      "loss": 1.1395,
+      "step": 196
+    },
+    {
+      "epoch": 0.22800925925925927,
+      "grad_norm": 0.13101308047771454,
+      "learning_rate": 7.731481481481483e-06,
+      "loss": 1.1486,
+      "step": 197
+    },
+    {
+      "epoch": 0.22916666666666666,
+      "grad_norm": 0.13381895422935486,
+      "learning_rate": 7.719907407407407e-06,
+      "loss": 1.1791,
+      "step": 198
+    },
+    {
+      "epoch": 0.23032407407407407,
+      "grad_norm": 0.14860618114471436,
+      "learning_rate": 7.708333333333334e-06,
+      "loss": 1.2094,
+      "step": 199
+    },
+    {
+      "epoch": 0.23148148148148148,
+      "grad_norm": 0.14094901084899902,
+      "learning_rate": 7.69675925925926e-06,
+      "loss": 1.1111,
+      "step": 200
+    },
+    {
+      "epoch": 0.2326388888888889,
+      "grad_norm": 0.14977441728115082,
+      "learning_rate": 7.685185185185185e-06,
+      "loss": 1.1361,
+      "step": 201
+    },
+    {
+      "epoch": 0.2337962962962963,
+      "grad_norm": 0.14465676248073578,
+      "learning_rate": 7.673611111111113e-06,
+      "loss": 1.1081,
+      "step": 202
+    },
+    {
+      "epoch": 0.2349537037037037,
+      "grad_norm": 0.13928347826004028,
+      "learning_rate": 7.662037037037037e-06,
+      "loss": 1.1567,
+      "step": 203
+    },
+    {
+      "epoch": 0.2361111111111111,
+      "grad_norm": 0.13289427757263184,
+      "learning_rate": 7.650462962962964e-06,
+      "loss": 1.0925,
+      "step": 204
+    },
+    {
+      "epoch": 0.23726851851851852,
+      "grad_norm": 0.13851501047611237,
+      "learning_rate": 7.638888888888888e-06,
+      "loss": 1.1707,
+      "step": 205
+    },
+    {
+      "epoch": 0.23842592592592593,
+      "grad_norm": 0.1386367678642273,
+      "learning_rate": 7.627314814814816e-06,
+      "loss": 1.1997,
+      "step": 206
+    },
+    {
+      "epoch": 0.23958333333333334,
+      "grad_norm": 0.140579491853714,
+      "learning_rate": 7.615740740740741e-06,
+      "loss": 1.0898,
+      "step": 207
+    },
+    {
+      "epoch": 0.24074074074074073,
+      "grad_norm": 0.14284437894821167,
+      "learning_rate": 7.6041666666666666e-06,
+      "loss": 1.1325,
+      "step": 208
+    },
+    {
+      "epoch": 0.24189814814814814,
+      "grad_norm": 0.1468944400548935,
+      "learning_rate": 7.592592592592594e-06,
+      "loss": 1.1007,
+      "step": 209
+    },
+    {
+      "epoch": 0.24305555555555555,
+      "grad_norm": 0.14520986378192902,
+      "learning_rate": 7.581018518518519e-06,
+      "loss": 1.0402,
+      "step": 210
+    },
+    {
+      "epoch": 0.24421296296296297,
+      "grad_norm": 0.1370912790298462,
+      "learning_rate": 7.569444444444445e-06,
+      "loss": 1.1189,
+      "step": 211
+    },
+    {
+      "epoch": 0.24537037037037038,
+      "grad_norm": 0.13841593265533447,
+      "learning_rate": 7.557870370370372e-06,
+      "loss": 1.1722,
+      "step": 212
+    },
+    {
+      "epoch": 0.2465277777777778,
+      "grad_norm": 0.13426385819911957,
+      "learning_rate": 7.546296296296297e-06,
+      "loss": 1.077,
+      "step": 213
+    },
+    {
+      "epoch": 0.24768518518518517,
+      "grad_norm": 0.14030449092388153,
+      "learning_rate": 7.534722222222223e-06,
+      "loss": 1.1423,
+      "step": 214
+    },
+    {
+      "epoch": 0.2488425925925926,
+      "grad_norm": 0.13870425522327423,
+      "learning_rate": 7.523148148148148e-06,
+      "loss": 1.1222,
+      "step": 215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1389496624469757,
+      "learning_rate": 7.511574074074075e-06,
+      "loss": 1.1281,
+      "step": 216
+    },
+    {
+      "epoch": 0.2511574074074074,
+      "grad_norm": 0.13513287901878357,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.0742,
+      "step": 217
+    },
+    {
+      "epoch": 0.2523148148148148,
+      "grad_norm": 0.12963634729385376,
+      "learning_rate": 7.4884259259259265e-06,
+      "loss": 1.1002,
+      "step": 218
+    },
+    {
+      "epoch": 0.2534722222222222,
+      "grad_norm": 0.14197136461734772,
+      "learning_rate": 7.476851851851853e-06,
+      "loss": 1.2041,
+      "step": 219
+    },
+    {
+      "epoch": 0.25462962962962965,
+      "grad_norm": 0.14973148703575134,
+      "learning_rate": 7.465277777777778e-06,
+      "loss": 1.1425,
+      "step": 220
+    },
+    {
+      "epoch": 0.25578703703703703,
+      "grad_norm": 0.15027864277362823,
+      "learning_rate": 7.453703703703704e-06,
+      "loss": 1.1197,
+      "step": 221
+    },
+    {
+      "epoch": 0.2569444444444444,
+      "grad_norm": 0.14004850387573242,
+      "learning_rate": 7.442129629629629e-06,
+      "loss": 1.0865,
+      "step": 222
+    },
+    {
+      "epoch": 0.25810185185185186,
+      "grad_norm": 0.14548063278198242,
+      "learning_rate": 7.4305555555555565e-06,
+      "loss": 1.1288,
+      "step": 223
+    },
+    {
+      "epoch": 0.25925925925925924,
+      "grad_norm": 0.15253789722919464,
+      "learning_rate": 7.418981481481482e-06,
+      "loss": 1.1755,
+      "step": 224
+    },
+    {
+      "epoch": 0.2604166666666667,
+      "grad_norm": 0.17614801228046417,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 1.1337,
+      "step": 225
+    },
+    {
+      "epoch": 0.26157407407407407,
+      "grad_norm": 0.14582227170467377,
+      "learning_rate": 7.395833333333335e-06,
+      "loss": 1.127,
+      "step": 226
+    },
+    {
+      "epoch": 0.26273148148148145,
+      "grad_norm": 0.15011462569236755,
+      "learning_rate": 7.38425925925926e-06,
+      "loss": 1.1503,
+      "step": 227
+    },
+    {
+      "epoch": 0.2638888888888889,
+      "grad_norm": 0.14777828752994537,
+      "learning_rate": 7.372685185185186e-06,
+      "loss": 1.1979,
+      "step": 228
+    },
+    {
+      "epoch": 0.2650462962962963,
+      "grad_norm": 0.14377012848854065,
+      "learning_rate": 7.361111111111112e-06,
+      "loss": 1.159,
+      "step": 229
+    },
+    {
+      "epoch": 0.2662037037037037,
+      "grad_norm": 0.14066831767559052,
+      "learning_rate": 7.3495370370370375e-06,
+      "loss": 1.083,
+      "step": 230
+    },
+    {
+      "epoch": 0.2673611111111111,
+      "grad_norm": 0.14046640694141388,
+      "learning_rate": 7.337962962962964e-06,
+      "loss": 1.0942,
+      "step": 231
+    },
+    {
+      "epoch": 0.26851851851851855,
+      "grad_norm": 0.14818687736988068,
+      "learning_rate": 7.326388888888889e-06,
+      "loss": 1.1081,
+      "step": 232
+    },
+    {
+      "epoch": 0.26967592592592593,
+      "grad_norm": 0.14957569539546967,
+      "learning_rate": 7.314814814814816e-06,
+      "loss": 1.1607,
+      "step": 233
+    },
+    {
+      "epoch": 0.2708333333333333,
+      "grad_norm": 0.14794419705867767,
+      "learning_rate": 7.303240740740741e-06,
+      "loss": 1.1061,
+      "step": 234
+    },
+    {
+      "epoch": 0.27199074074074076,
+      "grad_norm": 0.14263461530208588,
+      "learning_rate": 7.291666666666667e-06,
+      "loss": 1.0914,
+      "step": 235
+    },
+    {
+      "epoch": 0.27314814814814814,
+      "grad_norm": 0.1496988832950592,
+      "learning_rate": 7.280092592592594e-06,
+      "loss": 1.1051,
+      "step": 236
+    },
+    {
+      "epoch": 0.2743055555555556,
+      "grad_norm": 0.17018254101276398,
+      "learning_rate": 7.268518518518519e-06,
+      "loss": 1.1102,
+      "step": 237
+    },
+    {
+      "epoch": 0.27546296296296297,
+      "grad_norm": 0.13826943933963776,
+      "learning_rate": 7.256944444444445e-06,
+      "loss": 1.1153,
+      "step": 238
+    },
+    {
+      "epoch": 0.27662037037037035,
+      "grad_norm": 0.15133726596832275,
+      "learning_rate": 7.245370370370371e-06,
+      "loss": 1.1704,
+      "step": 239
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 0.1380927413702011,
+      "learning_rate": 7.233796296296297e-06,
+      "loss": 1.0637,
+      "step": 240
+    },
+    {
+      "epoch": 0.2789351851851852,
+      "grad_norm": 0.17274874448776245,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 1.0886,
+      "step": 241
+    },
+    {
+      "epoch": 0.2800925925925926,
+      "grad_norm": 0.1451474279165268,
+      "learning_rate": 7.210648148148148e-06,
+      "loss": 1.1426,
+      "step": 242
+    },
+    {
+      "epoch": 0.28125,
+      "grad_norm": 0.15506966412067413,
+      "learning_rate": 7.199074074074075e-06,
+      "loss": 1.1539,
+      "step": 243
+    },
+    {
+      "epoch": 0.2824074074074074,
+      "grad_norm": 0.14218544960021973,
+      "learning_rate": 7.1875e-06,
+      "loss": 1.1328,
+      "step": 244
+    },
+    {
+      "epoch": 0.2835648148148148,
+      "grad_norm": 0.14230510592460632,
+      "learning_rate": 7.1759259259259266e-06,
+      "loss": 1.1231,
+      "step": 245
+    },
+    {
+      "epoch": 0.2847222222222222,
+      "grad_norm": 0.15011471509933472,
+      "learning_rate": 7.164351851851853e-06,
+      "loss": 1.1324,
+      "step": 246
+    },
+    {
+      "epoch": 0.28587962962962965,
+      "grad_norm": 0.17919589579105377,
+      "learning_rate": 7.152777777777778e-06,
+      "loss": 1.174,
+      "step": 247
+    },
+    {
+      "epoch": 0.28703703703703703,
+      "grad_norm": 0.22477856278419495,
+      "learning_rate": 7.141203703703704e-06,
+      "loss": 1.1293,
+      "step": 248
+    },
+    {
+      "epoch": 0.2881944444444444,
+      "grad_norm": 0.15485665202140808,
+      "learning_rate": 7.129629629629629e-06,
+      "loss": 1.0675,
+      "step": 249
+    },
+    {
+      "epoch": 0.28935185185185186,
+      "grad_norm": 0.16213078796863556,
+      "learning_rate": 7.1180555555555565e-06,
+      "loss": 1.1293,
+      "step": 250
+    },
+    {
+      "epoch": 0.29050925925925924,
+      "grad_norm": 0.15522325038909912,
+      "learning_rate": 7.106481481481482e-06,
+      "loss": 1.1387,
+      "step": 251
+    },
+    {
+      "epoch": 0.2916666666666667,
+      "grad_norm": 0.16681715846061707,
+      "learning_rate": 7.0949074074074075e-06,
+      "loss": 1.0447,
+      "step": 252
+    },
+    {
+      "epoch": 0.29282407407407407,
+      "grad_norm": 0.15302035212516785,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 1.1575,
+      "step": 253
+    },
+    {
+      "epoch": 0.29398148148148145,
+      "grad_norm": 0.1592303216457367,
+      "learning_rate": 7.07175925925926e-06,
+      "loss": 1.1439,
+      "step": 254
+    },
+    {
+      "epoch": 0.2951388888888889,
+      "grad_norm": 0.14766466617584229,
+      "learning_rate": 7.060185185185186e-06,
+      "loss": 1.1439,
+      "step": 255
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 0.14591576159000397,
+      "learning_rate": 7.048611111111112e-06,
+      "loss": 1.0393,
+      "step": 256
+    },
+    {
+      "epoch": 0.2974537037037037,
+      "grad_norm": 0.1551518440246582,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 1.0774,
+      "step": 257
+    },
+    {
+      "epoch": 0.2986111111111111,
+      "grad_norm": 0.15124709904193878,
+      "learning_rate": 7.025462962962963e-06,
+      "loss": 1.1139,
+      "step": 258
+    },
+    {
+      "epoch": 0.29976851851851855,
+      "grad_norm": 0.15364578366279602,
+      "learning_rate": 7.013888888888889e-06,
+      "loss": 1.091,
+      "step": 259
+    },
+    {
+      "epoch": 0.30092592592592593,
+      "grad_norm": 0.15513962507247925,
+      "learning_rate": 7.002314814814816e-06,
+      "loss": 1.1245,
+      "step": 260
+    },
+    {
+      "epoch": 0.3020833333333333,
+      "grad_norm": 0.15260472893714905,
+      "learning_rate": 6.990740740740741e-06,
+      "loss": 1.0679,
+      "step": 261
+    },
+    {
+      "epoch": 0.30324074074074076,
+      "grad_norm": 0.15631963312625885,
+      "learning_rate": 6.979166666666667e-06,
+      "loss": 1.0904,
+      "step": 262
+    },
+    {
+      "epoch": 0.30439814814814814,
+      "grad_norm": 0.16324004530906677,
+      "learning_rate": 6.967592592592594e-06,
+      "loss": 1.122,
+      "step": 263
+    },
+    {
+      "epoch": 0.3055555555555556,
+      "grad_norm": 0.1569405198097229,
+      "learning_rate": 6.956018518518519e-06,
+      "loss": 1.1128,
+      "step": 264
+    },
+    {
+      "epoch": 0.30671296296296297,
+      "grad_norm": 0.14927881956100464,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 1.1037,
+      "step": 265
+    },
+    {
+      "epoch": 0.30787037037037035,
+      "grad_norm": 0.16223756968975067,
+      "learning_rate": 6.932870370370371e-06,
+      "loss": 1.1296,
+      "step": 266
+    },
+    {
+      "epoch": 0.3090277777777778,
+      "grad_norm": 0.15246330201625824,
+      "learning_rate": 6.9212962962962974e-06,
+      "loss": 1.121,
+      "step": 267
+    },
+    {
+      "epoch": 0.3101851851851852,
+      "grad_norm": 0.15738119184970856,
+      "learning_rate": 6.909722222222223e-06,
+      "loss": 1.021,
+      "step": 268
+    },
+    {
+      "epoch": 0.3113425925925926,
+      "grad_norm": 0.14696645736694336,
+      "learning_rate": 6.898148148148148e-06,
+      "loss": 1.1129,
+      "step": 269
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.17972029745578766,
+      "learning_rate": 6.886574074074075e-06,
+      "loss": 1.1459,
+      "step": 270
+    },
+    {
+      "epoch": 0.3136574074074074,
+      "grad_norm": 0.15355713665485382,
+      "learning_rate": 6.875e-06,
+      "loss": 1.0327,
+      "step": 271
+    },
+    {
+      "epoch": 0.3148148148148148,
+      "grad_norm": 0.14485421776771545,
+      "learning_rate": 6.863425925925927e-06,
+      "loss": 1.0728,
+      "step": 272
+    },
+    {
+      "epoch": 0.3159722222222222,
+      "grad_norm": 0.14966461062431335,
+      "learning_rate": 6.851851851851853e-06,
+      "loss": 1.11,
+      "step": 273
+    },
+    {
+      "epoch": 0.31712962962962965,
+      "grad_norm": 0.15380822122097015,
+      "learning_rate": 6.840277777777778e-06,
+      "loss": 1.0322,
+      "step": 274
+    },
+    {
+      "epoch": 0.31828703703703703,
+      "grad_norm": 0.15833279490470886,
+      "learning_rate": 6.828703703703704e-06,
+      "loss": 1.108,
+      "step": 275
+    },
+    {
+      "epoch": 0.3194444444444444,
+      "grad_norm": 0.15459351241588593,
+      "learning_rate": 6.817129629629629e-06,
+      "loss": 1.0904,
+      "step": 276
+    },
+    {
+      "epoch": 0.32060185185185186,
+      "grad_norm": 0.1778038591146469,
+      "learning_rate": 6.8055555555555566e-06,
+      "loss": 1.1272,
+      "step": 277
+    },
+    {
+      "epoch": 0.32175925925925924,
+      "grad_norm": 0.16406135261058807,
+      "learning_rate": 6.793981481481482e-06,
+      "loss": 1.05,
+      "step": 278
+    },
+    {
+      "epoch": 0.3229166666666667,
+      "grad_norm": 0.18400990962982178,
+      "learning_rate": 6.7824074074074075e-06,
+      "loss": 1.0778,
+      "step": 279
+    },
+    {
+      "epoch": 0.32407407407407407,
+      "grad_norm": 0.157948300242424,
+      "learning_rate": 6.770833333333334e-06,
+      "loss": 1.0552,
+      "step": 280
+    },
+    {
+      "epoch": 0.32523148148148145,
+      "grad_norm": 0.14986403286457062,
+      "learning_rate": 6.75925925925926e-06,
+      "loss": 1.0509,
+      "step": 281
+    },
+    {
+      "epoch": 0.3263888888888889,
+      "grad_norm": 0.1901959925889969,
+      "learning_rate": 6.747685185185186e-06,
+      "loss": 1.0861,
+      "step": 282
+    },
+    {
+      "epoch": 0.3275462962962963,
+      "grad_norm": 0.15678079426288605,
+      "learning_rate": 6.736111111111112e-06,
+      "loss": 1.0942,
+      "step": 283
+    },
+    {
+      "epoch": 0.3287037037037037,
+      "grad_norm": 0.16509470343589783,
+      "learning_rate": 6.7245370370370375e-06,
+      "loss": 1.0301,
+      "step": 284
+    },
+    {
+      "epoch": 0.3298611111111111,
+      "grad_norm": 0.16847574710845947,
+      "learning_rate": 6.712962962962963e-06,
+      "loss": 1.1624,
+      "step": 285
+    },
+    {
+      "epoch": 0.33101851851851855,
+      "grad_norm": 0.17270340025424957,
+      "learning_rate": 6.701388888888889e-06,
+      "loss": 1.0494,
+      "step": 286
+    },
+    {
+      "epoch": 0.33217592592592593,
+      "grad_norm": 0.16031427681446075,
+      "learning_rate": 6.689814814814816e-06,
+      "loss": 1.1025,
+      "step": 287
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.15844310820102692,
+      "learning_rate": 6.678240740740741e-06,
+      "loss": 1.1172,
+      "step": 288
+    },
+    {
+      "epoch": 0.33449074074074076,
+      "grad_norm": 0.18013358116149902,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.143,
+      "step": 289
+    },
+    {
+      "epoch": 0.33564814814814814,
+      "grad_norm": 0.15711136162281036,
+      "learning_rate": 6.655092592592594e-06,
+      "loss": 1.0963,
+      "step": 290
+    },
+    {
+      "epoch": 0.3368055555555556,
+      "grad_norm": 0.1526433676481247,
+      "learning_rate": 6.643518518518519e-06,
+      "loss": 1.1131,
+      "step": 291
+    },
+    {
+      "epoch": 0.33796296296296297,
+      "grad_norm": 0.1604623794555664,
+      "learning_rate": 6.631944444444445e-06,
+      "loss": 1.091,
+      "step": 292
+    },
+    {
+      "epoch": 0.33912037037037035,
+      "grad_norm": 0.1599881798028946,
+      "learning_rate": 6.620370370370371e-06,
+      "loss": 1.1546,
+      "step": 293
+    },
+    {
+      "epoch": 0.3402777777777778,
+      "grad_norm": 0.1568020135164261,
+      "learning_rate": 6.608796296296297e-06,
+      "loss": 1.0707,
+      "step": 294
+    },
+    {
+      "epoch": 0.3414351851851852,
+      "grad_norm": 0.16622503101825714,
+      "learning_rate": 6.597222222222223e-06,
+      "loss": 1.0423,
+      "step": 295
+    },
+    {
+      "epoch": 0.3425925925925926,
+      "grad_norm": 0.1662890464067459,
+      "learning_rate": 6.5856481481481484e-06,
+      "loss": 1.0784,
+      "step": 296
+    },
+    {
+      "epoch": 0.34375,
+      "grad_norm": 0.24495013058185577,
+      "learning_rate": 6.574074074074075e-06,
+      "loss": 1.0781,
+      "step": 297
+    },
+    {
+      "epoch": 0.3449074074074074,
+      "grad_norm": 0.15532761812210083,
+      "learning_rate": 6.5625e-06,
+      "loss": 1.0132,
+      "step": 298
+    },
+    {
+      "epoch": 0.3460648148148148,
+      "grad_norm": 0.1543862670660019,
+      "learning_rate": 6.550925925925926e-06,
+      "loss": 1.1089,
+      "step": 299
+    },
+    {
+      "epoch": 0.3472222222222222,
+      "grad_norm": 0.1753946989774704,
+      "learning_rate": 6.539351851851853e-06,
+      "loss": 1.1471,
+      "step": 300
+    },
+    {
+      "epoch": 0.34837962962962965,
+      "grad_norm": 0.15991893410682678,
+      "learning_rate": 6.5277777777777784e-06,
+      "loss": 1.109,
+      "step": 301
+    },
+    {
+      "epoch": 0.34953703703703703,
+      "grad_norm": 0.15706345438957214,
+      "learning_rate": 6.516203703703704e-06,
+      "loss": 1.0623,
+      "step": 302
+    },
+    {
+      "epoch": 0.3506944444444444,
+      "grad_norm": 0.16499797999858856,
+      "learning_rate": 6.504629629629629e-06,
+      "loss": 1.0779,
+      "step": 303
+    },
+    {
+      "epoch": 0.35185185185185186,
+      "grad_norm": 0.1579602062702179,
+      "learning_rate": 6.493055555555557e-06,
+      "loss": 1.0858,
+      "step": 304
+    },
+    {
+      "epoch": 0.35300925925925924,
+      "grad_norm": 0.16340512037277222,
+      "learning_rate": 6.481481481481482e-06,
+      "loss": 1.1147,
+      "step": 305
+    },
+    {
+      "epoch": 0.3541666666666667,
+      "grad_norm": 0.16150407493114471,
+      "learning_rate": 6.4699074074074076e-06,
+      "loss": 1.0838,
+      "step": 306
+    },
+    {
+      "epoch": 0.35532407407407407,
+      "grad_norm": 0.1647614687681198,
+      "learning_rate": 6.458333333333334e-06,
+      "loss": 1.1434,
+      "step": 307
+    },
+    {
+      "epoch": 0.35648148148148145,
+      "grad_norm": 0.1532718688249588,
+      "learning_rate": 6.44675925925926e-06,
+      "loss": 1.0938,
+      "step": 308
+    },
+    {
+      "epoch": 0.3576388888888889,
+      "grad_norm": 0.15940245985984802,
+      "learning_rate": 6.435185185185186e-06,
+      "loss": 1.0954,
+      "step": 309
+    },
+    {
+      "epoch": 0.3587962962962963,
+      "grad_norm": 0.1621445119380951,
+      "learning_rate": 6.423611111111112e-06,
+      "loss": 1.1109,
+      "step": 310
+    },
+    {
+      "epoch": 0.3599537037037037,
+      "grad_norm": 0.15937702357769012,
+      "learning_rate": 6.4120370370370375e-06,
+      "loss": 1.0635,
+      "step": 311
+    },
+    {
+      "epoch": 0.3611111111111111,
+      "grad_norm": 0.16414445638656616,
+      "learning_rate": 6.400462962962963e-06,
+      "loss": 1.0817,
+      "step": 312
+    },
+    {
+      "epoch": 0.36226851851851855,
+      "grad_norm": 0.16262070834636688,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 1.1235,
+      "step": 313
+    },
+    {
+      "epoch": 0.36342592592592593,
+      "grad_norm": 0.23703633248806,
+      "learning_rate": 6.377314814814816e-06,
+      "loss": 1.0968,
+      "step": 314
+    },
+    {
+      "epoch": 0.3645833333333333,
+      "grad_norm": 0.1544935554265976,
+      "learning_rate": 6.365740740740741e-06,
+      "loss": 1.0329,
+      "step": 315
+    },
+    {
+      "epoch": 0.36574074074074076,
+      "grad_norm": 0.1603870689868927,
+      "learning_rate": 6.354166666666667e-06,
+      "loss": 1.1552,
+      "step": 316
+    },
+    {
+      "epoch": 0.36689814814814814,
+      "grad_norm": 0.16536477208137512,
+      "learning_rate": 6.342592592592594e-06,
+      "loss": 1.0654,
+      "step": 317
+    },
+    {
+      "epoch": 0.3680555555555556,
+      "grad_norm": 0.153824120759964,
+      "learning_rate": 6.331018518518519e-06,
+      "loss": 1.025,
+      "step": 318
+    },
+    {
+      "epoch": 0.36921296296296297,
+      "grad_norm": 0.17179211974143982,
+      "learning_rate": 6.319444444444445e-06,
+      "loss": 1.1973,
+      "step": 319
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 0.16214361786842346,
+      "learning_rate": 6.307870370370371e-06,
+      "loss": 1.0792,
+      "step": 320
+    },
+    {
+      "epoch": 0.3715277777777778,
+      "grad_norm": 0.17154665291309357,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 1.1232,
+      "step": 321
+    },
+    {
+      "epoch": 0.3726851851851852,
+      "grad_norm": 0.16794797778129578,
+      "learning_rate": 6.284722222222223e-06,
+      "loss": 1.0727,
+      "step": 322
+    },
+    {
+      "epoch": 0.3738425925925926,
+      "grad_norm": 0.16869844496250153,
+      "learning_rate": 6.2731481481481485e-06,
+      "loss": 1.0605,
+      "step": 323
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.1624298244714737,
+      "learning_rate": 6.261574074074075e-06,
+      "loss": 1.1402,
+      "step": 324
+    },
+    {
+      "epoch": 0.3761574074074074,
+      "grad_norm": 0.16999289393424988,
+      "learning_rate": 6.25e-06,
+      "loss": 1.0476,
+      "step": 325
+    },
+    {
+      "epoch": 0.3773148148148148,
+      "grad_norm": 0.18032115697860718,
+      "learning_rate": 6.238425925925926e-06,
+      "loss": 1.0555,
+      "step": 326
+    },
+    {
+      "epoch": 0.3784722222222222,
+      "grad_norm": 0.18404638767242432,
+      "learning_rate": 6.226851851851853e-06,
+      "loss": 1.1022,
+      "step": 327
+    },
+    {
+      "epoch": 0.37962962962962965,
+      "grad_norm": 0.17180074751377106,
+      "learning_rate": 6.2152777777777785e-06,
+      "loss": 1.1233,
+      "step": 328
+    },
+    {
+      "epoch": 0.38078703703703703,
+      "grad_norm": 0.16245652735233307,
+      "learning_rate": 6.203703703703704e-06,
+      "loss": 1.088,
+      "step": 329
+    },
+    {
+      "epoch": 0.3819444444444444,
+      "grad_norm": 0.16371048986911774,
+      "learning_rate": 6.1921296296296294e-06,
+      "loss": 1.0627,
+      "step": 330
+    },
+    {
+      "epoch": 0.38310185185185186,
+      "grad_norm": 0.17475458979606628,
+      "learning_rate": 6.180555555555557e-06,
+      "loss": 1.0689,
+      "step": 331
+    },
+    {
+      "epoch": 0.38425925925925924,
+      "grad_norm": 0.1612567901611328,
+      "learning_rate": 6.168981481481482e-06,
+      "loss": 1.0404,
+      "step": 332
+    },
+    {
+      "epoch": 0.3854166666666667,
+      "grad_norm": 0.16201715171337128,
+      "learning_rate": 6.157407407407408e-06,
+      "loss": 1.0846,
+      "step": 333
+    },
+    {
+      "epoch": 0.38657407407407407,
+      "grad_norm": 0.15960504114627838,
+      "learning_rate": 6.145833333333334e-06,
+      "loss": 1.0497,
+      "step": 334
+    },
+    {
+      "epoch": 0.38773148148148145,
+      "grad_norm": 0.1676989644765854,
+      "learning_rate": 6.134259259259259e-06,
+      "loss": 1.0672,
+      "step": 335
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 0.17084524035453796,
+      "learning_rate": 6.122685185185186e-06,
+      "loss": 1.0504,
+      "step": 336
+    },
+    {
+      "epoch": 0.3900462962962963,
+      "grad_norm": 0.20204032957553864,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 1.0935,
+      "step": 337
+    },
+    {
+      "epoch": 0.3912037037037037,
+      "grad_norm": 0.16617536544799805,
+      "learning_rate": 6.0995370370370376e-06,
+      "loss": 1.0946,
+      "step": 338
+    },
+    {
+      "epoch": 0.3923611111111111,
+      "grad_norm": 0.17326408624649048,
+      "learning_rate": 6.087962962962963e-06,
+      "loss": 1.072,
+      "step": 339
+    },
+    {
+      "epoch": 0.39351851851851855,
+      "grad_norm": 0.1672402322292328,
+      "learning_rate": 6.0763888888888885e-06,
+      "loss": 1.0738,
+      "step": 340
+    },
+    {
+      "epoch": 0.39467592592592593,
+      "grad_norm": 0.18285104632377625,
+      "learning_rate": 6.064814814814816e-06,
+      "loss": 1.0381,
+      "step": 341
+    },
+    {
+      "epoch": 0.3958333333333333,
+      "grad_norm": 0.19718050956726074,
+      "learning_rate": 6.053240740740741e-06,
+      "loss": 1.0386,
+      "step": 342
+    },
+    {
+      "epoch": 0.39699074074074076,
+      "grad_norm": 0.1670585423707962,
+      "learning_rate": 6.041666666666667e-06,
+      "loss": 1.1144,
+      "step": 343
+    },
+    {
+      "epoch": 0.39814814814814814,
+      "grad_norm": 0.1710127294063568,
+      "learning_rate": 6.030092592592594e-06,
+      "loss": 1.1076,
+      "step": 344
+    },
+    {
+      "epoch": 0.3993055555555556,
+      "grad_norm": 0.1666959673166275,
+      "learning_rate": 6.018518518518519e-06,
+      "loss": 1.0994,
+      "step": 345
+    },
+    {
+      "epoch": 0.40046296296296297,
+      "grad_norm": 0.16433195769786835,
+      "learning_rate": 6.006944444444445e-06,
+      "loss": 1.0441,
+      "step": 346
+    },
+    {
+      "epoch": 0.40162037037037035,
+      "grad_norm": 0.1659228801727295,
+      "learning_rate": 5.995370370370371e-06,
+      "loss": 1.0886,
+      "step": 347
+    },
+    {
+      "epoch": 0.4027777777777778,
+      "grad_norm": 0.19117744266986847,
+      "learning_rate": 5.983796296296297e-06,
+      "loss": 1.0332,
+      "step": 348
+    },
+    {
+      "epoch": 0.4039351851851852,
+      "grad_norm": 0.17095550894737244,
+      "learning_rate": 5.972222222222222e-06,
+      "loss": 1.1132,
+      "step": 349
+    },
+    {
+      "epoch": 0.4050925925925926,
+      "grad_norm": 0.1593662053346634,
+      "learning_rate": 5.9606481481481485e-06,
+      "loss": 1.0432,
+      "step": 350
+    },
+    {
+      "epoch": 0.40625,
+      "grad_norm": 0.257305383682251,
+      "learning_rate": 5.949074074074075e-06,
+      "loss": 0.9819,
+      "step": 351
+    },
+    {
+      "epoch": 0.4074074074074074,
+      "grad_norm": 0.17508187890052795,
+      "learning_rate": 5.9375e-06,
+      "loss": 1.0263,
+      "step": 352
+    },
+    {
+      "epoch": 0.4085648148148148,
+      "grad_norm": 0.17764556407928467,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 1.0585,
+      "step": 353
+    },
+    {
+      "epoch": 0.4097222222222222,
+      "grad_norm": 0.1692568063735962,
+      "learning_rate": 5.914351851851853e-06,
+      "loss": 1.0111,
+      "step": 354
+    },
+    {
+      "epoch": 0.41087962962962965,
+      "grad_norm": 0.1803068220615387,
+      "learning_rate": 5.9027777777777785e-06,
+      "loss": 1.0209,
+      "step": 355
+    },
+    {
+      "epoch": 0.41203703703703703,
+      "grad_norm": 0.17242421209812164,
+      "learning_rate": 5.891203703703704e-06,
+      "loss": 1.0509,
+      "step": 356
+    },
+    {
+      "epoch": 0.4131944444444444,
+      "grad_norm": 0.17684854567050934,
+      "learning_rate": 5.8796296296296295e-06,
+      "loss": 1.0462,
+      "step": 357
+    },
+    {
+      "epoch": 0.41435185185185186,
+      "grad_norm": 0.16622677445411682,
+      "learning_rate": 5.868055555555557e-06,
+      "loss": 1.0227,
+      "step": 358
+    },
+    {
+      "epoch": 0.41550925925925924,
+      "grad_norm": 0.18189331889152527,
+      "learning_rate": 5.856481481481482e-06,
+      "loss": 1.0007,
+      "step": 359
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 0.17341043055057526,
+      "learning_rate": 5.844907407407408e-06,
+      "loss": 1.0132,
+      "step": 360
+    },
+    {
+      "epoch": 0.41782407407407407,
+      "grad_norm": 0.17046672105789185,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 1.0724,
+      "step": 361
+    },
+    {
+      "epoch": 0.41898148148148145,
+      "grad_norm": 0.1714986264705658,
+      "learning_rate": 5.8217592592592594e-06,
+      "loss": 1.0817,
+      "step": 362
+    },
+    {
+      "epoch": 0.4201388888888889,
+      "grad_norm": 0.17200608551502228,
+      "learning_rate": 5.810185185185186e-06,
+      "loss": 1.1255,
+      "step": 363
+    },
+    {
+      "epoch": 0.4212962962962963,
+      "grad_norm": 0.16442537307739258,
+      "learning_rate": 5.798611111111112e-06,
+      "loss": 1.0655,
+      "step": 364
+    },
+    {
+      "epoch": 0.4224537037037037,
+      "grad_norm": 0.17119750380516052,
+      "learning_rate": 5.787037037037038e-06,
+      "loss": 1.1056,
+      "step": 365
+    },
+    {
+      "epoch": 0.4236111111111111,
+      "grad_norm": 0.26399606466293335,
+      "learning_rate": 5.775462962962963e-06,
+      "loss": 1.0784,
+      "step": 366
+    },
+    {
+      "epoch": 0.42476851851851855,
+      "grad_norm": 0.17613111436367035,
+      "learning_rate": 5.7638888888888886e-06,
+      "loss": 1.0825,
+      "step": 367
+    },
+    {
+      "epoch": 0.42592592592592593,
+      "grad_norm": 0.1671566367149353,
+      "learning_rate": 5.752314814814816e-06,
+      "loss": 1.0748,
+      "step": 368
+    },
+    {
+      "epoch": 0.4270833333333333,
+      "grad_norm": 0.1717667132616043,
+      "learning_rate": 5.740740740740741e-06,
+      "loss": 1.0559,
+      "step": 369
+    },
+    {
+      "epoch": 0.42824074074074076,
+      "grad_norm": 0.1810443252325058,
+      "learning_rate": 5.729166666666667e-06,
+      "loss": 1.1115,
+      "step": 370
+    },
+    {
+      "epoch": 0.42939814814814814,
+      "grad_norm": 0.1782471239566803,
+      "learning_rate": 5.717592592592593e-06,
+      "loss": 1.1441,
+      "step": 371
+    },
+    {
+      "epoch": 0.4305555555555556,
+      "grad_norm": 0.16452831029891968,
+      "learning_rate": 5.706018518518519e-06,
+      "loss": 1.0284,
+      "step": 372
+    },
+    {
+      "epoch": 0.43171296296296297,
+      "grad_norm": 0.1789056956768036,
+      "learning_rate": 5.694444444444445e-06,
+      "loss": 1.0889,
+      "step": 373
+    },
+    {
+      "epoch": 0.43287037037037035,
+      "grad_norm": 0.18854652345180511,
+      "learning_rate": 5.682870370370371e-06,
+      "loss": 0.982,
+      "step": 374
+    },
+    {
+      "epoch": 0.4340277777777778,
+      "grad_norm": 0.16741988062858582,
+      "learning_rate": 5.671296296296297e-06,
+      "loss": 1.0511,
+      "step": 375
+    },
+    {
+      "epoch": 0.4351851851851852,
+      "grad_norm": 0.17495352029800415,
+      "learning_rate": 5.659722222222222e-06,
+      "loss": 1.0276,
+      "step": 376
+    },
+    {
+      "epoch": 0.4363425925925926,
+      "grad_norm": 0.18103112280368805,
+      "learning_rate": 5.6481481481481485e-06,
+      "loss": 1.0774,
+      "step": 377
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.17093877494335175,
+      "learning_rate": 5.636574074074075e-06,
+      "loss": 1.0617,
+      "step": 378
+    },
+    {
+      "epoch": 0.4386574074074074,
+      "grad_norm": 0.17672328650951385,
+      "learning_rate": 5.625e-06,
+      "loss": 1.1172,
+      "step": 379
+    },
+    {
+      "epoch": 0.4398148148148148,
+      "grad_norm": 0.17157655954360962,
+      "learning_rate": 5.613425925925926e-06,
+      "loss": 1.0651,
+      "step": 380
+    },
+    {
+      "epoch": 0.4409722222222222,
+      "grad_norm": 0.21019726991653442,
+      "learning_rate": 5.601851851851853e-06,
+      "loss": 1.0853,
+      "step": 381
+    },
+    {
+      "epoch": 0.44212962962962965,
+      "grad_norm": 0.16854119300842285,
+      "learning_rate": 5.5902777777777785e-06,
+      "loss": 1.0568,
+      "step": 382
+    },
+    {
+      "epoch": 0.44328703703703703,
+      "grad_norm": 0.17988835275173187,
+      "learning_rate": 5.578703703703704e-06,
+      "loss": 1.0607,
+      "step": 383
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.1783677488565445,
+      "learning_rate": 5.5671296296296295e-06,
+      "loss": 1.135,
+      "step": 384
+    },
+    {
+      "epoch": 0.44560185185185186,
+      "grad_norm": 0.1827668398618698,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 1.0959,
+      "step": 385
+    },
+    {
+      "epoch": 0.44675925925925924,
+      "grad_norm": 0.1617071032524109,
+      "learning_rate": 5.543981481481482e-06,
+      "loss": 1.0016,
+      "step": 386
+    },
+    {
+      "epoch": 0.4479166666666667,
+      "grad_norm": 0.17546425759792328,
+      "learning_rate": 5.532407407407408e-06,
+      "loss": 1.0316,
+      "step": 387
+    },
+    {
+      "epoch": 0.44907407407407407,
+      "grad_norm": 0.19249460101127625,
+      "learning_rate": 5.520833333333334e-06,
+      "loss": 1.0625,
+      "step": 388
+    },
+    {
+      "epoch": 0.45023148148148145,
+      "grad_norm": 0.18015442788600922,
+      "learning_rate": 5.5092592592592595e-06,
+      "loss": 1.0718,
+      "step": 389
+    },
+    {
+      "epoch": 0.4513888888888889,
+      "grad_norm": 0.17945371568202972,
+      "learning_rate": 5.497685185185185e-06,
+      "loss": 1.0278,
+      "step": 390
+    },
+    {
+      "epoch": 0.4525462962962963,
+      "grad_norm": 0.21291273832321167,
+      "learning_rate": 5.486111111111112e-06,
+      "loss": 1.0924,
+      "step": 391
+    },
+    {
+      "epoch": 0.4537037037037037,
+      "grad_norm": 0.18178021907806396,
+      "learning_rate": 5.474537037037038e-06,
+      "loss": 1.0163,
+      "step": 392
+    },
+    {
+      "epoch": 0.4548611111111111,
+      "grad_norm": 0.16679397225379944,
+      "learning_rate": 5.462962962962963e-06,
+      "loss": 1.0027,
+      "step": 393
+    },
+    {
+      "epoch": 0.45601851851851855,
+      "grad_norm": 0.19789078831672668,
+      "learning_rate": 5.451388888888889e-06,
+      "loss": 1.0478,
+      "step": 394
+    },
+    {
+      "epoch": 0.45717592592592593,
+      "grad_norm": 0.183577299118042,
+      "learning_rate": 5.439814814814816e-06,
+      "loss": 1.0869,
+      "step": 395
+    },
+    {
+      "epoch": 0.4583333333333333,
+      "grad_norm": 0.17376714944839478,
+      "learning_rate": 5.428240740740741e-06,
+      "loss": 1.0763,
+      "step": 396
+    },
+    {
+      "epoch": 0.45949074074074076,
+      "grad_norm": 0.17424362897872925,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 1.1034,
+      "step": 397
+    },
+    {
+      "epoch": 0.46064814814814814,
+      "grad_norm": 0.18567875027656555,
+      "learning_rate": 5.405092592592593e-06,
+      "loss": 1.1487,
+      "step": 398
+    },
+    {
+      "epoch": 0.4618055555555556,
+      "grad_norm": 0.1731259822845459,
+      "learning_rate": 5.3935185185185194e-06,
+      "loss": 1.0371,
+      "step": 399
+    },
+    {
+      "epoch": 0.46296296296296297,
+      "grad_norm": 0.1738322377204895,
+      "learning_rate": 5.381944444444445e-06,
+      "loss": 1.0892,
+      "step": 400
+    },
+    {
+      "epoch": 0.46412037037037035,
+      "grad_norm": 0.181026428937912,
+      "learning_rate": 5.370370370370371e-06,
+      "loss": 1.0335,
+      "step": 401
+    },
+    {
+      "epoch": 0.4652777777777778,
+      "grad_norm": 0.1937108188867569,
+      "learning_rate": 5.358796296296297e-06,
+      "loss": 1.0659,
+      "step": 402
+    },
+    {
+      "epoch": 0.4664351851851852,
+      "grad_norm": 0.1845736801624298,
+      "learning_rate": 5.347222222222222e-06,
+      "loss": 1.0769,
+      "step": 403
+    },
+    {
+      "epoch": 0.4675925925925926,
+      "grad_norm": 0.17622292041778564,
+      "learning_rate": 5.335648148148148e-06,
+      "loss": 1.0344,
+      "step": 404
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 0.19735385477542877,
+      "learning_rate": 5.324074074074075e-06,
+      "loss": 1.1107,
+      "step": 405
+    },
+    {
+      "epoch": 0.4699074074074074,
+      "grad_norm": 0.1723097264766693,
+      "learning_rate": 5.3125e-06,
+      "loss": 1.0776,
+      "step": 406
+    },
+    {
+      "epoch": 0.4710648148148148,
+      "grad_norm": 0.1863865703344345,
+      "learning_rate": 5.300925925925926e-06,
+      "loss": 1.0645,
+      "step": 407
+    },
+    {
+      "epoch": 0.4722222222222222,
+      "grad_norm": 0.18787802755832672,
+      "learning_rate": 5.289351851851853e-06,
+      "loss": 1.0585,
+      "step": 408
+    },
+    {
+      "epoch": 0.47337962962962965,
+      "grad_norm": 0.17600968480110168,
+      "learning_rate": 5.2777777777777785e-06,
+      "loss": 1.0306,
+      "step": 409
+    },
+    {
+      "epoch": 0.47453703703703703,
+      "grad_norm": 0.1793355494737625,
+      "learning_rate": 5.266203703703704e-06,
+      "loss": 1.0337,
+      "step": 410
+    },
+    {
+      "epoch": 0.4756944444444444,
+      "grad_norm": 0.17563559114933014,
+      "learning_rate": 5.2546296296296295e-06,
+      "loss": 1.0991,
+      "step": 411
+    },
+    {
+      "epoch": 0.47685185185185186,
+      "grad_norm": 0.17623300850391388,
+      "learning_rate": 5.243055555555556e-06,
+      "loss": 1.0204,
+      "step": 412
+    },
+    {
+      "epoch": 0.47800925925925924,
+      "grad_norm": 0.17649118602275848,
+      "learning_rate": 5.231481481481482e-06,
+      "loss": 1.0523,
+      "step": 413
+    },
+    {
+      "epoch": 0.4791666666666667,
+      "grad_norm": 0.1824215203523636,
+      "learning_rate": 5.219907407407408e-06,
+      "loss": 1.0839,
+      "step": 414
+    },
+    {
+      "epoch": 0.48032407407407407,
+      "grad_norm": 0.18569543957710266,
+      "learning_rate": 5.208333333333334e-06,
+      "loss": 1.0277,
+      "step": 415
+    },
+    {
+      "epoch": 0.48148148148148145,
+      "grad_norm": 0.18239271640777588,
+      "learning_rate": 5.1967592592592595e-06,
+      "loss": 0.9784,
+      "step": 416
+    },
+    {
+      "epoch": 0.4826388888888889,
+      "grad_norm": 0.17993135750293732,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 1.0564,
+      "step": 417
+    },
+    {
+      "epoch": 0.4837962962962963,
+      "grad_norm": 0.18158768117427826,
+      "learning_rate": 5.173611111111112e-06,
+      "loss": 1.0332,
+      "step": 418
+    },
+    {
+      "epoch": 0.4849537037037037,
+      "grad_norm": 0.1836976557970047,
+      "learning_rate": 5.162037037037038e-06,
+      "loss": 1.0209,
+      "step": 419
+    },
+    {
+      "epoch": 0.4861111111111111,
+      "grad_norm": 0.19026342034339905,
+      "learning_rate": 5.150462962962963e-06,
+      "loss": 1.1446,
+      "step": 420
+    },
+    {
+      "epoch": 0.48726851851851855,
+      "grad_norm": 0.17658670246601105,
+      "learning_rate": 5.138888888888889e-06,
+      "loss": 0.9938,
+      "step": 421
+    },
+    {
+      "epoch": 0.48842592592592593,
+      "grad_norm": 0.1957542598247528,
+      "learning_rate": 5.127314814814816e-06,
+      "loss": 1.0233,
+      "step": 422
+    },
+    {
+      "epoch": 0.4895833333333333,
+      "grad_norm": 0.18295548856258392,
+      "learning_rate": 5.115740740740741e-06,
+      "loss": 1.0643,
+      "step": 423
+    },
+    {
+      "epoch": 0.49074074074074076,
+      "grad_norm": 0.18762163817882538,
+      "learning_rate": 5.104166666666667e-06,
+      "loss": 1.0762,
+      "step": 424
+    },
+    {
+      "epoch": 0.49189814814814814,
+      "grad_norm": 0.18123242259025574,
+      "learning_rate": 5.092592592592593e-06,
+      "loss": 1.0738,
+      "step": 425
+    },
+    {
+      "epoch": 0.4930555555555556,
+      "grad_norm": 0.17455293238162994,
+      "learning_rate": 5.081018518518519e-06,
+      "loss": 1.0094,
+      "step": 426
+    },
+    {
+      "epoch": 0.49421296296296297,
+      "grad_norm": 0.1735786497592926,
+      "learning_rate": 5.069444444444445e-06,
+      "loss": 1.0997,
+      "step": 427
+    },
+    {
+      "epoch": 0.49537037037037035,
+      "grad_norm": 0.18896614015102386,
+      "learning_rate": 5.057870370370371e-06,
+      "loss": 1.0226,
+      "step": 428
+    },
+    {
+      "epoch": 0.4965277777777778,
+      "grad_norm": 0.2275708019733429,
+      "learning_rate": 5.046296296296297e-06,
+      "loss": 1.0348,
+      "step": 429
+    },
+    {
+      "epoch": 0.4976851851851852,
+      "grad_norm": 0.1853165179491043,
+      "learning_rate": 5.034722222222222e-06,
+      "loss": 1.0053,
+      "step": 430
+    },
+    {
+      "epoch": 0.4988425925925926,
+      "grad_norm": 0.1842850297689438,
+      "learning_rate": 5.023148148148148e-06,
+      "loss": 1.0847,
+      "step": 431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19081000983715057,
+      "learning_rate": 5.011574074074075e-06,
+      "loss": 1.0571,
+      "step": 432
+    },
+    {
+      "epoch": 0.5011574074074074,
+      "grad_norm": 0.18483759462833405,
+      "learning_rate": 5e-06,
+      "loss": 1.017,
+      "step": 433
+    },
+    {
+      "epoch": 0.5023148148148148,
+      "grad_norm": 0.18229275941848755,
+      "learning_rate": 4.988425925925927e-06,
+      "loss": 1.0299,
+      "step": 434
+    },
+    {
+      "epoch": 0.5034722222222222,
+      "grad_norm": 0.19052374362945557,
+      "learning_rate": 4.976851851851852e-06,
+      "loss": 1.1137,
+      "step": 435
+    },
+    {
+      "epoch": 0.5046296296296297,
+      "grad_norm": 0.19372689723968506,
+      "learning_rate": 4.9652777777777786e-06,
+      "loss": 1.0611,
+      "step": 436
+    },
+    {
+      "epoch": 0.5057870370370371,
+      "grad_norm": 0.18391257524490356,
+      "learning_rate": 4.953703703703704e-06,
+      "loss": 1.0225,
+      "step": 437
+    },
+    {
+      "epoch": 0.5069444444444444,
+      "grad_norm": 0.18629541993141174,
+      "learning_rate": 4.94212962962963e-06,
+      "loss": 0.9979,
+      "step": 438
+    },
+    {
+      "epoch": 0.5081018518518519,
+      "grad_norm": 0.19393537938594818,
+      "learning_rate": 4.930555555555556e-06,
+      "loss": 1.1086,
+      "step": 439
+    },
+    {
+      "epoch": 0.5092592592592593,
+      "grad_norm": 0.1992591768503189,
+      "learning_rate": 4.918981481481482e-06,
+      "loss": 1.0497,
+      "step": 440
+    },
+    {
+      "epoch": 0.5104166666666666,
+      "grad_norm": 0.18729424476623535,
+      "learning_rate": 4.907407407407408e-06,
+      "loss": 1.0624,
+      "step": 441
+    },
+    {
+      "epoch": 0.5115740740740741,
+      "grad_norm": 0.18621307611465454,
+      "learning_rate": 4.895833333333333e-06,
+      "loss": 1.01,
+      "step": 442
+    },
+    {
+      "epoch": 0.5127314814814815,
+      "grad_norm": 0.1837424784898758,
+      "learning_rate": 4.8842592592592595e-06,
+      "loss": 1.055,
+      "step": 443
+    },
+    {
+      "epoch": 0.5138888888888888,
+      "grad_norm": 0.2010790854692459,
+      "learning_rate": 4.872685185185186e-06,
+      "loss": 1.1718,
+      "step": 444
+    },
+    {
+      "epoch": 0.5150462962962963,
+      "grad_norm": 0.1856854408979416,
+      "learning_rate": 4.861111111111111e-06,
+      "loss": 1.0231,
+      "step": 445
+    },
+    {
+      "epoch": 0.5162037037037037,
+      "grad_norm": 0.18698154389858246,
+      "learning_rate": 4.849537037037038e-06,
+      "loss": 1.0363,
+      "step": 446
+    },
+    {
+      "epoch": 0.5173611111111112,
+      "grad_norm": 0.18074962496757507,
+      "learning_rate": 4.837962962962963e-06,
+      "loss": 1.0585,
+      "step": 447
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 0.18344129621982574,
+      "learning_rate": 4.8263888888888895e-06,
+      "loss": 1.0549,
+      "step": 448
+    },
+    {
+      "epoch": 0.5196759259259259,
+      "grad_norm": 0.1772606372833252,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 1.1409,
+      "step": 449
+    },
+    {
+      "epoch": 0.5208333333333334,
+      "grad_norm": 0.18440352380275726,
+      "learning_rate": 4.803240740740741e-06,
+      "loss": 1.0651,
+      "step": 450
+    },
+    {
+      "epoch": 0.5219907407407407,
+      "grad_norm": 0.2034968137741089,
+      "learning_rate": 4.791666666666668e-06,
+      "loss": 0.9843,
+      "step": 451
+    },
+    {
+      "epoch": 0.5231481481481481,
+      "grad_norm": 0.18375763297080994,
+      "learning_rate": 4.780092592592593e-06,
+      "loss": 1.0571,
+      "step": 452
+    },
+    {
+      "epoch": 0.5243055555555556,
+      "grad_norm": 0.18968044221401215,
+      "learning_rate": 4.768518518518519e-06,
+      "loss": 1.0843,
+      "step": 453
+    },
+    {
+      "epoch": 0.5254629629629629,
+      "grad_norm": 0.18977278470993042,
+      "learning_rate": 4.756944444444445e-06,
+      "loss": 1.076,
+      "step": 454
+    },
+    {
+      "epoch": 0.5266203703703703,
+      "grad_norm": 0.191218301653862,
+      "learning_rate": 4.7453703703703705e-06,
+      "loss": 1.1139,
+      "step": 455
+    },
+    {
+      "epoch": 0.5277777777777778,
+      "grad_norm": 0.18904373049736023,
+      "learning_rate": 4.733796296296297e-06,
+      "loss": 1.1205,
+      "step": 456
+    },
+    {
+      "epoch": 0.5289351851851852,
+      "grad_norm": 0.19077414274215698,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 1.086,
+      "step": 457
+    },
+    {
+      "epoch": 0.5300925925925926,
+      "grad_norm": 0.19545550644397736,
+      "learning_rate": 4.710648148148149e-06,
+      "loss": 1.0117,
+      "step": 458
+    },
+    {
+      "epoch": 0.53125,
+      "grad_norm": 0.19482102990150452,
+      "learning_rate": 4.699074074074074e-06,
+      "loss": 1.0768,
+      "step": 459
+    },
+    {
+      "epoch": 0.5324074074074074,
+      "grad_norm": 0.1793646365404129,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 1.0542,
+      "step": 460
+    },
+    {
+      "epoch": 0.5335648148148148,
+      "grad_norm": 0.18208812177181244,
+      "learning_rate": 4.675925925925927e-06,
+      "loss": 1.0722,
+      "step": 461
+    },
+    {
+      "epoch": 0.5347222222222222,
+      "grad_norm": 0.19631284475326538,
+      "learning_rate": 4.664351851851852e-06,
+      "loss": 1.0069,
+      "step": 462
+    },
+    {
+      "epoch": 0.5358796296296297,
+      "grad_norm": 0.1896466761827469,
+      "learning_rate": 4.652777777777779e-06,
+      "loss": 1.0232,
+      "step": 463
+    },
+    {
+      "epoch": 0.5370370370370371,
+      "grad_norm": 0.17967751622200012,
+      "learning_rate": 4.641203703703704e-06,
+      "loss": 1.0982,
+      "step": 464
+    },
+    {
+      "epoch": 0.5381944444444444,
+      "grad_norm": 0.19527268409729004,
+      "learning_rate": 4.62962962962963e-06,
+      "loss": 1.0464,
+      "step": 465
+    },
+    {
+      "epoch": 0.5393518518518519,
+      "grad_norm": 0.18822702765464783,
+      "learning_rate": 4.618055555555556e-06,
+      "loss": 1.0881,
+      "step": 466
+    },
+    {
+      "epoch": 0.5405092592592593,
+      "grad_norm": 0.19704404473304749,
+      "learning_rate": 4.606481481481481e-06,
+      "loss": 1.0765,
+      "step": 467
+    },
+    {
+      "epoch": 0.5416666666666666,
+      "grad_norm": 0.1972915083169937,
+      "learning_rate": 4.594907407407408e-06,
+      "loss": 1.1231,
+      "step": 468
+    },
+    {
+      "epoch": 0.5428240740740741,
+      "grad_norm": 0.18384619057178497,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 1.0372,
+      "step": 469
+    },
+    {
+      "epoch": 0.5439814814814815,
+      "grad_norm": 0.19371986389160156,
+      "learning_rate": 4.5717592592592595e-06,
+      "loss": 1.0317,
+      "step": 470
+    },
+    {
+      "epoch": 0.5451388888888888,
+      "grad_norm": 0.21294532716274261,
+      "learning_rate": 4.560185185185186e-06,
+      "loss": 0.9952,
+      "step": 471
+    },
+    {
+      "epoch": 0.5462962962962963,
+      "grad_norm": 0.19161927700042725,
+      "learning_rate": 4.548611111111111e-06,
+      "loss": 1.0672,
+      "step": 472
+    },
+    {
+      "epoch": 0.5474537037037037,
+      "grad_norm": 0.18557506799697876,
+      "learning_rate": 4.537037037037038e-06,
+      "loss": 1.0299,
+      "step": 473
+    },
+    {
+      "epoch": 0.5486111111111112,
+      "grad_norm": 0.19008249044418335,
+      "learning_rate": 4.525462962962963e-06,
+      "loss": 1.0675,
+      "step": 474
+    },
+    {
+      "epoch": 0.5497685185185185,
+      "grad_norm": 0.19142603874206543,
+      "learning_rate": 4.5138888888888895e-06,
+      "loss": 1.0979,
+      "step": 475
+    },
+    {
+      "epoch": 0.5509259259259259,
+      "grad_norm": 0.19404229521751404,
+      "learning_rate": 4.502314814814815e-06,
+      "loss": 0.9621,
+      "step": 476
+    },
+    {
+      "epoch": 0.5520833333333334,
+      "grad_norm": 0.19568336009979248,
+      "learning_rate": 4.490740740740741e-06,
+      "loss": 1.0802,
+      "step": 477
+    },
+    {
+      "epoch": 0.5532407407407407,
+      "grad_norm": 0.1997448205947876,
+      "learning_rate": 4.479166666666667e-06,
+      "loss": 1.1484,
+      "step": 478
+    },
+    {
+      "epoch": 0.5543981481481481,
+      "grad_norm": 0.2235734760761261,
+      "learning_rate": 4.467592592592593e-06,
+      "loss": 1.0239,
+      "step": 479
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 0.20642857253551483,
+      "learning_rate": 4.456018518518519e-06,
+      "loss": 1.0456,
+      "step": 480
+    },
+    {
+      "epoch": 0.5567129629629629,
+      "grad_norm": 0.1832839995622635,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.0074,
+      "step": 481
+    },
+    {
+      "epoch": 0.5578703703703703,
+      "grad_norm": 0.1859252154827118,
+      "learning_rate": 4.4328703703703705e-06,
+      "loss": 1.0592,
+      "step": 482
+    },
+    {
+      "epoch": 0.5590277777777778,
+      "grad_norm": 0.19628946483135223,
+      "learning_rate": 4.421296296296297e-06,
+      "loss": 0.9807,
+      "step": 483
+    },
+    {
+      "epoch": 0.5601851851851852,
+      "grad_norm": 0.18916712701320648,
+      "learning_rate": 4.409722222222222e-06,
+      "loss": 0.9956,
+      "step": 484
+    },
+    {
+      "epoch": 0.5613425925925926,
+      "grad_norm": 0.19245581328868866,
+      "learning_rate": 4.398148148148149e-06,
+      "loss": 1.0348,
+      "step": 485
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.18789884448051453,
+      "learning_rate": 4.386574074074074e-06,
+      "loss": 1.0984,
+      "step": 486
+    },
+    {
+      "epoch": 0.5636574074074074,
+      "grad_norm": 0.18986792862415314,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.9934,
+      "step": 487
+    },
+    {
+      "epoch": 0.5648148148148148,
+      "grad_norm": 0.1932685375213623,
+      "learning_rate": 4.363425925925927e-06,
+      "loss": 1.1058,
+      "step": 488
+    },
+    {
+      "epoch": 0.5659722222222222,
+      "grad_norm": 0.21794289350509644,
+      "learning_rate": 4.351851851851852e-06,
+      "loss": 1.0482,
+      "step": 489
+    },
+    {
+      "epoch": 0.5671296296296297,
+      "grad_norm": 0.19237902760505676,
+      "learning_rate": 4.340277777777779e-06,
+      "loss": 1.0242,
+      "step": 490
+    },
+    {
+      "epoch": 0.5682870370370371,
+      "grad_norm": 0.19967250525951385,
+      "learning_rate": 4.328703703703704e-06,
+      "loss": 1.0208,
+      "step": 491
+    },
+    {
+      "epoch": 0.5694444444444444,
+      "grad_norm": 0.18474088609218597,
+      "learning_rate": 4.31712962962963e-06,
+      "loss": 1.0251,
+      "step": 492
+    },
+    {
+      "epoch": 0.5706018518518519,
+      "grad_norm": 0.22464704513549805,
+      "learning_rate": 4.305555555555556e-06,
+      "loss": 0.9969,
+      "step": 493
+    },
+    {
+      "epoch": 0.5717592592592593,
+      "grad_norm": 0.1928233802318573,
+      "learning_rate": 4.293981481481481e-06,
+      "loss": 1.1077,
+      "step": 494
+    },
+    {
+      "epoch": 0.5729166666666666,
+      "grad_norm": 0.20419670641422272,
+      "learning_rate": 4.282407407407408e-06,
+      "loss": 1.0883,
+      "step": 495
+    },
+    {
+      "epoch": 0.5740740740740741,
+      "grad_norm": 0.21001799404621124,
+      "learning_rate": 4.270833333333333e-06,
+      "loss": 1.0765,
+      "step": 496
+    },
+    {
+      "epoch": 0.5752314814814815,
+      "grad_norm": 0.197916179895401,
+      "learning_rate": 4.2592592592592596e-06,
+      "loss": 1.0165,
+      "step": 497
+    },
+    {
+      "epoch": 0.5763888888888888,
+      "grad_norm": 0.18999773263931274,
+      "learning_rate": 4.247685185185186e-06,
+      "loss": 1.0272,
+      "step": 498
+    },
+    {
+      "epoch": 0.5775462962962963,
+      "grad_norm": 0.19263650476932526,
+      "learning_rate": 4.236111111111111e-06,
+      "loss": 0.985,
+      "step": 499
+    },
+    {
+      "epoch": 0.5787037037037037,
+      "grad_norm": 0.19697001576423645,
+      "learning_rate": 4.224537037037038e-06,
+      "loss": 1.1516,
+      "step": 500
+    },
+    {
+      "epoch": 0.5798611111111112,
+      "grad_norm": 0.20765507221221924,
+      "learning_rate": 4.212962962962963e-06,
+      "loss": 1.0993,
+      "step": 501
+    },
+    {
+      "epoch": 0.5810185185185185,
+      "grad_norm": 0.1897115409374237,
+      "learning_rate": 4.2013888888888896e-06,
+      "loss": 1.051,
+      "step": 502
+    },
+    {
+      "epoch": 0.5821759259259259,
+      "grad_norm": 0.20036956667900085,
+      "learning_rate": 4.189814814814815e-06,
+      "loss": 1.0293,
+      "step": 503
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 0.19671155512332916,
+      "learning_rate": 4.178240740740741e-06,
+      "loss": 1.0452,
+      "step": 504
+    },
+    {
+      "epoch": 0.5844907407407407,
+      "grad_norm": 0.1968628168106079,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.0635,
+      "step": 505
+    },
+    {
+      "epoch": 0.5856481481481481,
+      "grad_norm": 0.19419008493423462,
+      "learning_rate": 4.155092592592593e-06,
+      "loss": 1.0393,
+      "step": 506
+    },
+    {
+      "epoch": 0.5868055555555556,
+      "grad_norm": 0.20477062463760376,
+      "learning_rate": 4.143518518518519e-06,
+      "loss": 1.1189,
+      "step": 507
+    },
+    {
+      "epoch": 0.5879629629629629,
+      "grad_norm": 0.19754110276699066,
+      "learning_rate": 4.131944444444444e-06,
+      "loss": 1.0108,
+      "step": 508
+    },
+    {
+      "epoch": 0.5891203703703703,
+      "grad_norm": 0.19897371530532837,
+      "learning_rate": 4.1203703703703705e-06,
+      "loss": 0.9864,
+      "step": 509
+    },
+    {
+      "epoch": 0.5902777777777778,
+      "grad_norm": 0.19264736771583557,
+      "learning_rate": 4.108796296296297e-06,
+      "loss": 1.0757,
+      "step": 510
+    },
+    {
+      "epoch": 0.5914351851851852,
+      "grad_norm": 0.19793768227100372,
+      "learning_rate": 4.097222222222222e-06,
+      "loss": 1.0985,
+      "step": 511
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 0.22821103036403656,
+      "learning_rate": 4.085648148148149e-06,
+      "loss": 1.0662,
+      "step": 512
+    },
+    {
+      "epoch": 0.59375,
+      "grad_norm": 0.2039816677570343,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 1.0677,
+      "step": 513
+    },
+    {
+      "epoch": 0.5949074074074074,
+      "grad_norm": 0.19190923869609833,
+      "learning_rate": 4.0625000000000005e-06,
+      "loss": 1.052,
+      "step": 514
+    },
+    {
+      "epoch": 0.5960648148148148,
+      "grad_norm": 0.18609005212783813,
+      "learning_rate": 4.050925925925927e-06,
+      "loss": 1.0877,
+      "step": 515
+    },
+    {
+      "epoch": 0.5972222222222222,
+      "grad_norm": 0.1913410723209381,
+      "learning_rate": 4.039351851851852e-06,
+      "loss": 1.1173,
+      "step": 516
+    },
+    {
+      "epoch": 0.5983796296296297,
+      "grad_norm": 0.1951320916414261,
+      "learning_rate": 4.027777777777779e-06,
+      "loss": 1.032,
+      "step": 517
+    },
+    {
+      "epoch": 0.5995370370370371,
+      "grad_norm": 0.19345664978027344,
+      "learning_rate": 4.016203703703704e-06,
+      "loss": 1.0731,
+      "step": 518
+    },
+    {
+      "epoch": 0.6006944444444444,
+      "grad_norm": 0.19238971173763275,
+      "learning_rate": 4.00462962962963e-06,
+      "loss": 1.0494,
+      "step": 519
+    },
+    {
+      "epoch": 0.6018518518518519,
+      "grad_norm": 0.24165932834148407,
+      "learning_rate": 3.993055555555556e-06,
+      "loss": 1.0699,
+      "step": 520
+    },
+    {
+      "epoch": 0.6030092592592593,
+      "grad_norm": 0.19671660661697388,
+      "learning_rate": 3.9814814814814814e-06,
+      "loss": 0.9833,
+      "step": 521
+    },
+    {
+      "epoch": 0.6041666666666666,
+      "grad_norm": 0.20090967416763306,
+      "learning_rate": 3.969907407407408e-06,
+      "loss": 1.0286,
+      "step": 522
+    },
+    {
+      "epoch": 0.6053240740740741,
+      "grad_norm": 0.18830811977386475,
+      "learning_rate": 3.958333333333333e-06,
+      "loss": 0.9915,
+      "step": 523
+    },
+    {
+      "epoch": 0.6064814814814815,
+      "grad_norm": 0.2034299075603485,
+      "learning_rate": 3.94675925925926e-06,
+      "loss": 1.0386,
+      "step": 524
+    },
+    {
+      "epoch": 0.6076388888888888,
+      "grad_norm": 0.206466406583786,
+      "learning_rate": 3.935185185185186e-06,
+      "loss": 1.1197,
+      "step": 525
+    },
+    {
+      "epoch": 0.6087962962962963,
+      "grad_norm": 0.1990206092596054,
+      "learning_rate": 3.9236111111111114e-06,
+      "loss": 1.0646,
+      "step": 526
+    },
+    {
+      "epoch": 0.6099537037037037,
+      "grad_norm": 0.18752767145633698,
+      "learning_rate": 3.912037037037038e-06,
+      "loss": 1.0491,
+      "step": 527
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 0.18697869777679443,
+      "learning_rate": 3.900462962962963e-06,
+      "loss": 1.0638,
+      "step": 528
+    },
+    {
+      "epoch": 0.6122685185185185,
+      "grad_norm": 0.19752278923988342,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 1.0748,
+      "step": 529
+    },
+    {
+      "epoch": 0.6134259259259259,
+      "grad_norm": 0.21283791959285736,
+      "learning_rate": 3.877314814814815e-06,
+      "loss": 1.0623,
+      "step": 530
+    },
+    {
+      "epoch": 0.6145833333333334,
+      "grad_norm": 0.19169341027736664,
+      "learning_rate": 3.865740740740741e-06,
+      "loss": 1.0359,
+      "step": 531
+    },
+    {
+      "epoch": 0.6157407407407407,
+      "grad_norm": 0.2018590122461319,
+      "learning_rate": 3.854166666666667e-06,
+      "loss": 0.9971,
+      "step": 532
+    },
+    {
+      "epoch": 0.6168981481481481,
+      "grad_norm": 0.22165873646736145,
+      "learning_rate": 3.842592592592592e-06,
+      "loss": 1.0325,
+      "step": 533
+    },
+    {
+      "epoch": 0.6180555555555556,
+      "grad_norm": 0.2014155089855194,
+      "learning_rate": 3.831018518518519e-06,
+      "loss": 1.0155,
+      "step": 534
+    },
+    {
+      "epoch": 0.6192129629629629,
+      "grad_norm": 0.19486472010612488,
+      "learning_rate": 3.819444444444444e-06,
+      "loss": 1.0517,
+      "step": 535
+    },
+    {
+      "epoch": 0.6203703703703703,
+      "grad_norm": 0.19350384175777435,
+      "learning_rate": 3.8078703703703705e-06,
+      "loss": 1.0234,
+      "step": 536
+    },
+    {
+      "epoch": 0.6215277777777778,
+      "grad_norm": 0.20624063909053802,
+      "learning_rate": 3.796296296296297e-06,
+      "loss": 1.0635,
+      "step": 537
+    },
+    {
+      "epoch": 0.6226851851851852,
+      "grad_norm": 0.19752593338489532,
+      "learning_rate": 3.7847222222222224e-06,
+      "loss": 1.048,
+      "step": 538
+    },
+    {
+      "epoch": 0.6238425925925926,
+      "grad_norm": 0.19572751224040985,
+      "learning_rate": 3.7731481481481487e-06,
+      "loss": 1.0476,
+      "step": 539
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.19526030123233795,
+      "learning_rate": 3.761574074074074e-06,
+      "loss": 1.0884,
+      "step": 540
+    },
+    {
+      "epoch": 0.6261574074074074,
+      "grad_norm": 0.1917031854391098,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.0593,
+      "step": 541
+    },
+    {
+      "epoch": 0.6273148148148148,
+      "grad_norm": 0.18925105035305023,
+      "learning_rate": 3.7384259259259264e-06,
+      "loss": 1.056,
+      "step": 542
+    },
+    {
+      "epoch": 0.6284722222222222,
+      "grad_norm": 0.18812283873558044,
+      "learning_rate": 3.726851851851852e-06,
+      "loss": 1.0288,
+      "step": 543
+    },
+    {
+      "epoch": 0.6296296296296297,
+      "grad_norm": 0.2086893916130066,
+      "learning_rate": 3.7152777777777783e-06,
+      "loss": 1.0505,
+      "step": 544
+    },
+    {
+      "epoch": 0.6307870370370371,
+      "grad_norm": 0.2322862148284912,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 1.0552,
+      "step": 545
+    },
+    {
+      "epoch": 0.6319444444444444,
+      "grad_norm": 0.19495348632335663,
+      "learning_rate": 3.69212962962963e-06,
+      "loss": 1.085,
+      "step": 546
+    },
+    {
+      "epoch": 0.6331018518518519,
+      "grad_norm": 0.19767415523529053,
+      "learning_rate": 3.680555555555556e-06,
+      "loss": 1.0229,
+      "step": 547
+    },
+    {
+      "epoch": 0.6342592592592593,
+      "grad_norm": 0.19750790297985077,
+      "learning_rate": 3.668981481481482e-06,
+      "loss": 1.0217,
+      "step": 548
+    },
+    {
+      "epoch": 0.6354166666666666,
+      "grad_norm": 0.20190097391605377,
+      "learning_rate": 3.657407407407408e-06,
+      "loss": 1.0161,
+      "step": 549
+    },
+    {
+      "epoch": 0.6365740740740741,
+      "grad_norm": 0.2032175362110138,
+      "learning_rate": 3.6458333333333333e-06,
+      "loss": 1.0353,
+      "step": 550
+    },
+    {
+      "epoch": 0.6377314814814815,
+      "grad_norm": 0.19069179892539978,
+      "learning_rate": 3.6342592592592596e-06,
+      "loss": 1.0275,
+      "step": 551
+    },
+    {
+      "epoch": 0.6388888888888888,
+      "grad_norm": 0.20992456376552582,
+      "learning_rate": 3.6226851851851855e-06,
+      "loss": 1.0731,
+      "step": 552
+    },
+    {
+      "epoch": 0.6400462962962963,
+      "grad_norm": 0.2105102688074112,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 1.0781,
+      "step": 553
+    },
+    {
+      "epoch": 0.6412037037037037,
+      "grad_norm": 0.19200275838375092,
+      "learning_rate": 3.5995370370370374e-06,
+      "loss": 1.0076,
+      "step": 554
+    },
+    {
+      "epoch": 0.6423611111111112,
+      "grad_norm": 0.2056274116039276,
+      "learning_rate": 3.5879629629629633e-06,
+      "loss": 1.0997,
+      "step": 555
+    },
+    {
+      "epoch": 0.6435185185185185,
+      "grad_norm": 0.3178321421146393,
+      "learning_rate": 3.576388888888889e-06,
+      "loss": 1.0507,
+      "step": 556
+    },
+    {
+      "epoch": 0.6446759259259259,
+      "grad_norm": 0.19499000906944275,
+      "learning_rate": 3.5648148148148147e-06,
+      "loss": 1.1137,
+      "step": 557
+    },
+    {
+      "epoch": 0.6458333333333334,
+      "grad_norm": 0.19996462762355804,
+      "learning_rate": 3.553240740740741e-06,
+      "loss": 1.0567,
+      "step": 558
+    },
+    {
+      "epoch": 0.6469907407407407,
+      "grad_norm": 0.19035173952579498,
+      "learning_rate": 3.5416666666666673e-06,
+      "loss": 1.0325,
+      "step": 559
+    },
+    {
+      "epoch": 0.6481481481481481,
+      "grad_norm": 0.20647963881492615,
+      "learning_rate": 3.530092592592593e-06,
+      "loss": 1.0859,
+      "step": 560
+    },
+    {
+      "epoch": 0.6493055555555556,
+      "grad_norm": 0.20454923808574677,
+      "learning_rate": 3.5185185185185187e-06,
+      "loss": 1.07,
+      "step": 561
+    },
+    {
+      "epoch": 0.6504629629629629,
+      "grad_norm": 0.18973702192306519,
+      "learning_rate": 3.5069444444444447e-06,
+      "loss": 1.0253,
+      "step": 562
+    },
+    {
+      "epoch": 0.6516203703703703,
+      "grad_norm": 0.23206621408462524,
+      "learning_rate": 3.4953703703703706e-06,
+      "loss": 0.9817,
+      "step": 563
+    },
+    {
+      "epoch": 0.6527777777777778,
+      "grad_norm": 0.19497150182724,
+      "learning_rate": 3.483796296296297e-06,
+      "loss": 0.9765,
+      "step": 564
+    },
+    {
+      "epoch": 0.6539351851851852,
+      "grad_norm": 0.20704008638858795,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 1.0923,
+      "step": 565
+    },
+    {
+      "epoch": 0.6550925925925926,
+      "grad_norm": 0.19126930832862854,
+      "learning_rate": 3.4606481481481487e-06,
+      "loss": 1.0344,
+      "step": 566
+    },
+    {
+      "epoch": 0.65625,
+      "grad_norm": 0.3226393163204193,
+      "learning_rate": 3.449074074074074e-06,
+      "loss": 1.0769,
+      "step": 567
+    },
+    {
+      "epoch": 0.6574074074074074,
+      "grad_norm": 0.22098763287067413,
+      "learning_rate": 3.4375e-06,
+      "loss": 1.0469,
+      "step": 568
+    },
+    {
+      "epoch": 0.6585648148148148,
+      "grad_norm": 0.20191290974617004,
+      "learning_rate": 3.4259259259259265e-06,
+      "loss": 1.0734,
+      "step": 569
+    },
+    {
+      "epoch": 0.6597222222222222,
+      "grad_norm": 0.2134922593832016,
+      "learning_rate": 3.414351851851852e-06,
+      "loss": 1.0846,
+      "step": 570
+    },
+    {
+      "epoch": 0.6608796296296297,
+      "grad_norm": 0.23931817710399628,
+      "learning_rate": 3.4027777777777783e-06,
+      "loss": 1.0915,
+      "step": 571
+    },
+    {
+      "epoch": 0.6620370370370371,
+      "grad_norm": 0.3414127230644226,
+      "learning_rate": 3.3912037037037038e-06,
+      "loss": 1.0368,
+      "step": 572
+    },
+    {
+      "epoch": 0.6631944444444444,
+      "grad_norm": 0.2182944118976593,
+      "learning_rate": 3.37962962962963e-06,
+      "loss": 1.0002,
+      "step": 573
+    },
+    {
+      "epoch": 0.6643518518518519,
+      "grad_norm": 0.1967781037092209,
+      "learning_rate": 3.368055555555556e-06,
+      "loss": 1.0334,
+      "step": 574
+    },
+    {
+      "epoch": 0.6655092592592593,
+      "grad_norm": 0.2108740508556366,
+      "learning_rate": 3.3564814814814815e-06,
+      "loss": 1.1083,
+      "step": 575
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.20377075672149658,
+      "learning_rate": 3.344907407407408e-06,
+      "loss": 0.9726,
+      "step": 576
+    },
+    {
+      "epoch": 0.6678240740740741,
+      "grad_norm": 0.19917164742946625,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.0444,
+      "step": 577
+    },
+    {
+      "epoch": 0.6689814814814815,
+      "grad_norm": 0.20163099467754364,
+      "learning_rate": 3.3217592592592597e-06,
+      "loss": 1.0322,
+      "step": 578
+    },
+    {
+      "epoch": 0.6701388888888888,
+      "grad_norm": 0.20033225417137146,
+      "learning_rate": 3.3101851851851856e-06,
+      "loss": 1.0193,
+      "step": 579
+    },
+    {
+      "epoch": 0.6712962962962963,
+      "grad_norm": 0.1936756819486618,
+      "learning_rate": 3.2986111111111115e-06,
+      "loss": 1.0447,
+      "step": 580
+    },
+    {
+      "epoch": 0.6724537037037037,
+      "grad_norm": 0.19549590349197388,
+      "learning_rate": 3.2870370370370374e-06,
+      "loss": 1.0219,
+      "step": 581
+    },
+    {
+      "epoch": 0.6736111111111112,
+      "grad_norm": 0.20389345288276672,
+      "learning_rate": 3.275462962962963e-06,
+      "loss": 1.0208,
+      "step": 582
+    },
+    {
+      "epoch": 0.6747685185185185,
+      "grad_norm": 0.2609894275665283,
+      "learning_rate": 3.2638888888888892e-06,
+      "loss": 1.0721,
+      "step": 583
+    },
+    {
+      "epoch": 0.6759259259259259,
+      "grad_norm": 0.19846400618553162,
+      "learning_rate": 3.2523148148148147e-06,
+      "loss": 1.0241,
+      "step": 584
+    },
+    {
+      "epoch": 0.6770833333333334,
+      "grad_norm": 0.20081739127635956,
+      "learning_rate": 3.240740740740741e-06,
+      "loss": 1.01,
+      "step": 585
+    },
+    {
+      "epoch": 0.6782407407407407,
+      "grad_norm": 0.19961774349212646,
+      "learning_rate": 3.229166666666667e-06,
+      "loss": 1.0194,
+      "step": 586
+    },
+    {
+      "epoch": 0.6793981481481481,
+      "grad_norm": 0.20194299519062042,
+      "learning_rate": 3.217592592592593e-06,
+      "loss": 1.0244,
+      "step": 587
+    },
+    {
+      "epoch": 0.6805555555555556,
+      "grad_norm": 0.2130938321352005,
+      "learning_rate": 3.2060185185185188e-06,
+      "loss": 1.1111,
+      "step": 588
+    },
+    {
+      "epoch": 0.6817129629629629,
+      "grad_norm": 0.21115541458129883,
+      "learning_rate": 3.1944444444444443e-06,
+      "loss": 1.0162,
+      "step": 589
+    },
+    {
+      "epoch": 0.6828703703703703,
+      "grad_norm": 0.22813038527965546,
+      "learning_rate": 3.1828703703703706e-06,
+      "loss": 1.0524,
+      "step": 590
+    },
+    {
+      "epoch": 0.6840277777777778,
+      "grad_norm": 0.20492292940616608,
+      "learning_rate": 3.171296296296297e-06,
+      "loss": 1.0527,
+      "step": 591
+    },
+    {
+      "epoch": 0.6851851851851852,
+      "grad_norm": 0.2057611495256424,
+      "learning_rate": 3.1597222222222224e-06,
+      "loss": 1.0563,
+      "step": 592
+    },
+    {
+      "epoch": 0.6863425925925926,
+      "grad_norm": 0.19977176189422607,
+      "learning_rate": 3.1481481481481483e-06,
+      "loss": 1.0076,
+      "step": 593
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.20001326501369476,
+      "learning_rate": 3.1365740740740742e-06,
+      "loss": 1.0837,
+      "step": 594
+    },
+    {
+      "epoch": 0.6886574074074074,
+      "grad_norm": 0.20452041923999786,
+      "learning_rate": 3.125e-06,
+      "loss": 1.0414,
+      "step": 595
+    },
+    {
+      "epoch": 0.6898148148148148,
+      "grad_norm": 0.19678117334842682,
+      "learning_rate": 3.1134259259259265e-06,
+      "loss": 1.0396,
+      "step": 596
+    },
+    {
+      "epoch": 0.6909722222222222,
+      "grad_norm": 0.2082507163286209,
+      "learning_rate": 3.101851851851852e-06,
+      "loss": 0.9543,
+      "step": 597
+    },
+    {
+      "epoch": 0.6921296296296297,
+      "grad_norm": 0.20555655658245087,
+      "learning_rate": 3.0902777777777783e-06,
+      "loss": 1.0255,
+      "step": 598
+    },
+    {
+      "epoch": 0.6932870370370371,
+      "grad_norm": 0.2342251092195511,
+      "learning_rate": 3.078703703703704e-06,
+      "loss": 1.0321,
+      "step": 599
+    },
+    {
+      "epoch": 0.6944444444444444,
+      "grad_norm": 0.19482319056987762,
+      "learning_rate": 3.0671296296296297e-06,
+      "loss": 1.0691,
+      "step": 600
+    },
+    {
+      "epoch": 0.6956018518518519,
+      "grad_norm": 0.20958203077316284,
+      "learning_rate": 3.055555555555556e-06,
+      "loss": 1.1018,
+      "step": 601
+    },
+    {
+      "epoch": 0.6967592592592593,
+      "grad_norm": 0.20291192829608917,
+      "learning_rate": 3.0439814814814815e-06,
+      "loss": 1.0635,
+      "step": 602
+    },
+    {
+      "epoch": 0.6979166666666666,
+      "grad_norm": 0.2140427827835083,
+      "learning_rate": 3.032407407407408e-06,
+      "loss": 1.0796,
+      "step": 603
+    },
+    {
+      "epoch": 0.6990740740740741,
+      "grad_norm": 0.2033246010541916,
+      "learning_rate": 3.0208333333333334e-06,
+      "loss": 1.0493,
+      "step": 604
+    },
+    {
+      "epoch": 0.7002314814814815,
+      "grad_norm": 0.19889415800571442,
+      "learning_rate": 3.0092592592592597e-06,
+      "loss": 0.9938,
+      "step": 605
+    },
+    {
+      "epoch": 0.7013888888888888,
+      "grad_norm": 0.19974705576896667,
+      "learning_rate": 2.9976851851851856e-06,
+      "loss": 1.1022,
+      "step": 606
+    },
+    {
+      "epoch": 0.7025462962962963,
+      "grad_norm": 0.2727442681789398,
+      "learning_rate": 2.986111111111111e-06,
+      "loss": 1.0657,
+      "step": 607
+    },
+    {
+      "epoch": 0.7037037037037037,
+      "grad_norm": 0.27862659096717834,
+      "learning_rate": 2.9745370370370374e-06,
+      "loss": 0.9777,
+      "step": 608
+    },
+    {
+      "epoch": 0.7048611111111112,
+      "grad_norm": 0.1923118382692337,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 1.0446,
+      "step": 609
+    },
+    {
+      "epoch": 0.7060185185185185,
+      "grad_norm": 0.21940167248249054,
+      "learning_rate": 2.9513888888888892e-06,
+      "loss": 0.9947,
+      "step": 610
+    },
+    {
+      "epoch": 0.7071759259259259,
+      "grad_norm": 0.19208753108978271,
+      "learning_rate": 2.9398148148148147e-06,
+      "loss": 1.0289,
+      "step": 611
+    },
+    {
+      "epoch": 0.7083333333333334,
+      "grad_norm": 0.19762782752513885,
+      "learning_rate": 2.928240740740741e-06,
+      "loss": 0.9995,
+      "step": 612
+    },
+    {
+      "epoch": 0.7094907407407407,
+      "grad_norm": 0.20181262493133545,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.9995,
+      "step": 613
+    },
+    {
+      "epoch": 0.7106481481481481,
+      "grad_norm": 0.20299744606018066,
+      "learning_rate": 2.905092592592593e-06,
+      "loss": 1.0601,
+      "step": 614
+    },
+    {
+      "epoch": 0.7118055555555556,
+      "grad_norm": 0.23061317205429077,
+      "learning_rate": 2.893518518518519e-06,
+      "loss": 1.0337,
+      "step": 615
+    },
+    {
+      "epoch": 0.7129629629629629,
+      "grad_norm": 0.2045440375804901,
+      "learning_rate": 2.8819444444444443e-06,
+      "loss": 0.9872,
+      "step": 616
+    },
+    {
+      "epoch": 0.7141203703703703,
+      "grad_norm": 0.20532366633415222,
+      "learning_rate": 2.8703703703703706e-06,
+      "loss": 1.0355,
+      "step": 617
+    },
+    {
+      "epoch": 0.7152777777777778,
+      "grad_norm": 0.20266734063625336,
+      "learning_rate": 2.8587962962962965e-06,
+      "loss": 1.0675,
+      "step": 618
+    },
+    {
+      "epoch": 0.7164351851851852,
+      "grad_norm": 0.2020139992237091,
+      "learning_rate": 2.8472222222222224e-06,
+      "loss": 1.0446,
+      "step": 619
+    },
+    {
+      "epoch": 0.7175925925925926,
+      "grad_norm": 0.21043045818805695,
+      "learning_rate": 2.8356481481481484e-06,
+      "loss": 1.1025,
+      "step": 620
+    },
+    {
+      "epoch": 0.71875,
+      "grad_norm": 0.20197473466396332,
+      "learning_rate": 2.8240740740740743e-06,
+      "loss": 1.0526,
+      "step": 621
+    },
+    {
+      "epoch": 0.7199074074074074,
+      "grad_norm": 0.21696704626083374,
+      "learning_rate": 2.8125e-06,
+      "loss": 1.0345,
+      "step": 622
+    },
+    {
+      "epoch": 0.7210648148148148,
+      "grad_norm": 0.203839972615242,
+      "learning_rate": 2.8009259259259265e-06,
+      "loss": 1.035,
+      "step": 623
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 0.21520066261291504,
+      "learning_rate": 2.789351851851852e-06,
+      "loss": 1.033,
+      "step": 624
+    },
+    {
+      "epoch": 0.7233796296296297,
+      "grad_norm": 0.212286576628685,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 1.0467,
+      "step": 625
+    },
+    {
+      "epoch": 0.7245370370370371,
+      "grad_norm": 0.20607005059719086,
+      "learning_rate": 2.766203703703704e-06,
+      "loss": 1.0341,
+      "step": 626
+    },
+    {
+      "epoch": 0.7256944444444444,
+      "grad_norm": 0.2012479156255722,
+      "learning_rate": 2.7546296296296297e-06,
+      "loss": 1.0087,
+      "step": 627
+    },
+    {
+      "epoch": 0.7268518518518519,
+      "grad_norm": 0.2070990800857544,
+      "learning_rate": 2.743055555555556e-06,
+      "loss": 1.0397,
+      "step": 628
+    },
+    {
+      "epoch": 0.7280092592592593,
+      "grad_norm": 0.20724061131477356,
+      "learning_rate": 2.7314814814814816e-06,
+      "loss": 1.0123,
+      "step": 629
+    },
+    {
+      "epoch": 0.7291666666666666,
+      "grad_norm": 0.20699013769626617,
+      "learning_rate": 2.719907407407408e-06,
+      "loss": 0.992,
+      "step": 630
+    },
+    {
+      "epoch": 0.7303240740740741,
+      "grad_norm": 0.2006494551897049,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 1.0206,
+      "step": 631
+    },
+    {
+      "epoch": 0.7314814814814815,
+      "grad_norm": 0.2076648622751236,
+      "learning_rate": 2.6967592592592597e-06,
+      "loss": 1.0702,
+      "step": 632
+    },
+    {
+      "epoch": 0.7326388888888888,
+      "grad_norm": 0.1989884227514267,
+      "learning_rate": 2.6851851851851856e-06,
+      "loss": 1.0134,
+      "step": 633
+    },
+    {
+      "epoch": 0.7337962962962963,
+      "grad_norm": 0.20851609110832214,
+      "learning_rate": 2.673611111111111e-06,
+      "loss": 1.0791,
+      "step": 634
+    },
+    {
+      "epoch": 0.7349537037037037,
+      "grad_norm": 0.2005952000617981,
+      "learning_rate": 2.6620370370370374e-06,
+      "loss": 1.0168,
+      "step": 635
+    },
+    {
+      "epoch": 0.7361111111111112,
+      "grad_norm": 0.20755450427532196,
+      "learning_rate": 2.650462962962963e-06,
+      "loss": 1.0862,
+      "step": 636
+    },
+    {
+      "epoch": 0.7372685185185185,
+      "grad_norm": 0.20766755938529968,
+      "learning_rate": 2.6388888888888893e-06,
+      "loss": 1.0771,
+      "step": 637
+    },
+    {
+      "epoch": 0.7384259259259259,
+      "grad_norm": 0.2103908360004425,
+      "learning_rate": 2.6273148148148148e-06,
+      "loss": 1.0256,
+      "step": 638
+    },
+    {
+      "epoch": 0.7395833333333334,
+      "grad_norm": 0.20396418869495392,
+      "learning_rate": 2.615740740740741e-06,
+      "loss": 1.0897,
+      "step": 639
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 0.2057056427001953,
+      "learning_rate": 2.604166666666667e-06,
+      "loss": 1.0358,
+      "step": 640
+    },
+    {
+      "epoch": 0.7418981481481481,
+      "grad_norm": 0.20586839318275452,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 1.0551,
+      "step": 641
+    },
+    {
+      "epoch": 0.7430555555555556,
+      "grad_norm": 0.20847848057746887,
+      "learning_rate": 2.581018518518519e-06,
+      "loss": 1.0597,
+      "step": 642
+    },
+    {
+      "epoch": 0.7442129629629629,
+      "grad_norm": 0.19656863808631897,
+      "learning_rate": 2.5694444444444443e-06,
+      "loss": 0.9972,
+      "step": 643
+    },
+    {
+      "epoch": 0.7453703703703703,
+      "grad_norm": 0.20384305715560913,
+      "learning_rate": 2.5578703703703706e-06,
+      "loss": 1.087,
+      "step": 644
+    },
+    {
+      "epoch": 0.7465277777777778,
+      "grad_norm": 0.21364444494247437,
+      "learning_rate": 2.5462962962962966e-06,
+      "loss": 1.022,
+      "step": 645
+    },
+    {
+      "epoch": 0.7476851851851852,
+      "grad_norm": 0.20833688974380493,
+      "learning_rate": 2.5347222222222225e-06,
+      "loss": 1.0704,
+      "step": 646
+    },
+    {
+      "epoch": 0.7488425925925926,
+      "grad_norm": 0.20712623000144958,
+      "learning_rate": 2.5231481481481484e-06,
+      "loss": 1.075,
+      "step": 647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2294331043958664,
+      "learning_rate": 2.511574074074074e-06,
+      "loss": 1.0206,
+      "step": 648
+    },
+    {
+      "epoch": 0.7511574074074074,
+      "grad_norm": 0.21636289358139038,
+      "learning_rate": 2.5e-06,
+      "loss": 1.1148,
+      "step": 649
+    },
+    {
+      "epoch": 0.7523148148148148,
+      "grad_norm": 0.20249076187610626,
+      "learning_rate": 2.488425925925926e-06,
+      "loss": 1.0555,
+      "step": 650
+    },
+    {
+      "epoch": 0.7534722222222222,
+      "grad_norm": 0.19626903533935547,
+      "learning_rate": 2.476851851851852e-06,
+      "loss": 1.0323,
+      "step": 651
+    },
+    {
+      "epoch": 0.7546296296296297,
+      "grad_norm": 0.2054426223039627,
+      "learning_rate": 2.465277777777778e-06,
+      "loss": 1.0264,
+      "step": 652
+    },
+    {
+      "epoch": 0.7557870370370371,
+      "grad_norm": 0.2102532982826233,
+      "learning_rate": 2.453703703703704e-06,
+      "loss": 1.0552,
+      "step": 653
+    },
+    {
+      "epoch": 0.7569444444444444,
+      "grad_norm": 0.20038294792175293,
+      "learning_rate": 2.4421296296296298e-06,
+      "loss": 1.0043,
+      "step": 654
+    },
+    {
+      "epoch": 0.7581018518518519,
+      "grad_norm": 0.20885545015335083,
+      "learning_rate": 2.4305555555555557e-06,
+      "loss": 1.0417,
+      "step": 655
+    },
+    {
+      "epoch": 0.7592592592592593,
+      "grad_norm": 0.20562267303466797,
+      "learning_rate": 2.4189814814814816e-06,
+      "loss": 0.9974,
+      "step": 656
+    },
+    {
+      "epoch": 0.7604166666666666,
+      "grad_norm": 0.21955883502960205,
+      "learning_rate": 2.4074074074074075e-06,
+      "loss": 1.0758,
+      "step": 657
+    },
+    {
+      "epoch": 0.7615740740740741,
+      "grad_norm": 0.29182884097099304,
+      "learning_rate": 2.395833333333334e-06,
+      "loss": 1.059,
+      "step": 658
+    },
+    {
+      "epoch": 0.7627314814814815,
+      "grad_norm": 0.20799702405929565,
+      "learning_rate": 2.3842592592592593e-06,
+      "loss": 0.9904,
+      "step": 659
+    },
+    {
+      "epoch": 0.7638888888888888,
+      "grad_norm": 0.19996783137321472,
+      "learning_rate": 2.3726851851851852e-06,
+      "loss": 1.0222,
+      "step": 660
+    },
+    {
+      "epoch": 0.7650462962962963,
+      "grad_norm": 0.20935088396072388,
+      "learning_rate": 2.361111111111111e-06,
+      "loss": 1.0626,
+      "step": 661
+    },
+    {
+      "epoch": 0.7662037037037037,
+      "grad_norm": 0.2047765702009201,
+      "learning_rate": 2.349537037037037e-06,
+      "loss": 1.0108,
+      "step": 662
+    },
+    {
+      "epoch": 0.7673611111111112,
+      "grad_norm": 0.22284026443958282,
+      "learning_rate": 2.3379629629629634e-06,
+      "loss": 1.0378,
+      "step": 663
+    },
+    {
+      "epoch": 0.7685185185185185,
+      "grad_norm": 0.2159646898508072,
+      "learning_rate": 2.3263888888888893e-06,
+      "loss": 1.0571,
+      "step": 664
+    },
+    {
+      "epoch": 0.7696759259259259,
+      "grad_norm": 0.21102292835712433,
+      "learning_rate": 2.314814814814815e-06,
+      "loss": 1.0601,
+      "step": 665
+    },
+    {
+      "epoch": 0.7708333333333334,
+      "grad_norm": 0.24397940933704376,
+      "learning_rate": 2.3032407407407407e-06,
+      "loss": 1.0433,
+      "step": 666
+    },
+    {
+      "epoch": 0.7719907407407407,
+      "grad_norm": 0.2021220624446869,
+      "learning_rate": 2.2916666666666666e-06,
+      "loss": 1.0553,
+      "step": 667
+    },
+    {
+      "epoch": 0.7731481481481481,
+      "grad_norm": 0.20541134476661682,
+      "learning_rate": 2.280092592592593e-06,
+      "loss": 1.0218,
+      "step": 668
+    },
+    {
+      "epoch": 0.7743055555555556,
+      "grad_norm": 0.2077786922454834,
+      "learning_rate": 2.268518518518519e-06,
+      "loss": 1.0973,
+      "step": 669
+    },
+    {
+      "epoch": 0.7754629629629629,
+      "grad_norm": 0.21917060017585754,
+      "learning_rate": 2.2569444444444448e-06,
+      "loss": 1.0308,
+      "step": 670
+    },
+    {
+      "epoch": 0.7766203703703703,
+      "grad_norm": 0.2081642746925354,
+      "learning_rate": 2.2453703703703707e-06,
+      "loss": 1.0731,
+      "step": 671
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 0.20313401520252228,
+      "learning_rate": 2.2337962962962966e-06,
+      "loss": 1.1032,
+      "step": 672
+    },
+    {
+      "epoch": 0.7789351851851852,
+      "grad_norm": 0.2097117006778717,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.0215,
+      "step": 673
+    },
+    {
+      "epoch": 0.7800925925925926,
+      "grad_norm": 0.20481272041797638,
+      "learning_rate": 2.2106481481481484e-06,
+      "loss": 1.0594,
+      "step": 674
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 0.21563556790351868,
+      "learning_rate": 2.1990740740740743e-06,
+      "loss": 0.996,
+      "step": 675
+    },
+    {
+      "epoch": 0.7824074074074074,
+      "grad_norm": 0.21871314942836761,
+      "learning_rate": 2.1875000000000002e-06,
+      "loss": 1.0673,
+      "step": 676
+    },
+    {
+      "epoch": 0.7835648148148148,
+      "grad_norm": 0.21116910874843597,
+      "learning_rate": 2.175925925925926e-06,
+      "loss": 1.0941,
+      "step": 677
+    },
+    {
+      "epoch": 0.7847222222222222,
+      "grad_norm": 0.20714472234249115,
+      "learning_rate": 2.164351851851852e-06,
+      "loss": 1.0641,
+      "step": 678
+    },
+    {
+      "epoch": 0.7858796296296297,
+      "grad_norm": 0.217753604054451,
+      "learning_rate": 2.152777777777778e-06,
+      "loss": 1.0157,
+      "step": 679
+    },
+    {
+      "epoch": 0.7870370370370371,
+      "grad_norm": 0.20730328559875488,
+      "learning_rate": 2.141203703703704e-06,
+      "loss": 1.0018,
+      "step": 680
+    },
+    {
+      "epoch": 0.7881944444444444,
+      "grad_norm": 0.21109417080879211,
+      "learning_rate": 2.1296296296296298e-06,
+      "loss": 0.9815,
+      "step": 681
+    },
+    {
+      "epoch": 0.7893518518518519,
+      "grad_norm": 0.221344456076622,
+      "learning_rate": 2.1180555555555557e-06,
+      "loss": 1.0124,
+      "step": 682
+    },
+    {
+      "epoch": 0.7905092592592593,
+      "grad_norm": 0.22139620780944824,
+      "learning_rate": 2.1064814814814816e-06,
+      "loss": 1.011,
+      "step": 683
+    },
+    {
+      "epoch": 0.7916666666666666,
+      "grad_norm": 0.21301385760307312,
+      "learning_rate": 2.0949074074074075e-06,
+      "loss": 1.0383,
+      "step": 684
+    },
+    {
+      "epoch": 0.7928240740740741,
+      "grad_norm": 0.20398671925067902,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 1.0404,
+      "step": 685
+    },
+    {
+      "epoch": 0.7939814814814815,
+      "grad_norm": 0.2034938633441925,
+      "learning_rate": 2.0717592592592593e-06,
+      "loss": 1.0328,
+      "step": 686
+    },
+    {
+      "epoch": 0.7951388888888888,
+      "grad_norm": 0.22071969509124756,
+      "learning_rate": 2.0601851851851853e-06,
+      "loss": 0.9852,
+      "step": 687
+    },
+    {
+      "epoch": 0.7962962962962963,
+      "grad_norm": 0.20521746575832367,
+      "learning_rate": 2.048611111111111e-06,
+      "loss": 0.9691,
+      "step": 688
+    },
+    {
+      "epoch": 0.7974537037037037,
+      "grad_norm": 0.20581713318824768,
+      "learning_rate": 2.037037037037037e-06,
+      "loss": 1.0328,
+      "step": 689
+    },
+    {
+      "epoch": 0.7986111111111112,
+      "grad_norm": 0.20280128717422485,
+      "learning_rate": 2.0254629629629634e-06,
+      "loss": 0.9418,
+      "step": 690
+    },
+    {
+      "epoch": 0.7997685185185185,
+      "grad_norm": 0.20858825743198395,
+      "learning_rate": 2.0138888888888893e-06,
+      "loss": 1.04,
+      "step": 691
+    },
+    {
+      "epoch": 0.8009259259259259,
+      "grad_norm": 0.21282228827476501,
+      "learning_rate": 2.002314814814815e-06,
+      "loss": 1.0959,
+      "step": 692
+    },
+    {
+      "epoch": 0.8020833333333334,
+      "grad_norm": 0.21245089173316956,
+      "learning_rate": 1.9907407407407407e-06,
+      "loss": 1.0266,
+      "step": 693
+    },
+    {
+      "epoch": 0.8032407407407407,
+      "grad_norm": 0.21784378588199615,
+      "learning_rate": 1.9791666666666666e-06,
+      "loss": 1.0592,
+      "step": 694
+    },
+    {
+      "epoch": 0.8043981481481481,
+      "grad_norm": 0.22031240165233612,
+      "learning_rate": 1.967592592592593e-06,
+      "loss": 1.0245,
+      "step": 695
+    },
+    {
+      "epoch": 0.8055555555555556,
+      "grad_norm": 0.20002427697181702,
+      "learning_rate": 1.956018518518519e-06,
+      "loss": 1.0165,
+      "step": 696
+    },
+    {
+      "epoch": 0.8067129629629629,
+      "grad_norm": 0.20793163776397705,
+      "learning_rate": 1.944444444444445e-06,
+      "loss": 1.0229,
+      "step": 697
+    },
+    {
+      "epoch": 0.8078703703703703,
+      "grad_norm": 0.20959703624248505,
+      "learning_rate": 1.9328703703703707e-06,
+      "loss": 1.0051,
+      "step": 698
+    },
+    {
+      "epoch": 0.8090277777777778,
+      "grad_norm": 0.20717735588550568,
+      "learning_rate": 1.921296296296296e-06,
+      "loss": 0.961,
+      "step": 699
+    },
+    {
+      "epoch": 0.8101851851851852,
+      "grad_norm": 0.20636354386806488,
+      "learning_rate": 1.909722222222222e-06,
+      "loss": 1.0221,
+      "step": 700
+    },
+    {
+      "epoch": 0.8113425925925926,
+      "grad_norm": 0.2062767595052719,
+      "learning_rate": 1.8981481481481484e-06,
+      "loss": 1.0733,
+      "step": 701
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.21050766110420227,
+      "learning_rate": 1.8865740740740743e-06,
+      "loss": 0.9686,
+      "step": 702
+    },
+    {
+      "epoch": 0.8136574074074074,
+      "grad_norm": 0.21685615181922913,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.9892,
+      "step": 703
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 0.24662281572818756,
+      "learning_rate": 1.863425925925926e-06,
+      "loss": 1.0709,
+      "step": 704
+    },
+    {
+      "epoch": 0.8159722222222222,
+      "grad_norm": 0.20373214781284332,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 1.025,
+      "step": 705
+    },
+    {
+      "epoch": 0.8171296296296297,
+      "grad_norm": 0.212567999958992,
+      "learning_rate": 1.840277777777778e-06,
+      "loss": 1.003,
+      "step": 706
+    },
+    {
+      "epoch": 0.8182870370370371,
+      "grad_norm": 0.20584706962108612,
+      "learning_rate": 1.828703703703704e-06,
+      "loss": 1.0821,
+      "step": 707
+    },
+    {
+      "epoch": 0.8194444444444444,
+      "grad_norm": 0.21041174232959747,
+      "learning_rate": 1.8171296296296298e-06,
+      "loss": 0.9875,
+      "step": 708
+    },
+    {
+      "epoch": 0.8206018518518519,
+      "grad_norm": 0.2137048840522766,
+      "learning_rate": 1.8055555555555557e-06,
+      "loss": 1.0338,
+      "step": 709
+    },
+    {
+      "epoch": 0.8217592592592593,
+      "grad_norm": 0.2729160785675049,
+      "learning_rate": 1.7939814814814816e-06,
+      "loss": 1.0076,
+      "step": 710
+    },
+    {
+      "epoch": 0.8229166666666666,
+      "grad_norm": 0.2052346020936966,
+      "learning_rate": 1.7824074074074073e-06,
+      "loss": 1.0193,
+      "step": 711
+    },
+    {
+      "epoch": 0.8240740740740741,
+      "grad_norm": 0.22163906693458557,
+      "learning_rate": 1.7708333333333337e-06,
+      "loss": 1.0455,
+      "step": 712
+    },
+    {
+      "epoch": 0.8252314814814815,
+      "grad_norm": 0.21060538291931152,
+      "learning_rate": 1.7592592592592594e-06,
+      "loss": 1.051,
+      "step": 713
+    },
+    {
+      "epoch": 0.8263888888888888,
+      "grad_norm": 0.21227489411830902,
+      "learning_rate": 1.7476851851851853e-06,
+      "loss": 1.0054,
+      "step": 714
+    },
+    {
+      "epoch": 0.8275462962962963,
+      "grad_norm": 0.2057754248380661,
+      "learning_rate": 1.7361111111111112e-06,
+      "loss": 1.0488,
+      "step": 715
+    },
+    {
+      "epoch": 0.8287037037037037,
+      "grad_norm": 0.21524278819561005,
+      "learning_rate": 1.724537037037037e-06,
+      "loss": 1.0146,
+      "step": 716
+    },
+    {
+      "epoch": 0.8298611111111112,
+      "grad_norm": 0.2293296605348587,
+      "learning_rate": 1.7129629629629632e-06,
+      "loss": 1.0471,
+      "step": 717
+    },
+    {
+      "epoch": 0.8310185185185185,
+      "grad_norm": 0.20466309785842896,
+      "learning_rate": 1.7013888888888891e-06,
+      "loss": 0.9884,
+      "step": 718
+    },
+    {
+      "epoch": 0.8321759259259259,
+      "grad_norm": 0.215054452419281,
+      "learning_rate": 1.689814814814815e-06,
+      "loss": 1.1239,
+      "step": 719
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.21369078755378723,
+      "learning_rate": 1.6782407407407408e-06,
+      "loss": 1.0047,
+      "step": 720
+    },
+    {
+      "epoch": 0.8344907407407407,
+      "grad_norm": 0.23838470876216888,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.9937,
+      "step": 721
+    },
+    {
+      "epoch": 0.8356481481481481,
+      "grad_norm": 0.20185384154319763,
+      "learning_rate": 1.6550925925925928e-06,
+      "loss": 0.9759,
+      "step": 722
+    },
+    {
+      "epoch": 0.8368055555555556,
+      "grad_norm": 0.20911818742752075,
+      "learning_rate": 1.6435185185185187e-06,
+      "loss": 1.0018,
+      "step": 723
+    },
+    {
+      "epoch": 0.8379629629629629,
+      "grad_norm": 0.20076234638690948,
+      "learning_rate": 1.6319444444444446e-06,
+      "loss": 0.9895,
+      "step": 724
+    },
+    {
+      "epoch": 0.8391203703703703,
+      "grad_norm": 0.21198947727680206,
+      "learning_rate": 1.6203703703703705e-06,
+      "loss": 0.9969,
+      "step": 725
+    },
+    {
+      "epoch": 0.8402777777777778,
+      "grad_norm": 0.20863303542137146,
+      "learning_rate": 1.6087962962962964e-06,
+      "loss": 1.0044,
+      "step": 726
+    },
+    {
+      "epoch": 0.8414351851851852,
+      "grad_norm": 0.20545004308223724,
+      "learning_rate": 1.5972222222222221e-06,
+      "loss": 1.0147,
+      "step": 727
+    },
+    {
+      "epoch": 0.8425925925925926,
+      "grad_norm": 0.2052348256111145,
+      "learning_rate": 1.5856481481481485e-06,
+      "loss": 0.9703,
+      "step": 728
+    },
+    {
+      "epoch": 0.84375,
+      "grad_norm": 0.2435932457447052,
+      "learning_rate": 1.5740740740740742e-06,
+      "loss": 1.0181,
+      "step": 729
+    },
+    {
+      "epoch": 0.8449074074074074,
+      "grad_norm": 0.22731885313987732,
+      "learning_rate": 1.5625e-06,
+      "loss": 1.1076,
+      "step": 730
+    },
+    {
+      "epoch": 0.8460648148148148,
+      "grad_norm": 0.208094984292984,
+      "learning_rate": 1.550925925925926e-06,
+      "loss": 1.0641,
+      "step": 731
+    },
+    {
+      "epoch": 0.8472222222222222,
+      "grad_norm": 0.20480291545391083,
+      "learning_rate": 1.539351851851852e-06,
+      "loss": 1.0031,
+      "step": 732
+    },
+    {
+      "epoch": 0.8483796296296297,
+      "grad_norm": 0.21013250946998596,
+      "learning_rate": 1.527777777777778e-06,
+      "loss": 1.0306,
+      "step": 733
+    },
+    {
+      "epoch": 0.8495370370370371,
+      "grad_norm": 0.211978942155838,
+      "learning_rate": 1.516203703703704e-06,
+      "loss": 1.0357,
+      "step": 734
+    },
+    {
+      "epoch": 0.8506944444444444,
+      "grad_norm": 0.22962923347949982,
+      "learning_rate": 1.5046296296296298e-06,
+      "loss": 1.0954,
+      "step": 735
+    },
+    {
+      "epoch": 0.8518518518518519,
+      "grad_norm": 0.2159680426120758,
+      "learning_rate": 1.4930555555555555e-06,
+      "loss": 1.0309,
+      "step": 736
+    },
+    {
+      "epoch": 0.8530092592592593,
+      "grad_norm": 0.20775794982910156,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.9869,
+      "step": 737
+    },
+    {
+      "epoch": 0.8541666666666666,
+      "grad_norm": 0.22548414766788483,
+      "learning_rate": 1.4699074074074074e-06,
+      "loss": 1.1086,
+      "step": 738
+    },
+    {
+      "epoch": 0.8553240740740741,
+      "grad_norm": 0.21975237131118774,
+      "learning_rate": 1.4583333333333335e-06,
+      "loss": 1.053,
+      "step": 739
+    },
+    {
+      "epoch": 0.8564814814814815,
+      "grad_norm": 0.20786741375923157,
+      "learning_rate": 1.4467592592592594e-06,
+      "loss": 1.0756,
+      "step": 740
+    },
+    {
+      "epoch": 0.8576388888888888,
+      "grad_norm": 0.20583467185497284,
+      "learning_rate": 1.4351851851851853e-06,
+      "loss": 1.0741,
+      "step": 741
+    },
+    {
+      "epoch": 0.8587962962962963,
+      "grad_norm": 0.2188231199979782,
+      "learning_rate": 1.4236111111111112e-06,
+      "loss": 0.9667,
+      "step": 742
+    },
+    {
+      "epoch": 0.8599537037037037,
+      "grad_norm": 0.20682111382484436,
+      "learning_rate": 1.4120370370370371e-06,
+      "loss": 1.0409,
+      "step": 743
+    },
+    {
+      "epoch": 0.8611111111111112,
+      "grad_norm": 0.21626965701580048,
+      "learning_rate": 1.4004629629629633e-06,
+      "loss": 1.0366,
+      "step": 744
+    },
+    {
+      "epoch": 0.8622685185185185,
+      "grad_norm": 0.21385551989078522,
+      "learning_rate": 1.3888888888888892e-06,
+      "loss": 1.091,
+      "step": 745
+    },
+    {
+      "epoch": 0.8634259259259259,
+      "grad_norm": 0.2148781716823578,
+      "learning_rate": 1.3773148148148149e-06,
+      "loss": 1.0033,
+      "step": 746
+    },
+    {
+      "epoch": 0.8645833333333334,
+      "grad_norm": 0.21703305840492249,
+      "learning_rate": 1.3657407407407408e-06,
+      "loss": 1.0228,
+      "step": 747
+    },
+    {
+      "epoch": 0.8657407407407407,
+      "grad_norm": 0.20812486112117767,
+      "learning_rate": 1.3541666666666667e-06,
+      "loss": 1.0227,
+      "step": 748
+    },
+    {
+      "epoch": 0.8668981481481481,
+      "grad_norm": 0.20546585321426392,
+      "learning_rate": 1.3425925925925928e-06,
+      "loss": 1.0391,
+      "step": 749
+    },
+    {
+      "epoch": 0.8680555555555556,
+      "grad_norm": 0.26289820671081543,
+      "learning_rate": 1.3310185185185187e-06,
+      "loss": 1.0451,
+      "step": 750
+    },
+    {
+      "epoch": 0.8692129629629629,
+      "grad_norm": 0.262515127658844,
+      "learning_rate": 1.3194444444444446e-06,
+      "loss": 1.0597,
+      "step": 751
+    },
+    {
+      "epoch": 0.8703703703703703,
+      "grad_norm": 0.21458126604557037,
+      "learning_rate": 1.3078703703703705e-06,
+      "loss": 1.0574,
+      "step": 752
+    },
+    {
+      "epoch": 0.8715277777777778,
+      "grad_norm": 0.20759519934654236,
+      "learning_rate": 1.2962962962962962e-06,
+      "loss": 1.0697,
+      "step": 753
+    },
+    {
+      "epoch": 0.8726851851851852,
+      "grad_norm": 0.21060067415237427,
+      "learning_rate": 1.2847222222222222e-06,
+      "loss": 0.956,
+      "step": 754
+    },
+    {
+      "epoch": 0.8738425925925926,
+      "grad_norm": 0.2437879592180252,
+      "learning_rate": 1.2731481481481483e-06,
+      "loss": 0.9785,
+      "step": 755
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.22363758087158203,
+      "learning_rate": 1.2615740740740742e-06,
+      "loss": 1.0751,
+      "step": 756
+    },
+    {
+      "epoch": 0.8761574074074074,
+      "grad_norm": 0.2119099348783493,
+      "learning_rate": 1.25e-06,
+      "loss": 1.0775,
+      "step": 757
+    },
+    {
+      "epoch": 0.8773148148148148,
+      "grad_norm": 0.2066664844751358,
+      "learning_rate": 1.238425925925926e-06,
+      "loss": 0.9457,
+      "step": 758
+    },
+    {
+      "epoch": 0.8784722222222222,
+      "grad_norm": 0.21327026188373566,
+      "learning_rate": 1.226851851851852e-06,
+      "loss": 1.0065,
+      "step": 759
+    },
+    {
+      "epoch": 0.8796296296296297,
+      "grad_norm": 0.22070683538913727,
+      "learning_rate": 1.2152777777777778e-06,
+      "loss": 1.0822,
+      "step": 760
+    },
+    {
+      "epoch": 0.8807870370370371,
+      "grad_norm": 0.5572031140327454,
+      "learning_rate": 1.2037037037037037e-06,
+      "loss": 1.0996,
+      "step": 761
+    },
+    {
+      "epoch": 0.8819444444444444,
+      "grad_norm": 0.20942085981369019,
+      "learning_rate": 1.1921296296296297e-06,
+      "loss": 1.0156,
+      "step": 762
+    },
+    {
+      "epoch": 0.8831018518518519,
+      "grad_norm": 0.20444652438163757,
+      "learning_rate": 1.1805555555555556e-06,
+      "loss": 1.0382,
+      "step": 763
+    },
+    {
+      "epoch": 0.8842592592592593,
+      "grad_norm": 0.2016180455684662,
+      "learning_rate": 1.1689814814814817e-06,
+      "loss": 1.0174,
+      "step": 764
+    },
+    {
+      "epoch": 0.8854166666666666,
+      "grad_norm": 0.20109869539737701,
+      "learning_rate": 1.1574074074074076e-06,
+      "loss": 1.0037,
+      "step": 765
+    },
+    {
+      "epoch": 0.8865740740740741,
+      "grad_norm": 0.23390746116638184,
+      "learning_rate": 1.1458333333333333e-06,
+      "loss": 1.0423,
+      "step": 766
+    },
+    {
+      "epoch": 0.8877314814814815,
+      "grad_norm": 0.21662883460521698,
+      "learning_rate": 1.1342592592592594e-06,
+      "loss": 1.03,
+      "step": 767
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.21152052283287048,
+      "learning_rate": 1.1226851851851853e-06,
+      "loss": 1.0689,
+      "step": 768
+    },
+    {
+      "epoch": 0.8900462962962963,
+      "grad_norm": 0.2138381153345108,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.9821,
+      "step": 769
+    },
+    {
+      "epoch": 0.8912037037037037,
+      "grad_norm": 0.2141958475112915,
+      "learning_rate": 1.0995370370370372e-06,
+      "loss": 0.9939,
+      "step": 770
+    },
+    {
+      "epoch": 0.8923611111111112,
+      "grad_norm": 0.21247626841068268,
+      "learning_rate": 1.087962962962963e-06,
+      "loss": 1.0159,
+      "step": 771
+    },
+    {
+      "epoch": 0.8935185185185185,
+      "grad_norm": 0.20598827302455902,
+      "learning_rate": 1.076388888888889e-06,
+      "loss": 1.0364,
+      "step": 772
+    },
+    {
+      "epoch": 0.8946759259259259,
+      "grad_norm": 0.2089841365814209,
+      "learning_rate": 1.0648148148148149e-06,
+      "loss": 1.0458,
+      "step": 773
+    },
+    {
+      "epoch": 0.8958333333333334,
+      "grad_norm": 0.20895791053771973,
+      "learning_rate": 1.0532407407407408e-06,
+      "loss": 1.0413,
+      "step": 774
+    },
+    {
+      "epoch": 0.8969907407407407,
+      "grad_norm": 0.22381016612052917,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 0.9736,
+      "step": 775
+    },
+    {
+      "epoch": 0.8981481481481481,
+      "grad_norm": 0.20567692816257477,
+      "learning_rate": 1.0300925925925926e-06,
+      "loss": 0.954,
+      "step": 776
+    },
+    {
+      "epoch": 0.8993055555555556,
+      "grad_norm": 0.2209719866514206,
+      "learning_rate": 1.0185185185185185e-06,
+      "loss": 1.0772,
+      "step": 777
+    },
+    {
+      "epoch": 0.9004629629629629,
+      "grad_norm": 0.2152528315782547,
+      "learning_rate": 1.0069444444444447e-06,
+      "loss": 1.0277,
+      "step": 778
+    },
+    {
+      "epoch": 0.9016203703703703,
+      "grad_norm": 0.2000735104084015,
+      "learning_rate": 9.953703703703704e-07,
+      "loss": 1.0367,
+      "step": 779
+    },
+    {
+      "epoch": 0.9027777777777778,
+      "grad_norm": 0.2128438502550125,
+      "learning_rate": 9.837962962962965e-07,
+      "loss": 0.9917,
+      "step": 780
+    },
+    {
+      "epoch": 0.9039351851851852,
+      "grad_norm": 0.21334069967269897,
+      "learning_rate": 9.722222222222224e-07,
+      "loss": 1.0571,
+      "step": 781
+    },
+    {
+      "epoch": 0.9050925925925926,
+      "grad_norm": 0.20610938966274261,
+      "learning_rate": 9.60648148148148e-07,
+      "loss": 1.0239,
+      "step": 782
+    },
+    {
+      "epoch": 0.90625,
+      "grad_norm": 0.21400775015354156,
+      "learning_rate": 9.490740740740742e-07,
+      "loss": 1.057,
+      "step": 783
+    },
+    {
+      "epoch": 0.9074074074074074,
+      "grad_norm": 0.22360017895698547,
+      "learning_rate": 9.375000000000001e-07,
+      "loss": 1.0245,
+      "step": 784
+    },
+    {
+      "epoch": 0.9085648148148148,
+      "grad_norm": 0.22137287259101868,
+      "learning_rate": 9.259259259259259e-07,
+      "loss": 0.9877,
+      "step": 785
+    },
+    {
+      "epoch": 0.9097222222222222,
+      "grad_norm": 0.2224033772945404,
+      "learning_rate": 9.14351851851852e-07,
+      "loss": 1.1311,
+      "step": 786
+    },
+    {
+      "epoch": 0.9108796296296297,
+      "grad_norm": 0.20913951098918915,
+      "learning_rate": 9.027777777777779e-07,
+      "loss": 0.9691,
+      "step": 787
+    },
+    {
+      "epoch": 0.9120370370370371,
+      "grad_norm": 0.21408195793628693,
+      "learning_rate": 8.912037037037037e-07,
+      "loss": 1.0495,
+      "step": 788
+    },
+    {
+      "epoch": 0.9131944444444444,
+      "grad_norm": 0.21985873579978943,
+      "learning_rate": 8.796296296296297e-07,
+      "loss": 1.0578,
+      "step": 789
+    },
+    {
+      "epoch": 0.9143518518518519,
+      "grad_norm": 0.21710622310638428,
+      "learning_rate": 8.680555555555556e-07,
+      "loss": 1.0115,
+      "step": 790
+    },
+    {
+      "epoch": 0.9155092592592593,
+      "grad_norm": 0.2131613790988922,
+      "learning_rate": 8.564814814814816e-07,
+      "loss": 0.9411,
+      "step": 791
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 0.2729133665561676,
+      "learning_rate": 8.449074074074075e-07,
+      "loss": 1.0428,
+      "step": 792
+    },
+    {
+      "epoch": 0.9178240740740741,
+      "grad_norm": 0.21577362716197968,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 1.0285,
+      "step": 793
+    },
+    {
+      "epoch": 0.9189814814814815,
+      "grad_norm": 0.21308249235153198,
+      "learning_rate": 8.217592592592593e-07,
+      "loss": 0.9985,
+      "step": 794
+    },
+    {
+      "epoch": 0.9201388888888888,
+      "grad_norm": 0.21895642578601837,
+      "learning_rate": 8.101851851851853e-07,
+      "loss": 1.1138,
+      "step": 795
+    },
+    {
+      "epoch": 0.9212962962962963,
+      "grad_norm": 0.2174258530139923,
+      "learning_rate": 7.986111111111111e-07,
+      "loss": 0.9322,
+      "step": 796
+    },
+    {
+      "epoch": 0.9224537037037037,
+      "grad_norm": 0.20960073173046112,
+      "learning_rate": 7.870370370370371e-07,
+      "loss": 1.0716,
+      "step": 797
+    },
+    {
+      "epoch": 0.9236111111111112,
+      "grad_norm": 0.2102329581975937,
+      "learning_rate": 7.75462962962963e-07,
+      "loss": 1.0035,
+      "step": 798
+    },
+    {
+      "epoch": 0.9247685185185185,
+      "grad_norm": 0.22140781581401825,
+      "learning_rate": 7.63888888888889e-07,
+      "loss": 1.0538,
+      "step": 799
+    },
+    {
+      "epoch": 0.9259259259259259,
+      "grad_norm": 0.2097562849521637,
+      "learning_rate": 7.523148148148149e-07,
+      "loss": 1.0062,
+      "step": 800
+    },
+    {
+      "epoch": 0.9270833333333334,
+      "grad_norm": 0.21233950555324554,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.9928,
+      "step": 801
+    },
+    {
+      "epoch": 0.9282407407407407,
+      "grad_norm": 0.223008394241333,
+      "learning_rate": 7.291666666666667e-07,
+      "loss": 1.0258,
+      "step": 802
+    },
+    {
+      "epoch": 0.9293981481481481,
+      "grad_norm": 0.21271130442619324,
+      "learning_rate": 7.175925925925927e-07,
+      "loss": 1.0305,
+      "step": 803
+    },
+    {
+      "epoch": 0.9305555555555556,
+      "grad_norm": 0.2107594758272171,
+      "learning_rate": 7.060185185185186e-07,
+      "loss": 0.9713,
+      "step": 804
+    },
+    {
+      "epoch": 0.9317129629629629,
+      "grad_norm": 0.21259775757789612,
+      "learning_rate": 6.944444444444446e-07,
+      "loss": 1.0238,
+      "step": 805
+    },
+    {
+      "epoch": 0.9328703703703703,
+      "grad_norm": 0.22627367079257965,
+      "learning_rate": 6.828703703703704e-07,
+      "loss": 1.0147,
+      "step": 806
+    },
+    {
+      "epoch": 0.9340277777777778,
+      "grad_norm": 0.22005265951156616,
+      "learning_rate": 6.712962962962964e-07,
+      "loss": 0.9688,
+      "step": 807
+    },
+    {
+      "epoch": 0.9351851851851852,
+      "grad_norm": 0.198886439204216,
+      "learning_rate": 6.597222222222223e-07,
+      "loss": 1.0142,
+      "step": 808
+    },
+    {
+      "epoch": 0.9363425925925926,
+      "grad_norm": 0.21684283018112183,
+      "learning_rate": 6.481481481481481e-07,
+      "loss": 0.9538,
+      "step": 809
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.2184874415397644,
+      "learning_rate": 6.365740740740741e-07,
+      "loss": 1.0073,
+      "step": 810
+    },
+    {
+      "epoch": 0.9386574074074074,
+      "grad_norm": 0.2754838764667511,
+      "learning_rate": 6.25e-07,
+      "loss": 0.9981,
+      "step": 811
+    },
+    {
+      "epoch": 0.9398148148148148,
+      "grad_norm": 0.21016815304756165,
+      "learning_rate": 6.13425925925926e-07,
+      "loss": 1.0259,
+      "step": 812
+    },
+    {
+      "epoch": 0.9409722222222222,
+      "grad_norm": 0.209848091006279,
+      "learning_rate": 6.018518518518519e-07,
+      "loss": 1.0265,
+      "step": 813
+    },
+    {
+      "epoch": 0.9421296296296297,
+      "grad_norm": 0.2134370058774948,
+      "learning_rate": 5.902777777777778e-07,
+      "loss": 1.0218,
+      "step": 814
+    },
+    {
+      "epoch": 0.9432870370370371,
+      "grad_norm": 0.21250751614570618,
+      "learning_rate": 5.787037037037038e-07,
+      "loss": 1.0203,
+      "step": 815
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 0.21034426987171173,
+      "learning_rate": 5.671296296296297e-07,
+      "loss": 1.0716,
+      "step": 816
+    },
+    {
+      "epoch": 0.9456018518518519,
+      "grad_norm": 0.20719684660434723,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.9431,
+      "step": 817
+    },
+    {
+      "epoch": 0.9467592592592593,
+      "grad_norm": 0.20388469099998474,
+      "learning_rate": 5.439814814814815e-07,
+      "loss": 1.0351,
+      "step": 818
+    },
+    {
+      "epoch": 0.9479166666666666,
+      "grad_norm": 0.20847779512405396,
+      "learning_rate": 5.324074074074074e-07,
+      "loss": 1.1189,
+      "step": 819
+    },
+    {
+      "epoch": 0.9490740740740741,
+      "grad_norm": 0.22145158052444458,
+      "learning_rate": 5.208333333333334e-07,
+      "loss": 1.067,
+      "step": 820
+    },
+    {
+      "epoch": 0.9502314814814815,
+      "grad_norm": 0.21092523634433746,
+      "learning_rate": 5.092592592592593e-07,
+      "loss": 1.0551,
+      "step": 821
+    },
+    {
+      "epoch": 0.9513888888888888,
+      "grad_norm": 0.20877982676029205,
+      "learning_rate": 4.976851851851852e-07,
+      "loss": 1.052,
+      "step": 822
+    },
+    {
+      "epoch": 0.9525462962962963,
+      "grad_norm": 0.21412430703639984,
+      "learning_rate": 4.861111111111112e-07,
+      "loss": 1.0205,
+      "step": 823
+    },
+    {
+      "epoch": 0.9537037037037037,
+      "grad_norm": 0.21050423383712769,
+      "learning_rate": 4.745370370370371e-07,
+      "loss": 0.9803,
+      "step": 824
+    },
+    {
+      "epoch": 0.9548611111111112,
+      "grad_norm": 0.21291252970695496,
+      "learning_rate": 4.6296296296296297e-07,
+      "loss": 1.0225,
+      "step": 825
+    },
+    {
+      "epoch": 0.9560185185185185,
+      "grad_norm": 0.20651710033416748,
+      "learning_rate": 4.5138888888888893e-07,
+      "loss": 1.0051,
+      "step": 826
+    },
+    {
+      "epoch": 0.9571759259259259,
+      "grad_norm": 0.31924715638160706,
+      "learning_rate": 4.3981481481481484e-07,
+      "loss": 0.9834,
+      "step": 827
+    },
+    {
+      "epoch": 0.9583333333333334,
+      "grad_norm": 0.21879231929779053,
+      "learning_rate": 4.282407407407408e-07,
+      "loss": 1.0335,
+      "step": 828
+    },
+    {
+      "epoch": 0.9594907407407407,
+      "grad_norm": 0.20607353746891022,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 1.0727,
+      "step": 829
+    },
+    {
+      "epoch": 0.9606481481481481,
+      "grad_norm": 0.2167244255542755,
+      "learning_rate": 4.0509259259259263e-07,
+      "loss": 1.0587,
+      "step": 830
+    },
+    {
+      "epoch": 0.9618055555555556,
+      "grad_norm": 0.23660607635974884,
+      "learning_rate": 3.9351851851851854e-07,
+      "loss": 1.0287,
+      "step": 831
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 0.21072295308113098,
+      "learning_rate": 3.819444444444445e-07,
+      "loss": 1.0098,
+      "step": 832
+    },
+    {
+      "epoch": 0.9641203703703703,
+      "grad_norm": 0.2014254480600357,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.9951,
+      "step": 833
+    },
+    {
+      "epoch": 0.9652777777777778,
+      "grad_norm": 0.20475035905838013,
+      "learning_rate": 3.5879629629629633e-07,
+      "loss": 0.953,
+      "step": 834
+    },
+    {
+      "epoch": 0.9664351851851852,
+      "grad_norm": 0.2787926495075226,
+      "learning_rate": 3.472222222222223e-07,
+      "loss": 1.0162,
+      "step": 835
+    },
+    {
+      "epoch": 0.9675925925925926,
+      "grad_norm": 0.21481136977672577,
+      "learning_rate": 3.356481481481482e-07,
+      "loss": 0.9762,
+      "step": 836
+    },
+    {
+      "epoch": 0.96875,
+      "grad_norm": 0.25820887088775635,
+      "learning_rate": 3.2407407407407406e-07,
+      "loss": 1.0382,
+      "step": 837
+    },
+    {
+      "epoch": 0.9699074074074074,
+      "grad_norm": 0.20729532837867737,
+      "learning_rate": 3.125e-07,
+      "loss": 1.041,
+      "step": 838
+    },
+    {
+      "epoch": 0.9710648148148148,
+      "grad_norm": 0.23142774403095245,
+      "learning_rate": 3.0092592592592594e-07,
+      "loss": 1.0654,
+      "step": 839
+    },
+    {
+      "epoch": 0.9722222222222222,
+      "grad_norm": 0.2782406210899353,
+      "learning_rate": 2.893518518518519e-07,
+      "loss": 0.9969,
+      "step": 840
+    },
+    {
+      "epoch": 0.9733796296296297,
+      "grad_norm": 0.2658650279045105,
+      "learning_rate": 2.7777777777777776e-07,
+      "loss": 1.122,
+      "step": 841
+    },
+    {
+      "epoch": 0.9745370370370371,
+      "grad_norm": 0.21585489809513092,
+      "learning_rate": 2.662037037037037e-07,
+      "loss": 0.9342,
+      "step": 842
+    },
+    {
+      "epoch": 0.9756944444444444,
+      "grad_norm": 0.20729321241378784,
+      "learning_rate": 2.5462962962962963e-07,
+      "loss": 0.9905,
+      "step": 843
+    },
+    {
+      "epoch": 0.9768518518518519,
+      "grad_norm": 0.20832544565200806,
+      "learning_rate": 2.430555555555556e-07,
+      "loss": 1.0455,
+      "step": 844
+    },
+    {
+      "epoch": 0.9780092592592593,
+      "grad_norm": 0.20782917737960815,
+      "learning_rate": 2.3148148148148148e-07,
+      "loss": 1.0264,
+      "step": 845
+    },
+    {
+      "epoch": 0.9791666666666666,
+      "grad_norm": 0.2169673889875412,
+      "learning_rate": 2.1990740740740742e-07,
+      "loss": 1.0178,
+      "step": 846
+    },
+    {
+      "epoch": 0.9803240740740741,
+      "grad_norm": 0.20999179780483246,
+      "learning_rate": 2.0833333333333333e-07,
+      "loss": 0.9974,
+      "step": 847
+    },
+    {
+      "epoch": 0.9814814814814815,
+      "grad_norm": 0.21232061088085175,
+      "learning_rate": 1.9675925925925927e-07,
+      "loss": 1.0486,
+      "step": 848
+    },
+    {
+      "epoch": 0.9826388888888888,
+      "grad_norm": 0.21031907200813293,
+      "learning_rate": 1.8518518518518518e-07,
+      "loss": 0.9641,
+      "step": 849
+    },
+    {
+      "epoch": 0.9837962962962963,
+      "grad_norm": 0.22616001963615417,
+      "learning_rate": 1.7361111111111115e-07,
+      "loss": 0.9608,
+      "step": 850
+    },
+    {
+      "epoch": 0.9849537037037037,
+      "grad_norm": 0.21996359527111053,
+      "learning_rate": 1.6203703703703703e-07,
+      "loss": 1.0732,
+      "step": 851
+    },
+    {
+      "epoch": 0.9861111111111112,
+      "grad_norm": 0.24073836207389832,
+      "learning_rate": 1.5046296296296297e-07,
+      "loss": 1.0632,
+      "step": 852
+    },
+    {
+      "epoch": 0.9872685185185185,
+      "grad_norm": 0.21955423057079315,
+      "learning_rate": 1.3888888888888888e-07,
+      "loss": 1.0028,
+      "step": 853
+    },
+    {
+      "epoch": 0.9884259259259259,
+      "grad_norm": 0.211786687374115,
+      "learning_rate": 1.2731481481481482e-07,
+      "loss": 1.0355,
+      "step": 854
+    },
+    {
+      "epoch": 0.9895833333333334,
+      "grad_norm": 0.21731968224048615,
+      "learning_rate": 1.1574074074074074e-07,
+      "loss": 1.041,
+      "step": 855
+    },
+    {
+      "epoch": 0.9907407407407407,
+      "grad_norm": 0.27528902888298035,
+      "learning_rate": 1.0416666666666667e-07,
+      "loss": 0.9799,
+      "step": 856
+    },
+    {
+      "epoch": 0.9918981481481481,
+      "grad_norm": 0.20832359790802002,
+      "learning_rate": 9.259259259259259e-08,
+      "loss": 0.9987,
+      "step": 857
+    },
+    {
+      "epoch": 0.9930555555555556,
+      "grad_norm": 0.22494304180145264,
+      "learning_rate": 8.101851851851852e-08,
+      "loss": 1.02,
+      "step": 858
+    },
+    {
+      "epoch": 0.9942129629629629,
+      "grad_norm": 0.2141268104314804,
+      "learning_rate": 6.944444444444444e-08,
+      "loss": 1.0446,
+      "step": 859
+    },
+    {
+      "epoch": 0.9953703703703703,
+      "grad_norm": 0.21849359571933746,
+      "learning_rate": 5.787037037037037e-08,
+      "loss": 1.1139,
+      "step": 860
+    },
+    {
+      "epoch": 0.9965277777777778,
+      "grad_norm": 0.21971233189105988,
+      "learning_rate": 4.6296296296296295e-08,
+      "loss": 1.021,
+      "step": 861
+    },
+    {
+      "epoch": 0.9976851851851852,
+      "grad_norm": 0.21371498703956604,
+      "learning_rate": 3.472222222222222e-08,
+      "loss": 1.0028,
+      "step": 862
+    },
+    {
+      "epoch": 0.9988425925925926,
+      "grad_norm": 0.21930085122585297,
+      "learning_rate": 2.3148148148148148e-08,
+      "loss": 1.0663,
+      "step": 863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2178175002336502,
+      "learning_rate": 1.1574074074074074e-08,
+      "loss": 0.9601,
+      "step": 864
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 864,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2938073541516984e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}